egulations Handbook of Deep Trade Agreements Handbook of Deep Trade Agreements Handbook of Deep Trade Agreements Handbook of Deep Trade Agreements Edited by Aaditya Mattoo, Nadia Rocha, and Michele Ruta © 2020 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW,Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 23 22 21 20 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Mattoo, Aaditya, Nadia Rocha, and Michele Ruta, eds. 2020. Handbook of Deep Trade Agreements. Washington, DC: World Bank. doi:10.1596/978-1-4648-1539-3. License: Creative Commons Attribution CC BY 3.0 IGO. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation.The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank.Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications,The World Bank Group, 1818 H Street NW,Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1539-3 ISBN (electronic): 978-1-4648-1554-6 DOI: 10.1596/978-1-4648-1539-3 Cover design: Guillermo Varela, PRADO Diseño (Uruguay) Cover images: Background: abstract art painting, © T30 Gallery / Shutterstock.com; top right: connecting dots and lines, © Artistdesign29 / Shutterstock.com; center: Manhattan midtown skyline, © Songquan Deng / Shutterstock.com; bottom: container ship, © MAGNIFIER / Shutterstock.com.All photos used with permission; further permission required for reuse. Interior design: Ramón Pérez, PRADO Diseño (Uruguay) Library of Congress Control Number: 2020938541. Handbook of Deep Trade Agreements HANDBOOK OF DEEP TRADE AGREEMENTS Edited by Aaditya Mattoo, Nadia Rocha, and Michele Ruta CONTENTS CONTRIBUTORS vii FOREWORD ix P. Lamy ACKNOWLEDGMENTS xii ABBREVIATIONS xiii OVERVIEW: THE EVOLUTION OF DEEP TRADE AGREEMENTS 1 A. Mattoo, N. Rocha, and M. Ruta 1. PREFERENTIAL TARIFFS 45 A. Espitia, A. Mattoo, M. Mimouni, X. Pichot, and N. Rocha 2. PREFERENCE UTILIZATION 73 J. Richtering and T. Verbeet 3. EXPORT RESTRICTIONS 87 M. Wu 4. SERVICES 111 B. Gootiiz, G. Jonetzko, J. Magdeleine, J. Marchetti, and A. Mattoo 5. INVESTMENT 143 J. Crawford and B. Kotschwar 6. MOVEMENT OF CAPITAL 177 D. Siegel, K. Gallagher, and R. Thrasher 7. INTELLECTUAL PROPERTY RIGHTS 201 M. Wu 8. VISAS AND ASYLUM 223 J. Pauwelyn, T. Nguyen, and K. Kamal 9. RULES OF ORIGIN 249 M. Angeli, J. Gourdon, I. Gutierrez, and P. Kowalski 10. TRADE FACILITATION AND CUSTOMS 293 E. Kieck CONTENTS 11. ANTI-DUMPING AND COUNTERVAILING DUTIES 319 T. Prusa 12. TECHNICAL BARRIERS TO TRADE 343 A. Espitia, S. Pardo, R. Piermartini, and N. Rocha 13. SANITARY AND PHYTOSANITARY MEASURES 367 S. Stone and F. Casalini 14. PUBLIC PROCUREMENT 391 A. Shingal and V. Ereshchenko 15. SUBSIDIES 427 L. Rubini 16. STATE-OWNED ENTERPRISES 463 L. Rubini and T. Wang 17. COMPETITION POLICY 505 M. Licetti, G. Miralles, and R. Teh 18. ENVIRONMENTAL LAWS 553 J. A. Monteiro and J. P. Trachtman 19. LABOR MARKET REGULATIONS 583 D. Raess and D. Sari APPENDIX. LIST OF FIGURES, TABLES, AND BOXES 607 vi Handbook of Deep Trade Agreements CONTRIBUTORS Marietta Angeli World Trade Institute, Geneva, Switzerland Francesca Casalini Graduate Institute for International and Development Studies, Geneva, Switzerland Jo-Ann Crawford World Trade Organization, Geneva, Switzerland Viktoriya Ereshchenko World Bank, Washington, DC, United States Alvaro Espitia World Bank, Washington, DC, United States Kevin Gallagher Global Development Policy Center, Boston University, Boston, Massachusetts, United States Batshur Gootiiz World Bank, Washington, DC, United States Julien Gourdon Organisation for Economic Co-operation and Development, Paris, France Israel Gutierrez Graduate Institute for International and Development Studies, Geneva, Switzerland Giulia Jonetzko World Bank, Washington, DC, United States Khalid Kamal Georgetown University Law Center, Washington, DC, United States Erich Kieck World Bank, Washington, DC, United States Barbara Kotschwar World Bank, Washington, DC, United States Przemek Kowalski Center for Social and Economic Research, Warsaw, Poland Martha Licetti World Bank, Washington, DC, United States Joscelyn Magdeleine World Trade Organization, Geneva, Switzerland Juan Marchetti World Trade Organization, Geneva, Switzerland Aaditya Mattoo World Bank, Washington, DC, United States Mondher Mimouni International Trade Centre, Geneva, Switzerland Graciela Miralles World Bank, Washington, DC, United States Jose-Antonio Monteiro World Trade Organization, Geneva, Switzerland Trung Nguyen Georgetown University Law Center, Washington, DC, United States Sharon Pardo Batalla, San José, Costa Rica vii CONTRIBUTORS Joost Pauwelyn Graduate Institute of International and Development Studies, Geneva, Switzerland Xavier Pichot International Trade Centre, Geneva, Switzerland Roberta Piermartini World Trade Organization, Geneva, Switzerland Thomas Prusa Rutgers University, New Brunswick, New Jersey, United States Damian Raess World Trade Institute, University of Bern, Bern, Switzerland Jurgen Richtering World Trade Organization, Geneva, Switzerland Nadia Rocha World Bank, Washington, DC, United States Luca Rubini University of Birmingham Law School, Birmingham, United Kingdom Michele Ruta World Bank, Washington, DC, United States Dora Sari University of Geneva, Geneva, Switzerland, and Harvard Law School, Cambridge, Massachusetts, United States Anirudh Shingal European University Institute, Florence, Italy Deborah Siegel Formerly of the International Monetary Fund, Washington, DC, United States Susan Stone Organisation for Economic Co-operation and Development, Paris, France Robert Teh World Trade Organization, Geneva, Switzerland Rachel Thrasher Global Development Policy Center, Boston University, Boston, Massachusetts, United States Joel Trachtman Tufts University, Medford, Massachusetts, United States Thomas Verbeet World Trade Organization, Geneva, Switzerland Tiffany Wang University of Bristol, Bristol, United Kingdom Mark Wu Harvard Law School, Cambridge, Massachusetts, United States viii Handbook of Deep Trade Agreements FOREWORD Preferential trade agreements (PTAs) are a critical part of the international trade architecture, but they remain poorly understood. In part, this is because there is surprisingly limited information on what their features are and how they work in practice. This disconnect became patently clear in the aftermath of the Brexit referendum, which has pulled the United Kingdom out of the European internal market, the most achieved form of PTA. As I have argued, getting Brexit done will be as difficult as getting an egg out of an omelet. This is because preferential trade agreements have changed dramatically and increased in complexity in recent years. This Handbook carefully documents this dramatic shift: the evolution of PTAs into “deep” trade agreements. The wealth of new data that detail the content of PTAs will be essential for researchers and practitioners in trade and beyond. It will help us address difficult questions on the design and effects of deep integration and disintegration and on the future of international economic governance at a moment when these issues are at the front line of the policy debate. Let me put things in perspective. Controversy over the effects of preferential trade agreements, whether they are good or bad, and particularly their relationship with the multilateral trade system, has persisted despite the evolution of PTAs. Since the early days of the post-war multilateral economic order, some trade experts have lamented that these arrangements would create discrimination across countries and fragment the world economy. Others (among which I include myself) have emphasized the complementarity between regionalism and multilateralism, stressing that preferential trade agreements could create a dynamic reform process leading to more, not less, global integration. These arguments revolved around the notion that PTAs were essentially about tariffs.This was true in the 1950s, but it is no longer true today. First, while tariff preferences are negotiated in all trade agreements, they matter much less today than they did 70 years ago. As the WTO noted in the World Trade Report 2011, a major reason is that most-favored-nation (MFN) tariffs (i.e., the tariffs applied to all trade partners in a non-discriminatory way) have declined over time, with more than half of global merchandise trade having applied MFN rates of zero. Second, more than tariffs, preferential trade agreements today are about regulatory measures and other so called non-tariff measures that were once the exclusive domain of domestic policy- making. As shown by recent research at the World Bank, over 50 percent of the close to 300 PTAs in force today cover policy areas such as competition, subsidies, and regulatory standards. ix FOREWORD For these reasons, “deep” trade agreements, as trade experts refer to this new class of agreements, are fundamentally different than the previous generation of PTAs.They aim not only to create market access between members but also to establish broader economic integration rights in goods, services, and factor markets. Deep agreements support these rights by regulating the behavior of importing and exporting governments. They frequently aim to improve efficiency and consumer or social welfare, as in the case of competition or environmental provisions. As noted by the authors in the Overview, ultimately deep trade agreements contribute to setting the rules of the game that define how economies integrate, function, and grow. The new evidence on the evolution of deep trade agreements should change the way we think about the international trade architecture. PTAs continue to play a critical role in creating market access through tariff reductions. In fact, PTAs have reduced trade-weighted average tariffs rates to less than 5 percent for more than two-thirds of countries. But what sets recent trade agreements, particularly post 2000, apart is the large increase in commitments in areas such as services, trade facilitation, investment, and movement of capital. Many of these areas are not covered in the WTO, and for those that are, PTAs often commit countries to deeper, more substantive integration of markets. Over the past two decades, PTAs have also seen an increase in regulatory requirements - the most striking and important of which is the increased emphasis on enforcement of rules and dispute settlement. Deep trade agreements are mostly driven by advanced economies, namely, the EU, the United States, and Japan. PTAs signed between advanced economies and between advanced and developing countries have deepened the most in the past 20 years.With very few exceptions, such as the Pacific Alliance, preferential trade agreements among developing countries have remained closer to the original purpose of trade agreements to grant reciprocal market access in goods. This is true also for China, whose PTAs have remained limited in scope. In this respect, while there are different approaches to deep integration in the EU, the United States, and Japan, a Chinese model has yet to emerge. The evolution in the institutions overseeing deep integration has corresponded to a broader evolution in the nature of international trade.The old world of trade was a world where production systems were national and where obstacles to trade were designed to protect domestic producers from foreign competition. By contrast, the new world is a world where production is transnational along global supply chains of goods and services and where obstacles to trade are aimed at protecting the consumer from risks. We are moving from the administration of protection – quotas, tariffs, and subsidies – to the administration of precaution – security, safety, health, and environmental sustainability. Indeed, much of this administration is what deep trade agreements are about. x Agreements Handbook of Deep Trade FOREWORD This changing content raises a series of new questions for policymakers and for researchers. What are the driving forces behind “deep” trade agreements and what are their effect? How should deep trade agreements be designed to promote welfare of all members and minimize discrimination of non-members? How can we find common ground between different integration approaches and use preferential agreements as laboratories for reform of the multilateral trade system? These are difficult questions, but they are also timely and indispensable questions. The technological innovations that led to the rise of global supply chains are here to stay. So is the increased attention that consumers pose to regulations that protect their health and the environment where they live. The “green wave” is coming fast and strong onto the shores of international trade. And while one would hope that the recent surge in international trade tensions would subside, the challenge of competing economic systems with different rules on subsidies, competition, and state-owned enterprises will remain.What these issues have in common is that they are about deep integration. Whether they are addressed multilaterally, regionally, or bilaterally remains a major question about the future of the governance of globalization. My conviction is that the new data and analysis in this Handbook will help us all, both trade experts and others, understand how trade rules can contribute to better harnessing globalization. Pascal Lamy Former Director General of the World Trade Organization xi ACKNOWLEDGMENTS This Handbook and the associated database are the result of a two-year work program. The editors are thankful to Caroline Freund, Bill Maloney, and Antonio Nucifora for their guidance and to Anabel Gonzalez and Jose Reis for their support in launching this work. Alvaro Espitia provided outstanding research assistance for several chapters of this Handbook and assisted the editors in its production.The chapters of this Handbook have been written in collaboration with experts from academia, the International Trade Centre (ITC), the Organisation for Economic Co-operation and Development (OECD), and the World Trade Organization (WTO). The editors are grateful for this fruitful collaboration. In addition, a number of colleagues at the World Bank and other institutions provided valuable comments on the different chapters, including Rohini Acharya, Chad Bown, Maggie Chen, Roberto Echandi, Michael Ferrantino, Jennifer Hillman, Bernard Hoekman, Nuno Limao, Bill Maloney, Brad McDonald, Mauricio Mesquita Moreira, Theodore Moran, Nora Neufeld, and Kamal Saggi. Deborah Davis edited the Handbook. The graphic concept, design, and layout were carried out by Guillermo Varela and Estudio Prado Diseño. Financial support from the Umbrella Trade Trust Fund is gratefully acknowledged. xii Handbook of Deep Trade Agreements ABBREVIATIONS ACP African, Caribbean, and Pacific countries AD Antidumping AEO Authorized Economic Operator AFTA ASEAN Free Trade Agreement AGOA African Growth and Opportunity Act ANZCERTA Australia–New Zealand Closer Economic Agreement APTA Asia-Pacific Trade Agreement ASEAN Association of Southeast Asian Nations ATIGA ASEAN Trade in Goods Agreement BIT Bilateral Investment Treaty BOP Balance of Payments CACM Central American Common Market CAFTA Central America Free Trade Agreement CAN Andean Community (Comunidad Andina) CARICOM Caribbean Community and Common Market CARIFORUM Caribbean Forum CBT Cross Border Trade CEFTA Central European Free Trade Agreement CEPT Common Effective Preferential Tariff CETA EU-Canada Comprehensive Economic and Trade Agreement CEZ Common Economic Zone CFM Capital Flow Measures CIF Cost, Insurance, and Freight CIS Commonwealth of Independent States CITES Convention on International Trade in Endangered Species COMESA Common Market for Eastern and Southern Africa CPTPP Comprehensive and Progressive Agreement for Trans-Pacific Partnership CSME Caribbean Single Market and Economy CTC Change in Tariff Classification CU Customs Union CVD Countervailing Duties DDA Doha Development Agenda DESTA Design of Trade Agreements DFQF Duty-Free Quota-Free DS Dispute Settlement DSM Dispute Settlement Mechanism EAC East African Community EAEC Eurasian Economic Community EAEU Eurasian Economic Union (superseded EAEC) EAP East Asia and Pacific EC European Commission ECA Economic Cooperation Area ECLAC Economic Commission for Latin America and the Caribbean ECOWAS Economic Community of West African States EEA European Economic Area xiii ABBREVIATIONS EPA Economic Partnership Agreement EPI Environmental Performance Index ERP Environment-Related Provision EFTA European Free Trade Association EU European Union FAO Food and Agriculture Organization FDI Foreign Direct Investment FET Fair and Equitable Treatment FOB Free on Board FTA Free Trade Agreement FTZ Free Trade Zone GATS General Agreement on Trade in Services (succeeded GATT) GATT General Agreement on Tariffs and Trade (1948–95) GI Geographical Indication GNI Gross National Income GPA Agreement on Government Procurement GSP Generalized System of Preferences HS Harmonized System (trade codes) I-TIP Integrated Trade Intelligence Portal IIA International Investment Agreement IMF International Monetary Fund IP Intellectual Property ISDS Investor-State Dispute Settlement ITC International Trade Centre IUU Illegal, Unreported, Unregulated (fishing) KORUS US-Korea Free Trade Agreement LABPTA Labor Agreements in Trade Agreements LAC Latin America and the Caribbean LAIA Latin American Integration Association LDCs Least-Developed Countries MARPOL International Convention for the Prevention of Pollution from Ships MCPAT Market Competition Policy Assessment Tool (World Bank Group) MEA Multilateral Environmental Agreement MENA Middle East and North Africa Mercosur Southern Common Market (South America) MFN Most Favored Nation MNP Movement of Natural Persons MRA Mutual Recognition Agreement MSG Melanesian Spearhead Group NAFTA North American Free Trade Agreement NBER National Bureau of Economic Research NT National Treatment NTMs Nontariff Measures OAS Organization of American States OCTs Overseas Countries and Territories OECD Organisation for Economic Co-operation and Development PICTA Pacific Island Countries Trade Agreement xiv ABBREVIATIONS Handbook of Deep Trade Agreements PPA Preferential Procurement Agreement PR Performance Requirement PSA Partial Scope Agreement PSR Product-Specific Rules PTA Preferential Trade Agreement RF Random Forest RGPA Revised Agreement on Government Procurement ROI Reasonable, Objective, and Impartial RoOs Rules of Origin RTA Regional Trade Agreement RVC Regional Value Content RWHI Regime-Weighted Harris Index SACU Southern African Customs Union SADC Southern Africa Development Community SAFTA South Asian Free Trade Agreement SBO Substantive Business Operations SDT Special and Differential Treatment SE State Enterprise SITA Sustainable Innovations in Trade Agreements SMEs Small and Medium Enterprises SOE State-Owned Enterprise SPARTECA South Pacific Regional Trade and Economic Cooperation in Asia-Pacific SPS Sanitary and Phytosanitary STC Specific Trade Concern STE State Trading Enterprise STRI Services Trade Restriction Index SWF Sovereign Wealth Fund TBT Technical Barrier to Trade TFA Trade Facilitation Agreement TFEU Treaty on the Functioning of the European Union TISA Trade in Services Agreement TTIP Transatlantic Trade and Investment Partnership TPP Trans-Pacific Partnership TPSEP Trans-Pacific Strategic Economic Partnership TREND Trade and Environment Database TRIMS Trade-Related Investment Measures TRIPS Trade-Related Intellectual Property Rights UN United Nations UNCED United Nations Conference on Environment and Development UNCTAD United Nations Conference on Trade and Development UNEP United Nations Environment Programme VC Value Content WAEMU West African Economic and Monetary Union WBG World Bank Group WCO World Customs Organization WO Wholly Obtained WTO World Trade Organization xv The Evolution of Deep Trade Agreements OVERVIEW The Evolution of Deep Trade Agreements A. Mattoo, N. Rocha, and M. Ruta 1 Handbook of Deep Trade Agreements OVERVIEW The Evolution of Deep Trade Agreements A. Mattoo*, N. Rocha* and M. Ruta* * World Bank, Washington, DC, United States CONTENTS O.1. INTRODUCTION 3 O.2. SCOPE AND METHODOLOGY 6 O.3. STYLIZED FACTS 12 O.4. THE CHALLENGE OF QUANTIFYING THE EFFECTS OF DTAs 19 O.4.1 Directly constructed indices 19 O.4.2 Indirectly estimated measures 20 O.4.3 Quantification challenges: some analytical issues going forward 22 O.5. CONCLUSIONS 23 ACKNOWLEDGMENTS 24 REFERENCES 25 ANNEX 26 2 The Evolution of Deep Trade Agreements O.1. INTRODUCTION This Handbook provides new data on and analysis of all preferential trade agreements (PTAs) that have been notified to the World Trade Organization (WTO), and highlights the emergence of deep trade agreement (DTA).1 DTAs are reciprocal agreements between countries that cover not just trade but additional policy areas, such as international flows of investment and labor, and the protection of intellectual property rights and the environment, amongst others. While these legal arrangements are still referred to as trade agreements, their goal is integration beyond trade or deep integration. DTAs aim at establishing five “economic integration” rights: free (or freer) movement of goods, services, capital, people, and ideas. DTAs also include enforcement provisions that limit the discretion of importing governments in these areas, as well as provisions that regulate the behavior of exporters. Preferential trade agreements have always been a feature of the world trading system but have become more prominent in recent years.The number of PTAs increased from 50 in the early 1990s to roughly 300 in 2019. All WTO members are currently party to one, and often several, PTAs. While WTO rules still form the basis of most trade agreements, PTAs have in some sense run away with the trade agenda. Traditional trade policy areas, such as tariff reduction or services liberalization, are now more frequently negotiated in regional contexts rather than at the WTO, with PTAs often going beyond what countries have committed to at the WTO. The result is that PTAs have expanded their scope. While the average PTA in the 1950s covered 8 policy areas, in recent years they have averaged 17. In other words, there is some preliminary evidence that PTAs are becoming DTAs, both on the intensive margin (specific commitments within a policy area) and the extensive margin (number of policy areas covered). In this Handbook, we do not draw a sharp distinction between DTAs and other PTAs. Rather, the aim is to demonstrate the progressive deepening of PTAs. Deep trade agreements matter for economic development. The rules embedded in DTAs, along with the multilateral trade rules and other elements of international economics law such as international investment agreements, influence how countries (and, hence, the people and firms that live and operate within them) transact, invest, work, and, ultimately, develop. 1 In the international economics and law literature, “PTA” is an umbrella term encompassing several types of reciprocal agreements between trading partners: regional trade agreements (RTAs), free trade agreements (FTAs), and customs unions (CUs). This definition differs from that of the World Trade Organization (WTO), which defines PTAs as agreements that grant unilateral (i.e., non-reciprocal) trade preferences such as the Generalized System of Preferences schemes, under which developed countries grant preferential tariffs to imports from developing countries. This study, following the definition from international economics and law, uses the term “PTA” to refer to all types of trade agreements, both within and across regions, and uses “DTA” to refer to PTAs that contain provisions aimed at deepening economic integration between trading partners. 3 Handbook of Deep Trade Agreements Trade and investment regimes determine the extent of economic integration, competition rules affect economic efficiency, intellectual property rights matter for innovation, and environmental and labor rules contribute to social and environmental outcomes. It is, therefore, vital that rules and commitments in DTAs be informed by evidence and shaped more by development priorities than by international power dynamics or domestic politics. An impediment to this goal is that data and analysis on trade agreements have not captured the new dimensions of integration, which makes it difficult to identify the content and consequences of DTAs. This Handbook takes a first step towards filling this important gap in our understanding of international economic law and policy. It presents detailed data on the content of the eighteen policy areas most frequently covered in PTAs, focusing on the stated objectives, substantive commitments, and other aspects such as transparency, procedures, and enforcement. In terms of the coverage of policy areas and the granularity of information within each area, this is the most comprehensive effort to date. Each chapter, authored by a leading expert in his or her field, explains in detail the methodology used to collect the information and provides a first look at the evidence in each policy area. The new data and analysis will inform experts and policymakers in their efforts to design, negotiate, and take advantage of DTAs that promote development. This information will also enable researchers to develop indicators on the depth of trade agreements in different policy areas, assess the similarities between these arrangements, and benchmark countries’ DTAs relative to their partners. It will also help identify the rules that benefit only participants and those that have large spillover effects on non-participants or excluded countries. Finally, the new data and analysis in this study will allow researchers to identify areas where there is de facto convergence across different players, thus facilitating the adoption of commonly agreed multilateral rules. This Handbook will lay the groundwork for new research in international economics and other fields. A large body of economic literature has looked at the effects of PTAs on international trade flows and on welfare.2 However, this literature has two important limitations.3 On the theoretical side, the study of PTAs is mostly based on Vinerian4 logic, which focuses exclusively on tariffs, thus by construction excluding deep integration issues. On the empirical side, attempts to quantify the effects of PTAs suffer from a measurement error problem, as studies generally rely on dummies to identify trade agreements or distinguish between broad categories of agreements such as FTAs or CUs. The new data will help theorists to model DTAs and help empirical economists to properly identify their effects. 2 See, e.g., Freund and Ornelas 2010, Limao 2017. 3 See, e.g., Baldwin 2010, Mattoo et al. 2017. 4 Viner 1950. 4 The Evolution of Deep Trade Agreements Beyond economics, the data will inform research in other fields, primarily international law and international relations, on important issues such as the commonality or divergence of the rules set in PTAs and how they could evolve in the future. The research in this Handbook is the result of collaboration among the World Bank, the International Trade Centre (ITC), the Organisation for Economic Co-operation and Development (OECD), the WTO, and experts from academic institutions. It builds on previous research by the World Bank and others.A first database on the content of deep trade agreements was published in 2017 with the goal of documenting how the policy areas covered by PTAs had increased over time (Hofmann, Osnago, and Ruta 2017). This dataset allowed researchers to construct a first series of indicators which capture the scope of trade agreements; i.e., what policy areas they cover.We refer to this as the extensive margin of PTA depth. Based on this first dataset, several research papers then looked, respectively, at the impact of deep trade agreements on trade, global value chains, and foreign direct investment (FDI), and the effect of breaking up such agreements.5 The data have also been extensively employed for policy advice by the World Bank in several developing countries in Africa, Latin America, East Asia, and the Balkans. The new data that we briefly review in this introduction and that are analyzed in detail in the individual chapters of this Handbook offer insights into a different dimension of PTAs’ depth. They capture the detailed commitments to establish and preserve the rights to economic integration, and the procedures, institutions, and enforcement mechanisms that countries set up to make deep integration work. The focus is therefore not on the extensive margin of integration (number of policy areas that are covered by the agreement), but on its intensive margin (the specific commitments within a policy area). While there are a number of individual studies that have documented the deepening of PTAs in specific areas, two major data collection projects—Dür, Baccini, and Elsig (2014) and Acharya (2016)—also aimed at documenting the specific commitments for a group of policy areas covered in PTAs. Both efforts have important merits. Dür, Baccini, and Elsig (2014) covered a large set of PTAs, including those that have been notified to the WTO but are no longer in force. Acharya (2016) provided a series of databases on the content of PTAs that go beyond specific policy areas and cover emerging issues such as e-commerce or the rules on dispute settlement in PTAs. Relative to these data collection projects, the new dataset is more comprehensive, both in terms of the number of policy areas covered and in terms of the information on detailed disciplines in each area. This introduction describes the scope and methodology underlying the research agenda on deep trade agreements. It also highlights a novel set of stylized facts that can be inferred from a first look at the new data collected as part of this project, and offers some insights into future applications and areas for analysis. 5 Mattoo et al. 2017, Mulabdic et al. 2017, Laget et al. 2018, Laget et al. 2019. 5 Handbook of Deep Trade Agreements O.2. SCOPE AND METHODOLOGY The number of policy areas covered by PTAs has increased in the last two decades. Up until the late 1990s, when the number of PTAs started increasing, the majority of new agreements covered fewer than 10 policy areas. Since the 2000s, most new PTAs have covered between 10 and 20 policy areas, with some having even more than 20 (Figure O.1). In a study of 28 trade agreements signed by the US and the EU, Horn et al. (2010) identify up to 52 policy areas that have been covered by at least one of the agreements. The inclusion of new policy areas in PTAs is not random. As shown in Mattoo et al. (2017), trade agreements covering few policy areas generally focus on traditional trade policy, such as tariff liberalization or customs (Table O.1). Agreements with broader coverage (between 10 and 20 policy areas) tend to include trade-related regulatory issues, such as subsidies or technical barriers to trade. Finally, agreements with more than 20 provisions often include policy areas that are not directly related to trade, such as labor, environment, and movement of people. Figure O.1: Number of policy areas covered in PTAs, 1970-2017 Source: WTO, Preferential Trade Agreements database, following Hofmann, Osnago, and Ruta 2017. The policy areas studied in this Handbook are those that appear most frequently in trade agreements.They include (a) a set of 18 policy areas that are covered in 20 percent or more of trade agreements notified to the WTO (Figure O.2): (b) tariffs on industrial and agricultural goods, which are covered by all trade agreements; (c) customs and export taxes, which are regulated in more than 80 percent of PTAs; (d) services and movements of capital, which are regulated in roughly half of the PTAs; and (e) environmental and labor issues, which are covered by around 20 percent of all trade agreements. Other issues that are sometimes (although infrequently) regulated in trade agreements, such as education, nuclear safety, and 6 The Evolution of Deep Trade Agreements Table O.1: Share of policy areas for different PTAs Source: Mattoo et al. 2017. 7 Handbook of Deep Trade Agreements human rights, are not included in this Handbook and could be the subject of future research. The focus on individual areas helps us to identify specific policies that are the object of negotiation but may obscure cross-cutting issues—such as electronic commerce—that may be disciplined under multiple policy areas. Figure O.2: Number of policy areas covered in PTAs, by policy Source: Deep Trade Agreements Database, based on Hofmann, Osnago, and Ruta 2017. The classification of policy areas used in Figure O.2 deviates slightly from the one of Horn et al. (2010).6 Specifically, for this Handbook, we decided to include rules of origin, a policy area that was absent from the Horn et al. (2010) classification, and to treat as a single policy area: (a) trade remedies, which include anti-dumping and countervailing measures; (b) investment, which includes the areas covered under the WTO’s Trade-Related Investment Measures, or TRIMs; and (c) intellectual property rights (IPR), which include the areas covered under the WTO’s Trade-Related Intellectual Property Rights, or TRIPs. Trade agreements are generally assessed in terms of the market access they create. Given the complexity of policy areas that are covered by DTAs, the metric of market access—while still important—appears inadequate. In this introduction, we propose to define deep trade agreements as international arrangements that aim to regulate three (partially overlapping) sets of policy areas (Figure O.3). • First, the core policy areas included in DTAs aim to establish five economic integration rights: free (or freer) movement of goods, services, capital, people, and ideas.7 6 The Horn et al. 2010 classification was used to collect data on the extensive margin of PTA depth. 7 We use the words “aim to establish” rather than “establish” for two main reasons. First, DTAs may cover only a subset of integration rights. Second, provisions may not be justiciable. A contribution of the new data is to identify the extent to which integration rights are established in PTAs. 8 The Evolution of Deep Trade Agreements The policy areas that directly impact these flows include: (a) tariffs and export taxes, which affect the movement of goods; (b) services, which regulate services trade flows; (c) investment and movement of capital, which affect the movement of capital; (d) visa and asylum, which regulate the movement of people; and (e) intellectual property rights, which influence the flows of ideas. • Second, DTAs also include policy areas that aim to support these economic integration rights by limiting government discretion. Actions by importing governments that limit international flows can be taken at the border and behind the border and are often of a regulatory nature. The policy areas that fall in this category are: (a) customs; (b) rules of origin; (c) trade remedies; (d) public procurement; (e) technical barriers to trade (TBT); (f) sanitary and phytosanitary measures (SPS); (g) state-owned enterprises (SOEs); (h) subsidies; and (i) competition policy.8 • Third, DTAs cover policy areas that aim to enhance social or consumer welfare by regulating the behavior of exporters. Policy areas such as environment and labor impose obligations on exporters to further consumer or social interests in importing countries. Rules in areas such as competition, SOEs, and subsidies can have a dual aspect: in addition to regulating action that undermines economic integration rights, they can aim to address distortionary actions that lower economic efficiency, thus hurting consumer or social welfare. For each policy area, the experts followed a uniform approach to coding.9 The coding templates encompass several common headings such as objectives and definitions, institutional framework, and an enforcement mechanism, plus a series of discipline-specific questions. Under each heading, questions on specific provisions in the agreement are formulated so that they can be answered with Yes/No. For some policy areas, additional information is provided at the provision level, including (a) the relationship between the coverage of the disciplines on and the corresponding regulation in the WTO; (b) the level of enforceability of each provision;10 (c) whether the specific 8 Some of these provisions apply only to cross-border trade in goods (e.g., customs, TBT, and SPS). Others can also apply to cross-border trade in services (e.g., public procurement and competition policy). In some cases, services-related provisions are included separately in a services agreement. 9 One exception is preferential tariffs. Differently from the other policy areas, tariff commitments apply at the product level. The information for this area is therefore collected at the country-pair-product level. For rules of origin a sub-sample of agreements in Latin America and East Asia, the dataset on regime-wide provisions is accompanied by a mapping of the rules of origin that apply at the product level. 10 The legal enforceability of the PTA provisions is coded according to the language used in the text of the agreements. It is assumed that commitments expressed with a clear, specific, and imperative legal language can more successfully be invoked by a complainant in a dispute settlement proceeding, and therefore are more likely to be legally enforceable. In contrast, unclearly formulated legal language might be related to policy areas that are covered but that might not be legally enforceable. 9 Handbook of Deep Trade Agreements Figure O.3: A classification of policy areas in DTAs commitment can be applied discriminatorily or whether it is de facto non-discriminatory. Finally, when applicable, for example, in services and government procurement, the coders included information at the sectoral level on exclusion of certain sectors from an agreement, or the applicability of an agreement to a specific industry. The analysis covers the realm of PTAs that are in force and notified to the WTO as of end-2017. The basis of the coding analysis is the legal text of the trade agreements and the relevant annexes that accompany the agreement (and have been notified to the WTO). This approach comes with two main limitations that should be clear to the user of the database. First, the focus on the legal text of the agreement implies that secondary law (the body of law that derives from the principles and objectives of the treaties) has not been coded. This is a concern particularly when assessing the depth of integration of the EU, since in most policy areas covered in this Handbook, EU institutions have used secondary law such as regulations, directives, and other legal instruments to pursue integration.11 Second, the focus on the legal text also excludes from consideration issues of implementation of the trade agreement into national laws and regulations or subsequent annexes that the parties might agree on which are not reported to the WTO. These are important areas for future research. Despite the similarity in the coding approach, policy areas differ widely from each other. First, some policy areas are inherently more complex than others and their description requires 11 Note that the figures and tables in this introduction refer to the EU as a single entity (i.e., the European Union agreement and enlargements are excluded) and report data for EU PTAs with third countries where this concern does not apply. 10 The Evolution of Deep Trade Agreements a larger number of questions to reflect the more detailed provisions. IPR has the highest number of provisions (120), while labor has the lowest (18). Second, some policy areas focus primarily on substantive provisions: specific commitments on integration, such as market access commitments, and specific obligations such as harmonization of standards. Others tend to have a larger number of procedural provisions, such as transparency provisions and procedural requirements. Table O.2 provides an overview, showing the heterogeneity across policy areas in these different dimensions and identifying the set of “substantive” provisions as those that require specific integration/liberalization commitments and obligations. Table O.2: Number of substantive and other provisions per policy area in all PTAs notified to the WTO Trade Facilitation and Customs Intellectual Property Rights Sanitary and Phytosanitary Labor Market Regulations Technical Barriers to Trade Discipline State-Owned Enterprises Procurement Subsidies Countervailing Duties Movement of Capital Environmental Laws Competition Policy Anti-dumping and Visa and Asylum Total provisions Measures Public Rules of Origin Export Taxes Investment Services Category Objectives 1 1 8 1 2 2 2 2 2 2 23 Scope and definitions 1 16 11 7 2 17 2 4 2 1 10 8 10 91 Transparency 4 9 3 13 3 8 6 3 10 7 4 5 1 76 Substantive commitments 17 19 13 59 3 6 2 19 20 4 3 8 11 27 20 12 243 Liberalization/ 14 8 11 19 4 1 3 4 3 13 80 Integration Conditions/ 3 11 2 40 3 2 1 16 16 1 3 8 11 27 7 12 163 Obligations Procedural requirements 17 8 12 3 28 10 3 28 2 2 17 130 Enforcement mechanism 1 3 8 22 1 2 4 5 7 5 4 1 63 Sectoral coverage 2 1 2 5 33 9 8 60 Specific coverage 2 1 13 9 8 2 8 1 44 Exceptions 5 6 2 35 4 4 3 1 3 63 Safeguards 1 31 10 1 43 Special and 7 2 2 11 differentiated treatment Institutional framework 1 1 2 2 2 6 2 11 2 1 2 32 Cooperation 2 3 1 8 3 3 1 1 1 5 4 1 33 Miscellaneous 9 6 5 2 2 1 25 Total provisions 46 64 57 95 120 30 52 51 34 59 100 36 54 35 48 38 18 937 Source: Deep Trade Agreements Database. 11 Handbook of Deep Trade Agreements We also make an effort to identify the set of provisions within each policy area that are essential to achieve the objectives of the agreement. The provisions we refer to as “essential” comprise the set of substantive provisions plus the disciplines among procedures, transparency, enforcement, or objectives which are viewed as indispensable and complementary to achieving the substantive commitments. Non-essential provisions are referred to as “corollary.” A caveat is that this exercise is based on the experts’ knowledge and, hence, is subjective. However, this approach has the advantage of limiting the dimensionality of the data in an informed way.12 O.3. STYLIZED FACTS A number of new stylized facts emerge from a preliminary analysis of the data. Each of the chapters in this Handbook provides a first look at the data by policy area. In this introduction, we present a bird’s-eye view of the entire dataset put together by experts. Given the differences among policy areas and among provisions within each policy area, this approach presents many quantification challenges, which are discussed below. In this section, we rely on simple counts of the provisions and on coverage ratios13 to investigate the evolution of the content of deep trade agreements. The underlying assumption in this approach is that deeper trade agreements imply a larger number of provisions. As shown in chapter 1,14 liberalization in PTAs has reduced trade-weighted average tariff rates to less than 5 percent for more than two-thirds of countries (Figure O.4).While there are still Figure O.4: Tariffs in PTAs and MFN tariffs Source: Chapter 1 by Espitia et al. 2020. 12 A statistical approach on how to assess the importance of specific provisions included in the different policy areas in explaining trade outcomes is presented in section O.4. 13 The coverage ratio is defined as the share of provisions for a policy area contained in a given agreement relative to the maximum number of provisions in that policy area or agreement. 14 Chapter 1 by Espitia et al. 2020. 12 The Evolution of Deep Trade Agreements pockets of high protection in some countries, most notably lower-income economies, PTAs have been broadly successful in committing national governments to maintaining low tariffs. Trade-weighted applied tariffs are, on average, 2.3 percentage points lower than average most-favored-nation (MFN) rates, with gaps of greater than 6 percentage points for countries like Tunisia, Morocco, Bosnia and Herzegovina, Namibia, and the Lao People’s Democratic Republic. So, while from an efficiency perspective, preferential tariff liberalization is inferior to non-preferential liberalization, the commitments countries have made in the network of preferential trade agreements may provide a safety net at a time when trade tensions are escalating and some countries are disregarding their multilateral commitments. The number of commitments that governments have made in trade agreements, particularly since the early 2000s, has increased over time. Figure O.5 shows how the coverage ratio has changed over time for the 17 policy areas analyzed in this Handbook (all but tariffs) in aggregate. With only a few exceptions, the majority of new PTAs signed after 2000 have a coverage ratio higher than 25 percent. This stands in sharp contrast to the trade agreements signed in the 1980s and 1990s, when coverage ratios were below 15 percent and, in many cases, even below 5 percent. The reduction in tariffs accomplished through preferential trade liberalization, together with the increased depth of agreements over time, suggests that countries that are willing to cut tariffs reciprocally may also be willing to accept deeper mutual commitments in other areas. Figure O.5: Number of agreements over time vs. average coverage ratio Note: Coverage ratio refers to the share of provisions contained in a given agreement relative to the maximum number of provisions. European Union agreement and enlargements excluded. While the overall number of provisions is suggestive, it can hide important elements of the evolution of deep trade agreements. First, as discussed above, some provisions imply substantive commitments while others concern broad objectives, definitions, or procedural matters. Second, deep trade agreements, as defined in this Handbook, do not only concern themselves with market access in goods, but also aim to establish freedom of mobility for 13 Handbook of Deep Trade Agreements services, capital, ideas, and people, as well as regulating policy areas that have an impact on consumer and/or social welfare, such as labor and the environment. To gain a better understanding of how the commitments in PTAs have changed over time, we look at the evolution of coverage ratios by policy area. Figure O.6 shows that the coverage of essential disciplines in PTAs has increased over time across all policy areas. This is most clearly the case for the policy areas aimed at facilitating the flows of goods (customs and trade facilitation), capital (investment and movement of capital), and services. IPR and movement of people (visa and asylum) also saw a steady but less remarkable increase in essential commitments over time. Along with economic integration rights, PTAs increasingly include essential commitments in policy areas that support these rights or impose obligations on exporters. The ones that appear to stand out are subsidies, competition, and SOEs, areas that are either excluded from the WTO or for which reform of multilateral rules is considered difficult. Interestingly, while essential commitments in labor have largely increased in recent years, this happened to a lesser extent for provisions on the environment. Figure O.6: Coverage ratios by policy area, over time Note: Coverage ratio by policy area refers to the share of provisions for a policy area contained in a given agreement relative to the maximum number of provisions in that policy area.Years refer to entry-into-force date. European Union agreement and enlargements excluded. The presumption is that the increase in the essential disciplines in deep PTAs has been driven by countries taking on more substantive commitments over time. Indeed, Figure O.7 shows 14 The Evolution of Deep Trade Agreements that this is the case, but it also uncovers interesting insights about the evolution of non- substantive commitments.We focus on the three (numerically) most relevant non-substantive provisions: procedural rules, transparency, and enforcement provisions. The deepening of substantive commitments has been accompanied by an increase in the number of corollary provisions, suggesting that achieving deeper commitments may require more procedural rules for implementation, transparency, and enforcement. A second insight is that, while these disciplines are all necessary to render substantive commitments in trade agreements effective, they have evolved differently in recent years. Starting in the early 2000s, the relevance of enforcement provisions in DTAs has increased disproportionally relative to procedural and transparency provisions. The growing enforcement capacity of DTAs may help explain the success of these institutional arrangements as tools for deep integration. Figure O.7: Substantive provisions and a breakdown of non-substantive provisions in PTAs, over time Note: Coverage ratio refers to the share of provisions for a policy area contained in a given agreement relative to the maximum number of provisions in that policy area.Years refer to entry-into-force date. European Union agreement and enlargements excluded. When we break down the trade agreements by level of development of the signatories, we observe two facts. First, the deepest PTAs are those involving developed economies, followed by PTAs between developed and developing economies. PTAs between developing countries are the shallowest. Indeed, there is a sizeable gap between average coverage ratios for the latter group of PTAs relative to the first two (Figure O.8). This could reflect a focus of negotiations on tariffs and traditional trade barriers, which are still high for several low-income economies. Second, in terms of composition, PTAs between developed countries and those between developed and developing economies include similar shares of provisions establishing economic integration rights, supporting these rights, and aiming to regulate exporters (Figure O.9). PTAs between developing countries are shallower across the board, with a stronger gap in areas such as environment and labor that aim at improving social welfare. 15 Handbook of Deep Trade Agreements Figure O.8: Inclusion of substantive commitments in PTAs, by level of development Note: Coverage ratio refers to the share of provisions for a policy area contained in a given agreement relative to the maximum number of provisions in that policy area.Years refer to entry-into-force date. European Union agreement and enlargements excluded. Figure O.9: Coverage ratio by type of policy and level of development Note: Coverage ratio refers to the share of provisions for a policy area contained in a given agreement relative to the maximum number of provisions in that policy area.Years refer to entry-into-force date. European Union agreement and enlargements excluded. We next analyze the depth of trade agreements by country. Here, we focus on the substantive commitments.15 As several countries have multiple agreements with different levels of depth, we present the average number of substantive commitments per country in panel a of Figure O.10 and the maximum number in panel b of Figure O.10. The main takeaway is that developing countries in Sub-Saharan Africa, the Middle East and North Africa, South America, South Asia, and, to a lesser extent, East Asia tend to have fewer substantive commitments in trade agreements relative to advanced economies. The few 15 Annex Tables O.A.1 and O.A.2 provide other indicators by PTA and by country. 16 The Evolution of Deep Trade Agreements exceptions include countries in South America that are signatories to the Pacific Alliance and other developing economies that have signed deep trade agreements with an advanced trade partner, such as Mongolia with Japan and Caribbean countries with the EU. In terms of depth as measured here, North America and Europe are the most integrated regions, through NAFTA and its successor agreement, and through the agreements the EU has signed with neighboring countries. East Asia is a region with a mixed profile: the network of ASEAN agreements includes most countries but tends to have fewer substantive commitments relative to North America and Europe, except for some countries such as Vietnam, which have signed on to the Comprehensive Agreement for the Trans-Pacific Partnership (with a coverage ratio of 61 percent). Figure O.10: Substantive provisions in PTAs by country a. Average number of provisions b. Maximum number of provisions 17 Handbook of Deep Trade Agreements With the increasing depth and complexity of trade agreements, both similarities and dissimilarities between PTAs could potentially increase. Older agreements that covered only preferential tariff liberalization and other aspects of market access tended to be very similar. As PTAs now cover more ground, there can be provisions that are included in two agreements, making them more similar, or there can be provisions that are covered by one PTA but not by another, making them more dissimilar. To capture this information, we construct a similarity index for DTAs, calculated as the ratio between the number of provisions for which two agreements have a “yes” (a measure of similarity) and the total number of provisions covered by the agreements, independently of whether they have the same answer or not. The closer the similarity index is to one (or zero), the more (or less) similar are the two DTAs; i.e., they include the same type(s) of provisions. Figure O.11 plots the degree of similarity for the PTAs signed by the three major trading blocs: the European Union, the United States, and Japan. Each color represents a DTA signed by a third country with the US (green bubbles), EU (purple bubbles), or Japan (pink bubbles). The size of the bubbles represents the depth of the agreements, measured as the number of provisions covered. Each agreement is connected to the one which is most similar within a trading bloc. The figure also links the three trading blocs, by connecting the pair of agreements that are the most similar between two blocs. As expected, within each bloc, DTAs are highly similar: up to 0.89 for the US (US-Peru; US-Colombia), up to 0.80 for the EU (EU-Moldova; EU-Ukraine), and up to 0.75 for Japan (Japan-Indonesia; Japan-Mongolia). This fact often reflects a “template effect,” where the EU, the US, and Japan tend to negotiate based on a template offered to third countries. Interestingly, the similarity of DTAs is relatively high even across blocs, although lower than within blocs. For example, the EU-Republic of Korea agreement shares more than 50 percent of the provisions with the Japan-Switzerland agreement (similarity index of 0.54) and with the US-Peru agreement (similarity index of 0.51). These results indicate that concerns about the fragmentation of the global trade system have some foundation (i.e., they do not share almost half of provisions), but also point to substantial similarities—based on which multilateral rules can be agreed upon. 18 The Evolution of Deep Trade Agreements Figure O.11: Similarity of agreements Note: The size of the bubbles represents the depth of a trade agreement, as captured by the number of provisions included in the agreement. Each edge connects an agreement with one that is most similar. Purple bubbles represent EU agreements with non-EU countries, green bubbles represent US agreements, and pink bubbles represent Japan’s agreements. O.4. THE CHALLENGE OF QUANTIFYING THE EFFECTS OF DTAs Quantification of the effects of DTAs poses a serious challenge. DTAs cover heterogeneous areas: tariffs, contingent protection, export taxes, customs procedures, technical barriers in goods; a wide range of restrictions across modes in services; investment measures, subsidies, procurement, state enterprises, competition policy affecting both trade and investment in goods and services, visas and asylum, and a range of regulatory requirements affecting labor mobility; and a variety of policies affecting the protection of intellectual policy rights and the environment. How can the diversity of policies be quantified and aggregated within separate areas? How can we aggregate across the different areas? We briefly discuss here two approaches to quantification—directly constructed indices and indirectly estimated measures—and some analytical issues going forward. O.4.1 Directly constructed indices The count variables and coverage ratios presented in the previous section are the simplest directly constructed indices of depth. They provide an immediate view of how commitments 19 Handbook of Deep Trade Agreements in PTAs have changed over time, across countries, and for subsets of provisions. Still, aggregate indicators based on some form of counting disregard the fact that DTAs cover multiple policy areas and sectors and that the “value” of each provision is unlikely to be the same even within the same policy area. In some cases, it may be possible to construct a hierarchy of measures. For example, in the areas of services and government procurement, provisions could be divided into three tiers. Tier 1 would comprise provisions ensuring market access and national treatment at entry. Tier 2 would comprise provisions on post-entry operation; e.g., preferences or offsets. Tier 3 would comprise procedural rules limiting discretion in licenses and awards. The construction of an index could then be lexicographic, in that we would consider first only differences between countries or sectors in Tier 1 and move to subsequent tiers only to break ties. Such an approach is ideally suited to the construction of an ordinal rather than cardinal (i.e., qualitative rather than quantitative) measure. A pragmatic approach to overcoming some of the constraints to constructing representative indices is to rely on experts’ judgment. This is the method adopted in this Handbook. The individual chapters will offer a disaggregated set of stylized facts for each policy area using count variables, coverage ratios, and the individual assessment of the authors of the key provisions in each policy area. We have already discussed the distinction between substantive and essential provisions. Some chapters go even beyond these categories. For instance, the chapter on SOEs (Rubini and Wang 2020) identifies four commitments concerning issues of ownership, discrimination, subsidization, and anti-competitive behavior as key. The chapter on technical barriers to trade (Espitia et al. 2020) identifies a subset of seven commitments which are key to achieving deep integration in the area of technical regulations.This type of information can be used in the estimation exercises we discuss below, as it allows the researcher to address problems associated with large numbers of possible variables at hand, such as multicollinearity (i.e., the high correlation between the different provisions within and across policy areas). O.4.2 Indirectly estimated measures These measures are obtained by estimating the impact of the provisions on a variable of interest. For example, we could infer the value of individual provisions by estimating their impact on bilateral trade, controlling for other influences. In principle, each binary element in the relevant DTA areas could be included in a country-product import regression as a right-hand variable while controlling for applied policies, including tariffs and non-tariff measures. Similar methods have been used to estimate the Overall Trade Restrictiveness Index.16 However, even for trade in goods we have limited degrees of freedom, and in other 16 Kee, Nicita, and Olarreaga 2009. 20 The Evolution of Deep Trade Agreements areas (such as services), we do not have sufficiently fine outcome data. In these areas, it may be necessary to take a hybrid approach, based on first constructing more aggregated indices. Another approach is to quantify the effects of DTAs and build indicators of depth using new statistical methods. As a first example, we employ machine learning techniques to detect the influential variables/provisions in DTAs for trade.17 “Machine learning” is a generic term referring to a wide variety of algorithms which detect a certain pattern from a large dataset, often referred to as “Big Data,” and make predictions based on that pattern. In this case, we use a method called Random Forest (RF) to calculate the importance of each variable/ provision for international trade flows.18 Specifically, we run as a first step a structural gravity model with the standard set of fixed effects and then use the residuals as the left-hand variable in the RF. Figure O.12 shows the boxplot of scores calculated by the RF of variables/provisions in PTAs belonging to the 17 (non-tariff) policy areas analyzed in this Handbook.19 The areas are colored according to their categorization into the three main groups illustrated in Figure O.3; dark green indicates policies that establish economic integration rights, purple is assigned to those supporting these rights, and light green to those that promote welfare. Each box shows the range of the first (25 percent) and third (75 percent) quartiles, and the black line in the box shows the median of the scores.The vertical lines extending from the box indicate the variability outside the above quartiles, and the dots outside of the line are regarded as outliers. Boxplots are ordered according to the magnitude of the median. Focusing on the entire set of PTAs, we find that provisions such as investment, subsidies, and services, and to a lesser extent, rules of origin and movement of capital have a median score above the overall score average, suggesting that these policy areas are good predictors of bilateral trade, after controlling for the usual gravity determinants of trade flows. Provisions in policy areas such as SPS, environmental laws, and visa and asylum are located at the other extreme of the distribution of median scores, suggesting a more limited role in predicting bilateral trade flows. The size of the boxes and the vertical lines also indicate that there are policy areas such as movement of capital and IPR for which the contribution to trade is more or less uniform across provisions. For other policy areas such as competition policy and SOEs, there is more heterogeneity within provisions in terms of their contribution to trade. 17 This exercise has been carried out in collaboration with Kazusa Yoshimura and Edith Laget. Parallel work by Breinlich et al. (2020) also uses machine learning techniques to precisely quantify the impact of individual provisions in trade agreements on trade flows. 18 RF is a frequently used machine learning algorithm that predicts a Y variable by combining the results from hundreds of regression/classification trees. It has the merit of not imposing a linear relationship between the Y and X variables, which is an advantage when analyzing the impact of a highly heterogenous set of variables, such as the provisions in PTAs. 19 A score should not be interpreted as a coefficient in a regression analysis. It measures how much the accuracy of the prediction for Y gets worse if the particular X variable is randomly permuted. 21 Handbook of Deep Trade Agreements Figure O.12: Boxplot of scores calculated by the RF of variables/provisions in PTAs O.4.3 Quantification challenges: some analytical issues going forward Looking ahead, there is a need for stronger analytical underpinnings for any quantification exercise. Ideally, the “value” of a commitment must be evaluated in light of the objective that the provision of the deep trade agreement is trying to achieve. In other words, depth indicators could use different weights, depending on whether the outcome variable is market access, welfare, or another metric. For trade policy, market access may seem to be the most obvious metric, but for intellectual property rights, welfare may be the more relevant. In still other areas, such as competition policy, both might be relevant: the market access measure would include only provisions restricting barriers to foreign entry and operation while the welfare measure would include provisions requiring action against anti-competitive behavior affecting consumers. 22 The Evolution of Deep Trade Agreements One indicator cannot provide a measure of both the trade distortions a country imposes on its trading partners (market access) and the trade distortions a country imposes on itself (welfare). For a market access-based measure in the goods context, the relevant question could be: what is the uniform tariff that if imposed on home imports instead of the existing structure of protection would leave aggregate imports at their current level? And for a welfare-based measure: what is the uniform tariff that if applied to imports instead of the current structure of protection would leave home welfare at its current level? The relationship between the two measures is likely to vary across policy areas: positive correlation for tariffs; perhaps negative for environmental standards; and ambiguous for intellectual property rights. A further issue relates to whether we should be interested in what legal commitments do to the level of a policy or to its variance. Provisions such as the elimination of tariffs, or of a national treatment rule in services or government procurement, fix the level of protection at zero. Provisions which legally bind policy (e.g., the permissible levels of fees, subsidies, or preferences) truncate the distribution of possible policy outcomes by reducing the variance and hence the expected level of protection. Provisions which reduce discretion, such as rules on customs valuation, licensing, or procurement procedures, narrow the distribution of possible policy outcomes. Finally, we also need to consider whether we should assess agreements per se or agreements relative to applied policies. If we have the relevant data, the mean and variance shift would ideally be assessed relative to the prevailing policy (and not just the law or policy on paper but how it is implemented). For example, a legally binding tariff at 10 percent might have a different value depending on whether the existing tariff was 5, 10, or 20 percent.The creation of new databases on applied policies in goods and services trade may facilitate such analysis. O.5. CONCLUSIONS The World Development Report 2009 made the case that “thicker” borders between countries hurt economic growth, especially in developing countries. Policies that directly or indirectly restrain the international mobility of goods, services, capital, people, and ideas limit, among other things, the scale of the market, which is vital for development.20 Deep trade agreements aim at establishing the rights of economic integration, protecting these rights from importing governments’ actions that could undo them, and regulating actions of exporters that can have negative welfare effects. These agreements have developed over time into a key institutional mechanism for countries to overcome the constraints to economic development created by the thick borders that fragment markets. 20 World Bank 2009. 23 Handbook of Deep Trade Agreements Of course, deep integration is not an end in itself. First, countries at different levels of development may have different institutional needs, and trade agreements still need to strike the right balance between rules in PTAs and the needed discretion at the national level to pursue desirable social objectives. Second, while many deep provisions may be de facto non- discriminatory and apply to members and non-members alike, there is still a tension between the proliferation of regional approaches and multilateral rules enshrined in the WTO. Therefore, from the perspective of both economic development and global governance, the efficient set of rules in DTAs is an empirical question. The wealth of information on the content of the policy areas commonly included in PTAs could provide new impetus to the analysis of the determinants and impact of deep trade agreements. Such analysis would also provide the necessary tools to further understand the opportunities and challenges that countries face in terms of negotiation and implementation of deep trade agreements. We suggest three areas of work going forward. A first step is to improve the measurement of the depth of trade agreements and quantification of its effects. Beyond simple count variables and coverage ratios, more work will be needed to develop new analytic methods to overcome the challenges discussed in the previous section. As shown, machine learning techniques may provide a useful, innovative approach. Second, the detailed information at the level of individual policy areas could inform a series of studies to assess how specific provisions impact trade and other relevant economic variables. As trade policy experts well understand, the devil is often in the details. Finally, the new data and analysis could provide essential information to policymakers on priorities for the negotiation and implementation of trade agreements: finding what potential partners include in their trade deals, identifying best practices in DTAs and areas where practices diverge or overlap across different players, and assessing gaps between international commitments and domestic legislation. ACKNOWLEDGMENTS We are highly indebted to Alvaro Espitia for providing background research for this introduction and more broadly for his outstanding contribution to the production of this Handbook. Kazusa Yoshimura and Edith Laget worked on the quantification exercise using machine learning techniques in Section O.4. 24 The Evolution of Deep Trade Agreements REFERENCES Acharya, R. (ed.) 2016. Regional trade agreements and the multilateral Mulabdic, A., A. Osnago, and M. Ruta. 2017. Deep integration trading system. Cambridge: Cambridge University Press. and UK-EU trade relations. Policy Research Working Paper 7947,World Bank,Washington, DC. Baldwin, R. 2010. 21st century regionalism: Filling the gap between 21st century trade and 20th century trade rules. Working Paper Rubini, L., and T. Wang. 2020. State-owned enterprises. Chapter 2010-31, Geneva Graduate Institute, Geneva. 15 in Handbook of Deep Trade Agreements,, edited by A. Mattoo, N. Rocha, and M. Ruta..Washington, DC:World Bank. Breinlich, H., V. Corradi, N. Rocha, M. Ruta, J. Santos Silva, and T. Zylkin. 2020. “Machine learning in international trade research: Viner, J. 1950. The customs union issue. New York: Carnegie Evaluating the impact of trade agreements.” Unpublished, Endowment for International Peace. University of Surrey and World Bank. World Bank. 2009. World development report 2009: Reshaping economic Dür, A., L. Baccini, and M. Elsig. 2014.The design of international geography. Washington, DC:World Bank. trade agreements: Introducing a new dataset. Review of International Organizations 9(3): 353-75. ———. 2019. World development report 2020: Trading for development in the age of global value chains. Washington, DC:World Bank. Espitia, A., A. Mattoo, M. Mimouni, X. Pichot, and N. Rocha. 2020. Preferential tariffs. Chapter 1 in Handbook of Deep Trade World Trade Organization. 2011. World trade report 2011:The WTO Agreements,, edited by A. Mattoo, N. Rocha, and M. Ruta.. and preferential trade agreements – From co-existence to coherence. Geneva: Washington, DC: World Bank. WTO. Espitia, A., S. Pardo, R. Piermartini, and N. Rocha. 2020. Technical barriers to trade. Chapter 12 in Handbook of Deep Trade Agreements, edited by A. Mattoo, N. Rocha, and M. Ruta.. Washington, DC: World Bank. Freund, C., and E. Ornelas. 2010. Regional trade agreements. Annual Review of Economics 2: 139-66. Hofmann, C., A. Osnago, and M. Ruta. 2017. Horizontal depth: A new database on the content of preferential trade agreements. Policy Research Working Paper 7981,World Bank,Washington, DC. Horn, H., P.C. Mavroidis, and A. Sapir. 2010. Beyond the WTO? An anatomy of EU and US preferential trade agreements. World Economy 33: 1565-88. Kee, Hiau Looi, Alessandro Nicita, and Marcelo Olarreaga. 2009. Estimating trade restrictiveness indices. Economic Journal 119 (534): 172–99. Laget, E., A. Osnago, N. Rocha, and M. Ruta. 2018. Deep trade agreements and global value chains. Policy Research Working Paper 8491, World Bank, Washington, DC. Laget, E., N. Rocha, and G. Varela. 2019. Deep trade agreements and foreign direct investments. Unpublished paper,World Bank, Washington, DC. Limao, N. 2016. Preferential trade agreements. In Handbook of commercial policy, vol. 1, edited by K. Bagwell and R. Staiger. Amsterdam: Elsevier. Mattoo, A., A. Mulabdic, and M. Ruta. 2017. Trade creation and trade diversion in deep agreements. Policy Research Working Paper 8206, World Bank, Washington, DC. Paper funded by the World Bank’s Strategic Research Program. 25 Handbook of Deep Trade Agreements ANNEX Annex Table O.A.1: Number of provisions included and coverage ratio – by agreement Entry Agreement PROVISIONS COVERAGE RATIO into Overall Substantive Overall Substantive Force (Year) (%) (%) 1958 EC Treaty 121 24 13.5 10.8 1960 European Free Trade Association (EFTA) 224 38 24.9 17.0 1961 Central American Common Market (CACM) 155 44 17.2 19.7 1971 EU - Overseas Countries and Territories (OCT) 57 11 6.3 4.9 1973 EC-Iceland 57 11 6.3 4.9 EC-Norway 57 9 6.3 4.0 EC-Switzerland-Liechtenstein 58 9 6.5 4.0 EC (9) Enlargement 127 31 14.1 13.9 Caribbean Community and 152 36 16.9 16.1 Community Market (CARICOM) 1976 APTA 68 10 7.6 4.5 1977 Australia - Papua New Guinea (PATCRA) 19 3 2.1 1.3 EC-Syrian Arab Republic 30 2 3.3 0.9 1981 South Pacific Regional Trade and Economic 21 2 2.3 0.9 Cooperation Agreement (SPARTECA) Latin American Integration Association 32 7 3.6 3.1 EC (10) Enlargement 114 30 12.7 13.5 1983 Australia-New Zealand (ANZCERTA) 108 35 12.0 15.7 1985 US-Israel 90 7 10.0 3.1 1986 EC Enlargement (12) 118 32 13.1 14.3 1987 Panama - Dominican Republic 20 1 2.2 0.4 1988 CAN 116 25 12.9 11.2 1989 Global System of Trade Preferences among 7 0 0.8 0.0 Developing Countries (GSTP) 1991 Lao PDR-Thailand 9 1 1.0 0.4 EU - Andorra 30 6 3.3 2.7 MERCOSUR 161 32 17.9 14.3 1992 Economic Cooperation Organization (ECO) 46 8 5.1 3.6 Turkey-EFTA 120 30 13.3 13.5 ASEAN free trade area 133 18 14.8 8.1 1993 Russian Federation - Uzbekistan 34 13 3.8 5.8 Russian Federation - Tajikistan 36 8 4.0 3.6 Russian Federation - Turkmenistan 40 13 4.4 5.8 Russian Federation - Azerbaijan 41 15 4.6 6.7 Faroe Islands - Norway 65 17 7.2 7.6 ECOWAS 99 18 11.0 8.1 EFTA-Israel 101 22 11.2 9.9 1994 MSG 26 3 2.9 1.3 Georgia - Russian Federation 42 10 4.7 4.5 CIS 109 23 12.1 10.3 COMESA 158 40 17.6 17.9 EEA 213 42 23.7 18.8 NAFTA 360 68 40.0 30.5 26 The Evolution of Deep Trade Agreements Entry Agreement PROVISIONS COVERAGE RATIO into Overall Substantive Overall Substantive Force (Year) (%) (%) 1995 South Asian Preferential Trade 11 2 1.2 0.9 Agreement (SAPTA) Faroe Islands - Switzerland 21 6 2.3 2.7 Kyrgyz Republic - Armenia 37 11 4.1 4.9 Ukraine-Turkmenistan 38 8 4.2 3.6 Kyrgyz Republic - Kazakhstan 39 7 4.3 3.1 Armenia - Moldova 39 10 4.3 4.5 EC Enlargement (15) 111 31 12.3 13.9 Colombia - Mexico 254 61 28.3 27.4 1996 Ukraine - Azerbaijan 34 6 3.8 2.7 Armenia - Turkmenistan 36 10 4.0 4.5 Ukraine - Uzbekistan 37 5 4.1 2.2 Georgia - Azerbaijan 38 11 4.2 4.9 Georgia - Ukraine 40 12 4.4 5.4 Armenia - Ukraine 40 7 4.4 3.1 Kyrgyz Republic - Moldova 42 10 4.7 4.5 EC-Turkey 110 40 12.2 17.9 1997 Russian Federation - Belarus - Kazakhstan 12 1 1.3 0.4 EAEC 52 8 5.8 3.6 EC-Faroe Islands 59 8 6.6 3.6 Turkey - Israel 74 15 8.2 6.7 EC-West Bank and Gaza 107 21 11.9 9.4 Canada - Israel 141 21 15.7 9.4 Canada - Chile 290 62 32.3 27.8 1998 PAFTA 11 0 1.2 0.0 Kyrgyz Republic - Uzbekistan 37 7 4.1 3.1 Georgia - Armenia 37 10 4.1 4.5 Ukraine-Kazakhstan 37 6 4.1 2.7 Kyrgyz Republic - Ukraine 39 10 4.3 4.5 EC-Tunisia 126 35 14.0 15.7 1999 Georgia - Kazakhstan 40 9 4.4 4.0 Economic and Monetary Community 57 14 6.3 6.3 of Central Africa (CEMAC) EFTA - West Bank and Gaza 100 24 11.1 10.8 EFTA - Morocco 130 31 14.5 13.9 Chile - Mexico 265 69 29.5 30.9 2000 Georgia - Turkmenistan 40 10 4.4 4.5 West African Economic and 61 12 6.8 5.4 Monetary Union (WAEMU) Turkey - FYR Macedonia 73 14 8.1 6.3 Southern African Development Community 84 22 9.3 9.9 East African Community (EAC) 103 32 11.5 14.3 EC-Mexico 121 20 13.5 9.0 27 Handbook of Deep Trade Agreements Entry Agreement PROVISIONS COVERAGE RATIO into Overall Substantive Overall Substantive Force (Year) (%) (%) 2000 EC-Morocco 122 27 13.6 12.1 EC-Israel 125 31 13.9 13.9 EC-South Africa 128 33 14.2 14.8 Israel - Mexico 139 18 15.5 8.1 2001 Armenia - Kazakhstan 38 11 4.2 4.9 India-Sri Lanka 39 8 4.3 3.6 Ukraine - FYR Macedonia 65 19 7.2 8.5 US-Jordan 80 22 8.9 9.9 EC-FYR Macedonia 171 46 19.0 20.6 New Zealand - Singapore 179 34 19.9 15.2 Dominican Republic - Central America 189 44 21.0 19.7 EFTA - Mexico 233 37 25.9 16.6 2002 Asia Pacific Trade Agreement (APTA) - 29 3 3.2 1.3 Accession of China EU-San Marino 32 11 3.6 4.9 Ukraine – Tajikistan 39 8 4.3 3.6 EFTA - FYR Macedonia 121 26 13.5 11.7 EFTA - Jordan 122 25 13.6 11.2 Canada - Costa Rica 155 35 17.2 15.7 EC-Jordan 162 46 18.0 20.6 Chile - Costa Rica (Chile - Central America) 230 49 25.6 22.0 Chile - El Salvador (Chile - Central America) 230 49 25.6 22.0 Japan-Singapore 237 29 26.4 13.0 2003 GCC 32 3 3.6 1.3 Pacific Island Countries Trade Agreement (PICTA) 37 1 4.1 0.4 India - Afghanistan 37 4 4.1 1.8 China - Macao SAR, China 60 14 6.7 6.3 Turkey - Bosnia and Herzegovina 77 16 8.6 7.2 EC-Lebanon 93 32 10.3 14.3 EFTA - Singapore 264 52 29.4 23.3 Panama - El Salvador (Panama - Central America) 271 59 30.1 26.5 EC-Chile 301 54 33.5 24.2 Australia-Singapore 303 76 33.7 34.1 2004 SACU 25 7 2.8 3.1 CEZ 45 13 5.0 5.8 China-Hong Kong SAR, China 64 12 7.1 5.4 EC Enlargement (25) 103 29 11.5 13.0 EC-Arab Rep. of Egypt 132 32 14.7 14.3 EFTA - Chile 232 38 25.8 17.0 Mexico - Uruguay 233 59 25.9 26.5 Panama - Taiwan, China 271 73 30.1 32.7 US-Singapore 318 56 35.4 25.1 28 The Evolution of Deep Trade Agreements Entry Agreement PROVISIONS COVERAGE RATIO into Overall Substantive Overall Substantive Force (Year) (%) (%) 2004 Chile-Rep. of Korea 330 55 36.7 24.7 US-Chile 348 64 38.7 28.7 2005 Pakistan - Sri Lanka 43 7 4.8 3.1 Ukraine - Moldova 65 14 7.2 6.3 Turkey - Tunisia 87 18 9.7 8.1 Turkey - West Bank and Gaza 104 15 11.6 6.7 Jordan - Singapore 120 17 13.3 7.6 EC-Algeria 132 37 14.7 16.6 EFTA - Tunisia 145 30 16.1 13.5 Thailand - New Zealand 178 28 19.8 12.6 China-ASEAN 184 36 20.5 16.1 Australia-Thailand 197 33 21.9 14.8 India-Singapore 197 37 21.9 16.6 Japan-Mexico 252 35 28.0 15.7 US-Australia 324 61 36.0 27.4 2006 India - Bhutan 8 1 0.9 0.4 Ukraine-Belarus 36 8 4.0 3.6 SAFTA 39 2 4.3 0.9 Russian Federation - Serbia 39 7 4.3 3.1 Iceland - Faroe Islands 66 18 7.3 8.1 Turkey - Morocco 85 18 9.5 8.1 Chile-China 130 22 14.5 9.9 EC-Albania 158 39 17.6 17.5 Guatemala – Taiwan, China 227 52 25.3 23.3 US-Bahrain 237 38 26.4 17.0 Japan-Malaysia 245 47 27.3 21.1 Trans-Pacific Strategic Economic Partnership 260 52 28.9 23.3 US-Morocco 273 52 30.4 23.3 Panama - Singapore 274 52 30.5 23.3 EFTA-Rep. of Korea 296 47 32.9 21.1 CAFTA-Dominican Republic 300 53 33.4 23.8 Rep. of Korea-Singapore 315 54 35.0 24.2 2007 Agadir Agreement 22 0 2.4 0.0 East African Community (EAC) - 40 14 4.4 6.3 Accession of Burundi Mauritius and Pakistan 42 7 4.7 3.1 Chile-India 70 11 7.8 4.9 Arab Rep. of Egypt - Turkey 84 18 9.3 8.1 EC Enlargement (27) 102 29 11.3 13.0 Turkey - Syrian Arab Republic 109 17 12.1 7.6 China-Pakistan 120 10 13.3 4.5 EFTA - Lebanon 142 36 15.8 16.1 29 Handbook of Deep Trade Agreements Entry Agreement PROVISIONS COVERAGE RATIO into Overall Substantive Overall Substantive Force (Year) (%) (%) 2007 EFTA - Arab Rep. of Egypt 150 37 16.7 16.6 CEFTA 153 55 17.0 24.7 Chile-Japan 244 39 27.1 17.5 Japan-Thailand 267 55 29.7 24.7 2008 Japan-ASEAN 67 8 7.5 3.6 Turkey - Albania 73 14 8.1 6.3 Turkey - Georgia 78 15 8.7 6.7 EFTA - SACU 108 21 12.0 9.4 Panama - Chile 145 27 16.1 12.1 EC-Montenegro 194 63 21.6 28.3 Brunei Darussalam - Japan 197 34 21.9 15.2 EC-Bosnia Herzegovina 204 57 22.7 25.6 Pakistan - Malaysia 212 37 23.6 16.6 Japan-Indonesia 222 38 24.7 17.0 Chile - Honduras (Chile - Central America) 226 49 25.1 22.0 El Salvador - Honduras - Taiwan, China 233 54 25.9 24.2 Panama - Costa Rica (Panama - Central America) 241 49 26.8 22.0 Japan-Philippines 251 60 27.9 26.9 China-New Zealand 256 50 28.5 22.4 EC-CARIFORUM 334 74 37.2 33.2 Nicaragua - Taiwan, China 334 95 37.2 42.6 2009 India - Nepal 19 0 2.1 0.0 MERCOSUR-India 52 7 5.8 3.1 EU-Côte d’Ivoire 78 11 8.7 4.9 EU - Papua New Guinea/Fiji 96 14 10.7 6.3 EU-Cameroon 108 18 12.0 8.1 Canada-EFTA 158 23 17.6 10.3 China-Singapore 214 39 23.8 17.5 Colombia - Northern Triangle 223 41 24.8 18.4 (El Salvador, Guatemala, Honduras) Panama - Nicaragua (Panama - Central America) 231 45 25.7 20.2 Panama - Guatemala (Panama - Central America 237 49 26.4 22.0 Panama - Honduras (Panama - Central America) 239 49 26.6 22.0 Japan-Vietnam 250 38 27.8 17.0 Peru - Chile 258 64 28.7 28.7 Chile - Colombia 266 60 29.6 26.9 US-Oman 286 59 31.8 26.5 Japan-Switzerland 292 47 32.5 21.1 Peru - Singapore 306 57 34.0 25.6 Chile-Australia 332 51 36.9 22.9 US-Peru 355 74 39.5 33.2 Canada-Peru 375 79 41.7 35.4 30 The Evolution of Deep Trade Agreements Entry Agreement PROVISIONS COVERAGE RATIO into Overall Substantive Overall Substantive Force (Year) (%) (%) 2010 Turkey - Montenegro 71 15 7.9 6.7 Turkey - Serbia 71 14 7.9 6.3 EFTA - Serbia 162 30 18.0 13.5 EFTA - Albania 173 29 19.2 13.0 ASEAN-India 175 29 19.5 13.0 ASEAN-Rep. of Korea 194 31 21.6 13.9 EU-Serbia 209 56 23.2 25.1 Chile - Guatemala (Chile - Central America) 228 49 25.4 22.0 New Zealand - Malaysia 245 44 27.3 19.7 Rep. of Korea - India 252 45 28.0 20.2 China-Peru 260 44 28.9 19.7 ASEAN-Australia-New Zealand 260 59 28.9 26.5 2011 South Asian FTA (SAFTA) - 30 0 3.3 0.0 Accession of Afghanistan Turkey - Jordan 82 16 9.1 7.2 Turkey - Chile 93 19 10.3 8.5 China - Costa Rica 176 23 19.6 10.3 Hong Kong SAR, China - New Zealand 211 42 23.5 18.8 India-Malaysia 213 32 23.7 14.3 India-Japan 234 40 26.0 17.9 EFTA - Peru 291 61 32.4 27.4 EFTA - Colombia 330 63 36.7 28.3 EU - Rep. of Korea 392 81 43.6 36.3 Canada - Colombia 401 79 44.6 35.4 Peru - Rep. of Korea 403 77 44.8 34.5 2012 Treaty on a Free Trade Area between members 37 5 4.1 2.2 of the Commonwealth of Independent States (CIS) El Salvador-Cuba 51 10 5.7 4.5 EU - Eastern and Southern Africa 78 16 8.7 7.2 States Interim EPA Chile - Malaysia 91 12 10.1 5.4 Canada - Jordan 120 29 13.3 13.0 EFTA - Montenegro 206 52 22.9 23.3 Chile - Nicaragua (Chile - Central America) 227 48 25.3 21.5 Peru - Mexico 236 45 26.3 20.2 Mexico - Central America 244 56 27.1 25.1 EFTA - Hong Kong SAR, China 295 54 32.8 24.2 EFTA - Ukraine 302 54 33.6 24.2 Panama - Peru 311 65 34.6 29.1 Japan - Peru 324 51 36.0 22.9 US - Panama 328 71 36.5 31.8 Rep. of Korea - US 344 68 38.3 30.5 US - Colombia 355 76 39.5 34.1 31 Handbook of Deep Trade Agreements Entry Agreement PROVISIONS COVERAGE RATIO into Overall Substantive Overall Substantive Force (Year) (%) (%) 2013 Turkey - Mauritius 55 7 6.1 3.1 EU (28) Enlargement 102 31 11.3 13.9 Ukraine - Montenegro 133 22 14.8 9.9 Rep. of Korea - Turkey 140 42 15.6 18.8 Gulf Cooperation Council (GCC) - Singapore 184 27 20.5 12.1 Malaysia - Australia 291 65 32.4 29.1 Costa Rica - Singapore 294 57 32.7 25.6 Costa Rica - Peru 301 64 33.5 28.7 New Zealand - Taiwan, China 340 70 37.8 31.4 Canada - Panama 345 74 38.4 33.2 EU - Central America 395 88 43.9 39.5 EU - Colombia and Peru 399 89 44.4 39.9 2014 Chile - Vietnam 75 5 8.3 2.2 Iceland - China 180 32 20.0 14.3 Hong Kong SAR, China - Chile 189 27 21.0 12.1 Switzerland - China 210 40 23.4 17.9 Singapore - Taiwan, China 274 40 30.5 17.9 EFTA - Central America (Costa Rica and Panama) 356 74 39.6 33.2 Canada - Honduras 357 74 39.7 33.2 EU - Georgia 376 96 41.8 43.0 EU - Moldova 402 94 44.7 42.2 Rep. of Korea - Australia 415 91 46.2 40.8 EU Ukraine 448 111 49.8 49.8 2015 Southern African Development Community 40 10 4.4 4.5 (SADC) - Accession of Seychelles Eurasian Economic Union (EAEU) - 114 38 12.7 17.0 Accession of Armenia Eurasian Economic Union (EAEU) - 117 40 13.0 17.9 Accession of Kyrgyz Republic Mexico - Panama 169 24 18.8 10.8 Rep. of Korea - Vietnam 186 18 20.7 8.1 Australia - China 202 39 22.5 17.5 EFTA - Bosnia and Herzegovina 206 48 22.9 21.5 Rep. of Korea - New Zealand 240 37 26.7 16.6 Eurasian Economic Union (EAEU) 243 63 27.0 28.3 China - Rep. of Korea 262 56 29.1 25.1 Japan - Australia 342 63 38.0 28.3 Canada - Rep. of Korea 415 93 46.2 41.7 2016 Pacific Alliance 110 27 12.2 12.1 Costa Rica - Colombia 172 28 19.1 12.6 Japan - Mongolia 210 42 23.4 18.8 Rep. of Korea - Colombia 267 49 29.7 22.0 2017 Trans-Pacific Partnership 486 136 54.1 61.0 32 The Evolution of Deep Trade Agreements Annex Table O.A.2: Average number of provisions included and coverage ratio – by country Number of PROVISIONS COVERAGE RATIO Economy Agreements Substantive Overall Substantive Overall (%) (%) Afghanistan 2 33.5 2.0 3.7 0.9 Albania 4 139.3 34.3 15.5 15.4 Algeria 2 71.5 18.5 8.0 8.3 Andorra 1 30.0 6.0 3.3 2.7 Angola 2 62.0 16.0 6.9 7.2 Antigua and Barbuda 1 152.0 36.0 16.9 16.1 Argentina 4 63.0 11.5 7.0 5.2 Armenia 9 77.9 22.2 8.7 10.0 Aruba 1 57.0 11.0 6.3 4.9 Australia 13 253.8 54.9 28.2 24.6 Austria 37 170.1 39.7 18.9 17.8 Azerbaijan 5 51.8 12.0 5.8 5.4 Bahamas, The 2 243.0 55.0 27.0 24.7 Bahrain 4 116.0 17.0 12.9 7.6 Bangladesh 5 35.4 3.4 3.9 1.5 Barbados 1 334.0 74.0 37.2 33.2 Belarus 7 88.4 24.4 9.8 11.0 Belgium 37 170.1 39.7 18.9 17.8 Belize 2 243.0 55.0 27.0 24.7 Benin 2 80.0 15.0 8.9 6.7 Bermuda 1 57.0 11.0 6.3 4.9 Bhutan 4 22.0 1.3 2.4 0.6 Bolivia 2 74.0 16.0 8.2 7.2 Bosnia and Herzegovina 4 160.0 44.0 17.8 19.7 Botswana 4 64.3 15.0 7.1 6.7 Brazil 3 81.7 15.3 9.1 6.9 Brunei Darussalam 9 217.3 44.8 24.2 20.1 Bulgaria 37 165.4 39.7 18.9 17.8 Burkina Faso 2 80.0 15.0 8.9 6.7 Burundi 3 100.3 28.7 11.2 12.9 Cabo Verde 1 99.0 18.0 11.0 8.1 Cambodia 6 168.8 30.2 18.8 13.5 Cameroon 2 82.5 16.0 9.2 7.2 Canada 12 300.3 64.4 33.4 28.9 Cayman Islands 1 57.0 11.0 6.3 4.9 Central African Republic 1 57.0 14.0 6.3 6.3 Chad 1 57.0 14.0 6.3 6.3 Chile 26 218.8 44.0 24.3 19.7 China 14 167.6 30.0 18.6 13.5 Colombia 12 243.8 50.4 27.1 22.6 Comoros 1 158.0 40.0 17.6 17.9 33 Handbook of Deep Trade Agreements Number of PROVISIONS COVERAGE RATIO Economy Substantive Agreements Overall Substantive Overall (%) (%) Congo, Dem. Rep. 3 94.0 24.0 10.5 10.8 Congo, Rep. 1 57.0 14.0 6.3 6.3 Costa Rica 13 246.8 51.1 27.4 22.9 Côte d’Ivoire 3 79.3 13.7 8.8 6.1 Croatia 37 170.1 39.7 18.9 17.8 Cuba 2 41.5 8.5 4.6 3.8 Cyprus 37 170.1 39.7 18.9 17.8 Czech Republic 37 165.4 39.7 18.9 17.8 Denmark 37 170.1 39.7 18.9 17.8 Djibouti 1 158.0 40.0 17.6 17.9 Dominica 2 243.0 55.0 27.0 24.7 Dominican Republic 3 181.0 39.7 20.1 17.8 Ecuador 2 74.0 16.0 8.2 7.2 Egypt, Arab Rep. 6 92.8 21.2 10.3 9.5 El Salvador 10 229.1 49.8 25.5 22.3 Equatorial Guinea 1 57.0 14.0 6.3 6.3 Eritrea 1 158.0 40.0 17.6 17.9 Estonia 37 165.4 39.7 18.9 17.8 Ethiopia 1 158.0 40.0 17.6 17.9 Faroe Islands 4 52.8 12.3 5.9 5.5 Fiji 4 45.0 5.0 5.0 2.2 Finland 37 165.4 39.7 18.9 17.8 France 37 170.1 39.7 18.9 17.8 French Polynesia 1 57.0 11.0 6.3 4.9 Gabon 1 57.0 14.0 6.3 6.3 Gambia, The 1 99.0 18.0 11.0 8.1 Georgia 8 86.4 21.6 9.6 9.7 Germany 37 170.1 39.7 18.9 17.8 Ghana 1 99.0 18.0 11.0 8.1 Greece 37 170.1 39.7 18.9 17.8 Greenland 1 57.0 11.0 6.3 4.9 Grenada 2 243.0 55.0 27.0 24.7 Guatemala 9 244.2 52.9 27.2 23.7 Guinea 1 99.0 18.0 11.0 8.1 Guinea-Bissau 2 80.0 15.0 8.9 6.7 Guyana 2 243.0 55.0 27.0 24.7 Haiti 1 152.0 36.0 16.9 16.1 Honduras 10 256.1 55.2 28.5 24.8 Hong Kong SAR, China 4 189.8 33.8 21.1 15.1 Hungary 37 165.4 39.7 18.9 17.8 Iceland 29 189.1 37.4 21.0 16.8 India 16 92.1 14.4 10.2 6.5 34 The Evolution of Deep Trade Agreements Number of PROVISIONS COVERAGE RATIO Economy Substantive Agreements Overall Substantive Overall (%) (%) Indonesia 7 176.4 31.3 19.6 14.0 Iran, Islamic Rep. 1 46.0 8.0 5.1 3.6 Iraq 1 11.0 0.0 1.2 0.0 Ireland 37 165.4 39.7 18.9 17.8 Israel 6 111.7 19.0 12.4 8.5 Italy 37 170.1 39.7 18.9 17.8 Jamaica 2 243.0 55.0 27.0 24.7 Japan 16 257.5 47.6 28.6 21.4 Jordan 8 89.9 19.4 10.0 8.7 Kazakhstan 10 73.7 19.6 8.2 8.8 Kenya 3 100.3 28.7 11.2 12.9 Kiribati 2 29.0 1.5 3.2 0.7 Korea, Rep. 17 267.5 50.4 29.8 22.6 Kuwait 3 75.7 10.0 8.4 4.5 Kyrgyz Republic 9 80.0 21.6 8.9 9.7 Lao PDR 9 124.3 21.7 13.8 9.7 Latvia 37 170.1 39.7 18.9 17.8 Lebanon 3 82.0 22.7 9.1 10.2 Lesotho 4 64.3 15.0 7.1 6.7 Liberia 1 99.0 18.0 11.0 8.1 Libya 2 84.5 20.0 9.4 9.0 Liechtenstein 27 194.0 38.3 21.6 17.2 Lithuania 37 165.4 39.7 18.9 17.8 Luxembourg 37 170.1 39.7 18.9 17.8 Macao SAR, China 1 60.0 14.0 6.7 6.3 Macedonia, FYR 5 116.6 32.0 13.0 14.3 Madagascar 4 90.0 22.0 10.0 9.9 Malawi 3 94.0 24.0 10.5 10.8 Malaysia 13 215.1 42.6 23.9 19.1 Maldives 3 26.7 1.3 3.0 0.6 Mali 2 80.0 15.0 8.9 6.7 Malta 37 165.4 39.7 18.9 17.8 Marshall Islands 1 21.0 2.0 2.3 0.9 Mauritius 6 76.2 17.0 8.5 7.6 Mexico 14 223.9 47.3 24.9 21.2 Micronesia, Fed. Sts. 2 29.0 1.5 3.2 0.7 Moldova 5 140.2 36.6 15.6 16.4 Mongolia 1 210.0 42.0 23.4 18.8 Montenegro 5 151.4 41.4 16.8 18.6 Morocco 6 107.2 21.3 11.9 9.6 Mozambique 2 62.0 16.0 6.9 7.2 Myanmar 6 168.8 30.2 18.8 13.5 35 Handbook of Deep Trade Agreements Number of PROVISIONS COVERAGE RATIO Economy Agreements Overall Substantive Overall Substantive (%) (%) Namibia 4 64.3 15.0 7.1 6.7 Nauru 2 29.0 1.5 3.2 0.7 Nepal 4 24.8 1.0 2.8 0.4 Netherlands 37 170.1 39.7 18.9 17.8 New Caledonia 1 57.0 11.0 6.3 4.9 New Zealand 12 232.0 49.1 25.8 22.0 Nicaragua 8 259.4 59.1 28.9 26.5 Niger 2 80.0 15.0 8.9 6.7 Nigeria 1 99.0 18.0 11.0 8.1 Norway 28 189.4 37.5 21.1 16.8 Oman 4 128.3 22.3 14.3 10.0 Pakistan 9 61.1 8.1 6.8 3.6 Panama 14 245.6 50.9 27.3 22.8 Papua New Guinea 5 39.8 4.6 4.4 2.1 Paraguay 3 81.7 15.3 9.1 6.9 Peru 16 285.2 60.3 31.7 27.0 Philippines 7 180.6 34.4 20.1 15.4 Poland 37 165.4 39.7 18.9 17.8 Portugal 37 170.1 39.7 18.9 17.8 Qatar 3 75.7 10.0 8.4 4.5 Romania 37 170.1 39.7 18.9 17.8 Russian Federation 12 67.9 19.1 7.6 8.6 Rwanda 3 100.3 28.7 11.2 12.9 Samoa 2 29.0 1.5 3.2 0.7 San Marino 2 240.0 61.0 26.7 27.4 Saudi Arabia 3 75.7 10.0 8.4 4.5 Senegal 2 80.0 15.0 8.9 6.7 Serbia 5 126.8 32.4 14.1 14.5 Seychelles 4 90.0 22.0 10.0 9.9 Sierra Leone 1 99.0 18.0 11.0 8.1 Singapore 22 238.1 45.3 26.5 20.3 Slovak Republic 37 170.1 39.7 18.9 17.8 Slovenia 37 165.4 39.7 18.9 17.8 Solomon Islands 3 28.0 2.0 3.1 0.9 South Africa 5 77.0 18.6 8.6 8.3 South Sudan 1 158.0 40.0 17.6 17.9 Spain 37 165.4 39.7 18.9 17.8 Sri Lanka 7 37.0 4.6 4.1 2.0 St. Kitts and Nevis 2 243.0 55.0 27.0 24.7 St. Lucia 2 243.0 55.0 27.0 24.7 St.Vincent and the Grenadines 2 243.0 55.0 27.0 24.7 Sudan 1 11.0 0.0 1.2 0.0 36 The Evolution of Deep Trade Agreements Number of PROVISIONS COVERAGE RATIO Economy Agreements Overall Substantive Overall Substantive (%) (%) Suriname 2 243.0 55.0 27.0 24.7 Swaziland 5 83.0 20.0 9.2 9.0 Sweden 37 165.4 39.7 18.9 17.8 Switzerland 30 192.0 37.5 21.4 16.8 Syrian Arab Republic 3 50.0 6.3 5.6 2.8 Taiwan, China 6 279.8 64.0 31.1 28.7 Tajikistan 3 42.3 8.0 4.7 3.6 Tanzania 4 66.8 19.5 7.4 8.7 Thailand 10 166.4 29.8 18.5 13.4 Togo 2 80.0 15.0 8.9 6.7 Tonga 2 29.0 1.5 3.2 0.7 Trinidad and Tobago 2 243.0 55.0 27.0 24.7 Tunisia 5 78.2 16.6 8.7 7.4 Turkey 19 85.9 18.5 9.6 8.3 Turkmenistan 6 50.0 11.5 5.6 5.2 Turks and Caicos Islands 1 57.0 11.0 6.3 4.9 Tuvalu 2 29.0 1.5 3.2 0.7 Uganda 3 100.3 28.7 11.2 12.9 Ukraine 14 67.9 13.7 7.5 6.1 United Arab Emirates 3 75.7 10.0 8.4 4.5 United Kingdom 37 165.4 39.7 18.9 17.8 United States 14 285.6 54.9 31.8 24.6 Uruguay 4 119.5 26.3 13.3 11.8 Uzbekistan 5 50.8 10.6 5.7 4.8 Vanuatu 3 28.0 2.0 3.1 0.9 Venezuela, RB 3 81.7 15.3 9.1 6.9 Vietnam 10 201.0 37.8 22.4 17.0 West Bank and Gaza 3 103.7 20.0 11.5 9.0 Yemen, Rep. 1 11.0 0.0 1.2 0.0 Zambia 3 94.0 24.0 10.5 10.8 Zimbabwe 4 90.0 22.0 10.0 9.9 37 Handbook of Deep Trade Agreements Annex Table O.A.3: Maximum number of provisions included and coverage ratio – by country Country PROVISIONS SUBSTANTIVE Agreement Provisions Coverage Agreement Provisions Coverage Ratio (%) Ratio (%) Afghanistan India - Afghanistan 37 4.1 India - Afghanistan 4 1.8 Albania EFTA - Albania 173 19.2 CEFTA 55 24.7 Algeria EC-Algeria 132 14.7 EC-Algeria 37 16.6 Andorra EU - Andorra 30 3.3 EU - Andorra 6 2.7 Angola Southern African 84 9.3 Southern African 22 9.9 Development Development Community Community Antigua CARICOM 152 16.9 CARICOM 36 16.1 and Barbuda Argentina MERCOSUR 161 17.9 MERCOSUR 32 14.3 Armenia EAEU 243 27.0 EAEU 63 28.3 Aruba EU - Overseas 57 6.3 EU - Overseas 11 4.9 Countries and Countries and Territories (OCT) Territories (OCT) Australia Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Austria EU Ukraine 448 49.8 EU Ukraine 111 49.8 Azerbaijan CIS 109 12.1 CIS 23 10.3 Bahamas, The EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Bahrain US-Bahrain 237 26.4 US-Bahrain 38 17.0 Bangladesh APTA 68 7.6 APTA 10 4.5 Barbados EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Belarus Eurasian Economic 243 27.0 Eurasian Economic 63 28.3 Union (EAEU) Union (EAEU) Belgium EU Ukraine 448 49.8 EU Ukraine 111 49.8 Belize EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Benin ECOWAS 99 11.0 ECOWAS 18 8.1 Bermuda EU - Overseas Countries 57 6.3 EU - Overseas Countries 11 4.9 and Territories (OCT) and Territories (OCT) Bhutan SAFTA 39 4.3 SAFTA 2 0.9 Bolivia CAN 116 12.9 CAN 25 11.2 Bosnia and EFTA - Bosnia 206 22.9 EC-Bosnia Herzegovina 57 25.6 Herzegovina and Herzegovina Herzegovina Botswana EFTA - SACU 108 12.0 Southern African 22 9.9 Development Community Brazil MERCOSUR 161 17.9 MERCOSUR 32 14.3 Brunei Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Darussalam Bulgaria EU Ukraine 448 49.8 EU Ukraine 111 49.8 Burkina Faso ECOWAS 99 11.0 ECOWAS 18 8.1 Burundi COMESA 158 17.6 COMESA 40 17.9 Cabo Verde ECOWAS 99 11.0 ECOWAS 18 8.1 Cambodia ASEAN-Australia- 260 28.9 ASEAN-Australia- 59 26.5 New Zealand New Zealand Cameroon EC-Cameroon 108 12.0 EC-Cameroon 18 8.1 Canada Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Cayman EU - Overseas 57 6.3 EU - Overseas 11 4.9 Islands Countries and Countries and Territories (OCT) Territories (OCT) 38 The Evolution of Deep Trade Agreements Country PROVISIONS SUBSTANTIVE Agreement Provisions Coverage Agreement Provisions Coverage Ratio (%) Ratio (%) Central Economic and 57 6.3 Economic and 14 6.3 African Monetary Community Monetary Community Republic of Central Africa of Central Africa (CEMAC) (CEMAC) Chad Economic and 57 6.3 Economic and 14 6.3 Monetary Community Monetary Community of Central Africa of Central Africa (CEMAC) (CEMAC) Chile Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 China China- 262 29.1 China- 56 25.1 Rep. of Korea Rep. of Korea Colombia Canada-Colombia 401 44.6 EU - Colombia and Peru 89 39.9 Comoros COMESA 158 17.6 COMESA 40 17.9 Congo, COMESA 158 17.6 COMESA 40 17.9 Dem. Rep. Congo, Rep. Economic and Monetary 57 6.3 Economic and Monetary 14 6.3 Community of Central Community of Central Africa (CEMAC) Africa (CEMAC) Costa Rica EU - Central America 395 43.9 EU - Central America 88 39.5 Côte d’Ivoire ECOWAS 99 11.0 ECOWAS 18 8.1 Croatia EU Ukraine 448 49.8 EU Ukraine 111 49.8 Cuba El Salvador-Cuba 51 5.7 El Salvador-Cuba 10 4.5 Cyprus EU Ukraine 448 49.8 EU Ukraine 111 49.8 Czech Republic EU Ukraine 448 49.8 EU Ukraine 111 49.8 Denmark EU Ukraine 448 49.8 EU Ukraine 111 49.8 Djibouti COMESA 158 17.6 COMESA 40 17.9 Dominica EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Dominican EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Republic Ecuador CAN 116 12.9 CAN 25 11.2 COMESA Egypt, Arab Rep. 158 17.6 COMESA 40 17.9 El Salvador EU - Central America 395 43.9 EU - Central America 88 39.5 Equatorial Economic and Monetary 57 6.3 Economic and Monetary 14 6.3 Guinea Community of Central Community of Central Africa (CEMAC) Africa (CEMAC) Eritrea COMESA 158 17.6 COMESA 40 17.9 Estonia EU Ukraine 448 49.8 EU Ukraine 111 49.8 Ethiopia COMESA 158 17.6 COMESA 40 17.9 Faroe Islands Iceland - Faroe Islands 66 7.3 Iceland - Faroe Islands 18 8.1 Fiji EU - Papua 96 10.7 EU - Papua 14 6.3 New Guinea/Fiji New Guinea/Fiji Finland EU Ukraine 448 49.8 EU Ukraine 111 49.8 France EU Ukraine 448 49.8 EU Ukraine 111 49.8 French EU - Overseas Countries 57 6.3 EU - Overseas Countries 11 4.9 Polynesia and Territories (OCT) and Territories (OCT) Gabon Economic and 57 6.3 Economic and 14 6.3 Monetary Community Monetary Community of Central Africa of Central Africa (CEMAC) (CEMAC) Gambia, The ECOWAS 99 11.0 ECOWAS 18 8.1 39 Handbook of Deep Trade Agreements Country PROVISIONS SUBSTANTIVE Agreement Provisions Coverage Agreement Provisions Coverage Ratio (%) Ratio (%) Georgia EU - Georgia 376 41.8 EU - Georgia 96 43.0 Germany EU Ukraine 448 49.8 EU Ukraine 111 49.8 Ghana ECOWAS 99 11.0 ECOWAS 18 8.1 Greece EU Ukraine 448 49.8 EU Ukraine 111 49.8 Greenland EU - Overseas 57 6.3 EU - Overseas 11 4.9 Countries and Countries and Territories (OCT) Territories (OCT) Grenada EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Guatemala EU - Central America 395 43.9 EU - Central America 88 39.5 Guinea ECOWAS 99 11.0 ECOWAS 18 8.1 Guinea-Bissau ECOWAS 99 11.0 ECOWAS 18 8.1 Guyana EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Haiti Caribbean Community 152 16.9 Caribbean Community 36 16.1 and Community and Community Market (CARICOM) Market (CARICOM) Honduras EU - Central America 395 43.9 EU - Central America 88 39.5 Hong Kong EFTA - Hong Kong SAR, 295 32.8 EFTA - Hong Kong SAR, 54 24.2 SAR, China China China Hungary EU Ukraine 448 49.8 EU Ukraine 111 49.8 Iceland EFTA - Central America 356 39.6 EFTA - Central America 74 33.2 (Costa Rica and Panama) (Costa Rica and Panama) India Rep. of Korea-India 252 28.0 Rep. of Korea-India 45 20.2 Indonesia ASEAN-Australia- 260 28.9 ASEAN-Australia- 59 26.5 New Zealand New Zealand Iran, Islamic Economic Cooperation 46 5.1 Economic Cooperation 8 3.6 Rep. Organization (ECO) Organization (ECO) Iraq PAFTA 11 1.2 PAFTA 0 0.0 Ireland EU Ukraine 448 49.8 EU Ukraine 111 49.8 Israel Canada - Israel 141 15.7 EC-Israel 31 13.9 Italy EU Ukraine 448 49.8 EU Ukraine 111 49.8 Jamaica EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Japan Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Jordan EC-Jordan 162 18.0 EC-Jordan 46 20.6 Kazakhstan Eurasian Economic 243 27.0 Eurasian Economic 63 28.3 Union (EAEU) Union (EAEU) Kenya COMESA 158 17.6 COMESA 40 17.9 Kiribati Pacific Island Countries 37 4.1 South Pacific Regional 2 0.9 Trade Agreement (PICTA) Trade and Economic Cooperation Agreement (SPARTECA) Korea, Rep. Canada - Rep. of Korea 415 46.2 Canada - Rep. of Korea 93 41.7 Kuwait Gulf Cooperation 184 20.5 Gulf Cooperation 27 12.1 Council (GCC) - Council (GCC) - Singapore Singapore Kyrgyz Eurasian Economic 243 27.0 Eurasian Economic 63 28.3 Republic Union (EAEU) Union (EAEU) Lao PDR ASEAN-Australia- 260 28.9 ASEAN-Australia- 59 26.5 New Zealand New Zealand Latvia EU Ukraine 448 49.8 EU Ukraine 111 49.8 Lebanon EFTA - Lebanon 142 15.8 EFTA - Lebanon 36 16.1 40 The Evolution of Deep Trade Agreements Country PROVISIONS SUBSTANTIVE Agreement Provisions Coverage Agreement Provisions Coverage Ratio (%) Ratio (%) Lesotho EFTA - SACU 108 12.0 Southern African 22 9.9 Development Community Liberia ECOWAS 99 11.0 ECOWAS 18 8.1 Libya COMESA 158 17.6 COMESA 40 17.9 Liechtenstein EFTA - Central America 356 39.6 EFTA - Central America 74 33.2 (Costa Rica and Panama) (Costa Rica and Panama) Lithuania EU Ukraine 448 49.8 EU Ukraine 111 49.8 Luxembourg EU Ukraine 448 49.8 EU Ukraine 111 49.8 Macao SAR, China - Macao SAR, China 60 6.7 China - Macao SAR, China 14 6.3 China Macedonia, FYREC-FYR Macedonia 171 19.0 CEFTA 55 24.7 Madagascar COMESA 158 17.6 COMESA 40 17.9 Malawi COMESA 158 17.6 COMESA 40 17.9 Malaysia Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Maldives SAFTA 39 4.3 SAFTA 2 0.9 Mali ECOWAS 99 11.0 ECOWAS 18 8.1 Malta EU Ukraine 448 49.8 EU Ukraine 111 49.8 Marshall South Pacific Regional 21 2.3 South Pacific Regional 2 0.9 Islands Trade and Economic Trade and Economic Cooperation Agreement Cooperation Agreement (SPARTECA) (SPARTECA) Mauritius COMESA 158 17.6 COMESA 40 17.9 Mexico Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Micronesia, Pacific Island Countries 37 4.1 South Pacific Regional 2 0.9 Fed. Sts. Trade Agreement Trade and Economic (PICTA) Cooperation Agreement (SPARTECA) Moldova EU - Republic 402 44.7 EU - Republic 94 42.2 of Moldova of Moldova Mongolia Japan - Mongolia 210 23.4 Japan - Mongolia 42 18.8 Montenegro EFTA - Montenegro 206 22.9 EC-Montenegro 63 28.3 Morocco US-Morocco 273 30.4 US-Morocco 52 23.3 Mozambique Southern African 84 9.3 Southern African 22 9.9 Development Community Development Community Myanmar ASEAN-Australia- 260 28.9 ASEAN-Australia- 59 26.5 New Zealand New Zealand Namibia EFTA - SACU 108 12.0 Southern African 22 9.9 Development Community Nauru Pacific Island Countries 37 4.1 South Pacific 2 0.9 Trade Agreement Regional Trade (PICTA) and Economic Cooperation Agreement (SPARTECA) Nepal SAFTA 39 4.3 SAFTA 2 0.9 Netherlands EU Ukraine 448 49.8 EU Ukraine 111 49.8 New EU - Overseas Countries 57 6.3 EU - Overseas Countries 11 4.9 Caledonia and Territories (OCT) and Territories (OCT) New Zealand Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Nicaragua EU - Central America 395 43.9 Nicaragua - Taiwan, China 95 42.6 Niger ECOWAS 99 11.0 ECOWAS 18 8.1 41 Handbook of Deep Trade Agreements Country PROVISIONS SUBSTANTIVE Agreement Provisions Coverage Agreement Provisions Coverage Ratio (%) Ratio (%) Nigeria ECOWAS 99 11.0 ECOWAS 18 8.1 Norway EFTA - Central America 356 39.6 EFTA - Central America 74 33.2 (Costa Rica and Panama) (Costa Rica and Panama) Oman US-Oman 286 31.8 US-Oman 59 26.5 Pakistan Pakistan - Malaysia 212 23.6 Pakistan - Malaysia 37 16.6 Panama EFTA - Central America 356 39.6 Canada - Panama 74 33.2 (Costa Rica and Panama) Papua New EU - Papua 96 10.7 EU - Papua 14 6.3 Guinea New Guinea/Fiji New Guinea/Fiji Paraguay MERCOSUR 161 17.9 MERCOSUR 32 14.3 Peru Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Philippines ASEAN-Australia- 260 28.9 Japan-Philippines 60 26.9 New Zealand Poland EU Ukraine 448 49.8 EU Ukraine 111 49.8 Portugal EU Ukraine 448 49.8 EU Ukraine 111 49.8 Qatar Gulf Cooperation 184 20.5 Gulf Cooperation 27 12.1 Council (GCC) - Council (GCC) - Singapore Singapore Romania EU Ukraine 448 49.8 EU Ukraine 111 49.8 Russian Eurasian Economic 243 27.0 Eurasian Economic 63 28.3 Federation Union (EAEU) Union (EAEU) Rwanda COMESA 158 17.6 COMESA 40 17.9 Samoa Pacific Island Countries 37 4.1 South Pacific Regional 2 0.9 Trade Agreement Trade and Economic (PICTA) Cooperation Agreement (SPARTECA) San Marino EU Ukraine 448 49.8 EU Ukraine 111 49.8 Saudi Arabia Gulf Cooperation 184 20.5 Gulf Cooperation 27 12.1 Council (GCC) - Council (GCC) - Singapore Singapore Senegal ECOWAS 99 11.0 ECOWAS 18 8.1 Serbia EU-Serbia 209 23.2 EU-Serbia 56 25.1 Seychelles COMESA 158 17.6 COMESA 40 17.9 Sierra Leone ECOWAS 99 11.0 ECOWAS 18 8.1 Singapore Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 Slovak Republic EU Ukraine 448 49.8 EU Ukraine 111 49.8 Slovenia EU Ukraine 448 49.8 EU Ukraine 111 49.8 Solomon Pacific Island Countries 37 4.1 MSG 3 1.3 Islands Trade Agreement (PICTA) South Africa EC-South Africa 128 14.2 EC-South Africa 33 14.8 South Sudan COMESA 158 17.6 COMESA 40 17.9 Spain EU Ukraine 448 49.8 EU Ukraine 111 49.8 Sri Lanka APTA 68 7.6 APTA 10 4.5 EC-CARIFORUM 334 St.Kitts and Nevis 37.2 EC-CARIFORUM 74 33.2 St. Lucia EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 St.Vincent and EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 the Grenadines Sudan PAFTA 11 1.2 PAFTA 0 0.0 42 The Evolution of Deep Trade Agreements Country PROVISIONS SUBSTANTIVE Agreement Provisions Coverage Agreement Provisions Coverage Ratio (%) Ratio (%) Suriname EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 Swaziland COMESA 158 17.6 COMESA 40 17.9 Sweden EU Ukraine 448 49.8 EU Ukraine 111 49.8 Switzerland EFTA - Central America 356 39.6 EFTA - Central America 74 33.2 (Costa Rica and Panama) (Costa Rica and Panama) Syrian Arab Turkey- 109 12.1 Turkey- 17 7.6 Republic Syrian Arab Republic Syrian Arab Republic Taiwan, China New Zealand - 340 37.8 Nicaragua - 95 42.6 Taiwan, China Taiwan, China Tajikistan EAEC 52 5.8 EAEC 8 3.6 Tanzania East African 103 11.5 East African 32 14.3 Community (EAC) Community (EAC) Thailand Japan-Thailand 267 29.7 ASEAN-Australia- 59 26.5 New Zealand Togo ECOWAS 99 11.0 ECOWAS 18 8.1 Tonga Pacific Island Countries 37 4.1 South Pacific Regional 2 0.9 Trade Agreement Trade and Economic (PICTA) Cooperation Agreement (SPARTECA) Trinidad EC-CARIFORUM 334 37.2 EC-CARIFORUM 74 33.2 and Tobago Tunisia EFTA - Tunisia 145 16.1 EC-Tunisia 35 15.7 Turkey Rep. of Korea 140 15.6 Rep. of Korea 42 18.8 - Turkey - Turkey Turkmenistan CIS 109 12.1 CIS 23 10.3 Turks and EU - Overseas Countries 57 6.3 EU - Overseas Countries 11 4.9 Caicos Islands and Territories (OCT) and Territories (OCT Tuvalu Pacific Island Countries 37 4.1 South Pacific Regional 2 0.9 Trade Agreement Trade and Economic (PICTA) Cooperation Agreement (SPARTECA) Uganda COMESA 158 17.6 COMESA 40 17.9 Ukraine EFTA - Ukraine 302 33.6 EFTA - Ukraine 54 24.2 United Arab Gulf Cooperation Council 184 20.5 Gulf Cooperation Council 27 12.1 Emirates (GCC) - Singapore (GCC) – Singapore United Kingdom EU Ukraine 448 49.8 EU Ukraine 111 49.8 United States NAFTA 360 40.0 US - Colombia 76 34.1 Uruguay Mexico - Uruguay 233 25.9 Mexico - Uruguay 59 26.5 Uzbekistan CIS 109 12.1 CIS 23 10.3 Vanuatu Pacific Island Countries 37 4.1 MSG 3 1.3 Trade Agreement (PICTA) V enezuela, RB MERCOSUR 161 17.9 MERCOSUR 32 14.3 Vietnam Trans-Pacific Partnership 486 54.1 Trans-Pacific Partnership 136 61.0 West Bank EC- 107 11.9 EFTA- 24 10.8 and Gaza West Bank and Gaza West Bank and Gaza Yemen, Rep. PAFTA 11 1.2 PAFTA 0 0.0 Zambia COMESA 158 17.6 COMESA 40 17.9 Zimbabwe COMESA 158 17.6 COMESA 40 17.9 43 Preferential Tariffs CHAPTER 1 Preferential Tariffs A. Espitia, A. Mattoo, M. Mimouni, X. Pichot, and N. Rocha 45 Handbook of Deep Trade Agreements CHAPTER 1 Preferential Tariffs A. Espitia*, A. Mattoo*, M. Mimouni†, X. Pichot†, and N. Rocha* * World Bank, Washington, DC, United States † International Trade Centre, Geneva, Switzerland CONTENTS 1.1. INTRODUCTION 47 1.2. A NEW DATABASE ON PREFERENCE MARGINS AND PREFERENTIAL TRADE 48 1.2.1 ITC database description 49 1.3. THE MFN LEGACY 51 1.4. HOW HAVE PREFERENTIAL AGREEMENTS CHANGED TRADE REGIMES? 53 1.4.1 Patterns of preferential liberalization 53 1.4.2 What did preferences do to tariff peaks? 56 1.4.3 How big is the preferential advantage? 57 1.4.4 How are preference margins distributed? 57 1.5. FROM PREFERENCES IN PRINCIPLE TO PREFERENCES IN PRACTICE 60 1.6. CONCLUSIONS 63 ACKNOWLEDGMENTS 64 REFERENCES 65 ANNEX 66 46 Preferential Tariffs 1.1. INTRODUCTION World trade is increasingly ruled by preferential trade agreements (PTAs), but their precise nature remains relatively opaque.This chapter assesses a central dimension of these agreements, the significance of tariff preferences, using a new database on preferential and nonpreferential most-favored-nation (MFN) tariffs, which represented approximately 97 percent of world imports in 2016. Countries around the world have increased their participation in preferential trade agreements, especially in the last two decades. From the 1950s onwards, the number of active PTAs increased steadily to almost 50 in 1990. Thereafter, PTA activity accelerated noticeably, with the number of PTAs more than doubling over the next five years and more than quadrupling by 2010, to reach close to 280 PTAs presently in force (Figure 1.1). Figure 1.1: Trade agreements have proliferated over time Source: Deep Trade Agreements Database. The existing literature suggests at least two reasons for the significant increase in the number of PTAs. First, the lack of progress in trade negotiations at the multilateral level has improved countries’ incentives to engage in bilateral or regional preferential negotiations.1 Second, the fear of losing market share by being excluded from existing PTAs has pushed more countries to sign PTAs—a “domino effect” of PTAs.2 This chapter addresses two main questions: (a) What is the legacy of unilateral and multilateral liberalization? and (b) how have preferential tariffs changed the trade regime? 1 Capling and Low 2010; Bhagwati 2008. 2 Baldwin and Jaimovich 2010. 47 Handbook of Deep Trade Agreements The chapter is organized as follows. Section 1.2 introduces the new database on tariffs and preference margins that was recently constructed by the International Trade Centre (ITC) and the World Bank (ITC/World Bank database). Section 1.3 describes the multilateral trade regime and the scope for further liberalization. Section 1.4 discusses how preferential tariffs have changed the trade regime. Section 1.5 illustrates the extent of preference utilization focusing on the European Union. Section 1.6 concludes. 1.2. A NEW DATABASE ON PREFERENCE MARGINS AND PREFERENTIAL TRADE This analysis is based on the new database, which includes information on MFN and preferential tariffs imposed at the Harmonized System (HS) 6-digit product level in 2016. The database was constructed by merging different sources of data.The ITC Market Access Map database was the main source of information on ad valorem equivalents at the HS 6-digit level for both MFN- applied tariffs and preferential tariffs by country pair. Imports in 2016 come from UN Comtrade.3 Information on PTAs in force during the same year comes from the newly constructed World Bank database on the content of PTAs.4 Table 1.1: Non–ad valorem tariffs and ad valorem equivalent (AVE) composition NAV tariff category Example Final AVE composition Specific tariff $2 per kg AVE of the specific tariff Compound tariff 10% plus $2 per kg Ad valorem component added to (or subtracted from) the AVE of the specific component Mixed tariff 30% or € 2 per kg, AVE of the specific component subject to whichever is highest the conditional choice expressed in the tariff Tariff rate quota 5% for imports within AVE depends on the real volume of imports quota and 20% for in the year of reference. The marginal level of out-of-quota imports protection of a tariff rate quota consists of the average of the inside and outside tariff rates if the import volume is less than or equal to 80% of the contingent, or the outside tariff if beyond Technical tariff 9% on dairy spreads, with Not calculated due to a lack of information a fat content between on technical product specifications 39% and 60% Source: ITC. 3 TRAINS and COMTRADE information are taken from the World Integrated Trade Solution (WITS) trade platform. 4 Hofmann, Osnago, and Ruta 2017. Database is available at http://data.worldbank.org/data-catalog/deep-trade-agreements. 48 Preferential Tariffs 1.2.1 ITC database description The ITC database includes customs duties at the national tariff line code (NTLC) for 201 reporters faced by 239 partners under MFN, non-MFN, and preferential regimes and tariff rate quotas. The database is continuously updated with tariff data that ITC collects directly from national authorities such as Customs authorities, ministries, and other government institutions. When national sources cannot provide ITC with the preferential rates under a preferential trade agreement that is known to be in force, then ITC obtains the missing information from the tariff phase-out schedules of the agreement. The ITC database contains pre-calculated ad valorem equivalents (AVE) for non-ad valorem duties and tariff rate quotas (TRQs) (Table 1.1). AVEs express non-ad valorem tariffs in percentage terms as follows: AVE = ( SP * XR ) * 100 UV where SP is the monetary value of duty per unit of imports, UV is the import unit value that is calculated as the ratio between the value of imports (V) and the quantity of imports (Q), and XR is the currency exchange rate when appropriate. The accuracy of the AVEs depends on the UV estimates, which are sensitive to variations in the data. ITC’s strategy for selecting the most accurate UV estimates is schematized in Annex Figure 1.A.1.5 Not all non-ad valorem tariffs can be converted into an ad valorem equivalent rate. This is the case for technical duties imposed on some products (Table 1.2). Nonetheless, those that can’t be converted represent only 1.7 percent of the country-pair-product observations in the database. To make the tariffs comparable across countries and sectors, AVEs are aggregated from the NTLC to HS6 by calculating the simple average of all underlying NTLC rates. If there is more than one preferential tariff under a given NTLC for a partner country, then the minimum rate is selected. The most-favored-nation tariff or the general tariff is used if no tariff preference is applicable. The resulting aggregated database includes information on the ad valorem equivalent at the 6-digit HS product level for both the maximum applied rate (MFN rates) and 5 The entire calculation process is detailed in ITC (2006, pp. 186–97). 49 Handbook of Deep Trade Agreements preferential tariffs for a total 199 reporters and 239 partners. Among the 199 reporters, 141 countries reported data for 2016, 7 for 2017, 20 for 2015, and 13 for 2014. For the remaining 18 countries, the most recent available information are for 2006-2013 (Figure 1.2).6 In terms of products, information is reported on all 5,203 HS6 level products (HS 2012 nomenclature). Table 1.2: Examples of technical duties Importing National Product Product description Custom duty country Code as reported Yemen, Rep. 22043000 Wine of fresh grapes, including fortified Prohibited wines; grape must other than that of heading 20.09; other grape must. Russian 8703329093 Motor cars and other motor vehicles principally € 2.2 per cm3 Federation designed for the transport of persons (other of engine volume than those of heading 8702), including station wagons and racing cars; Other vehicles, with compression-ignite. New Zealand 95081000 Roundabouts, swings shooting galleries, and The rates applicable other fairground amusements; travelling circuses to the separate and travelling menageries; travelling theatres components United States 91091010 Alarm clock movements, complete and 3.9% on the assembled, electrically operated, with movement + 5.3% optoelectronic display only on the battery Source: ITC. By construction, MFN tariffs between members of a customs union are not available in the database.7 This is the case for countries which are part of the European Union, the Southern Africa Customs Union (SACU), the Switzerland-Liechtenstein customs union, the Israel-West Bank and Gaza customs union, and the Eurasian Economic Union (EAEU). For this analysis, the missing MFN rate will be replaced by the MFN rate available from other partners as a notional MFN rate to be able to compute preferential margins. 6 For purpose of this analysis, countries with information before 2014 will be included, with the caveat that Panama and Trinidad and Tobago had agreements entering into force after their available data (Annex Table 1.A.2). 7 A member of a customs union does not apply MFN tariffs to other members of the customs union. 50 Preferential Tariffs The reporter-partner-product combinations covered in the database represented approximately 97 percent of world imports in 2016. Non-covered trade is mainly explained by the lack of information on trade flows, either from the reporter or partner country (1.3 percent), or by missing information on MFN rates (0.9 percent) or preferential tariffs (0.6 percent). The information on preferential tariffs covers 94 percent of PTAs notified to the WTO that are currently in force.8 Figure 1.2: Most recent tariff information Source: ITC/World Bank database. 1.3. THE MFN LEGACY MFN tariffs have progressively fallen since the establishment of the General Agreement on Tariffs and Trade (GATT) in 1948. Unilateral liberalization and eight rounds of multilateral trade negotiations have significantly reduced tariffs applied by WTO members. Applied MFN rates fell from levels between 12.5 and 15 percent in 1995 to lower than 10 percent during 2015 (Figure 1.3). Of the total value of imports, 42 percent trades free under MFN rates. Another 45 percent is subject to MFN rates below 10 percent, and only one-tenth to MFN rates above 10 percent. In terms of products, 24 percent of tariff lines are subject to zero MFN rates, 23 percent to MFN rates over 15 percent, and one-quarter to rates between 5 and 10 percent (Figure 1.4). 8 Although all 260 PTAs are included in the database, for 16 agreements (6 percent) information is not available for all partners; for example, in the COMESA agreement, information is missing for South Sudan. See Annex Table 1.A.8. 51 Handbook of Deep Trade Agreements Figure 1.3: Applied MFN rates have steadily declined over time Source: World Integrated Trade Solution. Note: To avoid sample selection bias, tariffs have been calculated for a balanced sub-sample of countries and missing data have been interpolated.The sub-sample includes 27 countries with applied MFN rates in at least 15 years between 1995 and 2015 (see Annex Table 1.A.3).The data used in the figure are simple averages and trade-weighted averages of MFN rates for all products. Figure 1.4: Almost two-thirds of imports by value are subject to MFN rates of less than 5 percent Source: Deep Trade Agreements Database. On average, agricultural imports are subject to higher MFN rates than manufactured goods and natural resources. Whereas more than half the imports of natural resources and around 42 percent of manufacturing goods are subject to a zero MFN rate, less than a quarter of agricultural imports benefit from duty-free treatment. At the same time, nearly 40 percent of agricultural imports are subject to MFN rates higher than 10 percent (Figure 1.5), compared to less than one-tenth of manufacturing imports. A higher share of tariff lines in agriculture is also subject to higher MFN rates, compared to manufacturing and natural resources (Figure 1.6). Nearly two-fifths of agricultural tariff lines and about one-fifth of manufacturing tariff lines are subject to MFN rates higher than 15 percent. 52 Preferential Tariffs Figure 1.5: MFN rates vary significantly across the three economic sectors (value of imports) Source: Deep Trade Agreements Database. Figure 1.6: MFN rates vary significantly across the three economic sectors Source: Deep Trade Agreements Database. 1.4. HOW HAVE PREFERENTIAL AGREEMENTS CHANGED TRADE REGIMES? Lack of progress in multilateral negotiations, among other reasons, has spurred tariff reductions through bilateral and regional preferential trade agreements. 1.4.1 Patterns of preferential liberalization In 2016, preferential trade agreements fully liberalized an additional 28 percent of global trade, bringing the level of global duty-free imports to 70 percent. Only 5.5 percent of global imports are subject to positive tariffs under PTAs, of which one-fifth receive no preferences at all (Figure 1.7), reducing the overall trade-weighted average tariff from 5.0 to 2.7 percent. 53 Handbook of Deep Trade Agreements In 2016, preferential trade agreements fully liberalized an additional 28 percent of global trade, bringing the level of global duty-free imports to 70 percent. Only 5.5 percent of global imports are subject to positive tariffs under PTAs, of which one-fifth receive no preferences at all (Figure 1.7), reducing the overall trade-weighted average tariff from 5.0 to 2.7 percent. Figure 1.7: More than half of the value of global trade took place under an agreement in 2015 Source: Deep Trade Agreements Database. The extent of preferential liberalization varies across countries, but more than two-thirds of countries have reduced trade-weighted average tariffs to less than 5 percent. Multilateral liberalization efforts have been driven mainly by high-income countries.This is reflected on their low preferential trade- weighted applied MFN rates, which are mainly below 5 percent (Figure 1.8). However, preferential liberalization has been widely spread across nations, with developing countries such as Rwanda, Burundi, and Uganda reducing their average preferential trade-weighted rates by 40 percent.9 Figure 1.8: Preferential liberalization has reduced trade-weighted average tariff rates to less than 5 percent for more than two-thirds of countries Source: Deep Trade Agreements Database. 9 Annex Figure 1.A.2. 54 Preferential Tariffs Liberalization efforts through PTAs are taking place across tariff lines, but countries are in general less willing to liberalize higher tariffs. While over three-quarters of tariff lines with MFN rates under 15 percent are fully liberalized, that is the case for only half of the lines with MFN rates over 15 percent. In fact, nearly one-quarter of tariff lines with MFN rates over 15 percent are completely excluded from preferential liberalization (Figure 1.9). Figure 1.9: Preferential agreements have reduced protection across the board but less so where tariffs are high Source: Deep Trade Agreements Database. Tariffs have been reduced across sectors but are still high for agricultural products. Agricultural sectors such as foodstuffs, animal and animal products, and vegetables (all with MFN trade- weighted averages over 15 percent) have seen tariff rates cut by half, but remain relatively high (Figure 1.10). On average, tariff reductions across sectors range between 32 and 62 percent. Figure 1.10: Tariffs have been reduced across sectors but are still high for agricultural products Source: Deep Trade Agreements Database. 55 Handbook of Deep Trade Agreements There is room left for further liberalization, especially in lower-income countries. Low- income and lower-middle-income countries still have trade-weighted preferential tariff levels over 5 percent on average (panel a of Figure 1.11). When preferential tariffs are split by level of development of importing and exporting countries, trade-weighted preferential tariffs imposed by South countries on the North and on the South are, respectively, more than 2.7 times and 2 times higher than those imposed by the North (panel b of Figure 1.11). Figure 1.11: There is room for further liberalization a. Especially in lower-income countries… b. …in their trade with both developing and developed nations Source: Deep Trade Agreements Database. 1.4.2 What did preferences do to tariff peaks? The analysis below focuses on “sensitive products,” defined as the sub-set of tariff lines that are subject to MFN rates above 15 percent.They are products particularly subject to competition and therefore have high tariff rates. Although preferential liberalization has targeted highly protected sectors, there remain pockets of protection in agricultural products, textiles, and footwear. Preferential tariff lines with MFN rates over 15 percent are mostly concentrated in apparel and agroindustry goods. Around half of those tariff lines have been fully liberalized through preferential trade agreements (Figure 1.12). While total liberalization in these industries has been mostly granted by developed nations, developing nations are still reluctant to grant liberalization in multilaterally sensitive products (see Annex Tables 1.A.4 and 1.A.5).This trend is maintained when tariff rates are weighted by a trade partner’s share of global trade at the product level,10 to control for the fact that lower tariffs can be granted on non-traded goods or to non-trading partners (see Annex Figure 1. A.3). 10 We use the following formula to calculate the trade-weighted tariff lines: wTik=TiK * ∑j SXjk , where TiK is the total k number of tariff lines of product k from country i. (TiK=∑j tij ) and SXjk is the share of country j of global exports of k ∑ X product k ( SXjk = i ij k ). ∑j∑iXij 56 Preferential Tariffs Figure 1.12: Although preferential liberalization has targeted highly protected sectors (MFN tariffs greater than 15 percent), agricultural products, textiles, and footwear remain pockets of protection Source: Deep Trade Agreements Database. 1.4.3 How big is the preferential advantage? The most common way to measure the advantage given by preferential access is through preference margins. Preference margins are traditionally calculated as the difference between the MFN applied rate and the preferential tariff.11 The average preferential margin in PTAs is low; more than a quarter of world trade is subject to an average preference margin of 7.4 percent. The average preferential margin is low because one-fifth of world trade under preferential agreements is already duty free and a further 2 percent of world trade is not at all liberalized. However, significant margins are applied to the trade that is liberalized under PTAs: the average preference is 7.4 percent for the 28 percent of world trade that is completely liberalized, and 6.4 percent for the remaining 3 percent that is partially liberalized (Table 1.3).12 1.4.4 How are preference margins distributed? Of the 31 percent of global trade subject to positive preference margins, 16 percent is subject to preferences below 5 percent, 10 percent to preferences between 5 and 10 percent, and 5 percent to preference margins over 10 percent (Figure 1.13). 11 MFN j MFN Traditional preference j margin = Tk,i - Tk,i , where Tk,i is the MFN rate applied by country k on product i and Tk,i is the preferential rate applied to country j. 12 The preferential margin is significantly larger if MFN bound rates instead of applied rates are used as a point of reference. The average preferences are, on average, 17.4 percent for the 28 percent of world trade that is completely liberalized, and 13.6 percent for 3 percent that is partially liberalized. 57 Handbook of Deep Trade Agreements Table 1.3: More than a quarter of world trade is subject to an average preference margin of 7.4 percent Share of global Avg. Bound Avg. Applied Avg. Applied Type of regime imports (%) MFN rate MFN rate Preferential Rate Trade not MFN rate > 0 21 27.9 9.9 covered by Zero-MFN rate 21 10.5 0.0 an Agreement Trade covered Zero-MFN rate 21 13.4 0.0 0.0 by an Total Liberalization 28 24.8 7.4 0.0 Agreement Partial Liberalization 3 27.7 14.1 7.7 No Liberalization 2 34.5 15.1 15.1 Source: Deep Trade Agreements Database. Figure 1.13: Distribution of preference margins Source: Deep Trade Agreements Database. Preferential margins vary significantly across economic sectors. Preferential liberalization efforts have been significant for sectors such as agroindustry and apparel, where initial trade-weighted MFN rates were above 10 percent. Over 45 percent of preferential trade in animals and animal products, foodstuffs, and textiles was subject to preferential margins over 10 percent (62, 47, and 46 percent, respectively). On the other hand, sectors such as machinery/electrical, transportation and raw hides, skins and leather, where initial MFN rates were moderate (between 5 and 10 percent), were mainly subject to preferential margins under 5 percent (see Annex Table 1.A.6). Given the proliferation of PTAs, the advantage conferred by a preferential tariff to a given exporter does not depend only on the difference between the MFN tariff and preferential rate, but also on tariffs faced by competing suppliers from other countries in the same market. The concept of competition-adjusted preference margins accounts for these other factors.13 Competition-adjusted preference margins are calculated as the percentage difference between the weighted average tariff 13 Low et al. 2009. 58 Preferential Tariffs rate applied to the rest of the world and the preferential rate applied to the beneficiary country, where weights are represented by trade shares in the preference-granting market.14 Unlike a traditional preference margin, a competition-adjusted preference margin can assume positive as well as negative values. A negative value indicates that in a specific market, a certain country faces worse market conditions than its trade competitors. In terms of competition-adjusted preference margins, relatively small shares of world trade receive a significant preferential advantage or suffer a significant preferential disadvantage. Specifically, only 5.2 of global trade benefited from a preferential advantage over 5 percent, and only 3.3 percent of global trade suffered from a preferential disadvantage higher than 5 percent (Figure 1.14). Figure 1.14: Most countries benefited from a competition-adjusted margin between -2 and 2 percent Source: Deep Trade Agreements Database. Lower-income countries tend to benefit the most from preferential access, with competition- adjusted margins over 3 percent for their exported products. About 15 percent of countries benefit from competition-adjusted margins of over 2 percent, while 84 percent of competition-adjusted preference margins are concentrated within the range of -2 percent and +2 percent (Figure 1.15). Nepal, Lesotho, and Afghanistan receive positive preferential margins of 8.9, 9.2, and 10.5 percent, respectively, whereas Cuba, American Samoa, and the Maldives pay 4 percent higher tariffs on their exports than the competition-adjusted levels.15 14 Competition-adjusted preference margin for product i granted to partner j by country = ∑ v X vk,i T k,i is the export-weighted (X in the formula denotes exports w j w v k = CAPM jk,i = T k,i - T k,i where T k,i ∑v Xvk,i of v into k) average tariff imposed by country k on all other exporting countries v (excluding country j) with j respect to product i. The preferential rate applied to country j is T k,i. 15 A similar result is obtained when import demand elasticities are also used as weights to aggregate preferential margins across products (see Annex Figure 1.A.4), in order to account for the fact that imports of some goods can be more responsive to changes in prices than others. See Nicita and Hoekman 2008. 59 Handbook of Deep Trade Agreements Figure 1.15: Lower-income countries tend to benefit the most in terms of competitive-adjusted margin Source: Deep Trade Agreements Database. 1.5. FROM PREFERENCES IN PRINCIPLE TO PREFERENCES IN PRACTICE So far, the analysis has been based on the preferential tariff rates that would in principle be levied on imports. However, not all imported products from preference-receiving sources are automatically eligible for preferential duties. If, for instance, a particular product does not comply with the origin rules specified in an agreement between two countries, its imports will be subject to the higher MFN duty. Preference utilization rates are defined at the HS6 level as the share of total imports in a specific category that enter a country under preferences, divided by the total imports from that source in the relevant category.16 In this section, we illustrate the extent of preference utilization, focusing on the European Union’s preferential trade.17 More than 80 percent of preferences granted by the EU were fully utilized in 2016. More than 70 percent of exports from least-developed countries to the European Union were eligible for preferences. In 2016, the rate of utilization of the duty-free preferential advantage 16 Note that the denominator of the utilization rate excludes all trade under zero MFN rates, and all trade in products under non-zero MFN rates for which no tariff preference is available. 17 Data on utilization rates come from Eurostat. 60 Preferential Tariffs provided by the “Everything But Arms” arrangement18 was equal to 94 percent. The share of exports from developing and developed countries which were eligible for preferences through non-reciprocal (GSP)19 as well as reciprocal (GPS+) agreements with the EU is much lower— equal to 18 and 16 percent, respectively. The rate of utilization of such preferences is still high, at above 80 percent (Figure 1.16). Figure 1.16: EU imports by tariff regime and country group (US$ million) Source: Statistics from Eurostat 2016. Note: (*) Excluding EU countries. Preference utilization rates vary widely across countries. Countries such as Bangladesh represent more than 60 percent of preferential trade from LDC countries to the EU, and have rates of utilization above 90 percent (see Figure 1.17). In contrast, countries such as Chad and Guinea- Bissau rarely use preferences provided through the EU’s EBA. Developing countries such as Sri Lanka used GSP preferences for only 55 percent of their eligible exports.A key explanation of the low utilization rates is restrictive rules of origin as well as the related administrative burden. In fact, 11 percent of Sri Lankan firms, interviewed in an ITC survey20 on non-tariff measures in 2011, considered rules of origin a recurrent problem. At the sector level, agricultural imports tend to have higher utilization rates than manufacturing and natural resources imports. Manufacturing sectors with the highest utilization rates are apparel (textiles, clothing, and leather) and wood and paper. The biggest import sector in terms of trade eligible for preferences is clothing. In 2016, the total amount of EU imports of clothing that were eligible for preferences amounted to US$56.5 billion.The rate of utilization of such preferences, with an average preference margin of 10 percent, was 85 percent.The sector with the highest utilization rate is dairy products.This is also the sector with the highest preference margin (Figure 1.18). 18 The EBA agreement allows all LDC-originating products except arms and ammunition to enter the EU market duty free. 19 General System of Preferences. 20 ITC 2011. 61 Handbook of Deep Trade Agreements Figure 1.17: Utilization of EU preferences by beneficiary countries Source: ITC’s calculation based on EUROSTAT 2016. (*) Utilization rate = Ratio (Exports that entered the EU under preferences; Total Exports eligible for preferences) Figure 1.18: Utilization rates vs. preferential margin Source: Statistics from EUROSTAT and Market Access Map database. Note: Product groups are based on multilateral trade negotiations categories (ITC 2017). 62 Preferential Tariffs Common reasons for tariff preferences not being fully utilized include small preferential margins, small shipment amounts, time sensitivity for certain goods, and transaction costs (lack of information, administrative burden). ITC business surveys on non-tariff barriers21 identified rules of origin, including origin certification, as one of the most common obstacles to trade perceived by SMEs in developing countries. Rules of origin are perceived to be burdensome more often in industrial sectors than in agriculture - 35 percent compared to 11 percent of all complaints. Most of the complaints are related not to the restrictiveness of the rules of origin per se, but rather to the procedural obstacles related to obtaining proof of origin. Typical procedural obstacles include delays in obtaining a certificate of origin, unusually high fees, the large number of required documents, numerous administrative windows involved, and mismatch between published information and actual requirements on the ground.22 Recent surveys have also identified lack of knowledge and awareness by businesses as one reason for the lack of utilization of preferences granted in PTAs.23 1.6. CONCLUSIONS MFN tariffs have progressively fallen since the establishment of the General Agreement on Tariffs and Trade (GATT) in 1948. Unilateral liberalization and eight rounds of multilateral trade negotiations have significantly reduced tariffs applied by WTO members over time from levels between 12.5 and 15 percent in 1995 to lower than 10 percent in 2015. In addition, countries around the world have increased their participation in PTAs, especially in the last two decades. From the 1990s onwards, the number of PTAs has almost quadrupled, from around 50 to close to 280 PTAs presently in force. Lack of progress in multilateral negotiations in recent years, among other reasons, has spurred tariff reductions through bilateral and regional preferential trade agreements. Three main findings emerge from this chapter on the significance of tariff preferences in a context of decreasing MFN-applied tariffs and PTA proliferation. First, preferential trade agreements, which now cover more than half of world trade, have significantly widened the scope of tariff-free trade.Whereas 42 percent of the total value of trade was traded free under MFN rates in 2016, PTAs have fully liberalized an additional 28 percent of global trade. In fact, only 5 percent of global imports are subject to positive tariffs under PTAs. 21 ITC 2015. 22 Specific examples include rejections in certain Arabic countries of certificates of origin qualifying under the Pan-Euro-Med origin protocol due to customs officers’ lack of knowledge; rejections due to minor mistakes in the certificate or in the documentary evidence; or the requirement of full translation, including of all technical terms. 23 Global Trade Management Survey 2015, 2016; PwC Australia 2018; Holmes and Jacob 2018. 63 Handbook of Deep Trade Agreements Second, the extent of preferential liberalization varies across countries and sectors. Around 70 percent of countries participating in PTAs have reduced trade-weighted average preferential tariffs to less than 5 percent, but there remain pockets of protection. Several lower-income countries still have trade-weighted average tariffs above 5 percent. And even PTAs have not been able to eliminate the high levels of protection for agricultural products, textiles, and footwear. Third, because one-fifth of world trade under preferential agreements is already duty free and another 2 percent has not been liberalized at all, more than a quarter of world trade is subject to an average preference margin of 7.4 percent. Once competition from both preferential and non-preferential sources is considered, however, only 5.2 percent of global exports benefited from a preferential advantage of over 5 percent, and only 3.3 percent of global exports suffered from a preferential disadvantage higher than 5 percent. These findings are based on potentially applied tariffs. In practice, preferential duties are not granted automatically to all potentially eligible products. An assessment of the scope of preference utilization for the sub-sample of EU imports from its trading partners suggests that the rate of utilization of preferences varies across countries and products. Key factors explaining low utilization rates include rules of origin as well as the related administrative burden and lack of knowledge of import and export processes. The stylized facts on the patterns and extent of preferential liberalization presented in this chapter provide the basis for future research on the implications and determinants of preferential tariffs. The relatively small extent of preference margins also suggests reasons for entering into PTAs beyond only preferential tariffs. ACKNOWLEDGMENTS We would like to thank Caroline Freund, Nuno Limao, and participants in the World Bank conference on the “Evolution of Deep Trade Agreements” for their comments.  64 Preferential Tariffs REFERENCES Holmes, P., and N. Jacob. 2018. Certificates and rules of origin:The Bagwell, K., and R. W. Staiger. 1999. An economic theory of experience of UK firms. Brighton: UK Trade Policy Observatory. GATT. American Economic Review 89(1): 215-248. ITC (International Trade Centre). 2006. World tariff profiles. Baier, S.L., and J.H. Bergstrand. 2004. Economic determinants Geneva: ITC. of free trade agreements. Journal of International Economics 64(1): 29-63. ———. 2011. Sri Lanka: Company perspectives. ITC Series on Non-tariff Measures (December 13). Baldwin, R., and D. Jaimovich. 2012. Are free trade agreements contagious? Journal of International Economics 88(1): 1-16. ———. 2015. The invisible barriers to trade: How businesses experience non-tariff measures. Geneva: ITC. Bhagwati, J. 2008. Termites in the trading system: How preferential agreements undermine free trade. Oxford: Oxford University ———. 2017. World tariff profiles. Geneva: ITC. Press. ———. No date. Market access map methodology user guide. Blanchard, E., and X. Matschke. 2010. U.S. multinationals and https://www.macmap.org/. preferential market access. Research Papers in Economics 8/10. http://repec.org/. Johnson, H. G. 1954. Increasing productivity, income-price trends Blanchard, E. J., C.P. Bown, and R.C. Johnson. 2016. and the trade balance. Economic Journal 64(255): 462-85. Global supply chains and trade policy. NBER Working Paper 21883. Cambridge, MA: National Bureau of Low, P., R. Piermartini, and J. Richtering. 2009. Multilateral Economic Research. solutions to the erosion of nonreciprocal preferences in nonagricultural market access. In Trade preference erosion: Broda, C., N. Limao, and D.E. Weinstein. 2008. Optimal tariffs Measurement and policy response, edited by B. Hoekman, W. and market power: The evidence. American Economic Review Martin, and C. Primo Braga. Washington, DC, and Basingstoke, 98(5): 2032-65. UK: World Bank and Palgrave Macmillan. Capling, A., and P. Low (eds.). 2010. Governments, non-state actors Magee, C. S. 2003. Endogenous preferential trade agreements: and trade policy-making: Negotiating preferentially or multilaterally? An empirical analysis. Contributions in Economic Analysis & Cambridge: Cambridge University Press. Policy 2(1). De Scitovszky, T. 1942. A reconsideration of the theory of Malthus, Thomas R. The grounds of an opinion on the policy tariffs. Review of Economic Studies 9(2): 89-110. of restricting the importation of foreign corn. Edgeworth, F.Y. 1994. Theory of international values. Economic Mill, J. S. 1844. Of the influence of consumption on production. Journal 4(16): 606-38. In Some unsettled questions of political economy. Eurostat database. 2016. https://ec.europa.eu/eurostat/data/database. PwC Australia. 2018. Free Trade Agreement Utilization Study. Fugazza, M., and A. Nicita. 2019. The value of preferential Torrens, R. 1815. Letters on commercial policy. London, 1833. market access. UNCTAD Blue Series on Policy Issues in International Trade and Commodities. World Bank. 2016. PTA database. Global Trade Management Survey. 2015. KPGM and Thomson World Trade Organization. 2011. The WTO and the preferential Reuters. trade agreements: From co-existence to coherence. Geneva: World Trade Organization. Grossman, G. M., and E. Helpman. 1995.Technology and trade. Handbook of International Economics 3: 1279-1337. Hofmann, C., A. Osnago, and M. Ruta. 2017. Horizontal depth: A new database on the content of preferential trade agreements. Policy Research Working Paper 7981, World Bank, Washington, DC. 65 Handbook of Deep Trade Agreements ANNEX Annex Figure 1.A.1: Reduction in trade-weighted tariffs is uniform across sectors 66 Preferential Tariffs Annex Figure 1.A.2: On average countries had reduced tariffs by half Source: Deep Trade Agreements Database. Annex Figure 1.A.3: Share of sector tariff lines weighted by partner’s share of global trade Source: Deep Trade Agreements Database. Note: (i) We use the following formula to calculate the trade-weighted tariff lines: wTik = TiK * ∑j SXjk , where TiK k is the total number of tariff lines of product k k from country i. (TiK = ∑j tij) and SXjk is the share of country j of ∑i Xij global exports of product k ( SXjk = ∑ ∑ X k ). j i ij 67 Handbook of Deep Trade Agreements Annex Figure 1.A.4: Competition-adjusted preference margin using import demand elasticities Source: Deep Trade Agreements Database. Note: Competition-adjusted preferential margins measuring the advantage that exports of country j have in ∑k ∑i Xjk,i ( ∑k X∑ jk,i k,i CAPMjk,i ) exporting its goods is calculated as: k Xjk,i k,i ,where CAPMjk,i is the competition-adjusted ∑k ∑i X jk,i preference margin for product i granted to partner j by country k. k,i is an estimate of the price elasticity of demand for an import. Weighted by the trade share of the country concerned and by total exports of country j. Annex Table 1.A.1: Agreements with partial information Agreement Missing Country Armenia - Turkmenistan Turkmenistan CIS Turkmenistan COMESA South Sudan EC-Faroe Islands Faroe Islands EU - Andorra Andorra EU-San Marino San Marino Faroe Islands - Norway Faroe Islands Faroe Islands - Switzerland Faroe Islands Georgia - Turkmenistan Turkmenistan Iceland - Faroe Islands Faroe Islands Pacific Island Countries Trade Agreement Faroe Islands Panama - Taiwan, China Taiwan, China Russian Federation - Turkmenistan Turkmenistan Ukraine-Turkmenistan Turkmenistan Source: Deep Trade Agreements Database. 68 Preferential Tariffs Annex Table 1.A.2: Comparison of available data and entry into force of last agreement Country Available data Entry in force of Country Available data Entry in force of last agreement last agreement Afghanistan 2013 2011 Mayotte 2013 Barbados 2013 2008 Micronesia, 2006 2003 Fed. States Equatorial Guinea 2007 1999 Panama 2013 2014 Eritrea 2006 1994 Papua New Guinea 2010 2009 Gambia,The 2012 1993 Sierra Leone 2006 1993 Iran, Islamic Rep. 2011 Suriname 2007 2008 Jamaica 2011 2008 Syrian Arab Republic 2013 2007 Kiribati 2006 2003 Trinidad and Tobago 2008 2008 Libya 2006 1998 Zambia 2013 2000 Source: Deep Trade Agreements Database. Annex Table 1.A.3: Countries with MFN information in at least 15 years between 1995 and 2015 Country Missing Country Missing Country Missing Argentina N/A El Salvador N/A Paraguay N/A Bolivia N/A Guatemala 1996 Peru 1996, 2012 Brazil N/A Japan N/A Singapore 2004 Canada N/A Korea, Rep. N/A Switzerland N/A Central 1996, 1998-2000, Madagascar 1999 Thailand 1996-1998, African Rep. 2014 2002, 2012 Chile 2014 Mauritius 2003 Tunisia 1996 – 1997, 1999, 2001, 2007, 2014 Colombia N/A Mexico N/A Turkey 2012, 2014 Ecuador 2013 Nicaragua N/A United States N/A Egypt, Arab Rep. 1996-1997 Norway N/A Uruguay 2003 Source: Deep Trade Agreements Database. 69 Handbook of Deep Trade Agreements Annex Table 1.A.4: Share of sectoral and development-level tariff lines (multilaterally sensitive preferentially free) Sector South-South South-North North-South North-North Animal & animal products 3.10% 0.83% 11.75% 13.06% Vegetable products 2.45% 1.00% 5.42% 6.67% Foodstuffs 3.49% 1.24% 19.37% 22.24% Mineral products 0.37% 0.27% 0.02% 0.26% Chemicals & allied industries 0.50% 0.47% 0.20% 0.27% Plastics / rubbers 1.29% 1.10% 0.03% 0.54% Raw hides, skins, leather, & furs 1.93% 1.30% 0.64% 1.96% Wood & wood products 1.49% 1.24% 0.17% 0.36% Textiles 2.61% 1.16% 0.80% 2.38% Footwear / headgear 3.69% 1.77% 10.10% 14.45% Stone / glass 2.28% 1.19% 0.09% 1.70% Metals 1.24% 0.91% 0.03% 0.53% Machinery / electrical 0.76% 0.50% 0.01% 0.33% Transportation 1.10% 0.72% 1.38% 1.84% Miscellaneous 2.14% 0.83% 0.13% 1.12% Source: Deep Trade Agreements Database. Annex Table 1.A.5: Share of sectoral and development-level tariff lines (excluded) Sector South-South South-North North-South North-North Animal & animal products 4.55% 7.25% 1.29% 5.26% Vegetable products 3.93% 4.82% 0.97% 4.38% Foodstuffs 5.74% 7.58% 1.10% 5.89% Mineral products 0.76% 0.42% 0.06% 0.90% Chemicals & allied industries 1.06% 0.69% 0.07% 0.93% Plastics / rubbers 2.75% 2.03% 0.06% 0.96% Raw hides, skins, leather, & furs 3.39% 2.76% 0.06% 1.48% Wood & wood products 3.05% 2.37% 0.05% 0.77% Textiles 5.10% 3.33% 0.02% 0.30% Footwear / headgear 6.19% 4.73% 0.07% 1.41% Stone / glass 3.93% 2.91% 0.04% 0.64% Metals 2.46% 1.45% 0.07% 1.01% Machinery / electrical 1.43% 1.04% 0.06% 0.93% Transportation 2.25% 2.19% 0.06% 0.99% Miscellaneous 3.64% 3.28% 0.05% 0.83% Source: Deep Trade Agreements Database. 70 Preferential Tariffs Annex Table 1.A.6: Share of preferential trade by preferential margin and MFN range MFN range (%) Less than 5 Between 5 and 10 Over 10 Preferential Margin (%) None Less than 5 None Less than 5 5-10 None Less than 5 5-10 Over 10 Animal & animal products 3.92 8.48 2.04 2.61 10.22 4.81 2.61 3.47 61.83 Vegetable products 1.68 17.98 6.67 2.96 28.05 9.22 2.20 2.67 28.58 Foodstuffs 1.07 12.78 2.40 3.17 21.46 5.19 4.37 2.31 47.26 Mineral products 7.28 74.07 2.70 0.51 11.42 1.45 0.54 1.03 1.00 Chemicals & allied industries 3.38 23.77 7.63 3.54 57.12 0.95 0.19 0.22 3.20 Plastics / rubbers 1.37 27.59 4.98 4.03 55.09 1.39 0.29 0.34 4.91 Raw hides, skins, leather, &furs 0.77 37.86 4.35 11.02 36.52 1.23 0.72 0.68 6.83 Wood & wood products 1.25 33.48 8.77 5.64 36.62 1.72 0.34 0.30 11.88 Textiles 0.27 9.26 3.13 4.27 24.12 6.09 6.05 1.08 45.73 Footwear / headgear 0.05 14.06 1.36 10.71 32.77 3.58 8.45 2.73 26.29 Stone / glass 0.19 48.48 7.90 3.90 29.11 1.60 0.58 0.51 7.73 Metals 1.78 47.24 7.82 3.73 31.19 1.59 0.37 0.52 5.76 Machinery / electrical 1.98 63.39 4.75 2.09 12.25 0.95 0.24 0.36 14.00 Transportation 2.33 45.57 1.84 0.93 33.44 2.06 0.80 0.74 12.29 Miscellaneous 1.11 48.48 3.84 8.94 14.60 1.20 1.01 0.72 9.20 Source: Deep Trade Agreements Database. Annex Table 1.A.7: Share of preferential trade by preferential margin and MFN range, by importing country group MFN range (%) Less than 5 Between 5 and 10 Over 10 Preferential Margin (%) None Less than 5 None Less than 5 5-10 None Less than 5 5-10 Over 10 Low income 0.12 0.71 23.96 5.78 16.67 24.67 2.86 2.50 22.72 Lower middle income 6.06 12.43 25.60 5.81 21.39 11.40 1.83 3.03 12.45 Upper middle income 4.27 23.81 6.54 6.23 23.72 4.33 1.60 1.18 28.33 High income 1.26 51.68 1.03 1.66 29.10 0.63 1.08 0.53 13.02 Source: Deep Trade Agreements Database. 71 Handbook of Deep Trade Agreements 72 Preference Utilization CHAPTER 2 Preference Utilization J. Richtering and T. Verbeet 73 Handbook of Deep Trade Agreements CHAPTER 2 Preference Utilization J. Richtering and T. Verbeet World Trade Organization, Geneva, Switzerland CONTENTS 2.1. INTRODUCTION 75 2.2. DEFINITION OF PREFERENCE UTILIZATION AND MEASUREMENT METHODOLOGIES 75 2.3. DATA SOURCES AND DATA LIMITATIONS 77 2.4. OVERVIEW OF PRELIMINARY RESULTS 78 2.5. CONCLUSIONS 83 ACKNOWLEDGMENTS 84 REFERENCES 85 74 Preference Utilization 2.1. INTRODUCTION International global trade in goods exceeded USD 19 trillion in 2018. The increase in trade volumes and values over the last three decades has been accompanied by an increasing proliferation of preferential trade arrangements (PTAs). Many of those agreements are reciprocal, offering negotiated tariff reductions and other trade-facilitating measures for both imports and exports, while others grant unilateral preferences to a specific group of countries, often to pursue a development objective through trade policy.The impact of tariff reductions on bilateral trade flows merits closer examination, as little is known about the degree to which trade is actually benefiting from preferential trade arrangements. The PTA Transparency Mechanism, established in December 2010 (WTO WT/L/806), requires WTO members granting non- reciprocal preferential schemes to notify the relevant trade statistics. These notifications have enabled the WTO to construct a database on preference utilization.1 Using this unique WTO database on preference utilization, this chapter gives an overview on the use of preferences by least-developed countries (LDCs). It starts with a short introduction defining preference utilization and the measurement methodologies, then proceeds in section 2.2 with an analysis on preference utilization by LDCs. These countries benefit from both (a) developed countries’ preferential duty schemes accorded under the Generalized System of Preferences (GSP);2 and (b) preferential market access by developing WTO Members authorized through a Waiver Decision (WT/L/759).3 Section 2.3 lists data sources and highlights data limitations. Section 2.4 provides new insights on the utilization of preferential trade agreements, based on import data for developed as well as key developing countries. We conclude that the failure to use preferential market access for products exported by LDCs increases the cost of these exports by hundreds of USD millions as a result of customs duties paid. 2.2. DEFINITION OF PREFERENCE UTILIZATION AND MEASUREMENT METHODOLOGIES A preferential tariff rate is a customs duty that is lower than the most-favored-nation (MFN) rate. It is granted by an importing country for a product, defined at the tariff line level, that originates from a specific exporter or group of exporters. Such a preference could be granted on a reciprocal basis, as in a regional trade agreement such as the “North American Free 1 Relevant information notified to the WTO Secretariat with regard to the PTA Transparency Mechanism can be found on the PTA database (http://ptadb.wto.org). 2 PTAs falling under paragraph 2 of the Decision of 28 November 1979 on Differential and More Favourable Treatment Reciprocity and Fuller Participation of Developing Countries (“Enabling Clause”), with the exception of regional trade agreements under paragraph 2(c) as described in the General Council Decision of 14 December 2006 (Transparency Mechanism for Regional Trade Agreements). 3 Initially adopted by Members in 1999 (WT/L/304), and extended in 2009 (through WT/L/759). 75 Handbook of Deep Trade Agreements Trade Agreement” (NAFTA) or a customs union such as the EU, or on a non-reciprocal basis, such as a Generalized System of Preferences (GSP). In the following, the focus will be on non-reciprocal preferences and their use by preference-receiving countries. Not all imports are subject to preferential tariff treatment. Tariff preferences can only be granted for tariff lines where the MFN duty is not zero. Furthermore, certain products may not be eligible to receive preferential treatment, as preferential regimes often do not cover all products exported from a preference-receiving country. Exclusion lists may contain only a few items (such as many LDC schemes), but can also show a large number of products (such as some developed country’s GSPs). Finally, some import transactions which are covered by a preferential regime may not benefit from a preferential rate for reasons linked to the fulfillment of certain requirements defined in the preferential agreement. For example, the shipment of a certain product may not have met certain rules of origin requirements or the exporter chooses not to use the preferential rate. We can illustrate the different forms of bilateral import flows (at the tariff line level) by assigning them to the following mutually exclusive types: A, B, C, or D (Figure 2.1). Figure 2.1: Categorization of preferential imports by duty schemes To determine the economic value of preferences to beneficiary countries, these preferences have to be used.“Preference utilization” can narrowly be defined as the degree to which imports that are eligible for preferences enter under these preferential rates. More broadly, one can look at preference utilization as the overall benefit that preferences bring to the exporting country. There are different concepts that can be used when measuring “preference utilization.” The most common measurement, and the one used for this analysis, is a ratio based on import value: imports that have reportedly benefited from a specific preferential duty scheme (C in Figure 2.1) in comparison to imports on all tariff lines eligible for preferential duty treatment (C+D in Figure 2.1). Other measures analyze the use of preferences in the context of overall bilateral imports, referred to as the “utility ratio.” However, it is often considered more appropriate to exclude 76 Preference Utilization MFN duty-free tariff lines from the calculation, as they do not offer any preference with regard to other exporters and would offer little insight into the utilization rates of preferences. A further option is to calculate preference utilization by examining actual customs duties paid.4 The interpretation of preference utilization ratios needs to take into account that some preference-granting WTO Members allow products from the same beneficiary country to enter on different (“overlapping”) preferential duties schemes.The following analysis separates non-utilized preferential imports into those benefiting from other preferential schemes and those that face MFN duties. The analysis of “forgone tariff duties” at the end of Section 2.4 estimates the sum of duties paid on products that are eligible for preferential duty treatment.5 2.3. DATA SOURCES AND DATA LIMITATIONS Information and data used in the analysis arise from notifications submitted by preference- granting WTO Members pursuant to the requirements of the Transparency Mechanism for Preferential Trade Arrangements (WT/L/806) established by General Council Decision on December 14, 2010.The data on preferential schemes have been notified under the “Enabling Clause” for developed country Members6 and the 1999 waiver for developing countries.7 Accordingly, WTO Members must notify, on an annual basis, tariff lines of non-reciprocal preferential duty schemes and report the corresponding value of imports by partner and duty scheme used. There are a number of data limitations which may affect the calculation of utilization rates. A specific product can sometimes benefit from two or more preferential duty schemes; for example, preferential market access can be granted through the GSP-LDC scheme and also through another PTA. The United States, for example, also grants preferential market access to many African LDCs through the “African Growth and Opportunity Act” (AGOA).8 For a global analysis on preferential market access of products exported by LDCs, we consider first the utilization rate of a particular GSP-LDC scheme, and complement with an additional category of “other non-reciprocal preferential duty schemes” from which an LDC beneficiary country can benefit. 4 See Keck and Lendle 2012. 5 This is similar to the “tariff exemption ratio” introduced by Hayakawa et al. 2018. 6 Paragraph 2 of the Decision of November 28, 1979, on Differential and More Favourable Treatment Reciprocity and Fuller Participation of Developing Members. 7 Waiver Decision initially adopted by Members in 1999 (WT/L/304), and extended until June 30, 2019 (WT/L/759). 8 A more detailed analysis on the EU and US preferential duty schemes is offered by Davies and Nilsson (2013). 77 Handbook of Deep Trade Agreements The analysis is based on aggregated annual averages and could conceal differences in preference utilization for some products, sectors, time periods, or beneficiary countries. Further, some notified preference utilization data are based on “requested” or “claimed” customs declarations which might have been rejected after customs clearance. In these cases, figures for preferential trade may be slightly overestimated. WTO Members listed in Table 2.1 have GSP-LDC or LDC-specific duty schemes in force, some for many years. The analysis of preference utilization for LDCs takes into account the latest year for which detailed import data are available, which for most countries is 2016; for Chile, China, and India, the reference year is 2015. Table 2.1: List of PTAs included in the analysis Provider PTA type In force since Australia GSP-LDC 01-Jan-74 Canada GSP-LDC 01-Jul-74 European Union GSP-LDC 01-Jul-71 Japan GSP-LDC 01-Aug-71 Norway GSP-LDC 01-Oct-71 Switzerland GSP-LDC 01-Mar-72 United States GSP-LDC 01-Jan-76 Chile LDC-specific 28-Feb-14 9 China LDC-specific 01-Jul-10 India LDC-specific 13-Aug-08 Korea, Rep. LDC-specific 01-Jan-00 Taiwan, China LDC-specific 17-Dec-03 Source: Preferential Trade Arrangements database (http://ptadb.wto.org). 2.4. OVERVIEW OF PRELIMINARY RESULTS Complementing a previous World Bank analysis,10 data provided by Australia, Canada, European Union, Japan and the United States allows to measure the actual extent of preferential trade. The import data detail which products have been imported and registered under the respective available duty schemes. Figure 2.2 provides an indication of the relative share of imports by duty schemes used. 9 Import data for China has been submitted to the WTO Secretariat, and complemented by LDC preferential duties sourced from ITC. 10 Espitia et al. 2018. 78 Preference Utilization Figure 2.2: Total imports by duty categories used, 2016 Source: WTO Integrated Database 2018. Despite the large proliferation of regional and bilateral trade agreements in recent years, import data for 2016 show that most trade takes place under WTO MFN rules. With some variation among the reporting countries, around 20 to 30 percent of imported goods are subject to MFN duties; the majority of trade in terms of value enters MFN duty free. Across all five economies, MFN dutiable and MFN duty-free imports cover more than 80 percent of all imports. The remaining imports use either preferential duty schemes accorded under regional or bilateral trade agreements, or non-reciprocal preferential trade arrangements such as GSP-LDC. It should be noted that in most cases imports under GSP-LDC schemes are duty free; however, there are also instances where the GSP-LDC rate is only a reduced rate compared to the MFN applied duty rate. Imports from beneficiaries of the GSP-LDC scheme are relatively small in comparison to overall imports (Figure 2.2). However, most of those markets are very important export destinations for products originating in LDCs. Therefore, the following analysis will focus on the utilization of preferences by LDCs when exporting to both developed and developing markets. With regard to the analytical challenge of overlapping preferential duty schemes, an additional category of “other preferential duty schemes” is introduced. Imports for products for which a preferential LDC duty exist are then categorized into imports entering under the LDC duty scheme and those imports entering under any other bilateral duty scheme.11 11 The Kommerskollegium National Board of Trade, Sweden, in cooperation with UNCTAD, presented an analysis of EU bilateral agreements, detailing preference utilization from both an exporter and importer perspective. See Kasteng and Inama 2018. 79 Handbook of Deep Trade Agreements WTO Members adopted, at the 6th Ministerial Conference in Hong Kong SAR, China, in December 2005, a decision on duty-free quota-free (DFQF) market access for LDCs.12 In 2019, the remaining number of dutiable tariff lines in developed markets for LDC products was rather limited. Australia, New Zealand, Norway, and Switzerland already provide full DFQF market access. Canada, Japan, and the European Union exclude a number of products but reach close to 100 percent of duty-free treatment. Furthermore, some developing countries offer significant DFQF market access to LDCs, some reaching close to full coverage. Figure 2.3: Imports from least-developed countries by duty categories, 2016 (or latest available year) Source: WTO Integrated Database 2019. Figure 2.3 presents import shares of products from LDCs when entering beneficiary-granting WTO Member countries. For developed country markets, it shows that the European Union is the most open market: 28 percent of LDC products exported to the European Union enter MFN duty free, more than 65 percent make use of LDC preferential market access, and some 2 percent of imports from LDCs enter under another preferential duty scheme. Across developed country markets, around 5 to 15 percent of imports from LDCs face MFN duties despite being eligible for preferential duty treatment. Only the United States shows a much larger share of other preferential duty schemes used by LDCs due to overlapping preferential 12 Annex F, WT/MIN(05)/DEC. 80 Preference Utilization duty regimes such as the African Growth and Opportunity Act (AGOA) and the Caribbean Basin Trade Partnership (CBPTA). In spite of these other preferential schemes, around 40 percent of LDC products enter the US market paying MFN duties. This is largely due to a higher number of products excluded from the GSP-LDC scheme, and a strong commercial presence of Asian LDC clothing exporters in the US market. Developing countries that offer preferential market access to LDCs present a more diverse picture with regard to the utilization of these duty schemes. Both Taiwan, China, and China offer MFN duty-free treatment to a majority of products exported by LDCs. The Republic of Korea offers significant market access through the LDC or other preferential duty scheme; however, a substantial share of products has no preference at all and faces MFN duties. Chile offers LDC treatment on nearly all tariff lines; however, half of the LDC imports are not able to utilize the preferences granted and are hence subject to MFN duties. A very similar conclusion can be drawn from the import statistics provided by India, where more than half of all LDC imports are subject to preferential tariff treatment. However, only a minor part of LDC exporters are able to take advantage of the preferential market access. Figure 2.4: Utilization of the LDC duty schemes in developed markets by product and duty categories, 201613 Source: WTO Integrated Database 2019. Figure 2.4 shows aggregated import data across all developed countries providing preferential market access through GSP-LDC schemes (as listed in Table 2.1, i.e., Australia, Canada, European Union, Japan, Norway, Switzerland, and the United States), broken down by product category.14 13 Data from Australia, Canada, European Union, Japan, Norway, Switzerland, and the United States. 14 For more information on Multilateral Trade Negotiations Product Categories and the alignment to the Harmonized System (HS), please consult: http://stat.wto.org/idbdata/MTN_product_classification_e.pdf. 81 Handbook of Deep Trade Agreements The vertical bars, representing sectors, indicate the proportion of LDC imports by duty scheme. For example, in Textiles, not all LDC textiles enter developed countries duty free; a small proportion pays MFN duties, while other textile products enter MFN duty free. The majority of LDC-produced textiles make use of the LDC duty scheme when exported, or use another preferential duty scheme. All product categories—with the exception of cotton—show a red bar, which depicts the proportion of imports that paid MFN duties irrespective of being eligible for preferential treatment. Sectors that show a relatively low utilization of preferences include “sugars and confectionery,” “non-electrical machinery,” “electrical machinery,” “transport equipment,” and “manufactures n.e.s.,” followed by “dairy products,” “textiles,” and other sectors shown in red in Figure 2.4. Figure 2.5: MFN duties paid on GSP-LDC preference-eligible imports by product category, 2016 (USD million, log scale) Source: WTO Integrated Database 2019. 82 Preference Utilization Figures 2.3 and 2.4 both present proportions and shares based on observed import values, but it needs to be noted that some sectors show a significantly higher annual export volume and value than others. LDC exports to beneficiary-granting WTO Members in 2016 are dominated by exports of “petroleum,” “clothing,” and “minerals and metals,” which account for around 80 percent of LDC products exported in terms of value to developed markets analyzed. Figure 2.5 provides an overview of MFN duties paid on LDC products that are eligible for preferential duty treatment. The values on the logarithmic scale present estimates of duties paid in developed country markets in the year 2016.The analysis reveals that some sectors are more prone to non-compliance or non-utilization than others, in particular the “clothing,” “sugars and confectionery,” and “petroleum” sectors. These three sectors are responsible for about 75 percent of all duties paid on products that are eligible for GSP-LDC preferential tariff treatment. Countries most affected overall, across all sectors, include Bangladesh, Cambodia, as well as Angola, Ethiopia, and Myanmar.   In the year 2016, duties of around USD 310 million were paid on products exported by LDCs to developed countries that could have benefited from preferential tariff treatment, most of these entering the European Union. This represents about 5 percent of all LDC imports in the developed markets which have been analyzed. 2.5. CONCLUSIONS Complementing the analysis on the scope of preferential trade agreements, this chapter examines preference utilization of non-reciprocal LDC schemes across various beneficiary- granting WTO Members.The data are sourced from notifications to the WTO, following the establishment of the Transparency Mechanism for Preferential Trade Arrangements in 2010, which enabled the WTO to construct a unique database on preference utilization. The data show that for developed countries, trade on an MFN basis, either MFN dutiable or MFN duty free, constitutes around 80 percent of total imports. The remaining imports benefit from bilateral or regional trade agreements. Only a relatively small share of LDC exports enters developed countries under non-reciprocal duty schemes such as GSP-LDC. Nevertheless, for many developing and least-developed countries, non-reciprocal preferential duty schemes offer important market access opportunities. The analysis on preference utilization reveals that LDC exporters make good use of preferential duty schemes. In developed preference-granting Member countries, close to 90 percent of total imports from LDCs are either MFN duty free or using the GSP-LDC preferential market access, apart from the United States, where other preferential duty 83 Handbook of Deep Trade Agreements schemes are used, in particular AGOA. Also, some developing countries offer significant market access opportunities for products from LDCs. A sectoral analysis across developed countries uncovers large differences in the utilization of preferences. Some sectors seem to allow a higher share of LDC exports to benefit from preferential tariff treatment, whereas other sectors benefit much less. LDC exporters not using preferential tariff treatment were paying MFN customs duties of more than USD 300 million in the year 2016. Helping least-developed countries make better use of preferential duty schemes will reduce those tariff-related trade costs. ACKNOWLEDGMENTS Several colleagues in the WTO Secretariat provided valuable input that greatly improved this chapter. We are immensely thankful for comments from Darlan F. Martí from the WTO Market Access Division, Harish Iyer from the WTO Development Division, and assistance in data preparation by the Integrated Database (IDB) team under the supervision of Adelina Mendoza. 84 Preference Utilization REFERENCES Candau, F., L. Fontagné, and S. Jean. 2004. The utilisation rate of preferences in the EU. Paper presented at the 7th Global Economic Analysis Conference, Washington, DC, June 17-19. Davies, E., and L. Nilsson. 2013.A comparative analysis of EU and US trade preferences for the LDCs and the AGOA beneficiaries. Working Paper 01-2013, EU Commission, Brussels. Espitia Rueda, A., A. Mattoo, M. Mimouni, X. Pichot, and N. Rocha. 2018. How preferential is preferential trade? Policy Research Working Paper WPS 8446,World Bank,Washington, DC. Hayakawa, K., F. Kimura, and N. Laksanapanyakul. 2018. Measuring the usage of preferential tariffs in the world. Review of World Economics 154(4) (November). Kasteng, J., and S. Inama. 2018. The use of the EU’s free trade agreement. Stockholm: UNCTAD–Kommerskollegium. Keck, A., and A. Lendle. 2012. New evidence on preference utilization. WTO Staff Working Paper ERSD-2012-12, World Trade Organization, Geneva. Laird, S. 2012. A review of trade preference schemes for the World’s poorest countries. ICTSD Issue Paper 25, International Centre for Trade and Sustainable Development, Geneva. WTO (World Trade Organization). 2011. The WTO and preferential trade agreements: From co-existence to coherence. Geneva: WTO. ———. 2019. Integrated database. http://tariffdata.wto.org/. 85 Export Restrictions CHAPTER 3 Export Restrictions M. Wu 87 Handbook of Deep Trade Agreements CHAPTER 3 Export Restrictions M. Wu* * Harvard Law School, Cambridge, Massachusetts, United States CONTENTS 3.1. INTRODUCTION 89 3.2. LITERATURE REVIEW 92 3.3. METHODOLOGY 94 3.4. FINDINGS 96 3.4.1. Prevalance 97 3.4.2. Quantitative restrictions on exports 100 3.4.3. Export taxes 103 3.4.4. Export licenses 106 3.4.5. Export certificates of origin and other administrative measures 107 3.5. CONCLUSIONS 108 ACKNOWLEDGMENTS 109 REFERENCES 110 88 Export Restrictions 3.1. INTRODUCTION International trade negotiators have focused heavily on lowering tariffs and other forms of protectionist barriers for imports, but have devoted much less attention to reducing similar barriers for exports. However, both distort trade. With the rise of preferential trade agreements (PTAs), legal disciplines on export restrictions have become more commonplace. Through the World Bank’s Deep Trade Agreements Database, we identified 246 PTAs with some legal provision that impacts the use of export restrictions. This chapter highlights the major findings of this study. At the onset, it may be worth considering why some trade negotiations aim to put in place legal disciplines on export restrictions.The General Agreement on Tariffs and Trade (GATT) already imposes certain limitations on export restrictions. Article XI of the GATT, which governs quantitative restrictions on trade, imposes a general prohibition on the use of export bans, quotas, licenses, and other forms of quantitative restrictions. However, the GATT stipulates a limited set of circumstances when this prohibition does not apply. Furthermore, it preserves the right of members of the World Trade Organization (WTO) to enact export duties, taxes, or other charges on goods destined for export. Finally, the general exceptions outlined in GATT Articles XX and XXI provide yet another set of circumstances when WTO members may be exempt from their obligations on export restrictions. In a sense, the existing system operates not to curtail export restrictions, but simply to transform such restrictions into a simpler, more transparent format. Just as the GATT sought to transform the multiple forms of import restrictions into import tariffs, it also sought to do so with export restrictions. By prohibiting export bans, quotas, and licenses, the GATT forces WTO members to resort to export taxes as the primary form for restricting exports. Reducing these restrictions into a quantifiable format then makes it easier for governments to bargain across products, with the hope of finding mutually beneficial outcomes in which the restrictions are lowered over time. In theory, trade negotiators could have focused on negotiating both lower import tariffs and export taxes. In reality, however, negotiators have focused much more on the latter. From a political economy standpoint, this makes sense, at least historically. Domestic industries cared much more about gaining access to foreign markets. They cared much less about whether foreign governments imposed high export taxes that dissuaded their foreign competitors from exporting competitive goods. With the rise of global value chains, however, export taxes and other forms of export restrictions have taken on added importance. High export taxes can restrict the outflow of certain critical inputs for global production chains, causing supply shortages and increasing 89 Handbook of Deep Trade Agreements costs. Particularly if the export restriction is enacted by a country that is a major supplier of a particular product on the world market, this can create distortions in the available supply across domestic and foreign markets, giving rise to price differences. Further, uncertainty over the future course of a government’s policies on export taxes can affect investment decisions and the design of production chains. Firms may choose to allocate parts of their downstream production chain to domestic producers to avoid the costs and additional uncertainty triggered by the export tax. Cognizant of this possibility, some governments have focused on export taxes and other forms of export restrictions as yet another instrument for driving industrial policy. In recent years, several high-profile cases concerning China’s export restrictions on raw materials1 have been litigated before the WTO Dispute Settlement Body. In addition to challenging export policies through litigation, trade policymakers are increasingly seeking greater discipline on export taxes and other forms of export restrictions permissible under existing WTO law. Given the lingering stalemate in multilateral negotiations, the vast majority of additional disciplines have emerged in PTAs. Not surprising, a key motivation for such disciplines is the desire to seek a deeper integration with trading partners, although the nature of the desired integration may differ. In one such scenario, the trading partner has no inherent problem with the industrial policy or other development-related policy of the country that imposes an export tax. Its goal is simply to ensure that its producers are not placed at a competitive disadvantage relative to domestic producers. In other words, it seeks to benefit from, rather than be harmed by, the industrial policy that the trading partner has put in place, via a deep-integration PTA. To understand this scenario, suppose Country A is a dominant supplier of widgets worldwide. It imposes an export tax on widgets, causing a price distortion in world markets. Widgets now cost less in Country A than they do in the rest of the world, including Country B. If the price distortion is large enough and widgets are an important enough input, then this export tax will induce a percentage of downstream producers reliant on widgets to move their production to Country A so as to lower their cost. This includes producers in Country B. By securing a commitment that Country A will not apply export taxes for widgets bound for Country B, the price distortion disappears for producers in Country B. They now have a comparative advantage over producers in the rest of the world, in line with producers in Country A. This saves jobs, because producers in Country B will no longer find it necessary to shift their production to Country A. More importantly, the combination of the export 1 These cases include China – Measures Related to the Exportation of Various Raw Materials (DS394, 395 & 398), China – Measures Related to the Exportation of Rare Earths, Tungsten and Molybdenum (DS431, 432 & 433), and China – Export Duties on Certain Raw Materials (DS508 & 509). 90 Export Restrictions tax plus a PTA guaranteeing no export taxes creates an incentive for producers in the rest of the world to move their production to Country B, so as to secure cheaper widgets. Notice, however, that this objective is achieved only if Country A agrees to a restriction on export taxes that is not applied on a most-favored-nation (MFN) basis. In the example above, Country B seeks the new disciplines on export restrictions in the PTA in order to maximize the benefits its producers will receive from the deeper integration with Country A. It is not seeking to change Country A’s overall policy. By contrast, an alternative scenario is one where Country B may have an ideological problem with the use of export restrictions as a whole. It is seeking to create deeper economic integration among like-minded countries that share its principles. In this instance, Country B may demand that Country A curtail its use of export restrictions as a condition for deeper integration with Country B’s economy. Here, Country B is seeking explicitly to change Country A’s trade practices, and by extension, economic policies altogether, through these new legal disciplines. It may even (but not necessarily) seek to have Country A extend these new obligations on an MFN basis to non-PTA partners. Both of these scenarios arrive at the same endpoint: a deeper integration between the economies of Country A and Country B as a result of the PTA. However, the two scenarios are driven by starkly different motivations. While in the former, countries accept the use of trade-distorting export policies to influence supply chain decisions, and simply seek to deepen their relationship with those advancing such policies, the latter scenario conceives of PTAs as helping to build the foundation for future multilateral rules designed to limit export-oriented protectionist policies. Because this study relies entirely on an analysis of treaty provisions, and because the crafting of such provisions does not allow us to readily identify which motivation is at work, it is difficult to posit the underlying motivations driving the rise in recent years of PTAs with rules on export restrictions. Both forces discussed above are likely at work in this push toward deeper integration, but their relative proportion is not possible to discern on the basis of the work to date. Such an analysis requires work beyond the scope of this study, including interviews with treaty negotiators. This issue is simply noted so that one does not inadvertently draw too expansive a conclusion on the basis of this study. Nevertheless, this study does lead to several interesting findings. They include the following: A substantial proportion of PTAs, well in excess of two-thirds, include some legal discipline on export restrictions and/or export taxes.The inclusion of such provisions dates back to the GATT era. Unlike some of the other elements of PTAs analyzed for this Handbook, which are growing in popularity as economies integrate more deeply, legal disciplines on export restrictions are not a new phenomenon. Rather, they are a classic element of preferential trade agreements. 91 Handbook of Deep Trade Agreements The three most common forms of export restrictions are requirements on: (a) export certificates of origin, (b) export taxes, and (c) quantitative restrictions on exports.2 Not surprisingly, these are also the most common forms of export restrictions for which additional disciplines are created through PTAs. As far as export taxes and certification requirements are concerned, the new legal disciplines found in PTAs are predominantly WTO+ in nature. However, the same is not true of quantitative restrictions on exports. Some PTAs merely incorporate WTO commitments without change, whereas others are WTO+ largely because they limit the scope of applicable exceptions. Unlike some other areas studied in this volume, it is not possible to identify approaches to export restrictions that are specific to a particular trading power. For the most part, there is no discernible American or European template that emerges as the basis for PTA rules. Nor are there apparent differences in regional approaches. However, it is clear that the major users of various forms of export restrictions—such as China, Argentina, Russia, and Vietnam— are among those most resistant to the creation of new rules constraining the use of export restrictions via PTAs. Nevertheless, even they have agreed in some instances to accept new rules in exchange for greater access to certain export markets. As compared to import restrictions, much less work has been done via trade negotiations to curtail protectionism in the form of export restrictions. However, as this chapter will illustrate, to the extent that new rules to address this barrier are being negotiated, this is being done through PTAs. The impact of this effort on global supply chains is still emerging and will continue to be felt in coming years. 3.2. LITERATURE REVIEW Several recent studies have focused on the issue of export restrictions in PTAs. A study conducted by the OECD examined a total of 93 regional trade agreements covering a wide range of geographies.3 The study’s sample included PTAs concluded by six major trading powers (United States, European Union, European Free Trade Area countries, China, Japan, and Canada) as well as a series of regional trade agreements in Africa, Asia, Latin America, and Eastern Europe/Central Asia. The study also included a handful of agreements concluded by developing countries across regions (e.g., India-MERCOSUR). In constructing their sample, the authors of the OECD study aimed for geographic and income-level diversity, with 70 of the 93 PTAs including at least one developing country, and 27 of the 93 PTAs being exclusively among developing countries. 2 Mendez Parra et al. 2016, p. 14. 3 Korinek and Bartos 2012. 92 Export Restrictions The study focused primarily on two forms of export restrictions: (a) quantitative restrictions on exports, and (b) export taxes. It found that more than four-fifths of the sampled PTAs include some language on quantitative restrictions, with the majority of these simply reaffirming the language found in GATT Article XI. Approximately one-sixth of the sampled PTAs include prohibitions on quantitative restrictions on exports which exceed those of the WTO. Somewhat more surprising was the finding that nearly one-quarter of the sampled PTAs contain disciplines on quantitative export restrictions that are weaker than the WTO requirements. In most instances, these WTO- provisions allow quantitative restrictions to be applied on bilateral trade for a wider range of goods than what is stipulated in the GATT. With regard to export taxes, the authors found that more than 70 percent of the sampled PTAs include some disciplines on export taxes. In general, the OECD study concluded that governments are striving to improve upon WTO disciplines on export restrictions in a variety of ways, without any one approach appearing to dominate. Many seek to include greater precision than what is contained in existing WTO rules—by using a list approach, by imposing a time limit, and/or by imposing a cap on the export tax. Others seek to further limit the scope of exceptions that can be applied to justify export restrictions. A second study was commissioned by the European Parliament’s Committee on Development.4 It examined the impact of the export restrictions contained in the EU’s PTAs with least-developed countries (LDCs).Within EU parlance, these are known as the Economic Partnership Agreements (EPAs) with the African, Caribbean, and Pacific (ACP) countries.The authors noted that the EU requirements differ for both export restrictions and export taxes, depending on the agreement. The study found that the enactment of stricter disciplines for export restrictions and export taxes is likely is to have a negative impact on consumer prices in the short term, which may be partially offset in the long term if the restrictions give rise to greater competitiveness. The overall impact, the authors suggested, will depend on the economic size of the country applying them. Tighter disciplines may cause LDCs to increase the rate of extraction of non-renewable resources. Given these findings, the authors recommended that the EU revisit its EPAs with selected ACP countries and consider introducing greater flexibility particularly for LDCs and more room for special and differential treatment, as it already does in some agreements. While there have been several other additional studies examining the trade impact of export taxes and export restrictions, few have examined the subject specifically in the context of PTAs.5 This study is an attempt to fill this gap. 4 Mendez Parra et al. 2016, p.14. 5 Examples of other studies that have analyzed the trade impact of particular export restrictions include Estrades et al. 2017, Laborde et al. 2013, Martin and Anderson 2011, and Solleder 2013. 93 Handbook of Deep Trade Agreements 3.3. METHODOLOGY This study draws from the World Bank’s Deep Trade Agreements Database of 295 PTAs signed through December 2016. Efforts were made to update the content of the PTA if it had undergone any changes during the ratification period or re-negotiations following the initial signing, as was the case for the Comprehensive and Progressive Trans-Pacific Partnership (CPTPP). However, the findings in this study do not reflect PTAs that were signed in 2017 or later (e.g., the EU-Japan Economic Partnership Agreement). A team of researchers at Harvard Law School identified 46 different possible elements of export taxes and other export-related disciplines within PTAs to be coded for this study. The work of identifying these fields and developing this template was an iterative process, whereby a non-random group of selected PTAs across time periods negotiated by a diverse set of governments was examined in batches. During each iteration, relevant provisions were identified. This allowed for the creation of a preliminary template, which was then revised again with the next batch, until the researchers were relatively confident that they had identified the relevant universe of potential elements and created a robust template. Based on this template, each agreement was coded by one researcher and subsequently checked by another. Each of the four researchers involved in the coding had previous knowledge of international trade law as well as practical experience working with the international trade practice of a major law firm. Errors and coding inconsistencies were subsequently resolved by the group as a whole, in consultation with the principal investigator. Altogether, the 46 fields coded can be divided into seven different categories: • Export Quotas and Quantitative Restrictions: Coding for whether the PTA mandates the elimination of all export quotas and quantitative restrictions across parties, or simply prevents the parties from imposing new ones. Also, coding for whether the disciplines take the form of scheduled commitments and/or exceptions. • Export Taxes: Coding for whether the PTA requires the elimination of all export taxes or simply prohibits the imposition of new ones. Also, coding for whether parties are required to schedule export taxes or whether the PTA simply delineates a series of exceptions to the general rule. • Export Price: Coding for whether the PTA prohibits the parties from imposing export price requirements. 94 Export Restrictions • Export Licensing: Coding for whether the PTA extends the application of the import licensing agreement to export licensing. Also, coding for whether the PTA requires the disclosure of certain contact points and procedures related to export licenses. • Administrative Fees and Formalities: Coding for whether the PTA places any restrictions on the forms of administrative fees that can be imposed on exports. Also, coding for whether the PTA imposes any procedural requirements related to administrative fees and formalities. • Export Certification of Origin: Coding for whether the PTA imposes any requirements concerning rules and procedures related to an export certification of origin. For example, certain PTAs might prescribe rules related to the issuance of such certification, or include a list of mandatory documentation. Also, coding for whether there are certain recordkeeping and notification requirements related to certification. • Investment Rules Related to Exports: Coding for whether the PTA’s investment chapter prohibits the imposition of export-related performance requirements. This might include, for example, a requirement that a certain proportion of outputs be allocated to the domestic market, or that the amount of permissible exports be linked to imports and/or investment inflows. • Non-Tariff Measures: Coding for whether the PTA contains a general prohibition on the imposition of non-tariff measures related to exports, or contains a narrower provision explicitly tailored to the port/point of departure of exports. • Agriculture-Specific Measures: Coding for whether the PTA includes particular provisions related to export restrictions on agricultural products, such as the requirement for advance notification of restrictions imposed for food security purposes, and/or for certain forms of export certification as attestation for sanitary and phytosanitary requirements. Beyond simply coding whether or not a particular provision is included in the PTA, the database also includes information about the enforceability of the provision.The researchers highlighted whether the language was binding with state-to-state dispute resolution, binding but without a formal dispute resolution mechanism, a best endeavor provision, or not binding whatsoever. In addition, the database notes whether the coverage of any particular provision applies only to specific sectors, rather than having general application. In very few instances was this limited application found to be the case. 95 Handbook of Deep Trade Agreements Altogether, the coding exercise found that 246 out of the 295 PTAs in the Deep Trade Agreements Database contain at least one provision concerning export taxes or other export restrictions. Given both the number of PTAs and the breadth of the export-related provisions examined, this database is believed to be the most comprehensive to date in terms of documenting the possible range of legal disciplines imposed on exports through PTAs. However, some limitations in the methodology ought to be noted: First, the methodology simply notes whether or not the PTA restricts the use of export taxes, export quotas, or other types export requirements. This is coded on a binary basis, in line with what was agreed upon among various investigators for the World Bank’s Deep Trade Agreements project. No effort was made to examine the scope of goods that are subject to such restrictions as a result of the PTA. Therefore, a PTA that includes a negative list of five products for which export taxes are permissible is coded identically to one that includes 100 products. Additional work would be needed to further differentiate among the various schedules included in the annexes of PTAs beyond what was done in this exercise. Second, the methodology does not take into account the volume of trade affected by a particular provision. For example, the impact of a provision requiring the elimination of all export taxes between two parties to a PTA is quite different if less than 0.01 percent of their bilateral trade is subject to such export taxes as opposed to 5 percent of bilateral trade. However, the information found within the PTA itself does not allow for a determination of the breadth of existing trade impacted by a particular provision. To do so would require additional work examining the precise nature of the taxes, quotas, or other restrictions in place at the time of the PTA’s entry into force. Finally, in the case of export taxes, the database also does not capture the average weighted applied rate for such taxes prior to and after the PTA’s entry into force. In the case of imports, this statistic is often cited to examine the impact of a PTA on reducing import barriers to market access. Similar work could also be done with regard to the impact of a PTA on reducing export taxes. 3.4. FINDINGS Since 2009, the OECD has collected information on export restrictions for raw materials, covering 66 metals and minerals in more than 84 countries (treating the EU as a single entity). The OECD has engaged in a similar effort for agricultural products, but for a smaller range of products.The OECD’s research has found that a sizeable number of WTO members maintain export restrictions of one form or another. A set of OECD researchers identified 96 Export Restrictions more than 900 instances between 2009 and 2012 in which export restrictions were introduced or tightened on industrial raw materials, and more than 300 such instances between 2007 and 2011 for agricultural commodities.6 Whereas the restrictions for agricultural products were often temporary in duration, those for raw materials tended to be medium or long term. The vast majority of export restrictions are applied by emerging economies and developing countries, including least-developed countries.7 Against this backdrop, it is not altogether surprising that PTAs seek to curb the use of export restrictions. 3.4.1 Prevalance In terms of the prevalence of these elements, more than three-quarters of the PTAs notified to the WTO have some form of legal discipline concerning export restrictions. This reflects a growing desire by governments to use PTAs as a mechanism to fill gaps within existing WTO law in this area. Several of the PTAs that incorporate provisions concerning export restrictions are economic integration agreements or customs union agreements that prohibited the imposition of such measures among its members. For example, Article 16 of the Treaty of Rome establishing the European Economic Community required the elimination of all customs duties on exports among its members. One of the earliest free trade agreements to incorporate a robust set of binding rules on export restrictions was the US-Israel FTA, which entered into effect in September 1985. Article 4 of that agreement prohibited the introduction of any “new customs duties on [. . .] exports or any charge having equivalent effect and any new quantitative restrictions on [. . .] exports or any measure having equivalent effect” that was in effect on the date of entry into force of the FTA and deemed “not inconsistent” with the GATT. Furthermore, Article 13 prohibited either party from imposing “as a condition of establishment, expansion or maintenance of investments by nationals or companies of the other Party, requirements to export any amount of production resulting from such investments.” Finally, Annex 3.9 enumerated certain requirements on authorities concerning export certificate of origin, including a baseline as to the information that must be contained within such certificates. Figure 3.1 notes the evolution in the number of treaties containing with export tax and/or other export-related provisions. As the number of PTAs have proliferated in the past decade, so too has the number of PTAs that address export-related elements of trade. Figure 3.1 also makes clear that the inclusion of such provisions in not entirely a recent phenomenon. Such 6 Fleiss et al. 2014. 7 For additional information highlighting patterns of use by developing countries, see Mendez Parra et al. 2016 at 11-17. 97 Handbook of Deep Trade Agreements provisions date back as far as the late 1950s, when treaties establishing economic integration areas required the elimination of export taxes between the parties. However, in recent years, the breadth and scope of the provisions included in such treaties have expanded as the number of PTAs have proliferated. Figure 3.1: Number of PTAs addressing export taxes and/or other export-related provisions Source: Deep Trade Agreements Database. Figure 3.2 highlights the percentage of PTAs that include some provision concerning export restrictions. Since 1990, more than two-thirds of PTAs that enter into force in any given five-year period have included some disciplines on export restrictions. The concept of Figure 3.2: Percentage of PTAs addressing export tax and/or other export-related restrictions, 1991-2015 Source: Deep Trade Agreements Database. 98 Export Restrictions including such provisions in a PTA was already established during the GATT era, before the recent explosion in PTAs. This should not be altogether surprising, since GATT Article XI is widely considered to be one the foundational tenets on which the GATT rests. Figure 3.3 reflects the same breakdown on the percentage of PTAs that include some provision concerning export restrictions by five-year period, but only for PTAs concluded exclusively among developing countries according to GATT Article XXIV (as opposed to the Enabling Clause). It shows that a slightly lower percentage of treaties containing additional rules on export restrictions, as compared to all treaties. Even so, more than half of all treaties concluded in this category contain some additional rules governing export restrictions. Figure 3.3: Percentage of PTAs concluded in developing countries addressing export taxes and/or other export restrictions, 1991-2015 Source: Deep Trade Agreements Database. Table 3.1 lists the relative frequency of the most commonly found provisions in PTAs coded for this study.While no particular type of provision commands widespread inclusion in almost all PTAs, what is striking is the fact that certain principles find their way into approximately two-thirds of all PTAs. These can be classified into three main types: (a) prohibitions on export taxes (either new or all); (b) prohibitions on export quotas (either new or all); and (c) requirements for export certificates of origin designed to prevent such certificates from being used as a non-tariff barrier. At first glance, there do not appear to be temporal patterns associated with the adoption of such provisions, nor are there patterns associated with the relative level of development of the parties to the treaties. Further in-depth examination ought to be conducted to check whether this is truly the case. 99 Handbook of Deep Trade Agreements Table 3.1: Most common provisions found in PTAs concerning export taxes and other export-related restrictions Type of Provision Frequency Prohibits all export taxes between/among the parties, either with or without reference to exceptions 70% Prohibits new export taxes, either with or without reference to exceptions 68% Imposes rules on government authorities for the issuance of export certification 65% Establishes recordkeeping requirements for the issuing authority for export certification of origin 64% Establishes penalties for false declarations related to export certification 62% Establishes a limitation on the period of validity for an export certificate of origin 61% Prohibits all export quotas / quantitative restrictions between/among the parties, with or without reference to exceptions 61% Includes a list of mandatory documentation required for export certification 58% Includes exemption from requiring export certification of origin 54% Prohibits new export quotas / quantitative restrictions between/among the parties 53% Prohibits an increase in the rate of any existing export tax 52% 3.4.2 Quantitative restrictions on exports GATT Article XI governs quantitative restrictions on trade, covering imports and exports alike. If one were to parse its language to highlight the legal discipline relevant for exports only, the provision would read: No prohibitions or restrictions other than duties, taxes or other charges, whether made effective through quotas, [. . .] export licenses or other measures, shall be instituted or maintained by any contracting party [. . .] on the exportation or sale for export of any product destined for the territory of any other contracting party. As mentioned earlier, a number of general and specific exceptions apply to this provision. Given the explicit and stringent WTO rules prohibiting the use of quantitative restrictions on exports except in limited circumstances, it should not be altogether surprising that few governments employ this approach. Nevertheless, a handful of instances do exist. Consider, first, why countries might choose to apply export restrictions on raw materials. For the vast number of the minerals and metals required for global industrial supply chains, OECD researchers have found that production is dominated by a handful of countries.8 If producers in such countries were able to coordinate among themselves and limit the supply on world markets, they could use their cartel power to control the price of the raw material. 8 For example, the top five producing countries account for 95 percent of the global production of antimony (with China alone accounting for 82 percent), 97 percent for lithium (with Chile alone accounting for 49 percent), and 99 percent for platinum group metals (with South Africa alone accounting for 59 percent). See Fleiss et al. 2014, p. 21. 100 Export Restrictions This would allow them to raise their overall profit. In addition, governments might try to use export restrictions to create an artificial price wedge between global and domestic prices. This could be part of a broader comprehensive strategy to entice foreign investment in sectors dependent on such minerals or metals as an input. Foreign firms might choose to shift their supply chains in order to take advantage of the price wedge and/or to obtain more stable access to the input. While these might, in theory, appear to be attractive reasons for defying explicit WTO rules, the reality is that very few countries have employed this tactic. For the period 2009 to 2012, OECD researchers found that China was the only WTO member to have applied quotas to control the export of minerals and metals.9 As noted earlier, these actions have been challenged in WTO litigation. In one instance, China’s action was in response to a territorial dispute with Japan.10 In other instances, the actions may have been part of an effort by China to curtail industrial pollution or to attract investment in key high-tech sectors.11 More recently, some have warned that China might consider employing this tactic in response to growing trade pressure from the United States.12 For agricultural products, the rationale for export quotas or other quantitative restrictions may differ. Governments may wish to create an artificial price wedge, not for the sake of attracting downstream agro-processing business, but simply to keep prices lower for domestic consumers. In analyzing quantitative restrictions for exports of agricultural bulk commodities, OECD researchers found that a larger number of countries use this tactic, including Argentina, Belarus, Russia, and Ukraine.13 Because a number of countries continue to use export quotas despite relatively clear prohibitions in WTO law except in particular circumstances, a substantial number of governments have found it necessary to incorporate such a provision in their PTAs. Approximately three-fifths of all PTAs analyzed for this study contain provisions addressing this issue. Many PTAs simply incorporate GATT Article XI outright as part of the treaty, without any major modification to the scope of permissible exceptions. Doing so renders the prohibition on quantitative restrictions on exports subject to dispute settlement proceedings under the PTA, and therefore, introduces yet another means of enforcing the WTO prohibition. 9 Ibid. 10 Bradsher 2010. 11 Morrison and Tang 2012. 12 Hornby and Sanderson 2019. 13 Fleiss et al. 2014, pp. 39-40, 60-61. 101 Handbook of Deep Trade Agreements In some instances, while the scope for legally permissible quantitative restrictions remains the same as under the GATT, the PTA imposes additional procedural requirements. For example, Article 99 of the Japan-Indonesia PTA requires that for quantitative restrictions on energy and mineral resources, the party imposing such a restriction shall provide relevant information to the other party as soon as possible and respond to any questions posed, with a view toward avoiding disruption of ordinary business activities between the parties. Some PTAs, however, achieve a WTO+ outcome by limiting or restricting the scope of exceptions for quantitative restrictions to a narrower set of circumstances than are set out in the GATT. An example is the Turkey-Georgia PTA. The language of its Article 7, which bans quantitative restrictions on exports, mirrors that of GATT Article XI:1. However, the PTA does not contain exceptions along the lines of GATT Article XI:2, which allows for temporary imposition of quantitative restrictions to prevent or relieve critical shortages of foodstuff or other products deemed essential, among other circumstances. Moreover, the general exceptions of the PTA are narrower than those found in GATT Article XX. As a result, the circumstances under which either party can impose export restrictions is narrower than would be the case absent the PTA. One might expect the inclusion of WTO+ provisions on export quotas to be more prevalent in PTAs involving the countries that apply export quotas with greater regularity. After all, according to one theory mentioned above, a PTA partner might wish to seek deeper integration with the quota-imposing trading partner, so as to ensure that it can access necessary raw materials or agricultural products. However, this study finds little evidence to support that theory. If anything, countries such as China are careful to not agree to WTO+ language concerning export restrictions. At best, they agree to incorporate GATT Article XI mutatis mutandis into the PTA. Larger emerging economies, which are also likely to resort to export quotas, can also exert their power in PTA negotiations to resist any WTO+ demands from their PTA partners that seek to curb their use. 3.4.3 Export taxes Whereas GATT Article XI prohibits export quotas, the same is not true of export taxes. Not surprisingly, these taxes are used with much greater frequency by WTO members. A study conducted by the WTO Secretariat14 estimated that approximately one-third of all WTO members employ export duties, including two-thirds of the LDCs that have been reviewed in the context of the Trade Policy Review Mechanism. Drawing on the OECD Inventory of Restrictions on Exports of Raw Materials and earlier analysis by OECD researchers, a study for the European Parliament found a sharp rise in 14 Piermartini 2014, p. 2. 102 Export Restrictions the use of export taxes since 2006.15 That study suggests several motivations for the use of export taxes, including a desire to increase revenue, keep domestic prices low, improve food security, and support industrial policy. During recent periods of volatility in agricultural commodity prices (2006-08 and 2010-11), export taxes were a popular form of export restrictions designed to guarantee domestic food supply and keep domestic prices insulated from external price spikes.16 The users of export taxes are primarily larger developing countries with the power to affect the terms of trade.17 Why is this the case? Any country enacting an export tax experiences two effects. The first is an efficiency effect, which results in a welfare loss caused by the production and consumptions distortions arising out of the export tax. This loss occurs regardless of size of the economy. The second effect is a positive terms-of-trade effect. In theory, an increase in the export tax could lead to an increase in the world price of the commodity, which results in a terms-of-trade gain. However, the ability to move the world price depends on the share of the overall market for the taxed product held by the country enacting the export tax. If its share is rather large, it can move the world price; however, if it is not, then it cannot. Therefore, only countries with a large share of the global market for a particular good will find it worthwhile to enact an export tax on that good, as only those countries can reap a terms-of-trade gain to offset the efficiency loss. The OECD study documented more than 900 instances of export taxes being imposed on raw materials from 2000 to 2012. In analyzing these data, the study found that Argentina was, far and away, the largest user of export taxes, accounting for nearly two-thirds of all instances of export taxes enacted during this period.18 Two other frequent users were China and Vietnam. The three countries collectively accounted for more than four-fifths of the worldwide use of export taxes on raw materials from 2000 to 2012.19 By far, almost all of the users of export taxes during this period were developing countries. Although Argentina, China, and Vietnam dominated, a wide range of other developing countries also used export taxes. For example, Cameroon applied an export tax for logs, Mozambique for cashews, and Pakistan for bones, hides, and skins.20 According to the OECD data, the most common category of products for which export taxes were applied in this period were cereals, oil seeds, edible vegetables, and edible fruit. 15 Mendez Parra et al. 2016. 16 Beckman et al. 2018; Estrades et al. 2017. 17 For an explanation as to why this is the case, see Piermartini 2004. 18 Mendez Parra et al. 2016, p. 14. 19 Ibid. 20 Piermartini 2004. 103 Handbook of Deep Trade Agreements Altogether, agricultural and food-related products accounted for the majority of export taxes on raw materials. The other major category included metals and minerals such as ores, base metals, and salts, as well as their downstream products such as iron and steel. Because the OECD data do not separate out the motivations for enacting export restrictions by type, in general, it is difficult to discern countries’ motivations for using export taxes and other types of restrictions. However, for countries where export taxes are either the dominant or only form of export restrictions, the government’s motivations are easier to discern. Two such countries are Vietnam and Argentina. The OECD data suggest that for Argentina, the key motivations for enacting the large set of export taxes appear to be to increase revenue and keep prices low.21 For Vietnam, the main drivers for export taxes are to secure domestic supply and for food security.22 With this background in mind, let us now turn to discuss certain trends that stand out when analyzing the numerous PTAs with restrictions on export taxes. Because the WTO agreements do not contain any outright disciplines prohibiting export taxes, note that all of the various forms of legal disciplines discussed below can be considered to be WTO+. As Table 3.1 makes clear, more than two-thirds of the PTAs examined in the sample contain some form of a general prohibition of export taxes between the two parties. In other words, the PTA extends the general ban on quantitative restrictions for exports to export taxes as well. Similar to the WTO’s legal discipline on quantitative restrictions, the treaty then includes a series of exceptions to the general prohibition. These exceptions are often based on the language of the exceptions found in GATT Article XI:2, and/or on the general exceptions contained in GATT Article XX. However, there is large variation in terms of how the legal discipline is constructed, especially with regard to the exceptions. In some instances, the PTA provides for only situational exceptions that are quite narrow. For example, the PTA between Taiwan, China, and Nicaragua requires that all export taxes between the two parties be eliminated unless a party confronts the exceptional circumstance of a domestic shortage or a domestic stabilization plan. The Japan-Mexico PTA does not allow export taxes to be applied even in the case of a domestic shortage, but it does provide for a few general exceptions such as for the protection of public morals or the protection of human, animal, or plant life or health. Another variation is when the PTA contains a general prohibition but exempts a vast category of products. One such category is agricultural products. For example, to date, the 21 Mendez Parra et al. 2016, p. 13. 22 Ibid. 104 Export Restrictions PTAs concluded by the European Free Trade Area (EFTA) exempt unprocessed agricultural goods in Harmonized System (HS) chapters 1-24, while several of the PTAs concluded by the EU make clear that the export tax prohibition applies only to industrial goods. Other common exemptions are for precious metals and precious stones, or other natural resources such as wood or logs. Yet another variation is when the PTA does not contain a categorical exception for particular goods, but instead includes a positive list of exempt goods that are identified according to their product specification codes. For example, Article 2.6 of the Costa Rica-Singapore PTA bans export taxes, but makes an exception for those designated in a separate annex that is part of the agreement. Those export taxes that are not scheduled in the annex must be eliminated. For each of these variations, the permissible export tax imposed in accordance with the exceptions may be subject to further limitations. Consider two examples of the types of limitations that may be written into a PTA. Under the original North America Free Trade Agreement (NAFTA), there was a one-year limit on how long an export tax may be applied. Another example is a requirement that the export price resulting from application of an export tax may not be higher than the comparable domestic price for the same good. This type of requirement can be found in NAFTA as well as in several other PTAs to which Canada is a party. Not all of the PTAs with a WTO+ legal discipline on export taxes contain a general prohibition. Some only go so far as to require that the parties not enact any new export taxes between them. Other PTAs allow the parties to maintain certain export taxes, but require that they not be raised. One or both of these requirements might also be found in instances where there is a general prohibition with exceptions, as well as in those PTAs without a general prohibition. In addition, some PTAs require that any existing export taxes be lowered or phased out according to a schedule contained in the PTA. The idea is to treat negotiations over export taxes in much the same manner as import duties. Furthermore, a phase-out provision provides domestic industry with time to adjust and may lower domestic opposition to new obligations on export taxes. Recall that three countries (Argentina, China, and Vietnam) account for a substantial portion of all export taxes imposed worldwide. Therefore, it may be worth elaborating on the PTA commitments of these particular countries. Several of the MERCOSUR PTAs to which Argentina is a party do not contain legal disciplines on export taxes. Those that do also include an extensive positive list of products for which export taxes can be maintained or even increased within bounds. For example, the trade agreement between MERCOSUR and Peru includes an annex of several hundred pages of products for which export taxes are permissible, which effectively allows Argentina to maintain its export taxes policy. 105 Handbook of Deep Trade Agreements Very few of the PTAs to which Vietnam is a party impose any obligations to eliminate or reduce export taxes. Most of the ASEAN PTAs to which Vietnam is a party do not contain such obligations. Nor is such an obligation found in Vietnam’s PTA with Japan.The one major trade agreement in which Vietnam has agreed to curtail its use of export taxes is the recent PTA concluded with the EU. Article 2.11 of that PTA requires that the parties not impose export duties on products unless they are scheduled in an appendix. Note that the corresponding appendix for the products for Vietnam totals more than forty pages. Nevertheless, it represents an upper bound on the number of products subject to an export tax, as well as on the amount of the tax itself, that Vietnam could impose on exports bound for the EU. The majority of China’s PTAs also do not contain any prohibition against export taxes. However, there are three PTAs for which this is not the case: China-Chile, China-New Zealand, and China-Peru. Given that these three countries are significantly smaller than China and therefore more likely to be at a disadvantage in their bilateral negotiations, it is interesting that they all chose to expend their negotiating capital to achieve this obligation. It is unclear why they did so. One possibility is that each country had particular vision about the type of value chain linkages it hoped to achieve with China as a result of the PTA. Further study is necessary to understand why China’s PTAs with these three trading partners deviate from the norm. 3.4.4 Export licenses Export licenses are also used to restrict exports. A common policy is to require that exporters obtain an export license, but to make its issuance subject to the discretion of government authorities. Through control of the volume of licenses granted and the speed with which applications are reviewed, government authorities can effectively control who exports a particular good, how much is exported, and to what countries it is exported. Not only does this process create additional cost for companies, but it also increases uncertainty in terms of the ability to obtain inputs for one’s supply chain. As a result, depending on how the non-automatic export license regime is administered, it too may induce shifts in production toward the country using this form of restriction, just as quantitative restrictions and export taxes may do. An OECD study found that at the HS6 product level, non-automatic export licensing requirements are the most common form of export restriction used by governments for minerals and mining products.23 At the time of the study, non-automatic export licenses were used by twenty-six countries, including nine that ranked among the top five producers worldwide for the product subjected to an export license. Among the most frequent users of export licenses were China, the Dominican Republic, Malaysia, the Philippines, and the Russian Federation While WTO law has rules governing import licenses, the same is not true for export licenses. 23 Fleiss et al. 2014, pp. 27-28. 106 Export Restrictions Not surprisingly, some governments have sought to enact WTO+ rules on export licenses through PTAs. Two recent developments are worth highlighting. First, a handful of PTAs in the Asia-Pacific region (e.g., Japan-Indonesia, Japan-Australia) have included a number of requirements on export licenses, which have also been incorporated into the CPTPP. One such obligation is a requirement that there be a contact point for information concerning export licenses. These PTAs also include several disclosure requirements, such as informing trading partners of the aggregate number of export licenses granted over a specified period, and of the types of measures taken in conjunction with export licensing procedures to restrict/stabilize domestic production or consumption of the relevant good. Second, in its most recent trade agreements with Colombia and Peru, Singapore, and Vietnam, the EU has successfully sought to extend the obligations of the WTO Import Licensing Agreement to export licenses. This obligation requires that non-automatic export licensing shall not have trade-restrictive or trade-distortive effects on exports additional to those caused by the imposition of the restriction. It also requires that the non-automatic licensing procedures shall correspond in scope and duration to the measure they are used to implement, and shall be no more administratively burdensome than absolutely necessary to administer the measure. In addition, a number of disclosure requirements exist, similar to those mentioned above for the CPTPP. Whether others will follow the EU in extending the obligations of the WTO Import Licensing Agreement to export licenses through PTAs remains to be seen. Overall, the inclusion of legal obligations concerning a trading partner’s export license regime is still a relatively new phenomenon. At this point, it remains somewhat limited. However, with its inclusion in the CPTPP and the recent moves by the EU, there exists a possibility that it could spread more widely in future PTAs. 3.4.5 Export certificates of origin and other administrative measures Finally, two other types of administrative measures also have the potential to affect exports. The first is the issuance of an export certificate of origin. These certificates are important in the context of PTAs, since the trading partners need to have some mechanism for validating the origin of the product, to ensure that it ought to receive preferential treatment. The issuance of an export certificate of origin serves as a means to prevent producers from non-PTA countries from passing off their goods as originating in order to take advantage of the lower duties.Therefore, it serves as an important tool for promoting deeper integration between or among the PTA trading partners. Because the issuance of an export certificate of origin may be subject to the discretionary power of the relevant authorities, it could be subject to possible mismanagement and abuse. Therefore, more than half of the PTAs examined include some form of legal obligations concerning origin certificates. The Australia–Chile FTA, for example, contains an annex that details a set of minimum requirements that must be included as part of the export certificate of origin, and another annex that includes samples in English and Spanish. It also specifies the period of validity of 107 Handbook of Deep Trade Agreements the certificate of origin. Finally, it discusses the conditions under which the exporter may complete and sign a certificate of origin when the exporter is not the producer of the good referred to in the certificate. All of these obligations serve to standardize and reduce any legal uncertainty around the export certificate of origin. Additionally, a majority of PTAs include explicit rules that highlight penalties associated with a false export certificate of origin. Some also contain rules that require notification of changes or cancellations of an export certificate of origin by either authority. The overall aim is to promote deeper integration between or among the trading partners, while also ensuring that standard procedures exist to safeguard the PTA from being abused by outside producers. Another measure that may affect exports is the export fees charged by the authorities that oversee exports. If these charges are too high, they too can act as an artificial distortion that curtails exports. In that sense, they can act in a manner similar to an export tax. To counter this possibility, more than 90 PTAs contain an explicit requirement that any export fees charged be based on the cost of services rendered. Among these, more than half also include a requirement that the trading partners publish all of their export fees. A small number also contain an obligation for advance notification of new export fees. Examples of such agreements include the Singapore–Peru FTA and the Comprehensive Economic Trade Agreement between the EU and Canada. A number of EU PTAs also contain an explicit prohibition against levying export fees on an ad valorem basis. This is designed to ensure that an export fee does not serve as a disguised export tax. Examples include the EU’s PTAs with Georgia, Moldova, Papua New Guinea, Singapore, the Republic of Korea, Ukraine, and Vietnam. 3.5. CONCLUSIONS Export taxes and other types of export restrictions remain widely used trade policy instruments. This is particularly true for developing countries, and particularly for agricultural products and extractive industries. Governments employ these policy instruments for different reasons. Some governments do so for food security purposes, others to raise revenue. Still others employ them as an element of industrial policy or to mitigate negative environmental impacts or promote sustainable development impacts. Whatever the rationale, the use of export restrictions can result in price distortions in world markets and harm neighboring countries.24 Researchers have found that contemporary export restrictions have contributed to spikes in international food prices and increased market instability in food. Export taxes, in particular, reduce global welfare. One study has 24 Anderson et al. 2010; Martin and Anderson 2011. 108 Export Restrictions suggested that their removal would lead to an overall welfare gain in excess of US$100 billion per year and expand world trade volumes by 2.8 percent.25 At present, except for GATT Article XI’s prohibition on quantitative exports, WTO law includes few rules governing export restrictions. Some have called for the WTO to take a more aggressive stance in monitoring export taxes and other forms of export restrictions, with the aim of spurring the development of additional multilateral rules.26 After all, it would be logical for such rules to be formulated in a multilateral context, especially given the widespread use of export restrictions among developing countries and the likely resistance of those countries to additional rules.27 However, in light of the WTO’s negotiating impasse, such calls have not resulted in any action. As has been true of so many other areas, the multilateral impasse has meant that the PTAs have emerged as the main arena for the development of new export rules. As this study highlights, a significant number of PTAs have incorporated a prohibition against quantitative restrictions on exports, sometimes with additional WTO+ rules. Many have also included rules on export taxes - ranging from prohibiting them outright between or among the trading partners to simply requiring that they be scheduled in an annex to the PTA. Another common feature found in PTAs is the inclusion of rules for the issuance of export certificates of origin. More recently, a handful of PTAs have also started to tackle the issue of creating rules to govern export licenses, in line with what exists in WTO law for import licenses. As global production becomes increasingly disaggregated across borders, trade can be distorted not only with tariff and non-tariff barriers on imports, but also those on exports. While the multilateral trade regime, in theory, exists to create rules to discipline both, those for exports have always lagged those for imports. This gap, however, is being filled by PTAs. As particular economies integrate more deeply with each other, they are also beginning to experiment with new rules to limit the use of export taxes and other types of export restrictions, in order to ensure that they are able to reap the full benefits of integration while still preserving their policy flexibility. ACKNOWLEDGMENTS The author extends his gratitude to the World Bank staff, the team of Harvard Law School students, and other participants at the project workshop held at the World Bank in February 2018 for their comments and useful insights. Special thanks to the OECD staff members in the Trade and Agricultural Directorate for their generosity in sharing their expertise. 25 Laborde et al. 2013. 26 For example, the E15 Expert Group on Trade and Investment in Extractive Industries called for the WTO to develop a centralized regime for the notification of all types of export restrictions and procedures for WTO members to consult with one another on export restrictions. The overall aim is to increase transparency and monitoring of such restrictions among WTO members. In addition, the group offered various proposals for possible negotiations to reduce export restrictions. See Espa 2015. 27 However, there are significant differences as to what this reform agenda ought to be. For an alternative view, see Karapinar 2011. 109 Handbook of Deep Trade Agreements REFERENCES Anderson, Kym, Will Martin, and Signe Nelgen. 2010. Export Morrison, Wayne, and Rachel Tang. 2012. China’s rare earth restrictions and food market instability. Vox EU. https://voxeu. industry and export regime: Economic and trade implications org/article/export-restrictions-and-food-market-instability. for the United States. Congressional Research Service Report R42510, Congressional Research Service, Washington, DC. Beckman, Jayson, Carmen Estrades, Manuel Flores, and Angel Aguiar. 2018. The impacts of export taxes on agricultural trade.Vox EU. Piermartini, Roberta. 2004. The role of export taxes in the https://voxeu.org/article/impacts-export-taxes-agricultural-trade. field of primary commodities.WTO Discussion Paper 4,World Trade Organization, Geneva. Bradsher, Keith. 2010.Amid tensions, China blocks vital exports to Japan.” New York Times, September 22. Solleder, Olga. 2013. Trade effects of export taxes. Graduate Institute of International Development Studies Working Espa, Ilaria. 2015. Export restrictions in relation to extractive Paper 08/2013, Graduate Institute of International industries.Think Piece for the E15 Expert Group on Trade and Development Studies, Geneva.. Investment in Extractive Industries. Estrades, Carmen, Manuel Flores, and Guillermo Lezama. 2017. The role of export restrictions in agricultural trade. International Agricultural Trade Research Consortium (IATRC) Commissioned Paper 20. Fleiss, Barbara, Christine Arriola, and Peter Lapis. 2014. Recent developments in the use of export restrictions. In OECD: Export Restrictions in Raw Materials Trade: Facts, Fallacies and Better Practices, 17-62. Paris: OECD Publishing. Hornby, Lucy, and Henry Sanderson. 2019. Beijing’s rare plan to strike back. Financial Times, June 5. Karapinar, Baris. 2011 Export restrictions and the WTO law: How to reform the “regulatory deficiency.” Journal of World Trade 45(6): 1139-55. Korinek, Jane, and Jessica Bartos. 2012. Multilateralising regionalism: Disciplines on export restrictions in regional trade agreements. OECD Trade Policy Paper 139, Organisation for Economic Co-operation and Development, Paris. Laborde, David, Carmen Estrades, and Antoine Bouet. 2013. A global assessment of the economic effects of export taxes. World Economy 36(10): 1333-54. Martin, Will, and Kym Anderson. 2011. Export restrictions and price insulation during commodity price booms. Policy Research Working Paper 5645, World Bank, Washington. Mendez Parra, Maximiliano, Samuel R. Schubert, and Elina Brutschin. 2016. Export taxes and other restrictions on raw materials and their limitation through free trade agreements: Impact on developing countries. European Parliament Paper EP/EXPO/B/DEVE/FWC/2013-08/LOT7/15, European Parliament, Brussels. 110 Services CHAPTER 4 Services B. Gootiiz, G. Jonetzko, J. Magdeleine, J. Marchetti, and A. Mattoo 111 Handbook of Deep Trade Agreements CHAPTER 4 Services B. Gootiiz*, G. Jonetzko*, J. Magdeleine†, J. Marchetti†, and A. Mattoo* * World Bank , Washington, DC, United States † World Trade Organization, Geneva, Switzerland CONTENTS 4.1. INTRODUCTION 113 4.2. PREVIOUS ATTEMPTS AT SURVEYING AND CODING SERVICES PTAs 113 4.3. THE NEW DATASET 119 4.3.1 Design section of the dataset 119 4.3.2 Commitments/reservations section of the dataset 120 4.4. OVERALL TRENDS IN PTAs WITH SERVICES 121 4.5. THE DESIGN OF SERVICES PTAs 124 4.5.1 Structure 125 4.5.2 Scope and Coverage 127 4.5.3 Core obligations 128 4.5.4 Movement of natural persons (MNP) 135 4.5.5 Beneficiaries 135 4.5.6 Exceptions 135 4.5.7 Safeguard mechanisms 136 4.5.8 Dispute settlement 137 4.6. CONCLUSIONS 137 ACKNOWLEDGMENTS 138 REFERENCES 139 ANNEX 140 112 Services 4.1. INTRODUCTION Over the last two decades, preferential trade agreements (PTAs) that liberalize trade between two or more economies have proliferated. Many of these include rules for the liberalization of trade and investment in services. As of end-December 2016, 144 PTAs containing such rules have been notified to the WTO.1 Despite this clear upward trend, there are still significant gaps in the collection and systematization of information on services PTAs for the purpose of policy analysis. This chapter presents a new, comprehensive database on the design and depth of 144 PTAs covering trade in services signed by 105 WTO members (considering the EU as one member). This new dataset on services PTAs covers the most important aspects of these agreements, from the framework and general rules to specific commitments on liberalization. There appears to be no equivalent dataset, in terms of either the scope or coverage of services PTAs. The first section of the dataset comprises the coding of the main architectural and design features of services PTAs, while the second comprises the coding of liberalization commitments/reservations made by each signatory under each of these PTAs.This chapter focuses on the former. It provides a first overview and analysis of the rules that the agreements create to enhance market access, including rules on data flows, state-owned enterprises, government procurement of services, and competition policy. The commitments made in the context of these PTAs will be the focus of future analysis. The chapter is organized as follows. The next section reviews previous attempts at coding different aspects of services PTAs. The third section describes the new dataset and explains the main coding assumptions. The fourth section provides a first overview of results and trends in the design of PTAs arising from the database. The final section concludes. 4.2. PREVIOUS ATTEMPTS AT SURVEYING AND CODING SERVICES PTAs A few attempts have been made in the recent past to collect and systematize information on services PTAs – what is generally called coding. These previous studies (Table 4.1) typically covered a limited number of agreements and restricted themselves to either a few architectural features (i.e., the main features of the rule book), or to assessments of the value added brought by PTA commitments in comparison with GATS commitments or Doha Development Agenda (DDA) offers.2 In addition, a database of PTAs covering goods, services, 1 The information has been drawn from the Regional Trade Agreements Information System (RTA-IS) maintained by the WTO, available at http://rtais.wto.org/UI/PublicMaintainRTAHome.aspx. 2 DDA offers refer to the offers of new or improved commitments submitted by WTO members since the start of the Doha round of multilateral trade negotiations (also known semi-officially as the Doha Development Agenda). The negotiations were formally launched at the 4th WTO Ministerial Conference, held in Doha, Qatar, in November 2001. Initial offers were submitted by end-March 2003, and revised offers were submitted by end-July 2006. 113 Handbook of Deep Trade Agreements or both is maintained by the WTO Secretariat, on the basis of WTO Members’ notifications to WTO. This database identifies the coverage of those agreements (goods/services), and provides links to their official texts and relevant annexes.3 Another relevant source of information is the Integrated Trade Intelligence Portal (I-TIP) kept by the WTO and the World Bank. I-TIP Services is a set of linked databases that provides information on WTO Members’ commitments under the GATS, services commitments in PTAs, applied services trade policies and regulations, and services statistics.The PTA module of this database allows users to access and search information on WTO Members’ commitments and reservations in agreements notified under GATS Article V . At the time of writing, it included information on commitments in around 95 PTAs. These commitments and reservations are coded by sector/subsector, mode of supply, type of limitation (market access, national treatment – NT), or obligation concerned (for negative-list-type agreements).4 In a series of papers using the same methodology,5 commitments made by 53 WTO members in 67 PTAs in modes 1 (cross-border supply) and 3 (commercial presence) were compared across 142 and 152 services subsectors, respectively.6 The comparison was run both across PTAs (i.e., commitments entered into by individual trading partners in the various PTAs to which they were parties), and between PTAs and GATS schedules/DDA offers (i.e., for each trading partner, the best commitment across its PTAs was compared with its latest DDA offer, or—in cases where no offer had been submitted in the DDA negotiations—compared with its GATS schedule). Apart from sectoral coverage, the studies focused on the value added from PTA commitments over GATS schedules/DDA offers.The value added was gauged by comparing, for each services subsector and mode under consideration (modes 1 and 3), whether the GATS commitment or DDA offer evolved from a partial commitment (i.e., a commitment with some market access limitation) to a full commitment (i.e., without any market access limitation), or from a partial commitment to a better partial commitment (i.e., with lesser limitations).7 These studies also 3 See footnote 1. 4 More information on I-TIP Services can be found at http://i-tip.wto.org/services/. 5 Roy et al. 2007, Marchetti and Roy 2009, Roy 2011. 6 The 4 modes of services supply in trade agreements are mode 1: cross-border supply; mode 2: consumption abroad; mode 3: commercial presence; and mode 4: presence of natural persons. 7 The evolution from a restrictive to a less restrictive commitment but without reaching full liberalization was assessed on the basis of an improved Hoekman methodology: GATS commitments were coded as 0 (unbound), 0.5 (partial), and 1 (no restrictions or full commitment).The movements between 0.5 and 1 were coded as half the difference between 0.5 and 1 (0.75). In cases of further improvements by the same trading partner in other PTAs, the new – better – commitment would be coded as half the difference between 0.75 and 1 (0.875).All the figures, that is, all the codes for all subsectors by each trading partner, were then aggregated and normalized to 100. The higher the value, the higher the value added provided by PTA commitments over GATS schedules/DDA offers.This index did not provide the actual level of restrictiveness of policies – in other words, it did not constitute a proper Services Trade Restrictiveness Index (STRI). The latest iteration of this dataset (2011) is available at http://www.wto.org/english/tratop_e/serv_e/dataset_e/dataset_e.htm. 114 Services looked into two design features of PTAs, namely, the liberalization approach (negative or positive list) and the existence of a GATS-type market access obligation for mode 3.8 Another study assessed the value added of PTAs in East Asia for the four modes of supply.9 Covering commitments in the four modes of supply in 154 subsectors, the exercise yielded 616 entries per PTA, which were classified into four categories: (a) subsectors and modes for which only a GATS commitment exists or a PTA does not offer any improvement (GATS only); (b) subsectors and modes for which a partial GATS commitment exists and a PTA eliminates or relaxes one or more remaining trade-restrictive measures (PTA improvements); (c) subsectors and modes for which no GATS commitment is available but a PTA commitment is made (PTA new sectors); and (d) subsectors and modes for which neither a GATS nor a PTA commitment exists (unbound). A PTA commitment was counted as an improvement over existing GATS commitments if at least one trade-restrictive measure was relaxed or eliminated. A follow-up paper by the same authors10 looked into some of the design features of the same sample of 25 East Asian PTAs, in particular (a) the scheduling approach (positive vs. negative list); (b) the treatment of investment (by looking at the definition of commercial presence in services chapter, the definition of investment in horizontal investment disciplines, and the relationship between services and horizontal investment disciplines); (c) the treatment of the movement of natural persons (by looking at the definition of mode 4, and the existence or not of a separate chapter or agreement related to the movement of natural persons); (d) rules of origin (for juridical and natural persons; (e) dispute settlement (state-to-state and investor-state); and (f) other elements (inclusion of provisions on recognition of other parties’ standards, domestic regulation, government procurement, subsidies, and emergency safeguard measures). A fourth study examined services commitments in 56 PTAs to which an OECD country is a party.11 The preferential content of those agreements, and the value added as compared to the 8 As explained in Roy et al. 2008, “[w]hile various PTAs still follow either the NAFTA or GATS structure…, a number of the PTAs reviewed in this chapter have evolved into a combination of the two approaches, the aim being to achieve greater coherence between services and investment disciplines so as to avoid discrepancies in the treatment of investment in goods and services or in the treatment of trade in services under different modes of supply. Combined approaches therefore seek to ensure that services trade under all modes of supply are subject to the same core disciplines and that mode 3 is covered by generic investment disciplines. In such cases, mode 3 is typically subject to some obligations in both the investment chapter and the services chapter. Unlike in NAFTA, mode 3 is subject to the services chapter’s disciplines on non-discriminatory quantitative restrictions, as in GATS (i.e. Article XVI). However, in addition to GATS and as in NAFTA, generic investment disciplines apply to mode 3. A number of the services PTAs reviewed in this chapter have adopted variants of such a combined approach; e.g. all the recent PTAs involving the United States….” 9 Fink and Molinuevo 2008a. The authors’ dataset is not publicly available. 10 Fink and Molinuevo 2008b. The authors’ dataset is not publicly available. 11 Miroudot, Sauvage, and Sudreau 2010. 115 Handbook of Deep Trade Agreements GATS, was assessed through an analysis of market access and national treatment commitments at the level of 155 services subsectors. Three levels of commitment (and value added) were distinguished: (a) status quo (when a subsector is “unbound” or when the commitment is the same as in GATS, this is the status quo and the PTA is not preferential); (b) GATS+ or preferential “binding” (where PTA commitments improve on existing GATS commitments or cover new sectors); and (c) GATS- (where PTA commitments are less stringent than GATS commitments). Additionally, partial commitments were further broken down according to nine categories, four of which correspond to market access (scope of subsector limited, restrictions on foreign ownership, quantitative restrictions on the service or service supplier, restrictions on the movement of people), and five of which correspond to national treatment (nationality/residency requirements and licensing; restrictions on the movement of people; discriminatory measures on subsidies or taxes; discriminatory measures on property/land; and other discriminatory measures). Finally, the study provided an overview of rules of origin for service providers and most-favored- nation (MFN) clauses in services chapters in order to see whether commitments granted might be extended to non-parties to minimize discrimination among foreign service suppliers. A similar study12 looked at the design features of about 55 PTAs, covering 13 aspects or provisions: MFN, national treatment, market access (non-discriminatory quotas), domestic regulation, emergency safeguards, subsidy disciplines, government procurement, rules of origin (denial of benefits), scope/coverage, negotiating modality (positive vs. negative list), treatment of investment in services, right of non-establishment, and ratchet mechanism (which implies that restrictions removed by a PTA signatory cannot be reintroduced, thus locking in reform undertaken subsequent to the agreement). A significant step in the codification of PTAs design features has been the Design of Trade Agreements (DESTA) project.13 DESTA is a comprehensive database that identifies and codes the main chapters, provisions, and features of PTAs. As of February 2017 (latest information available on the project’s website), DESTA researchers have manually coded design features for more than 620 agreements in force since 1945, of which 178 appear to have significant provisions on trade in services. The coding includes 8 basic aspects related to trade in services embedded in those PTAs: existence of substantive provisions on trade in services, liberalization approach (positive vs. negative list), existence of MFN, existence of NT, right of non-establishment, movement of natural persons, review provisions, and sectoral coverage.With the exception of sectoral coverage, most questions require a binary (yes-no) answer, but some can be answered in one of three ways (e.g., for the question on national treatment, the reply may be “0” if no national treatment clause is included in the service chapter, “1” if the national treatment clause is included in the service 12 Mattoo and Sauvé 2011. 13 https://www.designoftradeagreements.org/. 116 Services chapter but is limited in scope to specific sectors, and “2” if the national treatment clause is included in the service chapter and is not limited to specific sectors). When it comes to the review of provisions applicable to services trade, the most comprehensive exercise thus far14 identified 48 significant provisions in services PTAs, divided into 7 broad themes and further into sub-themes. The themes are architecture, scope, beneficiaries, core obligations, permissive provisions, domestic regulation and recognition, and institutional provisions (Table 4.2). Table 4.1: Previous datasets on services PTAs Study Number of Main features PTAs covered Fink and Molinuevo 25 in East Asia Liberalization valued added in 154 services subsectors (based on 2008a and 2008b W/120 classification) and 4 modes of supply identified under the (F&M) GATS. Key architectural elements: approach to scheduling commitments, treatment of investment and movement of natural persons, rules of origin, dispute settlement. Houde, Kolse-Patil, 20 (investment Key investment disciplines in PTAs’ investment and Miroudot 2007 disciplines) and services chapters. Coding of investment-related and 10 commitments/reservations in 12 big sectors of W/120 (commitments) (but only for 10 PTAs). Roy, Marchetti, and 67 (by 53 Liberalization value added in 152 services subsectors Lim 2007; Marchetti Members) in mode 3 and 142 services subsectors in mode 1 and Roy 2009; (based on W/120 classification). Roy 2011 (M&R) Miroudot, Sauvage, 56 (where Examines services schedules of commitmentsin 155 services subsectors and Sudreau 2010 an OECD (based on W/120 classification) in the 4 modes of supply. Partial (M&S&S) is a party) commitments are broken down according to nine categories of non- conforming measures (4 on market access and 5 on national treatment). Design of Trade 178 Identifies and codes 8 key variables: existence of substantive Agreements provisions on trade in services, liberalization approach (positive (DESTA) vs negative list), existence of MFN, existence of NT, existence Database of right of right of non-establishment, movement of natural (Baccini et al. 2011) persons, review provisions, sectoral coverage. Latrille and 80 Analysis of 48 key provisions structured under 7 themes Lee 2012 commonly found in PTAs: architecture, scope, (L&L) beneficiaries, core obligations, permissive provisions, domestic regulation, institutional provisions. Mattoo and 55 Looks into 13 key features: MFN, national treatment, market access Sauvé 2011 (nondiscriminatory quotas), domestic regulation, emergency (M&S) safeguards, subsidy disciplines, government procurement, rules of origin (denial of benefits), scope/coverage, negotiating modality (positive vs. negative list), treatment of investment in services, right of non-establishment, and ratchet mechanism. Source: Deep Trade Agreements Database. 14 Latrille and Lee 2012. 117 Handbook of Deep Trade Agreements Table 4.2: Design of PTAs: Coverage of issues in previous datasets / exercises M&R F&M M&S&S L&L M&S DESTA Number of PTAs covered 67 25 56 80 55 178 Coverage of modes of supply (incl. investment) X X X X Relationship between cross-border trade in services X (CBTS) and investment chapters Separate sectoral chapters/annexes X Sectoral exclusions X X Policy exclusions Gov. procurement X X X Subsidies X X X Liberalization approach Positive vs. negative X X X X X Standstill X Ratchet X X Market access X X X National treatment X X X MFN X X X X Prohibition of local presence requirement X X X Prohibition of performance requirements X (1) Export Local content Technology transfer Prohibition of nationality/residence req. X (2) Disciplines on monopolies Additional commitments New issues (e.g., cross-border data flows) Phase-in sectoral liberalization Domestic regulation X (1) X (1) Necessity test X Oblig. to inform on licensing decision X Oblig. to inform on application status X Single windows Time-bound decisionmaking ROI administration of regulations X Mutual recognition X X Transparency Publication X Allow prior comment X Independent authority Exceptions General Security Prudential Other Safeguards Emergency safeguard X X Renegotiation of commitment Balance of Payments (BoP) X Natural persons X (1) Specific provisions X X Coverage of specific categories X Coverage of employment Rules of origin Juridical persons X X X Natural persons X X Dispute settlement State-state Investor-state Other Note: Issues/areas identified in the first column are those covered by the Deep Trade Agreements Database; (1) the existence of this provision/discipline is only covered in general terms; (2) nationality or residence requirement for senior managers or members of the boards of directors. Definition: ROI = reasonable, objective, and impartial. Full title of datasets are provided in Table 4.1. 118 Services 4.3. THE NEW DATASET The dataset on which this chapter is based breaks new ground in the analysis of services PTAs by providing a comprehensive coding of both the design features of the sample of 144 services PTAs and the types of restrictions in their liberalization commitments. The dataset consists of two sections: the first comprises the coding of the main architectural and design features of the PTAs, while the second comprises the coding of liberalization commitments/reservations made by each signatory under each of the PTAs. The new dataset represents an improvement over previous coding attempts in terms of both the extensive margin (more PTAs are covered) and the intensive margin (more PTA aspects are covered and with further granularity). While it builds on information already contained in the PTA module of I-TIP Services, the new dataset includes additional PTAs, provides for a more detailed coding of commitments/reservations, and introduces a new framework for coding service provisions. 4.3.1 Design section of the dataset The first part of the dataset identifies 8 main areas or aspects of services PTAs: (a) structure; (b) scope and coverage; (c) substantive disciplines; (d) exceptions; (e) safeguard mechanisms; (f) movement of natural persons; (g) rules of origin; and (h) dispute settlement.These areas are further broken down into subareas, for a total of about 50 questions/variables for each PTA. The structure area identifies how the four modes of supply are covered in the agreement, and - in the case of the inclusion of an investment chapter or the existence of an investment protocol - if and how the hierarchy between chapters is defined in case of inconsistency between those chapters.The structure area also looks into how specific sectors are treated in the agreements, namely, as chapters or annexes (the former being an indication that the sectors are considered more substantial). The scope and coverage area focuses on the sectors included or excluded from the agreement, the policy exclusions (government procurement, subsidies, treatment of job seekers under mode 4), as well as the liberalization approach adopted (GATS-type, negative, or other approaches). With respect to the substantive disciplines, the database captures information on a diverse set of provisions related to market access (how it is defined); non-discrimination (MFN and NT); local presence requirements (for cross-border trade in services); performance requirements or obligations on members of firms’ boards of directors (for mode 3/investment in services); discipline of monopolies; and possibility of undertaking additional commitments15 or 15 “Additional commitments” are provided for in GATS-type agreements—they allow parties to undertaken additional obligations on measures not considered as market access or national treatment limitations. 119 Handbook of Deep Trade Agreements coverage of new issues (e.g., cross-border data flows). This section also deals with domestic regulation (e.g., procedural aspects of licensing, single window, mutual recognition, necessity test) and transparency (publication, prior comments on new regulations, and availability of appeal procedures). The template for this exercise also includes questions on (a) the existence of exception clauses (general, security, or prudential matters); (b) safeguard mechanisms (emergency issues, possibility of renegotiating commitments, balance-of-payments difficulties); (c) the movement of natural persons (analyzed by including questions relating to the scope of mode 4 in the agreements; (d) rules of origin for firms and natural persons; and (e) the type of dispute settlement possibilities foreseen in the agreements (state-state and/or investor-state). Annex Table 4.A.1 provides the complete questionnaire used to code the main design features of the agreements, while Table 4.2 (above) shows how these variables have been covered in previous datasets or exercises. In comparison with those previous datasets, the new dataset addresses the design of services PTAs much more comprehensively.16 In particular, it improves significantly on the coding of domestic regulation and transparency, performance requirements, coverage of new issues (such as cross-border data flows), dispute settlement, and additional commitments/phase-in liberalization commitments. 4.3.2 Commitments/reservations section of the dataset The second part of the dataset codes the commitments for all 144 PTAs in the sample. The analysis of the commitments will be presented in a future paper. For this analysis, the PTAs were divided into two categories on the basis of their liberalization approach - those following a GATS-type approach, and those following a negative-list-type approach. In the case of the former, all market access and national treatment commitments in the four modes of supply have been coded, identifying the level of commitment (full, unbound, or partial). Whenever a partial commitment was encountered, the dataset identifies the market access and/or national treatment limitation concerned. Market access limitations are those contained in the GATS; namely: (a) limitations on the number of suppliers; (b) limitations on the value of transactions/ assets; (c) limitations on the total number of operations; (d) limitations on the number of natural persons that may be employed; (e) requirements regarding types of legal entities or joint ventures; and f) foreign equity limitations. A category of “other” was added to make allowance for market access limitations not clearly falling into any of these six categories. In the case of national treatment limitations, scheduled limitations were allocated to any of the following 14 categories: (a) tax measures; (b) subsidies or grants; (c) other financial measures; (d) 16 Only one previous exercise - reported in Latrille and Lee 2012 - comes close in that regard, but the results were only summarized in narrative form in their paper, and the actual data were not presented. 120 Services nationality requirements; (e) residency requirements; (f) licensing, standards, and qualifications; (g) registration requirements; (h) authorization requirements; (i) performance requirements; (j) technology transfer requirements; (k) local content requirements; (l) ownership/rental of land/ property; (m) and other national treatment requirements. In the case of negative-list-type agreements, the basic assumption is that services that are not excluded from the sectoral coverage of the agreement or by virtue of Annex 2 reservations (i.e., reservations for future measures), or for which no reservations have been made in Annex 1 (i.e., existing non-conforming measures) are considered to be fully liberalized. For those services for which reservations have been filed (through either Annex 1 or Annex 2), the reservations have been allocated to the relevant modes of supply.17 In general, these agreements allow for the filing of reservations with regard to the following disciplines: market access, national treatment, MFN, local presence requirements, nationality/residence of boards of directors and/or managers, and performance requirements.The reservations on market access and national treatment have been allocated to the market access and national treatment categories identified above. 4.4. OVERALL TRENDS IN PTAs WITH SERVICES As can be seen in Figure 4.1, the trend towards the inclusion of services trade in PTAs intensified in the 2000s, probably as a reflection of services regulatory reform worldwide. From a political economy perspective, the relationship between this trend and developments in the Doha Round of negotiations (which started at the end of 2001) is unclear. Governments seem to have turned their attention to services at every negotiating front - whether multilateral or plurilateral or bilateral - and that may explain the increasing number of PTAs covering services trade even in the first half of 2000s, when there was still hope that the Doha negotiations could be concluded. As of 2006, it became clear that bilateral or plurilateral PTAs were the only channel through which governments liberalized, committed to further liberalization, or complemented unilateral efforts to liberalize trade in services. The obligation to notify these agreements to the WTO, by virtue of GATS Article V and the additional “RTA transparency mechanism” adopted by WTO in 2006, has made it possible to keep track of these trends. Arguably, services have become a major component of PTAs, featuring prominently in mega-regional negotiations such as the Trans-Pacific Partnership (CPTPP), Transatlantic Trade and Investment Partnership (TTIP), Regional Comprehensive Economic Partnership (RCEP), and in the broad, services-only Trade in Services Agreement (TISA) initiative (which remains unfinished and deadlocked at the time of writing). 17 Reservations to cross-border trade (CBT) have been allocated to modes 1, 2, and 4, or to modes 1 and 2 (depending on the definition of CBT used in the PTA). Reservations appearing in the investment chapter and concerning services have been all allocated to mode 3. 121 Handbook of Deep Trade Agreements Figure 4.1: Proportion of PTAs with services over time a. In percent b. In numbers Source: Deep Trade Agreements Database. Figure 4.2: PTAs with services by level of development a. In percent b. In numbers Source: Deep Trade Agreements Database. The majority of services PTAs have been signed between developed and developing countries and the share has remained persistent over time. However, the involvement of developing countries has been growing, as evidenced by the growing number of developed-developing PTAs as well as developing-developing PTAs. Especially since 2005, the share of services PTAs signed between developing countries has increased (Figure 4.2). 122 Services While before 2000, trade liberalization in general took place within regions or among neighboring economies (e.g., the EU, NAFTA, ASEAN, MERCOSUR), cross-regional agreements (between partners in different regions of the world) have become common. While this is a general trend, it was only natural in the case of services, where modes of supplying internationally are less dependent on proximity factors (technological means, foreign direct investment, and movement of people). Also, it appears that countries in Latin America and the Caribbean (LAC) and East Asia and the Pacific (EAP) are more likely to be involved in PTAs with services, as evidenced by much larger shares of PTAs signed by parties in these regions. The countries in South Asia, the Middle East and North Africa (MENA), and Sub-Saharan Africa have the lowest number of PTAs with services (Figure 4.3). Figure 4.3: Number of PTAs with service, intra-, and extraregional groupings a. Regional groupings: overall b. Regional groupings: by region Note: Cross-regional grouping means that at least one party to the PTA is outside the region of the other parties. If all parties are in the same region, they fall into an intra-regional grouping. In panel b, the same PTA can appear in more than one region, depending on the parties to the PTAs. If one of the parties to a PTA is in the OECD region and the other party is in the LAC region, the PTA is grouped under both regions. In panel b, LAC means Latin America and Caribbean, EAP means East Asia and Pacific, ECA means Eastern and Central Asia, SA means South Asia, MENA means Middle East and North America, and SSA means Sub-Saharan Africa. Over time, as shown in Figure 4.4, there seems to have been a premium placed on negative- list approaches to liberalization of trade in services (see the definition in the previous section). PTAs signed among high-income or developed countries are more likely to follow a negative-list approach. However, the trends are less clear in the case of developing countries: PTAs signed by developing countries with developed partners tend to follow either approach, while, interestingly, a significant number of PTAs between developing countries follow a negative-list approach. This latter development may be an indication of the political will and background - economies having already embarked on unilateral reform processes - underpinning negotiations between developing countries. 123 Handbook of Deep Trade Agreements Figure 4.4: Liberalization approaches Source: Deep Trade Agreements Database. 4.5. THE DESIGN OF SERVICES PTAs A basic distinction between services PTAs is based on their approach to liberalization commitments. In agreements following a “negative-list” approach, the relevant obligation (e.g., national treatment) will apply to all the services sectors falling under the purview of the chapter unless the party lists relevant non-conforming measures (for example, in the Annex on existing non-conforming measures) and/or identifies sectors or sub-sectors to which the obligation does not apply (for example, in the Annex on ¨future¨ measures). This is unlike agreements following the so-called “positive-list” approach, like the GATS, where the relevant obligation (e.g., national treatment) applies only to those sectors that are listed or committed in the Member’s schedule (positive-listing) and subject to any conditions and qualifications set out therein. An additional difference between both approaches concerns the obligations that may be subject to reservations. While positive-list-type agreements only allow for reservations on market access and national treatment, negative-list-type agreements allow for reservations to be filed with respect to not only market access and national treatment but also MFN, the obligation to forbid local presence requirements, the obligation to eliminate performance requirements, and the obligation not to request nationality/residency senior management personnel and members of boards of directors. Finally, negative-list- type agreements are usually accompanied by a ratchet mechanism which locks in future liberalization. The distinction between these two liberalization approaches is important 124 Services since the choice of liberalization approach may have an impact on the negotiation’s dynamic, the actual prerequisites in terms of parties’ preparation, and the actual effects - the negative-list-type agreements being considered more ambitious and therefore more demanding on the parties. There is also a third category of agreements. The “other” or hybrid category includes some EU-related agreements and others that do not fit either the positive- or negative-list category. The common features of these hybrid other agreements are a neither/nor approach to the scheduling of commitments, the absence of modes, and the use of alternative concepts such as freedom to provide services and freedom of establishment. The 144 PTAs with services components can be grouped into three main categories: those based on a GATS-type positive list (58), those with a negative-list approach similar to NAFTA, (75), and those that include elements from both these approaches as well as other characteristics (11). The main features of each type of PTA are analyzed below on the basis of the 8 broad areas identified in the dataset. As noted above, these areas are (a) structure; (b) scope and coverage; (c) substantive disciplines; (d) exceptions; (e) safeguard mechanisms; (f) movement of natural persons; (g) rules of origin; and (h) dispute settlement. Items within each area are analyzed based on the types of agreements (positive, negative, or other types). 4.5.1 Structure The structure of a services PTA refers to the way that services trade has been covered in the agreement; i.e., the definition and inclusion of different modes of supply (Table 4.3); relevance of the investment chapter and its relation to the services chapter and other services provisions (Table 4.4); and the existence of specific sectoral rules, either in separate chapters or in annexes/annotations to the main services chapter ( Table 4.5 and Table 4.6). Table 4.3: Structure of services trade, number of PTAs Category A. B. C. D. All 4 modes (M) All 4 modes (M) Chapter on cross-border Chapter on cross-border covered in a self- covered in a self- trade in services trade in services (M1, M2), contained chapter contained chapter (M1, M2, M4), plus plus an investment (plus an Annex on M4) chapter on investment (M3) chapter (M3), plus a and an additional and other chapters/ chapter on movement investment annexes on movement of persons (M4) chapter/protocol of persons Positive 10 46 0 2 Negative 3 9 59 4 Other 4 3 3 0 Total 17 58 62 6 125 Handbook of Deep Trade Agreements The positive-list PTAs are more likely to have a category A or B structure and the majority of negative-list agreements feature a category C structure (see Table 4.3). Whatever the approach, it is clear that trade in services is predominantly governed by a combination of chapters, and that disciplines on investment have become a major component not only of PTAs but also of the framework through which trade in services is liberalized. Most negative- list agreements (60 out of a total of 75) contain provisions clarifying the relationship between the investment chapter/protocol and the other chapters as indicated in Table 4.4. Table 4.4: Inclusion of provisions clarifying the hierarchy/relationship between the investment chapter/protocol and trade in services chapter Number of PTAs Positive 14 Negative 60 Other 2 In addition to the main provisions relating to services trade (whether covered in cross-border trade in services or investment chapters), often some sectoral disciplines are also included in trade agreements. The GATS contains three sectoral annexes on air transport, telecommunications, and financial services, which generally develop or clarify GATS provisions on sector-specific features (telecom and financial services annexes) or define the coverage of the sectors (financial services and air transport). As shown in Tables 4.5 and 4.6, services PTAs present a variety of approaches, including separate sector chapters or annexes that provide for trade and investment disciplines for specific sectors (financial services in US PTAs). Of the 144 services PTAs in the dataset, 57 percent contain sector-specific chapters for financial, telecommunications, or air transport services, and 53 percent have sector-specific annexes to a trade in services or investment chapter. Sector-specific rules contained in these chapters and annexes do not vary significantly from agreement to agreement. Table 4.5: Inclusion of separate sector-specific chapter Number of PTAs Positive 20 Negative 59 Other 3 Total 82 Table 4.6: Inclusion of services sector-specific annexes to a trade in services or invesment chapter Number of PTAs Positive 33 Negative 37 Other 6 Total 76 126 Services 4.5.2 Scope and coverage The scope and coverage of a services PTA refers to sectoral exclusions (e.g., services supplied in the exercise of governmental authority; air transport services, in line with the GATS18); the exclusion of specific policies (e.g., government procurement, subsidies, employment); and the general liberalization approach. 4.5.2.1 Sectoral exclusions In line with the GATS, 85 percent of PTAs (122 out of 144) exclude from coverage those services supplied in the exercise of governmental authority.19 Some PTAs use the GATS definition of governmental authority, and others use a non-exhaustive list of examples such as Republic of Korea-Singapore (Article 9.2.3) or Panama-Taiwan, China (Article 11.02.3 (c)). Some agreements use both the GATS criteria and the list of examples. In the case of air transport services, about 86 percent of the PTAs do not cover air traffic rights (cross-border air transport). However, about 31 percent cover air transport services beyond the three ancillary air transport services covered by the GATS (Table 4.7, column 4). The majority of these cases appear in the negative-list agreements and involve specialty air services relating to aerial work; i.e., services using a plane for purposes other than passenger or freight transport (such as for crop spraying, aerial photography, aerial advertisement). Notably, air transport provided via mode 3/commercial presence is not excluded from the investment chapters of the agreements. Table 4.7: Sectoral exclusions from services and/or investments sections Excluded services Excluded air Air transport services Excluded supplied in traffic rights covered beyond (a) other services exercise of (cross-border computer reservation governmental air transport) systems, (b) marketing authority and sale services, or (c) Number of PTAs maintenance and repair services Positive 50 47 8 27 Negative 66 71 37 54 Other 6 6 0 5 Total 122 124 45 86 18 The following air service sectors are generally excluded from both the GATS and PTAs: air traffic rights; air transport services beyond (a) computer reservations, (b) marketing and sale services, and (c) maintenance and repair services. 19 These services are not identified by means of a list of sectors but are rather characterized through a sort of test (services supplied in the exercise of governmental authority must be supplied “neither on a commercial basis nor in competition with one or more service suppliers”). [GATS Article I.3(c)]). Results, and therefore the actual sectoral coverage of the GATS, may differ by Member. For example, prison services in the US belong to merchant services and hence are subject to general services disciplines of the PTAs. This is not the case for other Members, for which prison services are exclusively provided by the public sector. 127 Handbook of Deep Trade Agreements Other excluded services sectors cover financial services, telecommunications services, and maritime cabotage services (i.e., maritime transport between two ports located within the same country). Some of these sectors are excluded from the general services disciplines of the PTAs but are covered by the sector specific chapters/annexes and rules, which go beyond the generic services disciplines or GATS disciplines. 4.5.2.2 Policy exclusions Taking the GATS as a benchmark, policy exclusions cover government procurement, subsidies, and employment on a permanent basis. In the template, allowance was made for other policy exclusions in a catch-all category simply called “others.”As shown in Table 4.8, in 90 percent of the agreements, government procurement is not covered by services disciplines (mainly MFN, market access and national treatment). About 80 percent of the PTAs exclude provisions on subsidies, and on employment on a permanent basis. The agreements that do not exclude these areas generally involve EU/EC and EFTA members. Table 4.8: Policy areas excluded from PTAs Government Subsidies Job seekers, Other procurement citizenship, residence, or employment on a Number of PTAs permanent basis Positive 55 43 41 5 Negative 73 69 68 2 Other 2 1 6 1 Total 130 113 115 8 4.5.3 Core obligations The core disciplines included in the dataset are market access, MFN, national treatment, and standstill and ratchet obligations, as well as obligations to avoid local presence, performance, and local content requirements. 4.5.3.1 Market access Market access is an obligation universally found in PTAs. The purpose of this question is to find out how the agreements define this obligation, which is aimed at curbing quantitative limits on market access. Specifically, do PTAs define market access limitations as in the GATS, where prohibitions focus on quantitative restrictions (whether in the form of quotas or economic needs tests), legal forms of entry or foreign equity limits (Box 4.1); or as in negative- type agreements (which have become a model for other countries), where prohibitions focus on non-discriminatory measures, thus excluding foreign equity ownership limits? 128 Services Box 4.1. Market access limitations in the GATS …on the number of service suppliers (quotas, monopolies, exclusive service suppliers) …on the total value of service transactions or assets …on the total number of service operations or on the total quantity of service output …on the total number of natural persons that may be employed in a particular service sector or company …on specific types of legal entity or joint venture requirements …on foreign equity ownership Out of 144 agreements, 56 define market access according to the GATS definition; 42 use the negative-list-type approach, therefore omitting foreign equity limitations; and 42 adopt other definitions. Most GATS-type or positive-list agreements use the market access definition with 6 limitations (Table 4.9). The negative-list-type agreements use the market access definition with 5 limitations. Surprisingly, the other category mostly comprises negative-list-type agreements; these use a definition of market access different from the first two categories or do not contain market access provisions. For example, Chile-Japan, Chile- Korea, and Japan-Mexico PTAs, which are all negative-list agreements, do not contain a provision on market access. Table 4.9: Obligation on market access Market access definition Market access definition Other Number of PTAs covering 6 limitations covering 5 limitations (different definitions) Positive 52 0 6 Negative 3 42 29 Other (incl. n/a) 1 0 8 Total 56 42 43 4.5.3.2 National treatment obligation The questionnaire asks whether there is an obligation on national treatment. Most of the PTAs (139 out of 144) have an obligation on national treatment, though it is defined somewhat differently depending on the type of agreement. In GATS-type agreements, national treatment is defined as treatment accorded to foreign services and service suppliers that is less favorable than treatment accorded to “like” domestic services and service suppliers. It covers both de jure and de facto treatment and has a provision indicating that the national treatment clause should not be read as requiring that foreign service providers be compensated for the inherent handicaps of being foreign service providers. The negative-list-type agreements 129 Handbook of Deep Trade Agreements define national treatment as less favorable treatment accorded to foreign services and service suppliers in comparison to treatment accorded, “in like circumstances,” to domestic services and service suppliers. That is, the likeness of services and service suppliers is replaced by the likeness of circumstances faced by the different services and service suppliers. In the EU-type agreements, national treatment obligations simply prohibit restrictions based on nationality. 4.5.3.3 MFN obligation Out of 144 agreements, 122 have an obligation to extend MFN benefits to trading partners in subsequent agreements, and 22 PTAs do not have such an obligation. Out of the 22 without an MFN obligation, 15 are positive-list agreements, 5 are negative-list agreements, and 2 are other types of agreements. 4.5.3.4 Status quo and ratchet obligations When a party to a PTA commits to a “standstill,” it means the measures/reservations listed per sector and mode, if they do not conform with the obligation concerned, will not become more restrictive in the future. “Ratchet” means that if the measure is amended in the future to become less restrictive, the new, more favorable treatment will set the benchmark for the standstill requirement and will thus become the new commitment. About 50 percent of the PTAs contain standstill and ratchet provisions. The vast majority of these PTAs are negative- list agreements, although a few positive-list PTAs (ASEAN-India, China-Australia, New Zealand-Singapore) also have standstill and ratchet obligations (Table 4.10). Table 4.10: Standstill and ratchet provisions Number of PTAs Standstill provision Ratchet provision Positive 3 3 Negative 69 68 Other 1 2 Total 73 73 4.5.3.5 Other obligations These include obligations to avoid local presence requirements, performance requirements (based on exports, local content, or technology transfer), nationality/residency requirements for senior management and boards of directors, and provisions on monopolies (Table 4.11). The prohibition on performance-related requirements mostly appear in the investment chapters of negative-list agreements. Both positive- and negative-list-type agreements generally contain provisions to discipline monopolies. Most positive-list agreements contain additional commitments, but only six negative-list agreements contain them (Table 4.12). 130 Services Table 4.11: Obligation not to have certain requirements concerning trade in services Local presence Export- related Local content Performance Technology Nationality requirement performance requirement requirements transfer requirement for Number in other areas board of directors or of PTAs senior management Positive 0 0 0 0 0 7 65 60 60 58 57 59 Negative Other 1 2 2 1 2 4 Total 66 62 62 59 59 70 Table 4.12: Number of PTAs with provisions on monopolies, additional commitments, and liberalization Number Discipline Additional Cover new Gradual of PTAs monopolies commitments areas liberalization Positive 45 47 13 16 Negative 53 6 26 14 Other 8 1 1 1 Total 106 54 40 31 4.5.3.6 Domestic regulation Various aspects of domestic regulation are of interest for trade in services, and have been the subject of much discussion and negotiations over the past two decades. Concerns about domestic regulation go beyond the market access and national treatment obligations addressed in different sections of PTAs, because even in the absence of market access limitations or outright discrimination, practices related to licensing, qualifications, or technical standards may still act as obstacles to foreign services and service suppliers. While domestic regulations are important to fulfill legitimate policy objectives and prevent undesirable practices, they may also lack objective and transparent licensing requirements or technical standards, or be characterized by discretionary procedures. Depending on the PTA, domestic regulation disciplines may be mandatory, voluntary, subject to reservations and limitations, or of a best-endeavor nature. The level of enforceability is an important factor for service suppliers seeking to operate in foreign markets. Out of the 144 agreements in the dataset, 119 include provisions relating to qualifications, licensing, and technical standards (Table 4.13). Out of the 119 PTAs that do contain such provisions, 26 refer to a necessity test; e.g., the obligation that licensing, qualification, and technical standards be not more burdensome/not more restrictive than necessary to ensure certain policy objectives such as quality of service, integrity of the profession, consumer protection, or environmental protection. There is no necessity test going beyond licensing, qualification, and technical standards. 131 Handbook of Deep Trade Agreements Table 4.13: Number of PTAs containing provisions on qualification, licensing, technical standards, and a necessity test Number of PTAs Provisions on qualification, Of which, those subject licensing, and technical standards to a necessity test Positive 47 8 Negative 70 16 Other 2 2 Total 119 26 In addition to the obligation for domestic regulations to meet the necessity test, the domestic regulation section of the questionnaire asks whether each PTA in the sample includes obligations to (a) decide on applications in a timely manner; (b) inform applicants regarding the decision of authorities or the status of the review; (c) establish a single window for submitting applications; and (d) administer laws and regulations in a reasonable, objective, and impartial manner.Tables 4.14-4.16 show the results of the coding exercise with regard to the different types of domestic disciplines. Table 4.14: Number of PTAs with obligation to inform on status of application Number of PTAs Inform on status of application Of which, mandatory Positive 51 49 Negative 62 62 Other 1 1 Total 114 112 Table 4.15: Number of PTAs with obligation to make decisions within a certain period of time Number of PTAs Make decisions within certain period of time Of which, mandatory Positive 36 33 Negative 56 53 Other 1 1 Total 93 87 Table 4.16: Obligation to administer the measures / laws / regulations in a reasonable, objective, and impartial manner Number of PTAs Administer in a reasonable, Of which, mandatory objective, and impartial manner Positive 49 48 Negative 70 47 Other 3 2 Total 122 97 132 Services Regarding the obligation to provide information on application decisions, 114 PTAs include such provisions, which for the vast majority of agreements (110) is a general and mandatory obligation, as in the GATS (Table 4.14). This is not the case for 4 agreements (e.g., Japan- Indonesia, which is on a best-endeavor basis). A smaller number, 93 PTAs, include a provision relating to the timing of licensing decisions (62 percent of GATS-type agreements, and 75 percent of negative-list agreements). For 89 of those, it is a mandatory obligation (Table 4.15). A larger number of PTAs (122) include an obligation to administer the measures/laws/ regulations in a reasonable, objective, and impartial manner. This can be found in most negative-list-type agreements (93 percent), but also in many GATS-type agreements (84 percent). It is interesting that this obligation is mandatory in almost all of the GATS-type agreements but in only two-thirds of the negative-list agreements (Table 4.16). It appears from the results that, except for ASEAN, none of the services-related chapters or annexes of PTAs has a provision for setting up a single window for applications, even though the single window is regularly raised by many commentators as a facilitation factor for trade in services. 4.5.3.7 Mutual recognition While 137 PTAs include provisions on recognition of standards, education, experience obtained, or licenses granted in certain jurisdictions (Table 4.17), only 44 percent are more binding than the voluntary treatment in the GATS. Out of the 60 PTAs with provisions that go beyond voluntary treatment, 28 are GATS-type and 25 are negative-list-type PTAs. Most of the other types of PTAs include provisions that go beyond voluntary treatment. The way recognition is articulated may differ among PTAs. While in many cases provisions mention the recognition of education, experience, or licenses obtained in other countries (whether parties or non-parties to the agreements), some others specify that the provisions on recognition only apply to experience or education obtained in the parties to the agreement (e.g., ASEAN, CARICOM). Table 4.17: PTAs with mutual recognition provisions Number Mutual recognition Of which, mandatory of PTAs provisions (general or on specific commitments), or best endeavor Positive 58 28 Negative 70 26 Other 9 7 Total 137 61 133 Handbook of Deep Trade Agreements 4.5.3.8 Transparency Transparency is an important pillar of the trading system and is necessary at various levels, ranging from (a) the opportunity to comment during the drafting of laws and regulations; to (b) the publication of finalized laws/regulations/policy guidelines; to (c) enforcement of rules and procedures; to (d) establishment of contact points for information requests and provision of information to oversight bodies. This study concentrated on two aspects of transparency - the opportunity for interested parties to comment on proposed and regulations; and the publication of legal texts or their availability to interested persons. One hundred thirty of the PTAs include provisions related to the publication of information on relevant laws and regulations, of which 113 make this obligation mandatory, as in the GATS. Concerning the latter point, the overall proportion of mandatory obligations is the same whether GATS-type or negative-list-type PTAs are considered. The three other types of PTAs that include those types of provisions all correspond to general obligations and have a mandatory nature (Table 4.18). All remaining PTAs refer to best endeavor. Table 4.18: Obligation to publish Number of PTAs Publish relevant laws and regulation Of which, mandatory Positive 52 45 Negative 75 65 Other 3 3 Total 130 113 Table 4.19: Obligation to provide for prior comment on proposed regulation Number of PTAs Provide for prior comment Of which, mandatory on new regulation Positive 31 10 Negative 56 12 Other 2 1 Total 89 23 Eighty-nine PTAs include provisions that foresee the possibility of interested parties providing prior comments on proposed regulation, which is not incorporated in the GATS. Around half the GATS-type PTAs and 75 percent of the negative-list agreements contain such provisions. Twenty-three agreements consider this a general obligation of a mandatory nature (Table 4.19). The clear majority of the remaining PTAs indicate a best-endeavor nature. There is no specific trend depending on the type of agreement. 134 Services 4.5.3.9 Independent authority for appeal procedures One hundred nineteen PTAs include an obligation to set up an independent authority to which an appeal can be brought. In all cases, the obligation is of a general and mandatory nature (Table 4.20). Table 4.20: Obligation to set up an independent authority for appeal procedures Number of PTAs Obligation to set up independent authority Of which, mandatory Positive 48 48 Negative 68 68 Other 3 3 Total 119 119 4.5.4 Movement of natural persons (MNP) There are various aspects in PTAs addressing the question of the movement and the temporary presence of persons. One hundred fifteen PTAs include specific provisions on the presence of natural persons, generally in the form of a chapter or an annex. Only 29 identify specific categories of professions (with specific provisions for those professionals); and only 20 go beyond GATS by covering permanent residency/ employment or job seekers (Table 4.21). Table 4.21: Treatment of the movement of natural persons Specific provisions Specific categories Coverage of permanent or Number of PTAs on presence of professionals temporary employment of natural persons Positive 44 11 3 Negative 60 17 16 Other 11 1 1 Total 115 29 20 4.5.5 Beneficiaries Beneficiaries may be referred to in terms of rules of origin. Unlike PTAs that cover trade in goods, services PTAs do not have a chapter on rules of origin. Instead, services PTAs have sections defining cases in which “denial of benefits” can be invoked. These sections have the effect of covering rules of origin for juridical persons and natural persons (Tables 4.22 and 4.23). 4.5.6 Exceptions Regardless of type of agreement, almost all PTAs include general exceptions and allow for security exceptions. Eighty-seven PTAs allow for prudential exceptions for financial services (Table 4.24). 135 Handbook of Deep Trade Agreements Table 4.22: Rules of origin for juridical persons Be owned or Be (i) owned or Be owned or Incorporated Incorporated Other controlled by controlled by controlled by under the under the persons of natural persons natural domestic law domestic law the other of the other persons of the of the party of the party party (whether party, OR (ii) be other party and have juridical or natural) owned or substantive AND have controlled by business substantial juridical operations in business persons of the the territory operations (in other party AND of a member the other party have substantial Number or a third party) business of PTAs operations Positive 13 32 5 3 2 3 Negative 19 17 6 16 17 0 Other 5 2 0 1 0 3 Total 37 51 11 20 19 6 Table 4.23: Rules of origin for natural persons Be a national Have the right Be a national of that Have a Other of the party of permanent party (and be center of (whether residence in resident there or in economic resident there that party a third party) or interest in or in a third have the right of the territory Number party) permanent residence of the party of PTAs in that party Positive 16 0 40 0 1 Negative 22 1 51 0 1 Other 4 0 4 0 3 Total 42 1 95 0 5 Table 4.24: Exceptions General Security Prudential exception Number of PTAs exceptions exceptions for financial services Positive 58 56 36 Negative 75 74 46 Other 10 9 5 Total 143 139 87 4.5.7 Safeguard mechanisms Only a small number of PTAs (36) have a provision allowing for emergency safeguard actions in specific sectors and modes. A smaller number of PTAs (mostly positive-list agreements) contain provisions allowing the renegotiation of specific commitments or reservations. In contrast, the majority of PTAs have a provision allowing measures to counter balance-of-payments difficulties (Table 4.25). 136 Services Table 4.25: Safeguard mechanisms Emergency Allowing renegotiation Provision allowing safeguard of commitments measures to counter Number of PTAs provision or reservations balance-of-payments difficulties Positive 25 20 54 Negative 8 6 57 Other 3 0 10 Total 36 26 121 4.5.8 Dispute settlement Most positive-list-type agreements (72 percent) refer only to state-to-state dispute settlement (DS). The remaining positive-list agreements also include investor-state dispute settlement for relevant mode 3 issues. Most negative-list agreements (84 percent) include provisions covering both state-to-state and investor-state dispute settlement (Table 4.26). Table 4.26: Dispute settlement Number of PTAs State-state settlement only State-state and investor-state settlement Positive 42 16 Negative 12 63 Other 7 3 Total 61 82 4.6. CONCLUSIONS Most preferential trade agreements negotiated in the last two decades - 144 notified to the WTO as of end-December 2016 - contain rules for the liberalization of trade and investment in services. Despite this clear upward trend, there are still significant gaps in the collection and systematization of information on services PTAs for the purpose of policy analysis. This chapter presents a new database on the design of PTAs covering services trade. This dataset, which contains information for 144 services PTAs signed by 105 economies, covers in a comprehensive manner the most important aspects of these agreements, from the design of the regulatory framework through to the specific commitments liberalization. The dataset consists of two sections: the first one comprises the coding of the main architectural and design features of these services PTAs, while the second comprises the coding of liberalization commitments/reservations made by each signatory under each of 137 Handbook of Deep Trade Agreements these PTAs. The new dataset improves over previous attempts in both the intensive margin (more PTAs are covered) and the extensive margin (more PTA aspects are covered and with further granularity). Based on these data, we provide a first overview and analysis of results and trends in the provisions included in such PTAs. At the time of writing, this chapter’s authors are working on a quantification exercise on the commitments made in PTAs with services. The exercise compares those with the levels of restriction identified in applied services policies. Preliminary findings show, first, that commitments under PTAs seldom go beyond countries’ applied policies and, therefore, the explicit liberalization resulting from the agreements is usually limited to a few members and a few areas. Second, the PTAs do enhance transparency and policy certainty because parties’ services commitments cover more trading partners and more sectors, and in some cases are closer to applied policies than their commitments under GATS. Finally, and importantly, the new rules that the agreements create, including on data flows, state-owned enterprises, government procurement and competition policy, could enhance access to service markets. ACKNOWLEDGMENTS We would like to thank Claudia Castillo Comabilla and Evgeniia Shannon for their assistance with the construction of the database as well as for helpful comments and suggestions.We would also like to thank our discussant Roberto Echandi, and other participants in the conference “Evolution of Deep Trade Agreements” for their comments. All errors remain the authors’. 138 Services REFERENCES Baccini, Leonardo, Andreas Dür, Manfred Elsig, and Karolina Houde, Marie-France, Akshay Kolse-Patil, and Sébastien Milewicz. 2011. The design of preferential trade agreements: Miroudot. 2007. The interaction between investment and A new dataset in the making. WTO Staff Working Paper services chapters in selected regional trade agreements. ERSD-2011-10, World Trade Organization, Geneva. OECD Trade Policy Working Paper 55, Organisation for Economic Co-operation and Development, Paris. Borchert, Ingo, Batshur Gootiiz, and Aaditya Mattoo. 2011. Services in Doha: What’s on the table? In Unfinished Business? The WTO’s Latrille, P., and J. Lee. 2012. Services rules in regional trade Doha Agenda, edited by Will Martin and Aaditya Mattoo. London: agreements: How diverse and how creative as compared to Center for Economic and Policy Research and World Bank. the GATS multilateral rules? WTO Staff Working Paper ERSD-2012-19, World Trade Organization, Geneva. Borchert, Ingo, Batshur Gootiiz, and Aaditya Mattoo. 2014. Policy barriers to international trade in services: Evidence from a Marchetti, J.A., and M. Roy. 2009. Services liberalization new database. World Bank Economic Review 28(1):162-88. in the WTO and in PTAs. In Opening markets for trade in services: Countries and sectors in bilateral and WTO negotiations, Deardorff, Alan V., and Robert M. Stern. 2008. Empirical edited by Juan A. Marchetti and Martin Roy. Cambridge: analysis of barriers to international services transactions and the Cambridge University Press. consequences of liberalization. In A Handbook of International Trade in Services, edited by Aaditya Mattoo, Robert M. Stern, Marchetti, Juan, and Martin Roy. 2013. The TISA initiative: and Gianni Zanini. Oxford: Oxford University Press. An overview of market access issues. Staff Working Paper ERSD-2013-11, World Trade Organization, Economic Dür, Andreas, Leonardo Baccini, and Manfred Elsig. 2014. The Research and Statistics Division, Geneva. design of international trade agreements: Introducing a new database. Review of International Organizations 9(3): 353-75. Mattoo, Aaditya, and Carsten Fink. 2004. Regional agreements and trade in services: Policy issues. Journal of Fink, Carsten, and Martin Molinuevo. 2008a. East Asian Economic Integration 19:742-79. preferential trade agreements in services: Liberalization content and WTO rules. World Trade Review 7(4):641-73. Mattoo, Aaditya, and Pierre Sauvé. 2011. The preferential liberalization of services trade. In Preferential Trade Agreement Policies ———. 2008b. East Asian free trade agreements in services: for Development – A Handbook, edited by Jean-Pierre Chauffour Key architectural elements. Journal of International Economic and Jean-Christophe Maur.Washington DC:World Bank. Law 11:2 (June). Miroudot, S., J. Sauvage, and M. Sudreau. 2010. Geloso Grosso, M., et al. 2015. Services trade restrictiveness Multilateralising regionalism: how preferential are services index (STRI): Scoring and weighting methodology. OECD commitments in regional trade agreements? OECD Trade Trade Policy Paper 177, OECD Publishing, Paris. DOI: Policy Paper 106. http://dx.doi.org/10.1787/5js7n8wbtk9r-en. OECD (Organisation for Economic Co-operation and Gootiiz, Batshur, and Aaditya Mattoo. 2017. Services in the Development). 2009. Methodology for deriving the STRI. Trans-Pacific Partnership: What would be lost? Journal of Paper presented at the OECD Experts Meeting on the World Trade 51(5):763-85. Services Trade Restrictiveness Index (STRI), Paris, July 2–3. Gootiiz, Batshur, and Aaditya Mattoo. 2017. Regionalism in Roy, Martin, Juan Marchetti, and Hoe Lim. 2007. Services services: A study of ASEAN. World Economy 40(3): 574-97. liberalization in the new generation of preferential trade doi:10.1111/twec.12328. agreements (PTAs): How much further than the GATS? World Trade Review 6(2):155-92. Hoekman, Bernard. 1996. Assessing the General Agreement on Trade in Services. In The Uruguay Round and the Roy, Martin. 2011. Services commitments in preferential Developing Countries, edited by Will Martin and L. Alan trade agreements: an expanded dataset. WTO Staff Working Winters. Cambridge: Cambridge University Press. Paper ERSD-2011-18, World Trade Organization, Geneva. Hufbauer, Gary Clyde. 2016. Liberalization of services trade. US International Trade Commission. 2016. Trans-Pacific Chapter 7 of Assessing the Trans-Pacific Partnership, Volume 1: Partnership Agreement: Likely impact on the US economy Market Access and Sectoral Issues. Washington, DC: Peterson and on specific industry sectors. Publication 4607, https:// Institute for International Economics. www.usitc.gov/publications/332/pub4607.pdf. 139 Handbook of Deep Trade Agreements ANNEX Annex Table 4.A.1: Template on architecture and design of services PTAs AREA SUB-AREA QUESTION Answer type yes=1; no=0; letters or NA) AGREEMENT Coverage of modes How is services trade contemplated in this agreement? Please choose one of the options. STRUCTURE of supply A) All 4 modes covered in a self-contained chapter (plus an Annex on M4, as in the GATS)? B) All 4 modes covered in a self-contained chapter (plus an Annex on M4) and an additional Investment chapter/protocol. A, B, C, D C) Chapter on cross-border trade in services (as in NAFTA, M1, M2, and M4), PLUS one chapter on investment (dealing with M3) and other annexes/chapters on movement of persons. D) Chapter on cross-border trade in services (M1 & M2), plus an investment chapter (M3), plus a chapter on movement of persons (M4), as in CETA. Does the agreement contain provisions clarifying the hierarchy/relationship between 1, 0, NA the investment chapter/protocol and the other chapters? Separate Chapters Are there separate sector-specific chapters (e.g., financial services, telecommunications)? 1 or 0 and/or Annexes (except Please list such sectors in the comments (please also consider those when replying to for movement of natural subsequent questions). persons) Are there sector-specific annexes to a chapter on trade in services (such as express 1, 0, NA delivery as an annex to CBTS chapter, financial services) or to a chapter on investment? Please list such sectors in the comments. SCOPE AND Sectoral exclusions Are any of the following services excluded from the 1) Chapter on CBTS and/or 2) 1, 0, NA COVERAGE Chapter on Investment? Please indicate the chapter in the comments 1, 0, NA Services supplied in exercise of governmental authority Air traffic rights (cross-border air transport) For air transport, are there services covered beyond (1) computer reservation systems, (2) 1, 0, NA marketing and sale services, or (3) maintenance and repair services? (e.g., specialty air services, airport operation services - please list in the comments) Other 1, 0, NA Policy exclusions Are any of the following policy areas excluded from the Agreement? If some of them (e.g., government procurement of services, or subsidies to service sectors) are included in chapters other than the CBTS chapter, please indicate so. Government procurement 1, 0, NA Subsidies 1, 0, NA Job seekers; citizenship, residence, or employment on a permanent basis. 1, 0, NA Other 1, 0, NA Liberalization approach In the case of disciplines subject to scheduling/reservations (i.e., market access), what is the approach followed? A) Positive list (as in GATS); B) Negative list (as in NAFTA); C) Other (including combinations of the previous ones depending on the discipline, e.g., positive list for MA and negative list for NT): If C, please give details in the comments A, B, C Does the agreement contain a standstill provision? 1 or 0 Does the agreement contain a ratchet provision - implying all unilateral liberalization is legally bound? 1 or 0 Market access How is the market access obligation defined?: A, B, C A:As defined in the GATS (by reference to 6 prohibited market access limitations) B:As defined in the US FTAs (by reference to 5 prohibited market access limitations, and omitting foreign equity limitations) C: Other (no provision on market access; used different definitions; or other reasons) SUBSTANTIVE Non-discrimination Does the agreement/services chapter contain an MFN provision? 1 or 0 DISCIPLINES Is there a national treatment (NT) obligation? 1 or 0 Others Does the agreement contain a prohibition of local presence requirement as a pre-condition 1, 0 or NA to supply services cross-border? Does the agreement contain obligations not to apply export related performance requirements? 1, 0 or NA Please provide comments on the particular obligation Does the agreement contain obligations not to apply local content related performance 1, 0 or NA requirements? Please provide comments on the particular obligation Does the agreement contain obligations not to apply technology transfer related 1, 0 or NA performance requirements? Please provide comments on the particular obligation Does the agreement contain obligations not to apply performance requirements in other areas? 1, 0 or NA Please provide comments on the particular obligation Is there a general or sector specific obligation not to require nationality or residency 1, 0 or NA requirements for senior managers and/or members of Board of Directors? Does the agreement contain provisions to discipline monopolies? 1, 0 or NA If yes, does it contain it to A. protect the interest of foreign suppliers, B. protect consumers? A, B, NA Can parties to the agreement make additional commitments? If yes, please specify? 1, 0, NA Are there other provisions that cover new issues (i.e., cross-border data flows)? 1, 0, NA Does the agreement include obligations to liberalize specific sectors/transactions gradually 1, 0 or NA over time (if yes, please specify)? 140 Services AREA SUB-AREA QUESTION Answer type yes=1; no=0; letters or NA) Domestic Does the agreement contain provisions on qualification, licensing, and technical standards? 1 or 0 Regulation (DR) If yes, are those measures (qualifications, licensing, and technical standards) 1 or 0 subject to a “necessity test”? If yes, does the necessity test apply to other types of measures 1 or 0 or NA (beyond licensing, qualifications, and technical standards)? Is there a provision requiring the Party’s competent authority to inform the applicant 1 or 0 of the decision concerning the application? Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation Is there a provision requiring the Party’s competent authority to provide information 1 or 0 concerning the status of the application? Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation Is there an obligation to set up a single window for submission of applications? 1 or 0 Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation Is the competent authority required to make the licensing decision within a certain period of time? 1 or 0 Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation Do the Parties have to administer the measures/laws/regulations in a reasonable, 1 or 0 objective, and impartial manner? Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation Does the agreement contain provisions on mutual recognition? 1 or 0 Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation Transparency Is there a provision requiring publications of relevant laws and regulations or making the 1 or 0 laws and regulations available to interested persons? Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation Is there an obligation to allow interested parties an opportunity for 1 or 0 prior comment on proposed regulation? Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation Is there an obligation to set up an independent authority to which an appeal could be launched? 1 or 0 Please indicate the nature of the discipline above: A. General obligation and mandatory nature A, B, C, D, NA B. Obligation subject to limitations or reservations C. General obligation, but best-endeavor nature D. Voluntary obligation 141 Handbook of Deep Trade Agreements AREA SUB-AREA QUESTION Answer type yes=1; no=0; letters or NA) EXCEPTIONS General Does the agreement include general exceptions? (GATS Article XIV list) If yes, please list the general exceptions which go beyond the GATS Article XIV list 1 or 0 Security Does the agreement allow for security exceptions? 1 or 0 Prudential Does the agreement contain a prudential exception for financial services? 1 or 0 Other Do other exceptions apply to services sectors or measures? 1 or 0 SAFEGUARD Emergency safeguard Is there a provision allowing emergency safeguard action in specific sectors and/or modes? 1 or 0 MECHANISMS Renegotiation Is there a provision allowing the renegotiation of specific commitments or reservations? 1 or 0 of commitments Balance-of-payments Is there a provision allowing measures to counter balance-of-payments difficulties? 1 or 0 MOVEMENT Are there specific provisions clarifying the scope of the presence of natural persons 1 or 0 OF NATURAL (e.g., Chapter/annex on temporary presence of business persons)? PERSONS If yes, does the chapter/annex on movement of natural persons cover specific categories 1, 0, NA of professionals? (e.g., architects, lawyers, and accountants)? If yes, does the chapter/annex on movement of natural persons cover permanent 1, 0, NA or temporary employment (i.e. beyond GATS mode 4)? RULES Rules of origin for To be considered a service supplier of a party to the agreement, in the case of the supply OF ORIGIN juridical persons of services through commercial presences, does a juridical person have to: A. Be owned or controlled by natural persons of the other party A, B, C, D, E B. Be owned or controlled by persons of the other party (whether juridical or natural) AND have substantial business operations (in the other party or a third party) C. Be owned or controlled by juridical persons of the other party AND have substantial business operations (in the other party, a third party, or WTO Members) D. Be (i) owned or controlled by natural persons of the other party, OR (ii) be owned or controlled by juridical persons of the other party, AND have substantial business operations (in the other party or a third party) E. Incorporated under the domestic law of the party; F. Incorporated under the domestic law of the party and have substantive business operations in the territory of a member G. Other (please specify in the comments) Rules of origin To be considered a service supplier of a party to the agreement, does a natural person have to: for natural persons A. be a national of the party (whether resident there or in a third party) A, B, C, D, E B. have the right of permanent residence in that party C. be a national of that party (and be resident there or in a third party) or have the right of permanent residence in that party D. have a center of economic interest in the territory of the party E. Other DISPUTE Please indicate which one of the following dispute settlement provision applies SETTLEMENT to the services agreement? A. State-state dispute settlement; A, B, C B. Investors-state dispute settlement; C. Both 142 Investment CHAPTER 5 Investment J. Crawford and B. Kotschwar 143 Handbook of Deep Trade Agreements CHAPTER 5 Investment J. Crawford* and B. Kotschwar† * World Trade Organization, Geneva, Switzerland † World Bank, Washington, DC, United States CONTENTS 5.1. INTRODUCTION 145 5.2. MAPPING OF INVESTMENT PROVISIONS 147 5.2.1 Definitions and scope 147 5.2.1.1 Definition of investment 147 5.2.1.2 Definition of investor 149 5.2.1.3 Scope of the investment chapter 150 5.2.2 Investment liberalization 150 5.2.3 Investment protection 151 5.2.3.1 Expropriation and compensation 151 5.2.4 Social and regulatory goals 152 5.2.5 Institutional framework and dispute settlement provisions 153 5.3. RESULTS FROM THE MAPPING 154 5.3.1 Global overview 154 5.3.1.1 Definitions and scope 155 5.3.1.2 Investment liberalization 157 5.3.1.3 Investment protection 160 5.3.1.4 Social and regulatory goals 163 5.3.1.5 Institutional framework and transparency 164 5.3.1.6 Dispute settlement 165 5.3.1.7 Enforceability 165 5.3.2 Content of investment provisions in PTAs over time and PTA regional groupings 166 5.3.3 Patterns of investment provisions across regional groupings 170 5.4. CONCLUSIONS 175 ACKNOWLEDGMENTS 175 REFERENCES 176 144 Investment 5.1. INTRODUCTION Over the past 60 years, states have created an extensive network of investment treaties that govern and protect international investment. Increasingly, these stand-alone investment agreements are being replaced by investment provisions in preferential trade agreements (PTAs).The inclusion of an investment chapter in the North American Free Trade Agreement (NAFTA) in the early 1990s provided a template for extending the scope of investment protections in a typical investment treaty to investment liberalization and regulation.1 The entry into force of NAFTA was followed by the establishment of the World Trade Organization (WTO), which for the first time integrated trade in services into the international trade regime, and placed limits on investment measures that might impede trade. Trade in services was liberalized through the General Agreement on Trade in Services (GATS), which provided the legal framework for WTO members to engage in the preferential liberalization of trade in services. Trade-distorting investment among WTO members was limited through the Trade- Related Investment Measures (TRIMS) Agreement. Following the entry into force of NAFTA and the GATS, trade negotiators increasingly began to incorporate into PTAs a broad set of investment provisions that liberalize, protect, and regulate investments. This has resulted in the combination of the investment protection elements traditionally found in bilateral investment treaties (BITs) being merged with the trade liberalization elements found in PTAs. Many PTAs that liberalize trade in services now have a distinct investment chapter that extends coverage of investment beyond the mode 32 services provision of the GATS (establishment of a commercial presence in a partner country) and regulates a broader investment framework that applies to goods, intellectual property, and, depending on how investment is defined, portfolio investment. PTAs are also at the forefront of a trend to incorporate sustainability goals in investment provisions. This chapter explores the evolution of trends and patterns in PTAs’ investment disciplines. The United Nations Conference on Trade and Development (UNCTAD) has carried out an extensive mapping project of international investment agreements (IIAs) that include both BITs and the investment chapters of PTAs. The scope of this study is narrower than UNCTAD’s in that it focuses only on PTAs.3 However, it is more comprehensive in that PTAs have been thoroughly coded, while UNCTAD has coded a relatively small percentage of the agreements included in its database. The template used in this study draws on two previous 1 Given that BITs are by nature bilateral and often time limited, the negotiation of investment provisions in PTAs provided efficiencies, particularly in the case of PTAs involving three or more parties. 2 The GATS defines four modes of service delivery. Mode 1 is defined as the cross-border provision of a service, mode 2 is the consumption of a service abroad, and mode 4 involves the movement of persons providing the service. 3 Data on these agreements are available through UNCTAD’s online database, http://investmentpolicyhub.unctad. org/IIA/mappedContent. 145 Handbook of Deep Trade Agreements studies, one of which developed a framework for the analysis of a sample of 52 PTAs,4 and one of which further developed that framework to analyze around 130 PTAs.5 The template and methodology devised for this study are described in detail in Section 5.2. The chapter analyzes the legal texts of 230 PTAs in force, of which 111 contain substantive provisions on investment.6 These 111 PTAs, most of which came into force over the past two decades (Figure 5.1), were then mapped to a matrix of 57 distinct types of investment provisions. The analysis focuses only on provisions in the investment chapters of PTAs and not on investment- related provisions that might be found in, for example, services chapters (in relation to mode 3), or in chapters with broad social and regulatory provisions related to labor, environment, or sustainable development. Future work within the scope of the World Bank deep integration project may allow for greater analysis of the investment provisions in these other chapters. The rest of the chapter is organized as follows. Section 5.2 describes the criteria used to map the investment provisions contained in the investment chapters of PTAs. Section 5.3 presents the results of the mapping exercise and provides a global overview of the evolution of investment provisions in PTAs. It also offers insights into common characteristics shared by groups of PTAs in particular regions of the world. Section 5.4 concludes. Figure 5.1: Number of PTAs that include investment provisions, 1958-2018 Source: WTO RTA database: http://rtais.wto.org, May 2018. 4 Kotschwar 2009. 5 Chornyi et al. 2016. 6 This set includes PTAs that specifically incorporate a BIT in the text of the PTA. For example, see Article 10.01 of Chile-Central America. If, however, the parties only reaffirm their commitments under a BIT without specifically incorporating it, we do not analyze its provisions. For example, see Article 89 of China-Costa Rica. Given its sui generis form, the EU treaty was not mapped in our analysis. 146 Investment 5.2. MAPPING OF INVESTMENT PROVISIONS This study aims to identify the main elements generally present in investment chapters of PTAs. Its purpose is to facilitate the analysis of trends and patterns in the use of PTAs to regulate investment and the impact of this approach on investment protection and liberalization. The template distinguishes five main categories of investment provisions, following the basic structure of most PTAs: (a) definitions and scope; (b) investment liberalization; (c) investment protection; (d) social and regulatory goals; and (e) institutional aspects and dispute settlement. In keeping with the overall approach of the Deep Integration project, the template includes a series of Yes/No questions signifying whether or not a particular provision is present in a PTA. The coding assigns a value of 1 for a “yes” response and a value of 0 for a “no” response.7 The characteristics of the different categories are discussed below. 5.2.1 Definitions and scope The first section of a PTA’s investment chapter typically stipulates which parties and assets are protected and sets parameters for those protections.The nature and scope of protections have changed over time as countries have responded to shifting definitions, dispute settlement (DS) cases, and other evolving dynamics. One trend has been a tightening of the definitions of “investor” and “investment,” in order to (a) narrow the scope for interpretation of the agreement; (b) limit protections only to investments made in accordance with host country law; (c) introduce certain objective factors to determine when an asset should be protected; and/or (d) exclude some types of assets used for non-business purposes, such as certain commercial contracts, loans and debt securities, and assets used for non-business purposes. Another trend has been to limit the geographical and temporal scope of trade-related measures. Questions in the template on the definition of investment and the scope of the investment agreement are detailed below. 5.2.1.1 Definition of investment How an investment is defined determines which assets receive the protections granted in the investment chapter, and may shape investors’ access to each other’s markets. Traditionally, 7 Unlike other chapters produced as part of the Deep Integration project, this study does not code for whether provisions go beyond WTO commitments on trade-related investment, due to the limited nature of the WTO agreements. For example, while the TRIMS Agreement recognizes that certain investment measures can restrict and distort trade, it does not regulate trade-related foreign investment. Rather, it focuses specifically on investment measures that infringe on Articles III and XI of the General Agreement on Tariffs and Trade (GATT); that is, on measures that discriminate between imported and exported products and/or create import or export restrictions. Another WTO agreement, the General Agreement on Trade in Services (GATS), governs only foreign direct investment (FDI) in services. Given the limited scope of the WTO rules on trade-related investment, a comparison with the investment chapters of PTAs is not of value for this study. 147 Handbook of Deep Trade Agreements investments - defined only as existing and future investments - were protected by bilateral investment treaties. This definition underwent a change in 1960, when the Germany- Malaysia BIT set the open-ended “asset-based” definition and coverage of investment that is still used in most PTAs. The asset-based definition covers every kind of asset, including both foreign direct investment (FDI) and portfolio investment. With the asset-based definition in place, however, other issues have arisen. In some cases, a joint committee’s interpretation of a particular investment provision in a PTA is in conflict with the definition of foreign investment under a party’s domestic laws.8 Such cases have resulted in some countries, in subsequent PTAs, modifying and narrowing the parameters of what constitutes an investment, explicitly setting out exceptions, clarifying conditions, or using specific language to detail the forms of investment covered under their agreements. Other PTAs contain additional provisions on the admission of foreign investments. For example, the investment chapter of the Costa Rica-Mexico Free Trade Agreement (FTA) specifically “...does not include capital movements that are mere financial transactions for speculative purposes, commercial contracts for the sale of goods or services, credits granted to a State, or loans that are not directly related to an investment....”9 The recent European Union (EU)-Republic of Korea FTA, which was concluded after the Lisbon Treaty came into force, also includes specific definitions affecting the scope of each set of rules.10 Other PTAs, such as the US-Canada FTA (US-CFTA), have a narrower, “enterprise-based” definition of investment, which involves the establishment or acquisition of a business enterprise, and allows for a foreign investor to have control over the enterprise. NAFTA, which superseded the US-CFTA, also uses an enterprise-based definition, but a broader, more open-ended one. Parties to PTAs have increasingly sought to strike a balance between having a comprehensive definition of investment and avoiding the coverage of assets that they do not intend to cover.11 Techniques to achieve this balance include: (a) applying protections only to investments made in accordance with host country law; (b) using a closed-list definition instead of an open-ended one; (c) excluding portfolio shares by restricting the asset-based approach to only direct investment; 8 Malik (2009), for example, cites a case in which claimants and respondents differed as to whether a contract for the performance of certain pre-inspection services was held to constitute an investment. Under the respondent’s domestic law, this activity would not have qualified. The joint commission found that it did. 9 See WT/WGTI/W/60 (Communication from Costa Rica to the WTO’s Working Group on the Relationship between Trade and Investment, dated 28 October 1998). 10 The Treaty of Lisbon, which entered into force in December 2009, gave the European Parliament more authority over trade. Before this time, individual EU Member States had negotiated commitments on treatment of investors through BITs, while the European Commission negotiated market access and pre-establishment provisions. Article 207 of the Lisbon Treaty shifted FDI to the exclusive competence of the European Community, bringing it under the umbrella of the common commercial policy. 11 Echandi 2009. 148 Investment (d) introducing investment risk and other objective factors to determine when an asset should be protected PTA; (e) excluding certain types of assets such as certain commercial contracts, certain loans and debt securities, and assets used for non-business purposes; (f) using a more selective approach to intellectual property rights as protected assets; and (g) dealing with the special problems of defining the investment in the case of complex group enterprises as investors.12 Template questions for the definition of investment Does the agreement use a broad, asset-based definition of investment (i.e., the type of definition found in most BITs, in which investment is described as “every kind of asset” or “any kind of asset,” with the listed categories only serving as examples of the types of assets covered)? Does the agreement use an “enterprise-based” definition of investment that applies only to a business or professional establishment in which the investor has majority ownership or exercises control (direct investment)? Does the agreement use a definition of investment that combines elements of both the “asset-based” and “enterprise-based” definitions (mixed definition)? Does the agreement use a definition of investment based on “commercial presence”? Does the definition of investment exclude portfolio investment? 5.2.1.2 Definition of investor How an investor is defined determines who has access to the rights and protections accorded in the PTA’s investment chapter. In some cases, the definition of investor will explicitly exclude or include citizens with dual nationality or those who have given up citizenship in the country negotiating the agreement. Such a provision will, for example, prevent or allow the investor recourse to the agreement’s investor-state dispute settlement (ISDS) mechanism.13 Template questions for the definition of an investor Does the agreement include a definition of “investor”? Rather than defining “investor,” does the agreement define “juridical” and “natural person”? Does the definition of investor cover permanent residents or those who have a “right of abode” (or other similar rights)? Does the definition of investor limit those of dual nationality to be exclusively a national of his or her dominant and effective nationality? Does the definition limit the scope of the term “investor” or “juridical/natural person” to entities engaging in “substantial business activities,” “real economic activity,” or similar term? 12 UNCTAD 2011. 13 Other types of dispute settlement are not included in the index as they are not core to the investment chapter. Investors want ISDS as it enables them to challenge the state if necessary. A state-to-state option does not carry the same weight. 149 Handbook of Deep Trade Agreements 5.2.1.3 Scope of the investment chapter The scope of a PTA’s investment chapter is determined, to begin with, by the number and identity of the states that are party to the PTA, as well as their territorial limits. The scope of the agreement may also refer to the inclusion or denial of benefits to certain parties. For example, the chapter may include provisions denying treaty protection to investors whose home state does not maintain diplomatic relations with the host state. It may also include provisions prohibiting third-country nationals who own or control the investor from gaining access to protection from a treaty to which they are not a party. Template questions defining the scope of an investment chapter Does the agreement contain a denial-of-benefits provision? Does the agreement cover both national and subnational levels? Does the agreement contain provisions in case the investment changes form? 5.2.2 Investment liberalization A significant change in PTAs compared to BITs is the inclusion of market access provisions; that is, obligations for the parties to liberalize their regulatory regimes with respect to foreign investment. A growing number of investment chapters of PTAs include liberalization commitments and extend investor protections to the pre-establishment phase. Parties commit, in these agreements, to remove restrictions on foreign investment in their respective economies and/or to provide protections for foreign investors seeking to enter their markets. Some issues raised by these provisions relate to the challenges of accurately assessing the costs and benefits of liberalizing different sectors and activities, and the extent to which governments can continue to use tools such as investment screens for national security and other reasons. An increasing number of investment chapters in PTAs also include a prohibition on performance requirements (PRs), a trend that began with NAFTA. Some PTAs simply make reference to the Trade-Related Investment Measures (TRIMS) Agreement, which prohibits local content requirements, trade-balancing requirements, export controls, and foreign exchange restrictions related to foreign exchange inflows attributable to an enterprise. Others, such as the PTAs concluded by the United States, Canada, and Japan, explicitly prohibit these requirements. Canada and US PTAs extend this prohibition to the pre-establishment phase. Other PTAs contain special provisions prohibiting nationality requirements for senior management of an enterprise but allowing nationality requirements for a majority of the investment’s board of directors. In some cases, pre-establishment commitments are taken only with respect to sectors/ industries specifically mentioned (positive list), or to all sectors/industries except those specifically excluded (negative list), or to a combination of both lists. 150 Investment Template questions relating to investment liberalization commitments National treatment (NT) Does the agreement provide national treatment (NT) in the pre-establishment/acquisition phase of the investment? Most-favored-nation (MFN) Does the agreement provide most-favored-nation (MFN) treatment in the pre-establishment/acquisition phase of the agreement? Does the agreement grant exceptions to the MFN clause? Performance requirements Does the investment chapter prohibit or limit the use of performance requirements? Senior management/boards Does the investment chapter contain a provision that entitles covered investors to make appointments to senior management positions and/or and members of the board of directors without regard to nationality? Non-derogation Does the investment chapter guarantee that if another international treaty to which the Contracting States are parties, or national legislation of the host State, provides for more favorable treatment of investors/investments, that other treaty or national legislation shall prevail in the relevant part over the provisions of the PTA? Scheduling and reservations Does the investment chapter take a positive-list approach to commitments? Does the investment chapter take a negative-list approach to commitments? 5.2.3 Investment protection Trade protection disciplines aim to afford investors explicit protection of their investments and recourse in case such investments are expropriated or otherwise compromised by the host state. Investment chapters in PTAs continue to emphasize investment protection, setting conditions for the expropriation of assets and the transfer of payments and profits. More recent PTAs also clarify the meaning of provisions dealing with absolute standards of protection, in particular the international minimum standard of treatment in accordance with international law. 5.2.3.1 Expropriation and compensation One of the basic objectives of the investment chapter is to protect the assets of the investor from uncompensated seizure or in the case of conflict in the host country. Modern PTAs typically ban host states from expropriating foreign investment unless the state meets four conditions.The expropriation must be (a) for a public purpose; (b) carried out in a nondiscriminatory manner; (c) in accordance with some kind of legal process; and (d) accompanied by payment of (usually) the full value of the expropriated asset, usually specified as of the date of the expropriation. The investment chapter may also contain provisions regarding acceptable valuation techniques and the payment of interest. Recent PTAs have also introduced clarifying language regarding 151 Handbook of Deep Trade Agreements indirect expropriation through interference with the rights of ownership. The lack of clarity concerning this issue has created difficulties over the past few decades. Many recent PTAs explicitly state that obligations regarding expropriation are intended to reflect the level of protection granted by customary international law. Some also include guidelines and criteria for determining whether an indirect expropriation has taken place in a particular situation. In accordance with customary international law, most investment chapters require that the host state provide foreign investors with fair and equitable treatment (FET) as well as full protection and security.The obligation to accord FET to investors and their property is one of the standards that has been most invoked in investor-state dispute settlement. FET is, in theory, an “absolute,” “non-contingent standard”14 intended to protect investors from arbitrary or discriminatory treatment by the host state. However, the meaning of fair and equitable treatment may not necessarily be the same in all the PTAs in which the phrase appears. Its interpretation may be influenced by the specific wording of the agreement, its context, the parties’ negotiating history, or other indications of their intent. PTAs tend to apply FET in one of several ways: (a) through an unqualified obligation provision; (b) by linking the FET obligation to international law; (c) by linking it to the minimum standard of treatment of aliens under customary international law; or (d) by including additional substantive content, such as the obligation not to deny justice in legal or administrative proceedings. Another element affecting the scope and coverage of trade-related investment in PTAs is the use of the so-called umbrella clause, which brings investor-state contracts under the umbrella of the agreement. The umbrella clause has been used since the 1950s but has gained prominence during the last 20 years as a result of several high-profile disputes.15 5.2.4 Social and regulatory goals In addition to liberalizing and protecting investment, the investment chapters of PTAs also incorporate flexibilities for public policy. Some PTAs include provisions or references to protection of the environment, fundamental labor principles and human rights, compliance with social corporate responsibility standards, or the participation of PTA parties in relevant organizations. A number of recent PTAs include an obligation to prohibit corrupt practices. Many investment chapters also recognize different circumstances and levels of development among parties by including provisions on cooperation and technical assistance. 14 I.e.,a standard that states the treatment to be accorded in terms whose exact meaning has to be determined by reference to specific circumstances of application, as opposed to the “relative” standards embodied in national treatment and MFN principles which define the required treatment by reference to the treatment accorded to other investment. See OECD (2004). 15 UNCTAD (2015) notes that the use of the umbrella class has declined in more recent PTAs. 152 Investment Template questions relating to investor protection National treatment Does the agreement cover the post-establishment phase of the investment? Does the agreement provide MFN treatment in the post-establishment phase of the investment? Minimum standard Does the agreement grant fair and equitable treatment (FET)? of treatment Does the FET clause expressly include a reference to a denial of justice? Does the FET clause prohibit arbitrary, unreasonable, or discriminatory measures? Does the FET clause include an explicit clarification that the breach of another provision in the PTA or a breach of another international agreement by a contracting party will not by itself constitute a breach of the FET standard? Does the FET clause provide that the finding of an FET violation must take into account the level of development of the host country? Does the FET clause reference customary international law? Expropriation Does the investment chapter cover direct expropriation? and compensation Does the investment chapter cover indirect expropriation? Does the provision on expropriation and compensation allow for a carve-out for compulsory licenses? Does the provision on expropriation and compensation allow for a carve-out for subsidies? Does the provision on expropriation and compensation allow for a carve-out for general regulatory measures to protect legitimate public welfare goals? Protection in case Does the clause on protection in case of armed conflict or strife provide of armed conflict for national treatment? or strife Does the clause on protection in case of armed conflict or strife provide for MFN treatment? Does the clause on protection in case of armed conflict or strife provide for compensation? Transfers Does the clause on protection in case of armed conflict or strife provide for the transfer of funds? Umbrella clause Does the investment chapter include an umbrella clause requiring the parties to respect or observe any obligation assumed by them with regard to a specific investment, thereby bringing contractual and other obligations under the umbrella of the PTA? Subrogation Does the investment chapter provide for a mechanism of subrogation, such that if an insurer covers the losses suffered by an investor in the host state, it acquires the investor’s full rights to bring claim? 5.2.5 Institutional framework and dispute settlement provisions The final set of coding questions aims to identify what types of measures are put in place to ensure transparency in the administration of investment provisions, and whether the dispute settlement mechanism includes procedures for investor-state and state-to-state dispute settlement. 153 Handbook of Deep Trade Agreements Template questions relating to social and regulatory goals Social and Does the investment chapter reference the right to regulate? regulatory goals Does the investment chapter refer to protection of the environment? Does the investment chapter refer to protection of human rights? Does the investment chapter contain a reference to labor? Does the investment chapter refer to corporate social responsibility? Does the investment chapter refer to sustainable development? Does the investment chapter refer to corruption? Technical cooperation/ Does the investment chapter include a commitment on technical cooperation? capacity building Does the investment chapter include a commitment on capacity building? Finally, the template includes a cross-cutting category that aims to capture the level of enforceability of the investment chapter. Four categories of enforceability are identified: (a) no dispute settlement (provisions in the investment chapter are not covered by the PTA’s dispute settlement mechanism); (b) diplomatic dispute resolution (the parties refer a dispute to the PTA’s joint committee or other institutional body rather than to an arbitral mechanism); (c) state-to-state arbitral mechanism only; and (d) state-to-state arbitral mechanism plus investor-state dispute settlement. Template questions relating to institutional aspects and dispute settlement provisions Institutional Does the investment chapter establish a joint committee or another framework/committee type of institutional framework? Transparency Does the investment chapter include commitments for prior comment? Does the investment chapter include an agreement to publish laws, regulations, and investment policies that affect investment? Does the investment chapter establish national inquiry points? State-to-state Does the investment chapter include a state-to-state mechanism for dispute settlement (e.g., arbitration) between the contracting parties? Investor-state Does the investment chapter include a mechanism for the settlement of dispute settlement dispute settlement covered investors and the host state (ISDS)? Mechanism for consultations Does the investment chapter include a consultation mechanism? 5.3. RESULTS FROM THE MAPPING This section provides the results of the mapping. It is divided into three sub-sections. The first provides a global overview of provisions in the investment chapter.The second analyzes the content of investment provisions in PTAs over time. The third explores whether PTA regional groupings (such as those that follow the NAFTA model) share common characteristics within regions. 5.3.1 Global overview Historically, investment agreements have focused on the protection of investors and their investments made in accordance with the laws and regulations of the host country. Once 154 Investment established, these investments are typically granted the same treatment as host country and most- favored-nation (MFN) investments. Increasingly, however, the investment chapters of PTAs have been innovating in this area, liberalizing investment flows by granting national treatment (NT) and MFN treatment to foreign investors in the pre-establishment phase, and thereby reducing barriers to the entry of foreign investments. 5.3.1.1 Definitions and scope The definitions set in the investment chapter of a PTA are used to determine what types of investors and investment are covered by the agreement’s investment framework.16 On the definition of investment, the mapping found that about 25 percent of the PTAs use a broad asset-based definition that includes both foreign direct investment (FDI) and portfolio investment (Figure 5.2). For instance, Article 135 of the China-New Zealand PTA states that investment means “every kind of asset invested, directly or indirectly, by the investors of a Party in the territory of the other Party including, but not limited to, the following: (a) movable and immovable property and other property rights such as mortgages and pledges; (b) shares, debentures, stock and any other kind of participation in companies….” This definition is used mostly by the Association of Southeast Asian Nations (ASEAN) and by China in their PTAs. The enterprise-based definition of investment, which was pioneered in NAFTA,17 accounts for fewer than 10 percent of PTAs, primarily those involving Canada. However, almost half the PTAs use elements of the NAFTA definition combined with elements of the asset-based definition of investment.This mixed definition has been used by the United States in all PTAs negotiated after NAFTA. For example, Article 10.27 of the US-Chile agreement defines investment as “every asset that an investor owns or controls, directly or indirectly... Forms that an investment may take include: (a) an enterprise; (b) shares, stock, and other forms of equity participation in an enterprise.…” The mixed definition of investment has also been adopted by a number of Latin American countries as well as Japan and the Republic of Korea. Finally, a definition of investment based on commercial presence, which was inspired by the services liberalization provision of the GATS, is used exclusively by the EU and the European Free Trade Association (EFTA) with third parties.18 This definition accounts for 16 For a full discussion of the scope and definitions of investment see UNCTAD 2011. 17 Article 1139 of NAFTA defines investment as “(a) an enterprise; (b) an equity security of an enterprise; (c) a debt security of an enterprise ….” 18 For example, Article 5.2 of EFTA-Colombia defines “commercial presence” as “any type of business establishment, including through: (i) the constitution, acquisition or maintenance of a juridical person, or (ii) the creation or maintenance of a branch or a representative office, within the territory of another Party for the purpose of performing an economic activity.” 155 Handbook of Deep Trade Agreements fewer than 20 percent of the PTAs surveyed. It is narrower in scope than the other definitions, typically having disciplines that govern market access but not investment protection. The commercial presence definition excludes both portfolio investment and intangible assets such as intellectual property rights.19 Slightly less than a third of the PTAs, mostly those negotiated by ASEAN, Japan, and China, have a provision stating that a change in the form in which assets are invested does not affect their character as investments. Only PTAs using an asset-based or mixed definition contain such a provision. As to the definition of investor, about 88 percent of PTAs use the term “investor” in their investment chapter, while the remaining 12 percent, all from the EU or EFTA, refer to “juridical” or “natural persons.” More than half of all PTAs include in their definition permanent residents and those having the right of abode in the PTA partner; most involve Canada, Chile, EFTA, Japan, New Zealand, and Panama. About 25 percent of the PTAs limit investors of dual nationality to being exclusively a national or his or her dominant and effective nationality. 20 Figure 5.2: Definitions and scope provisions of the investment chapter, share of PTAs (%) Source: Deep Trade Agreements Database. Some PTAs allow the host state to stipulate that only those engaging in substantive (or substantial) business operations (SBO) or with a real and continuous link to an enterprise may 19 Rather than specifically excluding portfolio investment from the definition of investment, some PTAs provide instead that the application of national treatment does not apply to portfolio investment. See, for instance, Article 75.2 of Japan-Malaysia. 20 This is normally the state to which the investor has stronger ties such as those of a personal, economic or political nature. Many of the PTAs of the United States, Canada, and Australia as well as some of the PTAs among Latin American countries have a provision on dual nationality. 156 Investment benefit from the provisions in the investment chapter.21 Thirty-five percent of PTAs contain such a provision; most involve the EU or EFTA, although Japan, India, and Peru are also starting to use SBO provisions. This limitation prevents investors or enterprises from maintaining a mailing address in the host country for the purpose of benefiting from the agreement. Another means for a host state to exclude certain entities from the benefits of the investment chapter is the denial-of-benefits clause.This provision has been used to exclude (a) enterprises having no substantive business operations in the host country’s territory;22 (b) parties that do not own or control an enterprise; (c) a party that does not have diplomatic relations with the host country; (d) prohibited transactions with a non-party to the agreement.23 More than two-thirds of PTAs have a denial of benefits provision. Unlike the substantive business operations clause, which can potentially apply to a broad range of entities, the denial-of-benefits clause is drafted to operate in narrowly defined circumstances, and the host state has the discretion whether or not to exercise it.24 A handful of PTAs, including ASEAN-China, China-Singapore, Panama-Peru, Korea-India, and Japan-Philippines, contain both an SBO clause and a denial-of-benefits clause. Particularly in the case of federal states, the investment chapter specifies the levels of government to which the provisions apply. More than three-quarters of PTAs contain a provision stating that a party’s obligations apply at the national and sub-national levels. Exceptions include some of the PTAs of Japan, Chile, and China. 5.3.1.2 Investment liberalization Unlike most stand-alone investment agreements, PTAs provide for investment liberalization as well as investment protection. Liberalization may include (a) protections during the pre- establishment or entry phase of investment, including national treatment, which requires the host state to remove all discriminatory market access barriers and allow foreign investors to invest on the same terms as domestic investors. Liberalization may also include (b) a prohibition or limitation of performance requirements; and (c) the ability of investors to appoint senior management and members of boards of directors without regard to nationality. 21 For example, see Article 4.2(p), Chapter 4 of EFTA-Colombia, which holds that a “juridical person of another Party” is constituted or otherwise organized under the laws of that other Party and is engaged in substantive business operations. Also, Article 10.1 of India-Republic of Korea defines an enterprise as “constituted or organized under the law of a Party, and its branch [is] located in the territory of a Party and carrying out substantial business activities there.” 22 See Article 11.14 of Costa Rica-Singapore. 23 See Article 10.12 of US-Peru. 24 See Chornyi et al. 2016, p. 15. 157 Handbook of Deep Trade Agreements The investment chapter may also grant exceptions to the MFN clause, which extends non- discriminatory treatment enjoyed by any other non-party to the PTA partner. Finally, the investment chapter may define the approach to scheduling commitments and reservations. The mapping shows that 88 percent of the PTAs provide for national treatment and 77 percent provide for MFN treatment in the pre-establishment/acquisition phase of the investment (Figure 5.3).25 However, more than half the PTAs that grant MFN treatment include one or more general exceptions to the MFN clause. These exceptions can take different forms. Some PTAs take an MFN exception for regional integration such that any preference extended to a third party as a result of engaging in another PTA is not automatically accorded.26 Another form of MFN exception occurs when the parties reserve the right to adopt or maintain any measure according differential treatment to third parties for certain sectors such as fisheries or maritime matters.27 A number of PTAs grant an exception that prevents an investor of a party to a PTA from using the MFN provision to benefit from more favorable conditions of access to investor-state dispute settlement than are provided under another investment agreement to which the investor is a party.28 A less common MFN exception is one in which the parties reserve the right to adopt or maintain any measure that accords differential treatment to (a) socially or economically disadvantaged minorities and ethnic groups; or (b) cultural industries related to the production of books, magazines, periodical publications, or printed or electronic newspapers and music scores.29 In addition to scheduling general exceptions to NT and MFN treatment, states can schedule sector-specific exceptions for lists of non-conforming measures.30 Investment chapters of PTAs also liberalize performance requirements (PRs); i.e., conditions that investors must meet before being allowed to operate a business or benefit from an incentive offered by the host state. The WTO TRIMS Agreement (which applies only to trade-related investment in goods) prohibits the use of domestic content requirements, restrictions on imports and exports related to local production, and foreign exchange restrictions. Some PTAs echo the prescriptions of the TRIMS Agreement by incorporation or reference. Others go beyond the TRIMS Agreement by applying disciplines to PRs for 25 Exceptions to national treatment include Chile-Central America; China’s PTAs with Peru, New Zealand, Pakistan, and Singapore; Colombia-Mexico; and Dominican Republic-Central America. Exceptions to MFN treatment include a number of PTAs of EFTA and India. 26 For example, see Article 96.3 of Japan-Thailand. For a full description of the effects of the regional economic integration organization (REIO) clause in investment agreements, see UNCTAD (2004). 27 For example, see Article 139 of China-New Zealand. 28 For instance, see Article 14.4 of Australia-Japan, which reads “Each Party shall accord to investors of the other Party and to covered investments treatment no less favourable than that it accords, in like circumstances, to investors of a non-Party and to their investments with respect to investment activities in its Area. Note: For greater certainty, this Article does not apply to dispute settlement procedures or mechanisms under any international agreement.” 29 See Article 131 of Peru-China. 30 The analysis did not extend to the examination of lists of non-conforming measures or reservations. 158 Investment Figure 5.3: Investment liberalization provisions, share of PTAs (%) Source: Deep Trade Agreements Database. both goods and services, or by adding additional limitations on, for example, forced technology transfer, the hiring of a certain number or percentage of nationals, or the exclusive supply of the goods or services produced. Two-thirds of PTAs contain disciplines on performance requirements (Figure 5.3). Those of the United States and Canada systematically include these disciplines, as do those of Japan, Korea, and Panama. PTAs without performance requirements include all the EU’s and EFTA’s PTAs with third parties, ASEAN’s PTAs with China and India, and China’s PTAs with Peru, Pakistan, and Singapore. The mapping also showed that 67 percent of PTAs include provisions that allow investors to appoint senior management and boards of directors (SMBD) without regard to nationality. The PTAs of the United States, Canada, Panama, and Australia systematically include such a provision. PTAs without an SMBD provision include those of Japan, the EU, China, and ASEAN. Further 17 percent of PTAs include non-derogation clauses, which guarantee (or do not prevent) an investor’s right to take advantage of another investment treaty between the parties that results in more favorable treatment. These include some PTAs involving the EU, ASEAN, and India. For instance, Article 90 of EU-Ukraine states that “Nothing in this Chapter shall be taken to limit the rights of investors of the Parties to benefit from any more favourable treatment provided for in any existing or future international agreement relating to investment to which a Member State of the European Union and Ukraine are parties.” A different formulation is found in India-Singapore, which states that if the legislation of either party or international obligations existing at present or established thereafter between the parties results in more favorable treatment to investors than the India-Singapore FTA, such position shall not be affected by the agreement.31 31 See Article 6.22 of India-Singapore. 159 Handbook of Deep Trade Agreements Different techniques are used to schedule commitments or reservations in the investment chapters of PTAs. A positive-list approach (similar to that used in the GATS for the scheduling of services commitments) implies that only the sectors listed in the schedule are subject to the agreement’s disciplines on investment, subject to any qualifications contained therein. A negative-list approach provides that the obligations in the investment chapter are applied to all sectors with the exception of those sectors appearing in the list (or lists) of non- conforming measures. A negative-list approach is more common and is used in 85 percent of PTAs surveyed, while the positive-list approach applies in 9 percent. In a few PTAs, one party uses a positive list to schedule commitments while the other uses a negative list for its non- conforming measures.32 5.3.1.3 Investment protection All agreements assessed in this exercise grant national treatment to investors from partner countries. While MFN treatment has generally gone hand-in-hand with NT (it was included in the first bilateral investment treaty, in 1959, between Germany and Pakistan), a number of PTAs that include NT exclude MFN or modify its application. As noted above, this exclusion is often done with the intention of preventing claimants from invoking treaties with third parties that include potentially more favorable provisions relating to protection standards or ISDS.33 Agreements that exclude MFN tend to be between the EU or EFTA and Latin American countries, or between Asian countries. The 2009 Economic Partnership Agreement (EPA) between Japan and Switzerland is one such example, in which parties pledge to make best efforts to accord each other any more favorable treatment granted under other agreements but explicitly exclude this as an obligation.34 In the ASEAN-Australia-New Zealand FTA (2010), the applicability of MFN is an item to be negotiated in the future.35 Most PTAs also cover direct and indirect expropriation, fair and equitable treatment, and currency transfers (Figure 5.4). Fair and equitable treatment is incorporated into 76 percent of the PTAs surveyed. FET can take different forms, each with implications regarding its scope and content.The most important distinction arises between FET provisions that are explicitly linked to the minimum standard of 32 See for example India-Korea and India-Singapore. 33 This is sometimes called “treaty shopping.” 34 Article 88 of the Japan-Switzerland EPA states that “If a Party accords more favourable treatment to investors of a non-Party and their investments by concluding or amending a free trade agreement, customs union or similar agreement that provides for substantial liberalisation of investment, it shall not be obliged to accord such treatment to investors of the other Party and their investments. Any such treatment accorded by a party shall be notified to the other Party without delay and the former Party shall endeavour to accord to investors of the latter Party and their investments treatment no less favourable than that accorded under the concluded or amended agreement.The former Party, upon request by the latter Party, shall enter into negotiations with a view to incorporating into this Agreement treatment no less favourable than that accorded under such concluded or amended agreement.” 35 Article 16 (2)(a) of the Investment chapter. 160 Investment Figure 5.4: Investment protection provisions, share of PTAs (%) Source: Deep Trade Agreements Database. treatment under customary international law, and those that include an unqualified formulation of the obligation. More recent PTAs have started to include some additional language in their investment chapter to clarify the meaning of the obligation. Of the agreements that include FET, about 66 percent reference international law, 65 percent prohibit breach of other treaty obligations, and 53 percent explicitly reference denial of justice (Figure 5.5). Figure 5.5: Fair and equitable treatment provisions, share of PTAs (%) Source: Deep Trade Agreements Database. Note: IIA = international investment agreement. One of the main purposes of international investment agreements is to protect investors’ assets. The bulk of the agreements provide for compensation and promise national treatment in case of armed conflict or strife in the host country. A smaller percentage, about half, guarantee the transfer of foreign investors’ funds in such cases. Most of the agreements in the sample include provisions against direct (75 percent) or indirect (74 percent) expropriation. Agreements by Central European countries and a majority of the EFTA agreements, which group investment together with services, do not include expropriation commitments (Figure 5.6). 161 Handbook of Deep Trade Agreements Figure 5.6: Provisions regarding protection of assets, share of PTAs Source: Deep Trade Agreements Database. Within the agreements containing provisions on direct expropriations, about 95 percent include carve-outs to protect public welfare goals, about 72 percent include carve-outs for compulsory licenses, and 8 percent include carve-outs for subsidies provided by the host country (Figure 5.7). An important protection for investors is the ability to transfer funds freely and under reasonable conditions. This topic is discussed in depth in Chapter 9 on movement of capital. Just over half (56 percent) of the agreements in the sample provide protection for transfer of funds. An umbrella clause provides that parties uphold any contractual obligation (beyond the PTA itself) that they have entered into with regard to investments of nationals of their PTA partners, thus broadening the coverage of a PTA and its dispute settlement mechanism. Only 7 percent of agreements - half of them in the East Asia and Pacific region - contain an umbrella clause. Figure 5.7: Provisions regarding protection of assets, share of PTAs Source: Deep Trade Agreements Database. 162 Investment 5.3.1.4 Social and regulatory goals Provisions under this heading may include the right to regulate, protection of the environment, human rights, labor, corporate social responsibility, sustainable development, and corruption. The right to regulate refers to the balance between investor protection and the state’s right to protect legitimate policy interests such as national security, public health and safety, or the environment. It has been the subject of some controversy.36 The right-to-regulate provision in investment chapters takes different forms. It is sometimes connected with a specific provision such as performance requirements. For instance, in US-Panama, the parties provide for certain exceptions from the general proscription on performance requirements to permit the adoption or maintenance of measures “necessary to secure compliance with laws and regulations that are not inconsistent with this Agreement; necessary to protect human, animal or plant life or health; or related to the conservation of living or non-living exhaustible natural resources.”37 In the PTAs of the United States, the provision on performance requirements contains a right to regulate exception, but there is no broad right to regulate provision in the investment chapter. The right to regulate is also associated with provisions on expropriation. For instance, in Japan’s PTAs, a party shall not expropriate or nationalize investments of the other party (or take measures tantamount to expropriation) except “for public purpose.”38 Beyond those linked specifically to performance requirements or expropriation, 73 percent of PTAs include a right to regulate provision in the public interest (Figure 5.8). Some PTAs incorporate elements of GATS Article XIV on exceptions, either directly in the investment chapter or through a general exception linked to the investment chapter.39 Others have more specific exceptions. In EU-Ukraine, for instance, the parties “retain the right to regulate and to introduce new regulations to meet legitimate policy objectives provided they are compatible with this Chapter.”40Another example is India-Singapore, which allows the parties to adopt, maintain or enforce any measure, on a non-discriminatory basis, that is consistent with the investment chapter and is “in the public interest, including measures to meet health, safety or environmental concerns.”41 36 For the debate on the right to regulate, see Gaukrodger 2017 and UNCTAD 2012. 37 See Article 10.9 of US-Panama. 38 We capture the carve-out for general regulatory measures to protect legitimate public welfare goals in the section on investment protection. 39 See Article 11 of Japan-Indonesia, in which, for the purposes of the investment chapter, “Articles XIV and XIV bis of the GATS are incorporated into and form part of this Agreement, mutatis mutandis.” 40 See Article 85.4 of EU-Ukraine. 41 See Article 6.10 of India-Singapore. 163 Handbook of Deep Trade Agreements Environmental concerns are subject to the right to regulate in 85 percent of all investment chapters.42 A common formulation is that the parties recognize that it is inappropriate to encourage investment by relaxing their environmental measures. For instance, in Chile-Japan the parties agree not to waive or otherwise derogate from environmental measures as an encouragement for establishment, acquisition, or expansion of investments.43 In EU-Colombia-Peru, subject to the requirement that such measures are not applied in a manner which would constitute arbitrary or unjustifiable discrimination between the parties, the parties may adopt measures “necessary to protect human, animal or plant life or health, including those environmental measures…relating to the conservation of living and non-living exhaustible natural resources, if such measures are applied in conjunction with restrictions on domestic investors or on the domestic supply or consumption of services.”44 References to the right to regulate in the areas of human rights, labor, corporate social responsibility, sustainable development, and corruption appear in fewer than 20 percent of investment chapters,45 although they may appear in other parts of the PTA not coded during this exercise. There were also few direct references to technical cooperation and capacity building in the investment chapter, although they may appear elsewhere in the PTA. 5.3.1.5 Institutional framework and transparency All PTAs establish some sort of administrative body charged with monitoring and implementing the agreement. In addition, about one-quarter of PTAs establish a specific committee responsible for monitoring implement of the investment chapter. For instance, the ASEAN-Australia-New Zealand agreement establishes a Committee on Investment to review implementation of the investment chapter, consider any matters referred to it, and report to the PTA’s joint committee as required.46 Japan’s PTAs also typically establish such a committee, while those of the EU, the United States, and Canada do not. With regard to transparency provisions, the analysis looked at whether the investment chapters include commitments for prior comment, agreements to publish, or the establishment of national inquiry points. The mapping found that about two-thirds of investment chapters establish a national inquiry point, while commitments for prior comment occur in 38 percent of PTAs and agreements to publish in just over 20 percent (Figure 5.9). 42 See, for instance, Article 10.11 of Nicaragua-Taiwan, China, which reads “Nothing in this Chapter shall be construed to prevent a Party from adopting, maintaining, or enforcing any measure otherwise consistent with this Chapter that it considers appropriate to ensure that investment activity in its territory is undertaken in a manner sensitive to environmental concerns.” 43 See Article 87 of Chile-Japan. 44 See Article 167 of EU-Colombia-Peru. 45 Provisions on labor may be contained in a separate chapter. For details, see Chapter 12 in this volume. 46 See Article 11.17 of ASEAN-Australia-New Zealand. 164 Investment Figure 5.8: Selected social and regulatory goals in the investment chapter, share of PTAs (%) Source: Deep Trade Agreements Database. Figure 5.9: Selected transparency provisions, share of PTAs (%) Source: Deep Trade Agreements Database. 5.3.1.6 Dispute settlement Dispute settlement is a key provision in investment chapters, particularly investor-state dispute settlement (ISDS) provisions, which allow investors to bring disputes regarding the treaty’s substantive provisions. Almost all PTAs provide for a mechanism for consultations and state-to-state dispute settlement, and 77 percent provide for ISDS (Figure 5.10). 5.3.1.7 Enforceability All 111 PTAs in the study have enforceable investment chapters, but the quality of enforcement varies depending on the forum and agency responsible. In the Australia-New Zealand Closer Economic Agreement (ANZCERTA), for example, the investment chapter does not provide access to the agreement’s state-to-state arbitral mechanism, and there is no ISDS provision. In the case of an investment dispute, investors need to seek recourse in a domestic court.47 Of 47 See Trakman 2014. 165 Handbook of Deep Trade Agreements the other 110 PTAs in the study, 77 percent have investment chapters that provide access to both a state-to-state arbitral mechanism and ISDS, and 19 percent have chapters that provide access to a state-to-state arbitral mechanism without ISDS. The remainder give the parties access to dispute resolution using diplomatic channels under the agreement. The EU and EFTA for the most part provide access only state-to-state arbitral mechanisms in their investment chapters, although EFTA’s PTAs with Singapore and Korea have ISDS. Australia went through a period of not including an ISDS provision in its PTAs but has accepted it in recent PTAs.48 Figure 5.10: Dispute settlement mechanisms, share of PTAs Source: Deep Trade Agreements Database. 5.3.2 Content of investment provisions in PTAs over time and PTA regional groupings To assess the evolution of investment chapters over time, the study generated a variable that captures the share of questions included in the template to which there were “yes” answers. The analysis found that average number of investment provisions in PTAs has increased since 1995. While agreements entering into force between 1995 and 1999 included, on average, 24 provisions, those entering into force in the period 2010-2017 averaged 27 (Figure 5.11). In the Western Hemisphere, the United States, Canada, and Peru are party to PTAs with the largest number of investment provisions (Figure 5.12), averaging 28-35 provisions across their PTAs. Those of Mexico, the Central American countries, Colombia, and Chile have PTAs averaging 21-28 provisions. In European PTAs, the average number of investment 48 In Australia’s PTAs with Japan (2015), Malaysia (2013), and the United States (2005) the parties have access only to a state-to-state arbitral mechanism, while in those with ASEAN-New Zealand (2010), Chile (2009), CPTPP (2018), Korea (2014), Singapore (2003), and Thailand (2005) an ISDS mechanism is also available. 166 Investment Figure 5.11: Boxplot of average number of Investment provisions in new PTAs over time Source: Deep Trade Agreements Database. Note: A boxplot is a standardized way of displaying the distribution of data based on the five-number summary: minimum, first quartile, median, third quartile, and maximum.The central rectangle spans the first quartile to the third quartile, the bold segment inside the rectangle shows the median, and “whiskers” above and below the box show the locations of the minimum and maximum. Outliers are plotted as individual points provisions is smaller (14-21 provisions), reflecting the absence of investment protections. Although some of EFTA’s PTAs have investment protection provisions, those of the EU do not. In Africa, the Economic Community of West African States (ECOWAS) and Morocco have a high average number of investment provisions (based on a single PTA). Most Asian countries have PTAs averaging 28-35 provisions. Figure 5.12: Average number of Investment provisions by country Source: Deep Trade Agreements Database. 167 Handbook of Deep Trade Agreements The fact that the number of investment provisions has increased over time does not necessarily mean they have become deeper in terms of content.49 To analyze the evolution of core (or deep) investment provisions over time, a simple index of 12 liberalization and protection provisions was constructed.50 The value of the index varies between 0 and 11.51 Broad definitions of investment and investor, and liberalization disciplines that apply in pre-establishment phase, are indicative of key provisions that enhance investors’ market access. Strong investor protections in the form of post- establishment national and MFN treatment, disciplines on fair and equitable treatment, expropriation, and investor-state dispute settlement are key provisions in the post-establishment phase. The incidence of core liberalization and protection provisions varies by country income groups.52 In South-South PTAs, the incidence of core investment liberalization provisions is low (17 percent) while protection provisions are high (100 percent).53 For North-North and North-South PTAs, the shares are roughly equal, with investment protection provisions, on average, more prevalent than those liberalizing investment (Figure 5.13).54 Figure 5.13: Incidence of core investment provisions (average) Source: Deep Trade Agreements Database. Note: Average over agreements in force during 2017. 49 Indeed, some of the provisions are for exceptions or carve-outs (e.g., to MFN and FET provisions) rather than liberalization efforts. 50 The core provisions in investment liberalization are (a) broad asset-based definition of investment; (b) elements of both the asset-based and enterprise-based definitions of investment; (c) definition of investors that covers permanent residents or right of abode; (d) national treatment (NT) in pre-establishment/acquisition phase; (e) MFN treatment in pre-establishment/acquisition phase; and (f) no performance requirement for senior management positions and boards of directors. For investment protection, the core provisions are (a) NT in the post-establishment phase; (b) MFN in the post-establishment phase; (c) fair and equitable treatment (FET); (d) protections against direct expropriation; (e) protections against indirect expropriation; and (e) ISDS. 51 As each PTA has only one definition of investment: either (a) asset-based or (b) asset-based and enterprise-based. 52 For Figures 5.13-5.15, North and South countries are defined following the World Bank country classification for 2017. South countries are composed of low-income and lower-middle-income economies, while North countries have upper-middle-income and high-income economies. Low-income economies are defined as those with a GNI per capita, calculated using the World Bank Atlas method, of US$1,005 or less in 2016; lower-middle- income economies are those with a GNI per capita between US$1,006 and US$3,955; upper-middle-income economies are those with a GNI per capita between US$3,956 and US$12,235; high-income economies are those with a GNI per capita of US$12,236 or more. 53 In the South-South category, there is one PTA in the sample. 54 In the North-North category, there are 72 PTAs, and in North-South category, there are 38 PTAs. 168 Investment The incidence of core investment liberalization provisions in South-South PTAs is constant over the three-decade period (reflecting the single PTA in the sample), while that of North- North PTAs has increased slightly (Figure 5.14). Since 2000, the incidence of investment liberalization provisions in North-South PTAs has increased slightly as well.55 Figure 5.14: Rising incidence of investment liberalization provisions over time, by level of development Source: Deep Trade Agreements Database. Note: Average over agreements in force during 2017. For investment protection, the incidence of core provisions in South-South PTAs is constant, while for North-North PTAs there is little fluctuation (Figure 5.15). Since 2000, the share of investment protection provisions in North-South PTAs has declined.56 Figure 5.15: Evolution of share of Investment protection provisions over time, by level of development Source: Deep Trade Agreements Database. Note: Average over agreements in force during 2017. 55 The period 1995-1999 has only a single PTA in the North-South category. 56 Only one PTA falls in the North-South category in the period 1995-2000. 169 Handbook of Deep Trade Agreements The share of investor-state dispute settlement provisions in South-South PTAs has remained constant, reflecting the single PTA in the sample. Since 2000, the incidence of North-South and North-North PTAs with ISDS provisions has fluctuated slightly (Figure 5.16). Figure 5.16: Evolution of share of ISDS provisions over time, by level of development Source: Deep Trade Agreements Database. Note: Average over agreements in force during 2017. 5.3.3 Patterns of investment provisions across regional groupings Tables 5.1-5.6, show patterns of investment provisions across geographic regions, the EU, and EFTA for the core issue areas (scope and definition of investment, national treatment, MFN treatment, protection against expropriation, social and regulatory goals, transparency, dispute settlement). The number of PTAs in each grouping varies considerably, from a single PTA in Sub-Saharan Africa (SSA) to 57 PTAs involving Latin American and Caribbean (LAC) countries. Inter-regional PTAs are repeated in two (or more) groupings.57 The most common provisions (those that occur in more than 60 percent of cases) are shaded in the darkest green, the least common (those occurring in less than 40 percent of cases) are shaded in the lightest green, and the rest (occurring between 40 and 60 percent of cases) are shaded in the middle shade of green. Definition of investment. In North America and East Asia and the Pacific (EAP), the mixed definition of investment (asset and enterprise based) is the most commonly used, reflecting the adoption of the NAFTA-type investment chapter in these regions (Table 5.1). In South Asia the broad asset-based definition is predominant, while in the EU and Central Asia, a definition of investment based on commercial presence is the norm. The exclusion of portfolio investment from the definition of investment occurs most frequently in PTAs involving the EU, Sub-Saharan Africa, EFTA, and Central Asia (and tends to go together with a definition of investment based on commercial presence). The PTAs of the EU, EFTA, and Central Asia define juridical persons rather than investors in their PTAs, while a majority of 57 For instance, the CPTPP agreement (Australia, Brunei Darussalam, Canada, Chile, Japan, Malaysia, Mexico, New Zealand, Peru, Singapore,Vietnam) appears in three regional groupings – North America, LAC, and East Asia and Pacific. 170 Investment PTAs involving East Asia and the Pacific, North America, and EFTA broaden their definitions of investor to include permanent residents. The scope of the investment framework can be tailored by the use of a denial of benefits or SBO clause (rarely both). Regional groupings that have adopted the NAFTA-type investment chapter show a preference for a denial-of- benefits clause, while a provision limiting investors to SBO predominates in the PTAs of the EU, EFTA, Central Asia, and Sub-Saharan Africa. Table 5.1: Patterns of investment provisions across regions (%), scope and definitions European Sub- East North Latin South Middle EFTA Central 1. SCOPE AND DEFINITIONS Union Saharan Asia & America America & Asia East & Asia Africa Pacific Caribbean North Africa Number of agreements with investment (14) (1) (55) (18) (57) (7) (4) (10) (9) Broad asset-based definition 0 0 33 0 18 71 0 20 11 Enterprise-based definition 0 100 2 33 16 0 0 0 0 Mixed definition of investment 0 0 62 67 53 29 50 20 0 Commercial presence definition 93 0 4 0 14 0 50 60 78 Portfolio investment excluded 93 100 4 0 14 0 50 60 78 Definition of Investor 43 100 98 100 91 100 75 40 33 Definition of juridical persons 57 0 2 0 9 0 25 50 67 Permanent residents included 0 0 65 56 49 29 0 70 11 Dual nationals/dominant 0 0 25 78 35 14 50 0 0 Investors limited to SBO 93 100 20 0 30 43 50 100 89 Denial of benefits 0 0 84 100 75 71 50 0 0 National & subnational levels 50 0 85 100 79 86 75 70 44 Investment changes form 0 0 51 0 18 0 10 0 Prudential carve-out 93 0 31 0 16 29 50 50 89 Source: Deep Trade Agreements Database. Note: Share of agreements is calculated over the number of agreements in force with investment provisions during 2017. Increasing foreign investment through the reduction of barriers is a key motivation for the inclusion of investment provisions in PTAs. States encourage FDI by granting foreign investors the same treatment as domestic investors (through national treatment provisions) or as investors of any other state (through MFN treatment). States grant each other national treatment on the entry of investment (i.e., in the pre-establishment phase) in all regions except Sub-Saharan Africa, where states retain discretion regarding the entry of investment (Table 5.2). In all the PTAs in North America, the Middle East and North Africa (MENA), and Central Asia, foreign investors benefit from the same treatment as domestic and other third- party investors. In other regions fewer PTAs grant MFN treatment in the pre-establishment phase, thus retaining policy space. Additional policy space is created through the inclusion of an MFN exceptions clause, which is used in most regional groupings except MENA. PTAs involving the EU and Central Asian countries do not have provisions prohibiting performance requirements, in contrast to regions such as North America, where such provisions are the norm. All North American PTAs contain a provision entitling covered investors to make appointments to senior management positions and/or the board of directors without regard to nationality. This practice has been adopted most frequently by PTAs in LAC and MENA. 171 Handbook of Deep Trade Agreements Table 5.2: Patterns of investment liberalization (percentage of PTAs by provision and region) 2. SCOPE AND DEFINITIONS European Sub- East North Latin South Middle EFTA Central Union Saharan Asia & America America & Asia East & Asia Africa Pacific Caribbean North Africa Number of agreements with investment (14) (1) (55) (18) (57) (7) (4) (10) (9) NT pre-establishment 100 0 91 100 84 86 100 100 100 MFN pre-establishment 79 0 75 100 74 14 100 40 100 Exceptions to MFN clause 29 100 45 44 46 29 0 40 33 Performance requirements 0 100 76 100 75 57 50 10 0 Senior management/boards 43 100 55 100 81 29 75 60 56 Non-derogation 50 0 16 0 14 43 25 20 33 Positive-list scheduling 43 0 9 0 7 14 0 0 11 Negative-list scheduling 36 100 91 100 88 86 100 100 44 Source: Deep Trade Agreements Database. Note: NT = national treatment. Share of agreements is calculated over the number of agreements in force during 2017. Table 5.3 shows a significant variation between the European-type agreements, signed by the EU and EFTA countries, and the North American model. The protection provisions used by LAC and EAP groupings tend to follow the North American model. Asian and Table 5.3: Patterns of investment protection (percentage of PTAs by provision and region) 3. INVESTMENT PROTECTION European Sub- East North Latin South Middle EFTA Central Union Saharan Asia & America America & Asia East & Asia Africa Pacific Caribbean North Africa Number of agreements with investment (14) (1) (55) (18) (57) (7) (4) (10) (9) NT post-establishment 100 100 100 100 100 100 100 100 100 MFN post-establishment 79 100 82 100 88 43 100 40 100 Fair and equitable treatment 7 100 93 100 84 86 50 40 22 FET clause refers to denial of justice 0 100 55 72 40 29 50 0 0 FET clause prohibits arbitrary measures 7 0 5 0 4 0 0 20 11 Breach of another IIA,not a breach of FET 0 0 64 89 53 43 50 0 0 FET violation development aspects 0 0 0 0 0 0 0 0 0 FET refers to customary intl. law 0 100 64 89 53 57 50 0 0 Direct expropriation 0 100 96 100 86 100 50 30 0 Indirect expropriation 0 100 95 100 86 100 50 30 0 Expropriation c/o compulsory licences 0 100 65 94 61 86 50 0 0 Expropriation c/o for subsidies 0 0 20 0 0 14 0 0 0 Expropriation c/o regulatory measures 0 100 93 100 86 86 50 30 0 Armed conflict provides for NT 0 100 89 100 86 100 50 30 0 Armed conflict provides for compensation 0 0 91 100 84 100 50 30 0 Armed conflict provides for transfer of funds 0 0 76 61 44 100 50 30 0 Transfers 0 100 80 89 49 86 50 20 0 Umbrella clause 7 100 11 0 0 14 0 30 11 Subrogation 0 0 85 39 47 100 0 30 0 Source: Deep Trade Agreements Database. Note: NT = national treatment; IIA = international investment agreement. Share of agreements is calculated over the number of agreements in force with investment provisions during 2017. 172 Investment EFTA agreements are more likely to include umbrella clauses (though they remain rare). The Central Asian PTAs offer FET but no provisions on expropriation, and otherwise are similar to the European PTAs. All PTAs surveyed offer national treatment on investments once established; North American, MENA, and Central Asia also all offer MFN protection on the operation and management of investments. This latter category varies across region, however, with South Asian and EFTA agreements providing MFN treatment in less than half the PTAs surveyed. Most PTAs provide for national treatment in the case of armed conflict or strife and for compensation should this happen. The sole SSA agreement does not include a provision for compensation or protection of transfers in the case of armed conflict or strife, whereas the majority of PTAs in North America, EAP, South Asia, and LAC do so. Over 80 percent of EAP, North American, and South Asian agreements – and 100 percent of SSA – protect companies’ ability to transfer funds, but less than half of PTAs in LAC do so in the investment chapter. Both direct and indirect expropriation is covered in nearly all non- European agreements, although it is covered in only two of the four MENA PTAs. Table 5.4 shows the considerable regional variation regarding the treatment of social and regulatory issues in investment chapters. A general right to regulate provision (not linked to performance requirements or expropriation provisions) is the norm in EFTA and EU PTAs and common in EAP and South Asia PTAs. PTAs in North America and Sub-Saharan Africa are outliers. The single SSA PTA scores highly on other social and regulatory goals. Apart from a reference to protection of the environment the PTAs of South and Central Asia and MENA do not include provisions on other social and regulatory goals. Table 5.4: Patterns of investment provisions across regions (%) – social and regulatory goals 4. SOCIAL AND European Sub- East North Latin South Middle EFTA Central REGULATORY GOALS Union Saharan Asia & America America & Asia East & Asia Africa Pacific Caribbean North Africa Number of agreements with investment (14) (1) (55) (18) (57) (7) (4) (10) (9) General right to regulate provision 86 0 84 39 65 86 50 100 78 Protection of the environment 50 100 93 100 84 86 75 100 44 Protection of human rights 0 100 2 28 7 0 0 0 0 Reference to labor 7 100 11 28 11 0 0 10 0 Reference to corporate social resp. 0 100 4 33 9 0 0 0 0 Reference to sustainable development 14 100 0 0 4 0 0 0 0 Reference to corruption 0 100 4 28 9 0 0 0 0 Technical cooperation 50 100 15 0 14 14 25 0 22 Capacity building 7 100 4 0 2 14 0 0 0 Source: Deep Trade Agreements Database. Note: Share of agreements is calculated over the number of agreements in force during 2017. 173 Handbook of Deep Trade Agreements Table 5.5 shows patterns of institutional frameworks and transparency across regions. The creation of a specific committee by the investment chapter is most common in EAP and South Asia. Provisions allowing for prior comment on laws and regulations affecting the investment chapter occur in the PTAs of EAP, LAC, and South Asia and to a lesser extent in North America. An agreement to publish such laws and regulations occurs more frequently though not in EU and MENA PTAs.The establishment of national enquiry points to respond to investment-related queries is common in PTAs across all regions, though less frequent in EU, EFTA, and Central Asian PTAs. Table 5.5: Patterns of institutional framework (percentage of PTAs by provision and region) 5. INSTITUTIONAL FRAMEWORK European Sub- East North Latin South Middle EFTA Central AND TRANSPARENCY Union Saharan Asia & America America & Asia East & Asia Africa Pacific Caribbean North Africa Number of agreements with investment (14) (1) (55) (18) (57) (7) (4) (10) (9) Does the investment chapter 7 0 40 11 18 43 0 10 0 establish a committee? Commitments for prior comment 0 0 51 6 47 43 0 0 0 Does the investment chapter 0 100 29 6 7 57 0 40 11 include agreements to publish? Does the investment chapter 43 100 69 83 72 71 75 20 33 establish national enquiry points? Source: Deep Trade Agreements Database. Note: Share of agreements is calculated over the number of agreements in force with investment provisions during 2017. Finally, as seen in Table 5.6, nearly all agreements provide for state-to-state dispute settlement and many provide for investor-state dispute settlement. The PTAs of the EU and Central Asia are the exception, though the EU has begun to include ISDS provisions in its more recent PTAs. Most PTAs across all regions also provide for a mechanism for consultations prior to launching ISDS or state-to-state proceedings. Table 5.6: Dispute settlement provisions (percentage of PTAs by provision and region) 6. DISPUTE SETTLEMENT European Sub- East North Latin South Middle EFTA Central Union Saharan Asia & America America & Asia East & Asia Africa Pacific Caribbean North Africa Number of agreements with investment (14) (1) (55) (18) (57) (7) (4) (10) (9) State-to-state dispute settlement 79 100 96 100 98 100 100 100 67 Investor-state dispute settlement 0 100 87 94 86 100 50 30 0 Mechanism for consultations 64 100 96 100 100 100 75 90 67 Source: Deep Trade Agreements Database. Note: Share of agreements is calculated over the number of agreements in force with investment provisions during 2017. 174 Investment 5.4. CONCLUSIONS This chapter presents a new dataset on the content of investment provisions in PTAs. It covers a total of 111 PTAs that entered into force between 1960 and 2017 and include distinct investment provisions. The analysis of this dataset reveals the following patterns: • The scope and depth of investment provisions has increased over time, although at a modest rate. • Most PTAs extend national and MFN treatment in the pre- establishment phase, while all provide for national treatment (and to a lesser extent MFN treatment) in the post-establishment phase. • A majority of PTAs offer investment protections in the form of provisions on expropriation and fair and equitable treatment. • The majority of PTAs include a broad “right to regulate” provision that allows the host state to override investment provisions for public interest or national security purposes. • Provisions aimed at protection of the environment occur in more than three-quarters of PTAs. • More than three-quarters of PTAs provide for investor-state dispute settlement. • PTA regional groupings demonstrate a number of common characteristics, particularly with regard to provisions on scope and definitions and investment liberalization and protection. Further research is needed to analyze the schedules of investment commitments and to further develop the indicators of deep liberalization. ACKNOWLEDGMENTS The authors would like to acknowledge the support of Claudia Hofmann and Alvaro Espitia Rueda in the coding of the agreements and preparation of graphics and Rohini Acharya, Nadia Rocha, and Michele Ruta for their comments. We would also like to thank Theodore Moran and participants in the World Bank conference “Evolution of Deep Trade Agreements” for their comments. 175 Handbook of Deep Trade Agreements REFERENCES Chornyi, V., M. Nerushay, and J. Crawford. 2016. A survey of investment provisions in regional trade agreements. WTO Staff Working Paper ERSD-2016-07, World Trade Organization, Geneva. Echandi, Roberto. 2009.The new generation of international investment agreements: Recent developments in the Asia- Pacific region. Asia Europe Journal 7: 127-44. Gaukrodger, D. 2017. The balance between investor protection and the right to regulate in investment treaties: a scoping paper. OECD Working Paper on International Investment 2017/02, OECD Publishing, Paris. Kotschwar, Barbara. 2009. Mapping investment provisions in regional trade agreements: Towards an international investment regime? In Regional Rules in the Global Trading System, edited by A. Estevadeordal, K. Suominen, and R. Teh. Cambridge: Cambridge University Press. OECD (Organisation for Economic Co-operation and Development). 2004.“Fair and Equitable Treatment Standard in International Investment Law.” OECD Working Papers on International Investment 2004/3, OECD Publishing, Paris. Trakman, L. 2014. Resolving investor-state disputes: Australia’s dilemma and choices. In Nye Perram, International Commercial Law and Arbitration: Perspectives (2014). UNCTAD (United Nations Conference on Trade and Development). 2004. The REIO exception in MFN treatment clauses. Geneva: United Nations. ———. 2011. UNCTAD Series on International Investment Agreements II: Scope and Definition. Geneva: United Nations. ———. 2015. Recent trends in IIAs and ISDS. UNCTAD IIAs Issues Note No 1 (February). Yannaca-Small, K. 2006. Interpretation of the umbrella clause in investment agreements. OECD Working Papers on International Investment 2006/03, OECD Publishing, Paris. 176 Movement of Capital CHAPTER 6 Movement of Capital D. Siegel, K. Gallagher, and R. Thrasher 177 Handbook of Deep Trade Agreements CHAPTER 6 Movement of Capital D. Siegel*, K. Gallagher†, and R. Thrasher† * Formerly of the International Monetary Fund, Washington, DC, United States † Global Development Policy Center, Boston University, Boston, United States CONTENTS 6.1. INTRODUCTION 179 6.2. THE DATABASE: METHODOLOGY FOR MAPPING 180 6.2.1 Structure 180 6.2.2 Commitments, key issues, and policy implications 180 6.2.2.1 The basic transfer obligation 180 6.2.2.2 Qualifications to free transfer rules – broad exclusions 185 6.2.2.3 Qualifications to free transfer rules – exceptions for macro-economic crises 185 6.2.2.4 Financial sector safeguards and exceptions for prudential measures 189 6.2.2.5 Qualifications to free transfer rules – sector-specific or general exceptions 191 6.2.2.6 Enforcement through dispute resolution 191 6.3. FREE TRANSFERS COMMITMENTS: SOME TRENDS 194 6.4. CONCLUSIONS 196 ACKNOWLEDGMENTS 196 REFERENCES 197 ANNEX 199 178 Movement of Capital 6.1. INTRODUCTION Capital flows are an increasingly important means of allocating savings, promoting growth, and facilitating balance-of-payments (BOP) adjustment. How preferential trade agreements (PTAs) regulate such flows impacts both commercial decisions on foreign investment and broad, multilateral policy decisions. Many authors have reviewed this topic from a policy and economic perspective.1 This study, in contrast, compares the provisions regarding transfers and capital flows in a wide range of PTAs, and offers some initial impressions of patterns and practices, with the aim of providing useful information for strategic business decisions and policy analysis. In the absence of a generalized multilateral agreement, the transfers and capital controls provisions in PTAs form a patchwork of obligations among participating countries. One important theme that emerges from this review is the tension between the market access goals of PTAs and the absence of a coherent multilateral regime to oversee the international effects of these provisions. In broad strokes, PTAs seek to promote the commercial interests of the parties by creating greater certainty for manufacturers, foreign investors, and service providers through increasingly sophisticated instruments. Provisions on transfers and capital flows have a legitimate role in fostering confidence about operating in new markets by reducing risks. However, there has been little analysis of how countries’ obligations are impacted by participating in several PTAs with different provisions on transfers and capital flows, particularly if restrictive rules in one treaty undermine more permissive rules in another. This chapter draws on a sample from the World Bank’s Deep Trade Agreements Database, a rich source of information on all preferential trade agreements, to carry out such an analysis. In recent PTA negotiations, governments are increasingly recognizing the need for policy space on capital flow measures, and are turning to the World Trade Organization’s (WTO’s) General Agreement on Trade in Services (GATS) model as a point of reference for negotiations on safeguards for economic exigencies.2 Information on these provisions in the database can underpin an important policy discussion about including exceptions to free transfer rules in PTAs for BOP difficulties or other forms of economic and financial distress. The scope for government discretion is of crucial importance given the prevalence of dispute settlement (DS) provisions in PTAs, particularly those involving investor-state dispute settlement (ISDS).Under these provisions,investors may challenge governments directly,and these governments could be liable for damages due to general restrictions that were imposed for policy reasons. Some PTAs even contain specialized provisions for investor-state arbitration that impact transfers and 1 For economic and theoretical analyses of capital liberalization as it impacts growth and development, see Frankel 2002, Ishii et al. 2002, Mattoo et al. 2006, Bhagwati 1998, and Blanchard and Ostry 2012. For analytical reviews of capital controls in international agreements, see Hagan 2000, Gallagher 2012, Siegel 2013, and Viterbo 2012. 2 For example, US-Republic of Korea, the EU-Canada Comprehensive Economic and Trade Agreement (CETA), and the draft Trans-Pacific Partnership (TPP) agreement. 179 Handbook of Deep Trade Agreements capital control measures, although their effectiveness varies. While investor-state arbitration is a valid investor protection measure, it creates liability for countries in ways that differ from other international agreements. In treaties calling for government-to-government dispute settlement, such as the original WTO agreement, governments have the opportunity to filter the disputes that they initiate, taking into account economic exigencies or allowing political solutions. 6.2. THE DATABASE: METHODOLOGY FOR MAPPING 6.2.1 Structure The database on transfers and capital controls in PTAs consists of six main headings corresponding to six main aspects of transfer and capital control commitments in PTAs. These commitments relate to: (a) the nature of the transfer obligation, including any distinctions between current and capital transactions as well as inflows and outflows; (b) broad exclusions from that obligation; (c) safeguards for handling BOP and macroeconomic crises; (d) exceptions for prudential measures in the financial sector; (e) general exceptions to the PTA commitments; and (f) dispute settlement provisions as they apply to the free transfers obligation (see the database outline at the end of this chapter).3 This study identifies more than 90 specific issues related to how PTAs address capital movements, then notes how each treaty handles these questions. For each PTA, the presence of a specific provision is coded “1” and the absence of the provision is coded “0.” If the provision is equal to the same provision in the WTO agreements, it is coded with (=). If it is more restrictive, it is coded with (+). If less restrictive, it is coded with (-). Where appropriate, comments are included to elaborate. 6.2.2 Commitments, key issues, and policy implications 6.2.2.1 The basic transfer obligation As transfer rules apply only to transactions that are liberalized under a PTA, they depend entirely on, and are derivative of, how those investments/services are defined and covered by the agreement. In other words, if the PTA does not require that a particular form of investment or financial service be allowed, there is no obligation to allow capital movements related to that action.4 The purpose of these rules is to provide confidence that the proceeds of the investment will be able to flow to the investor. Of the 284 treaties examined, approximately 3 Commitments (b) through (e) comprise the different types of qualifications to free transfer rules in PTAs.There is currently little jurisprudence interpreting these qualifications or how they apply in practice. 4 Similarly, under the IMF Articles, current transactions are required to be allowed only for legal trade transactions. If the Member restricts the underlying trade transaction (e.g., imports of a particular product), the free currency rules of the IMF Articles (Art.VIII, Section 2(a)) do not require that payments and transfers be made in conjunction with that transaction (Hagan 2000). 180 Movement of Capital Table 6.1: Free transfer commitments # PTAs with free transfer Total commitments (percent of total) (percent) Free transfers 145 N/A 51 Free transfers commitment in covered services sectors 89 61 Free transfers commitment in financial services 83 57 Free transfers commitment for covered investment capital flows 137 94 Without free transfer commitments 139 N/A 49 Source: Deep Trade Agreements Database. 145 (51 percent) have some commitment to maintain liberalized capital flows for covered transactions, while 139 treaties (49 percent) have no commitment to do so (see Table 6.1).5 The first question in the database asks simply whether the PTA contains a commitment on free transfers. Where the coding shows a “1” for the basic presence of the provision, other questions identify factors that indicate what it applies to, in large part based on its location in the treaty, including the following issues: • The most comprehensive treaties require free transfers in both bound services sectors (including financial services) and investment (defined broadly); this is common in the US-based RTAs. • Some treaties require free transfers only for financial flows relating to direct investment,6 while others cover only financial flows in bound services sectors.7 • A financial services annex may simply be an extension of the services chapter, adopting that standard. Other PTAs contain an independent chapter on financial services, which incorporates both the free transfers provisions of the investment and services chapters, depending on whether the financial flows are related to investment or trade in services.This latter model is common in Japanese agreements and a few others.8 • Further qualifications to the transfer requirements may be contained in these chapters or in dedicated “exceptions” chapters or annexes. • Dispute settlement regarding transfer rules may be covered within the chapters on investment or financial services, or under institutional provisions. • The text also varies from treaty to treaty.Table 6.2 highlights some of the most common models. While many treaties are written to mimic GATS Article XI, more comprehensive free transfer provisions (e.g., requiring transfers “without delay” in “freely usable currency”) are often found in an investment chapter.9 5 Prominent examples are the South Asian Free Trade Agreement (SAFTA) of 2006, the Chile-China Free Trade Agreement of 2010, the European Free Trade Association (EFTA)-Canada Free Trade Agreement of 2009, the Hong Kong SAR, China-New Zealand Free Trade Agreement of 2011, the Korea-Turkey Free Trade Agreement of 2013, and the Southern African Development Community (SADC) of 2000. 6 For example, EU-Algeria FTA, Canada-Chile FTA, and ASEAN FTA. 7 For example, Jordan-Singapore FTA and Australia-China FTA. 8 For example, Japan-Malaysia FTA, China-Korea FTA. 9 Some treaties contain GATS-equivalent language in the Services chapter with the comprehensive model in the Investment chapter – in that case, the reach of the article is different for the two chapters. 181 Handbook of Deep Trade Agreements Some broad patterns by region are apparent: Agreements involving the United States tend to be uniform, and have the broadest transfer commitment with limited exceptions; they thus involve high investor protection by limiting the policy space of governments (point 4 in Table 6.2). Agreements involving Canada and Latin American countries often follow the US model. Two basic models are prevalent in agreements to which the European Union is a party. One version applies only to avoiding restrictions on the free movement of capital relating to direct investment (point 2 in Table 6.2). Another applies more extensively to specified transactions (point 3 in Table 6.2). EFTA-based treaties tend to follow one of the EU models. Other regions reflect more diversity (e.g., point 1 in Table 6.2). Table 6.2: Free transfer provisions: examples 1. Services-only model (Jordan-Singapore Free Trade Agreement Article 4.12): “Except under the circumstances envisaged in Article 4.13 [Balance of Payments Safeguards Article], a Party shall not apply restrictions on international transfers and payments for current transactions relating to its specific commitments.” 2. Direct investment-only model (EU-CARIFORUM, Article 123): “With regard to transactions on the capital account of balance of payments, the Signatory CARIFORUM States and the EC Party undertake to impose no restrictions on the free movement of capital relating to direct investments made in accordance with the laws of the host country and investments established in accordance with the provisions of Title II, and the liquidation and repatriation of these capitals and of any profit stemming therefrom.” 3. Narrow/transitional investment transfers + services model (EU-Bosnia and Herzegovina Article 61): “1.With regard to transactions on the capital and financial account of balance of payments, from the entry into force of this Agreement, the Parties shall ensure the free movement of capital relating to direct investments made in companies formed in accordance with the laws of the host country and investments made in accordance with the provisions of Chapter II of Title V, and the liquidation or repatriation of these investments and of any profit stemming there from. “2. With regard to transactions on the capital and financial account of balance of payments, from the entry into force of this Agreement, the Parties shall ensure the free movement of capital relating to credits related to commercial transactions or to the provision of services in which a resident of one of the Parties is participating, and to financial loans and credits, with maturity longer than a year. “3. As from the entry into force of this Agreement, Bosnia and Herzegovina shall authorize, by making full and expedient use of its existing rules and procedures, the acquisition of real estate in Bosnia and Herzegovina by nationals of Member States. “Within six years from the entry into force of this Agreement, Bosnia and Herzegovina shall progressively adjust its legislation concerning the acquisition of real estate in Bosnia and Herzegovina by nationals of the Member States to ensure the same treatment as compared to its nationals. “The Parties shall also ensure, from the fifth year after the entry into force of this Agreement, free movement of capital relating to portfolio investment and financial loans and credits with maturity shorter than a year.” 4. Broader investment transfers + services model (US-Singapore Free Trade Agreement Article 8.10, 15.7, incorporated for financial services in 10.1.2): “Each Party shall allow all transfers and payments, [relating to the cross-border supply of services/relating to a covered investment] to be made freely and without delay into and out of its territory.” (This is then followed by a long list of the types of transfers that are specifically protected and a requirement that the transfers be made “in a freely usable currency”.) 182 Movement of Capital Specific aspects of the transfer requirement vary by agreement: Among the features explored are: how does the treaty address inflows versus outflows, and is there separate treatment of current versus capital transactions? In the GATS, for example, both issues have separate treatment. While the GATS require free transfers on all current transactions associated with the agreement, it is more limited with regard to only capital transfers in specific commitments. Furthermore, it addresses capital inflows and outflows separately depending on the nature, or “mode,” of the service covered. Following the GATS, many PTAs have evolved to cover sophisticated capital transactions. Their coverage ranges from specifying particular financial instruments to including generalized provisions on capital transactions.10 Nonetheless, many treaties do not distinguish between transfers for current or capital transactions and instead state generally that all transfers must be freely allowed for covered transactions. An even greater number do not distinguish between inflows and outflows, notwithstanding the important economic differences. In some cases, however, the transfer rules may apply differently to existing transactions and to new transactions undertaken after the agreement enters into force.11 6.2.2.1.1 Key issues and policy implications Users of the database who want to know the extent of capital account transactions liberalized under a particular PTA may wish to consider the country’s level of development. Some analysts argue that capital account liberalization in the context of bilateral and regional arrangements prioritizes investor protection over financial stability, and market access over appropriate sequencing. It is clear that a balanced approach to opening the capital account is necessary to reflect what has been learned in the post-war years from the experience with increased capital flows. Ideally, a country would sequence the opening of the capital account in line with its economic development and the strength of its economy and institutions. Given that the International Monetary Fund (the IMF or Fund) is charged with providing financing to address crises that may be caused by premature liberalization, it may be particularly appropriate for the Fund to play a central role in determining when liberalization supports - or undermines - the stability of members and the overall system. Alternative solutions such as in the Organisation for Economic Co-operation and Development (OECD), the WTO, or some combination thereof, have been proposed.12 10 The transfers and capital controls database notes where treaties have similar provisions to the GATS Market Access commitments (e.g., footnote to Article XVI concerning commitments made under modes 1 and 3). 11 Certain findings could not be adequately captured in the database. Documents identified as annexes or side letters were referenced in the comments column of the template, but specific reservations were not coded, and general lists of non-conforming measures were not cited. Furthermore, the transfer rules themselves are not divided by sectors, again because they are derivative of the underlying investments or services covered in the PTA. 12 See, e.g., Subramanian 2011. 183 Handbook of Deep Trade Agreements In 2012, the IMF published its Institutional View on the liberalization and management of capital controls (Inst’lView) in response to a call from the ministerial/cabinet-level International Monetary and Financial Committee for further work on a comprehensive, flexible, and balanced approach for the management of capital flows.13 The Inst’l View does not create any obligations, nor does it seek an uniform approach to managing capital controls; rather, it provides guidance. The key elements attempt to balance the tradeoffs between an open capital account and the regulation of what the IMF has termed “capital flow measures” (CFMs).14 In brief, the Inst’l View states that capital flows have substantial benefits for countries, including enhancing efficiency, promoting financial sector competitiveness, and facilitating greater productive investment and consumption smoothing.At the same time, it acknowledges risks to capital flows, while recognizing that benefits tend to accrue when countries have reached certain thresholds of development. Liberalization needs to be well planned, timed, and sequenced; and policy challenges to capital inflow surges or disruptive outflows need to be met with monetary, fiscal, and exchange rate management, as well as sound financial supervision and regulation and strong institutions. Other elements are discussed more below.The OECD is also actively working on this issue. Another key issue concerns overlap between a country’s obligations under the Fund’s Articles Agreement and under PTAs. Certain transactions that economists tend to think of as capital are defined under the Fund’s Articles as part of the obligation to avoid restrictions on payments and transfers for current international transactions (“exchange restrictions”). These include amortization on loans and moderate amounts of remittances. Still, exchange restrictions may be consistent with the Fund’s Articles because they are grandfathered in under the IMF Articles of Agreement, Article XIV, or have been approved under the Fund’s BOP policies or under the more limited exception for national security.15 PTAs could complicate the Fund’s approval functions, given that the authority to impose approved exchange restrictions constitutes a “right” under the Fund’s Articles and because Fund rules prevent discrimination among Fund members. Thus, the Fund would not be able to approve any such restriction if a party to a PTA applied a restriction differently among other parties and non-parties to the PTA because of the PTA’s obligations. The database thus flags when the PTA refers to the “rights and obligations” under the Fund’s Articles. An additional question, discussed below under the paragraphs on exceptions, is whether agreements that omit permission to apply financial safeguards could lead to increased need for Fund resources. 13 IMF 2012a. 14 Measures identified as designed to limit inflows include: (a) taxes on portfolio equity, (b) holding periods on central bank bonds, (c) limits on short-term foreign borrowing, (d) withholding tax on interest income on non- resident purchases of treasury and monetary stabilization bonds, and (e) fees on nonresident purchases of central bank paper. Measures identified as designed to limit outflows include: (a) limited bank withdrawals on transfers and loans in foreign currency, (b) stopping convertibility of domestic currency accounts for capital transactions, (c) waiting periods to convert proceeds of securities, (d) limits on forward transactions, and (e) export surrender requirements. IMF 2013, p. 24. 15 Decision 144-(52/51). 184 Movement of Capital 6.2.2.2 Qualifications to free transfer rules – broad exclusions The first set qualifications to the transfer rules16 involves matters that are broadly excluded from the general obligation to allow free transfers. In other words, the transfer obligation itself is defined so that certain kinds of transfers are not required to be freely made. For example, many treaties exclude measures concerning countries’ good faith application of laws such as bankruptcy, criminal law, and fraudulent practices.These provisions are largely non-controversial, as indicated by their widespread use across all treaties that contain free transfers provisions for covered investments. Again, the objective is to balance the confidence for investors and service providers with public policy matters. Some treaties have specific exceptions in annexes, where a country reserves the right to employ specific capital flow regulation measures. In these cases, the database only flags that such annexes exist, without recording each detailed reservation. 6.2.2.3 Qualifications to free transfer rules – exceptions for macro–economic crises Most exceptions in the PTAs for macroeconomic crises, if any, are modeled on the GATT/GATS. We found that 82 percent (119) of all treaties with capital account transfers commitments contain a safeguard protecting countries with balance-of-payments difficulties, “serious difficulties for operation of exchange rate policy or monetary policy,” or “other macro-economic difficulties” (see Table 6.3). For limitations on the use of safeguards, the database attempts to capture the extent to which the PTA refers explicitly back to the safeguard rules in the GATS or is quite similar in substance. Some variation on safeguards is regional and results from differing negotiating power and perceived role of these agreements in the global economy. For example, while agreements with the European Union tend to include a safeguard, the majority of recent US Bilateral Investment Treaties (BITs) and Free Trade Agreements prohibit governments from restricting capital flows by covered investments and exclude exceptions for economic crises. The US consulted, but essentially declined input from, 250 economists for attention to these global issues, and the recently issued model BIT continues to lack a safeguard for economic crises. This result is echoed in the investment chapters of many of the PTAs that have the US as a signatory.“The BITs and FTAs of other major capital exporters such as those negotiated by the UK, Japan, China, and Canada, either completely ‘carve out’ host country legislation on capital account regulations (therefore permitting them) or allow for a temporary safeguard on inflows and outflows to prevent or mitigate a financial crisis.”17 16 The database organizes qualifications to the transfer rules into four main categories: (a) broad exclusions, (b) exceptions for macro-economic crises, (c) safeguards for prudential measures related to the financial sector, and (d) other general or sector-specific exceptions. There is currently little jurisprudence interpreting these qualifications to the free transfer rules of these treaties and how they apply in practice. 17 Gallagher 2012, p. 125. 185 Handbook of Deep Trade Agreements Researchers should consider the “triggers” of the type of economic circumstances that might warrant safeguard measures. As noted, the texts of some PTAs that contain a safeguard may mirror the GATT’s BOP-based exception. Others reflect evolution in the economic circumstances and increased capital flows and may include language such as “external financial difficulties” or “serious difficulties for macroeconomic management.” Still, it has not been completely resolved whether this type of language covers both inflows and outflows or if it extends to regulating inflows necessary to stem asset bubbles and the buildup of debt in the economy.18 Some sample “triggers” are noted in Table 6.3. Table 6.3: Safeguards: sample triggers and disciplines # Treaties with free Comments transfer commitments (percent) Safeguards 119 82 Safeguard for balance-of- 108 74 payments difficulties Safeguard for “serious difficulties 48 33 for operation of exchange rate policy or monetary policy” or “other macroeconomic difficulties” Safeguard measures must 118 99% of treaties “not exceed those necessary” with safeguards to address the situation Safeguard measures may accord 91 76% of treaties priority to activities essential with safeguards to economic stability Safeguard must be notified to 107 90% of treaties the other parties’ ex post with safeguard Safeguard with most-favored-nation 99 83% of treaties requirement with safeguards Prudential measures exceptions 95 65 Prudential measures may 79 83% of treaties “not be used as a means of avoiding with prudential the Party’s commitments” or measures exception “a disguised restriction on trade” Source: Deep Trade Agreements Database. 18 Gallagher,Viterbo, and Anderson 2015. 186 Movement of Capital As under the GATT/GATS, safeguards are generally considered an exception to the general free transfers rule and thus include conditions or “disciplines” to ensure against abuse. This section first examines whether certain broad topics were excluded from the safeguard, such as where transfers could be not being restricted, even in circumstances of economic distress. Only seven PTAs carved out foreign direct investment and, in one case, parties were prohibited from using “dual or multiple exchange rates” as a safeguard measure.19 We found that none of the PTAs reviewed carved expropriation out of the safeguard, so the transfer of compensatory payments could be restricted if the safeguard were applied. Non-discrimination is usually a key feature of any permissible capital control measure. Nonetheless, these provisions are not necessarily as broad as the basic free transfers rule, which normally would apply on a non-discriminatory basis under the MFN or national treatment (NT) terms of the agreement. The PTAs vary in the extent to which economic exigencies may warrant discriminatory application of a restriction (i.e., different treatment for residents or non-residents). The IMF Inst’l View generally prefers currency-based measures but notes the possible need to resort to residency-based measures if the former are ineffective (e.g., limits on residents’ investments in financial instrument abroad, sale and repatriation of nonresidents’ investments in the country in foreign currency, and waiting periods to transfer proceeds).20 Other disciplines that mirror the GATT/GATS include, for example, that restrictions should “not exceed those necessary in the circumstances” (99 percent), or “avoid unnecessary damage to the commercial, economic and financial interests of other Parties” (87 percent). Researchers should be aware that the “necessity” test in other contexts has been judged to require that no less restrictive solution was available.Additionally, 66 percent of treaties require parties to consult with their treaty partners before imposing safeguards and 90 percent require notification as soon as possible after the measures are deployed. Conversely, a small majority of treaties add some flexibility by allowing countries to “accord priority to activities essential to [their] economic stability” (see Table 6.3).21 The database also groups together any conditions that distinguished between inflows and outflows but does not address if there are quantitative limits on the restriction, as we found this to be rare. Important debate has centered on the permissible duration of any such restriction. Some commentators offer that such measures may need to be part of the regular medium-term policy “tool kit.”22 Additionally, some theoretical research shows that capital controls can be 19 Republic of Korea–Colombia Free Trade Agreement, Annex 8-C, Article 2(i). 20 IMF 2012a, p. 20. 21 Other conditions reflect the increasing conclusion of sophisticated financial instruments in capital flow measures. This is evidenced by the US-Korea FTA and the Canada-Korea FTA, in which Korea commits to rely principally on “price-based measures” as safeguards. 22 Anderson 2013. 187 Handbook of Deep Trade Agreements seen as correcting for market failure rather than being considered distortionary.23 In contrast, most of the PTAs call for the measure to be “temporary,” while differing on whether they specify particular timeframes or means to consult on the necessary duration of the measure. A final feature in this section is the extent to which Parties must notify each other or some consultative group about the restrictions. These provisions include how any consultations will take place, including whether the IMF is to be involved for economic analysis (as in the GATT, GATS, and draft MAI) or for a relationship to its financing function (as in NAFTA). 6.2.2.3.1 Key issues and policy implications Policymakers may want to consider the trade-offs by looking at restrictions relating to the macroeconomic, financial, and BOP implications of unrestricted capital flows, especially short- term flows. Safeguards recognize that countries may need to restrict transfers to stem capital flight, and to protect monetary reserves or guard against extreme exchange rate fluctuations, at least for a limited time to introduce appropriate adjustment policies, which impact global spillovers. Investors may argue that the PTA maximizes its goal of expanding investment and capital liberalization while minimizing the risk of restricted transfers. This rationale may apply more for short-term instruments, including speculative flows, than for instruments of longer maturities, which may not be affected by restrictions that are only temporary. But, will market participants focus on the added transfer risk due to possible restrictions for macroeconomic crises, over the benefits of the host country’s ability to maintain economic and financial stability? The latter could actually encourage investment. The expertise of any arbitrators in evaluating the triggering economic circumstances also matters. Many of the multilateral agreements recognize the international coherence of including a role for the IMF in such cases, such as in the GATT and the GATS under the WTO Agreements. The absence of a safeguard provision in PTAs could conflict with the functions of the IMF. One area is the connection with financing from the IMF. If a Party cannot protect its BOP position in a crisis, it could result in increased demands for Fund resources in the context of a program to resolve the BOP difficulties. Another area concerns dispute resolution, as discussed below. In the same vein, the fact that only some PTAs require the safeguard to be applied on an MFN basis highlights possible conflicts from the interaction of a patchwork of treaties. If a Party were to impose a safeguard permitted under one treaty (which demands MFN treatment), it would be important to consider how this impacts its obligations to a different trading partner under a separate treaty that may not allow safeguards. In other words, by 23 Subramanian 2012. 188 Movement of Capital applying the safeguard on an MFN basis, the party may violate another RTA. Conversely, a party may hesitate to employ a safeguard permitted under one treaty but prohibited under another. For example, even if Colombia is relatively free to employ capital controls (for BOP reasons) under its treaty with Canada, it may not do so because that same flexibility is not present in its treaty with the United States. Ninety-nine of those 119 safeguard provisions (83 percent) contain an MFN requirement. Some analytical work suggests a cooperative approach to create the best environment for both domestic and global economic development, while mitigating the effects of potentially destabilizing capital flows that were evidenced in the boom-bust cycles during and after the recent global financial crisis. This work has spanned proposals for countries to sequence capital liberalization based on their development level and to allow safeguards in extreme cases of economic stress. For example, looking at data on bank asset flows and capital account restrictions, some recent work considered “whether a cooperative approach to taming potentially destabilizing capital flows – by imposing capital account restrictions at both the source and the recipient country ends – may be feasible.”24 Furthermore, “coordination would likely need to involve both recipient countries (to minimize the risk of capital control wars and excessive mutual deflection of flows) and source countries (to ensure that they bear part of the cost burden when costs from controls are convex).”25 6.2.2.4 Financial sector safeguards and exceptions for prudential measures Many of the PTAs in this database cover financial services either under services generally or in a dedicated chapter. Transfer rules concerning capital flows are particularly relevant under the cross-border provision of financial services.Transfers are inherent in transactions ranging from the establishment of local branches in member countries (e.g., the acceptance and management of deposits), to the variety of financial instruments including loans and more sophisticated financial products.26 Again, the objective is to balance free transfers with necessary controls for the integrity of the financial system. In this area of financial sector commitments, the key exception to capital flow requirements on financial instruments is the allowance for “prudential measures.” There is no generally accepted “definition” of prudential measures, but the GATS annex is instructive: the equitable, nondiscriminatory, and good faith application of measures relating to soundness, integrity, or financial responsibility of financial institutions or cross-border financial service providers. Generally, the PTAs that contain a prudential safeguard use such language as: 24 Ghosh, Qureshi, and Sugawara 2014. 25 Blanchard and Ostry 2012. 26 As in other sections of this chapter, sub-headings are used to organize whether the PTAs have different treatment among different kinds of transactions, inflows or outflows, and rules regarding the duration of any such measure. 189 Handbook of Deep Trade Agreements “a Party shall not be prevented from adopting or maintaining measures for prudential reasons […] in order to ensure the integrity and stability of the financial system.” Some, however, have limiting language that focuses on individual financial institutions or cross-border financial service suppliers, which raises questions about whether generalized measures would be permitted.27 It is also important to note that 83 percent of the PTAs reviewed contain additional limiting language:“[w]here such measures do not conform to the provisions of this Agreement referred to in this paragraph, they shall not be used as a means of avoiding the Party’s commitments or obligations under such provisions.” This provision seems broad but may not be such a broad exception that engulfs the rule, as it could be interpreted to be limited to examining the justifiable need or intention behind the measure, rather than solely its broad effect. Prudential measures may be considered capital controls. For example, restriction on banks’ foreign borrowing (as in a levy on bank foreign exchange inflows) or required reserves on banks’ foreign exchange liabilities are both types of common prudential measures but also help to manage the capital account. Recent crises highlight this dilemma.The database maps both the exclusions and exceptions to transfer obligations in the financial services sector (see Table 6.3). One noteworthy exclusion is found in the PTAs that allow Parties to restrict transfers based on the application of their antifraud laws – sometimes expressed as measures “relating to the prevention of deceptive and fraudulent practices or to deal with the effect of a default on financial services contracts.” Such language could usefully allow coordination with anti-money-laundering efforts, such as “know your customer” rules. Allowing exceptions for prudential measures are another approach to ensuring financial stability and the safety of the financial sector. 6.2.2.4.1 Key issues and policy implications The IMF Inst’l View posits that, while not substituting for these macroeconomic and financial policy measures, CFMs could be useful for supporting macroeconomic policy adjustment and safeguarding financial system stability. It explains that if CFMs are used, they should seek to avoid discrimination based on residency, and the non-discriminatory measure that is effective should be preferred. One key message with regard to inflows is that CFMs should be targeted, transparent, and generally temporary, being lifted once the surge abates, in light of their costs. Taking account of the different circumstances involving controls of capital outflows, it states that CFMs should generally be used only in crisis (or imminent crisis) situations.28 Policy discussion on the Inst’l View is ongoing,29 including at the OECD. 27 “It is understood that the term “prudential reasons” includes the maintenance of the safety, soundness, integrity, or financial responsibility of individual financial institutions or cross-border financial service suppliers” (emphasis added). For example, US-Peru FTA, US-Morocco FTA, and US-Singapore FTA. 28 IM 2012a; see box 3, pp. 35-36. 29 See, e.g., Batista 2012; Fritz and Prates 2013; Gabor 2012; and Gallagher 2014. 190 Movement of Capital 6.2.2.5 Qualifications to free transfer rules – sector-specific or general exceptions This section covers the remaining exceptions that may apply to the basic transfer obligation.We looked at whether the treaty outright excludes certain sectors from the transfer requirement. As shown in Table 6.4, in 33 percent of the PTAs, monetary and exchange rate policy by a public entity is exempt from rules about capital account regulations.30 Also, many treaties contain an exception for “essential security” that is “self-judging.” In those cases, the text allows the Party imposing the measure to determine if there is a threat to essential security justifying derogations from the treaty obligations. Although it has not been litigated to date, there are some theories that financial collapse or instability may be a security issue within the meaning of that exception. Finally, slightly more than half of treaties contain a reference to “general exceptions” in the GATS, which makes exceptions for policy measures “necessary to protect human, animal or plant life or health.” This would not obviously cover situations of financial distress, but it is open to consider if it could be invoked in the direst of financial emergencies. Table 6.4: Carve-outs and general exceptions # Percent of treaties with free transfer commitments Carve-out for monetary or exchange 48 33% rate policy by public entity from transfers commitments31 Self-judging “essential security” 126 87% Reference to the “general exceptions” 84 58% in GATS32 Source: Deep Trade Agreements Database. 6.2.2.6 Enforcement through dispute resolution The final section of this database explores “enforcement” through dispute resolution. It covers both state-to-state dispute settlement and ISDS. The latter aims to provide covered investors protection from regulatory instability in the host state. It allows investors to seek redress directly against the host government for perceived treaty violations. Present in BITs for decades, it takes on new significance in the context of the range of capital and financial transactions now covered 30 See, e.g., Chile-Japan FTA Article 117.4, China-Korea FTA Art. 9.1(2), Malaysia-Australia FTA Annex on Financial Services Article 1.3, and Mexico-Panama FTA Article11.12.2. 31 In a number of cases (28), the treaty appears to carve out such a policy, but then explicitly prohibits the use of that carve-out for capital flow measures. 32 This number does not include treaties that have GATS-similar language but do not refer to the GATS explicitly. 191 Handbook of Deep Trade Agreements in PTAs. Thus, the difference in the type of governmental liability from other international agreements (with only state-to-state enforcement) has become more pronounced. ISDS could create additional difficulties for states already suffering from financial difficulties, given the sums involved in both funding the case and the potentially large awards. The database maps the provisions in PTAs that provide for ISDS as it may impact the free transfer provision. 99 percent of all treaties have some sort of dispute settlement option – at least at the state-to-state level, while 57 percent include ISDS (Table 6.5). In some cases, ISDS provisions vary in their breadth and scope.The US-Colombia FTA garnered particular attention when it entered into force in May 2012 due to special constitutional procedures in Colombia. Its content, however, closely aligned with the US approach. Although the ISDS provisions limit loss to reductions in the value of the transfer, and exclude lost profits in possible recovery, even this accommodation is limited to certain kinds of transfers and has further constraints on restrictions relating to outward payments and transfers. US-Chile and US Peru also limit compensation in certain investor-state cases, but these limits do not apply to controls on outflows.33 Finally, this section of the database identifies when the treaty has specialized dispute settlement rules applicable to transfers (often contained in Annexes). For example, the US-Singapore and US-Chile PTAs have a “cooling off ” period that operates to delay when an investor can initiate a claim. They still hold signatories liable to investors for (even temporary) restrictions that were imposed to resolve an economic and financial crisis, if a panel finds that the restrictions “substantially impede transfers.” The liability applies retroactively even if the restrictions have Table 6.5: Enforcement # Percent of treaties with Comments free transfer commitments Presence of 141 99% state-state disputes Presence of 83 57% investor-state disputes Requirement for 60% of treaties “special expertise” 57 with ISDS in arbitrators for financial for financial services disputes services disputes Disputes related 30% of treaties to prudential measures 28 with ISDS removed to a financial for financial authority of the parties services disputes Source: Deep Trade Agreements Database. 33 Anderson 2013, p. 86. 192 Movement of Capital been subsequently removed. The database refers researchers to an oft-mentioned “side letter” to the US-Singapore PTA which purports to clarify what measures are considered by the US government to “substantially impede transfers,” but it has minimum legal effect. It states a “rebuttable presumption” that certain forms and effects of restrictions “will be deemed [by the US] as not to substantially impede transfers,” including, for example, that the controls be non- discriminatory or price-based.This letter does not constrain individual investors from bringing a claim under the terms of the treaty, does not bind arbitral panels, and does not clarify how restrictions may indeed need to have substantial effects in order to serve their purpose.34 6.2.2.6.1 Key issues and policy implications The principal question as it relates to dispute settlement over transfers restrictions is whether and to what extent governments can be liable to individual investors when general economic difficulties require measures that may be in conflict with the PTAs. In treaties calling for state- to-state dispute settlement, such as the WTO, governments have the opportunity to filter the disputes that they initiate, taking into account economic exigencies or allowing political solutions. In PTAs that lack robust exceptions to transfer rules for economic exigencies, investors could seek compensation for damages resulting from policy measures that many economists and analysts would consider appropriate government policy.Thus, thoughtful writers have described how this potential liability could similarly cause a “chilling effect” on governments considering environmental or public health measures that also impact investments.35 Users of this database are encouraged to review if some of the “procedural protections” to ISDS actually serve the policy challenges in this area. Some of these concerns in connection with specialized side letters are mentioned above. Another issue is that the rules may require that the controls be consistent with IMF rules concerning transparency, limited duration, and avoidance of multiple exchange rate practices. One model is the US-Korea FTA (March 2012) that allows measures imposed pursuant to Article 6 of (Korea’s) Foreign Exchange Transactions Act” (Annex 11-G on Transfers), which accommodates Korea’s law.Yet, other such disciplines seem to limit the deferral to Korean law, by excluding payment or transfers for foreign direct investment or the rule that the measures may not otherwise interfere with investors’ ability to earn a market rate of return in the territory of the Republic of Korea. The trade-offs of ISDS should be considered along with the benefits of government filters that serve important diplomatic or economic benefits. In the WTO, for example, only governments may bring a claim under the dispute settlement procedures established by the WTO Agreements. Still, the US has vigorously pursued trade cases against China on the covered trade matters such 34 Siegel 2013, p. 76. 35 See, e.g., Gantz 2004, pp. 684-689. 193 Handbook of Deep Trade Agreements as anti-dumping or anti-subsidy provisions, but it did not include a petitioner’s claim that the exchange rate constituted a countervailable subsidy.36 It also declined to initiate a broader trade remedy case against China’s exchange rate or currency management,37 which is better managed through diplomatic channels and the work of the IMF. In its PTA with the US, Australia has negotiated to retain the right for the government to screen disputes. 6.3. FREE TRANSFERS COMMITMENTS: SOME TRENDS Combined, the commitments under these treaties (described in section 6.2) shape how treaty parties are able to regulate capital flows during economically stable times, as well as in times of crisis. It is clear, both from this chapter and others in this volume, that there is a trend toward more complex, deeper, and broader treaty commitments. In the context of free transfers, we can see that countries are increasingly binding their own hands by liberalizing capital flows (Figure 6.1), even in the 7 years following the financial crisis. The two final years shown on the chart seem to mark a change in the trend, although the data for those years are still incomplete. Another important trend is the strong prevalence of free transfers provisions in treaties where the parties represent of a mix of “developed” and “developing” countries.38 Indeed, 71 percent of North-South treaties from our data contain some level of free transfers commitment while only 64 percent of North-North treaties and a much smaller 30 percent of South-South treaties have the same provisions (Figure 6.2). This reality suggests both that (a) developing countries, when given the choice, tend to not demand liberalized capital flows in their treaties, while (b) developed countries tend to include them, especially when the trade partner is a developing country. If true, it would seem that the developed country partners have somewhat more power in treaty negotiations, making the concern for policy space for financial stability in developing countries even more crucial. 36 See e.g., Department of Commerce 2006. 37 In 2007, the US declined to pursue a petition under “Section 301” of the US trade laws. 38 The word “developed” has many meanings. In this case, we refer to countries considered “high income” by the World Bank data standards. All other countries are grouped together as “developing.” Admittedly, this is a rough gloss over a complex issue, but it begins to shed light on relationships between countries of varying development and income levels. 194 Movement of Capital Figure 6.1: Free transfers commitments over time Source: Deep Trade Agreements Database. Figure 6.2: Treaty commitment by income level Source: Deep Trade Agreements Database. 195 Handbook of Deep Trade Agreements 6.4. CONCLUSIONS The Deep Trade Agreements Database on the free transfer provisions of PTAs is important for researchers and policymakers because capital flows are an increasingly important means of allocating savings, promoting growth and facilitating BOP adjustment. These commitments create greater certainty for manufacturers, foreign investors, and service providers, given the increasing range and sophistication of the types of transactions covered in the PTAs. The data shows a number of trends ranging from common texts and regional approaches. At the same time, the transfers provisions also vary extensively across the PTAs in their scope and the transactions to which they apply. They also raise policy questions about the need to retain “policy space” generally, the importance of safeguard provisions to allow restrictions in times of economic crises, what the components of those safeguard provisions should be, and how they should be enforced, given the special features of investor-state dispute settlement. The creation and initial review of the database supported the theme that the patchwork of obligations under the PTAs is a poor substitute for a coherent international regime, under the aegis of the IMF or elsewhere. Even the extensive mapping herein doesn’t fully answer how the treaties actually reflect policy-making. In other words, most Parties are signatories to several treaties where they may have had differing negotiating strength and thus are subject to differing levels of obligation. It is also not clear how the treaties interact with each other. As mentioned, a country may not be able to employ provisions that are permissible under one PTA because the same provisions are restricted under another PTA (either generally or specifically) or would complicate their obligations under other multilateral agreements. Future work might consider how the database could inform a discussion of a multilateral framework, and how this topic is impacted by the distribution of multilateral corporations with locations globally. ACKNOWLEDGMENTS We would like to thank Brad McDonald for his valuable  comments and  the participants in the World Bank’s workshop “Evolution of Deep Trade Agreements” for helpful comments and suggestions.  196 Movement of Capital REFERENCES Abdelal, Rawi. 2007. Capital rules. Cambridge, MA: Harvard Gantz, David A. 2003. The evolution of FTA investment University Press. provisions: From NAFTA to the United States–Chile Free Trade Agreement. American University International Law Gallagher, Kevin P., Sarah Anderson, and Annamaria Viterbo. Review 19(4): 679-767. 2015. The Trans-Pacific Partnership and regulating capital flows. Global Economic Governance Initiative, Boston Gianviti, François. 1999. The International Monetary University. Fund and the liberalization of capital movements. Current Developments in Monetary Financial Law I: 7-16. Batista, P.N. 2012. The IMF, capital account regulation, and emerging market economies. In Regulating global capital flows for Ghosh, Atish, Mahvash S. Qureshi, and Naotaka Sugawara. long-run development, edited by Kevin Gallagher and Stephany 2014. Regulating capital flows at both ends. http://voxeu. Griffith-Jones,. Boston: Pardee Center, Boston University. org/article/regulating-capital-flows-both-ends. Bhagwati, Jagdish. 1998.The capital myth:The difference between Gold, J. 1977. International capital movements under the trade in widgets and dollars. Foreign Affairs 77(3): 7-12. Law of the International Monetary Fund. Pamphlet Series 21, International Monetary Fund, Washington, DC. Blanchard, Olivier, and Jonathan D. Ostry. 2012. The multilateral approach to capital controls. http://voxeu.org/ Hagan, Sean. 2000. Transfer of funds. UNCTAD Series, article/multilateral-approach-capital-controls (December 11). United Nations, Geneva. Department of Commerce, International Trade IMF (International Monetary Fund). 1976. Proposed second Administration. 2006. Notice of initiation of countervailing amendment to the Articles of Agreement of the International duty investigations: coated free sheet paper from the People’s Monetary Fund—a report by the Executive Directors to the Republic of China, Indonesia, and the Republic of Korea. 71 Board of Governors. Fed. Reg. 68546 (November 27). ———. 1994.The relationship of the World Trade Organization EUR-Lex: Access to European law. N.d. https://eur-lex. with the Fund. http://archivescatalog.imf.org/detail.aspx. europa.eu/homepage.html. ———. 1997. Capital movements under an amendment of the Frenkel, Roberto. 2002. Capital market liberalization and Articles—concepts of “international capital movements” and economic performance in Latin America. In International “restrictions”: selected issues for discussion (SM/97/146, Sup. capital markets and systems in transition, edited by John 1). http://archivescatalog.imf.org/brief.aspx. Eatwell and Lance Taylor. Oxford: Oxford University Press. ———. 1997. Communiqué of the Interim Committee of the Board of Governors of the International Monetary Fund. http:// Fritz, Barbara, and Daniela Prates. 2013. The new IMF www.imf.org/en/news/articles/2015/09/28/04/51/cm970921 approach to capital account management and its blind spots: (September). Lessons from Brazil and South Korea. http://www.tandfonline. com/doi/abs/10.1080/02692171. ———. 2007. Review of the 1977 decision—proposal for a new decision. http://www.imf.org/external/pp/longres.aspx?id=4191. Gabor, Daniela. 2012. Managing capital accounts in emerging markets: Lessons from the global financial crisis. http://www. ———. 2010a. The Fund’s role regarding cross-border tandfonline.com/doi/abs/10.1080/00220388.2011.649257. capital flows. http://www.imf.org/external/np/pp/eng/ 2010/111510.pdf. Gallagher, Kevin P. 2012. Capital account regulations and the trading system. In Regulating global capital flows for long- ———. 2010b. The Fund’s mandate—the legal framework. run development, edited by Kevin Gallagher and Stephany http://www.imf.org/external/pp/longres .aspx?id=4421. Griffith-Jones, Boston: Pardee Center, Boston University. ———. 2012a.The liberalization and management of capital ———. 2014. Contesting the governance of capital flows at the IMF. flows: an institutional view. http://www.imf.org/external/ http://onlinelibrary.wiley.com/doi/10.1111/gove.12100/full. pp/longres.aspx?id=4720. 197 Handbook of Deep Trade Agreements ———. 2012b. Liberalizing capital flows and managing outflows. http://www.imf.org/external/np/pp/eng/2012/031312.pdf. ———. 2012c. Modernizing the legal framework for surveillance—an integrated surveillance decision. http:// www.imf.org/external/ns/search.aspx. ———. 2013. Guidance note for the liberalization and management of capital flows. http://www.imf.org/ external/pp/longres.aspx?id=4759. Ishii, Shogo, Karl Habermeier, Bernard Laurens, John Leimone, Judit Vadasz, and Jorge Ivan Canales-Kriljenko. 2002. Capital account liberalization and financial sector stability. IMF Occasional Paper 211, International Monetary Fund, Washington, DC. Mattoo, Aaditya, Randeep Rothindran, and Arvind Subramanian. 2006. Measuring services trade liberalization and its impact on economic growth: an illustration. Journal of Economic Integration 21(1): 64-98. Organisation for Economic Co-operation and Development. N.d. http://www.oecd.org/investment/internationalinv estmentagreements/multilateralagreementoninvestment.htm and http://www.oecd.org/investment/investment-policy/codes.htm. Salacuse, Jeswald W. 2015. The law of investment treaties. Oxford: Oxford University Press. Siegel, Deborah. 2002. Legal aspects of the IMF/WTO relationship: The Fund’s Articles of Agreement and the WTO agreements. American University International Law Review 3(3) (July). ———. 2004. Using free trade agreements to control capital account restrictions: summary of remarks on the relationship to the mandate of the I.M.F. ISLA Journal of International and Comparative Law 10(2) (Spring). ———. 2013. Capital account regulations and the trading system: a comparability review. Pardee Center, Boston University, Boston. Subramanian, Arvind. 2012. International regulation of the capital account. In Regulating global capital flows for long-run development, edited by Kevin Gallagher and Stephany Griffith- Jones,. Boston: Pardee Center, Boston University. Viterbo,Annamaria. 2012. International economic law and monetary measures: Limitations to states’ sovereignty and dispute settlement. Cheltenham, UK: Edward Elgar. World Trade Organization. N.d. https://www.wto.org/english/res_e/res_e.htm. 198 Movement of Capital ANNEX Database Outline A. Free Transfers i. Types of transactions identified for free transfers ii. Transactions involving inflows iii. Transactions involving outflows B. Qualifications to Free Transfer Rules – Broad Exclusions i. Good-faith, non-discriminatory application of laws C. Qualifications to Free Transfer Rules – Safeguards i. Carve-outs from safeguards ii. Requirements for safeguards iii. Notification and consultation requirements for safeguards iv. Duration of safeguard rules D. Qualifications to Free Transfer Rules – Prudential Measures i. Requirements/limitations for prudential measures ii. Duration of prudential measures E. Qualifications to Free Transfer Rules – Other Exceptions i. Sector-specific exceptions ii. Duration for sector-specific exceptions iii. General exceptions iv. Duration for general exceptions F. Dispute Resolution as Applied to Capital Movements i. Specialized procedures for financial services ii. Other special procedures 199 Handbook of Deep Trade Agreements 200 Intellectual Property Rights CHAPTER 7 Intellectual Property Rights M. Wu 201 Handbook of Deep Trade Agreements CHAPTER 7 Intellectual Property Rights M. Wu* * Harvard Law School, Cambridge, Massachusetts, United States CONTENTS 7.1. INTRODUCTION 203 7.2. LITERATURE REVIEW 205 7.3. METHODOLOGY 207 7.4. FINDINGS 211 7.4.1. Prevalence in PTAs 211 7.4.2. US PTAs 213 7.4.3. European Union PTAs 216 7.4.4. European Free Trade Area PTAs 218 7.4.5. PTAs of advanced Asian economies 219 7.5. CONCLUSIONS 220 ACKNOWLEDGMENTS 221 REFERENCES 222 202 Intellectual Property Rights 7.1. INTRODUCTION Among the most controversial and intensely scrutinized elements of preferential trade agreements (PTAs) are their obligations concerning intellectual property rights (IPR). The worries associated with these provisions are plentiful: Will they raise the prices of pharmaceutical drugs or agricultural seeds? Will they limit the rights of creative artists? Do they impair the ability of indigenous communities to commercialize traditional forms of knowledge? Such concerns often galvanize opposition to trade agreements, particularly in developing countries. Given such concerns, why then do countries agree to strong IPR provisions in trade agreements? One answer is that these protections may be beneficial for economic development. The empirical evidence on this proposition, however, is mixed. One study1 found that strengthening IPR is only one of a broad set of factors necessary for developing countries to attract inward investment and technology transfer. Moreover, if stronger IPR protections are truly good for a country’s long-term prospects, then the government could simply seek to act unilaterally. At best, a trade agreement serves as a mechanism to enlist more interest groups to overcome opposition and to raise the cost of possible future back-tracking. However, the benefits of strengthening IPRs and enacting other forms of trade liberalization to attract technology transfer will vary depending on the developing country’s profile.2 In most instances where IPR provisions are included as part of a trade agreement, there is a larger trade-off at work. The stronger IPR provisions are often conceptualized as part of a bigger bargain. For example, that bargain might involve developing countries obtaining increased market access for agricultural products, raw materials, and low-cost manufactured goods in exchange for advanced economies gaining greater IPR protection and market access for services. After all, the Uruguay Round itself involved the drafting of a new Agreement on Trade-Related Intellectual Property Rights (the TRIPS Agreement) as well as a new General Agreement on Trade in Services (GATS) as part of the overall bargain. In the quarter century since the TRIPS Agreement was drafted, advanced economies have sought to further strengthen trade rules governing IPR through their preferential trade agreements. Some developing countries have succumbed, lured by the temptation of preferential market access terms for their exports. In modeling such agreements, one study3 found that stronger IPR provisions are most likely where the developing country’s imitative capacity is neither too high nor too low. 1 Maskus 2005. 2 Hoekman, Maskus, and Saggi 2005; Maskus, Saggi, and Puttitanun 2005; Branstetter, Foley, and Saggi 2010. 3 Hoekman and Saggi 2007. 203 Handbook of Deep Trade Agreements Once a developing country has succumbed to TRIPS-plus IPR provisions in exchange for better terms for its textiles or agricultural exports, other countries tend to follow. Not wanting Country A’s producers to gain an advantage of its own exporters, Country B agrees to the same bargain. Over time, a game of competitive liberalization has played itself out, leading to a series of PTAs with new TRIPS-plus rules. Competitive liberalization and market access are only part of the story, however. As technology and globalization contribute to increased disaggregation of production, another dynamic has developed - one centered around global value chains.As several authors have highlighted,4 a range of developing countries now recognize attachment to global value chains as a critical mechanism for sparking a virtuous cycle for economic and social development. To be competitive in the global race to direct regional value chains through their territory, however, countries need to embrace a range of policy actions. Not only must they lower tariffs and non-tariff barriers to allow for the increased flow of inputs, but they also align their domestic rules and regulations to attract investment. Firms based in advanced economies that spearhead these global value chains increasingly rely upon IPR as a key source of their competitiveness. They therefore seek assurances that the markets with which they are integrating will protect their intellectual property at the level necessary for the firm to generate a positive economic return over time. This desire to embed within global value chains has also has fueled the expansion of PTAs with TRIPS-plus rules.There is support for the idea5 that such PTAs have a positive impact on trade in high-IP goods among certain middle-income countries, but that the dynamic is complex. Multiple factors, therefore, explain the proliferation of PTAs with robust IPR provisions in deep-integration trade agreements. This chapter provides an overview of this phenomenon. It draws on a new database developed for this study, the content of which is explained below. A preliminary analysis of the data suggests that as of the late 2010s, four major hubs serve as the engines for this phenomenon. The first, and perhaps most well-known, is the United States, which has sought to advance a wide range of TRIPS-plus rules through its deep-integration PTAs. A second is the European Union, whose policies have also shifted in this direction over the past decade. Two others which have not received as much attention are the European Free Trade Area (EFTA) and a set of advanced Asian economies, both of which have also pushed forward with their own deep-integration trade agenda with robust TRIPS-plus IPR rules. Each hub has played a relatively distinct role in the process of driving an IPR-oriented deep integration trade agenda. Because extensive work has already been done by others on the IPR chapters of PTAs, this chapter opens with a review of the key findings of the major studies to date. It then 4 See, e.g., Baldwin 2016. 5 Maskus and Ridley 2017. 204 Intellectual Property Rights provides an overview of the elements of the new database constructed as part of this study. Finally, it offers a few summary statistics before shifting to highlight the salient features of the four hubs, each of which has relied upon IPR provisions to facilitate a vision of deeper integration with its trading partners. 7.2. LITERATURE REVIEW Over the past decade, the substance of IPR chapters of PTAs has been the focus of a number of studies. One of the most comprehensive was undertaken by the World Trade Organization (WTO).The authors of that study6 examined 245 PTAs notified to the WTO and in force as of February 2014, and found that slightly more than 70 percent included IP provisions. This share increases dramatically for recent trade agreements, with more than 90 percent of those concluded after 2009 containing an IP chapter. The study also provides a comprehensive description of the types of IP commitments in trade agreements, as well as the frequency of specific types of commitments. Of those PTAs that include IP provisions, the most common are “softer” ones that simply reaffirm existing commitments and promote cooperation. More than 70 percent of PTAs with IP provisions contain a statement affirming a general commitment to IP protection; more than 60 percent affirm the TRIPS Agreement; nearly half include a reference to World Intellectual Property Organization (WIPO) treaties; and approximately three-quarters include a statement on assistance, cooperation, or coordination. Some provisions explicitly promote technical assistance and capacity building between advanced and developing countries. Regarding specific areas of IPR, the authors of the WTO study found that the most common provisions concern geographical indications (GIs), copyrights, trademarks, patents, and new plant varieties. The next most common relate to industrial design, followed by traditional knowledge and genetic resources.Among the least common IP provisions are those concerning domain names, layout designs of integrated circuits, and encryption program-carrying satellite signals, all of which are found in fewer than one-fifth of all PTAs that include IP provisions. The WTO study also analyzed the prevalence of eleven provisions related to pharmaceuticals. Not surprisingly, these provisions are most common in agreements involving only developed countries, and to a much lesser extent, in agreements between developed and developing countries. The most common provisions, found in approximately one-third of the PTAs with IP provisions, are those concerning patentability criteria or patent subject matter, and those concerning compulsory licensing. Approximately one-quarter of the PTAs include a provision concerning data protection. 6 Valdés and McCann 2014. 205 Handbook of Deep Trade Agreements Based on a weighting of the 32 different types of IPR provisions analyzed, the authors generated an aggregate overall score for each agreement. On the basis of this score, the agreements were sorted into one of three categories according to their level of IP content - high, moderate, or negligible. The 176 agreements fell roughly equally into the three categories. Interestingly, however, while the overwhelming majority of PTAs signed by the United States and EFTA can be classified as high content, fewer than half of those signed by the EU and Japan meet this threshold. Concurrent with the WTO study, another study7 analyzed 256 PTAs notified to the WTO as of March 2013. Although drawing from the data source as the WTO study, the author of this second study adopted a slightly less restrictive view of what constitutes a PTA with an IPR provision. Consequently, this study included analysis of 25 PTAs that the authors of the WTO study would later exclude in their analysis. The second study also focused more on substantive provisions concerning enforcement rather than IPR provisions impacting public health.That study found similar patterns in terms of the IP subject areas covered by the PTAs, but also found a high variance in the number of IP provisions in PTAs between developing countries. About half of developing country PTAs included only one article within the treaty itself, while half included a specific annex or chapter on IP. A more recent study8 reviewed 357 PTAs that discuss IPR in a general or specific manner. This study focused on measuring three different concepts: the degree of IPR protection, the degree of IPR enforcement, and the coherence with multilateral IPR rules. For each concept, the authors devise an index with a series of questions to measure the variable. Not surprisingly, the authors found that PTAs involving the United States have the highest degree of IPR protection, followed by those with Japan, the Republic of Korea, and EFTA. North-South PTAs have the highest degree of protection, while South-South PTAs have the lowest, with intra-African PTAs lacking any IPR protection. In terms of enforcement, the patterns are roughly the same, but with Japan’s containing the strongest provisions. Finally, PTAs involving EFTA, Japan, and Korea have the highest level of coherence with multilateral treaties—exceeding those of the US and EU. Intra-European treaties also include many more requirements for accession to IP-related treaties than similar PTAs in the Americas or Asia. In addition to the three comprehensive analyses discussed above, several other studies have focused on IPR provisions in PTAs concentrated in a given region. One study examined 42 PTAs involving the Asia-Pacific region, finding great variation across these agreements.9 7 Seuba 2013. 8 Elsig and Surbeck 2016, based on the dataset developed by Dür, Baccini, and Elsig 2014 as part of the Design of International Trade Agreements (DESTA) project. 9 Puutio 2013. 206 Intellectual Property Rights It found that agreements between Asia-Pacific countries and the US or EU tend to have the most inflexible IPR provisions, demanding high standards even from poorer developing countries. By contrast, agreements involving Japan, Singapore, and Korea tend to show greater flexibility across the range of negotiating partners. Many PTAs involving China contain some mention of IPR, but the depth of commitments required is not necessarily very high. Another study focused on Asia10 discusses how PTAs interact with other regional initiatives, such as efforts by the Association of Southeast Asian Nations to increase IPR harmonization in the region. Two other studies11 provided excellent overviews of the salient features of the TRIPS-plus provisions found in US, EU, and other PTAs. Another study12 examined the degree of norm conflict and coherence across not only PTAs, but also bilateral investment treaties.Yet another study focused on specific TRIPS-plus provisions have affected the development and implementation of IPR regimes in developing countries, leading to shared challenges across countries.13 7.3. METHODOLOGY This study arises out of the database constructed as part of the World Bank’s Evolution of Deep Trade Agreements project. The study is based on the 295 PTAs signed through December 2016, and does not reflect PTAs signed in 2017 or later (e.g., the EU-Japan Economic Partnership Agreement). Efforts were made to update the contents of the PTA if it underwent any changes during the ratification period or renegotiations following the initial signing, as was the case for the Comprehensive and Progressive Trans-Pacific Partnership (CPTPP). A team of seven researchers at Harvard Law School identified possible elements of IPR-related disciplines within the PTAs; these elements were then coded. The process of identifying these fields and developing the coding template was an iterative process, whereby non-random groups of selected PTAs across time periods, negotiated by a diverse set of governments, were examined in batches. During each iteration, relevant provisions were identified.This allowed for the creation of a preliminary template, which was then revised again with the next batch, until the researchers were relatively confident that they had identified the relevant universe of potential elements and created a robust template. Based on this template, each agreement was coded by one researcher and subsequently checked by another. Each team of researchers involved at least one individual with experience 10 Quirino, Fider, and Gaba 2009. 11 Fink 2011; Roffe and Spennemann 2014. 12 Ruse-Khan 2011. 13 Biadleng and Maur 2011. 207 Handbook of Deep Trade Agreements in intellectual property law and one with experience in international trade law. Errors and coding inconsistencies were subsequently resolved by the group as a whole, in consultation with the principal investigator.Altogether, a total of 120 fields were coded for each agreement. These fields can be grouped across a range of categories, described below: • Accession to/ratification of existing international IP agreement(s): Coding for whether the agreements require the parties to accede to 15 international IP agreements, with each listed as a separate field. Examples include the Patent Cooperation Treaty, the Union for the Protection of NewVarieties of Plants (UPOV) Convention of 1991, the WIPO Copyright Treaty, the WIPO Performances and Phonogram Treaty, and the Trademark Law Treaty. • Incorporation of existing international IP agreements: Coding for whether the PTA explicitly incorporates the TRIPS Agreement and/or other multilateral IP agreements to which both parties are a party. Doing so renders these agreements subject to the dispute settlement mechanisms of the PTA. • Exhaustion: Coding for whether the PTA either provides for national exhaustion of IPRs or preserves the flexibility for each country to determine its own exhaustion scheme. • Trademarks: Coding for 16 fields that capture the range of trademark-related obligations found in PTAs, including many TRIPs-plus obligations. Some of the fields concern the scope of trademarks for which protection must be given (e.g., sound marks, collective marks) or the terms of protection. Others concern systemic issues governing trademarks (e.g., classification system, recordal) or procedural issues related to cancellation. • Geographical indications (GIs): Coding for a number of possible strategies to address GIs in PTAs. Examples include stipulating the scope of protection for GIs, designating a list of specific GIs to be protected, and stipulating that GIs can be registered or protected through a trademark system. • Copyright and related rights: Coding of 14 different fields to capture the various requirements in PTAs concerning copyright and related rights, including a term of protection beyond that required by the TRIPS Agreement, broadcast rights, and various exclusive rights.The coding also covers whether or not the PTA addresses newer issues not discussed in the TRIPS Agreement, such as requirements for digital rights management and protection against the circumvention of technological protection measures. • Patents: Coding of 15 different fields capturing a range of TRIPS-plus provisions related to patents.These types of provisions are among the most controversial in PTAs.They cover a wide range of issues including new use and/or new process patents for a known product, adjustment of the length of patent term, patent linkage, and patent revocation. 208 Intellectual Property Rights • Data protection/protection of undisclosed information: Coding for whether the PTA includes TRIPS-plus provisions that provide for a minimum term of protection for undisclosed test or other data for a new pharmaceutical product, agricultural chemical, and biologics.These types of provisions are also highly controversial, as countries may use PTAs that include these provisions to further specify the requirement of Article 39 of the TRIPS Agreement. • Industrial designs: Coding for whether the PTA requires protection of industrial designs, including design systems. • Biodiversity/traditional knowledge: Coding for whether the PTA includes a provision recognizing the importance of biodiversity and/or traditional knowledge, including commitments to preserve and protect. • Domain names/country names: Coding for whether the PTA includes any provision for settling disputes related to country-code, top-level domain names, as well as any provision to prevent the misleading commercial use of a country name. • Enforcement: Coding for more than 20 enforcement-related provisions found in various IPR chapters to capture whether the PTA reiterates and/or elaborate upon the various enforcement requirements discussed in the TRIPS Agreement. • Transparency/cooperation: Coding for whether the PTA includes requirements for greater transparency of registrations for trademarks, GIs, industrial designs, and new plant varieties. Also coding of four additional fields to attempt to capture any requirement for greater cooperation and/or harmonization among PTA partners on IPR-related issues. Beyond simply coding for whether or not a particular provision is included in the PTA, the database also includes information about the enforceability of the provision. Researchers highlighted whether or not the PTA contained binding language with state-to-state dispute settlement (DS); or contained binding language (“must,” “shall”) but without a formal DS mechanism, or included a best-endeavor provision, or included no binding language whatsoever. Altogether, 105 PTAs were coded. Note that the number of PTAs coded in this study is significantly fewer than that of the previous major studies, some of which examined more than twice as many PTAs, for two reasons. First, the overall goal of the World Bank project has been to analyze deep-integration trade agreements. Whereas earlier studies examined all PTAs with any mention of IPR, this study chose to exclude trade agreements with only shallow IPR commitments (e.g., a general provision discussing IPR) that did 209 Handbook of Deep Trade Agreements not promote any meaningful integration. Second, the specific goal of this study has been to provide a broader and more comprehensive analysis of the specific IPR commitments in PTAs than what exists to date. Unlike other previous major studies, this study did not seek to create yet another numerical index to measure the depth of all PTAs with IPR provisions. Instead, a decision was made to examine in-depth the most critical subset of PTAs, with the greatest variety of TRIPS+ provisions, to understand how this facilitates trade integration. Finally, a few limitations in the coding methodology should be noted. First, in line with what was agreed upon among the principal investigators for the Deep Integration project, this study employs a binary coding methodology. Consequently, it is only able to capture whether or not a particular IPR-related provision is included in a given PTA. Except for enforceability, it does not distinguish among variations in the legal commitment, as would be possible using a non-binary variable. For example, a PTA in which the term of protection for undisclosed data is five years and one in which is ten years are coded the same (as 1), even though the depth of commitment clearly differs between these agreements. Another example is that the database does not distinguish between a PTA with a long list of specific GIs to be protected and one with a short list; both would be coded as simply including a list. To compensate for this limitation, the template includes additional columns to allow for a form of non-binary coding. Second, the various fields are not of near-equivalent importance, and the database is not intended to be used to measure the strength or depth of a PTA’s IPR content through a simple counting exercise, or through one which applies weights to the various binary fields. Different fields present a vastly different impact on populations. For example, new IPR requirements on pharmaceuticals can exert an extremely different impact on citizens’ welfare than those for industrial designs. The hope is that future researchers will use the information captured in this database to tailor an approach specific to their specific research question, rather than simply add up the number of commitments to reflect a proxy for depth. In most instances, the IPR chapters of the 105 PTAs were written in English. However, in instances where the Spanish or French text was authoritative, a researcher fluent in the language was assigned to perform the coding exercise, with another researcher conversant in the language assigned to check the coding. While this approach sought to minimize coding errors, some errors may exist because of the complexity of the IPR provisions and the varied approaches taken in drafting the legal text. As these are discovered, the coding will be updated in the database. 210 Intellectual Property Rights 7.4. FINDINGS This study examines the specific modes through which particular major trading powers - the US and European countries - have sought to use IPR-related provisions to advance their vision of deeper integration with their trading partners. Although there are some similarities in the objectives sought, there are also vast differences among these approaches. This section seeks to draw attention to such differences. 7.4.1 Prevalence in PTAs The inclusion of IPR-related provisions in PTAs is a relatively recent phenomenon (Figure 7.1). Prior to the WTO’s creation in 1995, a handful of PTAs, such as the North American Free Trade Agreement, included an IPR chapter.These helped to lay the groundwork for the TRIPS Agreement. Subsequently, IPR provisions have become more commonplace. Figure 7.1: Percentage of PTAs with IPR-related provisions, 1991-2005 Several factors precipitated their growing presence in trade agreements. First, not entirely satisfied with the outcome of the TRIPS bargain, interest groups in developed countries pressed their governments to expand on IPR commitments through the pursuit of TRIPS- plus obligations in PTAs. Second, with the expansion of offshoring and outsourcing, stronger IPR protection became a major market access and investment concern for multinational corporations. Finally, the growth of digital technologies has also contributed to the increased economic importance of IPRs to firms engaged in cross-border trade. 211 Handbook of Deep Trade Agreements Until recently, these provisions were found primarily in PTAs involving developed countries Figure 7.2 shows that through 2000, only about 10 percent of PTAs between developing countries included any form of IPR-related provisions through 2000. This percentage increased between 2001 and 2010, but only slightly. Figure 7.2: Percentage of PTAs concluded among developing countries with IPR-related provisions, 1991-2005 After 2010, however, the prevalence of IPR-related provisions in PTAs between developing countries increased significantly, to about 40 percent of all such PTAs between 2011 and 2015. This important shift suggests that at least a number of developing countries have internalized the need for IPRs to allow for deeper integration via PTAs. Why might this be the case? One explanation is that as firms in developing countries have become more deeply embedded in global value chains over the past decade, their needs have come to reflect more closely that of firms in advanced economies.Therefore, they too seek higher levels of protection for their firms via trade agreements. Participation in global value chains shifts the economic interest of firms in developing countries, and/or causes these firms to internalize norms held by other firms with which they have developed linkages. Another explanation is that the shift has been government driven rather than firm driven. This may be the case because the TRIPS Agreement does not contain a most-favored-nation (MFN) exception for PTAs, akin to GATT Article XXIV or GATS Article V. Therefore, once a country agrees to a higher level of IPR protection in a trade agreement, it must extend this higher level of protection to not just its PTA partner, but to all WTO members on an MFN basis. Therefore, even if Country B agrees to a higher TRIPS-plus standard in a PTA 212 Intellectual Property Rights at the behest of Country A, it may nevertheless turn around and demand the same standard of Country C in a subsequent trade agreement. After all, through its PTA with Country A, it has already granted such a benefit to Country C; it simply seeks a reciprocal arrangement. This logic is in line with the findings of an early study14 that PTAs are clearly a driver for IP reform in developing countries. That study suggested that entering into a PTA with robust IPR obligations can lead the government of a developing country to revisit a number of matters, including trade agreements with third parties. With these trends in mind, the study now examines specific forms of deep-integration IPR chapters that have proliferated in recent years. 7.4.2 US PTAs The quintessential model of a TRIPS-plus IPR chapter is that sought by the United States in its PTAs. Several earlier studies15 have examined the salient features of this model in detail. The aim here is not to provide a comprehensive accounting of these features, but instead offer a sense of the depth of commitments sought by the US on IPRs. As a condition for deeper integration with the US economy, US trade negotiators typically demand that PTA partners agree to an extensive number of IPR-related obligations. In short, they demand that the PTA partners bring themselves to a higher level of IPR protection, including adopting rules that may be similar to those of the US. Not surprisingly, this has engendered much controversy. Although the level of economic development differs quite significantly across trading partners, the types of IPR commitments sought by US negotiators are relatively consistent. Why is this the case? One recent study16 has argued that US trade negotiators have little flexibility to do otherwise because their hands are tied by Congress. Although Congress has the constitutional power to regulate foreign commerce, Congress has regularly delegated this power to the Executive Branch through the grant of Trade Promotion Authority. In doing so, Congress lays out explicit objectives for what it expects of US trade negotiators. In its latest incarnation, the 2015 Bipartisan Congressional Trade Priorities and Accountability Act states that one of the principal negotiating objectives shall be “ensuring that the provisions of any trade agreement governing intellectual property rights that is entered into by the United States reflect a standard of protection similar to that found in United States law.”17 14 Bladgleng and Maur 2011. 15 Including, most notably, Roffe, Spennemann, and von Braun 2010. 16 Claussen 2018. 17 Bipartisan Congressional Trade Priorities and Accountability Act of 2015, §2(b)(5)(A)(i)(II). 213 Handbook of Deep Trade Agreements This includes “providing for strong protection for new and emerging technologies and new methods of transmitting and distributing products embodying intellectual property, including in a manner that facilitates legitimate digital trade.”18 Yet another requirement is “to secure fair, equitable, and nondiscriminatory market access opportunities for United States persons that rely upon intellectual property protection.”19 In light of these requirements, the US has set forth a relatively clear template for the IPR chapters in its PTAs, and this template has evolved only moderately over time. Several scholars20 have argued that the consistency of IPR chapters in US PTAs is due to the fact that the negotiators are captured by private industry. They have highlighted how the various US trade advisory committees with privileged access to trade negotiators are disproportionately represented by lobbyists from the pharmaceutical, entertainment, and software industries. Until the membership of these committees is more balanced, the content of US IPR chapters is likely to remain relatively consistent across agreements, reflecting the TRIPS-plus demands sought by such industries. Among the most controversial demands are the TRIPS-plus provisions sought for patents and pharmaceutical test data. Most US PTAs require patent term extension to compensate for delays under certain circumstances, such as during the regulatory approval process. Some PTAs also clarify the circumstances under which a compulsory license can be issued pursuant to the TRIPS Agreement; these are most restrictive in the US PTAs concluded with other advanced economies. Several US PTAs also expand the scope of patentability. For example, some require that patents be made available for new uses or for new methods of a known product. One area where there is considerable variation across US PTAs is with regard to whether the trading partner must provide patents for plants and animals. Some US PTAs require that this be the case, whereas others ask only that the PTA partner make “reasonable efforts” to do so. In addition, whereas the TRIPS Agreement requires only that test data submitted to obtain marketing approval be protected against “unfair commercial use,” US PTAs set forth an explicit minimum term of protection for undisclosed test or other data. This includes data on new agricultural chemicals as well as pharmaceutical products. Furthermore, in some PTAs, additional protection may be triggered by the provision of “new clinical information.” Over the years, the language of the requirement for pharmaceuticals has evolved to include a new indication, formulation, or administration method for an already-approved pharmaceutical product as well as for a pharmaceutical product containing a new chemical entity. Most recently, the US has sought a minimum term of protection for biologics as well. 18 Ibid, §2(b)(5)(A)(ii). 19 Ibid, §2(b)(5)(B). 20 See, e.g., Sell 2003; Moberg 2014. 214 Intellectual Property Rights Another area where the US PTAs have been a trailblazer is with respect to new technologies. The US requires that its PTA partner adopt a range of technological protection measures to guard against infringement, including banning circumvention devices and providing for criminal liability in the case of willful infringement for commercial purposes. Several US PTAs incorporate elements of the US Digital Millennium Copyright Act, requiring that trading partners limit the liability of internet service providers as long as they take down infringing content upon notification. In addition, US PTAs are among the very few that contain an explicit requirement for the settlement of disputes related to top-level domain names (e.g., .com, .net). On copyright term, since the US-Singapore PTA was signed in November 2000, the US has insisted that the term be the life of the author plus 70 years. Where the copyright term is decided based on other criteria, the US has insisted that it be at least 70 years from the publication or creation of the work. Certain PTAs, such as the US-Oman PTA, have resulted in even longer terms. With regard to trademarks, the US has also demanded that its PTA partners adopt certain TRIPS-plus requirements. Some examples include a requirement that trademarks include collective and certification marks, that certain types of signs must be eligible for trademark protection, and that sound and scent marks must be eligible for trademark protection. In addition, the US has also demanded the adoption of certain procedural requirements to allow for examination of/opposition to a trademark application and an application to cancel a trademark. In addition, the US has sought to require the establishment of an electronic trademark system and to prohibit recordal of a trademark license to establish license validity or as a condition for use. Finally, on enforcement, the US has sought TRIPS-plus obligations that expand on what it views as one of the weaker elements of the TRIPS Agreement. For example, several US PTAs require that border authorities shall have ex officio authority to detain suspected counterfeit or pirated goods, and to order their destruction. In addition, border authorities are required, under US PTAs, to allow for application by the rights holder to detain and suspend the release of any infringing good. Many US PTAs also require that infringing goods, if not destroyed, must be disposed of outside of the normal channels of commerce. In sum, the US has pursued an aggressive effort to elevate and expand the standards of IPR protection through its PTAs, in line with what Congress has demanded of US trade negotiators. Countries aspiring to deeper economic integration with the US through a trade agreement know well in advance what types of IPR commitments are expected of it in exchange. The end result is that developing countries have been confronted with a difficult choice as far as whether this trade-off is worthwhile, given the uncertainty of future benefits arising out of having a preferential trade arrangement with the US. Some eventually determine that the 215 Handbook of Deep Trade Agreements price is too high, especially given domestic political sensitivities surrounding pharmaceutical products, traditional knowledge, and biodiversity. For instance, exploratory negotiations with the Southern African Customs Union and with Thailand failed to progress to actual PTAs, in part due to IPR demands by US negotiators. Other developing countries agree to US demands begrudgingly, because, although they view the IPR concessions as costly, they deem them to be outweighed by the benefits of deeper integration with the US. Finally, some countries do so willingly because they view the higher IPR standards as in their long-term interests. The net effect is that US PTAs have become an important instrument to persuade (or pressure) a limited set of countries to adopt stronger IPR standards than are provided for by the TRIPS Agreement, especially in light of the negotiating stalemate at the WTO. 7.4.3 European Union PTAs European Union PTAs also use the allure of preferential trade access to demand higher IPR standards from trading partners. Over time, the scope of EU negotiating issues for IPR has expanded, as well as the depth of the TRIPS-plus provisions sought through the negotiations. Two factors have driven this evolution. First, there has been a shift in the EU’s conception of the role of trade agreements in managing its bilateral economic relationships with the African, Caribbean, and Pacific (ACP) countries, many of which are former colonies of EU member states. Originally, these trade relationships were managed through a series of non-reciprocal commitments, as set forth in the Lomé Convention signed in 1975 between the European Economic Community and 71 ACP countries, and the subsequent Conotou Agreement between the EU and ACP countries, signed in 2000. These agreements sought very little in the way of TRIPS-plus obligations. However, in 2006, the European Commission put forward the “Global Europe” strategy, through which trade agreements would serve as a tool to deepen Europe’s economic relationships and global competitiveness. Stronger and more robust IPR rules came to be viewed as a necessary vehicle to deepen the trade and investment relationships necessary to achieve this vision. Soon after this shift occurred, two studies21 highlighted the potential for a wider range of issues to be considered in the EU’s negotiations with ACP countries. A later study22 then explored the bargaining power of developing countries vis-à-vis the EU in these negotiations, and how asymmetries in these power relationships enabled the EU to impose standards similar to those found in its legislation without giving much consideration to the domestic conditions of developing countries. 21 Shabalala and Bernasconi 2007; Third World Network 2009. 22 Moerland 2017. 216 Intellectual Property Rights Second, over the same period of time, the EU accelerated its conclusion of PTAs with upper- middle-income developing countries and advanced economies (e.g., Korea, Singapore, and Canada), many of which were already used to embracing high-standard IPR chapters as part of their PTAs. This was, in part, to ensure that European exporters were not placed at a competitive advantage vis-à-vis American or Japanese exporters, which were also deepening their economic relationships with key countries. This shift in the profile of PTA partners has also contributed to the expanding scope and depth of IPR provisions sought in the EU’s PTAs. Traditionally, the IPR provisions of EU PTAs have focused most actively on obligations concerning geographical indications. One concern of European producers of agricultural products as well as wine and spirits has been the relative laxness of countries, particularly in the so-called “New World,” to allow for use of geographical names as long as the use does not mislead consumers as to the true origin and nature of the product. Some examples of the type of use that European producers deem problematic include “Champagne-like sparkling wine” or “locally made Parma ham.” Many EU PTAs have included a list of specific GIs that must be protected by both parties. As a result, non-original producers in the PTA partner must phase out their use of the geographical name altogether. Several EU PTAs also require that the PTA partner establish a dedicated system for GI protection, as opposed to doing so via the trademark system. This is in stark contrast to the approach taken by US PTAs. While TRIPS-plus commitments on GIs remain an important negotiating priority for the EU, other TRIPS-plus obligations have also entered into EU PTAs that are similar to those found in US PTAs. For example, several EU PTAs now require accession to the International Convention for the Protection of New Plants (UPOV Convention). They also seek patent term extension in the case of unreasonable delays. In addition, several stipulate terms for the protection of undisclosed test data, similar to terms found in US PTAs. The same is true for copyright-related provisions. Recent EU PTAs stipulate a minimum copyright term and require that certain rights be provided to performers of unfixed (live) performances, similar to US PTAs. In addition, EU PTAs also include TRIPS-plus provisions to guard against circumvention of technological protection measures and against alteration of rights management information. EU PTAs have also aggressively expanded on IPR-related obligations related to enforcement. These obligations have been heavily influenced by the EU’s own approach, notably including Directive 2004/48 and two regulations dealing with border measures.23 The enforcement obligations seek to expand the authority of border authorities and judicial authorities in combating infringement, including of pirated copyright products, counterfeit goods, and goods that violate a design right or GI. 23 Fink 2011. 217 Handbook of Deep Trade Agreements One area where the EU has been more open to developing countries’ interests, if only indirectly, is with regard to provisions on biodiversity and traditional knowledge. For example, as part of the sustainable development provisions in a PTA, there may be a mention of the Convention on Biological Diversity and the need for appropriate measures to preserve traditional knowledge, including working towards the development of internationally agreed sui generis models for their legal protection. Lastly, it is worth noting that EU PTAs can serve as a mechanism for dealing with intra-EU issues concerning IPR. Consider the example of the World Intellectual Property Organization (WIPO) Copyright Treaty and the WIPO Performances and Phonograms Treaty. Several EU member-states had signed but not ratified the treaties by the time they entered into force in 2002. As part of the EU-Chile PTA, however, the parties agreed to accede to the treaties by January 1, 2007. At the time of the negotiations, Chile had already ratified the agreement, so the main burden of this obligation fell upon the EU. Following EU enlargement in 2004 and 2007, a situation arose in which some EU member states had already ratified the treaties whereas others had not. However, earlier PTA commitments had already locked the entire EU into doing so. This helped spur the EU as a whole to ratify the treaties in late 2009, albeit nearly three years later than the deadline set forth in the EU-Chile PTA. 7.4.4 European Free Trade Area PTAs The four countries that account for EFTA - Iceland, Liechtenstein, Norway, and Switzerland - all operate as part of the European Single Market, but not as part of the European Union’s customs union. Despite the relatively small size of their economies, the EFTA nations serve as yet another hub for PTAs with deep-integration IPR chapters. Although they share many points of overlap with the US and EU, the EFTA PTAs nevertheless have their own distinctive qualities, which allow them to emerge as yet another model. While much of the focus has been on the EU’s policies toward advancing TRIPS-plus provisions in its PTAs, the EFTA nations have advanced a similar agenda with more than 20 countries.These include a wide range of developing countries, such as Mexico, Morocco, Montenegro, and Ukraine. Certainly, a number of the IPR provisions found in the EFTA PTAs resemble those of the US and EU. For example, EFTA nations also seek for their trading partners to sign on to the WIPO internet treaties as well as the UPOV Convention. They also seek to expand IPR enforcement provisions, including clearly stipulating the range of power and authority of judicial and border authorities to combat infringement. However, a few distinctions are worth noting. In contrast to US and EU PTAs, the IPR provisions in EFTA PTAs do not focus as heavily on copyright-related issues. With the exception of the EFTA PTAs with the Balkan countries, many of the other EFTA PTAs do not insist on a copyright term beyond that provided by the TRIPS Agreement. Nor do they clarify the rights associated with unfixed performances, an issue that is addressed in most US and many EU PTAs. 218 Intellectual Property Rights As is true of the US and EU PTAs, EFTA PTAs also include a TRIPS-plus obligation for patent term extension in the case of unreasonable regulatory delays. They also include requirements for the protection of undisclosed test data for agricultural chemicals and pharmaceuticals. Obligations concerning test data exclusivity are found in a much greater proportion of EFTA PTAs than EU PTAs, reflecting the relative importance of pharmaceutical exports for certain EFTA countries, most notably Switzerland. Despite overall similarity in the scope of coverage for pharmaceutical issues, there are two interesting divergences found in particular EFTA trade agreements.The first is the inclusion of a disclosure requirement for the origin or source of genetic material in patent applications in the EFTA-Colombia PTA. This type of disclosure has been a long-sought-after goal of developing countries that are concerned with multinational corporations deriving unfair profits from biological material and/or traditional knowledge.24 It is notable that EFTA nations were open to Colombia’s request for inclusion of a disclosure requirement whereas the US and EU were not. The second concerns the inclusion of a compensatory alternative to a fixed term of exclusivity for undisclosed test data.This can be found in the EFTA-Korea PTA. Some scholars25 have advocated the inclusion of this option to foster competition, and by implication, lower drug prices. Altogether, the EFTA PTAs represent yet another model for how advanced economies have sought to promote deeper integration with TRIPS-plus provisions, while seeking to accommodate the interests of trading partners. While much of the spotlight has been on the US and EU, the EFTA countries have managed to carve out their own distinctive approach. 7.4.5 PTAs of advanced Asian economies Finally, a number of advanced economies in the western Pacific have emerged as yet another focal point for IPR chapters with robust TRIPS-plus rules. These include Australia, Japan, Singapore, and Korea. Many of these countries have PTAs with one another. Many also have PTAs with the US and/or the EU. Through these interactions, they have emerged as leading proponents of robust TRIPS-plus rules in PTAs. The priorities of the four advanced Asian economies are not necessarily identical. Japan, for example, has been a leading proponent of plant patenting,26 although Japanese negotiators traditionally have not insisted on accession to a wide range of IPR treaties as an offensive demand for their PTAs, but they are not averse to their inclusion, if the PTA partner should so request. By contrast, Australia has insisted on accession to a range of treaties in its PTA with the Association of Southeast Asian Nations (ASEAN), including the WIPO internet treaties and the UPOV Convention. 24 Fink 2011. 25 For example, Reichman 2004. 26 Lindstrom 2010. 219 Handbook of Deep Trade Agreements Over time, there has been considerable variance among these PTAs. For example, the Australia- China PTA is much lighter in terms of IPR demands, even though the PTA itself is clearly intended to foster closer integration between the two economies. Overall, however, as these Asian countries have engaged in greater number of PTAs, they have gradually become accustomed to more robust IPR rules and have sought to advance them on their own accord in PTAs with other countries. With the conclusion of the CPTPP by eleven Asia-Pacific nations (including Australia, Japan, and Singapore), there is now much greater convergence along the lines of a single set of IPR rules. The CPTPP includes one of the highest-standard TRIPS-plus IPR chapters of any PTA. While the withdrawal of the US from the agreement led to the suspension of ten IPR provisions, the vast majority of the original IPR chapter remained squarely in place.27 This includes several TRIPS-plus requirements on trademarks, patent revocation, and enforcement powers for judicial and border authorities. While one might have once supposed the high-standard IPR rules to be American-led demands, the CPTPP experience makes clear that this is not the case. Rather, the CPTPP represents rules that advanced Asian economies have internalized and are advancing on their own accord as they integrate among themselves and with others in the Asia-Pacific. 7.5. CONCLUSIONS With the growing importance of knowledge-driven innovation, one of the core elements of a deep integration trade agreement is a set of robust rules governing IPR. Dissatisfied with what they viewed as inadequate protections arising out of the TRIPS Agreement, several advanced economies have sought to advance TRIPS-plus rules in their PTAs. As this chapter has discussed, the move toward broader and more enforceable IPR commitments has been driven by four different sets of WTO members: the US, the EU, EFTA, and a group of advanced economies in the western Pacific. While there is considerable overlap in what they seek, there are also important differences in areas such as GIs, biodiversity, and biologics. In the coming decade, three important questions are likely to rise to the fore: The first is whether any additional hubs will develop beyond the four mentioned above. To date, all of the hubs pushing forward with TRIPS-plus rules in PTAs are composed of advanced economies. However, as this chapter has noted, the frequency with which IPR-related provisions are included in PTAs concluded among developing countries has increased since 2011. Will this eventually result in a group of developing countries with a deep integration trade agenda to develop their own hub (e.g., a Pacific alliance in Latin America)? Or will the major centers for the development of TRIPS-plus rules remain in the advanced economies? 27 Including provisions concerning patent term adjustment, biologics, copyright term, trademark protection, rights management information, protection of encrypted program-carrying satellites, legal remedies, and safe harbors. See Trans-Pacific Partnership Ministerial Statement, November 11, 2017. 220 Intellectual Property Rights A second question is whether there will be further consolidation of the existing models of IPR-related provisions that have arisen out of the four hubs. Already, there are cross-regional PTAs forming across some of these hubs, especially between the advanced Asian economies and the other hubs. With the US and EU already actively exploring the possibility of a trans-Atlantic PTA and the US open to rejoining the CPTPP, additional possibilities exist for further deep integration. How will this affect the development of TRIPS-plus rules and norms? Will this lead to even greater harmonization amongst the major economies? If so, will this be along the lines established in CPTPP, given its first-mover status as a mega- regional PTA, or be based on another model? Finally, how will the growing expectation among some economies that some robust IPR rules must be in place affect the future of deep integration trade relationships? For example, divergent viewpoints over the depth of TRIPS-plus IPR provisions have emerged as a sticking point in the Regional Comprehensive Economic Partnership (RCEP) negotiations among some Asian and Pacific states. Will this lead some major developing countries, such as India or South Africa, to eventually succumb to PTAs with strong IPR obligations? Or will they choose to remain outside of the universe of deep integration PTAs that tie together global value chains because they deem the costs of joining to be too high? Furthermore, as the 2018-19 Sino-US trade war has highlighted, the failure to have an enforcement mechanism for agreed-upon IPR rules can generate economic uncertainty and temporary disruptions for a deeply integrated economic relationship. PTAs serve a valuable role in mitigating this risk. If a deep integration proceeds between two trading partners without a PTA, will this mean more frequent recourse to power to resolve IPR-related tensions? What is clear is that IPR-related provisions are likely to continue serving as a flashpoint for deep integration PTAs, much as they have done so far. The purpose of this study has been to build a database that captures the richness and depth of such provisions. While much of the existing academic literature has focused on pharmaceuticals and copyright as the crux of this flashpoint, there is much more in the way of other issues, such as TRIPS-plus provisions on enforcement or industrial design, which may also impact the formation of value chains. The hope is that this database will allow academic researchers, policymakers, and civil society groups to better explore the depth of these phenomena, so that they can answer the questions over trade and IPR that inevitably will arise as technology drives deeper economic integration. ACKNOWLEDGMENTS The author extends his gratitude to the World Bank staff, the team of Harvard Law School students, and other participants at the project workshop held at the World Bank in February 2018 for their comments and useful insights. Special thanks to Carsten Fink, Keith Maskus, Pedro Roffe, and the WTO staff for sharing their expertise from prior projects and to Bernard Hoekman for his suggestions on the focus of this chapter. 221 Handbook of Deep Trade Agreements REFERENCES Baldwin, Richard. 2016. The great convergence: Information technology Moberg, Katrina. 2014. Private industry’s impact on US trade law and the new globalization. Cambridge, MA: Belknap Press. and international intellectual property law:A study of post-TRIPS US bilateral agreements and the capture of USTR. Journal of the Biadgleng, Ernias Tekeste, and Jean-Christophe Maur. 2011. Patent and Trademark Office Society 96(2): 228-56. The influence of preferential trade agreements on the implementation of intellectual property rights in developing Moerland, Anke. 2017. Do developing countries have a say? countries: A first look. UNCTAD–ICTSD Project on IPRs Bilateral and regional intellectual property negotiations with and Sustainable Development, Issue Paper 33. the EU. International Review of Intellectual Property and Competition Law 48(7): 760-83. Branstetter, Lee, C. Fritz Foley, and Kamal Saggi. 2010. Has the shift to stronger intellectual property rights promoted technology transfer, Puutio, Teemu Alexander. 2013. Intellectual property rights FDI, and industrial development?WIPO Journal 2 (1): 93-98. in regional trade agreements of the Asia-Pacific economies. Working Paper 124, Asia-Pacific Research and Training Claussen, Kathleen. 2018. Separation of trade law powers.Yale Network on Trade (ARTNeT), Bangkok. Journal of International Law 43(2): 315-54. Puutio, T. Alexander, and Luca Parisotto. 2015. Intellectual Dür, Andreas, Leonardo Baccini, and Manfred Elsig. 2014. The property rights in the Asia-Pacific trade context. UNESCAP design of international trade agreements introducing a new Trade, Investment, and Innovation Working Paper 2. database. Review of International Organizations 9(3): 353-75. Quirino, Aleli Angela G., Alex Ferdinand S. Fider, and John Elsig, Manfred, and Jenny Surbeck. 2016. Intellectual property Paul M. Gaba. 2009. Regional agreement affecting intellectual rights and preferential trade agreements. NCCR Trade property in ASEAN. Paper prepared for AIIPI Forum. Regulation Working Paper 2016/2. Roffe, Pedro, and Christoph Spennemann. 2014. Preferential Fink, Carsten. 2011. Intellectual property rights. In Preferential trade agreements and intellectual property rights. In Intellectual trade agreement policies for development:A handbook, edited by Jean- property rights: Legal and economic challenges for development, edited Pierre Chauffour and Jean-Christophe Maur.Washington, DC: by Mario Cimoli et al. Oxford: Oxford University Press. World Bank. Roffe, Pedro, Christoph Spennemann, and Johanna von Braun. Hoekman, Bernard, Keith E. Maskus, and Kamal Saggi. 2005. 2010. Intellectual property rights in free trade agreements: Moving Transfer of technology to developing countries: Unilateral and beyond TRIPS minimum standards. In research handbook on the multilateral policy options. World Development 33: 1587-1602. protection of intellectual property under WTO rules, edited by Carlos M. Correa. Cheltenham, UK: Edward Elgar. Hoekman, Bernard, and Kamal Saggi. 2007. Intellectual property provisions in North-South trade agreements. CEPR Ruse-Khan, Henning Grosse. 2011 Protecting intellectual Working Paper 6460, Centre for Economic Policy Research, poperty under BITs, FTAs, and TRIPS: Conflicting regimes London. or mutual coherence. In Evolution in investment treaty law and arbitration, edited by Chester Brown and Kate Miles. Cambridge: Lindstrom, Beatrice. 2010. “Scaling back TRIPS-plus: An analysis of Cambridge University Press. intellectual property provisions in trade agreements and implications for Asia and the Pacific.” International Law and Politics 42(3): 917-80. Sell, Susan. 2003. Private power, public law: The globalization of intellectual property rights. Cambridge: Cambridge University Press. Maskus, Keith E. 2005. The role of intellectual property rights in encouraging foreign direct investment and technology transfer. Sueba, Xavier. 2013. Intellectual property in preferential trade In Intellectual property and development, edited by Carsten Fink and agreements: What treaties, what content? Journal of World Keith E. Maskus.Washington, DC:World Bank. Intellectual Property 16(5-6): 240-61. Maskus, Keith E., and William Ridley. 2017. Intellectual Third World Network. 2009. EU EPAs: Economic and Social property-related preferential trade agreements and the Development Implications. Report manuscript. composition of trade. Unpublished paper. Valdés, Raymundo, and Maegan McMann. 2014. Intellectual Maskus, Keith E., Kamal Saggi, and Thitima Puttitanun. 2005. property provisions in regional trade agreements: Revision and Patent rights and international technology transfer through update. In Regional trade agreements and the multilateral trading direct investment and licensing. In International public goods system, edited by Rohini Acharya. Cambridge: Cambridge and transfer of technology under a globalized intellectual property University Press. regime, edited by Keith E. Maskus and Jerome H. Reichman. Cambridge: Cambridge University Press. 222 Visa and Asylum CHAPTER 8 Visa and Asylum J. Pauwelyn, T. Nguyen, and K. Kamal 223 Handbook of Deep Trade Agreements CHAPTER 8 Visa and Asylum J. Pauwelyn*, T. Nguyen†, and K. Kamal† * Graduate Institute of International and Development Studies, Geneva, Switzerland † Georgetown University Law Center, Washington, DC, United States CONTENTS 8.1. INTRODUCTION 225 8.2. CONTEXT AND RELEVANT LITERATURE 225 8.2.1 The WTO and beyond 226 8.2.2 Selected literature 227 8.3. DEFINITIONS 228 8.4. METHODOLOGY FOR MAPPING 229 8.4.1 Pre-selection of PTAs to be coded 229 8.4.2 Questions coded for the selected PTAs 230 8.4.3 Stylized facts and insights derived from the coding 231 8.4.3.1 PTAs covering visa and asylum – absolute number and evolution over time 231 8.4.3.2 Which economies have concluded visa and asylum PTAs? 233 8.4.3.3 What types of movement of persons are covered? 235 8.4.3.4 How is the movement of persons facilitated? 237 8.4.3.5 What exceptions to movement of persons are provided for? 240 8.4.3.6 Do visa and asylum PTAs refer to other migration instruments? 240 8.4.3.7 Enforceability of visa and asylum provisions 241 8.4.3.8 Do “visa and asylum” provisions benefit non-parties? 243 8.5. MAIN FINDINGS AND CONCLUSIONS 244 ACKNOWLEDGMENTS 246 REFERENCES 247 224 Visa and Asylum 8.1. INTRODUCTION This chapter assesses the presence, depth, and geographical distribution of visa and asylum provisions in preferential trade agreements (PTAs). The aim is to map the extent to which PTAs include migration-related provisions that go beyond what is covered by the WTO. The WTO’s GATS agreement includes liberalization commitments with respect of services supplied through the movement of natural persons (GATS, mode 4) when, for example, an engineer travels abroad to supervise the construction of a dam (supply of construction services). This comes in addition to trade in services supplied cross-border (mode 1, e.g., telecom services), via consumption abroad (mode 2, e.g., tourism) and by means of commercial presence abroad (modes 3, basically, foreign direct investment in services). This chapter examines the extent to which PTAs address migration beyond GATS mode 4 (so-called WTO-extra provisions). Section 8.2 provides context and surveys relevant literature. Section 8.3 sets out the definitions used in this chapter. Section 8.4 describes the methodology adopted for the mapping of relevant PTAs. Section 8.5 describes stylized facts and insights derived from the coding of 100 PTAs. Section 8.6 offers a summary of main findings and a conclusion. 8.2. CONTEXT AND RELEVANT LITERATURE Openness through the movement of goods, services, capital, or people should, according to standard economic theory, affect prices and wages in the same way: by benefiting the abundant factor of production in the country while hurting the scarce factor.1 Liberalizing the flow of goods, services, or people should, therefore, produce similar gains (and related costs).Yet, actual trade agreements tend to focus more on liberalizing the flow of goods (and to some extent services) and relatively little on the movement of people. One economist put it bluntly: If international policymakers were really interested in maximizing worldwide efficiency, they would spend little of their energies on a new trade round or on the international financial architecture.They would all be busy at work liberalizing immigration restrictions.2 According to one calculation, a modest increase in industrial countries’ quotas on incoming temporary workers (an increase equal to 3 percent of the current work force) would have a global welfare increase far greater than the projected benefits of completing the WTO’s Doha Round, on which scores of diplomats have been laboring (unsuccessfully) for the last 17 years.3 1 See Peters 2015, Bradford, 2013, Trachtman 2009. 2 Rodrik 2001. 3 Trachtman 2014. 225 Handbook of Deep Trade Agreements Political realities largely explain the resistance, especially in wealthy developed countries, to liberalizing the flow of workers. As an expert in international law and economics points out, “[t]he distributive problem … is that most of the increased welfare stays in the hands of the migrants … citizens of wealthy states fear that increased immigration will either reduce wages or jobs, disproportionately absorb state funds for public services and transfer payments, or both.”4 Indeed, the global welfare gains from economic migration (which there undoubtedly are) have been unequally distributed within states and across nations. This often means that “migration is a one-way street,”5 with little room for reciprocity in a migration-only agreement. That, in turn, could lead one to think that linkage is the way forward: for example, wealthy countries admitting more immigrants from poorer countries in exchange for market access for goods, services, and investment from the wealthy countries into the poorer countries. How much of this linkage in trade agreements has actually occurred? This is the question examined in this chapter, from the vantage point of PTAs in the World Bank’s PTA database. 8.2.1 The WTO and beyond In the WTO’s General Agreement on Trade in Services (GATS), some linkage has happened through the liberalization of movement of natural persons supplying a service in another country (GATS, mode 4). However, this liberalization is limited to people working in one country as a service supplier (or employed by a service supplier) and then being allowed to supply their services to or in another country.6 The focus is on - mostly high-skilled - people who are already employed (or self-employed) in the country of origin; it excludes actual migration on a more permanent basis as well as people moving cross-border to look for work.7 “GATS is thus concerned with facilitating the provision of services by foreign entities, not with facilitating access to employment in the local labor market.”8 In addition, actual commitments under GATS mode 4 are country and sector-specific and so far rather limited.9 This raises the question of how much further PTAs can go in liberalizing the movement of persons.They can, of course, add to existing WTO commitments by enhancing mode 4 services commitments (such provisions would be coded as WTO+). More interestingly, PTAs could also provide for WTO-extra liberalization or facilitation of migration, that is, by means outside the 4 Trachtman 2018. 5 Ibid., at 482. 6 See GATS Annex on Movement of Natural Persons Supplying Services under the Agreement, para. 1: “This Annex applies to measures affecting natural persons who are service suppliers of a Member, and natural persons of a Member who are employed by a service supplier of a Member, in respect of the supply of a service.” 7 See GATS Annex on Movement of Natural Persons Supplying Services under the Agreement, para. 2 : “The Agreement shall not apply to measures affecting natural persons seeking access to the employment market of a Member, nor shall it apply to measures regarding citizenship, residence or employment on a permanent basis.” 8 Nonnenmacher 2012. 9 See Panizzon 2010. 226 Visa and Asylum current mandate of the WTO (i.e., beyond GATS mode 4).10 One study listed “visa and asylum” as a WTO-extra field, without explicitly defining what it includes. In an examination of 14 EU PTAs and 14 US PTAs, the study found that “visa and asylum” appeared in only 3 EU PTAs and were not at all present in US PTAs, and that “typically the obligations are not enforceable.”11 In addition to the WTO and PTAs, migration is also (and predominantly) addressed in bilateral agreements or Memorandums of Understanding (MoUs) between states.12 The UN is currently working on a Global Compact for Safe, Orderly, and Regular Migration.13 In January 2018, the African Union adopted a Protocol on Free Movement of Persons,14 illustrating how migration is often addressed in regional economic communities (RECs) that seek deeper economic integration between like-minded countries within a given region.15 As a recent UNCTAD report concludes: labor migration […] is increasingly driven by the interaction of migration-related agreements at the following three levels: (i) multilateral opening up of labor markets via [the WTO’s GATS] mode 4 and similar […] regional economic community initiatives [such as the EU or African Union]; (ii) European Union mobility partnerships; and (iii) bilateral migration management agreements […]the latter [bilateral agreements] represent the most comprehensive regulation of migration currently available in treaty law […] non-trade, bilateral migration agreements are the main channels for recruiting low-skilled migrants, whereas trade agreements, including under mode 4, tend to address highly skilled segments of the labor market.16 8.2.2 Selected literature The link between trade and migration has been widely addressed in the literature. A 2015 study, for example, finds that trade and immigration policy are “profoundly interrelated,” but in a rather surprising way: comparing, in particular, the 19th century (generally a period of open immigration but relatively closed trade) to the period after World War II (when most states opened trade but continued to restrict immigration), the study concludes that “increasing trade openness has led to increasingly restrictive immigration.”17 Another study, rather than 10 Horn et al. 2010. Following their definition, facilitation or liberalization of migration would fall under “provisions regarding commitments in policy areas not covered by the current mandate of the WTO,” hence WTO-extra. 11 Ibid. 12 See, for example, Saéz 2013. 13 See https://refugeesmigrants.un.org/migration-compact. 14 For a discussion, see https://issafrica.org/iss-today/a-new-dawn-for-african-migrants. 15 For an overview of migration issues covered in such RECs see IOM 2007. See also Pécoud et al. 2017. 16 UNCTAD 2018, p. 55. 17 Peters 2015. 227 Handbook of Deep Trade Agreements comparing openness of trade versus migration policies, looks at PTAs and the impact they have on migration flows.18 Using a gravity model and OECD International Migration Statistics (IMS), the study finds that a “mutual PTA stimulates international migration flows among member countries by almost 17.5 percent” and that “this effect increases up to 28 percent if the PTA includes visa and asylum provisions” (merely replicating GATS, without including WTO-extra migration provisions, in contrast, and somewhat surprisingly, “deters bilateral migration flows”). In legal scholarship, global migration law has recently attracted considerable attention. A recent symposium addressed the role of international law in governing migrant movements, and questioned whether there is actually a coherent field of global migration law.19 One contributor noted that “migration is governed by an eclectic set of superimposed norms that are scattered throughout a wide array of overlapping fields (human rights law, trade law, humanitarian law, labor law, refugee law, maritime law, etc.).”20 An earlier study had already referred to this situation as “substance without architecture.”21 Others have warned against the urge to ring- fence migration as a separate field: “[A] critical characteristic of migration is that it involves people in all their complexity, and with all their complex needs. Therefore, migration, perhaps more than any other field of international law, is difficult to separate as a body of law.”22 8.3. DEFINITIONS This chapter examines which of the 279 PTAs in the World Bank’s database (listing PTAs in force and notified to the WTO up to November 2015)23 address visa and asylum. For purposes of this study, these are considered to be WTO-extra provisions; i.e., going beyond what is covered under GATS mode 4. That said, visa and asylum address very different things than so-called labor provisions. Labor provisions in PTAs are not concerned with liberalizing the movement of people. Rather, they seek to impose certain minimum standards of labor protection in the domestic laws of the PTA parties. Whereas visa and asylum provisions aim at facilitating (certain types of) movement of people to take advantage of, for example, wage differentials, labor provisions assume that workers stay in their country of origin and aim at protecting them in that country (in no small measure to limit the competitive impact of lower wages or lower labor standards). 18 Orefice 2015. 19 American Journal of International Law 2017-18. https://www.cambridge.org/core/journals/american-journal- of-international-law/ajil-unbound-by-symposium/framing-global-migration-law. 20 Chétail 2017. 21 Aleinikoff 2007. 22 Trachtman 2018, p. 481. 23 The database is available at https://datacatalog.worldbank.org/dataset/content-deep-trade-agreements. 228 Visa and Asylum International migration law has been defined as “the set of international rules governing the movement of persons between states and the legal status of migrants within host states.”24 Typically, three types of movement of persons are distinguished: (a) refugees and asylum seekers; (b) migrant workers; (c) smuggled and trafficked migrants. The International Organization for Migration (IOM) defines a migrant as: any person who is moving or has moved across an international border or within a State away from his/her habitual place of residence, regardless of (1) the person’s legal status; (2) whether the movement is voluntary or involuntary; (3) what the causes for the movement are; or (4) what the length of the stay is.”25 Involuntary migrants have been defined by the United Nations as refugees: i.e., “[s]omeone who has been forced to flee his or her country because of persecution, war, or violence.”26 This study seeks to identify all migration-related provisions in PTAs, including those addressing refugees and asylum seekers and including both regular and irregular flows. 8.4. METHODOLOGY FOR MAPPING The mapping of migration provisions in PTAs proceeded in two steps. First, the list of 279 PTAs in the Deep Integration database was narrowed down to those PTAs that were expected to address migration issues.This first step reduced the total number of PTAs to 100. Second, those remaining 100 PTAs were coded using a template consisting of 30 “yes” or “no” questions. 8.4.1 Pre-selection of PTAs to be coded To develop the list of PTAs to be coded for migration provisions, the following methodology was used: First, several key terms (nine in total) were selected that signal migration elements: “investors,” “key personnel,” “visa,” “migration,” “asylum,” “refugee,” “citizenship,” “employment” and “residence.”These terms were then searched in the PTA database available at http://mappinginvestmenttreaties. com/rta/. This database includes a larger number of PTAs than the World Bank database, and also, unlike the World Bank database, allows for 24 Chetail 2017. 25 See https://www.iom.int/who-is-a-migrant. 26 See https://www.unrefugees.org/refugee-facts/what-is-a-refugee/. Article I.A(2) of the 1951 Refugee Convention provides the complete definition of “refugee” as someone who ”owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence as a result of such events, is unable or, owing to such fear, is unwilling to return to it.” 229 Handbook of Deep Trade Agreements word searches across all PTAs. An important limitation, however, was that these word searches were available only for the text of the PTA itself and excluded protocols, side agreements, and other instruments outside of the main text. The searches were further limited to PTAs in the English language, as the database consulted only includes English-language PTAs. The search yielded 175 PTAs that contained at least one of the nine key terms. The text of each of the 175 PTAs was then searched manually to see if the detected key terms were, indeed, indicative of the inclusion of relevant migration provisions.27 Based on this search, the following types of PTAs were eliminated: (a) those that contained irrelevant key term hits; (b) PTAs no longer in force (such as the US-Canada PTA, superseded by NAFTA); and, to ensure consistency with other chapters of this Handbook, (c) PTAs that did not appear in the World Bank’s Deep Integration database. Since migration provisions appear, in particular, in deep regional economic communities (RECs)28 such as the European Union (EU) or East African Community (EAC), such RECs (including their annexes and protocols) were manually double-checked (as some did not turn up following the standard key term search) to see whether they include migration-related provisions (this also allowed us to include some PTAs that are not originally in English, such as the Commonwealth of Independent States, or CIS). This led us to add six additional PTAs (all RECs, such as MERCOSUR, CARICOM, CIS, and SADC). Finally, many PTAs signed between countries in the Latin America region are only available in Spanish. A manual word search for the nine key terms in Spanish was conducted for those PTAs in the database of the Organization of American States (OAS), available at http://www.sice.oas.org/agreements_e.asp. This led the inclusion of 15 PTAs in Spanish. The end result was the list of 100 agreements coded for migration provisions. 8.4.2 Questions coded for the selected PTAs To produce the dataset on migration provisions in PTAs, a template was developed containing 30 yes/no questions relating to migration under six different headings (Table 8.1). Heading I (questions 1 & 2) assesses whether the PTA sets out general migration goals/objectives. Headings II and III set out the core substantive questions: Heading II (questions 3-11) seeks to ascertain what types of movement of people are covered by the PTA (beyond 27 Examples of hits in the word-search that turned out to be irrelevant for our purposes were words like “envisaged” which incudes the term “visa” but is obviously not relevant. Other times the word “residence” was detected but upon closer inspection it was used in the context of tax-related provisions, or we found the word “employment” in the context of broader objectives of alleviating unemployment (not migration related). 28 To detect such “deep” RECs, that is, PTAs between countries in a given region that integrate more deeply (e.g., by forming a customs union), we used the WTO’s list of so-called “plurilateral regional trade agreements” (available at https://www.wto.org/english/tratop_e/region_e/rta_plurilateral_map_e.htm) but deleted from that list agreements between states not within the same region/continent. 230 Visa and Asylum GATS mode 4 service suppliers), ranging from investors to refugees. Heading III (questions 12-20) examines how or by what means the movement of persons is liberalized or facilitated, be it through mutual recognition schemes (question 18); provisions that facilitate the application, processing, or renewal of visa/migration formalities (questions 12-17) or, importantly, reserving a number of visas (quota) for the PTA partner (question 20). Heading IV (questions 21-25) relates to exceptions or limits to commitments in the PTA. Heading V (questions 26-27) examines whether the PTA explicitly links itself to other migration-related instruments, be they bilateral or multilateral. Heading VI (questions 28-30) examines the institutional set-up and whether the PTA allows for “retaliation” (or suspension of benefits) in case of non-compliance with migration-related provisions. Note that the 30 questions were developed in an iterative way: after constructing a number of questions in the abstract, a selection of relevant PTAs were examined to see what issues they covered, and on that basis the questionnaire was further developed. Next to each of these six vertical headings, six horizontal columns were added, with a box for each PTA to be filled out as follows: (i) Yes or no answer to each of the 30 questions; (ii) If yes, the specific provision in the PTA where the norm29 can be found; (iii) If yes, and where applicable, the specific annex or other agreement where the norm can be found; (iv) Comments (v) Enforceability of the norm: non-binding (0), best endeavor (1), binding with no dispute settlement (DS) (2), binding with state-to-state DS (3), binding with private-state DS (“private” could be a private investor or migrant claimant) (4), binding, both state-to-state and private-state DS (5). (vi) Benefits to non-members: are commitments/concessions “excludible,” that is, reserved for PTA parties (1), or “non-excludible,” that is, can also benefit non-parties (2). 8.4.3 Stylized facts and insights derived from the coding 8.4.3.1 PTAs covering visa and asylum – absolute number and evolution over time All 100 PTAs selected pursuant to the methodology include a positive answer to at least one of the questions in the template (excluding exceptions or limitations).30 That 29 The procedures and standards for coding the agreements were guided by the Migration Codebook, which defines what counts as “norms,” “migration,” and WTO-extra in PTAs with migration provisions.The dataset was developed using both the Migration Codebook and the answers to questions in the template. Ten PTAs were coded by two different coders early on to enhance understanding of the coding exercise. 30 To avoid including PTAs that only provide for exceptions or limitations but not any positive coverage of visa and asylum issues, all 100 PTAs were double-checked to ensure that they included a positive answer to at least one of the questions other than questions 21-25. 231 Handbook of Deep Trade Agreements Table 8.1: Migration questionnaire template I. Migration goals/objectives 1. Does the agreement call for freedom of movement of workers/people 2. Does the agreement call for regulatory cooperation or harmonization in migration regulation? II. Coverage and types of (WTO-extra) movement of natural persons 3. Does the agreement address the movement of investors/key personnel related to investment? 4. Does the agreement address the movement of non-commercial visitors? 5. Does the agreement address the movement of migrant workers already employed by a company in the country of destination? 6. Does the agreement address the movement of migrant workers seeking employment in the country of destination? 7. Does the agreement positively address or facilitate persons obtaining residency in either party? 8. Does the agreement positively address or facilitate persons obtaining nationality/citizenship in either party? 9. Does the agreement address the movement of the dependents of natural persons? 10. Does the agreement address the movement of undocumented migrant workers? 11. Does the agreement address the movement of refugees? III. Facilitation of the (WTO-extra) movement of natural persons 12. Does the agreement encourage parties to expedite the application procedures for immigration formalities for natural persons? 13. Does the agreement limit the time for processing applications requesting temporary entry of natural persons? 14. Does the agreement limit the fees for processing applications for temporary entry of natural persons? 15. Does the agreement encourage parties to publish online if possible or otherwise make publicly available information regarding the current requirements for temporary entry? 16. Does the agreement encourage parties to provide facilities for online lodgment and processing (electronic visa)? 17. Does the agreement provide an entry/visa denial explanation mechanism? 18. Does the agreement provide a mutual recognition scheme (on qualifications, training, work experience)? 19. Does the agreement provide a visa extension or renewal mechanism? 20. Does the agreement provide a quota on number of visas to be issued to natural persons of parties? IV. Exceptions and limitations 1. Does the agreement explicitly exclude questions or measures regarding employment on a permanent basis? 2 22. Does the agreement explicitly exclude questions or measures regarding residency? 23. Does the agreement explicitly exclude questions or measures regarding nationality/citizenship? 24. Does the agreement specifically allow parties to bar entry of natural persons based on public security/ order reasons? 25. Does the agreement allow parties to undertake temporary safeguard measures to bar entry of natural persons? V. Reference to other international instruments 2 6. Does the agreement refer to bilateral agreements related to migration concluded by the parties? 27. Does the agreement refer to multilateral agreements relating to migration, refugees, or trafficking? VI. Institutional arrangements and dispute settlement 8. Does the agreement set up a dedicated organ or sub-committee to oversee migration issues? 2 29. Does the agreement encourage parties to undertake mutually agreed cooperation activities? 30. Does the agreement allow for retaliation to ensure compliance with the dispute settlement system’s outcomes in migration disputes? Note: The 30 questions were developed in an iterative way: after constructing a number of questions in the abstract, a selection of relevant PTAs were examined to see what issues they covered, and on that basis the questionnaire was further developed. 232 Visa and Asylum is, all 100 PTAs, or about 36 percent of PTAs in the World Bank database, include visa and asylum provisions. Of the 30 RECs in the World Bank database, 17 (or 57 percent) include visa and asylum provisions. This confirms that deep economic integration agreements among more like-minded countries within a given region tend to more often include visa and asylum provisions. The inclusion of visa and asylum provisions in PTAs shows an upward trend, particularly since the early 2000s (Figure 8.1), excluding exceptions and limitations (questions 21-25). Figure 8.1: Number of PTAs that include migration provisions over time 8.4.3.2 Which economies have concluded visa and asylum PTAs? The economy with the most visa and asylum PTAs is unquestionably the EU, with 28 PTAs, or 27 percent of the total amount. In second place comes Japan, with 12 PTAs. The United States has four (Table 8.2). The finding about the US contradicts previous studies, which concluded that US PTAs do not include visa and asylum provisions.31 These provisions were found in the annexes or protocols, which the other studies did not examine, even though that is often exactly where visa and asylum provisions are found. 31 See, for example, Horn et al. 2010, Table 5. 233 Handbook of Deep Trade Agreements Table 8.2: Top 10 economies with visa and asylum PTAs Economy No. of PTAs 1 EU 28 2 Japan 12 3 Chile 11 4 Korea, Rep. 10 5 Mexico 9 6 Australia 7 7 Panama 6 8 China 6 9 Colombia 5 10 US 4 Turning to country groupings, the overall World Bank database (279 PTAs) consists of 51 percent North-North PTAs, 42 percent North-South PTAs, and 7 percent South-South PTAs. South is defined as all low-income and lower-middle-income economies, and North as all upper-middle-income and high-income, based on GDP per capita World Bank classifications. To create this classification at the PTA level, if a PTA includes only North (South) economies, it is classified as North-North (South-South); if there is at least one North and one South economy as part of the PTA, it is classified as North-South. Looking at those PTAs that include visa and asylum provisions, 63 percent are North-North, 34 percent are North-South, and only 3 percent are South-South (Figure 8.2). This means that, relatively speaking, North-North PTAs tend to more often include visa and asylum provisions. These provisions are present in 44 percent of all North-North PTAs, 29 percent of all North-South PTAs, and only 15 percent of all South-South PTAs. Figure 8.2: Economies with “visa and asylum” PTAs (World Bank classification) The finding that North-North PTAs top the visa and asylum list may be somewhat surprising since, as discussed above, most of the potential gains from migration are between 234 Visa and Asylum economies at different levels of development (e.g., with large wage differentials).This finding may confirm, however, that migration in PTAs tends to be focused on high-skilled workers, between more developed economies facilitating mutual access to select types of service suppliers or specialized workers, as is the case for the EU. That South-South PTAs are at the bottom of the visa and asylum list is also interesting. It shows that even between poorer economies, it has proven hard to include migration issues. A recent UNCTAD report focusing on migration in Africa finds, however, that “most migration in Africa today is taking place within the continent” and argues that “this intra-African migration is an essential ingredient for deeper regional and continental integration.”32 Our findings suggest, however, that most South-South PTAs (85 percent) do not, at present, cover migration. It is important to note that the regional distribution of PTAs very much depends on the country classification used. If North and South are defined not in terms of the World Bank’s GDP per capita classification, but rather using the WTO’s classifications, based on self- selection by the countries themselves,33 the picture is quite different (Figure 8.3). Based on the WTO classifications - North as developed economies plus the Commonwealth of Independent States (developing Europe); and South as all other WTO members (developing countries excluding Europe) - 47 percent of visa and asylum PTAs are North-South, 38 percent are South-South, and only 15 percent are North-North. Figure 8.3: Economies with “visa and asylum” PTAs (WTO classification) 8.4.3.3 What types of movement of persons are covered? The coverage of different types of movements of people in PTAs was assessed under Heading II, questions 3-11 (Table 8.3). 32 UNCTAD 2018. 33 See the classification in : WTO Secretariat, Participation of Developing Economies in the Global Trading System, Committee on Trade and Development, WT/COMTD/W/230, 7 November 2017, at pp. 60-62. 235 Handbook of Deep Trade Agreements Table 8.3: Types of movement of persons covered in PTAs over time Pre-1995 1996-2000 2001-2010 2011-2015 All periods Total number of PTAs 51 36 128 64 279 in WB database 100% 100% 100% 100% 100% II. Coverage and types of movement 10 4 48 34 96 of natural persons (WTO-extra) 20% 11% 38% 53% 34% 3. Does the agreement address 5 3 45 32 85 the movement of investors/key 10% 8% 35% 50% 30% personnel related to investment? 4. Does the agreement address 7 2 4 2 15 the movement of non-commercial 14% 6% 3% 3% 5% visitors? 5. Does the agreement address 7 1 7 4 19 the movement of migrant workers 14% 3% 5% 6% 7% already employed by a company in the country of destination? 6. Does the agreement address 7 2 4 1 14 the movement of migrant workers 14% 6% 3% 2% 5% seeking employment in the country of destination? 7. Does the agreement positively 6 2 3 1 12 address or facilitate persons obtaining 12% 6% 2% 2% 4% residency in either party? 8. Does the agreement positively 2 0 1 0 3 address or facilitate persons obtaining 4% 0% 1% 0% 1% nationality/citizenship in either party? 9. Does the agreement address 6 0 7 17 30 the movement of the dependents 12% 0% 5% 27% 11% of natural persons? 10. Does the agreement address 3 1 6 2 12 the movement of undocumented 6% 3% 5% 3% 4% migrant workers? 11. Does the agreement address 2 2 2 3 9 the movement of refugees? 4% 6% 2% 5% 3% 236 Visa and Asylum Under movement of persons, three features stand out: • First, the most covered types of movement are (a) movement of investors/key personnel related to investment (30 percent of all PTAs); and (b) movement of dependents (11 percent of all PTAs). In PTAs concluded after 2011, these types of movements were covered in 50 and 27 percent of PTAs, respectively. • Second, PTAs, even recent ones, only rarely address (a) non-commercial visitors (5 percent of total); (b) undocumented migrant workers or refugees (4 and 3 percent, respectively); and (c) facilitation of nationality/citizenship (1 percent of total). • Third, the coverage of migrant workers (already employed or seeking employment in the destination country), at 7 and 5 percent of the total, has actually gone down compared to pre-1995 PTAs, when migrants were covered by 14 percent of PTAs. The same is true for PTAs facilitating residency, covered in 4 percent, compared to 12 percent before 1995. This downward trend might be explained by the higher proportion of RECs in pre-1995 PTAs, which often include these types of provisions. The types of movement of persons covered depends on the type of PTA. Using GDP-per- capita-based World Bank classifications for South (low-income and lower-middle-income economies) and North (all other economies), the figure below refers to the percentage of visa and asylum PTAs (100 in total) that include the relevant provision (Figure 8.4). Figure 8.4: Types of movement covered by type of PTA 8.4.3.4 How is the movement of persons facilitated? The ways that PTAs facilitate the movement of persons, and how they have done so over time, is shown in Table 8.4. 237 Handbook of Deep Trade Agreements Table 8.4: Methods of facilitating movement of persons in PTAs over time Pre-1995 1996-2000 2001-2010 2011-2015 All periods Total number of PTAs in WB database 51 36 128 64 279 100% 100% 100% 100% 100% III. Facilitation of the movement 9 5 51 35 100 of natural persons (WTO-extra) 18% 14% 40% 55% 36% 12. Does the agreement encourage 3 2 36 24 65 parties to expedite the application 6% 6% 28% 38% 23% procedures for immigration formalities for natural persons? 13. Does the agreement limit the time 2 1 3 8 14 for processing applications requesting 4% 3% 2% 13% 5% temporary entry of natural persons? 14. Does the agreement limit the 2 2 30 17 51 fees for processing applications 4% 6% 23% 27% 18% for temporary entry of natural persons? 15. Does the agreement encourage 5 4 27 28 64 parties to publish online if possible 10% 11% 21% 44% 23% or otherwise make publicly available information regarding the current requirements for temporary entry? 16. Does the agreement encourage 0 0 2 3 5 parties to provide facilities for online 0% 0% 2% 5% 2% lodgment and processing (electronic visa)? 17. Does the agreement provide 3 3 8 14 28 an entry / visa denial explanation 6% 8% 6% 22% 10% mechanism? 18. Does the agreement provide 7 3 30 22 62 a mutual recognition scheme 14% 8% 23% 34% 22% (on qualifications, training, work experience)? 19. Does the agreement provide a visa 2 0 19 15 36 extension or renewal mechanism? 4% 0% 15% 23% 13% 20. Does the agreement provide 1 1 4 2 8 a quota on number of visas to be issued to natural persons of parties? 2% 3% 3% 3% 3% 238 Visa and Asylum Here as well, three features stand out: • First, the most commonly included methods of facilitating movement of persons have to do with (a) facilitating immigration procedures (online publication of and expediting of procedures (each 23 percent of total PTAs); (b) limiting fees (18 percent of total); and (c) mutual recognition of qualifications, training, or work experience (22 percent). The inclusion of such provisions has increased significantly over time (for example, online publication of immigration procedures is covered in 44 percent of post-2011 PTAs). • Second, PTAs only rarely cover the more specific commitments of (a) time limitations for processing applications (5 percent of total); (b) electronic visa facilities (2 percent of total); or (c) visa quota allocations to PTA partners (3 percent). Interestingly, of the 8 PTAs that allocate visa quotas to PTA partners, 4 have the United States as a party, confirming the high level of US engagement with this type of migration provision in PTAs. • Third, provisions on (a) visa extension or renewal, as well as (b) explanation of entry/visa denials, are still only covered in a small sub-set of all PTAs (13 and 10 percent, respectively), but there is a clear upward trend over time (covered in 23 and 22 percent of post-2011 PTAs). Figure 8.5 looks at the methods of facilitating movement of persons covered in PTAs depending on the type of PTA, using the GDP-per-capita-based World Bank classifications. Percentages refer to the percentage of visa and asylum PTAs (100 in total) which both (a) include the relevant provision and (b) are a PTA of the particular type. Figure 8.5: Methods of facilitation covered by type of PTA 239 Handbook of Deep Trade Agreements 8.4.3.5 What exceptions to movement of persons are provided for? As more migration-related provisions have been included in PTAs, the percentage of PTAs including exceptions or limitations on migration has also increased. All exceptions are now commonly included, except for temporary safeguard measures related to migration (only 5 percent of total). Since most PTAs do not seriously open migration flows, such safeguards were probably not considered when the PTAs were drafted. Migration safeguards were included most often in pre-1995 PTAs (14 percent), mostly in RECs such as the EU or EU association agreements. Another feature that stands out is that whereas all 100 PTAs selected for this analysis cover visa and asylum issues, 72 explicitly exclude employment on a permanent basis. This confirms that migration in PTAs is largely limited to temporary movement and excludes permanent employment (Table 8.5). Table 8.5: Exceptions or limitations regarding migration in PTAs over time Pre-1995 1996-2000 2001-2010 2011-2015 All periods Total number of PTAs 51 36 128 64 279 in WB database 100% 100% 100% 100% 100% IV. Exceptions and limitations 8 4 40 34 86 16% 11% 31% 53% 31% 21. Does the agreement explicitly 2 2 35 33 72 exclude questions or measures regarding 4% 6% 27% 52% 26% employment on a permanent basis? 22. Does the agreement explicitly 2 1 34 31 68 exclude questions or measures 4% 3% 27% 48% 24% regarding residency? 23. Does the agreement explicitly 2 0 29 30 61 exclude questions or measures regarding 4% 0% 23% 47% 22% nationality / citizenship? 24. Does the agreement specifically 8 3 28 22 61 allow parties to bar entry of natural 16% 8% 22% 34% 22% persons based on public security/order reasons? 25. Does the agreement allow 7 2 2 2 13 parties to undertake temporary 14% 6% 2% 3% 5% safeguard measures to bar entry of natural persons? 8.4.3.6 Do visa and asylum PTAs refer to other migration instruments? Given that migration issues are covered in a variety of instruments - not only PTAs but also bilateral agreements on migration and multilateral treaties on refugees or migrant workers 240 Visa and Asylum - one could imagine that cross-references to such other instruments would be common in PTAs. However, the coding shows the opposite: only in exceptional cases is reference made in PTAs to other international instruments on migration, be they bilateral (3 percent) or multilateral (4 percent). In other words, to the extent that PTAs address migration, they do so in relative isolation, most often without linking back or making reference to other instruments in the dispersed field of global migration law. Table 8.6: Reference to international instruments on migration in PTAs over time Pre-1995 1996-2000 2001-2010 2011-present All periods Total number of PTAs 51 36 128 64 279 in WB database 100% 100% 100% 100% 100% V. Reference to other 4 1 5 6 16 international instruments 8% 3% 4% 9% 6% 26. Does the agreement refer 4 0 2 3 9 to bilateral agreements related 8% 0% 2% 5% 3% to migration concluded by the parties? 27. Does the agreement refer 0 1 5 4 10 to multilateral agreements relating 0% 3% 4% 6% 4% to migration, refugees or trafficking? 8.4.3.7 Enforceability of visa and asylum provisions In one of the horizontal columns of the template, the “bindingness,” or degree of enforceability, of each individual provision was coded under headings I, II, and III, using six different levels (0 to 5). Not surprisingly, provisions setting out broader migration goals/objectives (Heading I, questions 1 and 2) are for the most part “best efforts” provisions (Figure 8.6). More interestingly, however, is that a large majority of provisions under Heading II, types of movement, and Heading III, facilitation of movement, are legally binding, with many subject to state-to-state dispute settlement, and a small number even subject to both private-state and state-to-state DS. Previous studies have claimed or assumed that visa and asylum provisions are “typically … not enforceable.”34 Our data point in a different direction. When it comes to types of movement of persons (Heading II), the large majority of provisions on investors/key personnel related to investment (question 3) and dependents (question 9) are binding with state-to-state dispute settlement.This also represents the largest 34 For example, Horn et al. 2010. 241 Handbook of Deep Trade Agreements group for provisions on non-commercial visitors and migrant workers (already employed or seeking employment in the destination country; questions 4, 5, and 6). In contrast, only a small minority of provisions on undocumented migrants and refugees are subject to dispute settlement (questions 10 and 11). Turning to methods of facilitating movement of persons (Heading III), most provisions are legally binding but not subject to dispute settlement. The one provision that stands out here is visa quotas (question 20), included in only 8 PTAs, of which half (4) are subject to state-to-state dispute settlement. Provisions that limit processing fees (question 14) are subject to dispute settlement more often than other methods of facilitation (15 out of 52). An outlier in the other direction are provisions on mutual recognition, a majority of which are “best efforts” only. Figure 8.6: Enforceability of “visa and migration” provisions Where state-to-state dispute settlement is provided for the enforcement of visa and asylum provisions, two elements are worth mentioning: pattern of practice and exhaustion of administrative remedies. These preconditions for filing a dispute settlement case were first 242 Visa and Asylum introduced in NAFTA (Article 1606) and are now present in 47 of the 100 PTAs with visa and asylum provisions. In particular: • a pattern of practice must be demonstrated; a single refusal to grant temporary entry to a business person or investor is not sufficient to file a case; • available administrative remedies, including domestic remedies in the country in question, must have been exhausted before the person(s) involved can file a state-to-state dispute case. Unlike visa and asylum cases, most other PTA disputes do not require a showing of pattern of practice or exhaustion of domestic remedies.35 Not only are many visa and asylum provisions legally binding and subject to state-to- state dispute settlement, but where state-to-state DS is provided, it is often followed up by retaliation or trade sanctions if there is a finding of violation and the losing country fails to implement the ruling. Question 30 in the questionnaire asks exactly this question. In 60 out of the 100 PTAs, the answer is positive - retaliation or suspension of concessions is provided for the enforcement of at least one visa and asylum provision in the PTA. This hardens the enforceability of visa and asylum provisions. 8.4.3.8 Do “visa and asylum” provisions benefit non-parties? PTAs are not as preferential as they used to be.36 Deep PTAs often include regulatory or other provisions that also benefit non-PTA parties. Tariff concessions, the core of first- generation PTAs, are “excludible,” using rules of origin to reserve preferential treatment to PTA parties only. In contrast, many second- and third-generation PTAs include regulatory, intellectual property, investment, labor, environment, or competition standards that apply to traders or businesses regardless of whether they are a party to the PTA. This means that PTA provisions apply on an MFN basis to everyone, and that non-PTA parties can actually benefit from PTAs concluded by other economies without having to make concessions themselves. Visa and asylum provisions are an exception to this trend. The last horizontal column in the coding sheet asks, for each of the visa and asylum provisions, does the provision benefit third countries (i.e., is it “non-excludible),” or does it benefit only PTA parties (is it “excludible”)? The results for the substantive provisions under Heading II (types of movement) and Heading 35 In some PTAs, disputes under labor or environmental provisions may require something similar to a pattern or practice. See, for example, CAFTA-DR, Article 16.2(a), which requires “a sustained or recurring course of action or inaction.” 36 Pauwelyn 2017. 243 Handbook of Deep Trade Agreements III (methods of facilitation) are shown below (Figure 8.7). Not surprisingly, most visa and asylum provisions are, like tariff concessions, excludible; that is, they are reserved to the PTA parties only. In practice, this means that only persons from the PTA party are granted access or can benefit from mutual recognition of credentials, a country-specific visa quota, or faster or cheaper immigration-processing procedures. One provision that stands out is the publication of immigration information (Heading III, question 15). Here, many publication commitments benefit everyone, including non-PTA parties (29 are “non-excludible”). In other cases, however, transparency or disclosure of information remains limited to cases where the request is made by persons from a PTA party (32 are “excludible” in this sense). In sum, whereas many provisions in deep PTAs may work on an MFN basis and also benefit non-PTA parties, visa and asylum provisions continue to operate mostly on a discriminatory basis and benefit PTA parties only. Figure 8.7: Do “visa and asylum” provisions benefit non-parties? 8.5. MAIN FINDINGS AND CONCLUSIONS Numbers. Of the 279 PTAs in the World Bank database, visa and asylum provisions are present in 100, or 36 percent. Regional economic communities, in particular, tend to include visa and asylum provisions; they are present in 57 percent of the RECs in the World Bank database. There is a clear upward trend of PTAs including such provisions, especially since the 2000s. 244 Visa and Asylum Geographical distribution. The EU has, by far, the most PTAs with visa and asylum provisions (28 PTAs). Japan is second (12 PTAs), and the United States is also in the top 10 (4 PTAs). In terms of regional distribution, and using World Bank GDP per capita country classifications, most PTAs with visa and asylum provisions are North-North (63 percent, or 44 percent of all North-North PTAs in the database); only a very small amount are South-South (3 percent, or 15 percent of all South-South PTAs in the database), indicating that to the extent that PTAs address migration, it tends to be between more developed countries and focused on high- skilled individuals. South-South PTAs focus more generally on WTO-extra elements (e.g., adding commitments under GATS, mode 4), less on WTO-extra provisions (e.g., by adding provisions on migration not currently covered by the WTO’s mandate). Types of movement of persons. The focus of PTAs on specialized or high-skilled individuals is confirmed when looking at the types of movement of persons addressed: many PTAs address the movement, for example, of investors or dependents; very few cover migrant workers looking for employment, undocumented migrants or refugees. Importantly, of the 100 PTAs covering visa and asylum, 72 explicitly exclude employment on a permanent basis. In other words, migration in PTAs focuses on high-skilled, specialized individuals and temporary (not permanent) movement of persons. How movement is liberalized. In terms of methods used in PTAs to facilitate the movement of persons, the focus is on immigration procedures (publication of requirements, expediting procedures, limiting fees) and mutual recognition of qualifications, training, or work experience. Only rarely do PTAs actually open borders; e.g., by requiring free movement of workers in RECs or reserving a visa quota for PTA parties. Interestingly, where a PTA does include such visa quotas, this often (50 percent) involves the United States as a party. Enforceability and third parties. Visa and asylum provisions in PTAs are often legally binding, contrary to what earlier studies have found. In addition, they are regularly made subject to state-to-state dispute settlement and where this is done, violations are backed up by retaliation or trade sanctions (in 60 of the 100 PTAs coded). However, in many cases a dispute can only be filed if a “pattern of practice” can be demonstrated (an individual violation does not suffice) and once domestic remedies have been exhausted (47 of the 100 PTAs coded include these two pre-conditions). Unlike many regulatory-type provisions in deep PTAs, most visa and asylum provisions operate on a discriminatory or “excludible” basis, in that they are reserved to PTA parties and do not automatically benefit third parties. Lack of systemic integration. Migration is addressed with a variety of instruments other than PTAs, especially bilateral migration agreements and multilateral treaties on refugees or migrants. Only very rarely do PTAs refer to such other instruments. In conclusion, PTAs do increasingly cover visa and asylum provisions and migration-related issues. However, other than in some RECs, they do so in a “thin” way, focusing on procedural 245 Handbook of Deep Trade Agreements issues rather than genuinely opening migration flows. This is especially true in North- North PTAs (only rarely in South-South PTAs) and with a heavy focus on the temporary movement of high-skilled or specialized individuals, most often excluding the movement of people looking for work or permanent employment. In this sense, although cross-issue bargains could be struck within PTAs so as to liberalize not only goods and services but also the movement of people, the contribution of PTAs to the broader system of liberalizing or managing migration flows remains limited.   ACKNOWLEDGMENTS This chapter is based on work done in the TradeLab International Economic Law Practicum held at Georgetown University Law Center, in Spring 2018. We would like to thank Wenni Zhang for her work as part of the Practicum team and Tuan Doan for assistance with coding Spanish-language treaties, as well as the World Bank team, especially Nadia Rocha and Michele Ruta, for their input, feedback, and continuous support throughout this project. 246 Visa and Asylum REFERENCES Aleinikoff, T. Alexander. 2007. International legal norms on investment law, edited by Thilo Rensmann. Berlin: Springer. migration: Substance without architecture. In International migration law: Developing paradigms and key challenges, edited Pécoud, A., J. Gartland, K. Neyts, S. Nita and P. De by R. Cholewinski, R. Perruchoud, and E. MacDonald. The Lombaerde (eds.). 2017. Migration, free movement Hague: T.M.C. Asser Press. and regional integration. http://unesdoc.unesco.org/ images/0026/002606/260669E.pdf. American Journal of International Law. 2017-18. Unbound Symposium on Framing Global Migration Law. https:// Peters, Margaret E. 2015. Open trade, closed borders: Immigration www.cambridge.org/core/journals/americanjournal-of- in the era of globalization. World Politics 67: 114-54. international-law/ajil-unbound-by-symposium/framing- global-migration-law. Rodrik, Dani. 2001. Comments at a conference on immigration and the welfare state, June 23. http:// Bradford, Anu. 2013. Sharing the risks and rewards of economic ksghome.harvard.edu/~drodr ik/Comments%20on%20 migration. University of Chicago Law Review 80(1), article 3. Immigration%20Conference.doc. Chétail, V. 2017. The architecture of international migration Sáez, Sebastián (ed.). 2013. Let workers move: Using bilateral law:A deconstructivist design of complexity and contradiction. labor agreements to increase trade in services. Washington, DC: Symposium on Framing Global Migration Law. American World Bank. Journal of International Law Unbound 111: 18-23. Trachtman, Joel P. 2009. The international law of economic migration: ———. 2017. Conceptualizing international migration law. Toward the fourth freedom. Kalamazoo, MI: Upjohn Press. https:// Proceedings of 110th annual meeting of American Society of International research.upjohn.org/cgi/viewcontent.cgi?article=1030& Law, 201-4. https://www.cambridge.org/core/journals/proceedings context=up_press. -of-the-asil-annual-meeting/article/remarks-by-vincent-chetail /9FB716E1DEC2321D7A777D05DB208693#. ———. 2014. The role of international law in economic migration. NYU working paper. http://www.law.nyu. Horn, H., P. C. Mavroidis, and A. Sapir. 2010. Beyond edu/sites/default/files/upload_documents/economic%20 the WTO? An anatomy of EU and US preferential trade migration%20trachtman%20nyu.pdf. agreements. World Economy 33 (11). ———. 2018. Is migration a coherent field of international IOM (International Organization for Migration). 2007. law? The example of labor migration. Symposium on Framing International dialogue on migration: Intersessional workshop Global Migration Law, Part III. American Journal of International on free movement of persons in regional integration processes Law Unbound 111: 481-86. (June 18-19). Supplemental materials available at http://www. iom.int/jahia/webdav/site/myjahiasite/shared/shared/ Refugees and Migrants. 2017. Global compact for migration. mainsite/microsites/IDM/ workshops/free_movement_of_ https://refugeesmigrants.un.org/migration-compact. persons_18190607/idm2007_handouts.pdf. Tsion, Abebe T. 2018. A new dawn for African migrants. Nonnenmacher, Sophie. 2012. International trade law and Institute for Security Studies, January 31. https://issafrica.org/ labour mobility. In Foundations of international migration law, iss-today/a-new-dawn-for-african-migrants. edited by Brian Opeskin, Richard Perruchoud, and Jillyanne Redpath-Cross. Cambridge: Cambridge University Press. UNCTAD (United Nations Conference on Trade and Development). 2018. Economic development in Africa report Orefice, G. 2015. International migration and trade 2018: Migration for structural transformation. http://unctad.org/ agreements: The new role of PTAs. Canadian Journal of en/pages/PublicationWebflyer.aspx?publicationid=2118. Economics 48 (1): 310-34. WTO Secretariat. 2017. Participation of developing Panizzon, M. 2010. International law of economic migration: economies in the global trading system. Committee on Trade A ménage à trois? GATS mode 4, EPAs and bilateral migration and Development, November 7, WT/COMTD/W/230. agreements. Journal of World Trade 44 (6). Pauwelyn, Joost. 2017. Not as preferential as you may think: How mega-regionals can benefit third countries. In Mega- regional trade agreements and the future of international trade and 247 Handbook of Deep Trade Agreements 248 Rules of Origin CHAPTER 9 Rules of Origin M. Angeli, J. Gourdon, I. Gutierrez, and P. Kowalski 249 Handbook of Deep Trade Agreements CHAPTER 9 Rules of Origin M. Angeli*, J. Gourdon†, I. Gutierrezø, and P. Kowalski^ * World Trade Institute, Geneva, Switzerland † Organisation for Economic Co-operation and Development, Paris, France ø Graduate Institute for International and Development Studies, Geneva, Switzerland ^Center for Social and Economic Research, Warsaw, Poland CONTENTS 9.1. INTRODUCTION: RULES OF ORIGIN IN THE CONTEXT OF GLOBAL VALUE CHAINS 251 9.2. LITERATURE REVIEW ON EXISTING CLASSIFICATIONS OF ROOS 252 9.3. INTRODUCING THE OECD DESTA CLASSIFICATION OF ROOS 254 9.3.1 Classification of product-specific rules of origin (PSR) 254 9.3.1.1 Wholly obtained 255 9.3.1.2 Substantial transformation 255 9.3.1.3 Refinements 258 9.3.2 Classification of regime-wide RoOs 259 9.3.2.1 Certification 259 9.3.2.2 Verification 261 9.3.2.3 Cumulation 262 9.3.2.4 Methods of estimating value content 265 9.3.2.5 Other aspects 268 9.4. STYLIZED FACTS 271 9.4.1 Product-specific rules of origin 272 9.4.1.1 Historical review of PTAs through their PSR provisions 272 9.4.1.2 Differences across regions 275 9.4.1.3 Heterogeneity 277 9.4.1.4 Distribution of PSRs across sectors 279 9.4.2 Regime-wide RoOs matter 280 9.4.2.1 The certification process 281 9.4.2.2 Important mechanisms:The cumulation and de minimis rules 283 9.4.2.3 The burden of calculating VC 285 9.4.2.4 Additional flexibility mechanisms 287 9.5. CONCLUSIONS 288 ACKNOWLEDGMENTS 289 REFERENCES 290 ANNEX 291 250 Rules of Origin 9.1. INTRODUCTION: RULES OF ORIGIN IN THE CONTEXT OF GLOBAL VALUE CHAINS Rules of origin (RoOs) are critical components of preferential trade agreements (PTAs), but their implications for the development of global value chains (GVCs) are not well understood. RoOs establish the conditions that products must meet to be eligible for preferential market access. In cases where products are produced entirely within the boundaries of one country, these rules prevent trade deflection by ensuring that products from outside the PTA do not enter the duty-free zone through the country with the lowest external tariff. Although RoOs are indispensable in PTAs, if they are too complex or at odds with methods of production used in modern GVCs, they may actually undo the benefits of these trade agreements. In particular, if they are too restrictive, they may prevent producers from taking advantage of the preference and from developing regional value chains within the PTA. Some market participants may simply be unable to meet RoO requirements because of technological or managerial constraints and may therefore be at a competitive disadvantage and ultimately even exit the market. Others may decide not to utilize preferences because the preference margin between the multilateral and the PTA tariff is smaller than the anticipated cost of RoO compliance. The underutilization of PTA preferences has been empirically analyzed and found to be prevalent for sensitive sectors such as agriculture and textiles.1 RoOs generate two types of costs: (1) those related to reducing the sourcing of inputs from trading partners not covered by the PTA, even if those suppliers are less expensive or provide higher-quality goods (i.e., production costs); and (2) those related to proving the origin of products (administrative costs). Restrictive RoOs can increase both types of costs to the point where the cost of compliance exceeds the benefit of the preferences conferred by the agreement, leading producers to choose not to use the preferences. For example, estimates for Latin America show that RoOs undo a relatively significant portion of the positive trade effect of agreements, especially for trade in intermediate products.2 Overall, there are three main concerns related to RoOs: (1) Their restrictiveness: “prohibitive RoOs” undo trade liberalization benefits that would normally stem from tariff concessions, particularly in the context of GVCs. (2) Their heterogeneity: if a country has several different RoO regimes in its agreements with different countries, producers may either have to split production to use different input mixes to export to different partners or concentrate on the market with less restrictive RoOs. 1 Manchin 2006; Bureau et al. 2006. 2 Cadestin et al. 2016. On average, RoOs are estimated to have tariff equivalents of around 11 and 9 percent, respectively, for intra- and extra-trade agreement imports of intermediate products. 251 Handbook of Deep Trade Agreements In any case, they have to navigate a more complex RoO system, which is costly. (3) The burden on sourcing: RoOs can constrain sourcing decisions and prevent final producers from choosing the most efficient input suppliers from around the world.3 The constraints presented by RoOs can generally be reduced directly by simplifying existing product-specific RoOs at the tariff line level and bringing them more in line with actual business practices. They can also be reduced indirectly, through the use of various auxiliary mechanisms, also called regime-wide RoOs, such as the adoption of de minimis, cumulation, and business-friendly RoO certification schemes. The utility of such indirect schemes has been confirmed by existing empirical literature; cumulation schemes are particularly effective.4 There are signs that countries around the world are becoming more aware of the importance of flexible sourcing schemes; RoOs often get renegotiated and simplified, and new auxiliary mechanisms are being included. 9.2. LITERATURE REVIEW ON EXISTING CLASSIFICATIONS OF ROOS Given the empirical evidence for the trade effect of RoOs, a number of RoO restrictiveness indices have been established in academia to generate numerical data on RoO restrictiveness.They conduct a textual assessment of ex-ante RoO restrictiveness based on the sourcing decisions they permit or prevent.While an assessment of the export costs of specific RoOs would be interesting, the need for extensive polling of exporting companies and their knowledge of the origin provisions, as well as the lack of PTA-specific customs data, would make such an assessment prohibitive. The most prominent methodology for classifying product-specific rules (PSR) across products and across agreements was developed by Estevadeordal (2000) and establishes an ordinal index of restrictiveness of RoOs.5 Methodologically, the index codifies individual PSR provisions and adds or subtracts a value based on the provision’s restrictiveness. The PSR provisions are codified into 20 variables along the common structure of PSR: (1) changes in tariff classification, and exceptions on permitted inputs; (2) value content requirements; (3) technical requirements, (4) products that must be made entirely within the parties to be deemed originating (wholly obtained) Estevadeordal’s index was further expanded by nine additional categories of PSR, covering particular types of exceptions to change in tariff classification (CTC) rules in chapters, headings, and subheadings of PTAs.6 3 Conconi et al. 2018. 4 Augier, Gasiorek, and Tong 2005; Estevadeordal and Suominen 2008; Park and Park 2009. 5 Estevadeordal 2000 and its extensions: Estevadeordal and Suominen 2006 and 2008. 6 Estevadeordal, Harris, and Suominen 2009. 252 Rules of Origin Most recently the Regime-Weighted Harris Index (RWHI) has expanded the measurement of PSR by weighting PSR restrictiveness by the applicable regime-wide RoO provisions.7 This approach is based on the observation that the ex-ante restrictiveness of product- specific RoOs are best understood in conjunction with the cumulation, de minimis, and certification requirements contained in regime-wide RoOs. By first subtracting an estimate for the cumulation zone from one and multiplying it by the RWHI, the restrictiveness index gets smaller with an increase in the cumulation zone. The index is then weighted by the de minimis provision; i.e., the maximum level of non-originating materials permitted before the origin status of the final good is affected. The RWHI becomes smaller the more non-originating materials are de minimis permitted. Third, an ordinal value is added for the type of origin certification prescribed, assigning 0 index points for the least restrictive self-certification and 8 for the most restrictive, comprising a two-step public-and-private origin certification. The World Customs Organization (WCO)8 advanced this classification further by incorporating 20 different regime-wide provisions mapped across 55 existing PTAs. In addition, the Australian Productivity Commission9 developed a classification of 10 different types of regime-wide RoOs in order to construct an index of restrictiveness for such provisions in 10 existing PTAs. The MAST Group10 also developed a taxonomy of RoOs. Unlike existing indices that code RoO provisions as mostly dichotomous (1 or 0), the MAST taxonomy assigns codes of 1 to 5 digits across 40 categories for preferential and non-preferential RoOs and has not been applied yet to data. These examples go to show that, first, quantitative data on RoO design and impacts are scarce. Second, where they exist and are publicly available, datasets are limited to a number of provisions or covered PTAs. Third, the trade effects of RoOs have largely been analyzed through ex-ante restrictiveness measures based on text analysis. The analysis presented here aims to remedy that. By covering all relevant PSR provisions, both general and PSR, the dataset will inform researchers and academics alike on the design variation of RoOs. In addition, the proposed heterogeneity index adds a new angle to our understanding of RoO trade effects. 7 Kelleher 2012. 8 World Customs Organization 2014. 9 Australian Productivity Commission 2014. 10 The MAST Group includes the OECD, World Trade Organization, International Trade Centre, and World Bank and is spearheaded by UNCTAD. Its RoO project is ongoing. 253 Handbook of Deep Trade Agreements 9.3. INTRODUCING THE OECD DESTA CLASSIFICATION OF ROOS Each PTA contains a protocol or chapter in the main agreement detailing and identifying the criteria that test whether a product has PTA nationality and is thus eligible for preferential tariff treatment. Most PTAs today also contain annexes for product-specific RoOs.These product- specific origin criteria are typically identified at the HS4-digit level, or even at the HS6-digit level,11 turning these protocols into very technical and often complex instruments. Given that the WTO Agreements do not prescribe a specific design for RoO and origin procedures, PTA parties are, in practice, free to negotiate RoOs without any legal constraints, which has resulted in a “spaghetti bowl” of overlapping and intertwining RoOs and procedures. As a consequence, different RoO regimes often coexist in a single country, so producers need to comply with different value chain restrictions and certification requirements. In other words, the complexity of RoOs alone can entail significant administrative and operational costs for producers and for national Customs authorities. Building on the existing literature, this chapter develops a new classification of RoOs. The methodology follows a questionnaire approach that uses questions with either binary or multiple detailed responses encompassing the two broad types of RoOs: product-specific rules and regime wide. PSRs are classified into 40 different categories on the basis of answers to 14 broad questions and their combinations. Regime-wide RoOs are classified into 30 detailed categories according to their characteristics within 8 broad types of regime-wide provisions. 9.3.1 Classification of product-specific rules of origin (PSR) Unless otherwise specified, 1 denotes a “yes” answer to the given question, while 0 denotes a “no” answer. The square-bracketed terms below refer to the corresponding variables in the collected dataset, as defined in Annex Table 9.A.1, List of Codes and Definitions. The first question refers to the presence or absence of a PSR: Does the agreement contain product-specific rules of origin? [SR_psr]. The WTO Rules of Origin Agreement and the WCO Kyoto Convention12 recognize two basic criteria for determining origin: wholly obtained and substantial transformation. These are discussed in turn below. 11 The Harmonized System (HS) is an international standardized nomenclature for the classification of traded products developed by the World Customs Organization. The first two digits (HS2) refer to the chapter where the goods are classified. The next two digits (HS4) refer to groupings within that chapter. HS6 refers to the specific products. 12 The International Convention on the Simplification and Harmonization of Customs Procedures, 1974. 254 Rules of Origin 9.3.1.1 Wholly obtained The wholly obtained (WO) criterion specifies that the country of origin of a product or commodity is the country where it has been wholly produced (or grown, harvested or extracted for non-manufactured products). In this case, the origin requirement is met if a product or commodity does not use any foreign components or materials. Question: Is the product’s origin defined as wholly obtained? [SR_who] EFTA-Central America (2014), Annex I on Rules or Origin and Methods of Administrative Cooperation, Article 2 of Annex I: Origin Criteria: For the purposes of this Agreement, a product shall be considered as originating in a Party if: (1) it has been wholly obtained in a Party, in accordance with Article 3 (Wholly Obtained Products); (2) the non-originating materials used in the working or processing of that product have undergone sufficient working or processing in a Party, in accordance with Article 4; or (3) it has been produced in a Party exclusively from materials originating in one or more Parties. 9.3.1.2 Substantial transformation The substantial transformation criterion specifies that the country of origin is the country where the last substantial transformation took place, and this transformation must be sufficient to give the commodity its essential character. Question: Is the product’s origin defined through substantial transformation criteria? [SR_stc] Russian Federation-Serbia (2006), Article 4(1): Criterion of Sufficient Processing (treatment): Product is considered to be subjected to sufficient processing or treatment in one of the State Parties, if such a product is processed or treated and the value of materials used in this process (raw materials, semi-finished and finished goods) originates from other countries (other than State Parties), or the value of materials of unknown origin does not exceed 50 percent of the value of the exported goods. There are three distinct sets of criteria to express “substantial transformation”: Change of tariff classification (which can be at the chapter, heading, or subheading level); value content requirement (different methodologies depending on whether the focus is on originating or the non-originating materials); or a technical requirement (such as transformation by chemical reaction). According to the value content (VC) criterion, the exported good must reach a threshold percentage value of locally or regionally produced inputs. 255 Handbook of Deep Trade Agreements Question: Is the product’s origin defined through a value content requirement? [SR_vcr] South Asian Free Trade Agreement (2012), Product Specific Rules Under SAFTA Rules of Origin, Explanatory Notes (4): The DVA in percentage shall mean the minimum value addition in the Exporting Contracting State, calculated as per the following formula: DVA = FOB value of the export product - value of non-originating materials × 100 FOB value of the export product. For the VC entry, two additional subentries are distinguished with respect to the methods of calculation and the reference values: Question: Are there different ways of calculating value content requirement? [SR_vcr_cal] Questions on Method 1: What is Method 1 for calculating VC? [SR_vrc_meth1] What is the percentage of value content required under Method 1? [SR_vrc_perc1] Questions on Method 2: What is Method 2 for calculating VC? [SR_vrc_meth2] What is the percentage of value content required under method 2? [SR_vrc_perc2] On VC methodologies. Republic of Korea-Australia FTA (2014), Chapter 3 (Rules of Origin and Origin Procedures), Article 3.3(1) (Regional Value Content): Where Annex 3-A specifies a regional value content requirement, the regional value content shall be calculated in accordance with one of the following methods: (1) build-down method R VC = AV– VNM x 100 AV (2) build-up method R VC = VOM x 100 AV where: R VC is the regional value content, expressed as a percentage; AV is the adjusted value of the good and shall be the FOB value of the good determined in accordance with the Customs Valuation Agreement, inclusive of the cost of transport and insurance to the port or site of final shipment abroad; VNM is the value of non-originating materials (including materials of undetermined origin) used in the production of the good; and VOM is the value of originating materials used in the production of the good. 256 Rules of Origin On VC thresholds. Korea-Australia FTA (2014), Annex 3-A (Schedule of Product Specific Rules): 87.07.10: No change in tariff classification required, provided there is a regional value content of not less than: (1) 35 percent under the build-up method; or (2) 40 percent under the build-down method. Alternative to the VC criterion, a product can be considered to have undergone substantial transformation by undergoing a CTC. To change a product into a different product category, the exported good must have a different tariff classification from any imported inputs. Question: Is the product’s origin defined through a change in tariff classification? [SR_ctc] This category can be further broken down by the level of aggregation at which the change in tariff classification must occur: Question: Is the product’s origin defined through a change in chapter? [SR_cc] CPTPP (2018), Annex 3-D (Product-Specific Rules of Origin), Chapter 27 Note 1(1) Notwithstanding the applicable product-specific rule of origin, a good of chapter 27 that is the product of a chemical reactionis an originating good if the chemical reaction occurred in the territory of one or more of the Parties. Note: The Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) is a trade agreement among Australia, Brunei Darussalam, Canada, Chile, Japan, Malaysia, Mexico, New Zealand, Peru, Singapore and Vietnam. CPTPP (2018), Annex 3-D (Product-Specific Rules of Origin) 05.01 - 05.11: A change to a good of heading 05.01 through 05.11 from any other chapter. Question: Is the product’s origin defined through a change in heading? [SR_ch] CPTPP (2018), Annex 3-D (Product-Specific Rules of Origin) 1208.90: A change to any other good of subheading 1208.90 from any other heading. Question: Is the product’s origin defined through a change in subheading? [SR_cs] CPTPP (2018), Annex 3-D (Product-Specific Rules of Origin) 0801.32: A change to a good of subheading 0801.32 from any other subheading. 257 Handbook of Deep Trade Agreements Under the technical requirement (TR) criterion, the exported good must have undergone specified manufacturing or processing operations which are deemed to confer origin of the country in which they were carried out. Question: Is the product’s origin defined through a technical requirement? [SR_tr] 9.3.1.3 Refinements Combinations and alternatives The previous three criteria (change of tariff classification, value content requirement, technical requirement) are used in existing trade agreements by themselves or in combination with other criteria, or as alternative criteria. Question: Do two or more origin criteria apply cumulatively? [SR_cum] CPTPP(2018), Annex 3-D (Product-Specific Rules of Origin) 1901.20: A change to a good of subheading 1901.20 containing more than 30 percent by dry weight of rice flour from any other chapter, provided that the value of non-originating rice flour of subheading 1102.90 does not exceed 30 percent of the value of the good. Question: Do two or more origin criteria apply alternatively? [SR_alt] CPTPP (2018), Annex 3-D (Product-Specific Rules of Origin) 1515.19: A change to a good of subheading 1515.19 from any other chapter; or No change in tariff classification required for a good of subheading 1515.19, provided there is a regional value content of not less than 40 percent under the build-down method. Exceptions Exceptions can be attached to a particular CTC requirement, generally prohibiting the use of non-originating materials from a particular HS subheading, heading, or chapter for goods supposed to qualify via CTC, and thereby making the requirement more restrictive. Depending on the type and number of product lines being excluded from a CTC, an exception can lead to a de facto prohibition of imports if exporters wish to qualify for preferential tariff treatment.13 Question: Are one or more HS codes or product groups explicitly excluded from being used as inputs for originating goods? [SR_ctc_exc] CPTPP (2018), Annex 3-D (Product-Specific Rules of Origin) 1102.90:A change to a good of subheading 1102.90 from any other chapter, except from heading 10.06. 13 Conconi et al. 2018. 258 Rules of Origin 9.3.2 Classification of regime-wide RoOs Although products in different PTAs may have the same product-specific rules, the effective restrictiveness of these rules may differ depending on the type of regime-wide provisions. 9.3.2.1 Certification The certification and customs procedures specified in the trade agreements in relation to PSRs determine how compliance with a specific rule is demonstrated and verified, in terms of the required documents and customs procedures. The various preference schemes define different procedures for certification of origin, and establish a system of checks on the authenticity of claims for preferential treatment. The type of certificate issuing body Certification requirements specify who holds the burden of providing and securing information about the origin of the good. There are various bodies, such as authorities, exporters, producers, importers, and designated private bodies, that can typically issue the certificate of origin. All the RoO regimes rely on one or more types of issuing bodies. Question: Can the certificate be issued on the basis of self-certification by the exporter/producer/ importer without need for authentication by the competent authority? [SR_cer_sel] CPTPP (2018), Chapter 3 (Rules of Origin and Origin Procedures), Article 3.21 (Basis of a Certification of Origin) 1. Each Party shall provide that if a producer certifies the origin of a good, the certification of origin is completed on the basis of the producer having information that the good is originating. 2. Each Party shall provide that if the exporter is not the producer of the good, a certification of origin may be completed by the exporter of the good on the basis of: (a) the exporter having information that the good is originating; or (b) reasonable reliance on the producer’s information that the good is originating. 3. Each Party shall provide that a certification of origin may be completed by the importer of the good on the basis of: (a) the importer having documentation that the good is originating; or (b) reasonable reliance on supporting documentation provided by the exporter or producer that the good is originating. Question: Does the certificate have to be issued by competent authorities of the exporting party, including customs administrations, other government authorities or designated private ones? [SR_cer_adm] Japan-Philippines (2008), Chapter 3 (Rules of Origin), Article 41(1) (Certificate of Origin) The certificate of origin referred to in paragraph 1 of Article 40 shall be issued by the competent governmental authority of the exporting Party on request having been made in writing by the exporter or its authorized agent. 259 Handbook of Deep Trade Agreements Question: Is there a possibility to combine self-certification with administrative certification? [SR_cer_two] EFTA-Central America (2014), Annex I (Rules or Origin and Methods of Administrative Cooperation), Article 20(1) (Approved Exporter) The competent authority of the exporting Party may authorize any exporter, hereafter referred to as “approved exporter,” who makes frequent shipments of originating products under this Agreement, to make out origin declarations irrespective of the value of the products concerned. The validity period of the certificate of origin This indicator denotes the time limit allowed for the importer/exporter to conclude the importation of goods under the certificate of origin from the date of issuance.This certificate must be submitted to the Customs authorities of the importing country at the time of the importation and within the validity period. Question: What is the length of the validity period for the certificate of origin? [SR_cer_val] Canada-Honduras (2014), Chapter 5 (Customs Procedures), Article 5.2(6) (Certificate of Origin) A Party shall ensure that the Certificate of Origin is accepted by its Customs administration for at least 1 year after the date on which the Certificate of Origin was signed. The record keeping period This entry captures the time period during which exporters, producers, or importers should maintain documents or background information relating to the origin of the goods. This period ensures that Customs authorities can check and control documents during a certain period of time. Question: What is the length of the record keeping period? [SR_cer_rec] Canada-Honduras (2014), Chapter 5 (Customs Procedures), Article 5.6(1) (Records) Each Party shall provide that an exporter or a producer in its territory that completes and signs a Certificate of Origin must maintain in its territory for 5 years after the date on which the Certificate of Origin was signed, or for a longer period specified by the Parties, records relating to the origin of a good for which preferential tariff treatment was claimed in the territory of the other Party. Exemption Most PTAs include a provision regarding an exemption from certifying RoOs, but there is great diversity as to the maximum monetary amount up to which the certificate is not 260 Rules of Origin required. Moreover, some PTAs provide that the importing country can waive the requirement for proof of origin in accordance with its laws and regulations without stipulating a specific amount. Question: Is there a certificate exemption? [SR_cer_exe1] Question: What is the threshold for exemption in $US ? [SR_cer_exe2] Korea-Colombia (2016), Chapter 3 (Rules of Origin and Origin Procedures), Article 3.21 (Waiver of Certificate of Origin) Notwithstanding Article 3.19, a certificate of origin shall not be required where: (1) the customs value of the importation does not exceed US$1,000 or its equivalent amount in the currency of the importing Party, or such higher amount as may be established by the importing Party; or (2) the importing Party has waived the requirement for a certificate of origin in accordance with its laws and regulations. Minor amendments Whenever the certificate contains errors, some PTAs allow some minor amendments to be made ex post, while others require the issuance of a new certificate and invalidation of the former one. Question: Is there a possibility to amend minor errors? [SR_cer_err] Chile-Vietnam (2014), Annex 4-A (Operational Certification Procedures), Rule 5 (Treatment of Erroneous Declaration n the Certificate of Origin) The Customs Authority of the importing Party will disregard minor errors, such as slight discrepancies or omissions, typographical errors, and information which falls outside the designated box, provided that these minor errors do not affect the authenticity of the Certificate of Origin. 9.3.2.2 Verification The certificate often needs to be checked by the customs authorities of the importing member, in order to ensure that benefits are not unduly accorded to goods that do not comply with the origin requirements. Thus, there must be a system in place in order to ensure that the submitted origin-related information is accurate. Depending on the authority undertaking the verification, the verification systems can be classified into three different types. Direct verification means that the competent authority of the importing country directly conducts an audit or verification to an exporter/producer in the territory of the exporting country. 261 Handbook of Deep Trade Agreements Question: Is there a direct verification of the certificate? [SR_ver_dir] Indirect verification means that the competent authority of the exporting country undertakes verification upon request from the Customs authority of the importing country. Thus, this form of verification is built upon mutual administrative assistance of the competent authorities. Question: Is there an indirect verification of the certificate? [SR_ver_ind] Under a combined verification, the direct verification method is usually required only in exceptional cases, typically when the competent authority of the importing country is not satisfied with the outcome of the verification conducted by the Customs authority of the exporting country. Question: Is there a combined verification of the certificate? [SR_ver_two] Canada-Israel (1997), Ch. 5 on Customs Procedures, Art. 5.6 (1): Origin Verifications For purposes of determining whether a good imported into its territory from the territory of the other Party qualifies as an originating good, a Party may, through its customs administration, conduct a verification of origin, subject to paragraph 2, by means of: (1) written questionnaires to an exporter or a producer in the territory of that other Party for purposes of obtaining the information on the basis of which a Certificate of Origin referred to in Article 5.1 was completed and signed; (2) visits to the premises of an exporter or a producer in the territory of that other Party for purposes of reviewing the records referred to in Article 5.5 and to observe the facilities used in the production of the good; or (3) such other procedures as the Parties may agree. 9.3.2.3 Cumulation Cumulation refers to a diverse range of provisions which allow producers to consider materials purchased from outside of the PTA area as originating in the area for the purpose of determining origin. Cumulation is also relevant for the application ofVC methods involving cost calculations. Cumulation schemes can differ with respect to (a) which PTA members can cumulate (the so-called “quantitative” cumulation aspect), and to (b) the process and additional requirements for cumulation (“qualitative” aspect).14 Organized along these dimensions, cumulation can take four principal forms: bilateral (partial), full, diagonal, and cross-cumulation. 14 Inama 2011a; Cadestin et al. 2016. 262 Rules of Origin Bilateral (partial) cumulation Bilateral cumulation allows for cumulation of inputs between any two countries belonging to the same PTA, and permits each of them to treat products that originate in the other partner (according to the PSR stipulated in the PTA) as if they were its own. This applies only to originating goods and follows the same PSR as for originating goods. Bilateral cumulation between two members of the same PTA provides for the use of originating inputs from PTA partner countries. Under this scheme, only originating products or materials can benefit from the bilateral cumulation, and only inputs originating in one member country shall be considered as originating inputs in the other country. Question: Does the agreement allow for bilateral or partial cumulation? [SR_cum_bil] Japan-Philippines (2008), Chapter 3 (Rules of Origin), Article 30(1) (Accumulation) For the purposes of determining whether a good qualifies as an originating good of a Party, an originating good of the other Party which is used as a material in the production of the good in the former Party may be considered as an originating material of the former Party. Full cumulation requires that any processing activities carried in any PTA partner country can be counted as qualifying content regardless of whether the processing is sufficient to confer originating status to the material themselves. For PTA partners only, this can apply to non-originating goods (goods imported from outside the PTA, possibly within the provisions of another PTA) if the PTA with partner countries follows the same type of PSRs. In practice, this means that all operations carried out in the participating countries of a free trade zone (FTZ) are taken into account for origin determination purposes. Hence, non- originating imports with originating content from one party can be used by the other Party if the origin requirements are fulfilled within the FTZ as a whole. Question: Does the agreement allow for full cumulation? [SR_cum_ful] Diagonal cumulation refers to situations where there are three or more countries that have concluded bilateral or regional PTAs with each other. With any PTA partners and the same product-specific rules, this normally applies only to originating goods. 263 Handbook of Deep Trade Agreements Under this scheme, imported inputs originating in any PTA partner country can be counted as qualifying content when used in a country’s exports to the PTA area. In order to be applicable, diagonal cumulation requires that all trade agreements contain identical PSRs. Question: Does the agreement allow for diagonal cumulation? [SR_cum_dia] EFTA-Central America (2014), Annex I (Rules or Origin and Methods of Administrative Cooperation), Article 6(5) (Accumulation of Origin) When the EFTA States and Costa Rica or Panama have established a preferential trade agreement with a same non-party country or group of countries, the products or materials from that same non-party country or group of countries used in the manufacture of a product in their territories, may be considered as originating from the Party, as long as they comply with the specific rules of origin for thatproduct or material under this agreement. Cross-cumulation refers to situations where there are three or more countries that have concluded bilateral or regional PTAs with each other. This allows cumulation with any PTA partners even if there are no common PSRs, and it applies to all goods. This scheme refers to situations where at least three participating countries agree to merge individual overlapping bilateral treaties so that inputs can be sourced anywhere within the network. It is therefore one of the most flexible cumulation schemes. Question: Does the agreement allow for cross cumulation? [SR_cum_cro] EU-Papua New Guinea-Fiji (2009), Protocol II (Concerning the Definition of the Concept of “Originating Good” and Methods of Administrative Cooperation), Article 4 bis (Cumulation with Neighbouring Developing Countries). At the request of the Pacific States, materials originating in a neighboring developing country, other than an ACP State, can be considered as materials originating in a Pacific State when incorporated into a product obtained there. It shall not be necessary that such materials have undergone sufficient working or processing. For any cumulation method, detailed testing of the source of inputs can influence the restrictiveness of the origin rules. For example, tolerance or de minimis rules stipulate the conditions for the use of materials supplied by non-PTA members.These conditions are usually expressed as a maximum percentage of non-originating materials. In practice, the de minimis rule offers the possibility to comply with the CTC rule as long as the non-originating amount that does not undergo transformation does not exceed the stipulated threshold. Question: Does the agreement contain de minimis provisions? [SR_cum_min1] Question: What is the de minimis percentage? [SR_cum_min2] 264 Rules of Origin Korea-Colombia (2016), Chapter 3 (Rules of Origin and Origin Procedures),Article3.7(1) (De Minimis) A good that does not satisfy a change in tariff classification requirement pursuant to Annex 3-A is nonetheless originating if the value of all non-originating materials that have been used in the production of the good and that do not undergo the applicable change in tariff classification does not exceed 10 percent of the adjusted value of the good, provided that: (1) the value of such non-originating materials shall be included in the value of non-originating materials for any applicable regional value content requirement; and (2) the good meets all other applicable requirements in this Chapter. Similarly, according to the absorption principle, when a non-originating intermediate material acquires originating status by satisfying an initial PSR test, this material is considered to be fully originating once incorporated into a final product. The entire value of such intermediate materials is to be regarded as “originating” for purposes of determining the originating status of the final good. Question: Does the agreement include absorption provisions? [SR_cum_abs] CPTPP (2018),Annex 3-D (Product-Specific Rules of Origin),Section A(2) (General Interpretative Notes) Under this Annex, a good is an originating good if it is produced entirely in the territory of one or more of the Parties by one or more producers using non-originating materials, and: (1) each of the non-originating materials used in the production of the good satisfies any applicable change in tariff classification requirement, or the good otherwise satisfies the production process requirement, regional value content requirement, or any other requirement specified in this Annex; and (b) the good satisfies all other applicable requirements of Chapter 3 (Rules of Origin and Origin Procedures). 9.3.2.4 Methods of estimating value content Formulation of regional value content In cases where value content is a criterion for determining origin, a good is considered to be substantially transformed when the manufacturing operation conducted in the territory of a contracting party increases the value of the product, irrespective of changing its tariff classification. The increased value is usually expressed by an ad valorem percentage that can be calculated in one of two different ways: the regional content method or the import contents method. 265 Handbook of Deep Trade Agreements The regional content method imposes a minimum requirement on the value added within the region (in the form of material or processes) for the final product to be considered originating. This method requires a comparison between the value added exclusively in the region and the value of the final product, and can be calculated using different methodologies. Question: Is the value content requirement calculated as a minimum regional content requirement using a build-down calculation? [SR_vcr_rbd] Question: Is the value content requirement calculated as a minimum regional content requirement using a build-up calculation? [SR_vcr_rbu] Question: Is more than one calculation method permitted to determine the regional value content? [SR_rvc_alt] Dominican Republic-Central America-United States (2006), Chapter 4 (Rules of Origin and Origin Procedures), Article 4.2(1,3) (Regional Value Content) Where Annex 4.1 specifies a regional value content test to determine whether a good is originating, each Party shall provide that the importer, exporter, or producer may use a calculation of regional value content based on one or the other of the following methods: Method Based on Value of Non-Originating Materials (“Build-down Method”) RVC = AV - VNM x 100 AV or Method Based on Value of Originating Materials (“Build-up Method”) RVC = VOM x 100 AV or Method for Automotive Products (“Net Cost Method”) RVC = NC - VNM x 100 NC where RVC is the regional value content, expressed as a percentage; AV is the adjusted value; NC is the net cost of the good; VNM is the value of non-originating materials that are acquired and used by the producer in the production of the good (VNM does not include the value of a material that is self-produced); and VOM is the value of originating materials acquired or self-produced, and used by the producer in the production of the good. 266 Rules of Origin The import contents method establishes a maximum allowance for non-originating materials, implying that a final product can be considered as originating if foreign inputs do not exceed a certain threshold in terms of value. This method requires a comparison between the value of the imported inputs or inputs of undetermined origin and the value of the final product. Question: Is the value content requirement calculated through import content? [SR_vcr_imc] Australia- Papua New Guinea (1976), Article 4 (Rules of Origin of Goods) Goods shall be treated as originating in a Member State if those goods are: (A) the unmanufactured raw products of that Member State, or (B) manufactured goods in relation to which (i) the process last performed in the manufacture was performed in that Member State, and (ii) the expenditure (1) on material that is of Member State origin, (2) on labor, factory overheads and inner containers that are of Member State origin, or (3) partly on such material and partly on such other items of factory cost, is not less than one- half of the factory or works costs at the time of exportation. Notwithstanding the provisions of paragraph 1 of this Article, the Member States may agree to treat particular goods or classes of goods as originating in a Member State provided that, in the case of manufactured goods, the process last performed in the manufacture was performed in the territory of the exporting Member State. Some PTAs allow a combination of both approaches. Question: Is the value content requirement calculated through both regional and import content? [SR_vcr_ric] The valuation of non-originating materials The price basis used for the calculation of the value of final product also influences the ability to use non-originating inputs. This basis may be established at different moments in the value chain.There are four principal approaches to calculating value added.The ex-work cost (or factory cost) approach, for example, may exclude some costs, such as packing cost or packing material. The ex-works price approach uses the price paid to the manufacturer which is undertaking the last working or processing, provided the price includes the value of all products used in the manufacturing. The FOB price approach includes all costs incurred by placing the goods on board an exporting ship, aircraft, or other vehicle in addition to the ex-works price. Question: Is the price basis for the content threshold requirement the ex-works cost? [SR_vcr_cst] Question: Is the price basis for the content threshold requirement the FOB/net price? [SR_vcr_fnt] Question: Is the price basis for the content threshold requirement the ex-works price? [SR_vcr_prc] 267 Handbook of Deep Trade Agreements Question: Is the price basis for the content threshold requirement the FOB price? [SR_vcr_fob] Australia-China (2015), Chapter 3 (Rules of Origin and Implementation Procedures), Article 3.5(2) (Regional Value Content) The value of the non-originating materials shall be: (1) the CIF value of imported materials, determined in accordance with the Customs Valuation Agreement; or (2) the value determined in accordance with the Customs Valuation Agreement when the non- originating materials are acquired within the territory of that Party, not including freight, insurance, packing costs and any other costs incurred in transporting, within the Party’s territory, the non-originating materials to the location of the producer. The percent requirement is sometimes not defined at the product level but applied to all products covered by the PTA with a value content (VC) test. Question: What is the percentage of value content required? [SR_vcr_per] Australia-Papua New Guinea (1976), Article 4 (Rules of Origin of Goods) Goods shall be treated as originating in a Member State if those goods are: (A) the unmanufactured raw products of that Member State, or (B) manufactured goods in relation to which (i) the process last performed in the manufacture was performed in that Member State, and (ii) the expenditure (a) on material that is of Member State origin, (b) on labor, factory overheads and inner containers that are of Member State origin, or (c) partly on such material and partly on such other items of factory cost, is not less than one-half of the factory or works costs at the time of exportation. Notwithstanding the provisions of paragraph 1 of this Article, the Member States may agree to treat particular goods or classes of goods as originating in a Member State provided that, in the case of manufactured goods, the process last performed in the manufacture was performed in the territory of the exporting Member State. Question: What is the percentage of value content required with alt method? [SR_vcr_per2] 9.3.2.5 Other aspects Duty drawback Duty drawback allows tariffs due on imported materials used in the production of export items to be waived or refunded. Accordingly, the no-drawback rule means that there is 268 Rules of Origin no refund of duties paid for input materials from third countries which are used for the final products. The objective is to ensure that there will be equal treatment between goods manufactured and traded in the domestic market and those that will be exported to trade partner countries. Allowing for the possibility of drawback duties could result in the exported goods being cheaper than the same good sold on the domestic market. Question: Does the agreement contain drawback rules? [SR_drb] Question: Does the agreement allow drawback? [SR_dba] Question: Does the agreement prohibit drawback? [SR_dbp] Australia-China (2015), Chapter 4 (Customs Procedures and Trade Facilitation), Article 4.12 (Temporary Admission of Goods) 1. Each Party shall allow, as provided for in its laws and regulations, goods to be brought into its territory conditionally relieved, totally or partially, from payment of import duties and taxes if such goods: (1) are brought into its territory for a specific purpose; (2) are intended for re-exportation within a specific period; and (3) have not undergone any change except normal depreciation and wastage due to the use made of them. 2. A Party shall not apply any import duties or taxes on containers, pallets or packing material used in the transportation of goods. The concept of fungible goods or fungible materials Fungible goods or fungible materials are those that are interchangeable for commercial purposes insofar as their properties are essentially identical.When using originating and non-originating fungible materials, manufacturers are not required to stock those materials separately in order to trace them back to their different origins, but are rather permitted the use an inventory management method in accordance with the Generally Accepted Accounting Principles. Question: Does the agreement allow for joint storage of originating and non-originating inputs when these inputs are interchangeable? [SR_fng] CPTPP (2018), Chapter 3 (Rules of Origin and Origin Procedures), Article 3.12 (Fungible Goods or Materials) Each Party shall provide that a fungible good or material is treated as originating based on the: (1) physical segregation of each fungible good or material; or (2) use of any inventory management method recognized in the Generally Accepted Accounting Principles if the fungible good or material is commingled, provided that the inventory management method selected is used throughout the fiscal year of the person that selected the inventory management method. 269 Handbook of Deep Trade Agreements Advance rulings These rulings are binding official decisions issued by a competent authority which provides the applicant with an assessment of the origin prior to an import or export transaction for a specified period.The origin legislation of some countries sets out the legal basis for issuing advance rulings in a specific rule, while other countries deal with this matter under their general Customs Law. Question: Does the agreement allow for advance rulings? [SR_adr] Korea-Colombia (2016), Chapter 4 (Customs Administration and Trade Facilitation), Article 4.9(1) (Advance Rulings) Each Party shall issue, through its Customs Authority, prior to the importation of a good into its territory, a written advance ruling upon written request of an importer in its territory, or an exporter or producer in the territory of the other Party with regard to: (1) tariff classification; (2) the application of customs valuation criteria for a particular case, in accordance with the provisions of the Customs Valuation Agreement; (3) whether a good is originating in accordance with Chapter 3 (Rules of Origin and Origin Procedures); and (4) such other matters as the Parties may agree. Transshipment rule This rule implies derogation to the direct transport rule of the origin legislation, and so establishes the conditions by which a good may maintain its originating status when transported through the territory of a non-Party to the Agreement. Question: Does the agreement contain a transshipment rule? [SR_trs] CPTPP (2018), Chapter 3 (Rules of Origin and Origin Procedures), Article 3.18 (Transit and Transshipment) 1. Each Party shall provide that an originating good retains its originating status if the good has been transported to the importing Party without passing through the territory of a non-Party. 2. Each Party shall provide that if an originating good is transported through the territory of one or more non-Parties, the good retains its originating status provided that the good: (1) does not undergo any operation outside the territories of the Parties other than: unloading; reloading; separation from a bulk shipment; storing; labeling or marking required by the importing Party; or any other operation necessary to preserve it in good condition or to transport the good to the territory of the importing Party; and (2) remains under the control of the Customs administration in the territory of a non-Party. 270 Rules of Origin Review and appeal Importers, exporters, and producers are entitled to request a review of decisions rendered by Customs authorities with respect to origin determination. In terms of the legal basis for review and appeal, there are two approaches. Some agreements contain specific provisions on review and appeal. In other cases, review and appeal are instead regulated in national legislation. Question: Does the agreement contain specific review and appeal mechanisms? [SR_rev] Chile-Vietnam (2014), Chapter 5 (Customs Administration), Article 5.4 (Review and Appeal) 1. Each Party shall ensure that with respect to its determinations on customs matters, in accordance with the Party’s domestic laws and regulations, importers in its territory have access to: (1) administrative review independent of the official that issued the determination; and (2) judicial review of the determination or decision taken at the final level of administrative review. 2. Notice of the decision on appeal shall be given to the appellant and the reasons for such decision shall be provided in writing. 9.4. STYLIZED FACTS Using the classifications described above, RoOs have been mapped for various PTAs: 160 PTAs with product-specific RoOs and 250 PTAS with regime-wide provisions.The present section surveys preferential RoO regimes of several PTAs signed by Latin American (LAC) and East Asian and Pacific (EAP) countries within their respective regions and with extra- regional partners. All PTAs are thus classified into four groups depending on members they cover: Intra-EAP, Intra-LAC, Extra-EAP, and Extra-LAC. The aim is to show how the taxonomy of RoOs described above and the data collected using this taxonomy can be utilized to compare the product-specific and the regime-wide RoOs negotiated by different countries or regional groupings in terms of rules content and rules administration. The overriding objective of classifying the different RoO provisions is to enable empirical investigation of the restrictiveness and trade effects of different RoOs and thus to inform the debate on necessary reforms in existing and upcoming PTAs. The application of different product-specific rules PSR to similar products in different PTAs, the lack of a harmonized preferential RoO across the different agreements, and the increasing globalization of production create opportunities for countries to use their RoOs to implement trade policies in a biased manner. In such globalized context, there is no single correct definition of origin, and the dependence of the origin of a product on the formulation and application of the applicable RoOs often interferes with the usual business and market practices and the ways technology is being used. 271 Handbook of Deep Trade Agreements 9.4.1 Product-specific rules of origin 9.4.1.1 Historical review of PTAs through their PSR provisions A brief historical review of the regional integration experiences of EAP and LAC is useful to understand the rationale for RoO regimes negotiated in these regional contexts, and identify some commonalities and differences and their most salient features. Latin America (LAC) The PTAs signed by Latin American countries have been influenced by two different families of agreements - intra-regional and those negotiated with extra-regional partners. One way of analyzing the different regimes in force is by comparing the principal features of three main regimes, which are used as reference frameworks: the Latin American Integration Association (LAIA), the North American Free Trade Agreement (NAFTA), and the Central American Common Market (CACM; see Abreu (2016)). Traditional Latin American trade agreements are based mostly on the LAIA model, which provides that at least two countries in the region can enter into an agreement that grants tariff preferences to participants, thereby establishing an economic cooperation area (ECA). The first official set of RoO criteria under the LAIA model was established in 1987 (Resolution 78). These rules were brief and general in scope, using a general rule applicable across the board for all tariff items. The origin criteria stipulated that the CIF value at the port of destination, or at the maritime port of materials from third countries, must not exceed 50 percent of the FOB export value of the final good, or 60 percent in the case of countries with relatively less developed economies. It should also be noted that regarding cumulation, the rule provided that domestic or local value only applies to value generated in countries belonging to a given ECA. In this case, the general criteria of origin were relatively easy to meet, making several goods traded under the ECAs eligible for customs preferences. This approach, however, also had some negative aspects, including the possibility of contradictory interpretations, which hampered the predictability of trade and production conditions in member countries.15 As a result, some of the agreements based on the LAIA model have modified their regimes in recent years. Many countries have decided to apply rules that are more selective and less uniform than those of LAIA Resolution 78, while at the same time preserving CTC as the basic qualification criterion and rejecting a multiplicity of “rule families” at the tariff line level, which often occur in new-generation trade agreements. 15 The LAIA model is the point of reference for RoOs used in the Andean Community (CAN) and the Southern Common Market (Mercosur), as well as the agreements between countries in the two groupings. These ECAs provide for the establishment of free trade areas among its parties, and eliminate duties and other barriers to trade. 272 Rules of Origin The NAFTA model has inspired a new generation of trade agreements by the US, Canada, and Mexico. The RoO regimes in these agreements typically require a change of chapter, heading, subheading, or item, depending on the product in question. Additionally, many products combine the change of tariff classification with an exception, regional value content, or technical requirement. Finally, the CACM model uses an RoO regime that is an eclectic construction with some characteristics of Mercosur and some of NAFTA. These RoOs mostly require a change in tariff classification criteria, but unlike in Mercosur, the change must take place at the chapter, heading, or subheading level, depending on the product. It must be noted that the Central America Free Trade Agreement (CAFTA) contains a multilateralism principle under which it coexists with the CACM. Thus, Central American producers can freely choose between the CACM or CAFTA RoOs when exporting to other markets. The parallel existence of so many different RoO regimes in Latin America closely relates to the story of trade interests that were at stake in the past. In the case of NAFTA and agreements modeled on NAFTA, since the RoOs were negotiated for specific products, the requirements are quite clear and leave no room for conflicting interpretations. Nonetheless, one of the disadvantages of NAFTA is that every rule is quite long and detailed, potentially turning the negotiation of new trade agreements into a long and complicated process. Additionally, NAFTA families of RoOs do not have much impact on Latin American producers, since most large exporters sell a limited range of products and only have to be familiar with RoOs for the goods that matter to them. Conversely, these rules have been constraining for exporters that sell a wide range of goods or export goods other than commodities. East Asia and the Pacific (EAP) East Asia and the Pacific is a heterogeneous region with several economic and political heavyweights, including Japan, China, and the US (which has a strong commercial presence there); some mid-sized but politically and economically sophisticated partners such as Korea; and several smaller and very diverse countries at different levels of economic development. At the same time, and in contrast to other regions, the literature recognizes that the two main regional powers - Japan and China - have so far not played an important role in promoting regional integration. One of the most important regional integration projects was the creation of the Association of South East Asian Nations (ASEAN) in 1967, with Indonesia, Malaysia, the Philippines, Singapore, and Thailand as members, joined by Brunei Darussalam in 273 Handbook of Deep Trade Agreements 1984. Reflecting the rejuvenated will to enhance regional cooperation, the ASEAN Free Trade Area (AFTA) was formed in 1993, and subsequently Vietnam, the Lao People’s Democratic Republic, Myanmar, and Cambodia joined between 1995 and 1999. This regional approach was largely inspired by the creation of the EU and NAFTA, since ASEAN nations also wanted to reap the benefits of regional integration through enhanced economic relations and institutional cooperation. A proliferation of PTAs in EAP has added complexity to the region’s RoO regimes, notably through trade agreements signed with extra-regional partners. Even though many countries in the region have agreements similar to NAFTA and EU types of treaties, ASEAN has been the key reference point for RoO negotiations in the region. In fact, ASEAN has developed its own RoO model and has sought to promote it in other agreements with its regional trading partners.16 ASEAN also put in place the ASEAN Trade in Goods Agreement (ATIGA) in 2010. Before ATIGA was created, the region’s trade rules were set by the ASEAN Free Trade Agreement– Common Effective Preferential Tariff (AFTA–CEPT), which adopted the regional value content (RVC) rule. This approach was meant to be more liberal and straightforward than product-specific RoOs, which were seen as potentially very limiting. Over time, however, the CTC approach became dominant due to some practical problems with implementing the RVC approach. ATIGA introduced further improvements to liberalize and simplify the RoOs, and has refined them on a product-by-product basis rather than reforming the overall framework. In general, in East Asia and the Pacific, regionalism is a relatively recent phenomenon, but after the creation of AFTA in 1993, the drive for regional trade liberalization accelerated. However, many sub-regional specificities remain. Many of the main integration schemes in the region such as AFTA, ASEAN-China, ASEAN- Korea, and the South Pacific Regional Trade and Economic Cooperation in Asia-Pacific (SPARTECA) establish an across-the-board value content rule with relatively few exceptions. In the cases of AFTA and ASEAN-China, there is a simple and uniform format for RoOs, since CTC is not necessary and TRs are not mentioned. Instead RoOs only require 40 percent local content. Conversely, Japan and Korea rely on the NAFTA framework, including CTC and VC ratios, which makes them more restrictive than AFTA and ASEAN-China.The RoOs of the Japan-Singapore Economic Partnership Agreement (JSEPA) are also complex, but many of the rules provide for a simple change in heading. Additionally, for many products JSEPA introduces an alternative VC rule, granting flexibility to the originating criteria. 16 ASEAN is not a customs union and it enters into trade negotiations with other trading nations within Asia and with the rest of the world. 274 Rules of Origin 9.4.1.2 Differences across regions Figure 9.1 draws on the classification of RoOs presented at the beginning of this chapter and presents the characterization of the different PSR regimes across 143 PTAs in LAC and EAP for which data have been collected. A number of regularities can be observed. First, in both regions, the four types of qualification criteria used to determine origin - wholly obtained (WO), value content (VC), change in tariff classification (CTC), and technical regulations (TR) - can be used either uniformly or selectively. Due to the multiple influences of different families of RoOs, no single family clearly dominates, although change in heading (CH) and change in heading or value content (CH or VC) represent 75 percent of PSRs in intra-LAC PTAs, 28 percent of extra-LAC PTAs, 25 percent of intra-EAP PTAs, and 32 percent of extra-EAP PTAs. Furthermore, some clear regional patterns emerge. EAP countries tend to allow more alternative rules (“or”) with other EAP partners than with extra-regional partners and also contain fewer rules with exceptions (“EXC”) with regional partners, indicating less restrictive rules with the EAP partners. However, EAP countries include more TR rules, which are highly restrictive, in their PTAs with other EAP than with rest of the world (RoW). Figure 9.1: Distribution of product-specific rules of origin across regional PTAs Source: Deep Trade Agreements Database. Note: CC: change in chapter; CH: change in heading; CS: change in subheading; TR: technical requirement;VC: value content; WO: wholly obtained; EXC: exceptions.Vertical axis indicates the share of HS6 digit products covered (across 5200 products). Numbers in parentheses indicate the number of intra-regional and extra-regional PTAs in LAC and EAP . The second observed regularity is that LAC countries tend to allow more alternatives rules (“or”) with other LAC partners than with extra-LAC partners and also present fewer cumulated (“and”) and exception (“EXC”) types of rules with regional partners, indicating less restrictive rules with other LAC partners. LAC countries also include more TR rules, which are highly restrictive, in their PTAs with LAC partners than with RoW. 275 Handbook of Deep Trade Agreements In terms of assessing the restrictiveness of different PSRs, a brief summary of the main findings from the literature is given below. The overriding conclusion is that one cannot simply say whether a change of classification rule is more restrictive than a value test rule or a technical requirement test. It is also important to note that the choice of a change in chapter (CC) over a change in heading (CH) or change in subheading (CS) rule is due to the organization of the Harmonized Commodity Description and Coding System (Harmonized System, HS).As a general rule, the HS includes less- processed products (live animals, plants, raw materials followed by foodstuffs, chemical products) in the first chapters, followed by increasingly processed and complicated products (textiles, footwear, followed by household appliances, computers, vehicles). However, focusing on a subset of products for which alternative rules are provided,17 and without having to investigate in detail the text of those product-specific rules, it can be concluded that: • CTC stringency tends to be higher the lower the level of product classification transformation, as much greater transformation is usually required to change a product into a completely different product category. Therefore, an operation that results in a CC is more restrictive than one that results in a CH, which in turn is more restrictive one that results in a CS. • VC stringency rises with a higher share of local value-added required, as this requires more of the total cost to be spent on locally sourced inputs.This can be as restrictive as a CH test.18 • TR stringency rises the greater the technical change specified.The TR test is approximately as restrictive as a CC test.19 • Exceptions reduce the universe of permitted third-country inputs, and because the HS is not designed as a mechanism for the definition of PSRs, exceptions to a change of tariff classification rule are specified only in cases where they are actually meaningful for the production of the good in question, indicating a deliberate intention to be restrictive. • The mere existence of alternatives indicates a less restrictive environment than specifying only one rule. Indeed, the administrative burdens of qualifying for the tariff preference can be significantly different for different types of PSRs, and the mere fact that alternatives are permitted shows a belief that some traders would suffer a lesser burden under one or the other alternative. • The existence of cumulated rules indicates a more restrictive environment. 17 It is reasonable to assume that alternative qualification methods for a given product should be of very similar restrictiveness. 18 Harris 2007. 19 Harris 2007. 276 Rules of Origin Box 9.1: CTC type rules versus RVC type rules Regulatory risk associated with uncertainty of origin would be expected to influence the way businesses act, thereby adding to the restrictiveness of an origin regime. It is therefore possible for origin regimes to be highly restrictive in the case of some CTC rules but still be relatively certain for other CTC rules. On the other hand, other methods that may appear less restrictive, including those based on an RVC requirement with a relatively low threshold, could be less certain because of exogenous factors (e.g., exchange rate fluctuations). Consequently, some studies consider the RVC test to be more restrictive while origin rules based on the application of CTC alone are considered relatively more certain. On the other hand, an RVC rule would allow the producer to source from any mix of foreign inputs, as long as the threshold is met for the overall originating content of the final product. Rules of origin can also restrict technical and organization change in the production of goods in member countries. For example, rules based on a VC threshold can restrict technological and organizational change where firms are close to the content threshold, while rules based on a CTC method to some extent accommodate better relevant technologies. Overall, there are significant costs associated with compliance and verification of the VC rule, as this requires sophisticated accounting systems and large customs capacity. Additionally, the VC method also tends to penalize efficient low-cost producers. In industries where labor and assembly costs are lower than in higher-cost but inefficient facilities, RoOs probably will not be met. Thus, this method seems somehow at odds with the supply necessities of global value chains (GVCs). 9.4.1.3 Heterogeneity A country can have different PTAs with different RoOs, creating confusing and sometimes conflicting sets of rules that are hard to navigate for many exporters and importers and raise the costs of trade. A significant consequence of the divergence of RoOs across agreements is that an exporter faces different RoOs in different markets. If a particularly restrictive market is important to a producer, it is possible that the producer may be forced to change its production process in order to comply. It is also possible that the compliance and administrative costs of membership in multiple agreements are greater than the sum of the costs of the individual agreements, due to the additional coordination required. In such cases, the existence of multiple agreements would add to the actual trade restrictiveness of individual agreements.To investigate this possibility, a new index was set up to capture the divergence of a country’s RoOs across its different trade agreements. A Shannon- type index20 was used to indicate the number of different RoOs which can apply to a product (HS6 digit level) in a country because of multiple PTAs with different PSRs. 20 Calculated as follows: S H = ∑ i=1- (Pi * ln Pi) where: H = the Shannon diversity index, Pi = fraction of the entire population made up of PSR i, S = numbers of existing PSR in the classification, and ∑ = sum from PSR type 1 to PSR type S. 277 Handbook of Deep Trade Agreements Because of the way the index is constructed, the restrictiveness of PSRs in a country is considered to be at its highest when its agreements invoke more than one method for determining origin. Membership in multiple agreements which use dissimilar rules is considered to be more restrictive, while membership in only a single agreement is considered the least restrictive according to this criterion. For example, a Shannon index of 1.50 can be converted into exp(1.50) = 4.5, which is equivalent in heterogeneity to a country with 4.5 different types of PSRs, on average, per product. In this way, heterogeneity can be compared across countries and sectors (Table 9.1). Chemicals, footwear, and textiles are the sectors with the highest diversity of PSRs. Countries such as Malaysia or Thailand, although they have fewer PTAs than Japan or China, still have higher scores on the heterogeneity index. This could indicate that they signed PTAs with countries that have more negotiating power and were able to impose their types of PSRs. This heterogeneity in PSRs can also be illustrated by comparing all South-South and North- North agreements compared to North-South. As shown in Figure 9.2, heterogeneity in PSRs has increased over time because of the addition of new PTAs, which often come with new Table 9.1: Heterogeneity scores for some EAP countries HS sections Philippines Cambodia China Korea ,Rep. Indonesia LaoPDR Vietnam Japan Thailand Malaysia Average Live animals 0.78 0.78 0.70 0.73 0.78 0.89 1.04 0.29 0.75 1.07 0.76 Vegetable 0.93 0.93 0.86 0.75 0.95 1.08 1.06 0.40 0.87 1.21 0.89 Mach & electrical 0.75 0.75 0.98 1.15 0.85 1.02 0.93 1.10 1.26 1.33 0.94 Raw hides 1.00 1.00 0.76 0.74 1.00 1.14 1.14 0.64 1.25 1.38 1.00 Wood 1.04 1.04 0.89 0.82 1.12 1.17 1.26 0.71 0.90 1.29 1.03 Mineral 0.92 0.92 1.08 0.88 1.14 1.14 1.24 1.29 1.19 1.66 1.09 Stone 0.90 0.90 1.00 0.93 1.13 1.13 1.10 1.55 1.38 1.59 1.10 Pulp of wood 0.91 0.91 1.14 0.76 1.14 1.14 1.13 1.68 1.21 1.60 1.10 Miscellaneous 0.91 0.91 1.19 1.20 1.11 1.08 1.13 1.52 1.53 1.52 1.13 Fats and oils 1.06 1.06 0.88 1.73 1.12 1.20 1.21 0.80 1.09 1.51 1.15 Vehicles 0.95 0.95 1.15 1.15 1.23 1.22 1.18 1.63 1.52 1.51 1.17 Plastics 1.00 1.00 1.11 0.96 1.26 1.26 1.28 1.55 1.52 1.50 1.19 Optical 1.07 1.07 1.12 1.32 1.24 1.31 1.26 1.47 1.55 1.63 1.21 Prepared foods 1.21 1.21 0.92 1.44 1.29 1.38 1.34 0.83 1.30 1.62 1.21 Base metal 1.05 1.05 1.12 1.06 1.31 1.25 1.30 1.71 1.41 1.74 1.22 Chemical 1.14 1.14 0.95 1.32 1.37 1.35 1.29 1.36 1.59 1.56 1.24 Footwear 1.27 1.27 1.29 1.13 1.36 1.49 1.47 1.09 1.44 1.38 1.29 Textiles 1.47 1.47 1.25 1.81 1.57 1.65 1.50 1.65 1.64 1.70 1.48 Average 0.98 0.98 1.03 1.09 1.15 1.17 1.20 1.23 1.30 1.51 Source: Deep Trade Agreements Database. 278 Rules of Origin rules for new pairs of partners. However, intra-North PTAs tend to be more homogeneous than intra-South PTAs, possibly because intra-South includes both intra-Asia and intra- Latin America, which have different approaches to PSRs. In addition, there has been a large increase since 2000 in the heterogeneity of PSRs in North-South PTAs. Therefore, those types of agreements potentially raise more concerns in terms of compliance with RoOs and utilization of tariff preferences. Figure 9.2: Heterogeneity of product-specific rules of origin over time, average across all six-digit products Source: Deep Trade Agreements Database. Note: Heterogeneity score on the vertical axis, average across all 5200 HS products. 9.4.1.4 Distribution of PSRs across sectors Being a sensitive sector and containing several unprocessed products, agriculture relies heavily on restrictive RoOs such as CC and WO. For other unprocessed products such as rawhide and skins and base metals, there is a clear dominance of CH rules. Finally, in more technical industries, such as machinery and electrical equipment, vehicles, and optical, RoOs rely more on VC requirements. Interestingly, RoOs in textiles and footwear are characterized by exceptions to the CTC rule (often change in heading) and the technical requirements rule, indicating a voluntary complication of the compliance process for exporters (see Box 9.2). It is also interesting that CH and VC rules seem equally distributed across sectors. 279 Handbook of Deep Trade Agreements Box 9.2: Case study 1 - CTC rules in textiles CTC rules often vary regionally to reflect the resources available within the area covered by the trade agreement and country-specific sensitivities, and to incentivize production within that region.This can be seen, for example, in the rules of origin for textile and apparel products (chapters 50-63 in the Harmonized System). Some rules of origin for textiles and apparel are referred to as “fiber forward,” which means that the fiber has to have originated within the trade area in order for the final product to have originating status under the agreement. Such rules tend be applied to textile and apparel goods made of fibers (such as cotton) which are abundant in the region. Other rules confer origin at the next stage of processing, allowing fiber to be sourced from outside the trade area but conferring origin based on where the yarn (such as wool or polyester) is produced; these are “yarn forward” rules. Another approach is to confer origin based on where further processing takes place, such as where a garment is knitted, woven, or cut and sewn; these are “single transformation” rules, and tend to be used for products made of fibers (such as silk) that are largely produced outside of the area covered by the trade agreement. For example, in X agreement, a woman’s knit blouse would be classified in 61.06. Its applicable rule of origin would imply that wool blouses need to comply with a “yarn forward rule,” while silk blouses need to comply with a “single transformation rule,” as shown: 61.05-61.06 A change to heading 61.05 through 61.06 from any other chapter, except from heading 51.06 through 51.13, 52.04 through 52.12, 53.07 through 53.08 or 53.10 through 53.11, Chapter 54, or heading 55.08 through 55.16 or 60.01 through 60.02, provided that the good is both cut (or knit to shape) and sewn or otherwise assembled in the territory of one or more of the Parties. This means that if a wool blouse is made inside the trade region entirely from wool yarn (heading 51.06) imported from a non-Party, this product would be considered non-originating, as it follows a “yarn forward rule.” If, on the other hand, a silk blouse is made entirely from silk (heading 50.04) from a non-Party, this product would be originating, as it follows a “single transformation rule.” Cotton yarn is classified in 52.05. Its rule of origin is a “fiber forward rule,” as shown: 52.01-52.07 A change to heading 52.01 through 52.07 from any other chapter, except from heading 54.01 through 54.05 or 55.01 through 55.07. If cotton yarn is made entirely in the region from cotton fiber (heading 52.01), this product would only be originating if the fiber comes from any Party to the Agreement, as it follows a “fiber forward rule.” Other agreements have different approaches, such as establishing a “yarn forward rule” for all products but allowing the use of very specific non-originating materials found in a List of Short Supply of Products, while still conferring origin. 9.4.2 Regime-wide RoOs matter PTAs also include regime-wide rules that apply to all products and can either increase or decrease the restrictiveness of the product-specific criteria. Different types of regime-wide rules have different associated costs. 280 Rules of Origin 9.4.2.1 The certification process The more numerous the bureaucratic hurdles and the higher the costs for an exporter to obtain an origin certificate, the lower the incentives to seek PTA-conferred preferential treatment.21 Self-certification largely minimizes the involvement of the competent authority in the issuing process. Certification by the importer (such as in the Trans-Pacific Partnership, TPP) is considered a more flexible option than certification by the exporter/producer (e.g., NAFTA), which in turn is more flexible than certification by the authority or an authorized exporter scheme.The data show a striking difference in certification rules between intra- and extra-regional PTAs for both EAP and LAC (Table 9.3). Indeed, both regions have adopted the most restrictive approach to certification for their intra-regional PTAs, which seems to indicate a lack of confidence in the certification operators of their regional trade partners. Table 9.2: Distribution of product-specific rules of origin across HS sections Share of HS 6 digit products covered by different types of PSR (in percentage) HS Section CC CC_ CC & CC or CH CH_ CH & CH or CS CS_ CS & CS or TR VC WO EXC VC/ TR VC/TR EXC C/TR VC/TR V EXC VC/TR VC/TR Live animals 39 7 0 1 10 1 0 13 0 0 0 1 3 5 20 Vegetable prod 38 1 0 2 12 1 0 18 2 0 0 1 3 6 16 Fats and oils 29 7 3 4 10 4 1 19 2 0 0 0 4 7 9 Processed food 22 9 2 3 18 3 2 20 2 0 0 1 5 9 5 Minerals prod 12 0 0 1 35 0 2 26 1 1 1 2 5 11 3 Chemical prod 1 0 0 0 17 1 5 25 20 1 1 9 5 14 1 Rubber and plastics 2 0 1 1 31 1 12 29 5 0 1 2 4 12 1 Raw hide skins 21 1 0 1 34 3 2 23 2 0 0 1 3 8 1 Wood 6 0 1 1 46 2 1 25 1 0 0 2 3 11 1 Paper 11 0 0 1 38 2 2 26 1 0 1 3 3 12 0 Textiles 9 8 12 3 13 13 7 12 0 0 0 0 14 8 1 Footwear 9 0 0 1 29 14 6 19 2 1 2 1 5 10 1 Stone and cement 9 0 0 1 31 8 3 30 2 0 1 2 3 9 1 Precious stones 15 0 0 1 30 5 2 24 2 0 1 5 4 9 2 Base metals 8 1 1 1 33 3 2 25 4 0 2 4 4 10 1 Mach & Electrical 0 0 0 0 12 1 6 28 11 1 2 12 4 22 1 Motor vehicles 1 0 1 1 21 2 10 26 3 0 1 3 4 26 1 Optical medicals 1 0 3 0 16 1 7 29 9 0 2 7 4 19 1 Miscellaneous 7 0 4 1 22 0 3 34 8 1 2 5 2 11 0 TOTAL 11 2 3 1 21 4 4 23 6 0 1 4 6 12 3 Source: Deep Trade Agreements Database. Note: CC: change in chapter; CH: change in heading; CS: change in subheading; TR: technical requirement;VC: value content; WO: wholly obtained; EXC: exceptions. 21 The WCO guidelines on certification of origin encourage, in the first instance, the use of self-certification by the importer or exporter. 281 Handbook of Deep Trade Agreements The validity period for certification matters, since it allows the parties a longer period to conclude the importation process, claim preferential treatment, and correct any issues that may arise. Here again, it can be observed that for both regions, the validly is slightly less for intra-regional PTAs than for extra-regional PTAs. The recordkeeping requirement of the certification process is important to guarantee that producers and authorities have access to origination data in case a verification question arises. However, recordkeeping requires a large amount of effort and resources, especially for producers.22 On the one hand, producers need to maintain and manage records on origin information. On the other hand, manufacturers must maintain and manage records on their suppliers and must examine the bill of materials of the exported product in order to establish the origin and value of individual inputs.23 This recordkeeping requirement is roughly similar across PTAs in both regions. Exemptions and amendments for minor errors help to ease the potential restrictiveness of RoOs. Indeed, with exemptions rules, if the threshold for exemptions is sufficiently high, small parcels of exports, often from SMEs, do not have to go through the complicated RoO certification process. In the same vein, in case of minor errors on the certificate, the authority would allow the certificate to be modified or reissued instead of automatically denying preferential treatment. Latin American PTAs have clearly adopted a less restrictive threshold of exemptions than EAP, and again, for both regions, the threshold is more restrictive (lower) for extra-regional than for intra-regional PTAs. As for amendments of minor errors, the adoption of this rule is below 50 percent in general, which is well below the average for PTAs in the EAP region. Table 9.3: Provisions on certification aspects Agreements Self- Authority Combined Average validity Average record Average amount Amendment certification certification certification in months keeping in years for exemptions for minor errors INTRA LAC 29% 64% 7% 13 4,3 1087 19% INTRA EAP 5% 95% 0% 11 3,6 629 43% LAC with EXT 32% 34% 34% 19 4,2 1167 49% EAP with EXT 22% 39% 39% 14 4,2 943 46% Others 11% 18% 71% 6 3,4 1363 81% Source: Deep Trade Agreements Database. Note: Others = PTAs between non-LAC or non-EAP countries. 22 The software that companies use to retain the data on each transaction needs to be modified or instructed to retain the origin information specified in each different preferential agreement. 23 In practice, this means that the exporting manufacturer is obliged to maintain detailed origin communications and receive origin certificates from its suppliers. 282 Rules of Origin The verification process also matters in terms of RoO burden. In that domain the indirect verification (generally in the form of a detailed questionnaire sent to the exporting Party) can be considered as less restrictive and less costly than direct verification (generally in the form of a visit by the importing authorities to the exporting Party). Indirect verification however is rarely used in intra PTA for Latin America, and more generally, it is lower for EAP and LAC PTAs than for other PTAs. Table 9.4: Provisions on verification process Agreements Direct (importer) Indirect (exporter) Combined NTRA LAC 79% 5% 5% INTRA EAP 43% 33% 19% LAC with EXT 41% 51% 5% EAP with EXT 43% 31% 4% Others 17% 79% 0% Source: Deep Trade Agreements Database. Note: Others = PTAs between non-LAC or non-EAP countries. 9.4.2.2 Important mechanisms: The cumulation and de minimis rules For cumulation, the flexibility criteria are fairly straightforward. Cumulation provisions determine which products, which processes, and which countries are able participate in the elaboration of the product and still be considered “originating.” Bilateral cumulation is the minimum standard and is more restrictive than diagonal, which allows the inclusion of trading partners from PTAs, providing the PSRs are similar. Diagonal is, in turn, more restrictive than full cumulation, which allows the use of non-originating material within the PTA. This provision is, in turn, less flexible than cross cumulation, which allows the use of non-originating goods from other trading partners in other PTAs even if the PSRs differ.24 Here there is a clear regional pattern, where EAP seems to have more flexible cumulation rules compared to Latin America, which has more flexible mechanisms (see Box 9.3). The de minimis rule, also called the tolerance rule, can be found in several origin regimes. It is an important flexibility provision, in that a product is considered to have complied with the strict CTC rule as long as the value of the non-originating component does not exceed, for example, 10 percent of the value of the final good. As a general practice, the more extensive the de minimis rule, the more liberal the RoO 24 Thus it is considered the most flexible cumulation scheme, although it has not been largely implemented. 283 Handbook of Deep Trade Agreements regime. Overall, the average de minimis percentage is roughly the same across types of PTAs, though it is significantly lower for intra-PTAs in Latin America.25 The absorption rule allows the use of even more non-originating input. It is used across PTAs, actually less in EAP and LAC than in other regions, and with significant differences in Latin America between intra- and extra-regional PTAs. Between de minimis and absorption, de minimis tests are the more liberal option because they provide the greatest scope for raising the level of originating content. Table 9.5: Provisions on cumulation Quantitative (whom to cumulate) Intra PTA Inter PTA Intra PTA Inter PTA Average Absorption Qualitative (how to cumulate) Same PSR/OR Same PSR/OR Same PSR/ALL Any PSR/ALL de minimis (%) rule Agreements NO Bilateral Diagonal Full Cross INTRA LAC 7% 88% 3% 2% 0% 8,5 14% INTRA EAP 0% 71% 10% 19% 0% 10,0 24% LAC with EXT 0% 90% 5% 7% 5% 9,9 51% EAP with EXT 15% 78% 0% 7% 0% 9,8 28% Others 0% 91% 38% 19% 1% 10,1 71% Source: Deep Trade Agreements Database. Note: Others = PTAs between non-LAC or non-EAP countries. OR: cumulation for originating goods only; ALL: cumulation for all goods; Same PSR: cumulation allowed if same PSR applies; Any PSR: cumulation allowed even if PSR differs; Intra PTA: cumulation allowed for import from other partners in the PTA; Inter PTA: cumulation allowed even for input from countries out the PTA but in a PTA with one of the members. Box 9.3: Case study 2 - The facilitating role of cumulation schemes in Mexico and Central America At the end of the 1990s, Mexico signed agreements with Costa Rica (1995), Nicaragua (1998), and the Northern Triangle of El Salvador, Guatemala, and Honduras (2001). These agreements did not allow cumulation among the six countries, which had the negative effect of segmenting the value chains. For example, exports of chocolates from Costa Rica would not face tariffs when imported into Mexico as long as they were produced entirely in Costa Rica, but the same chocolates would face a tariff if they used cocoa paste from Honduras. In 2011, the six countries signed a new agreement that enabled full cumulation under a single set of RoOs. This gave firms more flexibility in sourcing their inputs. This example shows that a more systematic approach to RoOs may be needed and that promoting more cumulation of origin across the many bilateral and regional trade agreements in Latin American could have promising results. This is the approach followed by the Pacific Alliance, in which each member must have bilateral agreements with all other Alliance members as a prerequisite for membership, with the objective of full harmonization. Firms can take advantage of the differences in input prices across locations, resulting in more cross-border production within the region and higher competitiveness of their products. Source: Blyde 2014. 25 The threshold is generally established in weight for textile and apparel products, and it often does not cover agricultural products. 284 Rules of Origin 9.4.2.3 The burden of calculating VC Calculating value content to determine origin is a complex undertaking, and some methods are more complex than others, depending on (1) the formulation of regional value content (regional content or import content); (2) the method of estimating regional content (build-up or build-down); and (3) the valuation of the non-originating material (price basis). These conditions have several broad implications: A percentage criterion based on regional content is considered more restrictive than a specification based on the imported content because it may facilitate manipulation and add to compliance and administration costs.26 The ex-work cost (or factory cost) build-up affords the least flexibility to businesses in making input choices. It is also administratively more complicated because eligible expenditures are aggregated (i.e., built up, from component expenditures). This method is treated as most restrictive. The transaction value method (FOB) affords the most flexibility to firms in making their input choices. Since the method permits the producer to count all of its costs and profit as territorial, the required percentage of regional value content under this method is higher than under the net cost method. The net cost method generally requires a lower threshold than the build-down method, as it takes into account only costs that are directly related to the elaboration of the good. However, it represents a higher administrative cost, as detailed records of the expenditure for each product have to be kept by the producer (see Box 9.4). Regimes that allow traders to choose the cost basis they apply are considered least restrictive. It is clear from Table 9.6 that import content - the most flexible calculation method - is widely used in other parts of the world but less in Latin American and even more rarely in East Asian PTAs, which rely more on regional content. The restrictiveness of such value content rule lies more in the percentage, which can be same for all products covered by the PTA or be product specific. 26 Bearing in mind that the restrictiveness lies in the percentage attached to the VC rule, a RoO using the import content method will be more liberal the higher the maximum percentage allowed for non-originating materials. By the same token, an RoO using the regional content method will be more liberal the lower the minimum requirement for domestic content. 285 Handbook of Deep Trade Agreements Table 9.6: Provisions on calculation of value content Agreements Calculation method Price basis % product (Regional content vs. import content) specific RC RC RC Import Regional and FOB Fob/Net Ex works Ex works Build down Build up Both content import content price cost INTRA LAC 38% 0% 9% 43% 0% 71% 14% 7% 0% 52% INTRA EAP 43% 10% 24% 5% 14% 95% 0% 0% 0% 48% LAC with EXT 29% 12% 17% 39% 0% 59% 12% 27% 0% 85% EAP with EXT 27% 16% 22% 13% 1% 66% 0% 12% 3% 63% Others 10% 4% 9% 73% 1% 22% 2% 73% 0% Source: Deep Trade Agreements Database. Note: Others = PTAs between non-LAC or non-EAP countries. Box 9.4: Case study 3 - Net cost method and traced material list in automotive goods Rules of origin for automotive goods tend to be much discussed during trade negotiations. This is not surprising, given the implications of such rules for the competitiveness of this important manufacturing sector, as well as for the steel, glass, auto parts, and other related industries. The net cost methodology of determining origin is often preferred by auto-producing countries, as it presents a more precise value of the merchandise by not including costs that are not directly related to the production of the good. The nature of global value chains, and the high degree of integration in some regional blocs, have enabled these regions to become much more competitive vis-à-vis other auto-producing countries. Given the strong impact rules of origin can have in the industry, they can sometimes become very complex, such as the inclusion of tracing requirement which is used exclusively for the automobile sector, as in the following example: Most light vehicles for the transportation of persons are classified under heading 8703, and the applicable rule of origin is: 8703.21-8703.9042: A change to subheading 8703.21 through 8703.90 from any other heading, provided there is a regional value content of not less than 62.5 percent under the net cost method. Additionally, under the regime-wide rules of origin, it is specified that for automotive products, the value of the non-originating materials will be limited to the articles in the list of Traced Materials. Tracing guarantees a more accurate measure of the regional value content by tracking the non- originating components of each subassembly incorporated into the final product, which would avoid a “roll-up” whereby a product is counted as strictly originating or strictly non-originating regardless if it is a mixture of both originating and non-originating components. The Traced Materials list contains the major components of the vehicle such as engines (8407), electric motors (8501), car seats (9401), rubber hoses (4009), and so on. A car would therefore comply with the RVC requirement if the overall non-originating value (that is, the components brought from a non-Party) of the Traced Materials does not exceed 37.5 percent of the net cost of the vehicle in this particular example. Any non-originating component not included in the list (such as nuts and bolts, GPS systems, Bluetooth) can be incorporated without negatively affecting the RVC content of the vehicle. 286 Rules of Origin Table 9.7: Provisions on other aspects Drawback Fungible Advance rulings Transshipment rule Review (appeal) Agreements Prohibited Allowed INTRA LAC 2% 2% 50% 74% 21% 24% INTRA EAP 0% 5% 81% 76% 33% 33% LAC with EXT 15% 0% 90% 76% 66% 59% EAP with EXT 6% 1% 72% 61% 54% 37% Others 58% 0% 64% 89% 61% 51% Source: Deep Trade Agreements Database. Note: Others = PTAs between non-LAC or non-EAP countries. 9.4.2.4 Additional flexibility mechanisms There are also other provisions that can provide more flexibility to the RoO regime. Duty drawback schemes selectively lower the cost of inputs used to produce goods for export and can be especially important for the sourcing of intermediate goods outside the regional trading area. Indeed, when non-originating materials are used in the production of a final product that is geared for export, the duty drawback provision basically provides a waiving or a repayment of duties applicable to the non-originating materials used. Nevertheless, in order to discourage the use of non-originating materials in production processes, several PTAs do not allow duty drawback, raising the cost of exporting to member economies and encouraging firms to purchase inputs from potentially higher-cost local sources. Allowing fungible goods/materials can also lower the cost of inputs by allowing for the use of accounting methods that trace the origin of the fungible inputs without the obligation to keep originating and non-originating materials physically separated.27 Most of the PTAs in LAC and EAP offer this option, though it is less present in intra-regional PTAs for Latin America. Advance rulings are considered a highly efficient tool to ensure the proper implementation and application of administrative procedures. Among other benefits, advance rulings provide transparency in the laws, regulations, and practices regarding RoOs.They are present in two- thirds of the PTAs in LAC and EAP. Review and appeal provisions grant the right to request a second review of decisions by the Customs authority and therefore add transparency and accountability to the RoO process. They are not yet very widespread in the PTAs of LAC and EAP, although many PTAs in LAC with an extra-regional partner have adopted this scheme. 27 Some PTAs require a specific authorization from the Customs authorities for accounting segregation, and limit the application of this method to such cases where maintaining physical segregation would result in considerable costs or material difficulties for the producer. 287 Handbook of Deep Trade Agreements From the discussion above, one can tentatively construct an index of restrictiveness in regime- wide RoOs to then observe the evolution of such restrictiveness over time. The methods for calculating the restrictiveness score are presented in the annex. Based on the different provisions in PTAs and on those scores, the figure below displays the evolution since 1990 of the average restrictiveness across all existing PTAs at a certain year for North-North, South- South and North-South agreements (Figure 9.3). The regime-wide RoOs appear to be more restrictive in South-South PTAs, though restrictiveness significantly decreases over time. In parallel, the level of restrictiveness in North-South PTAs tends to converge with that in North-North PTAs. Figure 9.3: Index of restrictiveness for regime wide rules of origin over time Source: Deep Trade Agreements Database. Note: See Annex Table 9.A.2 for explanation of index calculations. 9.5. CONCLUSIONS Rules of origin give meaning to preferential trade agreements and are their vital element. RoOs are static administrative rules according to which a product imported from a PTA partner can be deemed eligible for preferential market access. With increasing fragmentation of production in global value chains, however, and with dynamic technological and business developments that often require location and sourcing adjustments in order to stay competitive, RoOs - instead of helping deliver the gains from PTAs - may actually undo their benefits. That this may indeed be the case is strongly suggested by the coexistence of different product-specific RoOs (PSR) for the same products across the different PTAs, but often also across the different PTAs signed by one country. There are also important differences in regime-wide RoOs, which set conditions for cumulation across different PTAs 288 Rules of Origin as well as for administrative procedures that must be met to prove origin. It is for this reason that there is a strong interest in learning more about which RoOs are the least restrictive and thus most suitable for ensuring gains from PTAs in the increasingly complex global economy. Classification of the complex rules and the availability of adequate data are prerequisites for the required empirical analysis. Building on the existing literature, this chapter describes a new classification of rules of origin that has been developed for this purpose and applied to the RoO regimes of 250 existing PTAs.The methodology follows a questionnaire-type approach, in which approximately fifty questions with either binary or multiple detailed answers are devised to classify product- specific and regime-wide rules. Using this approach, product-specific rules are classified into 40 different categories and regime-wide rules into a further 30 categories. To show how the new taxonomy and the collected data can be used, the chapter surveys the RoO regimes of a number of PTAs signed by Latin American and East Asian countries, both within the region and with extra-regional partners. Drawing on historical background of how the RoOs in these regions were negotiated, the chapter uses the RoO data collected to reveal differences in these types of PTAs, and show how the different families of PTAs have influenced each other and shaped RoOs seen today. Some interesting similarities can be seen between the two regions, with, for example, both East Asian and Latin American agreements tending to allow more alternative rules and fewer exceptions, which suggests a recognition of need for flexibility. Agreements in both these regions also trend to include more technical regulation rules, in their intra-regional agreements as compared to extra- regional agreements. Both regions also tend to have more restrictive regime-wide rules on, for example, certification or document validity periods in their intra-regional agreements, reflecting the possibility that PTAs between relatively less developed regional trading partners may be more restrictive to compensate for the lack of trust and appropriate institutions. An example of an interesting difference is that East Asian agreements generally have more flexible cumulation rules compared to agreements in Latin America. This taxonomy and these data can readily be used to empirically investigate the effects of PTAs and RoOs in these and other regions, and it is hoped that they will be useful building blocks for the empirical and policy analysis going forward. ACKNOWLEDGMENTS This chapter has benefited from the excellent contributions of Anmol Kaur and Sebastian Vallejo. We are also grateful to Charles Cadestin, Clarisse Legendre, and Emanuele Mazzini for important data analysis assistance and to Darlan Marti (WTO) for the information on the WTO agreement on RoOs. 289 Handbook of Deep Trade Agreements REFERENCES Abreu, M.D. 2016. Preferential rules of origin in regional Harris, J. 2007. Measurement and determination of rules of trade agreements. In Regional trade agreements and the origin in preferential trade agreements (PTAs). Ph.D. thesis, multilateral trading system, edited by R. Acharya. Cambridge: University of Maryland, College Park. http://hdl.handle. Cambridge University Press. net/1903/7180. Augier, P., M. Gasiorek, and C.L. Tong. 2005. The impact of Inama, S. 2011a.The reform of the EC GSP rules of origin: Per rules of origin on trade flows. Economic Policy 20(43). aspera ad astra? Journal of World Trade 45 (3): 577-603. Australian Productivity Commission. 2004. Restrictiveness ———. 2011b. Rules of origin in international trade. Cambridge: index for preferential rules of origin. Supplement to Productivity Cambridge University Press. Commission Research Report, Rules of origin under the Australia–New Zealand Closer Economic Relations Trade Kelleher, Sinéad. 2012. Playing by the rules? The Agreement. Productivity Commission, Canberra, June. development of an amended index to measure the impact of rules of origin on intra-PTA trade flows. Working Paper Bureau, Jean-Christophe, Raja Chakir, and Jacques Gallezot. Series, UCD Centre for Economic Research, University 2007. The utilization of trade preferences for developing College Dublin, Dublin. countries in the agri-food sector. Journal of Agricultural Economics 58(2): 175–98. Manchin, Miriam. 2006. Preference utilization and tariff reduction in EU imports from ACP countries. World Blyde, J. S. 2014. Synchronized factories: Latin America and the Economy 29(9): 1243–66. Caribbean in the era of global value chains.New York: Springer. Park, I., and S. Park. 2009. Consolidation and harmonization Cadestin, C., J. Gourdon, and P. Kowalski. 2016. Participation of regional trade agreements (PTAs): A path toward global in global value chains in Latin America: Implications for free trade. MPRA Paper 14217, University Library of trade and trade-related policy. OECD Trade Policy Papers Munich, Munich. 192.http://dx.doi.org/10.1787/5jlpq80ts8f2-en. World Customs Organization. 2014. Comparative study on Cadot, O., C. Carrere, J. de Melo, and B.Tumurchudur. 2006. certification of origin. Product-specific rules of origin in EU and US preferential trading arrangements: An assessment. In The origin of goods: A conceptual and empirical assessment of rules of origin in PTAs, edited by O. Cadot, A. Estevadeordal, A. Suwa-Eisenmann, and T.Verdier. Oxford: Oxford University Press. Conconi, P., M. Garcia-Santana, L. Puccio, and R.Venturini. 2018. From final goods to inputs: The protectionist effect of rules of origin. American Economic Review 108 (8): 2335-65. Estevadeordal, A. 2000. Negotiating preferential market access. Journal of World Trade 34(1): 141–66. Estevadeordal, A., J. Harris, and K. Suominen. 2007. Multilateralizing preferential rules of origin around the world. Washington, DC: Inter-American Development Bank. Estevadeordal, A., and K. Suominen. 2006. Rules of origin: A world map. In The origin of goods: A conceptual and empirical assessment of rules of origin in PTAs, edited by O. Cadot, A. Estevadeordal, A. Suwa-Eisenmann, and T. Verdier. Oxford: Oxford University Press. Estevadeordal, A., and K. Suominen (eds.). 2008. Gatekeepers of global commerce: Rules of origin and international economic integration. Washington, DC: Inter-American Development Bank. 290 Rules of Origin ANNEX Annex Table 9.A.1: List of codes and definitions Variable Question asked Area: PSR criteria SR_psr Does the agreement contain product-specific rules of origin? SR_who Is the product’s origin defined as wholly obtained? SR_stc Is the product’s origin defined through substantial transformation criteria? Area: Methods of estimating value content SR_vcr Is the product’s origin defined through a value content requirement? SR_vcr_cal Are there different ways of calculating value content requirement? SR_vrc_meth1 What is Method 1 for calculating VC? SR_vrc_perc1 What is the percentage of value content required under Method 1? SR_vrc_meth2 What is Method 2 for calculating VC? SR_vrc_perc2 What is the percentage of value content required under Method 2? Area: Methods of change in tariff classification SR_ctc Is the product’s origin defined through a change in tariff classification? SR_cc Is the product’s origin defined through a change in chapter? SR_ch Is the product’s origin defined through a change in heading? SR_cs Is the product’s origin defined through a change in subheading? Area: Other SR_tr Is the product’s origin defined through a technical requirement? SR_cum Do two or more origin criteria apply cumulatively? SR_alt Do two or more origin criteria apply alternatively? SR_ctc_exc Are one or more HS codes or product groups explicitly excluded from being used as inputs for originating goods? Area: Certification SR_cer_sel Can the certificate be issued on the basis of self-certification by the exporter/producer/importer without need for authentication by the competent authority? SR_cer_adm Does the certificate have to be issued by competent authorities of the exporting party, including customs administrations, other government authorities, and designated private ones? SR_cer_two Is there a possibility to combine self-certification with administrative certification? SR_cer_val What is the length of the validity period for the certificate of origin? SR_cer_rec What is the length of the record keeping period? SR_cer_ex1 Is there a certificate exemption? SR_cer_ex2 What is the threshold for exemption in $US? SR_cer_err Is there a possibility to amend minor errors? Area: Verification SR_ver_dir Is there a direct verification of the certificate? SR_ver_ind Is there an indirect verification of the certificate? SR_ver_two Is there a combined verification of the certificate? Area: Cumulation SR_cum_bil Does the agreement allow for bilateral or partial cumulation? SR_cum_dia Does the agreement allow for diagonal cumulation? SR_cum_ful Does the agreement allow for full cumulation? SR_cum_cro Does the agreement allow for cross cumulation? SR_cum_dm1 Does the agreement contain de minimis provisions? SR_cum_dm2 What is the de minimis percentage? SR_cum_abs Does the agreement include absorption provisions? Area: Methods of estimating value content SR_vcr_rec Is the value content requirement calculated through regional content? SR_vcr_rbd Is the value content requirement calculated as a minimum regional content requirement using a build-down calculation? SR_vcr_rbu Is the value content requirement calculated as a minimum regional content requirement using a build-up calculation? SR_vcr_alt Is more than one calculation method permitted to determine the Regional Value Content? SR_vcr_imc Is the value content requirement calculated through import content? SR_vcr_ric Is the value content requirement calculated through both regional and import content? SR_vcr_cst Is the price basis for the content threshold requirement the ex-works cost? SR_vcr_fnt Is the price basis for the content threshold requirement the FOB/net price? SR_vcr_prc Is the price basis for the content threshold requirement the ex-works price? SR_vcr_fob Is the price basis for the content threshold requirement the FOB (free on board) price? SR_vcr_per What is the percentage of value content required? SR_vcr_per2 What is the percentage of value content required with alt method? Area: Other aspects SR_drb Does the agreement contain drawback rules? SR_dba Does the agreement allow drawback? SR_dbp Does the agreement prohibit drawback? SR_fng Does the agreement allow for joint storage of originating and non-originating inputs when these inputs are interchangeable? SR_adr Does the agreement allow for advance rulings? SR_trs Does the agreement contain a transhipment rule? SR_rev Does the agreement contain a specific review and appeal mechanisms? 291 Handbook of Deep Trade Agreements Annex Table 9.A.2: Index for restrictiveness in regime-wide provisions (1: more restrictive, 0: less restrictive) Provisions types Categories Corresponding Index of restrictiveness Certification body Self/Administration/Both 0/1/1 Validity period Less 1 year/ 1 year/ more than 1 year 1 / 0.5 / 0 Record keeping period Less 4 years/ 4 years/ more than 4 years 0 / 0.5 / 1 Exemption No / <1000 $US / >1000$US 1 / 0 / 0.5 Minor errors Yes/No 0/1 Verification Indirect/Direct 0/1 Cumulation type No/ bilateral/ diagonal/ Full/ Cross 1 / 0.8 / 0.5 / 0.5 / 0 De minimis No / 7% / >10% 1 / 0.5 / 0 Absorption rule No/Yes 1/0 Duty drawbacks Not mentioned /Prohibited/ Allowed 0.5 / 1 / 0 Fungible goods No/Yes 1/0 Advance ruling No/Yes 1/0 292 Trade Facilitation and Customs CHAPTER 10 Trade Facilitation and Customs E. Kieck 293 Handbook of Deep Trade Agreements CHAPTER 10 Trade Facilitation and Customs E. Kieck World Bank, Washington, DC, United States CONTENTS 10.1. INTRODUCTION 295 10.2. TRADE FACILITATION DEFINED 296 10.3. METHODOLOGY 297 10.4. TRENDS AND PATTERNS 300 10.4.1 General remarks 300 10.4.2 Internet publication 307 10.4.3 Prior publication and opportunity to comment 307 10.4.4 Designation of enquiry points 308 10.4.5 Advance rulings 308 10.4.6 Appeal or review 309 10.4.7 Disciplines on fees and charges, and penalties 309 10.4.8 Release and clearance 309 10.4.9 Border agency cooperation 310 10.4.10 Formalities 310 10.4.11 Transit 310 10.4.12 Exchange of information 310 10.4.13 Customs unions 311 10.4.14 Origin administration 312 10.4.15 Technical assistance and capacity building 314 10.4.16 Institutional 314 10.5. CONCLUSIONS 315 ACKNOWLEDGMENTS 316 REFERENCES 317 294 Trade Facilitation and Customs 10.1. INTRODUCTION The rapid growth in international trade over the past few decades is largely the result of technological developments and trade liberalization efforts at the multilateral, regional, bilateral, and national levels.1 These efforts have mainly focused on the reduction of tariffs to promote economic development, although “… progress in trade facilitation is still slow in many countries - and [has been] hampered by high costs and administrative difficulties at the border.”2 In managing the movement of goods across their borders, countries apply controls that serve various public policy aims. These include policy objectives such as collecting duties and taxes, protecting the economy from illicit trade practices, and safeguarding society and the environment from dangerous goods.3 These controls are undertaken by various regulatory, fiscal, and border control agencies and are often outdated, with overly bureaucratic clearance processes that pose greater barriers to trade than tariffs.4 These constraints are often exacerbated by the uneven use of technology and procedures by agencies where some have automated systems and apply modern techniques such as risk management whilst others are paper-based and apply transactional, “total control/inspection” approaches. Inefficient organizational processes and weak administrative capacity, combined with little or no interagency coordination and inadequate infrastructure and equipment, increase compliance costs and add to delays and unpredictability when moving goods across borders. As a result, countries have increasingly started to focus on facilitating legitimate trade through national reforms and international trade negotiations. At the multilateral level, the Trade Facilitation Agreement (TFA) of the World Trade Organization (WTO) entered into force on February 22, 2017, following years of negotiations. The first multilateral trade agreement to be finalized since the establishment of the WTO in 1995, the TFA builds on the provisions in the General Agreement on Tariffs and Trade (GATT) related to freedom of transit, fees, and formalities, and the administration of trade regulations (Articles V,VIII, and X of GATT 1994). In addition to the multilateral negotiations at the WTO, countries have also included trade facilitation provisions in their regional and bilateral trade agreements. These agreements commonly take the form of free trade agreements and customs union agreements. 1 IMF 2001. 2 McLinden et al. 2011. 3 World Customs Organization 2008. 4 Mustra 2011. 295 Handbook of Deep Trade Agreements 10.2. TRADE FACILITATION DEFINED According to the OECD, “trade facilitation covers all the steps that can be taken to smooth and facilitate the flow of trade … including product testing and impediments to labor mobility …”5 For this chapter, the WTO definition of trade facilitation will be used; namely, the simplification, modernization, and harmonization of export and import processes. The commonly accepted aim of trade facilitation is to “simplify and streamline international trade procedures to allow the easier flow of trade across borders and thereby reduce the costs of trade.”6 Countries introduce trade facilitation reforms to achieve various policy goals. These include attracting investment and manufacturing to create jobs; reducing trade costs for importers, exporters, and consumers of goods; and participating in global value chains. According to the WTO, full implementation of the TFA will reduce global trade costs by an average of 14.3 percent and will result in export gains of between US$750 billion and US$1 trillion per annum, depending on a number of factors.7 Trade facilitation is often associated with the activities of a national customs administration.The central role of Customs is recognized in the TFA, and most provisions in Section I of the TFA deal with customs matters. The TFA also recognizes that other government agencies, both at and away from the border, have an impact on international trade, and therefore introduces the concept of Border Agency Cooperation in Article 8, and the requirement to review formalities and documentation in Article 10. There are also other articles covering other regulatory and border agencies. In Article 4, for example, paragraphs 1 to 5 cover procedures for appeal or review applicable to Customs, and paragraph 6 encourages Members to apply the provisions of this Article to administrative decisions of “… a relevant border agency other than Customs.” This chapter focuses on trade facilitation provisions in PTAs in the broader sense (more than customs) as well as provisions in PTAs that relate to customs matters. Prior to the emergence of the current concept and understanding of trade facilitation, contracting parties to PTAs included customs-related provisions in their PTAs, and these tended to focus on two main areas that were more focused on compliance than facilitation: the administration of preferential rules of origin (mainly through the issuing and processing of certificates of origin) to ensure that only qualifying goods receive preferential tariff treatment; and mutual administrative assistance to support customs enforcement. The TFA contains measures aimed at both facilitating trade (Articles 1 to 11) and promoting compliance and customs cooperation (Article 12). 5 OECD 2005. 6 Congressional Research Service 2017. 7 WTO 2015.Values for full implementation of trade data were applied for years 2003-2011. 296 Trade Facilitation and Customs 10.3. METHODOLOGY During negotiations and before its entry into force, the TFA had already influenced the inclusion of trade facilitation provisions in PTAs,8 and is continuing to be used by PTA negotiators to guide them on trade facilitation elements and content. This is resulting in greater convergence between the TFA and the trade facilitation provisions of PTAs. The template for the trade facilitation dataset was therefore developed based largely on the TFA structure and elements, but as the dataset covers the trade facilitation provisions of PTAs, additional elements were included. For the elements related to the TFA, the template was largely informed by a working paper of the WTO Secretariat that reviews 217 preferential trade agreements and analyzes their trade facilitation provisions compared to the TFA, with specific reference to parallels, additions, and overlaps.9 The same paper was used to verify the findings of the dataset. In addition to the WTO Secretariat working paper, the World Customs Organization (WCO) also produced a research paper that specifically identifies trends and patterns of customs-related trade facilitation measures in recent PTAs.10 In addition, two other studies, by UNCTAD and OECD, were also considered in preparing the template.11 The template covers most of the elements of the TFA and, in some instances, goes into more detail than provided for in the TFA. For example, in relation to the TFA’s single window provisions, the template adds two additional elements - the inter-operability of single window systems and the establishment of a common single window system. The template consists of the following eight main sections, most of which are divided into subsections: I. Transparency A. Publication and availability of information B. Opportunity to comment, information before entry into force, and consultations C. Advance rulings D. Procedures for appeal or review 8 WTO 2014. Also, the WTO defines PTAs as reciprocal trade agreements between two or more partners, including free trade agreements and customs union agreements. 9 WTO 2014. 10 WCO 2014. 11 UNCTAD 2011; OECD 2002. 297 Handbook of Deep Trade Agreements II. Fees and formalities A. Disciplines on fees and charges imposed on or in connection with importation and exportation and penalties B. Release and clearance of goods C. Border agency cooperation D. Movement of goods intended for import under Customs control E. Formalities connected with importation, exportation, and transit III. Transit A. Freedom of transit IV. Customs and other forms of trade facilitation cooperation A. Exchange of information B. Other V. Customs union specific VI. FTA specific VII. Technical assistance and capacity building VIII. Institutional arrangements For each section or subsection, the template lists specific elements. These are then unpacked into individual items (bullet points), where relevant, in order to identify the scope of an element and the commitment of the parties in relation to a specific element or item. Sections V and VI of the template deal with trade facilitation provisions that are specific to the two main types of PTAs - customs unions and free trade areas - and are not covered in the TFA. These sections have been included since they cover elements that have a non-tariff impact on (a) the movement of goods across borders, and (b) the customs and border management activities of the parties to the PTAs. For customs union agreements, the template focuses on three issues that are aimed at forming an impression of the level of integration of a customs union agreement, with the aim of determining how each issue contributes to the facilitation of trade. These issues are (a) the legal arrangements agreed to by the parties to apply a common external tariff and regulate other customs and cross-border trade matters (e.g., a supranational customs legal framework, as in the European Union and the East African Community, or separate national customs laws, as in the Southern African Customs Union); (b) the 298 Trade Facilitation and Customs point of collection of customs duties and taxes (at port of entry into the union or the final destination); and (c) customs revenue arrangements (does each party retain what it collects, or are customs a source of revenue for the union?). When looking at these issues in a customs union agreement, an impression can be formed about the level of integration envisaged by the contracting parties. For free trade agreements, the template focuses on origin administration. The actual preferential rules of origin are not covered, as these, similar to customs duties, are related to trade, industrial, and fiscal policy matters. According to the International Chamber of Commerce (ICC), “… origin requirement procedures (e.g., supplier declarations) linked to PTAs are starting to form a behind- the-border barrier to trade.”12 For this reason, the actual administration of the rules of origin, such as the requirement to obtain and present a certificate of origin to customs, is closely related to trade facilitation and covered in the template and dataset. From a trade facilitation perspective, the elements of the FTA specific section aim to provide a sense of whether a PTA uses the traditional origin administration model13 or contains provisions that aim to reduce the administrative complexities associated with rules of origin and facilitate trade in goods. The latter include measures such as (a) using a commercial invoice declaration; (b) not requiring the issuing/endorsement of a certificate of origin by authorities in the exporting country or the submission of proof of origin to the Customs authority at import unless requested; and (c) waivers, approved exporter schemes and self-certification. The template focuses on core trade facilitation and customs issues and does not include the following, even though they may directly or indirectly impact on the movement of goods across borders:14 • sanitary and phytosanitary measures and technical barriers to trade; • common external tariff or tariff reduction issues; • elements that have implications for customs control or trade facilitation, such as intellectual property rights enforcement, electronic commerce as well as prohibitions and restrictions; and • legal and process-related provisions, such as dispute settlement, arbitration, and notification, which may also apply to trade facilitation provisions. The WTO Secretariat paper notes that an analysis of trade facilitation provisions in PTAs is hampered by the absence of consistent trade facilitation terminology. The paper also mentions that some of the above-mentioned excluded issues “… are considered part of the TF chapter in several agreements whereas they are treated in separate sections elsewhere.”15 12 ICC 2017. 13 This usually requires that a paper certificate of origin has to be issued/certified by the exporting customs administration or another authority and submitted as part of the customs declaration to the importing customs administration on a transactional basis. 14 These issues are also covered by other WTO agreements. 15 WTO 2014b. 299 Handbook of Deep Trade Agreements 10.4. TRENDS AND PATTERNS 10.4.1 General remarks The inclusion of trade facilitation provisions in PTAs notified to the WTO has evolved over time. Earlier PTAs had no or narrow trade facilitation provisions, but over time the inclusion and the range of TF provisions in PTAs have expanded (Figure 10.1).16 The solid black line is the average number of TF provisions in PTAs, and the dots refer to specific PTAs that are above average for a period. Figure 10.1: Average coverage ratio of TF provisions in new PTAs over time Source: Deep Trade Agreements Database Note: TF = trade facilitation. Solid black line shows average number of TF provisions in PTAs; dots refer to specific PTAs that are above average for a period. The WTO TFA negotiations have influenced the negotiation of trade facilitation provisions in PTAs. In fact, “… it can be reasonably assumed that many governments tended to implement certain TF measures negotiated at the WTO in bilateral or regional domains….”17 16 WTO 2014b; Congressional Research Service 2017. 17 WCO 2014b. 300 Trade Facilitation and Customs The TFA negotiations began after the adoption by the WTO’s General Council of the “July 2004 package,” which mandated negotiations on trade facilitation based on the modalities set out in Annex D of the General Council decision.18 Of the 267 PTAs reviewed for this chapter, the vast majority (260) have at least one trade facilitation provision. The EU-Republic of Korea agreement contains the highest number of TF provisions, at 36. Overall, PTAs concluded by the EU, EFTA, Canada, the US, Korea, Chile, and Peru tend to have more TF provisions than others. All the PTAs with 30 or more trade facilitation provisions entered into force after 2005. Those that entered into force after 2010 have an average of 21.9 trade facilitation provisions, compared to an average of 14.5 provisions for PTAs concluded between 2005 and 2009, 9.6 for those concluded between 2000 and 2004, and 6.8 for those between 1995 and 1999. Among the specific provisions used in the template, the most and least common are shown in Figures 10.2 and 10.3. Figure 10.2: Trade facilitation provisions with highest frequency Source: Deep Trade Agreements Database. Note: TF = trade facilitation. 18 Decision Adopted by the General Council on 1 August 2004, WT/L/579 dated 2 August 2004. 301 Handbook of Deep Trade Agreements Figure 10.3: Trade facilitation provisions with lowest frequency Source: Deep Trade Agreements Database. In the TFA, some of the most common category C measures notified by countries include the single window, authorized operators, risk management, internet publication, border agency cooperation, and advance rulings. These are more “complex provisions” because many WTO members require technical assistance and capacity building to implement them. The figure below demonstrates the extent to which they have been incorporated into PTAs (Figure 10.4). Figure 10.4: Share of PTAs with more “complex” provisions Source: Deep Trade Agreements Database. Note: Figures at bottom of each bar represent the number of PTAs with complex provisions. The past few decades has seen an increase not only in the number of trade facilitation provisions included in PTAs, but also in the diversity of countries that include them.The average number of trade facilitation provisions in PTAs by contracting group is shown in Figure 10.5. 302 Trade Facilitation and Customs Figure 10.5: Average number of trade facilitation provisions in PTAs, by level of development Most of the trade facilitation provisions found in PTAs relate to customs matters, but this is changing. There appears to be a recognition of the need to also include the activities of other agencies that impact trade. These include enforcement agencies at borders, such as the police, standards, veterinary, and phytosanitary authorities, as well as agencies that regulate cross-border trade through permits, licenses, and certificates. Trade facilitation provisions are found either in the general text of an agreement (usually in chapters dealing with trade in goods and rules of origin) or in a separate chapter, annex, or appendix. The trend in recent PTAs is to deal with trade facilitation separately.19 For the United States, earlier PTAs such as the US-Chile agreement contained a chapter on “Customs Administration” while more recent agreements such as US-Korea contain a chapter on “Customs Administration and Trade Facilitation.” The first EFTA free trade agreements contained customs provisions in the context of rules of origin and mutual administrative assistance; however, most EFTA agreements concluded after 2008 contain an annex on all aspects of trade facilitation.20 This trend is most likely the result of a combination of triggers such as the commencement of TFA negotiations and the recognition that measures other than tariffs and rules of origin also require special attention to expand trade and reduce costs between the contracting parties. Some of the trade facilitation provisions that cover more than customs matters are the requirement to publish all trade-related laws and regulations on the internet, coordination among border 19 WCO 2014b. 20 An exception is EFTA-Peru, signed in 2010, which contains an annex on customs procedures and trade facilitation. 303 Handbook of Deep Trade Agreements management agencies, and electronic trade single window provisions. As of 2017, 114 PTAs contain at least one of the three provisions, and 9 contain all three provisions (Figure 10.6). Figure 10.6: Key provisions other than customs, over time Source: Deep Trade Agreements Database. Most PTAs that commit the parties to use specific international standards refer to WCO instruments such as the Revised Kyoto Convention (RKC),21 the SAFE Framework of Standards, and the Harmonized System (HS).22 The influence of the World Customs Organization can also be seen in references to “authorized economic operator” (AEO) in some PTAs, as opposed to the “authorized operator” concept of the TFA. As a matter of interest, national customs laws, in addition to these two expressions, also use concepts such as “trusted trader” or “preferred trader.”23 Some trade facilitation measures, such as those relating to fees and charges, are dealt with in a very similar manner in different PTAs. This demonstrates the influence of longstanding GATT obligations (such as Article VIII) in informing the commitments of the contracting parties. In other cases, the same issue is dealt with very differently in different PTAs. In the case of expedited shipments, for example, the provisions of PTAs tend to vary.24 This can be attributed to the absence of these provisions in the WTO before the TFA entered into force. This variability contributes to the complexity 21 International Convention on the Simplification and Harmonization of Customs Procedures (as amended). 22 Harmonized Commodity Description and Coding System. 23 These types of designations generally aim to reward high levels of compliance (or compliance with the country’s standards), although the requirements and benefits may differ. 24 WCO 2014b. 304 Trade Facilitation and Customs of analyzing trade facilitation provisions across PTAs. Therefore, the actual content of the commitments, and not just the headings, should be scrutinized. All of this is further complicated by the depth and specificity of the commitments made by contracting parties on the same measure across PTAs. An interesting feature of trade facilitation provisions in PTAs is that they are mostly non- discriminatory.25 They tend to benefit not only the contracting parties but also third parties, mainly because of the impracticality of maintaining two (or more) separate trade facilitation regimes. For example, it makes little sense to apply risk management techniques only to goods imported from PTA contracting parties. There are, of course, notable exceptions such as the exchange of information for the purpose of mutual administrative assistance or the commitment to work towards the mutual recognition of each other’s AEO/AO arrangements. As with the TFA, PTA commitments on trade facilitation can be classified as either best endeavor or binding. In some instances, even the binding commitments are tempered with conditions (such as subject to national laws/available resources, provided all regulatory requirements are met or to be negotiated and agreed at a later stage). Any analysis should be careful to identify the actual commitments of the contracting parties. This is another complicating factor that hampers analysis. The trade facilitation provisions in some PTAs are general while others contain comprehensive and detailed commitments. Again, using the example of expedited shipments, PTAs to which the United States is a party tend to contain detailed provisions on expedited shipments,26 while PTAs to which the EU is a contracting party are more general.27 A further factor to keep in mind when analyzing the trade facilitation provisions of PTAs is that some PTAs commit the contracting parties to putting in place an indicative work program or empowering an institution or body such as a customs committee to develop trade facilitation measures. For example, the Southern African Customs Union Agreement (SACU) commits the parties to developing arrangements on customs cooperation in the form of annexes to the agreement. It is not always possible to locate these instruments as they are developed after the PTA enters into force. Some of these instruments, such as in 25 Duval et al. 2019, page 19. 26 Congressional Research Service 2017. 27 WTO 2014b. 305 Handbook of Deep Trade Agreements the SACU example, form part of the PTA once they are finalized (signed or ratified, as the case may be), but others may take the form of, for example, Memoranda of Understanding (MoUs) and not be, in treaty terms, part of the PTA. This is another factor that hampers a comprehensive analysis of trade facilitation provisions. The types of trade facilitation provisions in PTAs are determined by a range of factors. These include the overall aims and objectives of the contracting parties (including intentions on level of integration); the type of PTA; the number of contracting parties; the practical facilitation challenges that need to be addressed (for example, issues identified as bottlenecks by traders); and levels of development of the contracting parties. The WTO working paper elaborates on these factors.28 It is also necessary, when comparing the trade facilitation provisions of PTAs, to consider particular issues such as geography. By way of example, it may make sense for countries sharing a land border to agree to implement a one-stop border post. But where countries do not have a common border, then this will (most likely) not be found in the agreement. This equally applies to transit provisions. It has also been noted that customs union agreements mostly tend to be entered into by countries next to or in close proximity to each other.29 Related to this point is that the dataset should not only be looked at in a numerical sense (for example, PTA A contains 10 trade facilitation provisions while PTA B contains 5 provisions, and therefore PTA A is “deeper”). The variation in the number of provisions can be explained by some of the above-mentioned factors, such as geography. For a deeper analysis of trade facilitation provisions, there is a need to consider context, quality, level of commitment, and other factors. Some provisions make more of a difference than others. In other words, some trade facilitation provisions “weigh” more than others in a specific context and have a bigger impact on the actual facilitation of legitimate trade between the contracting parties. The challenge for contracting parties that are committed to a deep PTA is to identify the most important provisions and garner the political will and resources to implement them. These provisions are not easy to identify; therefore, it may be useful for the parties to agree to prior analyses such as time release studies and close consultation with importers, exporters, brokers, and logistics service providers. This will enable the contracting parties to identify particularly important provisions as well as the specificity required to address concerns and challenges. It is not always the most complex and expensive solutions that make the biggest difference; measures related to transparency are often regarded as more effective in reducing costs than measures related to fees and formalities.30 28 WTO 2014b. 29 Andriamananjara 2011. 30 Duval et al. 2016. 306 Trade Facilitation and Customs In the same vein, the extent to which provisions are binding and enforceable, and the implementation will and capacity of the parties, also determine the impact of trade facilitation measures, and are therefore also important factors in trade analysis. Enforceability generally depends on the inclusion of dispute settlement (DS) provisions in the PTA. The TFA is subject to the Dispute Settlement Understanding (DSU) of the WTO, which is stronger and more binding than dispute settlement provisions had been under the GATT. The TFA also creates a committee to review the operation and implementation of trade facilitation measures. In contrast, most PTAs do not have similar mechanisms to ensure implementation. The absence of such mechanisms may have resulted in an increased willingness of parties to commit to some TF measures, knowing there is little or no risk of a sanction. The rest of the chapter focuses on a number of selected trade facilitation provisions. 10.4.2 Internet publication Many PTAs initially contained provisions committing the parties to publish trade- related laws and procedures. Over time, this commitment was expanded to include publication either in print on the internet; and more recently, it has narrowed to publication only on the internet. Eighty-two PTAs - 55 North-North, 16 North- South, and 11 South-South - now have this internet-only provision. Figure 10.7: Share of agreements including provisions on internet publication by level of development Source: Deep Trade Agreements Database. Note: The figures listed at the bottom of each bar represent the number of PTAs with these provisions. 10.4.3 Prior publication and opportunity to comment The requirements to (a) consult stakeholders and provide opportunity to comment and (b) publish laws and regulations before implementation often appear together in PTAs such as the 307 Handbook of Deep Trade Agreements EU-Chile agreement. Other PTAs include a prior publication requirement but do not require consultation on laws and regulations. Still others that predate the TFA specifically refer to Article X of GATT 1994. This article not only deals with transparency but provides that rules can only be enforced if they were published prior to application.31 Article X does not require an opportunity to comment. The ASEAN-India Framework Agreement and Canada-Israel FTA are examples of PTAs that incorporate Article X by reference. These two requirements are also not always dealt with in a similar manner in PTAs. As noted above, most trade facilitation provisions in PTAs also benefit third parties, and this is the case when the contracting parties commit to publish trade-related requirements and provide an opportunity for comment before implementation. Article 73 of the China-Chile FTA illustrates an exception to this and provides that a party will provide a reasonable opportunity to the other party and interested persons of that contracting party to comment before implementation. 10.4.4 Designation of enquiry points The provisions of the TFA on enquiry points are broad in their coverage as they, amongst others, have to answer reasonable answer enquiries from “governments, traders, and other interested parties” on all matters mentioned in Article 1 (paragraph 1.1). In the PTAs that provide for enquiry points or contact points, their role is often limited to communication between the contracting parties; 76 PTAs now have this requirement. 10.4.5 Advance rulings Of the 267 PTAs in the World Bank’s Deep Integration database, 101 provide for advance rulings. Of these, the majority specify the matters on which advance rulings may be issued, and the language used tends to be binding. EFTA-Peru and Australia-Thailand take a narrow approach to advance rulings and provide for them on only one issue— tariff classification.32 Other PTAs, such as ASEAN-Australia-New Zealand, extend this requirement to all traditional customs areas; namely, tariff classification, customs valuation, and origin. A number of PTAs, while identifying the specific issues on which rulings may be issued, also create the space for the parties to agree on additional issues at a later date. In the case of US-Korea, a list of seven items is provided on which advance rulings can be requested, and other items can be added upon agreement of the parties. Of interest in this PTA is that the right to request an advance ruling is limited to importers of 31 UNECE, n.d. 32 However, the EFTA-Peru PTA commits the parties to endeavor to extend advance rulings to origin and others matters. 308 Trade Facilitation and Customs the importing party and exporters and producers of the exporting party. Similar approaches are followed in a number of other PTAs. In China-New Zealand, for example, “any person with a justifiable cause” may also request an advance ruling. This PTA also prescribes the time periods within which an advance ruling should be requested and issued, respectively. In other cases, the prescribed time periods are left to the applicable national laws of the parties. 10.4.6 Appeal or review The requirement to provide for appeal or review of Customs and other administrative decisions is one of the most common trade facilitation provisions, appearing in 123 of the PTAs reviewed. As with advance rulings, the provisions on appeal and review found in different PTAs vary in specificity. While some PTAs provide for review of customs decisions specifically, others provide more generally for an appeal/review mechanism regarding not only customs and trade facilitation, but other matters covered by the agreement. An example of an PTA with an appeal/review provision specific to customs matters is CAFTA- Dominican Republic; an example of a PTA with general review/appeal provisions is the Economic Partnership Agreement between Japan and Brunei Darussalam. 10.4.7 Disciplines on fees and charges, and penalties Eighty-seven of the 267 PTAs contain provisions on fees and charges, and 50 (about 20 percent) contain a provision on penalties. These agreements tend to either refer directly to GATT Article VIII or use Article VIII as a template. Only a handful of PTAs go beyond Article VIII.33 10.4.8 Release and clearance Pre-arrival processing is not a common measure and is mostly found in PTAs that entered into force after 2003. Fifty-five PTAs have pre-arrival processing. One hundred twelve of the 267 PTAs require parties to adopt or maintain risk management. Ninety percent (100 of the 112 PTAs) entered into force in 2002 or later. Thirty PTAs provide for post-clearance audits. On Authorized Operators (AOs), six of EFTA’s agreements provide for the negotiation of mutual recognition of AEO systems under the broader heading of release and clearance.34 33 WTO 2014b, page 23. 34 These are the PTAs concluded with Albania, Bosnia and Herzegovina, Georgia, Montenegro, Serbia, and Ukraine. 309 Handbook of Deep Trade Agreements 10.4.9 Border agency cooperation The requirement for cooperation between national border agencies is found in only a limited number of PTAs. Recently concluded EFTA agreements provide for either (a) cooperation among all border authorities within each party, and among the authorities of the different parties; or (b) simultaneous inspection by national border authorities when goods are imported or exported. 10.4.10 Formalities Of the 267 PTAs reviewed, 122 require periodic review of formalities and documents. Of these 122, 106 (88 percent) entered into force in 2003 or later. According to the WTO, more than half of PTAs provide for the simplification and/or reduction of formalities, and this is mostly stated in general terms.35 The use of international standards as the basis for import, export, or transit formalities and procedures is one of the most common TF provisions in PTAs that entered into force after 2003. Where specific standards are mentioned, they tend to be WCO instruments. Ninety of the 267 PTAs reviewed refer to the use of international standards. Only 20 PTAs have been found to contain a provision on establishing or maintaining a single window. Ten PTAs contain provisions regarding the use of pre-shipment inspection services, 16 contain provisions on customs brokers, and 72 contain provisions on temporary admission. 10.4.11 Transit Eight-five PTAs provide for freedom of transit. Of these, the European Union is a party to 27 (32 percent). 10.4.12 Exchange of information One of the first customs-related provisions included in PTAs related to the exchange of information between the customs agencies of the contracting parties. As a result, these provisions are found in more than 70 percent of PTAs. As with most other trade facilitation provisions, exchange of information provisions ranges from comprehensive and specific to more general and vaguer. The aim of these provisions is mostly to support customs enforcement. 35 WTO 2014b, page 29. 310 Trade Facilitation and Customs The shift towards a paperless clearance environment and the use of electronic data to support selectivity in customs clearance (i.e., focusing on high-risk goods and facilitating the release of low-risk goods) has led many customs administrations to consider how information and communication technology can be used to create “a global Customs network in support of the international trading system. [T]his implies the creation of an international e-Customs network that will ensure seamless, real-time and paperless flows of information and connectivity.”36 The Japan-Singapore PTA was the first to introduce the concept of paperless trading. It commits the parties to work towards the introduction of paperless trading between themselves as well as their respective private sector entities. 10.4.13 Customs unions Article XXIV of the GATT 1994 defines a customs union as an arrangement wherein the parties remove customs duties on goods originating in their respective territories and apply the same duties and regulations on goods from third parties. A customs union agreement represents “… a deeper form of integration than a free trade agreement, generally requiring … a greater loss of autonomy.”37 The vast majority (239, excluding 3 additional accessions) of PTAs notified to the WTO are free trade agreements, while only 17 are customs union agreements.38 Among theTF provisions specific to customs unions, legal harmonization and standardization are the most common, appearing in 8 agreements.These provisions, however, have different aims. The EU, Gulf Cooperation Council (GCC), and East African Community (EAC) agreements provide, respectively, for a Union Customs Code, a Common Customs Law, and a Customs Management Act,39 which have direct application in the territories of the contracting parties. Mercosur has developed a Common Customs Code which members have to incorporate into their national customs laws to ensure implementation. SACU, on the other hand, does not have a common customs legal framework with either direct application or incorporation into national law, but provides for the application of similar legislation related to customs and excise duties. One of the key issues in negotiating a customs union agreement relates to the collection and allocation of customs duties imposed on goods imported from third parties. The 36 World Customs Organization 2008. 37 Andriamananjara 2011. 38 These figures exclude various enlargement and accession agreements. 39 Yasui 2014b. 311 Handbook of Deep Trade Agreements outcome of this is indicative of the level of integration envisaged by the contracting parties. In the case of the EU, customs duties collected on imports belong to the EU (“own resources”) and are collected by members on behalf of the EU. In other customs unions, customs duties are revenues that belong to the members. A unique arrangement exists in SACU, where customs duties (and excise duties) are shared according to a revenue-sharing formula. Other customs union agreements provide for the “final destination” principle, according to which the duties belong to the country of destination of the goods.40 Most of the 8 agreements contain a provision specifying where customs and other duties and taxes are to be collected; these include the EC Treaty and Enlargement instruments, as well as the Common Economic Zone (CEZ). The final destination principle and other measures such as the collection of value-added taxes and excise duties require (with the exception of the EU) that customs union members still maintain a level of physical customs controls for goods moving between their territories.41 As a result, some customs unions have developed specific trade facilitation measures to expedite the movement of goods between them. The Mercosur countries have agreed on a number of intra-union border posts where integrated controls are to be applied. The Andean Community has a high-level working group that oversees the Community Policy for Border Integration and Development.42 The EAC has adopted a Single Customs Territory policy, which aims to increase the interconnectivity of the customs systems used by EAC members and includes a payment system to manage the transfer of revenues among members. 10.4.14 Origin administration As with customs unions, Article XXIV of GATT 1994 also defines a free trade area. This is an arrangement whereby the parties eliminate customs duties and other restrictive regulations of commerce. However, in contrast to a customs union, the parties in a free trade area maintain their respective tariffs on goods imported from third parties. To ensure that the benefits of this arrangement are not extended to third parties, the contracting parties negotiate provisions related to rules of origin. The application of origin rules is supported by provisions related to the administration of rules of origin. These measures are often cited as non-tariff barriers. The most common method of substantiating a claim of origin is a certificate of origin issued by a specified competent authority of the exporter. Most free trade agreements require that this certificate be presented to the customs 40 Andriamananjara 2011, page 117. 41 Yasui 2014b. 42 Kieck and Maur 2011. 312 Trade Facilitation and Customs administration of the importing country upon request. Most agreements provide that a competent authority will issue certificates of origin and sometimes this is a specific entity or it is left to the contracting parties to notify each other of the details of their respective competent authorities or authorized bodies. These are mostly customs administrations, chambers of commerce, and the like. In support of the origin administration requirements, FTAs usually have origin verification measures or mutual administrative assistance provisions, which allow the importing party to send a request for verification of origin to the competent authority of the exporting party. The proliferation of free trade agreements has resulted in a significant increase in the issuing of certificates of origin, which has added to the cost of doing business. This has compelled some countries to look at various options from a trade facilitation perspective to simplify origin administration systems. The WCO issued its “Guidelines on Certification of Origin” in July 2014 and mentions four systems that have been introduced through FTAs to move away from certification by competent authorities to self-certification, namely approved exporter, registered exporter, fully exporter-based, and importer-based systems. Examples of each are provided below. The EU-Korea PTA provides that an exporter can issue an origin declaration after being granted approved exporter status by its respective national customs administration. In the case of the Comprehensive Economic and Trade Agreement between the EU and Canada, EU exporters can apply to be a registered exporter, in which case the origin status of their goods is evidenced by an invoice statement. The US-Korea PTA provides that an exporter can issue a certification of origin, either written or electronically. There is no prescribed format for the certificate and it will be recognized by the importing party as long as it contains certain elements that are specified in the agreement. The most liberal form of self-certification is the importer-based system, such as the US-Australia PTA. In this agreement, the importer claiming a duty preference for imported goods does not have to submit a certificate of origin. However, the importing customs administration may request the importer to provide a statement setting out the reasons for the qualification of the goods. In some cases, the trade agreements of a country can have different origin administration systems in place, depending on a number of factors, including (a) the date when the agreement entered into force (older agreements tended to require a certificate of origin issued by an export competent authority); and (b) the ability and “comfort” of the import customs administration to apply a risk-based approach to identify non-compliance. For example, Australia’s free trade agreements cover three different origin administration systems, and one of these can be further divided into four sub-systems (Table 10.1). 313 Handbook of Deep Trade Agreements Table 10.1: Origin administration systems in free trade agreements to which Australia is a party METHOD AGREEMENT Importer-based, knowledge by the importer • Australia-US FTA Written declaration / declaration of origin • Australia-New Zealand Closer Economic filled out by the exporter or producer Relations Trade Agreement • Australia-Chile Free Trade Agreement • South Pacific Trade and Economic Cooperation Agreement • Malaysia-Australia Free Trade Agreement Certificate of origin • can be used for multiple shipments • for each shipment supplemented by written declaration • Singapore-Australia Free Trade Agreement for each shipment • Korea-Australia Free Trade Agreement • for each shipment or written • Japan-Australia Economic Partnership Agreement declaration for each shipment • China-Australia Free Trade Agreement • for each shipment, but a written declaration • Thailand-Australia Free Trade Agreement can be used in case of an advance ruling on • ASEAN-Australia-New Zealand Free Trade Area the origin of the good Source: Adapted from Tramby 2017, page 5. 10.4.15 Technical assistance and capacity building One hundred of the 267 PTAs provide for the parties to support each other in general and specific ways to facilitate trade. These include technical assistance, advisory services, training, study visits, and exchange or secondment of officials to strengthen cooperation or partnerships between the contracting parties. While most of the PTAs that provide for specific technical assistance are concluded between developed and developing nations, a few are between developing countries. These include Pakistan-Malaysia, Peru-Chile, and Dominican Republic-Central America. 10.4.16 Institutional With regard to institutional arrangements,176 of the 267 PTAs contain provisions for the establishment of a structure such as a working group or committee to achieve specified goals. Goals may include, for example, developing a customs or trade facilitation work program, preparing an instrument or activity provided for in the agreement, or monitoring the implementation of relevant provisions. Only 18 of the 267 PTAs provide for a mechanism to consult the private sector, of which one entered into force before 1995 and one between 2000 and 2004. 314 Trade Facilitation and Customs 10.5. CONCLUSIONS The concept of trade facilitation in PTAs has evolved over time, from only limited provisions on customs matters to the inclusion of more comprehensive provisions on regulatory and border matters that have an impact on the cross-border movement of goods. This shift includes an expansion in scope, to encompass not only cooperation between the contracting parties to ensure proper application of the agreement, but also issues of concern to the private sector, such as increasing transparency, promoting certainty and predictability, and simplifying official processes and requirements.The latter are aimed at reducing the complexity, cost, and time required to comply with international trade rules.This holds true for contracting parties from both developed and developing countries. Increasingly, there is a recognition of the need to include trade facilitation provisions to optimize the gains from PTAs. The expansion of the concept of trade facilitation and the number of trade facilitation provisions included in PTAs have resulted from a number of factors. These include the negotiation of the WTO Trade Facilitation Agreement, which increased the awareness and understanding of trade facilitation.This agreement, even before its entry into force, has influenced and will continue to influence the negotiation of trade facilitation provisions in PTAs. Expansion of the concept of trade facilitation has been accompanied by an increase in the average number of provisions included in the agreements. Further, as a result of the influence of the WTO TFA, there is an increasing convergence of these provisions, in terms of both their subject matter and language in the PTA, although significant divergence remains. Technological advancements have also influenced the design of trade facilitation provisions in PTAs, such as the inclusion of trade information portals, single window systems, and the advance electronic processing of declarations. This is expected to continue. The use of technology supports the modernization of customs and other regulatory and border agencies and expands the range of trade facilitation provisions. A number of countries are looking into the use of new technologies such as blockchain to support their international trade and border management activities. In the case of free trade agreements, interesting developments are taking place to modernize and simplify the administration of preferential rules of origin. Business practices will continue to evolve as customs administrations increasingly make better use of automation and build their risk management and post-clearance audit capacity. The modernization of customs administrations will most likely also impact customs unions, especially the collection of duties and taxes, and hopefully will result in further simplification of processes, declarations, and documentary requirements. 315 Handbook of Deep Trade Agreements Finally, the ultimate test of the trade facilitation provisions that contracting parties include in their PTAs is the extent of implementation. In this regard, most PTAs fall short of the dispute settlement provisions that govern the WTO TFA, and there is usually very little legal recourse in case of a lack of implementation. Very often, this is as a result of limited capacity or resources. As most PTAs do not provide for technical assistance and capacity building, this is an area where WTO members can play a significant role in supporting the efforts of less developed countries to introduce the reforms necessary to facilitate legitimate trade. ACKNOWLEDGMENTS I would like to thank Nora Neufeld for her pioneering work in mapping the WTO TFA and regional trade agreements, and for sharing her insights. A special thanks also goes to Mintewab Gebre Woldesenbet for her support and comments. Finally, let me thank my colleagues Bill Gain and Ankur Huria for their support. All errors are my own.   316 Trade Facilitation and Customs REFERENCES Andriamananjara, Soamiely. 2011. Customs unions. In Preferential UNCTAD (United Nations Conference on Trade and trade agreement policies for development: A handbook. Washington, Development). 2011. Trade facilitation in regional trade DC:World Bank. agreements. Trade and Transport Facilitation Series 3, UNCTAD/DTL/TLB/2011/1. Geneva: UNCTAD. Congressional Research Service. 2017.WTO Trade Facilitation Agreement R4477 (March 3). UNECE (United Nations Economic Commission for Europe). No date. Trade facilitation implementation guide, Article X. Duval, Yann, Nora Neufeld, and Chorthip Utoktham. 2016. http://tfig.unece.org/. Do trade facilitation provisions in regional trade agreements matter? Impact on trade costs and multilateral spillovers. WCO (World Customs Organization). 2008. Customs in ARTNet Working Paper Series 164. Economic and Social the 21st century. WCO, Brussels. Commission for Asia and the Pacific (ESCAP), Bangkok. ———. 2014a. Customs administrations operating under ICC (International Chamber of Commerce). 2017. Business customs union systems. WCO Research Paper 29, WCO, recommendations on rules of origin in preferential trade Brussels. agreements: Policy statement. Document 104-89, ICC, Paris ———. 2014b. Trade facilitation in regional trade agreements. IMF (International Monetary Fund). 2001. Global trade WCO Research Paper 30,WCO, Brussels. liberalization and the developing countries. Issues Brief 01/08, November. ———. 2014c. Guidelines on certification of origin, WCO, Brussels. Kieck, Erich, and Jean-Christophe Maur. 2011. Regional integration and customs unions. In Border management WTO (World Trade Organization). 2014a. Agreement modernization, edited by Gerard McLinden, Enrique Fanta, on trade facilitation: Decision of the General Council David Widdowson, and Tom Doyle. Washington, DC: World WT/L/940. WTO, Brussels. Bank. ———. 2014b. Trade facilitation provisions in regional trade McLinden, Gerard, Enrique Fanta, David Widdowson, and agreements: traits and trends. WTO Staff Working Paper Tom Doyle (eds.). 2011. Border management modernization. ERSD-2014-01, WTO, Brussels. Washington, DC: World Bank. ———. 2015. World Trade Report. https://www.wto.org/ Moise, Evdokia. 2002. The relationship between regional trade english/res_e/booksp_e/world_trade_report15_e.pdf. agreements and the multilateral trading system. Document TD/TC/WP(2002)17/FINAL. Organisation for Economic Co-operation and Development, Paris. Mustra, Monica Alina. 2011. Border management modernization and the trade supply chain. In Border management modernization, edited by Gerard McLinden, Enrique Fanta, David Widdowson, and Tom Doyle.Washington, DC:World Bank. OECD (Organisation for Economic Co-operation and Development). 2002. The relationship between regional trade agreements and the multilateral trading system. Chapter on trade facilitation. Working Party of the Trade Committee. Document TD/TC/WP(2002)17/FINAL (April 11). ———. 2005.The costs and benefits of trade facilitation. Policy Brief. Tramby, Gabrielle. 2017. Administration of rules of origin. Presentation delivered at the WCO Origin Conference, Department of Immigration and Border Protection, Government of Australia, Canberra. 317 Handbook of Deep Trade Agreements 318 Anti-Dumping and Countervailing Duties CHAPTER 11 Anti-Dumping and Countervailing Duties T. Prusa 319 Handbook of Deep Trade Agreements CHAPTER 11 Anti-Dumping and Countervailing Duties T. Prusa Rutgers University, New Brunswick, NJ, United States CONTENTS 11.1. INTRODUCTION 321 11.1.1 Survey of analytical and policy discussions surrounding trade remedies 323 11.1.2 Legal issues surrounding trade remedies in PTAs 324 11.1.3 Demand for trade remedies in PTAs 324 11.1.4 Welfare effects of trade remedy provisions in PTAs 325 11.2. METHODOLOGICAL APPROACH TO THE MAPPING 326 11.2.1 Templates used for the mapping 326 11.2.2 Anti-dumping provisions 326 11.2.3 Countervailing duties 330 11.3. ANALYSIS 332 11.4. CONCLUSIONS 335 ACKNOWLEDGMENTS 335 REFERENCES 336 ANNEX 337 320 Anti-Dumping and Countervailing Duties 11.1. INTRODUCTION This chapter examines two types of contingent protection provisions in preferential trade agreements (PTAs) - anti-dumping (AD) provisions and countervailing duties (CVD). These provisions are referred to as contingent because the World Trade Organization (WTO) agreement specifies that they may be imposed only when trade reaches a certain volume.1 Although other forms of protection often garner more attention, the provisions mapped in this chapter account for most of the discretionary border protection beyond WTO-negotiated tariff rates.2 Official statistics on other border measures are not reported to the WTO, but extrapolating from a review of US border measures, it is doubtful that there are more than a few hundred disputes involving all other trade statutes combined. AD and CVD provisions can be levied on exporters who engage in what the importing party believes are unfair trading practices that cause material injury to domestic producers.These unfair trading practices can take the form of the exporting party (a) selling products below what the importing party believes are normal prices, or (b) benefiting from government-provided subsidies. AD duties could be levied in the former case and CVDs in the latter. Before levying either of these duties, however, the complaining country must show that the unfairly traded goods have caused material injury.The rules governing how governments should calculate normal and export prices, compute the differences, determine subsidy amounts, assess the injury or impact of the unfair trade, and determine how long the punitive duties can last are all contained in the General Agreement on Tariffs and Trade (GATT) and the WTO agreements.3 Whatever the conditions under which they can be triggered,both AD and CVD represent international agreement about what means a country can use to temporarily increase the level of trade protection for an injured domestic industry. However, given that both types of provisions are governed by existing GATT/WTO agreements, PTAs without such provisions issues simply fall back on their existing WTO commitments to address unfair trade issues. In other words, WTO rules prevail unless the PTA imposes additional contingent protections, or unless the PTA prohibits the use of contingent protections against other PTA members.As a general rule, contingent provisions typically make it more difficult to levy duties on PTA members than do the WTO rules. In addition, if a PTA’s contingent provisions on margins and subsidies are too strict, some exporters may be disinclined to join the PTA. 1 Global safeguards are another type of contingent provision, but they are not considered in this study. Under the WTO Safeguards Agreement, a WTO member may restrict imports of a product temporarily (take “safeguard” actions) if its domestic industry is seriously injured or threatened with serious injury caused by a surge in imports. 2 Bown 2011 documents that approximately 90% of trade value subject to discretionary protection is due to AD or CVD protection. 3 Article VI of the GATT provides for the right of contracting parties to apply AD measures; the WTO’s Agreement on Subsidies and Countervailing Measures addresses multilateral disciplines regulating the provision of subsidies and the use of countervailing measures to offset injury caused by subsidized imports. 321 Handbook of Deep Trade Agreements Contingent protection provisions in PTAs vary with the size of the PTA and the degree of integration, geographic scope, and level of economic development of its members. Contingent provisions also vary across PTAs for the same country. Further, a country’s attitude toward PTA provisions is a moving target.The United States’ stance, for instance, has clearly evolved, and its most recent PTAs lack many provisions included in its early PTAs. The overall trend is for PTAs to use fewer contingent protection provisions and to tighten restrictions on their use. A large number of PTAs have adopted rules that tighten discipline on the application of such measures, while more than half include some additional rules on applying them and 10 percent of PTAs have eliminated contingent protection provisions altogether. In the case of ADs, key provisions increase the de minimis volume, tighten dumping margin requirements, and shorten the duration for applying ADs relative to the WTO Anti-Dumping Agreement. With respect to CVDs, few PTAs contain rules that meaningfully curb subsidies or state aid. An argument for not including contingent protection provisions in PTAs is that economic impact of subsidies is rarely confined to just intra-PTA trade - subsidies affect global trade.4 However, PTAs that do have CVDs tend to add to them.5 A number of PTAs that include ADs and CVDs give a role to regional institutions to investigate claims or review the final determinations of national authorities. There is a theoretical presumption and some empirical evidence to suggest that these regional bodies reduce the frequency of anti-dumping and subsidy claims and of final determinations against PTA members.6 The complicated pattern of inclusion of these provisions threatens the delicate give-and- take balancing of incentives that is the crux of the GATT/WTO agreements. An ongoing policy concern is that the elastic and selective nature of trade remedies may lead to more discrimination, with reduced trade remedy actions against PTA partners but a greater frequency of such actions against non-members.The adoption of PTA-specific trade remedy rules increases this risk of discrimination, with trade remedies against PTA members being abolished outright or being subjected to greater discipline. In turn, this makes it more difficult for non-PTA members to agree to WTO liberalization, as the requisite quid pro quo from PTA members may not be realized. Said differently, market access that non-PTA members thought they had secured in prior WTO rounds may be eroded not primarily because of discriminatory tariffs but rather because of contingent protection rules. 4 While the work of Bown and Crowley 2007 demonstrates that anti-dumping duties also have a global trade effect, the difference is that the subsidy itself can affect exports to many markets. 5 While most PTAs that have AD provisions also have CVD provisions, they may not have the same injury provisions, even without the same PTA. 6 Blonigen 2005. 322 Anti-Dumping and Countervailing Duties 11.1.1 Survey of analytical and policy discussions surrounding trade remedies Why do trade agreements need trade remedy provisions? One explanation lies in the political economy of protectionism. The long-term process of tariff liberalization in the post-World War II era has successfully reduced tariff rates to very low levels worldwide. Despite, or perhaps because of, this liberalization, import-competing sectors continue to have an incentive to secure protection through whatever means are available. Although trade remedy measures are typically administered by bureaucracies that appear to be insulated from political pressure, influence can be brought to bear on them indirectly through the shaping of laws and regulations that govern their work.7 One advantage offered by administered protection to import-competing sectors is that it is inherently biased in their favor - it is a channel for complaints about an excess of import competition, not the lack of competition. By design, the trade remedy bureaucracy can only impose protection and not remove it (other than that which it imposes itself). A second explanation is that trade remedy measures are a pragmatic tool to deal with the political demands for protection that trade liberalization provokes.8 Trade liberalization may lead to costs of adjustment, and if nothing is done to manage those costs, political pressure may build to a point where protectionist forces would be able to engineer a permanent reversal of liberalization. The introduction of trade remedy measures in a trade agreement may be thought of as anticipating the possibility of such difficult adjustments and the demands for protection to which they give rise. Trade remedy measures provide a means to deflate these demands through a temporary reversal of liberalization. This implies that the depth of liberalization that a trade agreement can achieve may depend on whether there are escape clauses that allow governments to depart temporarily from their liberalization commitments under well-defined and circumscribed conditions. Whereas the use of such measures may result in welfare losses during periods when the level of protection is temporarily increased, the deeper liberalization that is embedded in the trade agreement means that these losses could be outweighed by long-term welfare gains. Paradoxically, these arguments suggest that PTAs should generally make it easier for member states to grant contingent protection. Empirically, however, trade remedy rules in PTAs generally work in the opposite direction: they often make it more difficult to grant protection. With respect to ADs, the inclusion of such rules is consistent with the view that dumping is driven by closed home markets.9 However, the elimination of barriers to intra-PTA trade reduces the ability of firms to dump because they no 7 Finger, Hall, and Nelson 1982. 8 Jackson 1997. 9 Mastel 1988. 323 Handbook of Deep Trade Agreements longer have a protected home market where they can earn supernormal profits. More generally, the opening of markets via PTA preferences reduces the ability for countries to price discriminate. This logic is also consistent with the lack of CVD rules in PTAs. Because most PTAs have failed to strengthen anti-subsidy rules, the notion that there will be fewer subsidies in PTA blocs, and in turn less need for CVDs, is not supported. Nevertheless, political pressure often makes it difficult to include CVD rules in PTAs. 11.1.2 Legal issues surrounding trade remedies in PTAs Because PTAs have the objective of dismantling barriers to trade between members, one might expect PTA members to abolish the use of trade remedies. In fact, there are those who view the elimination of trade remedies, in particular AD actions, as a requirement under Article XXIV of GATT 1994, which deals with customs unions and free trade areas. Paragraph 8(b) of GATT Article XXIV requires WTO members that form a preferential trade agreement to eliminate duties and other regulations restricting trade.10 One economist11 interprets the reference to other regulations restricting trade to include trade remedies, specifically AD actions. This view is strengthened by the fact that paragraph 8(b) of GATT Article XXIV allows PTA members to exclude, when necessary, certain GATT articles from the general requirement to eliminate other regulations restricting trade.12 It would have been easy to include GATT Article VI (Anti-dumping and Countervailing Duties) in the excluded GATT articles, if that had been the intention of the framers of the GATT. 11.1.3 Demand for trade remedies in PTAs As noted above, the elimination of intra-PTA tariffs may create new demands for the protective effects of trade remedies. For a government entering into a PTA, import- competing sectors need to be given assurance that they have the means to protect themselves from the unanticipated consequences of the intra-PTA liberalization program. Retaining trade remedies in the PTA serves the useful purpose of securing political support for the agreement. 10 Article XXIV 8(b) states that “A free-trade area shall be understood to mean a group of two or more customs territories in which the duties and other restrictive regulations of commerce (except, where necessary, those permitted under Articles XI, XII, XIII, XIV, XV, and XX) are eliminated on substantially all the trade between the constituent territories in products originating in such territories.” 11 Marceau 1994. 12 The GATT articles not covered by the requirement to eliminate “other regulations restricting trade” include articles XI (General Elimination of Quantitative Restrictions), XII (Restrictions to Safeguard the Balance of Payments), XIII (Nondiscriminatory Administration of Quantitative Restrictions), XIV (Exceptions to the Rule of Nondiscrimination), XV (Exchange Arrangements), and XX (General Exceptions). 324 Anti-Dumping and Countervailing Duties In these circumstances, trade remedies might be akin to other provisions in PTAs that limit or delay the possible deleterious effects of the PTA’s liberalization. For example, long transition periods, complicated rules of origin, and carve-outs for sensitive sectors all aim to cushion the effects of the PTA by drawing out the process of tariff elimination. Similarly, trade remedies achieve a cushioning effect by specifying a set of conditions - injury to a domestic industry - under which the regional liberalization program may be temporarily suspended or partially reversed. 11.1.4 Welfare effects of trade remedy provisions in PTAs Whereas abolishing trade remedies on PTA partners’ imports will most likely increase intra-bloc trade, this does not necessarily mean that abolition would raise welfare.The ambiguity of the welfare impact stems from the well-known insight that preferential trade arrangements have both trade- creation and trade-diversion effects.13 The impetus given to intra-bloc trade by the abolition of trade remedy actions may be at the expense of cheaper imports that come from nonmembers. There is a danger, in fact, that as intra-bloc trade expands because of falling intra-bloc tariffs, contingent protection may be increasingly directed at the imports of nonmembers. Some have argued that as a result of its elastic and selective nature, administered protection can increase the risk of trade being diverted away from PTAs.14 Therefore - in addition to discrimination against nonmembers introduced by preferential tariffs - the establishment of PTAs can lead to more discrimination through more frequent trade remedy actions. The elastic and selective nature of trade remedy protection allows for nonmembers to be targeted more frequently, and without necessarily requiring adherence to special PTA rules on trade remedies.Thus, one key conclusion from this hypothesis is that in a world teeming with PTAs, there is greater need for stronger multilateral discipline on trade remedies. To the extent that PTAs adopt special or additional rules on trade remedy actions against members’ trade, they can effectively increase the level of discrimination against nonmembers. Increased discrimination could increase when PTA members abolish trade remedy actions against other PTA members but not against nonmembers’ trade. It could also occur when PTA members strengthen discipline on trade remedy actions against PTA members but not against the trade of nonmembers. Moves to strengthen discipline on, or abolish, trade remedy actions against PTA partners may initially appear to be good for trade. However, once again the welfare effects are ambiguous. Such rules may simply lead to intra-regional imports substituting for cheaper imports from nonmembers (i.e., trade diversion). Because PTAs may thrust members into the world of the second best, actions that look as if they will lead to an increase in economic efficiency may achieve exactly the opposite effect. 13 Viner 1950. 14 Bhagwati 1992, 1993; Bhagwati and Panagariya 1996. 325 Handbook of Deep Trade Agreements 11.2. METHODOLOGICAL APPROACH TO THE MAPPING The primary purpose of the mapping exercise is to understand the nature of contingent protection rules in PTAs, recognizing that there may be a difference between the legal language in the provisions and how they are implemented. While this study does not account for factors that could affect implementation, such as institutional setting and administrative procedures, the mapping of trade remedy provisions as they were negotiated and written illuminates trends that can be predictive of future trade disputes and ways they can be resolved both within and outside of intra-PTA agreements. Nevertheless, future studies will need to examine the factors that account for differences in implementation, to ascertain what part they play in determining the trade and welfare effects of trade remedy actions.15 11.2.1 Templates used for the mapping The mappings of the PTAs surveyed for this chapter are drawn almost exclusively from the legal texts of the 283 PTAs in the World Bank’s new Deep Integration Database. The list of the PTAs appears in Annex Table 11.A.1.To my knowledge, this database of contingent protection rules in PTAs is the most comprehensive available.The database includes all of the economically large PTAs and the most active users of contingent protection. 11.2.2 Anti-dumping provisions A two-level template was developed for the comparative analysis of AD provisions. Level 1: The first level of the template classifies AD provisions in PTAs into three mutually exclusive categories (Table 11.1). The first category includes those that disallow AD actions among the PTA members. The second category includes PTAs that have no such prohibition and have no specific language or provisions on AD. The third category is made up of PTAs that allow AD against PTA members and contain specific provisions on AD. If specific AD rules apply to PTA members, the second level of the template maps specific provisions of the agreement. 15 Two significant studies related to these issues are Blonigen and Prusa (2003) and Horlick and Vermulst (2005). Blonigen and Prusa (2003) identify differences in AD practices across countries in the areas of transparency of proceedings, collection of AD duties before or after an injury determination is made, levying of the full or partial dumping margin, and other differential applications of punitive actions. In a similar vein, Horlick and Vermulst (2005) review the AD practices in ten major countries - Australia, Brazil, China, the EC, India, Indonesia, Mexico, South Africa,Thailand, and the United States - and find variations in institutional practices in the areas of procedures, determination of dumping margins, and injury determinations. They argue that the increasing use of constructed normal values gives too much discretion to AD authorities in determining the existence of dumping and conclude that there is too much administrative discretion in the determination of injury, injury margins, and causation. 326 Anti-Dumping and Countervailing Duties Table 11.1: Summary of AD/CVD rules in PTAs AD CVD Prohibited 21 (7.4%) 14 (4.9%) No rules 55 (19.4%) (35.0%) 99 Rules 207 (73.1%) 170 (60.1%) Rules (weak) 98 (34.6%) 105 (37.1%) Rules (strong) 109 (38.5%) 65 (23.0%) TOTAL 283 283 By this categorization, nearly three quarters of the PTAs have AD rules that go beyond WTO anti-dumping rules, and about one fifth make no mention of AD rules. Only a small number (21) prohibit the use of AD provisions. These are Australia- New Zealand (ANZCERTA); Canada-Chile; China-Hong Kong SAR, China; China-Macao SAR, China; the Common Economic Zone (CEZ); the original EC and its various expansions; the European Economic Area (EEA); the European Free Trade Association (EFTA); EFTA-Bosnia and Herzegovina; EFTA-Chile; EFTA- Hong Kong SAR, China; EFTA-Montenegro; EFTA-Serbia; and EFTA-Ukraine. In the case of the EEA, the prohibition on ADs applies only to trade of goods that fall under chapters 25 to 97 of the Harmonized Tariff System, while AD measures can still be taken against agricultural and fishery goods. The Chile-Mexico Free Trade Agreement (FTA), when it came into force in 1999, stipulated that future negotiations between the partners would lead to the removal of AD measures. However, this removal has not been achieved. Of the 183 PTAs with language pertaining to AD rules, 81 have weak or modest rules. While these PTAs are properly denoted as having rules, it is unclear how vague language or simply referencing WTO anti-dumping provisions amounts to a significant change from WTO practice in how PTA members should conduct AD investigations or determine findings. This is a question for future research. Level 2: For PTAs that contain specific AD rules, the second level of the template maps these specific provisions in some detail. Patterned after the Anti-dumping Agreement of the WTO, it includes elements such as determination of dumping, determination of injury, evidence, provisional measures, duration and review of AD duties and price undertakings, and notification and consultation. In addition, the template includes elements that are either unique to some agreements or that have been highlighted in the literature. 327 Handbook of Deep Trade Agreements AD – Level 2 Mapping 1 Determination of dumping 1.1 export price less than comparable price when destined for consumption in the exporting country 1.2 if there are no sales in the normal course of trade in the domestic market of the exporting country 1.3 a comparable price of the like product when exported to an appropriate third country 1.4 cost of production in the country of origin plus a reasonable amount 1.5 non-market economies 2 Determination of injury 2.1 volume of dumped imports 2.2 price effects of dumped imports 2.3 the consequent impact of dumped imports on the domestic industry - material injury 2.4 causality 2.5 material injury 3 Definition of domestic industry 4 Mutually acceptable solution (1=yes, 0=no) • if yes, length of period (days) 5 Initiation and conduct of investigations 5.1 on behalf of the domestic industry if collective output constitutes more than 50% of total 5.2 no initiation if the collective output is less than 25% of total 5.3 de minimis dumping margin 5.4 de minimis dumped volume 6 Evidence 7 Provisional measures 8 Price undertakings 9 Imposition and collection of anti-dumping duties 9.1 duty shall not exceed the margin of dumping 9.2 lesser duty rule 9.3 collection on a non-discriminatory basis 10 Retroactivity 11 Duration and review of anti-dumping duties and price undertakings 12 Duration 12.1 established period 12.2 review 13 Public notice and explanation of determinations 14 Anti-dumping action on behalf of a third country 15 Joint body/committee 15.1 conducts investigations and decides on AD duties 15.2 review/remand final determinations 15.3 other 16 Notification/consultation (1=yes, 0=no) • if yes, length of period (days) 17 Dispute settlement 18 In accordance with GATT Article VI / AD Agreement 328 Anti-Dumping and Countervailing Duties Discussion: Some of the AD language included in PTAs makes PTAs less attractive than simply relying on the WTO Anti-dumping Agreement. For example, the Andean Community requires a higher de minimis volume (6 percent) and mandates a shorter period (3 years) for applying AD than does the WTO Anti-Dumping Agreement. The New Zealand-Singapore PTA has a higher de minimis dumping margin (5 percent) and a higher de minimis volume requirement (5 percent) than the WTO benchmark. The Southern Common Market (Mercosur) of South America also limits the duration of AD duties (to 3 years, compared to 5 years in the WTO agreement). Of the PTAs that include anti-dumping provisions, many also include joint oversight bodies, which tend to reduce the amount of AD activity between members countries. Forty of the 283 PTAs included in the survey give a role to joint oversight bodies or committees to conduct investigations and/or review the final determinations of national authorities. Included in this group are the Andean Community (CAN), Central American Common Market (CACM), Caribbean Community (CARICOM), the North American Free Trade Agreement (NAFTA), Canada-Chile, and Canada-Costa Rica.16 Of the 40 PTAs with joint oversight bodies, 8 are customs unions, 29 are free trade agreements, and two are of an unknown type.17 Some of these groupings, especially if member states are in the same region, have a history of relying heavily on regional institutions in the integration process.The Andean Community and CARICOM, in particular, are composed of small member states and are accustomed to pooling expertise and resources. For example, in the context of the current WTO negotiations related to the Sanitary and Phytosanitary (SPS), Technical Barriers to Trade (TBT), and Trade-Related Intellectual Property Rights (TRIPS) agreements, CARICOM countries have tabled proposals that will allow the WTO to designate a regional body to carry out the functions necessary to implement these agreements. TheseWTO agreements have implementation obligations that seem to pose very high hurdles for developing countries, particularly for the smallest ones.This explanation has some similarity to the argument that a small state’s decision to form, expand, or join a regional organization is based on reduced negotiating costs and increased bargaining power rather than on the traditional costs and benefits of trade integration.18 However, the use of regional bodies to lower the cost of AD actions can also constrain the ability of domestic producers to inveigle a compliant national investigating authority to find in their favor in dumping cases. 16 In the Andean Community, the Secretary General is given the authority to open and conduct AD investigations and decide on provisional and final AD duties. In CACM, the Secretariat for Central American Economic Integration (SIECA) is the regional body given the authority to conduct AD investigations. In CARICOM, one of the regional organs - the Council for Trade and Development (COTED) - has the authority to conduct AD investigations, authorize member-states to apply AD measures, and keep such measures under review. In the case of NAFTA, the establishment of binational panels can be requested by any of the members to review final AD determinations. 17 Information on the PTA type was missing for a number of agreements. 18 Andriamananjara and Schiff 1999. 329 Handbook of Deep Trade Agreements Almost all of the PTAs to which the EC is a party contain specific anti-dumping language and also have joint oversight bodies. When (or even before) an AD petition is initiated, the oversight body is informed, and the parties attempt to reach a mutually satisfactory solution. If no solution is found, the action (investigation or final determination) proceeds. The oversight body can take provisional AD measures if a delay would lead to material injury of one of the parties. For EC-centered PTAs (i.e., PTAs involving the EC), most establish joint committees to oversee implementation of the agreement, but apart from serving as a forum for consultations or notification, these committees do not play an important role in how an AD action unfolds. EFTA-centered PTAs that involve the same trading partners as the EC-centered PTAs exhibit similar characteristics. With the exception of NAFTA, the PTAs entered into by the United States (US-Australia, US-Bahrain, US-Chile, US-Colombia, US-Israel, US-Jordan, US-Republic of Korea, US- Morocco, US-Oman, US-Panama, US-Peru, US-Singapore) have no specific provisions on AD. Significantly, all but US-Israel PTA were negotiated after NAFTA. The change in the United States’ position after NAFTA likely reflects unhappiness by large AD users, and in turn unhappiness by key members of Congress, over the perceived loss of autonomy in applying AD against NAFTA partners. The large number of PTAs (228) that have either abolished AD actions against PTA members or have drawn up specific rules on AD actions against PTA members should raise some concern about increased discrimination, whether de facto or de jure, against nonmembers. From a welfare standpoint, increased discrimination raises the likelihood of greater trade diversion to outside of the trade bloc. 11.2.3 Countervailing duties The two-level template developed for the comparative analysis of CVD provisions is similar to the AD template, except that the CVD template also asks about the presence of a common policy or program on subsidies and about any disciplines that are imposed on the use of subsidies and state aid. Under WTO multilateral rules, CVDs can be levied on imports that benefit from subsidies if they cause or threaten material injury to an established domestic industry, or if they materially retard the establishment of a domestic industry. If PTA members have a common policy on subsidies or state aid or are able to agree on additional disciplines that apply to subsidies or state aid, they may be able to dispense with the use of CVDs. However, absent a common subsidy policy or additional disciplines on subsidies, it is unlikely that the provisions governing CVDs in the PTA will depart from WTO rules or practice. Even if the PTA members have a common policy on subsidies or state aid, it may be difficult for them to negotiate CVD rules because the economic impact of subsidies is global and not confined to intra-PTA trade. 330 Anti-Dumping and Countervailing Duties The first level of the CVD template classifies CVD provisions in PTAs into three mutually exclusive categories. The first consists of PTAs that disallow CVD actions against PTA members.The second category includes PTAs with no specific CVD provisions.The third are PTAs with specific CVD rules. PTAs with additional information about regional disciplines on subsidies and state aid are also included in the third category. Based on these categories, about 60 percent of the PTAs have additional CVD rules and about 35 percent have no CVD rules (Table 11.2). Only 14 PTAs (about 5 percent of the sample) have abolished CVDs. These are Canada-Chile; Chile-Japan; China-Hong Kong SAR, China; China-Macao SAR, China; the original EC and its various expansions; the European Economic Area; and the European Free Trade Association (EFTA), However, in the case of EFTA and the EEA, CVDs are disallowed only for products falling under chapters 25 to 97 of the Harmonized Tariff System; that is, CVDs can be applied to agricultural and fishery products. Eighty-two of the 143 PTAs with language pertaining to CVD rules appear to have weak or modest rules. The second and more detailed level of the CVD template involves determining whether certain provisions are present in the third category of PTAs and is patterned after the Subsidies and Countervailing Measures (SCM) Agreement of the WTO. As CVD rules do not have the same granular language as AD rules, it was not possible to map CVD rules in the same way as AD rules, coding for elements as initiation and subsequent investigation, evidence, consultation, and determination of injury.Therefore, the CVD mapping focused on whether the PTAs had the more limited set of provisions listed below: CVD – Level 2 Mapping 1 Mutually acceptable solution (1=yes, 0=no) • if yes, length of period (days) 2 Joint body/committee 1.1 conducts investigations and decides on CVD rules 1.2 reviews/remands final determinations 1.3 other 3 Notification/consultation (1=yes, 0=no) • if yes, length of period (days) 4 In accordance with GATT Article XVI Discussion: The great majority of the surveyed PTAs either have no specific CVD provisions (99 PTAs) or have specific provisions that allow the use of CVD measures (170 PTAs), while the remainder specifically disallow the use of CVD measures. Of the PTAs with specific provisions on CVDs, 105 have what can be considered weak provisions, stating only that all CVD actions should be in accord with GATT Article VI and the SCM Agreement. Under a stronger standard, only 65 PTAs include any detailed provisions on CVD actions. Of these 65, only 14 include provisions that allow for a joint oversight body or committee to conduct CVD investigations or review and remand final CVD determinations.The 14 PTAs with such provisions are the Andean 331 Handbook of Deep Trade Agreements Community, Australia-China, CARICOM, Central American Common Market, EFTA-North Macedonia, EFTA-Jordan, EFTA-Morocco, EFTA-Tunisia, EFTA-Turkey, EU-Colombia and Peru, Korea-Vietnam, NAFTA, Pacific Island Countries Trade Agreement (PICTA) and Southern African Customs Union (SACU). The mapping indicates that there has been little change in CVD rules since they were originally set in the WTO agreements, likely due to a lack of consensus on additional curbs to subsidies or state aid. The only explicit provisions on CVD in the PTAs surveyed relate to the prohibition or elimination of export subsidies for agricultural products, and to the prohibition of state aid if it distorts competition. Thirty-four PTAs have language related to one or both issues; however, 17 of these 34 are various EC expansions, EFTA, or PTAs involving the EC or EFTA. PTAs with rules on subsidies and state aid are much more likely to also have rules governing the use of CVD; more than 75 percent of PTAs with rules restricting subsidies also prohibit CVD or have rules limiting their use. Many PTA that do not explicitly prohibit state aid will still include a general statement against state aid that distorts competition. This is particularly true of countries that have not put subsidy programs on the table in their PTA negotiations, and thus may feel a continuing need for CVD as a weapon to wield against such support. Although it is possible to agree to a reduction or elimination of subsidies in an PTA negotiation, part of the trade benefits from such an action could be captured by nonmembers.The reluctance to give away trade benefits to nonmembers may explain why the only meaningful negotiation on further reductions in agricultural subsidies is occurring at the multilateral level. The free rider problem might also explain why there are so few PTA rules on CVD. Specifically, in contrast to AD, where pricing issues can be viewed as market specific, the economic consequences of a subsidized industry are likely go beyond any single market and any single PTA. Hence, it may well make less sense to include CVD provisions in PTAs, given the nature of the distortion. 11.3. ANALYSIS Correlation across contingent policy rules: The Level 1 mapping of AD and CVD rules in PTAs clearly shows a correlation between the two types of rules (Table 11.2). For instance, PTAs that prohibit the use of AD are also likely to prohibit the use of CVD. Examples include Canada-Chile; China-Hong Kong SAR, China; China-Macao SAR, China; EEA; EFTA; and the EC and its various enlargements. All of these PTAs are characterized by deep integration, suggesting that the extent of a PTA’s rules on AD and CVD is likely consistent with the overall depth of integration. Likewise, PTAs with AD rules, including weak rules, tend to also have CVD rules; and PTAs that do not have AD rules almost never have CVD rules, although the correlation is somewhat weaker. 332 Anti-Dumping and Countervailing Duties Table 11.2: AD/CVD rules in PTAs PTAs with “weak” rules included PTAs with “weak” rules excluded AD AD, AD, AD AD, AD, prohibited no rules rules prohibited no rules rules CVD prohibited 13 1 0 CVD prohibited 13 1 0 CVD, no rules 2 53 44 CVD, no rules 2 53 28 CVD, rules 6 1 163 CVD, rules 4 0 54 Patterns over time: The Level 1 mappings show that the propensity of PTAs to include AD and CVD rules has increased over time (Table 11.3). Until 2000, about 45 percent of PTAs included AD rules and about 36 percent had CVD rules. Since 2000, in contrast, more than 80 percent of PTAs have included AD rules and 60 percent have included CVD rules. Table 11.3: AD/CVD rules over time AD % of PTAs with rules Prohibited No rules Rules Total Include “weak” rules Exclude “weak” rules 1958-1995 8 18 21 47 45% 36% 1996-2000 1 19 16 36 44% 33% 2001-2005 5 6 41 52 79% 40% 2006-2010 2 6 70 78 90% 33% 2011- 5 3 48 56 86% 48% Total 21 52 196 269 73% 38% CVD % of PTAs with rules Prohibited No rules Rules Total Include “weak” rules Exclude “weak” rules 1958-1995 7 23 17 47 36% 23% 1996-2000 1 31 4 36 11% 8% 2001-2005 3 18 31 52 60% 10% 2006-2010 2 16 60 78 77% 21% 2011- 1 8 47 56 84% 43% Total 14 96 159 269 59% 22% Patterns by PTA type: Broadly speaking, PTAs can be categorized as Customs Unions, Free Trade Agreements, or Partial Scope Agreements. Some PTAs are also referred to as Economic Integration Agreements (EIAs) because of provisions explicitly covering trade in services. Table 11.4 presents a break-out of the Level 1 mapping by agreement type. Several lessons can be drawn from the table. First, the extent of the PTA rules on AD and CVD is likely consistent with the overall depth of integration. For example, I found no examples of partial scope agreements where either AD or CVD was prohibited. In fact, only two partial scope agreements had anything more that weak rules for AD and CVD. Second, customs unions are far more likely to prohibit the use of AD and/or CVD. This makes sense as one would expect the depth of integration to be greater for customs unions. However, caution is warranted in 333 Handbook of Deep Trade Agreements inferring too much from this result as the finding is almost entirely driven by the EC and its various expansions. Third, free trade agreements generally have rules, especially those that have deeper integration (as reflected in also have service trade agreements). Table 11.4: AD/CVD rules by PTA type AD Prohibited No rules Rules Total Customs Union 3 8 6 17 Customs Union & EIA 6 3 3 12 Free Trade Agreement 4 30 70 104 FTA & EIA 8 5 109 122 Partial Scope Agreement 0 6 8 14 Total 21 52 196 269 CVD Prohibited No rules Rules Total Customs Union 3 10 4 17 Customs Union & EIA 5 3 3 11 Free Trade Agreement 0 56 48 104 FTA & EIA 5 20 97 122 Partial Scope Agreement 0 7 7 14 Total 14 96 159 269 Comparing EU and US rules: The European Community/Union and the United States have been two of the most active partners in trade agreements, with 44 and 13 PTAs, respectively. The two leaders clearly have different views on AD and CVD rules in agreements. The EC/EU has prohibited the use of AD and CVD in 8 of its agreements and imposed significant (non-weak) AD rules in 24 agreements and significant CVD rules in 12 agreements. By contrast, the US has never signed an agreement prohibiting either AD or CVD and has been a part of only one agreement - NAFTA - with substantial (non- weak) rules for either AD or CVD. While their mixed evidence (at best) that NAFTA rules have changed AD and CVD outcomes, it does appear that NAFTA rules have changed the filing patterns, with both Canada and Mexico now somewhat less likely to be subject to investigations. Whatever the actual effect of the NAFTA rules, the effect on US negotiators is quite clear - AD and CVD rules have not been included in any PTAs since NAFTA. In the EC/EU, on the other hand, negotiators have been much more open to including meaningful provisions in their agreements. For example, 18 EC/EU agreements have rules directing the parties to try and find a mutually acceptable outcome before initiating a formal investigation. Most agreements specify a 30-day negotiating window. Fifteen EC/EU agreements have rules (often multiple rules) governing how and when an oversight body might conduct or review an investigation. 334 Anti-Dumping and Countervailing Duties 11.4. CONCLUSIONS Rules governing some aspects of trade remedies are relatively common in preferential trade agreements. Trade remedies provide governments entering into a PTA with a useful policy tool to manage trade adjustment and the political pressure for protection that it creates. They also make it easier to obtain political support for the agreement. The PTA, in turn, makes possible a more liberal trade regime, despite the episodic recourse to protection during economic downturns. At least some WTO members, however, most notably the United States, view trade remedy rules as a serious problem and are reluctant to include them in their agreements. On the other hand, any increase in intra-PTA trade brought about by greater discipline on trade remedy actions may simply be substituting for cheaper imports from nonmembers. In light of both arguments and in the absence of further evidence, it appears that the welfare effects of additional AD and CVD rules are unclear. Based on the result of this mapping, less than one-tenth of the PTAs surveyed have dispensed with at least one type of trade remedy.What these PTAs seem to share in common is a greater level of integration (“deep” integration) as evidenced either by the adoption of common or harmonized behind-the-border policies and high shares of intra-regional trade. There appears to be a large number of PTAs that have adopted PTA-specific rules that have tightened discipline on the application of these remedies on PTA members. In the case of AD, for example, we noted that some specific provisions tightened discipline by increasing de minimis volume and dumping margin requirements and shortening the duration for applying AD duties relative to the WTO Anti-Dumping Agreement.The possible contribution by regional bodies to reducing action against PTA members has also been discussed. In the EC-centered and EFTA-centered PTAs, members acting through a regional body notify and consult one another to arrive at a mutually acceptable outcome short of applying the measure. In the Andean Community, CACM, CARICOM, NAFTA, and the West African Economic and Monetary Union (UEMOA), regional bodies have the authority to conduct their own investigations or to review conclusions reached by national bodies. In the case of CVDs, there is little evidence of major innovations in CVD rules and practice by past and present PTAs. We suspect that a major reason for this is the absence of agreements in the PTA on meaningful or significant curbs on subsidies or state aid.  ACKNOWLEDGMENTS I would like to thank Robert Teh for numerous impactful discussions related to PTAs and contingent protection measures. I have learned a great deal from his first hand experiences. Chad Bown and the editors of this volume also provided helpful comments. All remaining errors are my own. 335 Handbook of Deep Trade Agreements REFERENCES Andriamananjara, Soamiely, and Maurice Schiff. 1999. Regional groupings among microstates. Policy Research Working Paper 1922,World Bank,Washington, DC. Bhagwati, Jagdish. 1992. Regionalism versus multilateralism. World Economy 15(5): 535-56. ———. 1993. Regionalism and multilateralism: an overview. In New dimensions in regional integration, edited by Jaime de Melo and Arvind Panagariya. Cambridge: Cambridge University Press. Bhagwati, Jagdish, and Arvind Panagariya (eds.). 1996. The economics of preferential trade agreements. Washington, DC: AEI Press. Blonigen, Bruce A. 2005. The effects of NAFTA on antidumping and countervailing duty activity. World Bank Economic Review 19: 407-24 (December). ———. 2006. Evolving discretionary practices of U.S. antidumping activity. Canadian Journal of Economics 39(3): 874-900. Blonigen, Bruce A., and Thomas J. Prusa. 2003. Antidumping. In Handbook of international economics, edited by E. Kwan Choi and James Harrigan. Malden, MA: Blackwell. Bown, Chad P., 2011. Taking stock of antidumping, safeguards and countervailing duties, 1990-2009. World Economy 34(12). Bown, Chad P . and Meredith A. Crowley. 2007. Trade deflection and trade depression. Journal of International Economics 72(1): 176-201. Estevadeordal, Antoni, Kati Suominen, and Robert Teh (eds.). 2009. Regional rules in the global trading system. Cambridge: Cambridge University Press. Finger, J.M., Keith H. Hall, and Douglas R. Nelson. 1982. The political economy of administered protection. American Economic Review 72: 452-66. Horlick, Gary, and EdwinVermulst. 2005.The 10 major problems with the anti-dumping instrument: An attempt at synthesis. Journal of World Trade 39: 67-73. Jackson, John H. 1997. The world trading system: Law and policy of international economic relations. Cambridge, MA: MIT Press. Marceau, Gabrielle (ed.). 1994. Anti-dumping and anti-trust issues in free-trade areas. Oxford: Clarendon Press. Mastel, Gregory. 1998. Antidumping laws and the U.S. economy. Armonk, NY: M.E. Sharpe. Viner, 1950. The customs union issue. New York: Carnegie Endowment for International Peace. 336 Anti-Dumping and Countervailing Duties ANNEX Annex Table 11.A.1: List of PTAs and AD/CVD rules (1/6) PTA Year Type AD CVD Andean Community (CAN) 1988 CU Rules Rules Aqadir Agreement No Rules No Rules Armenia - Kazakhstan 2001 FTA No Rules No Rules Armenia - Moldova 1995 FTA No Rules No Rules Armenia - Russian Federation 1993 FTA No Rules No Rules Armenia - Turkmenistan 1996 FTA No Rules No Rules Armenia - Ukraine 1996 FTA No Rules No Rules ASEAN - Australia - New Zealand 2010 FTA & EIA Rules* Rules* ASEAN - China 2005 FTA & EIA Rules* No Rules ASEAN - India 2010 FTA & EIA No Rules No Rules ASEAN - Japan 2008 FTA Rules* Rules* ASEAN - Korea, Rep. 2010 FTA & EIA Rules* Rules* ASEAN Free Trade Area (AFTA) 1992 FTA Rules* Rules* Asia Pacific Trade Agreement (APTA) 1976 PSA Rules* Rules* Asia Pacific Trade Agreement (APTA) Accession of China 2002 PSA No Rules No Rules Australia - Chile 2009 FTA & EIA Rules* No Rules Australia - China Rules* Rules Australia - New Zealand (ANZCERTA) 1983 FTA & EIA Prohibited No Rules Australia - Papua New Guinea (PATCRA) 1977 FTA Rules Rules Brunei Darussalam - Japan 2008 FTA & EIA Rules* Rules* Canada - Chile 1997 FTA & EIA Prohibited Prohibited Canada - Colombia 2011 FTA & EIA Rules Rules* Canada - Costa Rica 2002 FTA Rules No Rules Canada - Honduras 2014 FTA & EIA Rules* Rules* Canada - Israel 1997 FTA Rules* No Rules Canada - Jordan 2012 FTA Rules* Rules* Canada - Korea, Rep. 2015 FTA & EIA Rules Rules Canada - Panama 2013 FTA & EIA Rules* Rules* Canada - Peru 2009 FTA & EIA Rules Rules* Caribbean Community and Common Market (CARICOM) 1973 CU & EIA Rules Rules Central American Common Market (CACM) 1961 CU Rules Rules Central European Free Trade Agreement (CEFTA) 2006 2007 FTA Rules* No Rules Chile - China 2006 FTA & EIA Rules* No Rules Chile - Colombia 2009 FTA & EIA Rules* Rules* Chile - Costa Rica (Chile - Central America) 2002 FTA & EIA Rules* Rules* Chile - El Salvador (Chile - Central America) 2002 FTA & EIA Rules* Rules* Chile - Guatemala (Chile - Central America) 2010 FTA & EIA Rules* Rules* Chile - Honduras (Chile - Central America) 2008 FTA & EIA Rules* Rules* Chile - India 2007 PSA Rules* Rules* Chile - Japan 2007 FTA & EIA No Rules Prohibited Chile - Malaysia 2012 FTA Rules* Rules* Chile - Mexico 1999 FTA & EIA Rules* No Rules Chile - Nicaragua (Chile - Central America) 2012 FTA & EIA Rules* Rules* Chile - Vietnam 2014 FTA Rules* Rules* China - Costa Rica 2011 FTA & EIA Rules No Rules China - Hong Kong SAR, China 2004 FTA & EIA Prohibited Prohibited China - Korea, Rep. Rules Rules China - Macao SAR, China 2003 FTA & EIA Prohibited Prohibited China - New Zealand 2008 FTA & EIA Rules No Rules 337 Handbook of Deep Trade Agreements Annex Table 11.A.1: List of PTAs and AD/CVD rules (2/6) PTA Year Type AD CVD China - Singapore 2009 FTA & EIA Rules No Rules Colombia - Mexico 1995 FTA & EIA Rules Rules Colombia - Northern Triangle (El Salvador, Guatemala, Honduras) 2009 FTA & EIA Rules* Rules* Common Economic Zone (CEZ) 2004 FTA Prohibited No Rules Common Market for Eastern and Southern Africa (COMESA) 1994 CU Rules Rules* Commonwealth of Independent States (CIS) 1994 FTA Rules Rules Costa Rica - Colombia Rules* Rules* Costa Rica - Peru 2013 FTA & EIA Rules* Rules* Costa Rica - Singapore 2013 FTA & EIA Rules Rules Dominican Republic - Central America 2001 FTA & EIA Rules* Rules* Dominican Republic - Central America United States Free Trade Agreement 2006 FTA & EIA Rules* No Rules (CAFTA-Dominican Republic) East African Community (EAC) 2000 CU & EIA No Rules No Rules East African Community (EAC) - Accession of Burundi and Rwanda 2007 CU No Rules No Rules EC (10) Enlargement 1981 CU Prohibited Prohibited EC (12) Enlargement 1986 CU Prohibited Prohibited EC (15) Enlargement 1995 CU & EIA Prohibited Prohibited EC (25) Enlargement 2004 CU & EIA Prohibited Prohibited EC (27) Enlargement 2007 CU & EIA Prohibited Prohibited EC (9) Enlargement 1973 CU Prohibited Prohibited EC Treaty 1958 CU & EIA Prohibited Prohibited Economic & Monetary Community of Central Africa (CEMAC) 1999 CU No Rules No Rules Economic Community of West African States (ECOWAS) 1993 CU Rules No Rules Economic Cooperation Organization (ECO) 1992 PSA No Rules No Rules EFTA - Albania 2010 FTA Rules* Rules EFTA - Bosnia and Herzegovina 2015 FTA Prohibited Rules EFTA - Canada 2009 FTA Rules* Rules EFTA - Central America (Costa Rica and Panama) 2014 FTA & EIA Rules Rules EFTA - Chile 2004 FTA & EIA Prohibited Rules* EFTA - Colombia 2011 FTA & EIA Rules Rules EFTA - Egypt, Arab Rep. 2007 FTA Rules* Rules* EFTA - Hong Kong SAR, China 2012 FTA & EIA Prohibited Rules* EFTA - Israel 1993 FTA Rules* Rules EFTA - Jordan 2002 FTA Rules Rules EFTA - Korea, Rep. 2006 FTA & EIA Rules Rules EFTA - Lebanon 2007 FTA Rules* Rules EFTA - Mexico 2001 FTA & EIA Rules Rules* EFTA - Montenegro 2012 FTA Prohibited Rules EFTA - Morocco 1999 FTA Rules Rules EFTA - North Macedonia 2002 FTA Rules Rules EFTA - Peru 2011 FTA Rules Rules EFTA - SACU 2008 FTA Rules Rules EFTA - Serbia 2010 FTA Prohibited Rules EFTA - Singapore 2003 FTA & EIA Rules* Rules* EFTA - Tunisia 2005 FTA Rules Rules EFTA - Turkey 1992 FTA Rules Rules EFTA - Ukraine 2012 FTA & EIA Prohibited Rules EFTA - West Bank and Gaza 1999 FTA Rules Rules 338 Anti-Dumping and Countervailing Duties Annex Table 11.A.1: List of PTAs and AD/CVD rules (3/6) PTA Year Type AD CVD Egypt, Arab Rep. - Turkey 2007 FTA Rules* Rules* El Salvador - Cuba 2012 PSA Rules* Rules* El Salvador - Honduras - Taiwan, China 2008 FTA & EIA Rules Rules EU - Albania 2006 FTA & EIA Rules* Rules* EU - Algeria 2005 FTA Rules Rules* EU - Andorra 1991 CU No Rules No Rules EU - Bosnia and Herzegovina 2008 FTA Rules* Rules* EU - Cameroon 2009 FTA Rules Rules EU - CARIFORUM States EPA 2008 FTA & EIA Rules Rules EU - Central America 2013 FTA & EIA Rules Rules EU - Chile 2003 FTA & EIA Rules* Rules* EU - Colombia and Peru 2013 FTA & EIA Rules Rules EU - Côte d’Ivoire 2009 FTA Rules Rules EU - Eastern and Southern Africa States Interim EPA 2012 FTA Rules Rules EU - Egypt, Arab Rep. 2004 FTA Rules* Rules* EU - Faroe Islands 1997 FTA Rules No Rules EU - Georgia 2014 FTA & EIA Rules Rules EU - Iceland 1973 FTA Rules No Rules EU - Israel 2000 FTA Rules No Rules EU - Jordan 2002 FTA Rules No Rules EU - Korea, Rep. 2011 FTA & EIA Rules Rules EU - Lebanon 2003 FTA Rules* Rules* EU - Mexico 2000 FTA & EIA Rules* No Rules EU - Montenegro 2008 FTA & EIA Rules* Rules* EU - Morocco 2000 FTA Rules No Rules EU - North Macedonia 2001 FTA & EIA Rules No Rules EU - Norway 1973 FTA Rules No Rules EU - Papua New Guinea - Fiji 2009 FTA Rules Rules EU - Rep. of Moldova 2014 FTA & EIA Rules Rules EU - San Marino 2002 CU No Rules No Rules EU - Serbia 2010 FTA & EIA Rules* Rules* EU - South Africa 2000 FTA Rules Rules EU - Switzerland - Liechtenstein 1973 FTA Rules Rules* EU - Syrian Arab Republic 1977 FTA No Rules No Rules EU - Tunisia 1998 FTA Rules No Rules EU - Turkey 1996 CU Rules No Rules EU - Ukraine 2014 FTA & EIA Rules Rules EU - West Bank and Gaza 1997 FTA Rules No Rules EU – Overseas Countries and Territories (OCT) 1971 FTA No Rules No Rules EU (28) Enlargement 2013 CU & EIA Prohibited Prohibited Eurasian Economic Community (EAEC) 1997 CU No Rules No Rules Eurasian Economic Union (EAEU) 2015 CU & EIA Rules Rules Eurasian Economic Union (EAEU) Accession of Armenia 2015 CU & EIA No Rules No Rules Eurasian Economic Union (EAEU) Accession of the Kyrgyz Republic 2015 CU & EIA No Rules No Rules European Economic Area (EEA) 1994 EIA Prohibited Prohibited European Free Trade Association (EFTA) 1960 FTA & EIA Prohibited Prohibited Faroe Islands - Norway 1993 FTA Rules No Rules Faroe Islands - Switzerland 1995 FTA No Rules No Rules Georgia - Armenia 1998 FTA No Rules No Rules Georgia - Azerbaijan 1996 FTA No Rules No Rules Georgia - Kazakhstan 1999 FTA No Rules No Rules 339 Handbook of Deep Trade Agreements Annex Table 11.A.1: List of PTAs and AD/CVD rules (4/6) PTA Year Type AD CVD Georgia - Russian Federation 1994 FTA No Rules No Rules Georgia - Turkmenistan 2000 FTA No Rules No Rules Georgia - Ukraine 1996 FTA No Rules No Rules Global System of Trade Preferences among Developing Countries (GSTP) 1989 PSA No Rules No Rules Guatemala - Taiwan, China 2006 FTA & EIA Rules Rules Gulf Cooperation Council (GCC) 2003 CU No Rules No Rules Gulf Cooperation Council (GCC) - Singapore 2013 FTA & EIA Rules* Rules* Hong Kong SAR, China - Chile 2014 FTA & EIA Rules Rules Hong Kong SAR, China - New Zealand 2011 FTA & EIA Rules Rules* Iceland - China 2014 FTA & EIA Rules* Rules Iceland - Faroe Islands 2006 FTA & EIA Rules No Rules India - Afghanistan 2003 PSA Rules No Rules India - Bhutan 2006 FTA No Rules No Rules India - Japan 2011 FTA & EIA Rules Rules* India - Malaysia 2011 FTA & EIA No Rules No Rules India - Nepal 2009 PSA No Rules No Rules India - Singapore 2005 FTA & EIA Rules No Rules India - Sri Lanka 2001 FTA Rules* No Rules Israel - Mexico 2000 FTA Rules* Rules* Japan - Australia 2015 FTA & EIA Rules* Rules* Japan - Indonesia 2008 FTA & EIA Rules* Rules* Japan - Malaysia 2006 FTA & EIA Rules* Rules* Japan - Mexico 2005 FTA & EIA Rules* Rules* Japan - Mongolia Rules* Rules* Japan - Peru 2012 FTA & EIA Rules* Rules* Japan - Philippines 2008 FTA & EIA No Rules Rules* Japan - Singapore 2002 FTA & EIA Rules* Rules* Japan - Switzerland 2009 FTA & EIA Rules* Rules* Japan - Thailand 2007 FTA & EIA Rules* Rules* Japan - Vietnam 2009 FTA & EIA Rules* Rules* Jordan - Singapore 2005 FTA & EIA Rules Rules* Korea, Rep. - Australia 2014 FTA & EIA Rules No Rules Korea, Rep. - Chile 2004 FTA & EIA Rules* Rules* Korea, Rep. - Colombia Rules Rules Korea, Rep. - India 2010 FTA & EIA Rules Rules* Korea, Rep. - New Zealand Rules Rules Korea, Rep. - Singapore 2006 FTA & EIA Rules Rules* Korea, Rep. - Turkey 2013 FTA Rules Rules Korea, Rep. - US 2012 FTA & EIA Rules Rules Korea, Rep. - Vietnam Rules Rules Kyrgyz Rep. - Armenia 1995 FTA No Rules No Rules Kyrgyz Rep. - Kazakhstan 1995 FTA No Rules No Rules Kyrgyz Rep. - Moldova 1996 FTA No Rules No Rules Kyrgyz Rep. - Ukraine 1998 FTA No Rules No Rules Kyrgyz Rep. - Uzbekistan 1998 FTA No Rules No Rules Lao People’s Democratic Republic - Thailand 1991 PSA No Rules No Rules Latin American Integration Association (LAIA) 1981 PSA No Rules No Rules Malaysia - Australia 2013 FTA & EIA Rules Rules* Mauritius - Pakistan 2007 PSA Rules* Rules* Melanesian Spearhead Group (MSG) 1994 PSA Rules Rules MERCOSUR - India 2009 PSA Rules* Rules* Mexico - Central America 2012 FTA & EIA Rules* Rules* 340 Anti-Dumping and Countervailing Duties Annex Table 11.A.1: List of PTAs and AD/CVD rules (5/6) PTA Year Type AD CVD Mexico - Panama Rules* Rules* Mexico - Uruguay 2004 FTA & EIA Rules* Rules* New Zealand - Malaysia 2010 FTA & EIA Rules Rules* New Zealand - Singapore 2001 FTA & EIA Rules Rules* New Zealand - Taiwan, China 2013 FTA & EIA Rules* Rules* Nicaragua - Taiwan, China 2008 FTA & EIA Rules Rules North American Free Trade Agreement (NAFTA) 1994 FTA & EIA Rules Rules Pacific Alliance No Rules No Rules Pacific Island Countries Trade Agreement (PICTA) 2003 FTA Rules Rules Pakistan - China 2007 FTA & EIA Rules* Rules* Pakistan - Malaysia 2008 FTA & EIA Rules* Rules* Pakistan - Sri Lanka 2005 FTA Rules* Rules* Pan-Arab FTA (PAFTA) 1998 FTA Rules No Rules Panama - Chile 2008 FTA & EIA Rules* Rules* Panama - Costa Rica (Panama - Central America) 2008 FTA & EIA Rules Rules* Panama - Dominican Republic Rules Rules* Panama - El Salvador (Panama - Central America) 2003 FTA & EIA Rules Rules* Panama - Guatemala (Panama - Central America) 2009 FTA & EIA Rules Rules* Panama - Honduras (Panama - Central America) 2009 FTA & EIA Rules Rules* Panama - Nicaragua (Panama - Central America) 2009 FTA & EIA Rules Rules* Panama - Peru 2012 FTA & EIA Rules* Rules* Panama - Singapore 2006 FTA & EIA Rules Rules* Panama - Taiwan, China 2004 FTA & EIA Rules Rules* Peru - Chile 2009 FTA & EIA Rules* Rules* Peru - China 2010 FTA & EIA Rules No Rules Peru - Korea, Rep. 2011 FTA & EIA Rules Rules Peru - Mexico 2012 FTA & EIA Rules Rules Peru - Singapore 2009 FTA & EIA Rules* Rules* Russian Federation - Azerbaijan 1993 FTA No Rules No Rules Russian Federation - Belarus - Kazakhstan 1997 CU No Rules No Rules Russian Federation - Serbia 2006 FTA Rules Rules Russian Federation - Tajikistan 1993 FTA No Rules No Rules Russian Federation - Turkmenistan 1993 FTA No Rules No Rules Russian Federation - Uzbekistan 1993 FTA No Rules No Rules Singapore - Australia 2003 FTA & EIA Rules Rules* Singapore - Taiwan, China 2014 FTA & EIA Rules* Rules* South Asian Free Trade Agreement (SAFTA) 2006 FTA Rules* Rules South Asian FTA (SAFTA) Accession of Afghanistan Rules Rules South Pacific Regional Trade and Economic Cooperation Agreement (SPARTECA) 1981 PSA Rules Rules Southern African Customs Union (SACU) 2004 CU Rules Rules Southern African Development Community (SADC) 2000 FTA No Rules No Rules Southern African Development Community (SADC) - Accession of Seychelles No Rules No Rules Southern Common Market (MERCOSUR) 1991 CU & EIA Rules Rules* Switzerland - China 2014 FTA & EIA Rules Rules Thailand - Australia 2005 FTA & EIA Rules No Rules Thailand - New Zealand 2005 FTA & EIA Rules* Rules* Trans-Pacific Partnership 2006 FTA & EIA Rules Rules* Trans-Pacific Strategic Economic Partnership Rules* Rules* Treaty on a free trade area between members of the CIS 2012 FTA Rules Rules Turkey - Albania 2008 FTA Rules* No Rules 341 Handbook of Deep Trade Agreements Annex Table 11.A.1: List of PTAs and AD/CVD rules (6/6) PTA Year Type AD CVD Turkey - Bosnia & Herzegovina 2003 FTA Rules No Rules Turkey - Chile 2011 FTA Rules* No Rules Turkey - Georgia 2008 FTA Rules* Rules* Turkey - Israel 1997 FTA Rules No Rules Turkey - Jordan 2011 FTA Rules* Rules* Turkey - Mauritius 2013 FTA Rules* No Rules Turkey - Montenegro 2010 FTA Rules* No Rules Turkey - Morocco 2006 FTA Rules Rules Turkey - North Macedonia 2000 FTA Rules No Rules Turkey - Serbia 2010 FTA Rules* No Rules Turkey - Syrian Arab Republic 2007 FTA Rules Rules* Turkey - Tunisia 2005 FTA Rules Rules* Turkey - West Bank and Gaza 2005 FTA Rules No Rules Ukraine - Azerbaijan 1996 FTA No Rules No Rules Ukraine - Belarus 2006 FTA Rules* No Rules Ukraine - Kazakhstan 1998 FTA No Rules No Rules Ukraine - Moldova 2005 FTA Rules* Rules* Ukraine - Montenegro 2013 FTA & EIA Rules No Rules Ukraine - North Macedonia 2001 FTA Rules* No Rules Ukraine - Tajikistan 2002 FTA No Rules No Rules Ukraine - Turkmenistan 1995 FTA No Rules No Rules Ukraine - Uzbekistan 1996 FTA No Rules No Rules US - Australia 2005 FTA & EIA Rules* Rules* US - Bahrain 2006 FTA & EIA Rules* Rules* US - Chile 2004 FTA & EIA Rules* Rules* US - Colombia 2012 FTA & EIA Rules* Rules* US - Israel 1985 FTA Rules* Rules* US - Jordan 2001 FTA & EIA No Rules No Rules US - Morocco 2006 FTA & EIA Rules* Rules* US - Oman 2009 FTA & EIA Rules* Rules* US - Panama 2012 FTA & EIA Rules* Rules* US - Peru 2009 FTA & EIA Rules* Rules* US - Singapore 2004 FTA & EIA Rules* Rules* West African Economic and Monetary Union (WAEMU) 2000 CU No Rules No Rules * Weak Rules 342 Technical Barriers to Trade CHAPTER 12 Technical Barriers to Trade A. Espitia, S. Pardo, R. Piermartini, and N. Rocha 343 Handbook of Deep Trade Agreements CHAPTER 12 Technical Barriers to Trade A. Espitia*, S. Pardo†, R. Piermartiniø, and N. Rocha* * World Bank, Washington, DC, United States † Batalla, San José, Costa Rica ø World Trade Organization, Geneva, Switzerland CONTENTS 12.1. INTRODUCTION 345 12.2. A NEW DATABASE ON TECHNICAL BARRIERS TO TRADE (TBT) 349 12.3. PATTERNS OF TBT INTEGRATION 353 12.3.1 How has the content of TBT provisions evolved over time? 354 12.3.2 Does the level of integration vary across regions? 357 12.3.3 What are the most common provisions in legally enforceable TBT agreements? 358 12.3.4 Are there families of PTAs? 360 12.3.5 Who integrates TBT provisions the most? 361 12.4. CONCLUSIONS 363 ACKNOWLEDGMENTS 364 REFERENCES 365 344 Technical Barriers to Trade 12.1. INTRODUCTION At a time when tariffs are progressively being reduced, there is concern that a less transparent form of protectionism may be replacing them. The number of product lines and shares of trade value covered by non-tariff measures (NTMs) increased between 1996 and 2008,1 and the number of NTMs introduced since the 2007-2009 financial crisis has increased as well. Some estimates show that NTMs are a more important component of trade costs than tariffs. One study2 finds that, averaging across countries, NTMs almost double the level of trade restrictiveness imposed by tariffs. NTMs take many forms: quotas, burdensome customs procedures, price controls, and technical barriers to trade (TBTs), including standards, technical regulations, and conformity assessment procedures. Standards and technical regulations specify the technical characteristics of a product (such as the level of safety of an electronic device). Conformity assessment procedures define the testing necessary to ensure that products adhere to established norms. There is evidence that technical barriers to trade and sanitary and phytosanitary (SPS) measures are more prevalent than other types of non-tariff measures.3 The average country imposes TBT measures on about 30 percent of traded products, and imposes SPS measures on about 15 percent.4 Surveys by the International Trade Centre (ITC) suggest that businesses perceive technical requirements and conformity assessment measures as the most burdensome forms Figure 12.1: TBT-related specific trade concerns submitted to WTO, 1995-2017 Source: WTO, Technical Barriers to Trade Information Management System (TBTIMS). 1 WTO (2012). 2 Kee et al. 2009. 3 UNCTAD 2012. Data collected for 30 developing countries plus the European Union and Japan suggest a significant prevalence of TBT and SPS measures over other NTMs. 4 WTO (2012). 345 Handbook of Deep Trade Agreements of NTMs, and that requirements are increasing. The incidence (frequency and coverage ratios) of NTMs show an upward trend in the number of SPS and TBT measures notified to the World Trade Organization (WTO).There is also a growing number of complaints known as Specific Trade Concerns (STCs), which may signal that these measures are increasingly perceived as trade-distortive. Figure 12.1 shows the cumulative number of STCs submitted to the WTO between 1995 and 2017. In general, technical regulations and conformity assessments are introduced not for the purpose of trade protection, but to further specific policy objectives such as health and safety. Conformity assessments are needed to evaluate a product, process, or service against specified requirements. Firms can also set their own standards to achieve efficiency or production objectives; or set compatibility standards to be able to mix and match alternative inputs, reduce inventory costs, and increase production flexibility. Firms can also set standards to enable them to exploit economies of scale, or signal to consumers the quality of their products. In all of these cases, standards fulfill a legitimate objective and are aimed at helping markets operate more efficiently. Yet technical regulations and conformity assessment procedures can also inhibit trade, and in some cases may be designed with that objective. Technical barriers to trade can impose disproportionate costs on foreign producers or create a disadvantage for foreign competitors. Although the same regulations apply to foreign and domestic producers, TBTs may increase costs for foreign companies relatively more than for domestic firms. For example, adapting a product to a new technical requirement may require an initial investment independent of the level of sales. Exporters will face the fixed costs of interpreting the regulation and bringing the product into conformity, and might also have higher marginal costs if the requirement results in a decreased scale of operation. NTMs that affect both intensive and extensive margin of trade are especially harmful for small firms.5 Conformity assessment procedures, like technical regulations, can also impose considerable costs on exporters. The 2011 National Trade Estimate Report on Foreign Trade Barriers (NTE Report) - an annual survey carried out by the United States Trade Representative to identify foreign barriers to US exports - offers several examples. It notes, for instance, that “Thailand imposes food safety inspection fees in the form of import permit fees on all shipments of uncooked meat. Currently, imports face fees of 5 baht per kilogram (approximately $160 per ton) for red meat (beef, buffalo meat, goat meat, lamb, and pork) and for offal, and 10 baht per kilogram ($320 per ton) for poultry meat. Fees for domestic meat inspections are much lower and are levied in the form of a slaughtering or slaughterhouse fee. The fees are $5 per ton for domestic beef; $21 5 Fontagné et al. 2015. This study focused primarily on SPS measures. 346 Technical Barriers to Trade per ton for poultry; $16 per ton for pork; and zero for offal.”6 Other costs imposed by conformity assessments include the cost of complying with administrative requirements, the cost of delays for inspection by the importing country authorities, the risk that the goods will be rejected by the importing country, and the cost of transporting them back. Lengthy certification procedures can also be an important obstacle to trade. For example, the 2011 NTE Report relates US industry concerns about lengthy approval procedures in Hong Kong SAR, China, related to new pharma products, which inhibits the ability of US firms to market products on a timely basis. Similarly, the NTE Report raises a concern over Paraguay’s “non-automatic import licenses on personal hygiene products, cosmetics, perfumes and toiletries, textiles and clothing, insecticides, agrochemicals, and poultry. Obtaining a license requires review by the Ministry of Industry and Commerce and sometimes by the Ministry of Health. The process is slow, taking up to 30 days for goods that require a health certification. Once issued, the certificates are valid for 30 days.”7 Over the past few decades, countries have reduced many obstacles to trade through multilateral, bilateral, and regional trade agreements. Most multilateral agreements are under auspices of the WTO. The WTO agreement that deals with technical barriers to trade in goods is the TBT Agreement, which went into force in 1995.8 This agreement recognizes the right of each country to take measures necessary to pursue national security; prevent deceptive practices; or protect human health or safety, animal or plant life, or health or the environment. However, these measures need be as non-disruptive to trade as possible. To this end, the TBT Agreement establishes a set of commitments among WTO Members, including on principles of integration, transparency, institutional and administrative setup, and cooperation among countries. First, it calls for WTO Members to use existing international standards as a basis for their technical regulations, and for Members to play a full part in the preparation of international standards by international standardization bodies (Article 2.4). Second, the TBT Agreement encourages countries to accept as equivalent the technical regulations of other Members if these regulations adequately fulfill the objectives of their own domestic regulations (Article 2.7). Third, the TBT Agreement encourages Members to enter into mutual recognition agreements (MRAs); that is, to recognize the tests and certifications of another country (Article 6). Fourth, the TBT Agreement requires that, before the adoption of a new technical regulation, Members publish and notify the WTO Secretariat (Article 2.9), and that a focal point exist in each country to satisfy reasonable enquiries and provide documents (Article 10). Fifth, the TBT Agreement 6 Extracted from WTO 2012, Box D5 7 Extracted from WTO 2012, Box D5. 8 Other WTO agreements containing standards-related provisions are the Agreement on the Application of Sanitary and Phytosanitary Measures (SPS Agreement) and the General Agreement on Trade in Services (GATS). The SPS Agreement applies to any measure designed to protect human or animal or plant life or health. The GATS contain standards-related provisions on services, specifically in Article VI, paras. 4 and 5. 347 Handbook of Deep Trade Agreements establishes that countries shall provide technical assistance to other WTO Members (Article 11). Finally, the TBT Agreement explicitly refers to the Dispute Settlement Body for consultation on TBT matters and resolutions of disputes (Article 14). In addition, the TBT Agreement establishes the Committee on TBT handle the administration of the Agreement (Article 13). The number of preferential trade agreements (PTAs) that include TBT provisions has been increasing over time. More than 90 percent of regional agreements included TBT-provisions in 2017, compared to 75 percent in 1995 and 82 percent in 2005 (see Figure 12.2). All agreements that have entered into force since 2007 include TBT provisions. In fact, the increasing importance of TBT obstacles to trade acts as a strong incentive for countries to enter into PTAs. Figure 12.2: Number of PTAs that include TBT provisions, 1958-2017 Source: Deep Trade Agreements Database. There are currently more than 260 PTAs that deal with preferential TBT rules, but there has been no comprehensive mapping of the TBT rules set by these PTAs. This chapter aims to fill this gap to enhance our understanding of the legal framework in which global trade takes place. By mapping the TBT provisions in all existing PTAs, we are able to address such questions as: (a) How much deeper than the WTO rules on TBTs do the PTA rules go? (b) What is the most common approach, and how have rules evolved over time? and (c) What economic or political economy factors explain the inclusion of specific provisions in PTAs? Previous research in this area has covered only a fraction of PTAs with TBT provisions. A mapping carried out in 2015, which covers 171 PTAs, provides the most granularity.9 For example, it distinguishes between enforceable and non-enforceable provisions, such as those 9 Piermartini and Budetta 2006 mapped TBT provisions in 58 PTAs. Molina and Khoroshavina 2015 mapped 171 PTAs. Data from the 2015 mapping can be found at http://rtais.wto.org/UI/PublicSearchByCr.aspx. 348 Technical Barriers to Trade that the parties agree not to submit for regional dispute settlement.10 Further, it not only captures whether a provision is deeper than the corresponding WTO provision, but also how much deeper. The database distinguishes among the cases in which the parties agree to promote the use of international standards, promote the creation of a compatible regional standard, or comply to a specific regional standard. The 2015 mapping also includes sector-level information on exclusion of certain sectors form the agreement or the applicability of the integration approach to a specific industry. Finally, the mapping of TBT disciplines in PTAs is complemented with the mapping of bilateral and regional MRAs to provide a more complete picture of the preferred method of integration in the area of conformity assessment.11 The rest of the chapter is organized as follows. Section 12.2 describes the methodology used to map the information contained in the text of PTAs, using a template developed for an earlier mapping exercise.12 It also gives some illustrative examples of how PTAs address the questions in the template. Section 12.3 describes the results that emerge from the mapping of TBTs in 269 preferential trade agreements, and provides descriptive statistics about the different approaches used across PTAs to reduce TBTs. It then analyzes the patterns of TBT integration to investigate (a) the evolution of TBT provisions in PTAs; (b) whether there are families of PTAs that share common characteristics; and (c) whether some factors are correlated with the characteristics of these agreements (for example, whether level of development and similarity in income levels among PTA members are correlated with the depth of the agreements). Section 12.4 concludes. 12.2. A NEW DATABASE ON TECHNICAL BARRIERS TO TRADE (TBT) The study maps technical barriers to trade in 269 preferential trade agreements that entered into force between 1960 and 2017. The mapping template is based on the provisions of the WTO TBT Agreement to allow for easy examination of the extent to which regional preferential rules on TBTs have progressed beyond WTO rules. The information collected relies solely on the legal text of the agreements.While this approach ensures comparability across PTAs, it does not take into account their practical implementation. The template (Table 12.1) identifies five types of provisions: those that (a) refer directly to WTO rules; (b) define the type of integration approach (harmonization or mutual recognition) chosen for standards, technical regulations, and conformity assessment procedures; (c) improve 10 In some trade agreements, such as Republic of Korea-Australia, Panama-Singapore, Canada-Costa Rica, Honduras-El Salvador-Taiwan, China, EFTA-Chile, and EFTA-Mexico, the parties agree not to have recourse to the PTA’s dispute settlement mechanism in the case of TBT-related disputes. 11 The dataset on the content of TBT provisions also includes a list of 51 mutual recognition agreements signed by countries that are part of a preferential trade agreement. 12 Piermartini and Budetta 2009. The template was developed for a survey of 70 PTAs signed by the EU, EFTA, US, and Mexico. 349 Handbook of Deep Trade Agreements transparency; (d) establish institutions or mechanisms to administer the agreement and solve disputes; and (e) foresee cooperation among regional partners on standards-related issues beyond trade-related targets and technical assistance. Section I of the template is designed to test whether a PTA is consistent with or goes beyond the provisions in the WTO’s TBT Agreement. The last question, “Does the agreement go beyond the TBT Agreement in terms of covered areas or sector-specific commitments?” evaluates whether the TBT disciplines included in the PTA go beyond the WTO commitments in terms of new areas covered and sector-specific commitments. For example, the Colombia-Mexico agreement is classified as going beyond the WTO since it includes commitments in areas such as labeling and marking: “Specific provisions are established for: (a) health protection, (b) risk assessment, (c) handling of hazardous substances and (d) labeling.”13 The US-Singapore agreement is also classified as going beyond the WTO, as it explicitly creates a committee to promote TBT integration in Medical Products: “The Parties establish the Medical Products Working Group and include the APEC work program on Standards and Conformance.”14 Section II of the template describes the type of integration approach - equivalence/mutual recognition or harmonization - adopted or encouraged by the PTAs. On mutual recognition, subsection (i) assesses whether the standards of a country are automatically accepted by the PTA partner; i.e., whether mutual recognition is in force. For example, all intra-EU agreements specify “the effective application of all Schengen rules in accordance with the agreed common standards and with fundamental principles.”15 Subsection (i) also collects information on whether the importing country needs to provide reasons for not accepting a standard as equivalent.This is not required by the WTO TBT Agreement, which states countries “shall give positive consideration to accepting as equivalent” technical regulations of another country, “provided that they are satisfied that these regulations adequately fulfil the objectives of their own regulation.”16 This requirement is, however, included in the Australia–US agreement, which states that “where a Party does not accept a technical regulation of the other Party as equivalent to its own, it shall, at the request of the other Party, explain its reasons. The Parties will, if they so agree, give further consideration to whether a Party should accept a particular regulation as equivalent to its own and consider establishing an ad hoc working group.”17 Other questions in the area of mutual recognition include whether countries commit to negotiating mutual recognition agreement of conformity assessment, for example, within a certain timeframe, and whether PTA member countries must participate in international accreditation agencies. 13 Free Trade Treaty between Colombia and Mexico. 1994. Articles 14-3, 14-14, 14-15 and 14-16. 14 United States – Singapore Free Trade Agreement. 2003. Chapter 6, Article 6.3 15 Enlargement of the European Union to 28-member states - Accession of Croatia. 2011. Article 4.2 16 Agreement on Technical Barriers to Trade. 1995. Article 2.7 17 A PTA is consider to promote harmonization if it includes language such as “the parties agree to bridge the gap, reduce divergence, or make compatible” their standards, technical regulations or conformity assessment procedures.” 350 Technical Barriers to Trade With regard to harmonization,18 subsection (ii) of the template includes questions about whether the PTA agreements: (a) define the standards to which parties shall harmonize; (b) promote the use or creation of regional standards;19 or (c) promote the use of international standards. A commitment to using international standards is deemed to exceed WTO rules, since those rules require only that WTO Members use existing international standards as a basis for their technical regulations. The Common Market for Eastern and Southern Table 12.1: Structure of the template I. Reference to WTO-TBT Agreement Does the agreement refer to the WTO TBT Agreement? Does the agreement use the same definitions as the TBT Agreement? Does the agreement go beyond the TBT Agreement in terms of coverage or sector-specific commitments? II. Integration Approach (A) Standards, (B) Technical Regulations, (C) Conformity Assessment (i) Mutual Recognition/Equivalence - Is mutual recognition/Equivalence in force? - Is the burden of justifying non-equivalence on the importing country? - Is there a time schedule for achieving mutual recognition? - Do parties participate in international or regional accreditation agencies? (ii) Harmonization - Are there specified existing standards to which countries shall harmonize? - Is the use or creation of regional standards promoted? - Is the use of international standards promoted? III. Transparency Requirements (i) Notification - Is the time period allowed for comments specified? - Is the time period allowed for comments longer than 60 days? (ii) Contact Points/Consultations for Exchange of Information IV. Institutions (i) Administrative Bodies - Is a regional body established? (ii) Dispute Settlement Mechanism - Is there a regional dispute settlement body? - Are there regional consultations foreseen to resolve disputes? - Is there a mechanism to issue recommendations? - Are recommendations mandatory? - Is the recourse to the dispute settlement for technical regulations disallowed? V. Further Cooperation Among Members (i) Common policy/standardization program (beyond trade-related objectives) (ii) Technical Assistance (iii) Metrology 18 A PTA is considered to promote harmonization if it includes language such as “the parties agree to bridge the gap, reduce divergence, or make compatible” their standards, technical regulations or conformity assessment procedures.” 19 Agreements that promote harmonization are deemed to promote the creation of a regional standard. 351 Handbook of Deep Trade Agreements Africa (COMESA) agreement, for example, includes an explicit commitment to adopt specific standards: “The Member States undertake to [. . .] adopt African regional standards and where these are unavailable, adopt suitable international standards for products traded in the Common Market.”20 Other agreements, such as that of the Caribbean Community (CARICOM), promote the creation of bodies to advance the use of international standards: “The Community shall promote the establishment of a regional standardizing body which shall, inter alia: a. facilitate implementation of the standardization programme; b. assist Member States in understanding and fulfilling their obligations under this Treaty and other international agreements.”21 Section III of the template focuses on the transparency and notification requirements of the PTAs.22 If the notification and comment period for a new regulation is more than 60 days from the time it is promulgated until it enters into force, the PTA is deemed to exceed the transparency requirements of the WTO Agreement. The WTO requires only that there be a reasonable time interval for notification and comments for all new technical regulations, but does not set a specific time period. However, the WTO’s TBT Committee “has recommended that the normal time limit for presentation of comments on notified technical regulations and conformity assessment procedures should be sixty days.”23 In 2005, Members allowed an average of 60.5 days for notification and comments.24 Section IV concerns the institutional and administrative structures set up by the PTAs, including dispute settlement mechanisms. This section includes questions on (a) whether PTAs establish regional committees, bodies, or consultation platforms for the administration of the agreement; (b) how disputes are resolved; (c) whether there is a dispute settlement body; and (d) to what extent countries are required to comply with technical recommendations. The questions aim to capture any potential gap between commitments that appear in the legal text and their practical implementation. Finally, Section V of the template contains information on whether the TBT aspects of the PTAs include provisions aimed at promoting common (regional) policymaking on standards beyond trade- related objectives, as well as provisions related to metrology and commitments for technical assistance. The template also includes a cross-cutting set of questions aimed at assessing the level of 20 Common Market for Eastern and Southern Africa (COMESA). 1993. Article 113. Paragraph b. 21 Protocol Amending the Treaty Establishing the Caribbean Community (Protocol III: Industrial Policy). 1998. Article 12. Paragraph 5. 22 Transparency and notification are particularly important when members of a PTA have very different standards and neither equivalence nor harmonization are feasible. In these cases, countries can still minimize the trade-reducing effect of different standards by increasing the transparency of their national standards and technical regulations. 23 Agreement on Technical Barriers to Trade. 1995.Annex 3. Code of good practice for the preparation, adoption and application of standards. Article L. 24 World Trade Organization 2006. 352 Technical Barriers to Trade enforceability of each of the TBT provisions included in the PTAs. Five categories of enforceability are identified: - Level 0: Non-binding and best endeavor: these provisions represent the weakest level of enforceability; the language used in the TBT commitments (“may,” “might,” “should”) is not legally binding on the signatory parties. - Level 1: Binding TBT commitments: these provisions use stronger language (“shall,” “will,” “agree,” “undertake,”“ensure,” “realize”) but are not subject to dispute settlement (DS).This means that TBT provisions can be specifically excluded from an agreement that allows for dispute settlement on other issues. For example, the Japan-ASEAN FTA states that “The dispute settlement procedures provided for in Chapter 9 shall not apply to TBT.”25 - Level 2: Binding and subject to state-to-state dispute settlements: where the PTA provides for the use of state-to-state DS for TBT disputes, these disputes must be resolved directly between the countries without the involvement of any private parties.26 Examples include the Colombia-Mexico and Costa Rica-Peru preferential trade agreements. Almost the totality of agreements that are binding and subject to DS have state-to-state DS. - Level 3: Binding and subject to private-to-private dispute settlements: in PTAs with these provisions, the agreement’s DS mechanism can be used to resolve TBT disputes involving private parties. These types of provisions are rare, as private agreements are usually regulated by contracts. The study found no examples of PTAs with this level of enforcement. - Level 4: Binding and subject to state and private DS: the study found only one agreement, the Colombia-Northern Triangle FTA, that has both state-state and state-private DS. 12.3. PATTERNS OF TBT INTEGRATION This section characterizes the results of the TBT mapping presented in Section 12.2. It assesses the evolution of TBT provisions in PTAs over time and investigates whether there are families of PTAs that share common characteristics across regions. The section also aims at providing insights on the most common types of TBT provisions in regional agreements, and on whether they are correlated to specific characteristics of the member countries. 25 Agreement on Comprehensive Economic Partnership among Japan and member states of the Association of Southeast Asian Nations. 2008. Article 42. 26 A state-to-state dispute can be settled by a joint committee or an arbitration panel. If a company or private actor is affected by a TBT provision of the PTA, that party shall communicate with the national institution charged with administration of the agreement. 353 Handbook of Deep Trade Agreements 12.3.1 How has the content of TBT provisions evolved over time? The level of enforceability of agreements that include TBT provisions has increased over time. To assess the evolution of enforceability, agreements have been grouped into three broad categories based on the TBT mapping results (see Section 12.2): non-enforceable, including all non-binding agreements (category 0 in the template); weakly enforceable, including binding provisions without dispute settlement (category 1 in the template); and binding with dispute settlement, or strictly enforceable TBT provisions (categories 2-4 in the template). This last category includes all agreements with state-to-state DS, and one agreement with both state-to-state and state-to-private DS. No agreements with private- to-private DS have been identified. As shown in Figure 12.3, the share of agreements with legally binding language in TBT provisions has increased from 50 percent before 1999 to more than 60 percent during 2000-2004, and to 75 percent since 2005. To be enforceable, even legally binding commitments must be subject to dispute settlement. Before 2000, fewer than half of the legally binding TBT provisions were subject to dispute settlement, while after 2010, three-quarters of them could be taken into court. The remainder of the analysis in this section focuses on strictly enforceable TBT provisions. While the enforceability of TBT provisions in PTAs has increased, the share of enforceable agreements that explicitly exclude TBT provisions has also increased. The share of agreements with exclusions in the TBT chapter of SPS-related and public procurement-related sectors was one-tenth before 1995, rising to one fifth in 1999 and to more than 60 percent during the last decade (Figure 12.4). This increase can be explained by the incorporation of additional and more specific provisions in PTAs in areas such as SPS and technical standards for public procurement. Figure 12.3: Enforceability of TBT provisions over time 354 Technical Barriers to Trade Figure 12.4: Exclusion of specific provisions in TBT agreements over time More than 90 percent of TBT agreements currently in force cover all the sectors of an economy, while about 8.5 percent apply only to specific sectors, and 4.2 percent exclude certain sectors from TBT integration. Sectors such as copper, electrical and electronic equipment, and telecommunications are more typically excluded. In the case of copper, agreements signed by Canada with Chile, Panama, and Peru in 1997, 2009, and 2013, respectively, excluded copper and other reserves for national industry from the agreement. Other sectors that are typically excluded from TBT integration include mannitol, sorbitol, and essential oils. Other agreements (around 3.2 percent) aim to ensure compatibility in these sectors (Figure 12.5), especially in countries where the supply chains in these industries are integrated (Japan-Thailand, Japan-Singapore, Japan-Philippines, Republic of Korea- Singapore, Singapore-Australia, New Zealand-Singapore). To assess the evolution of TBT integration, we generate a variable that captures the coverage of TBT provisions and find that the average number of provisions included in TBT agreements has been increasing since 1995. While PTAs that entered into force between 1995 and 1999 included 7 TBT provisions, on average, those that became effective between 2010 and 2017 included 13 provisions (Figure 12.6). European Union enlargements tend to be outliers in each period, with around 30 TBT provisions. In addition to increasing in number, TBT-related commitments in trade agreements have become, on average, progressively deeper in terms of their content.27 We classify the provisions as as shallow or deep (index values 0 to 6) according to whether they are key to achieving deep TBT integration. To capture the deepening of TBT provisions over time, we count the number of deep provisions in TBT agreements for the years 1995-2017. 27 Deep provisions include those that refer to: Is mutual recognition in force? Are there specified existing standards to which countries shall harmonize? Is the use or creation of regional standards promoted? Is the burden of justifying non-equivalence on the importing country? Is the time period allowed for comments specified? Is the time period allowed for comments longer than 60 days? Are recommendations mandatory? 355 Handbook of Deep Trade Agreements Figure 12.5: Sector-specific coverage of TBT provisions Source: Deep Trade Agreements Database. Note: (1) Agreements with no TBT provisions are excluded. (2) “Other sectors” before 1995 refers to “Motor Vehicles, Agricultural and Forestry Tractors, Lifting and Mechanical Handling Appliances, Household Appliances, Gas Appliances, Construction Plant and Equipment, Other Machines, Pressure Vessels, Measuring Instruments, Electrical Material, Textiles, Foodstuffs” in the EEA (European Economic Area) agreement. “Other sectors” after 1995 refers to organic chemicals (HS 2905.43, 2905.44); essential oils (HS 33.01); albuminoidal substances; modified starches; glues (HS 35.01 to 35.05); chemical products nec (HS 3809.10, 3824.60); raw hides, skins, and fur (HS 4101 to 41.03, 4301); and textiles (HS 50.01 to 51.03, 52.01 to 52.03, 53.01 and 53.02) in the Turkey-Montenegro PTA. Figure 12.6: Average coverage ratio of TBT provisions in new PTAs over time Source: Deep Trade Agreements Database. Note: (1) Agreements with no TBT provisions are excluded. (2) Boxplot is a standardized way of displaying the distribution of data based on the five-number summary: minimum, first quartile, median, third quartile, and maximum. The central rectangle spans the first quartile to the third quartile, the bold segment inside the rectangle shows the median, and “whiskers” above and below the box show the locations of the minimum and maximum. Outliers are plotted as individual points. 356 Technical Barriers to Trade Figure 12.7 shows that the average depth of TBT agreements almost doubled between 1995 and 2017. More recent agreements include a larger set of deeper commitments such as the definition of more than 60 days for comments, completion of recommendations, and mutual recognition of conformity assessments in the area of TBT integration. Figure 12.7: Average depth of TBT provisions in new PTAs over time Source: Deep Trade Agreements Database. Note: Agreements with no TBT provisions are excluded. 12.3.2 Does the level of integration vary across regions? During 2017, the European Union was a party to 36 agreements that include legally enforceable TBT provisions. The East Asia and Pacific (EAP) and Latin America and Caribbean (LAC) regions were involved in around 50 agreements that include legally enforceable TBT provisions, of which more than half were involving countries outside their region. Other regions with more than 25 agreements that include legally enforceable TBT provisions are the Middle East and North Africa (MENA), the European Free Trade Association (EFTA), and Eastern Europe and Central Asia (ECA), with 28, 36, and 37 agreements, respectively. The United States and Canada were, respectively, parties to 12 and 11 agreements, including the North American Free Trade Agreement (NAFTA), with legally enforceable TBT provisions. On average, European countries have signed agreements with deeper TBT provisions, largely due to their EU membership. European agreements in force and with a legally enforceable TBT chapter during 2017 had an average depth of 6 out of 8 (Figure 12.8). Agreements with a legally enforceable chapter on TBT signed by the United States and by Central American countries also tend to establish relatively deep relationships (average depth level of 3). PTAs signed by middle- and high-income countries in Latin America and East Asia such as Colombia, Peru, and Chile; and New Zealand, Korea, Japan, and China have levels of depth that vary between 2 and 4. In the Africa region, the Common Market for Eastern and Southern Africa (COMESA) agreement includes 8 core TBT provisions. 357 Handbook of Deep Trade Agreements Figure 12.8: TBT depth across countries, 2017 Source: Deep Trade Agreements Database. Note: Only enforceable agreements are shown. Agreements with no TBT provisions are excluded. 12.3.3 What are the most common provisions in legally enforceable TBT agreements? To remove technical barriers to trade, trade agreements must include solid institutional and administrative arrangements. Of the 183 agreements with a legally enforceable TBT chapter in existence in 2017, fully 97 percent establish a regional body, and 98 percent contain provisions for dispute settlement (Figure 12.9). However, only 44 percent of the latter include deep provisions that are subject to mandatory dispute settlement. Figure 12.9: Percentage of PTAs by TBT provision, 2017 Source: Deep Trade Agreements Database. Note: (1) Agreements with no TBT provisions are excluded. (2) Only legally enforceable agreements are considered. 358 Technical Barriers to Trade Although most agreements have provisions for transparency on TBT, very few have deep commitments on this issue.Transparency reduces costs and makes it more difficult for countries to introduce discriminatory regulation. Overall, 69 percent of the enforceable agreements include transparency provisions. However, these commitments are generally shallow: only 2 percent of agreements that require contact points and consultations for exchange of information have core provisions specifying the time period allowed for comments. Provisions related to cooperation in areas other than transparency are generally shallow and are included in fewer than 40 percent of enforceable agreements. About 37 and 39 percent of agreements include, respectively, commitments related to technical assistance and metrology. Only 14 percent include commitments to other common policy objectives such as the establishment of a Common Market. Integration of conformity assessment (testing) measures is more common than integration in standards and technical regulation. While 60 percent of agreements include enforceable provisions regarding conformity assessment, fewer than 50 percent include enforceable provisions for the integration of technical regulations (47 percent) and standards (36 percent). This pattern holds when only deep provisions are taken into account: while 57 percent of agreements contain deep provisions in conformity assessment, 43 and 25 percent, respectively, contain deep provisions in the areas of technical regulations and standards. With respect to the preferred integration approach, while equivalence and mutual recognition are more prevalent for the integration of technical regulations and conformity assessment procedures, harmonization is the preferred integration approach for standards. While 37 percent of the agreements mutually recognize their conformity assessment procedures through the inclusion of deep provisions, only 13 percent harmonize them and 7 percent accept both types of integration Figure 12.10: Percentage of PTAs by preferred integration approach, 2017 Source: Deep Trade Agreements Database. Note: (1) Agreements with no TBT provisions are excluded. (2) Figure considers only legally enforceable agreements. (3) Preferred approach means that a core provision is included in the PTA or MRA. 359 Handbook of Deep Trade Agreements (Figure 12.10).28 Equivalence is also the preferred integration approach for technical regulations: 26 percent of agreements include deep provisions on technical regulations mutually recognized, compared to 12 percent that harmonize and 4 percent that use both the approaches. In the area of standards, more than two-thirds of the agreements integrate through harmonization. 12.3.4 Are there families of PTAs? The preferred integration approach adopted by PTAs varies across regions, as does the pattern of deep commitments. Table 12.2 shows this pattern across regions. Provisions related with the institutional and administrative structures to deal with TBT concerns are common across the board. A major difference between North American and EU agreements is the integration approach preferred by member countries. North American PTAs usually favor equivalence/mutual recognition, whereas PTAs involving the EU tend to include harmonization provisions. The preference for harmonization in EU-related agreements is generally supported by the provision of technical assistance to help developing countries harmonize their products and processes. In general, intra-regional PTAs (among countries in the same region) tend to be deeper than extra-regional agreements. There are, however, cases in Central Asia and the Middle East where PTAs have deeper harmonization provisions in extra-regional agreements (Table 12.3). Table 12.2: Patterns of TBT integration across regions (percentage of PTAs/MRAs by provision and region) – only deep provisions, 2017 Source: Deep Trade Agreements Database. Note: (1) The most common provisions (those that occurred in over 60 percent of the cases) are shaded in dark green, the least common (those occurring in less than 40 percent of cases) are shaded in light green, and the rest (occurring between 40 and 60 percent of the cases are shaded in green. (2) Average depth is calculated as the average responses related to a specific provision. 28 It is important to highlight that one-third of agreements mutually recognize conformity assessment procedures through a mutual recognition agreement (MRA) rather than a preferential trade agreement. 360 Technical Barriers to Trade Table 12.3: Patterns of TBT integration in intra- and extra-regional agreements (percentage of PTAs by provision and region), 2017 Source: Deep Trade Agreements Database. Note: (1) The most provisions (those that covered over 60 percent of the cases) are shaded in dark green, the least (those covered less than 40 percent of cases) are shaded in light green, and the rest (covered between 40 and 60 percent of the cases) are shaded in green. (2) Average depth is calculated as the average responses related to a specific provision. 12.3.5 Who integrates TBT provisions the most? More similarly sized economies tend to have deeper agreements in terms of TBT integration. Economic theory suggests that the method and the extent to which TBTs are included in a PTA are likely to depend on the level of development of countries in the agreement.29 Economies that are more alike tend to share similar policy objectives and therefore similar types of standards, decreasing the cost of TBT integration. Higher- similarity indexes are associated with deeper TBT chapters (Figure 12.11). Geographic distance tends to influence TBT integration. For example, two families of EU agreements with developing countries can be distinguished. One group is with developing countries geographically close to the EU, such as those in the Mediterranean area. The other group of agreements is with geographically distant countries such as Chile and Mexico. The latter agreements do not include a provision requiring harmonization to European standards, while the former do. Agreements including countries with higher levels of development are likely to be deeper in terms of TBT integration. Figure 12.12 shows the average depth of agreements between countries with 29 Baldwin 2000. 361 Handbook of Deep Trade Agreements different levels of development, defined using the World Bank country classification for 2017. Higher levels of integration between developed countries might be explained by the fact that richer countries are associated with greater economic, social, and technological improvements and have more demanding consumers and higher environmental requirements. Figure 12.11: Correlations of similarity indexes with average number of provisions, 2017 Sources: WDI and CEPPI. Note: (1) Agreements with no TBT provisions are excluded. (2) GDP similarity is calculated following Helpman 1987: GDPi GDPj SIij=1-(GDP 2 2 i+GDPj) -( GDPi+GDPj ) where GDP is in real terms from date of entry into force. (3) Following Eager and Lanch 2008: Bilateral market size captures the economic size of countries in terms of their GDPs: BMSij=GDPi+GDPj. Figure 12.12: Number of TBT provisions, average over agreements in force during 2015 by level of development Source: Deep Trade Agreements Database. Note: (1) Agreements with no TBT provisions are excluded. (2) Developing countries are composed of low-income and lower-middle-income economies, while Developed countries are upper-middle-income and high-income economies. Low-income economies are defined as those with a GNI per capita, calculated using the World Bank Atlas method, of $1,005 or less in 2016; lower-middle-income economies have a GNI per capita between $1,006 and $3,955; upper- middle-income economies have a GNI per capita between $3,956 and $12,235; high-income economies are those with a GNI per capita of $12,236 or more 362 Technical Barriers to Trade While mutual recognition of standards is more likely to occur in agreements between developed countries, harmonization of standards is more common in agreements that include developing countries. Mutual recognition requires a certain degree of trust in another country’s ability to perform testing and adequately safeguard health and safety (Figure 12.13). Figure 12.13: Percentage of PTAs by TBT provision and development level, 2017 Source: Deep Trade Agreements Database. Note: Agreements with no TBT provisions are excluded. 12.4. CONCLUSIONS This chapter presents a new dataset on the content of TBT provisions in preferential trade agreements.The dataset covers a total 269 PTAs that entered into force between 1960 and 2017. The analysis of this new database reveals the following global patterns of TBT integration: • The level of enforceability of TBT provisions in trade agreements has increased over time. The share of PTAs with enforceable TBT provisions has increased from 50 percent before 2000 to 75 percent after 2010. Also, TBT agreements typically apply to all sectors: only a small share of them excludes specific sectors or apply to specific industries. • TBT-related commitments in trade agreements have become deeper over time. The depth of TBT agreements almost doubled between 1995 and 2017. Specifically, more recent agreements include a larger set of deeper commitments such as the definition of a time period for comments, mandatory completion of recommended actions, and mutual recognition of conformity assessments in the area of TBT integration. 363 Handbook of Deep Trade Agreements • Most agreements include provisions such as institutional and administrative structures and transparency. These provisions are generally less deep compared to commitments on integration of conformity assessments, technical regulations, and standards, which are present in 57 percent, 43 percent, and 25 percent, of TBT agreements, respectively. • While equivalence and mutual recognition are more prevalent for the integration of technical regulations and conformity assessments, harmonization is the preferred approach for standards. In addition, there are families of PTAs across regions. Specifically, while North American PTAs usually favor equivalence/mutual recognition, PTAs involving the EU tend to include harmonization provisions. • Larger and more similarly sized economies tend to have deeper TBT integration agreements, as do countries with higher levels of development.While mutual recognition of standards is more likely to occur in agreements between developed countries, harmonization of standards is more common in agreements that include developing countries.These findings are consistent with the increasing importance of TBT as obstacles to trade.They provide important insights into how countries have tackled TBT integration so far and how this process may evolve in the future. ACKNOWLEDGMENTS We would like to thank the Batalla Law Firm for their assistance with the construction of the database. A special thanks goes to Maggie Chen for her valuable comments as well as the participants in the World Bank’s workshop on the Evolution of Deep Trade Agreements for helpful comments and suggestions. 364 Technical Barriers to Trade REFERENCES Bacchetta, Marc, Cosimo Beverelli, Olivier Cadot, Marco ———. 2011. The WTO and preferential trade agreements: From Fugazza, Jean-Marie Grether, Matthias Helble, Alessandro co-existence to coherence. Geneva: World Trade Organization. Nicita, and Roberta Piermartini. 2012. A practical guide to trade policy analysis. Geneva: World Trade Organization. ———. 2012. Trade and public policies: A closer look at non-tariff measures in the 21st century. Geneva: World Trade Organization. Baier, S.L., and J.H. Bergstrand. 2007. Do free trade agreements actually increase members’ international trade? Journal of International Economics 71(1): 72-95. Baldwin, Richard. 2000. Regulatory protectionism, developing nations, and a two-tier world trade system. Discussion Paper Series 2574, Centre for Economic Policy Research, London. Egger, P., and M. Larch. 2008. Interdependent preferential trade agreement memberships: An empirical analysis. Journal of International Economics 76(2): 384-99. Fontagné, L., G. Orefice, R. Piermartini, and N. Rocha. 2015. Product standards and margins of trade: Firm-level evidence. Journal of International Economics 97(1): 29-44. Helpman, E. 1987. Imperfect competition and international trade: evidence from fourteen industrial countries. Journal of the Japanese and International Economies 1(1): 62-81. ITC (International Trade Centre). 2011. Sri Lanka: Company perspectives. Geneva: ITC. Looi Kee, H., A. Nicita, and M. Olarreaga. 2009. Estimating trade restrictiveness indices. Economic Journal 119(534): 172-99. Molina A. C., and V. Khoroshavina. 2015. TBT provisions in regional trade agreements: To what extent do they go beyond the WTO TBT Agreement? Staff Working Paper ERSD- 2015-09, Economic Research and Statistics Division, World Trade Organization, Geneva. Office of the U.S. Trade Representative. 2011. The 2011 national trade estimate report on foreign trade barriers. Washington, DC: U.S. Government Printing Office. Piermartini, Roberta, and Michele Budetta. 2009. A mapping of regional rules on technical barriers to trade. In Regional rules in the global trading system, edited by Antoni Estevadeordal, Robert Teh, and Kati Suominen. Cambridge: Cambridge University Press. World Trade Organization. 2006. Eleventh annual review of the implementation and operation of the TBT Agreement. Geneva: World Trade Organization. 365 Handbook of Deep Trade Agreements 366 Sanitary and Phytosanitary Measures CHAPTER 13 Sanitary and Phytosanitary Measures S. Stone and F. Casalini 367 Handbook of Deep Trade Agreements CHAPTER 13 Sanitary and Phytosanitary Measures S. Stone* and F. Casalini† * Organisation for Economic Co-operation and Development, Paris, France † Graduate Institute of International and Development Studies, Geneva, Switzerland CONTENTS 13.1. WHY DO SPS MEASURES MATTER? 369 13.1.1 Methodology 371 13.2. SPS PROVISIONS FOUND IN PTAs 372 13.2.1 General 372 13.2.2 SPS and WTO 373 13.2.3 Standards 374 13.2.4 Risk assessment 377 13.2.5 Audits/control inspection 378 13.2.6 Transparency 379 13.2.7 Institutions 380 13.2.8 International regulatory cooperation more generally 381 13.2.9 Other areas of cooperation 381 13.3. ANALYSIS 381 13.3.1 Consideration of WTO 382 13.3.2 Integration 384 13.3.2.1 Standards 384 13.3.2.2 Risk assessment 385 13.3.2.3 Audit/controls inspection 385 13.3.3 Transparency 386 13.3.4 Institutions 387 13.3.5 Further cooperation 387 13.4. CONCLUSIONS 388 ACKNOWLEDGMENTS 389 REFERENCES 390 368 Sanitary and Phytosanitary Measures 13.1. WHY DO SPS MEASURES MATTER? Food and agriculture exporters face a variety of costs when accessing markets for their goods. These costs go beyond tariffs and extend to a number of areas, including the cost of complying with a variety of regulatory measures and standards set by both public institutions and private organizations.While tariffs are generally published in national schedules and thus easily identified and quantified, these non-tariff measures (NTMs) can be more difficult to navigate and can have a much higher ultimate cost to suppliers. Among the most significant of the NTMs for food and farm products are sanitary and phytosanitary (SPS) measures. While SPS measures help protect producers and consumers (along with animals and plants) against certain health and safety risks, they are not always defined or implemented in an efficient or truly transparent manner. Procedures put in place under the WTO and, importantly for this work, preferential trade agreements (PTAs), have helped increase transparency. The WTO’s SPS Agreement provides a process through which member governments can work together to put forward more effective ways to enact and implement SPS measures. PTAs can play an important role in this process as well. Indeed, evidence has shown that, on average, PTAs can reduce the costs of SPS measures across trading partners.1 There are several ways that PTAs can contribute to the reduction of trade costs associated with SPS measures. First, the more PTAs streamline SPS requirements, the lower the compliance costs. Second, by providing better information on foreign products to consumers, PTAs tend to reduce the home bias among member countries. This translates into an increase in the demand for products covered by PTAs, thereby increasing output and potentially lowering the price impact of SPS measures. Finally, PTAs reduce protectionist-motivated distortions in the design of NTMs overall, including SPS. PTAs can also influence the development of SPS matters in a broader sense, by providing a structure within which to develop further trade-liberalizing agendas. This is done through provisions such as technical cooperation, sharing of information regarding the development of standards or regulatory approaches, or even joint development of standards and regulations. Indeed, there is evidence that mutual recognition of conformity assessment procedures within PTAs significantly lowers the trade costs associated with SPS measures.2 Recent research3 has found that including specific cooperative mechanisms in a PTA can have a large impact on its ability to increase trade flows, and that these mechanisms have the largest 1 Cadot and Gourdon 2016. 2 Cadot and Gourdon 2016. 3 Disdier et al. 2019. 369 Handbook of Deep Trade Agreements impact when they are legally enforceable. This work also found that SPS-related mechanisms promoting some level of regulatory cooperation are more significant than measures promoting cooperation on technical barriers to trade (TBT). SPS measures with legal enforceability and transparency have the strongest and most positive impact on trade flows. Of the large body of literature showing that SPS measures can have a significant impact on trade,4 more recent studies have found that the form in which SPS measures are incorporated into trade agreements also matters. However, much of this work is based on broad aggregate measures of SPS provisions and on a sub-sample of trade agreements. Comparisons across different agreements or across different SPS measures in agreements can be problematic. Thus, observations or conclusions often differ depending on the sample or time period examined. In order to maximize the performance of trade agreements, more detail is needed on exactly which provisions are most effective over time and in what circumstances. The work presented in this chapter is a step toward helping to complete this picture. The chapter outlines the methodology of documenting (i.e., coding) SPS measures within SPS provisions or chapters in all preferential trade agreements notified to the WTO over the past 5 decades, through 2016.The coding does not include any private standards.This work is part of the World Bank’s Deep Integration project, which aims to map the content of various disciplines across the entire set of PTAs to generate a new deep database on trade agreements. The coding presented here builds on and complements other efforts to map SPS provisions in PTAs,5 and efforts currently underway by the World Trade Institute (WTI).6 The chapter presents the coded measures and explanations of the process. It then provides some summary statistics and analysis of trends that result from that coding. This work goes beyond earlier efforts by increasing the granularity of measures coded, with the intention of providing the necessary information to pick up more precisely what types of provisions within SPS chapters are important in reducing trade costs.The coding attempts to capture whether and how strongly certain issues are promoted in the agreements, as well as allowing for cross-referencing regarding the context in which agreements mention certain issues - so, for example, that it is possible to see whether international standards are referenced as part of standards as well as part of risk assessment, and to what degree these provisions are binding. In addition, the provisions are coded so that future researchers attempting to gain insight into the degree to which cooperation on certain issues, such as regulation, are included in PTAs, will be able to cross-reference across the full set of agreements. Finally, the coding applies only to those issues covered in individual SPS chapters and does not cover SPS-related issues contained in other chapters. 4 Two recent efforts include Crivelli and Groeschl 2016 and Grant and Arita 2017. 5 Notably, Jackson and Vitikala 2016 and Piermartini and Budetta 2009. 6 See https://www.wti.org/research/res/ for more information. 370 Sanitary and Phytosanitary Measures SPS measures are applied to protect the health of a country’s citizens and ecosystems. According to Annex A of the WTO’s SPS Agreement,7 a sanitary and phytosanitary measure is any measure applied to: - protect animal or plant life or health within the territory of the Member from risks arising from the entry, establishment or spread of pests, diseases, disease-carrying organisms or disease-causing organisms; - protect human or animal life or health within the territory of the Member from risks arising from additives, contaminants, toxins or disease-causing organisms in foods, beverages or feedstuffs; - protect human life or health within the territory of the Member from risks arising from diseases carried by animals, plants or products thereof, or from the entry, establishment or spread of pests; or -prevent or limit other damage within the territory of the Member from the entry, establishment or spread of pests. The WTO Agreement states that SPS measures cover all relevant laws, decrees, regulations, requirements and procedures including, inter alia, end product criteria; processes and production methods; testing, inspection, certification, and approval procedures; quarantine treatments, including relevant requirements associated with the transport of animals or plants, or with the materials necessary for their survival during transport; provisions on relevant statistical methods, sampling procedures and methods of risk assessment; and packaging and labeling requirements directly related to food safety.8 It encourages the use of international standards and establishes a procedure to monitor the process of international harmonization. 13.1.1 Methodology This chapter draws upon the methodology of past analyses regarding agriculture and TBT provisions contained in PTAs.9 The database consists of all PTAs notified to the WTO, including annexes and side letters, when available. This covers a total of 283 agreements, of which 276 have been coded.10 Questions encompass the following broad categories: - Does the PTA contain SPS provisions? - Does the PTA affirm or reference the WTO SPS Agreement generally? - Does the PTA refer to or replicate the definition and or rules of SPS measures contained within the SPS Agreement? - Does the PTA contain details on the SPS provisions concerning standards, transparency, and risk assessment? - To what degree does the PTA encourage or require cooperation? 7 See https://www.wto.org/english/tratop_e/sps_e/sps_e.htm. 8 SPS Agreement, Annex A, paragraph 1. 9 Fulponi et al. 2011; Piermartini and Budetta 2009; Jackson and Vitikala 2016. 10 There are two agreements that include several bilateral treaties under broader headings (Chile-Central America, and Panama-Central America).Thus, eight treaties are redundant and have not been coded, leaving a total of 276 agreements. 371 Handbook of Deep Trade Agreements Each PTA is double-coded by separate researchers and results are cross-checked.The following section details how the provisions are coded for SPS chapters. Each agreement is treated as a separate entity and no attempts are made to cross-reference provisions within an agreement, nor across agreements. This is left to the discretion of future researchers. In addition, no attempts are made to judge the quality of a provision (only to note legal enforceability, as described below) nor the degree to which any provision is implemented. In general, the language used in the agreement is recorded as closely as possible. For example, if mutual recognition is referenced in standards, and equivalence is referenced in risk assessment, these are coded as different things.While in reality, these may be referencing the same procedure, the language the negotiators ultimately place in the agreement is respected and recorded. 13.2. SPS PROVISIONS FOUND IN PTAs 13.2.1 General The PTAs that include SPS provisions cover a variety of different substantive SPS areas. These range from entire chapters covering a comprehensive set of issues to agreements that contain a simple one-line reference to the WTO SPS Agreement. Most of the PTAs do not include a number of WTO SPS provisions nor detailed coverage of all the areas of the WTO SPS agreement. For example, provisions on standards and equivalence are present in many of the dedicated SPS chapters, but control and inspection provisions of the WTO Agreement are referred to less often. The project undertook coding of the complete list of PTAs as notified to the WTO as well as 30 additional Spanish language agreements. In order to be as consistent as possible with existing work and to contribute to the Deep Integration project, which codes across a number of different provisions, this work was informed by other SPS data collection efforts, notably the World Trade Institute and previous efforts undertaken by the WTO.11 The WTI has developed a new dataset, the Design of Trade Agreements (DESTA) database),12 which includes information on more than 620 PTAs across a number of provisions, including SPS. Much of the coding here uses concepts similar to those of these other efforts. The scope of the exercise is limited to those provisions or topics covered within the SPS chapter. That is, if an issue such as dispute settlement is covered outside the SPS chapter, it is not coded here. This allows the differentiation of topics within the context of the overall agreement, and also ensures that issues covered outside the topic of, but related to, SPS are not double-counted.The only exception is if there is direct reference in the SPS chapter to another chapter or annex. In this case, the provision is coded as if it were included in the SPS chapter. 11 Jackson and Vitikala 2016; Piermartini and Budetta 2009. 12 Dür et al. 2014. See https://www.designoftradeagreements.org/downloads/ for more information. 372 Sanitary and Phytosanitary Measures The SPS chapters or provisions are coded across three basic areas: - Existence is coded as a 1 if it is included in the agreement, 0 otherwise; - WTO is coded as to whether the contents of the provision are consistent with the WTO SPS Agreement (WTO =), go beyond what is included in the WTO SPS Agreement (WTO +), or fall short of what is included in the WTO Agreement (WTO -); and - Enforceability is rated on a scale of 1 to 3, where 1 is the weakest and 3 is the strongest. Each chapter of the Deep Integration project analyzes its subject with respect to the level of enforceability. In the context of the SPS chapter, this is interpreted more in the context of the binding agreement. Thus, an article or provision within the chapter is assigned a number representing the level of enforceability based on the binding nature of the language included. The three levels are: - When the provision is only suggested or noted. For example: “…shall endeavor to cooperate in the matters related to Harmonization.”13 This is considered the weakest level and is coded as (1); - When the provision clearly expresses a desire to engage or is encouraged. For example: “Where necessary and possible, the Parties agree that the provisions concerning special and differential treatment in the WTO SPS and TBT agreements are applicable to the trade between the Parties to this Agreement, including Pacific States that are not WTO members.”14 This is coded as (2); - When the language clearly indicates an obligation or understanding of undertaking action. For example: “Parties undertake to notify their proposed SPS measures to the contact points of the other Party at least sixty (60) days before they are adopted.”15 This is the strongest level and is coded as (3). The existence of a dispute settlement mechanism is coded separately. 13.2.2 SPS and WTO This section tracks the degree to which the WTO Agreement is recognized in a PTA. Each question is coded as a “1” if it is recognized and “0” otherwise.When reference is made to more general WTO agreements within the SPS chapter, or specific reference is made to agreements dealing with non-tariff measures and Article XX of the GATT, this is coded “1” as well. Does the agreement refer to the WTO SPS Agreement? Does the agreement use the same definitions as the SPS Agreement? Does the agreement use the same rules as the SPS Agreement? Are any specific annexes of SPS Agreement adopted? 13 Article 5-5, Japan-Mongolia 2016. 14 Article 36-4, EU-Papua-Fiji 2009. 15 Article 6-11, Peru-Singapore 2009. 373 Handbook of Deep Trade Agreements The PTAs are examined for detailed references to the WTO SPS Agreement and are coded across the four questions posed above. The first question establishes whether reference is made to the WTO Agreement. Many treaties simply mention one provision of the SPS Agreement without adopting the whole agreement.16 These are usually treaties involving parties that are not members of the WTO. Similarly, many treaties simply reaffirm the rights and obligations under the WTO SPS Agreement.17 13.2.3 Standards The next series of questions are coded with respect to how standards are addressed in the SPS chapter. Again, they are coded “1” if these standards are addressed and “0” otherwise. Do parties recognize the adaption to regional conditions (including regionalization, zoning and/or compartmentalization) Do parties reference international standards? Equivalence - Is equivalence recognized? - Are parties encouraged to take into account other parties’ standards when elaborating new standards? - Is the burden of justifying non-equivalence on the importing country? Mutual Recognition - Is mutual recognition recognized? - Is there a time schedule for achieving mutual recognition? Harmonization - Are there specified existing standards to which countries shall harmonize? - Is the creation of concerted/regional standards referenced? The coding attempts to capture the degree to which an agreement distinguishes between standards that are existing or new; international, regional, or domestic; or whether the agreement intends to develop new standards based on specific criteria. At the same time, the coding attempts to capture the nature of the cooperation on standards: equivalence, mutual recognition, or harmonization. Given that standards can often be used as a basis for regulation, the various forms in which standards are adopted in the PTA can still lead to the same regulatory outcome. The WTO defines equivalence for SPS measures as each party recognizing other parties’ measures as acceptable even if they are different, so long as an equivalent level of protection is provided.18 The focus is generally on regulatory outcomes; for instance, in monitoring 16 For example: “Member states shall, upon request, enter into consultation, with the aim of achieving agreements on equivalence of SPS in accordance with the WTO SPS Agreement,” Article 16 SADC Seychelles Accession 2015. 17 For example: “The Parties reaffirm and incorporate in this Chapter their existing rights and obligations with respect to each other under the SPS Agreement,” Article 6-4, Peru-Singapore 2009. 18 https://www.wto.org/english/thewto_e/glossary_e/glossary_e.htm. 374 Sanitary and Phytosanitary Measures the risk of disease or ensuring that a product is safe. Crucially, equivalence decisions are generally unilateral. Each party sets its own criteria for approval and terms of access. By contrast, mutual recognition is not a unilateral action but an agreement between two or more parties. The WTO does not define mutual recognition, but it can be thought of as each party keeping its own regulations and legal decisions but recognizing and upholding regulations and legal decisions taken by the other partners. This means that a product that is lawfully produced in the exporting country must be accepted in the importing state (unless it is considered to be a risk to public health). For example, the regulator of one party can test a product for certain agreed-to standards and this would be recognized as adequate by the other party or parties in the agreement. This approach saves the product from being tested multiple times. However, it does not harmonize the rules or involve one party in recognizing the other parties’ standards as adequate. The terms “equivalence” and “mutual recognition” are often used interchangeably. Parties can mutually recognize or consider as equivalent (a) the regulation itself; (b) the compliance techniques and/or risk assessment procedures, as well as the results of these processes; or (c) the regulation’s enforcement through judgments and arbitral awards.19 This work makes a distinction in the coding of equivalence and mutual recognition, depending on what term is used in the agreement (if the term mutual recognition is used, it is coded “1” for mutual recognition and “0” for equivalence). The rationale is that recognizing the equivalence of a standard is arguing that it is more or less the same, whereas mutual recognition contains no implied sameness, and more closely captures the idea of separate but equal. Equivalence focuses more on different standards that provide the same level of protection, whereas mutual recognition is less about the level of protection than about the product being lawfully produced in the exporting country and accepted in the importing country.20 Finally, for the purposes of this chapter, harmonization is coded when it is agreed that all parties apply the same standard or adopt the same regulation. This is a stronger interpretation of the term “harmonization” than in the WTO SPS Agreement (Article 3). The common standard referred to in the PTA can be an international standard defined by an international standard-setting organization or the national standard in force in one member country and adopted by others, or a regional standard that is commonly applied. The key difference is that there is only one standard at play. 19 Arvius and Jachia 2015; Kauffmann and Malyshev 2015. For a detailed analysis of mutual recognition and mutual recognition agreements, see Correia de Brito et al. 2016. 20 For further discussion of mutual recognition versus equivalence, see von Lampe et al. 2016. 375 Handbook of Deep Trade Agreements When several forms of cooperation are noted in an agreement, the prevailing approach is coded. For example, New Zealand-Singapore contains a mutual recognition provision that states “…where regulatory compliance is required and where there is equivalence of outcomes, each Party shall accept the standards of the other Party as equivalent to its own corresponding standards.”21 Thus, while the heading is “mutual recognition,” this provision is coded as equivalence. The main factor in determining whether these provisions are WTO=, WTO+, or WTO- lies in the level of enforcement. Given that the WTO Agreement encourages and supports harmonization of standards, any provision which is coded “1” for having references to harmonization, mutual recognition, or equivalence is WTO=. However, if the reference contains binding levels of enforcement, it is coded “3” and WTO+ For instance, the Australia- Thailand agreement states that “…the Parties shall endeavor to work towards the harmonization of SPS measures,”22 and is coded WTO=. Other treaties define a stronger level of enforceability for SPS measures without explicitly mentioning the word “harmonization.” This is the case for the EAEU treaty, which is coded as level of enforcement 3 and WTO+. 23 Other treaties explicitly refer to harmonization in accordance with international standards, with a level of enforceability that clearly tends toward harmonization and are coded as level 2 and WTO=.24 The most common example of treaties that relate harmonization to regional standards is the case of the EU. Even countries not a part of the EU often harmonize their SPS measures in accordance with EU standards.25 It is important to stress that not all the references to international standards entail a reference to harmonization.There are some treaties that reference international standards as the basis for equivalence,26 or for pest or disease-free areas.27 Further, treaties that reference international standards may differ in their level of enforceability. For instance, some treaties simply state that the parties shall “use as a reference international standards,”28 whereas others use a stronger 21 Article 42, New Zealand-Singapore 2001. 22 Article 6-05, Australia-Thailand 2015. 23 “Common veterinary requirements approved by the Commission shall be applied to goods and facilities subject to veterinary control,” Article 58-2, EAEU (italics added). 24 “To harmonize sanitary and phytosanitary measures as broadly as possible, the Parties shall base their sanitary and phytosanitary measures on international standards, guidelines and recommendations established by the Codex Alimentarius Commission (CAC), the World Organization for Animal Health (OIE).” Article 7-4, China- Switzerland 2008. 25 “Measures, concerning veterinary and phytosanitary control among the Parties, shall be harmonized on the basis of EU legislation.” Article 11-2, Turkey-Albania 2008. 26 ‘The Parties will perform consultations for the recognition of equivalence for SPS measures on the basis of international norms and recommendations.” Article 7-05, Peru-Mexico 2011. 27 “The decision to declare a pest disease free area will be made on the basis of the Zoosanitarian Code.” Appendix 4, Annex 4 Es 1-1, EU-Chile 2003. 28 “The Parties shall use international standards, or the relevant parts of international standards, as a basis for their technical regulations.” Article 51-01, China-Singapore 2003. 376 Sanitary and Phytosanitary Measures level of enforceability by stating that the SPS measures shall be applied “in conformity with international standards.”29 Some treaties use concepts other than harmonization, mutual recognition, or equivalence. This is the case for some agreements between the EU and Eastern European countries. For example, EU treaties with Georgia and Moldova use the notion of “approximation” rather than harmonization. There is no real consensus on the difference between these two ideas.30 For the purposes of this exercise, approximation is considered the same as harmonization but with a lower level of enforceability.31 For most treaties, the level of enforceability for mutual recognition provisions is low. The parties often promote mutual recognition for SPS measures by “trying to explore the possibilities of….,”32 or agree that “…the committee shall consider, as needed, the developments of guidelines and recommendations for mutual recognition agreements.”33 Only two PTAs make explicit reference to a mutual recognition agreement. Finally, most equivalence provisions also have a low level of enforcement. The parties simply agree to “give consideration to the recognition of equivalence if a request is raised by the other Party.”34 13.2.4 Risk assessment Do the parties recognize that SPS measures are based on documented and scientific (if not available, objective) evidence? Is the participation of interested parties referenced? - Is the burden of evaluating risk on the exporting country? - Is there reference to international standards/procedures? Under the WTO SPS agreement, members are permitted to assess the risk to human, animal or plant life, or health, and implement SPS measures accordingly. They are also permitted to ensure that the risk assessment techniques take into account those developed by relevant international organizations. This section reviews how risk assessment is addressed within the PTAs across the same enforcement levels as applied in the standards section. In most cases, the parties usually abide by the wording of the WTO. For instance, in the treaty between 29 “This agreement is applied to all the SPS measures that might affect trade … in conformity with the WTO SPS Agreement, Codex Alimentarius, IPPC, and OIE.” Article 6-01, Panama-Peru 2012. 30 For some jurists approximation and harmonization are synonyms and thus interchangeable. For others, there is ´ a subtle, but noted, difference. For more information, see Cemalovic ´ (2015). 31 “Georgia shall continue to gradually approximate its sanitary and phytosanitary, animal welfare and other legislative measures as laid down in Annex IV to this Agreement to that of the Union.” Article 55, EU-Georgia 2014. 32 Article 8-6, Chile-India 2007. 33 Article 504-2ci, Canada-Colombia 2011. 34 Article 5-6-2, Malaysia-Australia 2013. 377 Handbook of Deep Trade Agreements Mexico and Peru, the parties agree “The SPS measures will be based on scientific evidence…taking into account the norms, guidelines and recommendations of the relevant international organizations.”35 Other treaties establish a stronger level of cooperation by opening the door to participation of the parties in the risk assessment process of the other Party. However, those provisions are often accompanied by a low level of enforceability.36 The coding goes beyond what is in the WTO to assess the degree to which the risk assessment process is open and transparent. Thus, questions on allowing other interested parties, such as technical experts or business or civil society, are included. However, these provisions rarely have a strong level of enforceability.37 Finally, reference to standards is noted.These could include the parties’ acceptance of the OECD’s Principles of Good Laboratory Practices (GLP) for purposes of assessment and other uses relating to the protection of humans and the environment.38 13.2.5 Audits/control inspection Is there a provision on control inspection? Are there provisions for pre-certification processes for exporter firms? Are there provisions for advance rulings? Mutual Recognition - Is mutual recognition in force? - Does the importing party have the right to audit the exporting party’s competent authorities, inspection systems, or production procedures? Equivalence - Is the burden of justifying non-equivalence on the importing country? Harmonization Is the participation of interested parties referenced? - Are there specified existing standards to which countries shall harmonize? - Is the use or creation of regional standards promoted? - Is the use of international standards promoted? Similar to standards and risk assessment, the coding for audits and controls, as part of the more general area of conformity assessment, distinguishes among agreements that specifically reference mutual recognition, equivalence, and harmonization. In some agreements there is reference to the harmonization of inspection processes to a regionally defined standard.39 35 Article 7-6-1, Peru-Mexico 2011. 36 “The Parties will give the opportunity to the other Party to make comments on their risk assessment procedure in the conditions defined by the importing Party.” Article 6-7-4, Pacific Alliance 2014. 37 “The Parties will give the opportunity to the other Party to make comments on their risk assessment procedure in the conditions defined by the importing Party.” Article 6-7-4, Pacific Alliance 2014. 38 For more information see http://www.oecd.org/chemicalsafety/testing/oecdseriesonprinciplesofgoodlabora torypracticeglpandcompliancemonitoring.htm. 39 Article -03d of the Nicaragua-Taiwan, China, agreement notes the intention to harmonize standards for control inspection. 378 Sanitary and Phytosanitary Measures Special attention is paid to which party bears the burden of proof relating to the audit or inspection process. For harmonization, equivalence, and mutual recognition, it is always coded as WTO+, since the WTO SPS Agreement does not have any provision on harmonization, equivalence, or mutual recognition of audit/control and inspection processes.40 13.2.6 Transparency Is there a transparency provision? Is there a provision on exchange of information? Is there a provision on electronic publication? Is there a duty to translate the document into the language of the other party(ies)? Is there a limitation to the obligation to notify, for reasons of law enforcement, public interest or commercial interest? Do parties have to notify each other prior to the entry into force of a new standard or regulation? Is there a specified minimum time period for comments? Is there a derogation clause on the notification period for emergency? Does the agreement allow the participation of interested parties of the other party in the development of standards? Does the agreement specifically reference the participation of regulatory authorities of the other party in the development of standards? The coding makes a distinction between transparency and exchange of information. Transparency is defined by the WTO as “the degree to which trade policies and practices, and the process by which they are established, are open and predictable.”41 Coding for the exchange of information attempts to capture agreements that go beyond what is specified in the Annex B of the WTO’s SPS Agreement. Thus, for cases where there is an identified and binding process to notify the other party of information on trade policies and regulations, that provision is coded WTO+. PTA transparency provisions, as well, often integrate Annex B of the WTO SPS Agreement. Most agreements with this provision have a strong level of enforcement.42 Nevertheless, even though the transparency provision is an obligation stemming from the WTO SPS Agreement, some PTAs use language describing a lower level of enforceability. For instance, in the Chile and Mexico Agreement (two WTO members), the transparency provision does not contain language obligating the parties to any action in this area and thus is coded “1” and WTO-. 40 “From a date to be determined by the SPS Sub-Committee referred to in Article 65 of this Agreement, the Parties may agree on the conditions to approve each other’s controls referred to in Article 62(1)(b) of this Agreement with a view to adapt and reciprocally reduce, where applicable, the frequency of physical import checks for the commodities referred to in Article 60(2)(a) of this Agreement.” Article 63-5, EU-Georgia 2014. 41 https://www.wto.org/english/thewto_e/glossary_e/glossary_e.htm. 42 “It shall also provide information on measures according to the provisions of Annex B to the ASPS, and shall implement the relevant adjustment.” Article 8-10-1a, Guatemala-Taiwan, China, 2006. 379 Handbook of Deep Trade Agreements Those treaties stipulating an exchange of information have varying degrees of enforceability. Some PTAs discuss the potential to exchange information in the context of ad hoc committee decisions.43 For others, the parties clearly commit to an exchange of information on a more regular basis.44 13.2.7 Institutions Administrative Bodies Do the parties establish SPS contact/enquiry points? Do the parties establish an SPS committee? Is there a fixed periodic meeting for the committee? Is the SPS committee the designated first place for dispute resolution? Does the SPS committee have open proceedings? Do the parties establish a working group? Is there a mechanism to issue recommendations? Is there a mechanism mandated to issue administrative decisions? Is a body for administering the agreement established? Is recourse to dispute settlement for the SPS chapter disallowed? The coding under Institutions attempts to capture when structures are created as a direct result of the treaty. Thus there is a distinction made between committees and other forms of organization that the chapter or provision establishes. If a working party is established, even if it is given responsibilities typically associated with a committee, but no committee is established, the coding is “0” for the line on committees and “1” for the establishment of a working group.45 Agreements that set up a distinct institutional structure, while rare, are coded WTO+. The coding also explores the degree to which a committee is open by noting the existence of a provision allowing the possible participation of other groups or public interaction. This section also codes for the possibility of more formal interactions. For example, it asks whether there are provisions under which a body created by the treaty can issue decisions, and the level of commitment parties are expected to have to that decision.46 Finally, it codes for any specific instances that disallow the dispute settlement mechanism for SPS.47 43 “The SPS Coordinators’ functions shall include, among others: facilitating information exchange so as to enhance mutual understanding of each Party’s SPS measures and the regulatory processes that relate to those measures and their impact on trade in goods between the Parties.” Article 5-7-1c, Costa Rica-Singapore 2003. 44 “The Parties, through the contact points, shall exchange information relevant to the implementation of this Chapter on a uniform and systematic basis, to provide assurance, engender mutual confidence and demonstrate the efficacy of the programs controlled. Where appropriate, achievement of these objectives may be enhanced by exchanges of officials.” Article 87, New Zealand-China 2007. 45 See, for example, China-Singapore, Article 55, “Joint Working Groups.” 46 See Chile-Malaysia, Article 6.6. 47 For example, the Canada-Republic of Korea agreement specifically disallows dispute settlement.“This Chapter is not subject to Chapter Twenty-One (Dispute Settlement).” Article 5.4. 380 Sanitary and Phytosanitary Measures 13.2.8 International regulatory cooperation more generally Is there a general IRC clause/common policy/standardization program (beyond trade-related objectives)? Is there a provision on technical assistance? Is there a provision for technical consultation? A growing number of trade agreements have chapters on regulatory cooperation or other mechanisms to review issues in relation to the disciplines of the agreement. This section aims to identify mechanisms that support regulatory cooperation. For example, the Japan-Malaysia agreement states that the parties “… shall cooperate in the areas of SPS measures including capacity building, technical assistance and exchange of experts, subject to the availability of appropriated funds and the applicable laws and regulations of each Country.”48 While there is no specific research on the topic, it seems likely that such technical assistance or cooperation leads to or supports the development of cooperation across regulatory issues. 13.2.9 Other areas of cooperation Is there a provision on labeling, marking, and packaging? Is there a provision on traceability? Is coordination for participation in international or regional accreditation agencies referenced? Is testing data to be made available? Some agreements contain clauses or provisions that pertain to a number of other forms of cooperation. For instance, more recent treaties are likely to include provisions on the possibility of making testing data available.49 Other treaties have specific provisions on coordinated or joint participation in international fora. However, those provisions usually have a low level of enforcement.50 13.3. ANALYSIS The following analysis is based on the coding undertaken as described above. It is worth repeating that only topics found within the SPS chapter or provision are coded. If a topic related to SPS is dealt with in a chapter or provision separate and distinct from SPS, it is not coded. For example, the Japan-Australia Economic Partnership cites the use of international 48 Japan-Malaysia, Article 70, paragraph 2, 2006. 49 “That Party shall provide to the exporting Party in writing full explanations and supporting data used for the determinations and decisions covered by this Article.” Article 71, EU-Ukraine 2014. 50 “…committee will have the function to consult on issues positions and agendas for meetings at the Commission of SPS measures and other relevant organizations’ Article 23-2b, Turkey-Chile 2011. 381 Handbook of Deep Trade Agreements standards in TBT provisions, but not in its SPS chapter. Therefore, for SPS, the question Do parties reference international standards? has been has been coded as “0” or not found for SPS, for that agreement. 13.3.1 Consideration of WTO The majority of the 276 PTAs contain a general exception similar to Article XX(b) of the GATT, enabling the parties to adopt measures necessary to protect human, animal or plant life, or health, subject to the requirement that such measures are not applied in a manner which would constitute arbitrary or unjustifiable discrimination or a disguised restriction on trade. Such health-driven general exception clauses are clearly linked to sanitary and phytosanitary protection. Two hundred thirty-five, or 85 percent, of the 276 PTAs analyzed contain some form of SPS provision. Ninety-one of these agreements contain a detailed SPS chapter, while 85 agreements set out the parties’ SPS obligations in a more concise form, sometimes in one sentence.The SPS provisions found in the main text of the agreements are, in 42 agreements, complemented by annexes. Other instruments, such as Memorandums of Understanding,51 decisions,52 implementing arrangements,53 side letters,54 or joint statements,55 are also present in a number of agreements. With three exceptions,56 the preambles of the agreements analyzed do not explicitly reference SPS matters. The trend (Figure 13.1) shows that an increasing share of agreements include SPS provisions. By the 2010s, almost 98 percent of all agreements entering into force included an SPS provision. Just over half of the agreements (158) make specific reference to the WTO SPS Agreement. Of those, most (142) include specific reference to the Agreement’s rules, while a lesser amount mention definitions (63) or specific annexes (57) of the WTO Agreement. There is a strong distinction between countries, by level of development, over the inclusion of specific SPS provisions. Treaties are grouped into three broad categories: those where all parties are developing economies; those where at least one party is a developed economy (mixed); and those where all parties are developed economies (high income).Treaties between high-income members have the lowest incidence of SPS provision referenced, at 73 percent 51 See, for example, China-ASEAN. 52 See, for example, Southern Common Market (MERCOSUR). 53 See, for example, Trans-Pacific Strategic Economic Partnership. 54 See, for example, US-Australia. 55 See, for example, US-Morocco. 56 EC-OCT; Hong Kong SAR, China-New Zealand; US-Singapore. 382 Sanitary and Phytosanitary Measures Figure 13.1: Number of agreements with SPS provisions over time Source: Deep Trade Agreements Database. of all agreements, versus developing treaties, where 90 percent include such a provision (Figure 13.2). However, while very few of the treaties are between high-income economies, the majority of these contain an SPS provision. Harmonization57 and technical cooperation58 are other themes found rather frequently in these general SPS provisions, but the rest use these provisions to cover distinct issues. For example, the India-Nepal PTA limits itself to a reference to SPS certificates,59 the Iceland- Figure 13.2: Number of agreements with SPS provisions, by income Source: Deep Trade Agreements Database. 57 See, for example, COMESA, Articles 112 and 132; and CEFTA, Article 12. 58 See, for example, East African Community (EAC), Articles 105 and 108; and EC-Chile, Article 24. 59 India-Nepal, Article II.6. 383 Handbook of Deep Trade Agreements Faroe Islands agreement obliges the parties to establish border inspection posts,60 and many of the PTAs concluded by EFTA only state the general obligation to apply SPS regulations in a non-discriminatory manner and not to apply measures that have the effect of unduly obstructing trade.61 For example, EFTA-Colombia and EFTA-Peru state that the parties shall not use their SPS measures related to control, inspection, approval, or certification to restrict market access without scientific justification, and establish a forum for SPS experts. 13.3.2 Integration 13.3.2.1 Standards Only a small share of agreements specifically address international standards. Of the 63 agreements that do, only 10 have binding provisions (Figure 13.3). The majority, or 65 percent, are best endeavors. There are 69 agreements where equivalence in SPS standards is recognized, but again, the majority, 47 agreements, require only that countries use their best endeavor. Reference to specific existing standards to which countries shall harmonize occurs in 37 agreements, while mutual recognition is noted in just 15 agreements. There are 33 PTAs that allow for the creation of regional or specific standards, but more than half of these 33 suggest that such a course be undertaken when possible and do not require action. Figure 13.3: Standards cited by method and level of enforcement Source: Deep Trade Agreements Database. Standards provisions are rarely included in PTAs between high-income economies. The majority of these provisions, or 68 percent, are included in agreements between developing 60 Iceland-Faroe Islands, Article 6. 61 See, for example, EFTA-Israel, EFTA-Jordan, EFTA-Morocco, EFTA-West Bank and Gaza, EFTA-Colombia, and EFTA-Peru. 384 Sanitary and Phytosanitary Measures economies (Figure 13.4).When standards are included, they are most often addressed through equivalence provisions, but again, with low levels of commitment. Figure 13.4: Standards cited by method and level of development Source: Deep Trade Agreements Database. 13.3.2.2 Risk assessment Under provisions dealing with risk assessment, 56 PTAs specify that such procedures must be based on documented and scientific evidence, and the vast majority, over 85 percent, of these provisions are binding. In addition, most of these provisions are found in agreements among developing economies (58 percent). However, none of these 56 treaties reference the WTO Agreement. Only 26 agreements, less than 10 percent of the total number of agreements, include a reference to international standards under the area of risk assessment, and these cover the entire range of enforcement levels. Thus it may be inferred that the majority of PTAs rely on the WTO SPS Agreement to oversee risk assessment procedures, and it is only those countries that are not WTO members at the date of entry into force that specifically cite international standards in their treaties. There are very few agreements (only 9) that allow the participation of interested parties in the risk assessment process. And of these 9, there are no high-income economies. Four agreements between developing economies and five involving a mix of income levels include such a provision. But an open risk assessment process is a relatively recent development, with all but one agreement dating from 2005 and 4 dating from 2014. 13.3.2.3 Audit/controls inspection There are fewer than 50 agreements with a specific provision on audit/control procedures. However, the majority of those (42) are binding. Slightly more than half (27) are among 385 Handbook of Deep Trade Agreements developing economies and 20 between economies of different income levels.There is a small but significant number of agreements, 34, that allow the importing country to audit the exporting party’s competent authorities, inspection systems, or production procedures. Again, the majority of these provisions, 70 percent, are binding and evenly distributed between developing and mixed agreements. While almost half (47 percent) the agreements that allow the importing country to perform audits came into effect in the 2010s, this provision is also found in agreements dating back to the 1980s. 13.3.3 Transparency The importance of transparency in PTAs is well established, and is regularly noted as an important element of good regulatory practice. There are a relatively small number of agreements (64 in total) that contain a specific clause in the SPS chapter dealing directly with transparency. Given that many agreements do not have specific provisions dealing with transparency, this finding alone does not mean that transparency is not important in PTAs. Indeed, of the small number that have transparency provisions, the vast majority are binding (level 3 in Figure 13.5). More interesting, a larger number of SPS provisions within agreements (99) deal specifically with information exchange. When there is a provision for transparency or information exchange, it is almost always binding. There is also a small but significant number of agreements (18) that specifically stipulate the electronic exchange of information. However, in this case, it is often left to the best endeavors (level 2) of partners. As would be expected, the electronic exchange of information is a relatively new phenomenon, with the majority (67 percent) of agreements with this language coming into force after 2010. Figure 13.5: Distribution of level of enforcement across transparency provisions Source: Deep Trade Agreements Database. 386 Sanitary and Phytosanitary Measures Transparency provisions, like standards and risk provisions, are found predominantly in developing and mixed-income agreements. Information exchange is the most common form of transparency, and it occurs in agreements between developing economies and between mixed-income economies with equal frequency (Figure 13.6). Figure 13.6: Distribution of transparency provisions by income level Source: Deep Trade Agreements Database. 13.3.4 Institutions The number of agreements that establish a separate institutional structure related to SPS are few. There are 93 agreements that establish an SPS committee in the context of the PTA while a fewer number, 69, establish a working group.Very few (4) agreements establish a body to administer the SPS work within the agreement. There are 24 agreements that establish a mechanism mandated to issue administrative decisions. Half of the agreements that establish an SPS committee are dated after 2010 and are between developing economies. Only four agreements between high-income economies establish such a committee. Some agreements (27 in total) specifically designate the committee to be the first point for dispute resolution. The majority of these, 63 percent, are between developing economies. From the numbers presented here, while setting up a separate institutional structure is still relatively rare (only 34 percent of all agreements establish an SPS committee), the majority of PTAs that do are between developing economies. 13.3.5 Further cooperation Further cooperation is measured across a number of other dimensions.These include any regulatory cooperation mechanisms that are put in place (other than those already covered under areas such as inspection or risk assessment). In addition, technical assistance and consultations are also coded. 387 Handbook of Deep Trade Agreements There are 37 agreements that include language calling for common or consultative policy approaches, and of those, only two have binding commitments. In addition, there are 44 agreements calling for technical assistance (64 percent of which involve countries with different income levels), but only three have binding commitments. However, 74 agreements have a provision for technical consultation or cooperation, and half of those are between countries of mixed income levels. The level of commitment is fairly evenly spread across the three levels measured in this work, with a slightly higher share (40 percent) going to binding commitments (Figure 13.7). Figure 13.7: Provisions for technical cooperation/consultation by level of enforcement The other provisions are spread across levels of development, with the largest number of technical assistance (64 percent) and cooperation provisions (50 percent) being in the PTAs of mixed income levels. However, only 7 percent of agreements specifying technical cooperation are between high-income economies, and these are mainly non-binding provisions. 13.4. CONCLUSIONS The majority of trade agreements contain some reference to SPS measures, and these have been increasing over time. There is a clear distinction between the treatment of SPS provisions by level of economic development. Reference to SPS is found more frequently in agreements signed by developing economies, perhaps in an attempt to bolster inadequate domestic regulation. Provisions related to cooperation in regulatory measures are found in only a few agreements, but these are more recent and thus could indicate the beginning of a trend toward expanding 388 Sanitary and Phytosanitary Measures interest in international regulatory cooperation. Conversely, the recent agreements could indicate a trend toward competing regulatory approaches that could potentially introduce further heterogeneity into the rules of the trading system. The coding presented here attempts to provide an improved array of factors by which researchers can examine PTAs on matters dealing with SPS specifically and cooperation more generally. While studies have shown PTAs to have a positive impact on trade flows between member countries (Didier et al. 2019), the task is to understand which provisions have the largest impact. With data efforts such as the World Bank’s Deep Integration project, it will be possible to better tease out these impacts. ACKNOWLEDGMENTS   The authors wish to thank Juan C. Macias Gómez for coding assistance and many useful discussions; Clarisse Legendre for database work; and OECD colleagues, notably Martin van Lampe and Frank von Tongeren, as well as colleagues from the WTO, notably Lee Ann Jackson, for helpful comments. All remaining errors are our own. 389 Handbook of Deep Trade Agreements REFERENCES Arvius, C., and L. Jachia. 2015. Regulatory cooperation: A Kauffmann, C., and N. Malyshev. 2015. International regulatory Wikihow E15 initiative. http://e15initiative.org/publications/ cooperation: The menu of approaches. E15 Initiative ICTSD regulatory-cooperation-a-wikihow/. and World Economic Forum, Geneva. http://e15initiative.org/ wp-content/uploads/2015/09/E15-Regulatory-Kauffmann- Beghin, J., and J.C. Bureau. 2001. Quantitative policy analysis and-Malyshev-final.pdf. of sanitary and phytosanitary and technical barriers to trade. Economie Internationale 87:107-30. Piermartini, R., and M. Budetta. 2009. A mapping of regional rules on technical barriers to trade. In Regional rules in the Cadot, O., and J. Gourdon. 2016. Non-tariff measures, global trading system, edited by A. Estevadeordal, R. Teh, and preferential trade agreements and prices: New evidence. Review K. Suominen. Cambridge: Cambridge University Press. of World Economics 152: 227-49. doi:10.1017/CBO9780511814334.006. Ćemalović, Uroš. 2015, Framework for the approximation of von Lampe, M., K. Deconinck, and V. Bastien. 2016. Trade- national legal systems with the European Union’s Acquis: From related international regulatory cooperation: A theoretical a vague definition to jurisprudential implementation. Croatian framework. OECD Trade Policy Paper 195, OECD Publishing, Yearbook of European Law and Policy 11:241-58. Paris. https://doi.org/10.1787/3fbf60b1-en. Correia de Brito, A., C. Kauffmann, and J. Pelkmans. 2016. The contribution of mutual recognition to international regulatory co-operation. OECD Regulatory Policy Working Paper, OECD Publishing, Paris.http://dx.doi.org/10.1787/5jm56fqsfxmx-en. Crivelli, P., and J. Groeschl. 2016. The impact of sanitary and phytosanitary measures on market entry and trade flows. World Economy 39(3):444-73 (March). Disdier, A., S. Stone, and F. van Tongeren 2019. Trade and economic effects of IRC: Further empirical evidence from SPS and TBT provisions. OECD Trade Policy Paper 224, OECD Publishing, Paris. https://doi.org/10.1787/8648b6ca-en. Dür, A., L. Baccini, M. Elsig, and K. Milewicz. 2014.The design of international trade agreements: Introducing a new dataset. Review of International Organizations. https://doi.org/10.1007/ s11558-013-9179-8. Fulponi, L., M. Shearer, and J. Almeida. 2011. Regional trade agreements:Treatment of agriculture. OECD Food,Agriculture, and Fisheries Working Paper 44, OECD Publishing, Paris. Grant, Jason, and Shawn Arita. 2017. Sanitary and phyto- sanitary measures: Assessment, measurement, and impact. Commissioned Paper 259417, International Agricultural Trade Research Consortium, Saint Paul. Jackson, L.A., and H. Vitikala. 2016. Cross-cutting issues in regional trade agreements: Sanitary and phytosanitary provisions. In Regional trade agreements and the multilateral trading system, edited by Rohini Acharya. Cambridge and Geneva: Cambridge University Press and World Trade Organization. https://doi.org/10.1017/CBO9781316676493.009. 390 Public Procurement CHAPTER 14 Public Procurement A. Shingal and V. Ereshchenko 391 Handbook of Deep Trade Agreements CHAPTER 14 Public Procurement A. Shingal* and V. Ereshchenko† * European University Institute, Florence, Italy † World Bank, Washington, DC, Unites States CONTENTS 14.1. INTRODUCTION 393 14.2. LITERATURE REVIEW 395 14.3. NEW CLASSIFICATION TO MAP PROCUREMENT PROVISIONS IN PTAs 397 14.4. STYLIZED FACTS ON PROCUREMENT PROVISIONS IN PTAs 401 14.4.1 PTAs with no coverage of government procurement 403 14.4.2 PTAs with shallow government procurement provisions 404 14.4.3 PTAs with detailed provisions on government procurement 405 14.5. DETAILED ANALYSIS OF DEEP PROCUREMENT AGREEMENTS 410 14.5.1 Non-discrimination 410 14.5.2 Coverage 412 14.5.3 Procedural disciplines 414 14.5.4 Transparency (ex-ante and ex-post) 416 14.5.5 Dispute resolution 418 14.5.6 New issues 418 14.6. CONCLUSIONS 419 ACKNOWLEDGMENTS 421 REFERENCES 422 ANNEX 423 392 Public Procurement 14.1. INTRODUCTION Contestable government procurement markets account for an estimated 7-9 percent of GDP in developed1 countries and an estimated 9-20 percent in developing countries.2 Thus, the state has considerable influence over the allocation of resources in market economies through procurement. A prominent aspect of such procurement is the preference for domestic over foreign firms in the award of public contracts, regardless of cost and quality considerations.This home bias in public purchase decisions has nontrivial efficiency effects. A home bias can reduce trade flows and influence international specialization, especially in sectors where public demand is large relative to domestic output and which are characterized by monopolistic competition and increasing returns to scale.3 Given these adverse effects, non-discrimination in the award of public contracts is the cornerstone of most international rules on government procurement. These measures have included (a) efforts undertaken by the European Commission (EC) as a part of its internal market reform and deregulation programs; (b) the nonbinding proposals of the Asia-Pacific Economic Cooperation (APEC) and the model law proposed by the United Nations Commission on International Trade Law (UNCITRAL); (c) the plurilateral WTO Agreements on Government Procurement (GPAs);4 and (d) the rules already present and those being negotiated in various preferential trade agreements (PTAs). The last are the subject of this chapter. PTAs have become the main vehicle for extending procurement rules to countries not party to the GPAs.5 Existing work6 has identified over 40 PTAs that include commitments to open access to procurement contracts on a bilateral or regional basis and explicitly prohibit procurement practices that discriminate against foreign producers. PTAs with deep (or substantive) provisions on government procurement, which we refer to as deep procurement agreements (DPAs), have grown more popular over time, with more than half entering into effect since the year 2000. The proliferation 1 Trionfetti 2000, based on UN-OECD data. 2 OECD 2002. 3 Trionfetti 2000. 4 Especially the Uruguay Round GPA 1996 and Revised GPA 2012. 5 Hoekman 2015. 6 For instance, see Anderson et al. 2011, Ueno 2013, and Rickard and Kono 2014. 393 Handbook of Deep Trade Agreements of DPAs suggests that governments see them as a means for addressing procurement discrimination. DPAs explicitly forbid some or all forms of discrimination in public procurement. For instance, many forbid explicit “buy national” policies such as the 2009 “Buy American” provisions. These types of agreements also tend to prohibit price discrimination and a range of other policies such as local content requirements, which favor domestic firms. In general, though, PTAs vary greatly in their scope and coverage of procurement provisions. Some, in fact, either reflect the existing procurement policies of signatories or limit commitments to best-endeavor (non-binding, non-enforceable) clauses. Many of the more recent PTAs, however, include extensive procurement commitments and are also more enforceable, including through domestic bid-challenge mechanisms. As one study has noted, “The more ambitious PTAs go beyond commitments to remove discrimination in procurement and include language pertaining to the objectives of procurement policy (e.g., attaining best value for money); the use of new technologies, such as electronic procurement, provisions to create or strengthen national institutions that implement national procurement policies and associated reforms; how to address likely changes in the scope of transactions falling under the disciplines of the agreement as a result of privatization of government entities; and call for cooperation on the development of national procurement policies.”7 Given the proliferation of DPAs, attempts have been made to map procurement provisions in the full range of PTAs. A review of this literature in the following section finds that these attempts, as pioneering as they are, could be more comprehensive. The main purpose of this chapter is therefore to describe a new method of classifying government procurement provisions in PTAs and present stylized facts, based on this classification, for 283 PTAs notified to the World Trade Organization (WTO) as of March 2017. The rest of the chapter is structured as follows. The next section reviews earlier attempts in the literature to map the coverage of government procurement in PTAs. Section 14.3 introduces and describes the new methodology developed in this chapter to map government procurement provisions in PTAs, while Sections 14.4 and 14.5 provide stylized facts and detailed analysis on the basis of this classification. Section 14.6 concludes. 7 Hoekman 2015. 394 Public Procurement 14.2. LITERATURE REVIEW Perhaps the first attempts to compare government procurement provisions in PTAs were by Bourgeois et al. (2007) and Heydon and Woolcock (2009). Bourgeois et al. (2007) provided a comparative legal analysis of government procurement provisions in 27 PTAs as of the mid-2000s, looking inter alia at the scope and coverage of the procurement chapters in these PTAs and commitments made on tendering, qualification of suppliers, time limits, bid challenge, dispute settlement, and institutional features. Heydon and Woolcock (2009) provided a qualitative summary of procurement provisions in agreements negotiated by the US, EU, EFTA, Japan, and Singapore. Horn et al. (2009) examined 14 EU and US PTAs with WTO members, dividing the agreements into 52 policy areas that they classified as WTO+ and WTO-X. For each agreement, the authors identified the areas that were covered and whether the obligations were legally enforceable. Public procurement was classified as a WTO+ policy area in their analysis, with 50 percent of the observations being legally enforceable in EU PTAs against 93 percent in US PTAs. Shingal (2009) classified 119 PTAs into Groups (I-V) and categories: Basic, Comprehensive (minus), and Comprehensive on the basis of the coverage of public procurement provisions in the PTAs. Illustratively, PTAs classified as “Basic” included generic provisions on opening up procurement markets on a non-discriminatory and reciprocal basis and for developing rules, conditions, and practices on government procurement. Most of the EC agreements with countries in the Mediterranean and Africa belong to this category. In contrast, most of the agreements that Mexico, Singapore, and the US have entered into, as well as the EFTA-Chile agreement, were classified as “Comprehensive.” The UN Social and Economic Commission for Asia and the Pacific codes the presence (“yes”) or absence (“no”) of government procurement provisions in 244 PTAs that involve an Asia-Pacific country. More recently, Rickard and Kono (2014) construct a variable to denote 43 PPAs notified to the WTO, where a PPA is defined as a PTA with substantive provisions on government procurement. A more comprehensive treatment of this subject is provided in Anderson et al. (2011), Ueno (2013), Dür, Baccini, and Elsig (2014), and Gourdon and Messent (2017). Anderson et al. (2011) provide three levels of analyses. First, they classify 139 PTAs into three broad categories: (i) agreements between GPA Parties; (ii) agreements between a GPA Party and a non-GPA Party; and (iii) agreements between non-GPA Parties. Within each category, they then distinguish between: (a) PTAs incorporating government procurement chapters/ related schedules or having some provisions that include the liberalization of procurement markets as an objective; and (b) PTAs that do not include such commitments. They find 87 395 Handbook of Deep Trade Agreements agreements falling into the former category and provide a more detailed analysis of these 87 agreements in their second level of analyses by looking at eleven specific types of provisions and their coverage in each agreement, providing for examples of such provisions and giving a statistical overview on the occurrence of each provision in the agreements covered. In their third level of analysis, the authors compare and contrast the coverage commitments on government procurement in PTAs with those of the WTO Uruguay Round (UR) GPA. The eleven specific types of provisions that Anderson et al. (2011) focus on include: (i) provisions on national treatment (NT) and non-discrimination; (ii) provisions on most-favoured-nation (MFN) treatment; (iii) procedural provisions analogous to the GPA; (iv) requirements for the implementation of bid challenge procedures; (v) the availability of dispute settlement procedures (i.e., enforceability); (vi) provisions regulating the use of offsets; (vii) commitments to GPA accession; (viii) commitments regarding further negotiations; (ix) provisions ensuring integrity in procurement procedures; (x) cooperation; and (xi) establishment of a Joint Committee or other administering body. In another comprehensive treatment of this subject, Ueno (2013) examines the extent to which PTAs go beyond the WTO’s government procurement agreements, both UR GPA (1996) and the Revised GPA (RGPA; 2012), in 47 Organisation for Economic Co- operation and Development (OECD) PTAs and finds non-GPA countries to have achieved the general GPA level of market access commitments in their PTAs. The study provides a detailed analysis of coverage commitments (by entity, thresholds, and goods and services coverage) of government procurement in these 47 OECD member PTAs and then examines procurement provisions in these agreements in much detail, also providing a comparison with the relevant WTO procurement agreements. Ueno (2013) too considers eleven specific features in her analyses: (i) General principles (NT/non-discrimination and prohibition of offsets); (ii) Mechanisms supporting multilateralization (third-party MFN and future negotiation clauses); (iii) Information on procurement systems and opportunities; (iv) Qualification criteria; (v) Criteria for contract award; (vi) Use of information technology; (vii) Time periods; (viii) Transparency of decisions on contract awards; (ix) Domestic review; (x) Prevention of corruption; and (xi) Others (SME participation). 396 Public Procurement Using the same parameters as Ueno (2013), Gourdon and Messent (2017) have recently expanded on her analysis by including 13 more agreements, including those amongst non- OECD members. The authors find little variation in coverage across a party’s agreements for its central government entities, although the number of schedules with commitments in sub-central coverage is found to be slightly greater.The thresholds in the 13 new agreements also appear to be negotiated on a reciprocal basis, and are found to be closely related to those agreed in each country’s existing agreements. None of these studies, however, code the procurement provisions that they look at into an index to enable a quantitative comparison of the coverage of government procurement in PTAs. In contrast, Dür, Baccini, and Elsig (2014) have assembled DESTA, a database that has coded 587 PTAs up until June 2013 on 11 instruments of deep integration: market access, services, investment, procurement, SPS,TBT, dispute settlement, competition, trade defense, IPRs, and non-trade issues. Dür, Baccini, and Elsig (2014) code procurement provisions in PTAs on the basis of the following attributes framed as questions: whether there are any substantive provisions on procurement; whether there is national treatment, transparency, and coverage in terms of entities and goods/services; and whether any reference has been made to the GPA. Each question is coded between 0-2 and then a final composite index adds the responses to the individual questions in each case. The greater is the score of the final composite index, the deeper is the PTA in its coverage and treatment of government procurement. According to their classification, about 50 percent of the agreements have a reference to government procurement, but only 14 percent include substantive provisions, i.e., those going beyond stating adherence to the GPA or the desire to exchange information in this area. 14.3. NEW CLASSIFICATION TO MAP PROCUREMENT PROVISIONS IN PTAs Building on the existing literature, this chapter develops a new classification for mapping government procurement provisions in PTAs. To do so, we draw on two recent studies; 8 on the WTO Agreements on Government Procurement - the 1996 Uruguay Round GPA, on which most of the existing coverage of procurement in PTAs is based, and the Revised GPA (RGPA; 2012); as well as on the texts of “comprehensive” procurement chapters in representative PTAs such as those between the US and Singapore and Australia and Singapore. 8 Ueno 2013, and Gourdon and Messent 2017. 397 Handbook of Deep Trade Agreements Our methodology follows a questionnaire approach, in which questions have two types of responses: either binary or detailed. This approach enables classification at the extensive (“does a PTA have a detailed government procurement chapter/provisions?”) and intensive (“what are the salient features of the government procurement chapter/provisions in the PTA?”) margins. Our classification is based on eight broad themes incorporating one hundred questions, which cover the salient features of government procurement chapters/ provisions found in PTAs. The questions represent desirable characteristics that proscribe discrimination in the award of public contracts and/or lead to better value of money for the government. The eight broad themes, with the number of questions for each theme in parentheses, are: • Overview (4) • Non-discrimination (14) • Coverage (40) • Procedural disciplines (26) • Transparency (ex-ante 3, ex-post 4) • Dispute settlement (4) • New issues (5) The Overview theme includes four questions with binary responses. These questions provide a broad overview of the coverage of government procurement in a PTA: • Are provisions covering government procurement explicitly mentioned in the agreement? • Are the procurement provisions enforceable? • Is government procurement coverage detailed in the agreement? • Is this an agreement between GPA signatories? Non-discrimination includes 14 questions with binary responses that address different aspects of existing and prospective non-discrimination in the award of government contracts: • Does the agreement contain explicit provisions on - national treatment? - prohibition of offsets? - Most-favored-nation (MFN) treatment of third parties? - future negotiation of third parties? - review of commitments to expand coverage (more entities, more goods and services, lower thresholds)? - review of commitments to progressively reduce/eliminate discriminatory measures? • Does the agreement require rules of origin not to be different for procurement compared to those applied in the normal course of trade? • Are transitional measures allowed for developing country members of the agreement? 398 Public Procurement • Do transitional measures explicitly allow - price preferences? - offsets? - phased-in addition of specific entities or sectors? - a threshold that is higher than the permanent threshold? - delayed implementation periods? • Does the agreement include provisions for the extension of transitional measures and/or transition periods? Coverage includes 40 questions that require both binary and detailed responses and are at the core of our methodology to classify procurement provisions in PTAs: • Does the agreement cover central and subcentral governments and/or utilities? • What are the numbers of each entity covered by the agreement? • Does the agreement cover goods and/or services? • What is the number of aggregate goods sectors at the HS29-digit Chapter level and the number of aggregate services sectors as listed in the GATS W/12010 covered by the agreement? • Which aggregate goods and services sectors are covered by the agreement? • Are the threshold values for each Annex,11 goods and services higher, lower, or the same as in the WTO’s RGPA? • What are the threshold values for each Annex, goods, and services? • Are threshold values adjusted for inflation? • Does the agreement include unnecessary exceptions from coverage except those permitted by the RGPA? • Does the agreement include elaborate provisions for modification/rectification of coverage? Ex-ante transparency includes three questions with binary responses: • Does the agreement contain explicit provisions requiring that information on the procurement system (laws and regulations) be published? 9 The Harmonized Commodity Description and Coding System generally referred to as “Harmonized System”or simply “HS” is an international nomenclature for the classification of products developed by the World Customs Organization. The HS comprises approximately 5,300 article/product descriptions that appear as headings and subheadings, arranged in 99 chapters, grouped in 21 sections. 10 W/120 is a comprehensive list of services sectors and subsectors covered under the WTO’s General Agreement on Trade in Services (GATS). 11 Under GPA rules, only public procurement above stipulated thresholds is subject to international competitive bidding. These thresholds vary by type of procuring entity (central government, subcentral government, and utilities) and for goods, services, and construction services. GPA Members report central government entities covered by the rules of the agreement under Annex 1; subcentral government entities under Annex 2; and utilities under Annex 3. 399 Handbook of Deep Trade Agreements • Does the agreement contain explicit provisions requiring that notice of the intended/ planned procurement be published? • Are the notice details of the intended/planned procurement consistent with the requirements of Article VII:2 of the RGPA? Procedural disciplines contain 26 questions that require both binary and detailed responses: • Does the agreement contain explicit provisions on - conditions of participation? - qualification of suppliers? - technical specifications? - tender documentation? - time periods and deadlines? - negotiations? - limited tendering? - electronic auctions? - treatment of tenders and award of contracts? - transparency of procurement information? - ensuring integrity in procurement practices; e.g., by avoiding conflict of interest? • Are the provisions in each case consistent with the requirements of the RGPA? • Do participation conditions prohibit imposing conditions of previous awards? • Do requirements for the qualification of suppliers impose any limitations on the number of bidders? include explicit provisions on using selective tendering and multi-use lists? • How many days does the agreement allow for tender submission? publication of award information? • Does the treatment of tenders allow for protection and proper use of confidential information and intellectual property (IP) protection? Ex-post transparency includes four questions with binary responses: • Are there explicit provisions on information provided to bidders (results and reasons for non-selection)? • Are there explicit provisions on information provided to third parties (disclosure of information)? • Does the agreement contain explicit provisions on collection and reporting of statistics? • Are the provisions on collection and reporting of statistics consistent with Article XVI:4 of the RGPA? Dispute resolution also includes four questions: • Does the agreement provide for domestic review procedures? • Are the domestic review procedures consistent with Article XVIII of the RGPA? • Does the agreement contain explicit provisions on dispute settlement? • Is dispute settlement consistent with Article XX of the RGPA? 400 Public Procurement Finally, on New issues found in some of the more recent PTAs, there are five questions that require binary responses: • Does the agreement contain explicit provisions facilitating - e-procurement? - sustainable procurement? - participation of small and medium enterprises (SMEs)? - adoption of safety standards? - cooperation (as in Article 15.22 of the Trans-Pacific Partnership, TPP) in matters of public procurement? The themes in this questionnaire, especially the coverage of public procurement in PTAs by entity, goods and services, and threshold values, and the incorporation of new issues, yields a more comprehensive understanding of PTAs than provided by the existing literature. The next section presents stylized facts on the 283 WTO-notified PTAs in force as of March 2017 that were analyzed according to the eight broad themes in our classification. 14.4. STYLIZED FACTS ON PROCUREMENT PROVISIONS IN PTAs All PTAs can be classified into three groups according to their coverage of government procurement: (a) no coverage at all; (b) provisions on government procurement exist but are not detailed; and (c) detailed provisions on government procurement are included in the agreement. Of the 283 PTAs in force as of March 2017, 129 agreements (about 45 percent) have no provisions on government procurement; 70 agreements (25 percent) have shallow provisions; and 84 agreements (30 percent) have deep provisions (i.e., the group we refer to as DPAs). A complete list of PTAs in each category is presented in Annex Table 14.A.1. The three groups of PTAs have evolved over time (Figure 14.1). The majority of agreements concluded before 1995 did not have any procurement provisions. The period from 1995 to 2004 witnessed a decline in the number of no-coverage agreements entering into effect, followed by a surge in those agreements during 2005-2009 and another decline after 2010. The period 2005-2009 also saw the highest number of shallow procurement PTAs, followed by a decline thereafter. In contrast, the number of DPAs grew steadily over the years until 2014, and reached a maximum of 29 agreements entering into effect during 2010-2014. A majority of the DPAs have been concluded among high-income country partners or involve at least one high-income country (Figure 14.2). In fact, there are 34 DPAs between high-income countries, followed by 20 DPAs between high-income and upper-middle-income countries, and 10 DPAs between high-income and lower-middle-income countries.Thus, high-income countries exhibit the greatest propensity to conclude DPAs relative to all other income groups. 401 Handbook of Deep Trade Agreements Figure 14.1: Evolution of PTA groups by government procurement coverage over time Source: Deep Trade Agreements Database. Figure 14.2: Breakdown of PTA membership by procurement coverage and income Source: Deep Trade Agreements Database. Note: (1) Income based on World Bank income classification (WBIC) for 2017. (2) The figure includes only bilateral PTAs. (3) All EFTA and EU members are classified as high-income countries. (4) These agreements comprise 223 of the 283 WTO-notified agreements. The cohort of shallow procurement agreements is dominated by PTAs in which one partner is a high-income country (or trade bloc) and the other partner is an upper-middle-income country (21 agreements). Another 15 shallow agreements are between high-income and lower-middle-income countries, while 10 PTAs are between high-income countries. The participation of upper-middle- and lower-middle-income countries in shallow procurement agreements is higher than their participation in DPAs. 402 Public Procurement In contrast, the group of agreements with no procurement coverage exhibits a greater involvement of lower-middle-income countries. This is also the only cohort that includes agreements involving low-income countries (Afghanistan-India and India-Nepal). In the following subsections, we provide more descriptive statistics for each of the three PTA groups. A summary of these stylized facts is provided in Table 14.1. Table 14.1: Summary of stylized facts on government procurement provisions in PTAs TOTAL NUMBER of PTAs: 129 70 84 Provisions Shallow provisions Deep provisions (%) (%) (%) Share of total PTAs 45 25 30 Of which: Share of North-North 10 16 27 Share of North-South 24 71 51 Share of South-South  66 13 22 Share of cross-regional 38 54 84 Share of goods-only PTAs 64 65 8 Share of services-only PTAs 0 0 1 Share of goods + services PTAs 36 36 90 Share of PTAs before 2000 41 19 7 Share of GPA signatories, both parties 11 14 30 Share of GPA-signatories, one party 38 59 51 Source: Deep Trade Agreements Database. Note: North = OECD countries; South = non-OECD countries. 14.4.1 PTAs with no coverage of government procurement Non-OECD countries are relatively reluctant to open their procurement markets via PTAs. Among the 129 agreements with no provisions on government procurement, 85 PTAs are between South-South trading partners (a little over 65 percent of 129 PTAs), while the shares of North-South and North-North PTAs are 24 and 10 percent, respectively (Table 14.1). A large number of the South-South agreements with no coverage of government procurement includes agreements among former Soviet countries - e.g., Armenia-Kazakhstan, Georgia- Azerbaijan, and Kyrgyz Republic-Ukraine. Among North-North agreements without any reference to government procurement, nearly half relate to treaties that have enlarged EU membership over time. While most EU enlargement agreements do not explicitly cover government procurement, EU regulations have internal directives that set forth a comprehensive framework regulating government procurement in the common market. 403 Handbook of Deep Trade Agreements Most of the PTAs with no coverage of procurement have been signed with a member from within the region. Of the 129 PTAs without procurement coverage, 49 agreements (about 40 percent) are cross-regional, while 80 agreements (about 60 percent) are intra- regional. The majority of no-procurement-coverage PTAs entered into effect in the period up to the year 2000. Of the 72 PTAs signed before the year 2000, 53 agreements (74 percent) include no provisions on government procurement. In contrast, of the 211 PTAs signed during January 2000-March 2017, a much lower share (76 agreements, or 36 percent) have no provisions on government procurement. Most signatories of such PTAs are not members of the GPA. Only 14 of the 129 PTAs with no government procurement provisions involve both parties that are signatories to the Agreements of Government Procurement. The remaining 115 PTAs have at least one party that is not a signatory to the GPA. Finally, the bulk of these agreements have been negotiated under Article XXIV of the GATT. Of the 129 agreements with no provisions on government procurement, 83 agreements (64 percent) cover only goods, while the remaining 46 PTAs cover both goods and services. 14.4.2 PTAs with shallow government procurement provisions On the whole, government procurement is explicitly mentioned in 154 of the 283 WTO- notified PTAs up to March 2017 (almost 55 percent). Of these, 70 PTAs (25 percent of all 283 agreements) have only a shallow coverage of procurement. Some PTAs with shallow coverage of government procurement have only a single article (rather than a chapter) on the subject, and no binding commitments. The Tukey-Israel FTA (Box 14.1) is an example of a shallow procurement agreement. Box 14.1. Turkey-Israel FTA: An example of a shallow procurement agreement Article 24. Public Procurement 1.The Parties to this Agreement consider the effective liberalization of their respective public procurement markets an integral objective of this Agreement. 2. The Joint Committee will review progress in this area annually. In contrast, some agreements with shallow coverage of government procurement have a full chapter on the subject that addresses a number of issues (though the coverage is limited as compared to a GPA). For example, the agreement between Japan and Mongolia has a limited chapter on public procurement that covers procurement principles, exchange of information, further negotiations, and negotiations on non-discrimination, and provides for a subcommittee on government procurement. The group of shallow procurement PTAs is dominated by North-South agreements, which represent 50 of the 70 PTAs. Such agreements as US-Jordan, EFTA-Morocco, and Thailand- 404 Public Procurement Australia belong to the group of North-South PTAs with shallow provisions on government procurement. Only 11 PTAs between high-income countries fall into this group, including EFTA- Israel, EU-Turkey, and EU-Israel; as well as 9 South-South agreements, such as the Commonwealth of Independent States (CIS), the Melanesian Spearhead Group (MSG), and the West African Economic and Monetary Union (WAEMU).The shares of South-South and North-North PTAs within the shallow procurement PTAs are 13 and 16 percent, respectively (Table 14.1). Of the 70 shallow procurement agreements, the distribution is relatively balanced between cross-regional and intra-regional agreements (38 and 54 agreements, or 54 and 46 percent, respectively). Countries have tended to devote more consideration to government procurement in their PTAs concluded in the last two decades. Thirteen of the 70 shallow procurement accords (19 percent) were signed before 2000, and 57 (81 percent) in the period after 2000. In the majority of shallow procurement agreements, at least one party is not a signatory to the GPA, and only around 14 percent of the PTAs (10 by number) in this group are between parties that are both GPA signatories (e.g., EFTA-Turkey, EFTA-Israel, EU-Montenegro, Ukraine- Moldova, and EFTA-Montenegro). The bulk of these agreements have also been negotiated under Article XXIV of the GATT. Almost 65 percent of the shallow procurement agreements cover only goods, while the rest (25 PTAs) cover both goods and services. Box 14.2. Incorporation of GPA provisions in the Canada-Republic of Korea FTA Article 14.3: Scope This Chapter incorporates by reference the rights and obligations as listed in the Annex to the WTO Protocol Amending the GPA (hereinafter referred to as the “revised GPA”), with the exception of Articles V and XVIII through XXII. These rights and obligations apply mutatis mutandis to the procurement covered by Annexes 14-A through to 14-G. 14.4.3 PTAs with detailed provisions on government procurement This category includes PTAs with detailed clauses on government procurement and those that have an explicit reference to incorporating provisions of the GPA. Examples of the latter include the EFTA-Hong Kong SAR, China; Canada-Republic of Korea (Box 14.2); and EFTA-Canada agreements. Most DPAs include at least one OECD country as partner. Of the 84 DPAs, 18 agreements represent South-South partnerships, accounting for 22 percent of total DPAs, compared to 43 agreements between North-South partners and 23 between North-North countries (see Table 14.1). A majority of the South-South DPAs involve a Latin American country as a partner; for instance, Costa Rica-Peru, Panama-Guatemala, Costa Rica-Colombia, and Mexico-Central America. 405 Handbook of Deep Trade Agreements It appears that parties to cross-regional agreements tend to be more willing to open their procurement markets to foreign competition. An overwhelming majority (almost 85 percent) of the 84 DPAs and almost half of all PTAs (45 percent) are cross-regional, while only 10 percent of intra-regional agreements have detailed provisions on government procurement. However, the cohort of DPAs is dominated by the EU (9 agreements), EFTA (11 agreements), the US (14 agreements), and Chile (17 agreements) and the propensity of these partners to negotiate cross- regional accords more likely explains this particular stylized fact. The growing significance of government procurement over time is confirmed when considering DPAs. There were only 6 PTAs signed before the year 2000 that elaborated government procurement obligations in detail. The EFTA, NAFTA, EEA, US-Israel, Canada- Israel, and Canada-Chile pioneered the liberalization of government procurement by including comprehensive clauses on the subject in their PTAs prior to the year 2000. In contrast, there has been a surge in the number of such agreements (78) signed in the year 2000 and thereafter. GPA signatories seem to find it easier to negotiate DPAs, since they have already undertaken commitments to liberalize their procurement markets. There are 44 out of 283 PTAs in which both parties are GPA signatories, and 25 of these have detailed provisions on government procurement. That said, 16 of the 84 DPAs (19 percent) have been negotiated between partners that are not signatories to the GPA. Meanwhile, in 43 of the 84 DPAs, one party is not a signatory to the GPA, while in 25 agreements both parties are GPA signatories. An overwhelming majority of DPAs have been negotiated both under Article XXIV of the GATT and Article V of the GATS. Seventy-six out of 84 (around 90 percent) of the DPAs cover trade in both goods and services. There are also 7 agreements in this cohort which cover only trade in goods, and one (the European Economic Area) that covers only trade in services. In terms of coverage of goods sectors, 39 DPAs follow a negative list approach, covering all goods sectors with a list of exceptions. In 9 DPAs, at least one party’s commitments cover all goods sectors; Hong Kong SAR, China-Chile is the only DPA in which commitments of both parties cover all goods sectors. The most common exceptions include purchases by both Ministries of Defense in Japan-Singapore; commitments by Korea in Korea-Colombia; commitments by Singapore in Panama-Singapore; commitments by the US in US-Israel, NAFTA, and others; commitments by Canada in Canada-Israel, NAFTA, Canada-Korea, and others; purchases of agriculture-related products in US-Oman, US-Bahrain, and Central America Free Trade Agreement (CAFTA)- Dominican Republic; commitments by the US in US-Morocco; and commitments by Honduras in Canada-Honduras. In coverage of the services sectors, 20 DPAs follow a negative list approach and 22 follow a positive list approach, explicitly specifying the sectors to which government procurement provisions would apply. In another 18 DPAs, the commitments of one party follow a positive-list approach and the commitments of the other party follow a negative-list approach. In one DPA (Korea-Chile), the 406 Public Procurement commitments of both parties cover all services sectors. Similarly, the commitments of Chile in Chile- Australia and Chile-EFTA cover all services sectors. Some of the common exceptions in services sectors include research and development (US-Colombia, Canada-Panama, commitments of Colombia in EFTA-Colombia, and commitments of Canada in Canada-Peru); telecommunication services (US-Panama, Canada-Panama, and commitments of Canada in Canada-Honduras); and financial services (commitments of Korea in Peru-Korea, and commitments of Chile in US-Chile). The analysis also classified DPAs based on whether the majority of provisions restate the WTO obligation (WTO=), go beyond it (WTO+), or are more limited (WTO-).12 Alignment with WTO coverage could only be accurately assessed for 73 of the 84 DPAs whose text is in English.13 Of these, 32 were found to be equal to WTO coverage. In DPAs between GPA signatories, it is likely that the GPA was used as a reference for those accords, which explains their WTO= score. In contrast, coverage in 38 other DPAs was found to be more limited compared to the WTO, while in another three agreements - US-Chile (Box 14.3), US-Australia, and US-Peru - the coverage goes beyond the WTO. Box 14.3. Example of a procurement provision that goes beyond the WTO: US-Chile FTA Article 9.12: Ensuring Integrity in Procurement Practices Each Party shall adopt the necessary legislative or other measures to establish that it is a criminal offense under its law for: (a) a procurement official of that Party to solicit or accept, directly or indirectly, any article of monetary value or other benefit, for that procurement official or for another person, in exchange for any act or omission in the performance of that procurement official’s procurement functions; (b) any person to offer or grant, directly or indirectly, to a procurement official of that Party, any article of monetary value or other benefit, for that procurement official or for another person, in exchange for any act or omission in the performance of that procurement official’s procurement functions; and (c) any person intentionally to offer, promise or give any undue pecuniary or other advantage, whether directly or through intermediaries, to a foreign procurement official, for that foreign procurement official or for a third party, in order that the foreign procurement official act or refrain from acting in relation to the performance of procurement duties, in order to obtain or retain business or other improper advantage. 12 Coverage is classified as more limited than the GPA when (a) the PTA covers fewer provisions than the GPA; (b) the GPA provisions are not in the text of the PTA; and (c) the provisions of the PTA are not consistent with the corresponding provisions in the GPA. In contrast, coverage is defined as WTO+ when a majority of provisions in the PPA exceed those in the GPA in number and depth. 13 The remaining 11 agreements are in Spanish and it was not possible to code detailed responses for these PTAs with the same level of accuracy. These agreements are: Colombia-Northern Triangle (El-Salvador, Guatemala- Honduras); Costa Rica-Colombia; Costa Rica-Peru; Dominican Republic-Central America; Mexico-Central America; Panama-Costa Rica; Panama-Guatemala; Panama-Honduras; Panama-Peru; Chile-Colombia; and the Pacific Alliance. 407 Handbook of Deep Trade Agreements DPAs were also classified on the basis of enforceability.14 The classification depended on whether the majority of provisions in the agreement were found to be nonbinding, best-endeavor, binding but with no dispute settlement (DS), binding with state-to- state DS, binding with private DS, or binding with both state-to-state and private DS. Enforceability could be assessed accurately for only the 73 (of 84) DPAs whose texts are available in English. Of these 73 agreements, 61 (more than 80 percent) showed high levels of enforceability marked by binding obligations with some form of dispute settlement. Of these 61 DPAs, 47 agreements have provisions on state-to-state dispute settlement, and 14 provide for both state-to-state and private dispute settlements. Of these 14, the vast majority (12) have either the US or Canada as a party to the agreement. In another 10 DPAs, the majority of commitments were found to be nonbinding. In fact, a number of agreements have a majority of non-binding commitments despite having a clause related to the settlement of disputes. Examples include the Eurasian Economic Union and agreements between Panama-El Salvador, Chile-Costa Rica, EU-Georgia, and New Zealand-Singapore, among others. Finally, there are two PTAs whose provisions tend to follow legally binding language but have no enforcement mechanism. For instance, a chapter on Dispute Settlement in the Korea-Singapore agreement does not cover government procurement. Its chapter on Dispute Settlement lists the chapters that fall under the scope of its coverage, and Box 14.4. Non-applicability of DS chapter: Japan-Switzerland agreement Article 130: Existing Rights and Obligations 1.The rights and obligations of the Parties in respect of government procurement shall be governed by the Agreement on Government Procurement in Annex 4 to the WTO Agreement (hereinafter referred to as “the GPA”). 2. If the GPA is amended or is superseded by another agreement, “the GPA,” for the purposes of this Chapter, shall refer to the GPA as amended or such other agreement, as of the date on which such amendment or other agreement enters into force for both Parties. 3. Chapter 14 [DISPUTE SETLLEMENT] shall not apply to this Article. 14 In this analysis, we assess the overall enforceability of the PTA as measured by the modal value of the enforceability variable across the 100 questions in the questionnaire, combined with the existence of dispute settlement provisions. Enforceability based solely on the existence of a DS mechanism is evaluated in the previous section.This section assesses overall enforceability by evaluating: (a) the modal value of the enforceability variable; and (b) the interrelation between the language and the existence of a DS chapter. That is, either: (i) a DS chapter exists but the overall level of enforceability is low, as the majority of the provisions do not contain legally binding language; or (ii) most provisions have legally binding language but there is no DS chapter; or (iii) a DS chapter exists and the level of enforceability (mode) is high. 408 Public Procurement Government Procurement is not mentioned therein. Similarly, the chapter on Dispute Settlement in the Japan-Switzerland agreement specifically does not apply to government procurement (Box 14.4). In contrast, the agreement between Japan and Chile is an example of high enforceability with a provision for dispute settlement. The Dispute Settlement Chapter applies to the Government Procurement chapter, as it is not provided otherwise in the text of the agreement, and Article 175 reads that the chapter on Dispute Settlement shall apply unless otherwise provided for in this agreement.(Box 14.5). Box 14.5. Example of high levels of enforceability and DS provisions: Japan-Chile FTA Article 141:Tendering Procedures 1. Each Party shall ensure that the tendering procedures of its entities are applied in a non-discriminatory manner and in compliance with this Chapter. 2. Each Party shall ensure that its entities do not provide to any supplier information with regard to a specific procurement in a manner which would have the effect of precluding competition. Article 142: Qualification of Suppliers 1. In the process of qualifying suppliers, each Party shall ensure that its entities do not discriminate against suppliers of the other Party. Article 143: Notice of Procurement 1. For each case of intended procurement, each Party shall ensure that its entities make publicly available in advance in the appropriate publication listed in Part 7 of Annex 14.14. 2.The information in each notice of procurement shall include a description of the intended procurement, any conditions that suppliers must fulfill to participate in the procurement, the name of the entity, the address where all documents relating to the procurement may be obtained and the time-limits for submission of tenders. Finally, Figure 14.3 shows the frequency distribution of leading provisions in DPAs (e.g., those on technical specifications, national treatment, and domestic review) by level of development of the signatories. These frequently-used provisions on government procurement are observed mostly in North-South DPAs, which can be partly explained by the fact that North-South agreements (n=43) dominate the cohort of DPAs. In contrast, most leading government procurement provisions are observed in around 20 (of the 23) North-North DPAs and less than 10 (of the 18) South-South DPAs. 409 Handbook of Deep Trade Agreements Figure 14.3: Frequency distribution of leading provisions in DPAs by level of development Source: Deep Trade Agreements Database. 14.5. DETAILED ANALYSIS OF DEEP PROCUREMENT AGREEMENTS In this section, we provide a detailed analysis of the 73 English-language DPAs based on the six major themes (non-discrimination, coverage, procedural disciplines, transparency, dispute settlement, new issues) that were used to classify these accords. The frequency distribution of all provisions in DPAs across all six themes is shown in Annex Figure 14.A.1. 14.5.1 Non-discrimination The non-discrimination theme covers 14 aspects of non-discrimination in government procurement such as national treatment, MFN treatment of and future negotiation of third parties; prohibition of offsets; determination of rules of origin; existing transitional measures (price preferences, offsets, phased-in addition of specific entities or sectors, higher thresholds, and delayed implementation periods); and review of commitments to expand coverage and progressively reduce/eliminate discriminatory measures. 410 Public Procurement Figure 14.4 shows the frequency distribution of non-discrimination provisions in DPAs. The analysis reveals that no single DPA covers all 14 aspects of non-discrimination. The most provisions (12) are in the Trans-Pacific Partnership (TPP), the status of which is now uncertain. The two issues not included in the TPP are MFN treatment of third parties and progressive reduction of discriminatory measures. Figure 14.4: Distribution of non-discrimination procurement provisions in DPAs Source: Deep Trade Agreements Database. Other DPAs with relatively high coverage of non-discrimination issues include EU- Central America, EU-Moldova, and EU-Ukraine, each of which has 10 provisions, followed by EFTA-Central America (Costa Rica and Panama), with 9 provisions. Notably, all agreements with a large coverage of non-discrimination issues are between developing and developed country partners, wherein incorporated provisions relating to transitional measures are significant for developing country partners as they provide for various adjustments that can benefit developing countries, such as phased-in addition of sectors, delayed implementation, etc. Figure 14.4 also shows that about 45 percent of the 73 DPAs include between 3 and 4 of the 14 non-discrimination provisions (23 percent cover 4 provisions and 16 percent cover 3). Meanwhile, 67 of the 73 DPAs that cover at least some issues of non-discrimination were concluded in or after the year 2000, compared to only 6 such agreements before that year. In total, all 73 DPAs cover at least two aspects of non-discrimination. Most frequently included are national treatment, enshrined in 68 DPAs; prohibition of offsets, in 64 DPAs; and provisions requiring that rules of origin not be different from those in the normal course of trade, in 55 DPAs. The least-covered aspects of non-discrimination include MFN, transitional measures (price preferences and offsets), and extension of transitional 411 Handbook of Deep Trade Agreements periods. For instance, the MFN clause is reflected in only one DPA, the EAEU or the Eurasian Economic Union that includes Russia, Belarus, Kazakhstan, Armenia, and Kyrgyzstan as members (see Figure 14.5). Figure 14.5: Frequency distribution of non-discrimination provisions in DPAs Source: Deep Trade Agreements Database. Finally, five of the eight South-South DPAs that cover non-discrimination cover 3 of the 14 aspects. Among North-North agreements, 9 of the 23 DPAs (almost 45 percent) cover four of the assessed issues. The maximum number of issues covered in the North-North agreements is 8. As for North-South agreements, 8 of the 42 DPAs in this group cover 4 of the 14 aspects of non-discrimination. 14.5.2 Coverage The analysis of coverage is based on three questions related to procuring entities under Annexes 1, 2 and 3; one question each related to inflation, modification of coverage, and unnecessary exceptions; and 18 questions about whether thresholds for goods, services, and construction services under the three Annexes are higher or lower than in the GPA. In terms of coverage of procuring entities, 15 DPAs cover only Annex 1 entities, 20 DPAs cover both Annex 1 and Annex 2 entities, while the majority of the DPAs (44 agreements) cover entities listed under all three Annexes. Amongst North-North DPAs, 16 agreements (around 70 percent) cover entities under all three Annexes. Examples include Australia-Chile, 412 Public Procurement United States-Australia, and Canada-Israel. Only six North-North agreements, including Canada-Chile, Korea-New Zealand, and Korea-United States, do not extend coverage to entities under all three Annexes. For instance, in the Canada-Chile agreement, Annex K bis-01.1-1 and Annex K bis-01.1-2 cover central, regional, and other government entities for Chile. For Canada, however, these Annexes cover only central and other government entities, excluding regional government entities. Meanwhile, more than half of North-South DPAs (23 out of 42 agreements) and more than 60 percent of South-South DPAs (5 out of 8 accords) cover entities under all three Annexes. In terms of comparison with GPA-stipulated thresholds, thresholds for goods and services procurement by Annex 1 entities was not found to be higher than the GPA-stipulated thresholds for any DPA. For goods procurement by Annex 2 entities, only one agreement - that between the US and Colombia - has a threshold value higher than that stipulated by the US under the GPA.15 The agreement with the largest number of thresholds above GPA levels is between the US and Bahrain; it has threshold values higher than those (for the US) in the GPA in 4 cases - for Annex 1 construction services and for Annex 3 goods, services and construction services (Table 14.2). At the same time, the US-Bahrain agreement has several threshold values that are lower than GPA levels. Table 14.2: US-Bahrain agreements: Threshold values higher than GPA levels Threshold values under the Threshold values under the GPA US-Bahrain agreement Annex 1 construction services USD 7,611,532 USD 5,000,000 Annex 3 goods by a List A entity, USD 250,000, by a List B entity, USD 538,000 USD 250,000 or USD 400,000 Annex 3 services by a List A entity, USD 250,000, by a List B entity, USD 538,00 USD 250,000 or USD 400,000 Annex 3 construction services by a List A or a List B entity, USD 9,368,478 USD 5,000,000 Source: Deep Trade Agreements Database. There are 23 DPAs with thresholds equal to the GPA in goods, services, and construction services covered under Annexes 1, 2 and 3; the coverage is equal to that in the WTO in each case. Most of these agreements have the EFTA countries or the EU as a party, including EFTA-Colombia, EFTA-Korea, EU-Central America, EU–Ukraine, and EU-Chile. 15 Threshold levels are measured only for GPA signatories and Columbia is not a GPA signatory. 413 Handbook of Deep Trade Agreements Significantly, 27 DPAs have threshold values lower than those in the GPA in at least one area, and 7 of these have lower than GPA thresholds across all measured aspects; i.e., goods, services and construction services under Annexes 1-3. Six of these agreements have the US as a party; namely, US-Morocco, US-Panama, US-Peru, US-Chile, US-Singapore, and CAFTA-Dominican Republic. EU-Georgia is the only agreement not involving the US that stipulates threshold values for goods, services, and construction services across all Annexes that are lower than GPA levels. On the whole, DPAs that do not have thresholds higher than GPA thresholds comprise 11 South-South, 28 North-South, and 22 North-North agreements. Amongst DPAs that have threshold values equal to GPA levels, 52 percent (12 of 23) belong to the North-South group, while the remaining 48 percent are North-North agreements. Notably, threshold values are adjusted for inflation in only 37 DPAs.16 Out of 73 DPAs, 67 include provisions for modification/rectification of coverage and 63 exclude unnecessary exceptions from coverage except those permitted by the GPA. 14.5.3 Procedural disciplines Assessment of procedural disciplines covers the existence of procedural provisions in the text of an agreement and their consistency with the GPA. The assessment includes conditions for participation in a tender; requirements for tender documents, for qualification of suppliers and for negotiation; technical specifications; treatment of tenders and award of contracts; limited and selective tendering; electronic auctions; and integrity in procurement practices. The scoring for procedural disciplines ranges from 0 to 26. Of the 73 DPAs, 16 accords scored 18 out of 26 in terms of the coverage of procedural disciplines and their consistency with the GPA. The distribution of the number of procedural issues covered in DPAs is concentrated in the range of 12-23 (see Figure 14.6), with no single agreement covering all areas. Two agreements - EFTA-Colombia and EFTA-Peru - cover the highest number of procedural disciplines, 23 out of 26, followed by 6 DPAs that cover 22 issues.The latter include agreements between EFTA-Hong Kong SAR, China, EFTA-Ukraine, Canada-Korea, and EU-Korea, among others. Most DPAs with a high coverage of procedural disciplines are either North-South or North-North agreements. Within this distribution, 71 out of 73 agreements encompass provisions on technical specifications, of which 90 percent (64 out of 71) are GPA-consistent. More than 80 percent 16 Data on inflation-adjustment of threshold values is available for only 66 DPAs. 414 Public Procurement Figure 14.6: Distribution of procedural disciplines in DPAs Source: Deep Trade Agreements Database. of the 73 DPAs cover issues related to conditions of suppliers’ participation (with an 87 percent GPA consistency rate); treatment of tenders, award of contracts, and provisions on limited tendering (with a 98 percent consistency rate each); requirements for tender documentation (with a 75 percent consistency rate); and provisions on time periods and deadlines (Figure Figure 14.7: Frequency distribution of procedural disciplines in DPAs Source: Deep Trade Agreements Database. 415 Handbook of Deep Trade Agreements 14.7).With regard to the GPA consistency of the most common procedural disciplines (see Box 14.7 for an example in the context of NAFTA), the provision on tender documentation is consistent in 75 percent of the 73 DPAs, and the remaining frequently-used clauses are consistent in more than 85 percent of the 73 DPAs. In contrast, provisions on electronic auctions are contained in only 25 percent of DPAs; provisions ensuring integrity in procurement practices (e.g., by avoiding conflict of interest) are present in about 35 percent; and provisions on negotiations are reflected in 43 percent of DPAs. Finally, the highest average coverage of procedural issues, equal to 16.5, is observed among North-North DPAs. For North-South DPAs, the mean coverage stands at 15.4; and for the group of South-South DPAs, the average coverage of procedural disciplines is 12. Box 14.6. Example of a fully GPA-consistent provision on technical specification: NAFTA Article 1007:Technical Specifications 1. Each Party shall ensure that its entities do not prepare, adopt or apply any technical specification with the purpose or the effect of creating unnecessary obstacles to trade. 2. Each Party shall ensure that any technical specification prescribed by its entities is, where appropriate: (a) specified in terms of performance criteria rather than design or descriptive characteristics; and (b) based on international standards, national technical regulations, recognized national standards, or building codes. 3. Each Party shall ensure that the technical specifications prescribed by its entities do not require or refer to a particular trademark or name, patent, design or type, specific origin or producer or supplier unless there is no sufficiently precise or intelligible way of otherwise describing the procurement requirements and provided that, in such cases, words such as “or equivalent” are included in the tender documentation. 4. Each Party shall ensure that its entities do not seek or accept, in a manner that would have the effect of precluding competition, advice that may be used in the preparation or adoption of any technical specification for a specific procurement from a person that may have a commercial interest in that procurement. 14.5.4 Transparency (ex-ante and ex-post) The transparency assessment covers both ex-ante and ex-post issues.The three ex-ante issues relate to (a) publication of procurement laws and regulations; (b) publication of the notice of intended/planned procurement; and (c) consistency of the notice of the intended/planned procurement with the requirements of Art.VII:2 of the RGPA.The four ex-post issues cover (a) information provided to bidders (results and reasons for non-selection); (b) disclosure of information provided to third parties; (c) collection and reporting of statistics; and (d) consistency of such provisions with Art. XVI:4 of the RGPA. 416 Public Procurement More than 75 percent of the 73 DPAs cover all ex-ante transparency issues, while only 8 percent of the 73 DPAs cover all issues of ex-post transparency. Roughly half of all DPAs cover only 2 issues of ex-post transparency (Figure 14.8). Figure 14.8: Distribution of transparency-related provisions in DPAs Source: Deep Trade Agreements Database. Amongst issues of ex-ante transparency, the requirement to publish a notice of intended/planned procurement is reflected in more than 90 percent of all DPAs. Among ex-post transparency issues, provisions on information provided to bidders (results and reasons for non-selection) are incorporated in a majority of the DPAs (87 percent), while those on information provided to third parties can be observed in almost two-thirds of all DPAs. In contrast, the least common provision relates to the collection and reporting of statistics, which is enshrined only in 20 percent of DPAs. Thus, a very important element of ex-post transparency is largely ignored by signatories that otherwise negotiate deep commitments on government procurement in their trade agreements.17 On the whole, only 5 DPAs cover all ex-ante and ex-post transparency issues. In the North-North group, Canada-Korea and EU-Korea cover all transparency issues, while most other DPAs in this cohort cover at least 3 issues and the majority cover 5 to 6. In the North-South group, Japan- Mexico and EFTA-Hong Kong SAR. China, have extensive coverage of transparency issues. More than 70 percent of the DPAs in this group cover 4 to 5 issues. In the South-South group, the maximum number of transparency issues covered in an agreement is 5.The Panama-El Salvador accord is the only agreement across all income groups that does not cover any transparency issues. 17 A similar lack of statistical reporting by GPA signatories is documented in Shingal 2011, 2012, 2015. 417 Handbook of Deep Trade Agreements 14.5.5 Dispute resolution The dispute resolution theme covers domestic review procedures and their consistency with Art. XVIII of the GPA, as well as provisions on dispute settlement and their consistency with Art. XX of the GPA. More than 70 percent of the 73 DPAs cover all four issues related to dispute resolution, including domestic review procedures and dispute settlement, and the consistency of those provisions with the GPA. DPAs covering only two of the four issues constitute another 14 percent, as do PPAs covering three of the four dispute resolution issues. More specifically, provisions on dispute settlement are reflected in all DPAs except for Korea- Singapore, Japan-Switzerland, and Panama-El Salvador.The Korea-Singapore agreement lists the particular chapters to which dispute settlement procedures apply, and the government procurement chapter is not among them. The Japan-Switzerland agreement also excludes government procurement from dispute settlement (Box 14.4). In contrast, the Korea-Canada agreement specifically applies dispute settlement to government procurement provisions (Box 14.8). Box 14.7. Free trade agreement between Korea and Canada Chapter 21: Dispute Settlement Annex 21-A: Nullification and Impairment 1. If a Party considers that any benefit it could reasonably have expected to accrue to it under any provision of: (c) Chapter Fourteen (Government Procurement); is nullified or impaired as a result of the application of any measure that is not inconsistent with this Agreement, in the sense of Article XXIII(1)(b) of GATT 1994, Article XXIII (3) of GATS or Article XXII(2) of GPA, the Party may have recourse to dispute settlement under Section A of this Chapter [Chapter on Dispute Settlement]. 14.5.6 New issues The new issues theme covers a number of disciplines that have emerged in recent agreements, including those on e-procurement, sustainable procurement, SME participation, adoption of safety standards, and (as in the TPP) cooperation between the parties on matters of public procurement. 418 Public Procurement There is no agreement covering all five of these new issues. Of the 73 DPAs, 17 percent cover three new issues, while 27 percent and 38 percent cover one or two issues, respectively. Across all income groups (North-North, North-South, and South-South), most of the DPAs cover one to two new issues, whereas 12 DPAs (17 percent) do not cover any new issue. More specifically, provisions facilitating e-procurement can be observed in 60 percent of all DPAs, followed by clauses on facilitation of SME participation, which are reflected in just over half of the 73 DPAs. Provisions facilitating cooperation are in just over 40 percent of these agreements. Provisions on sustainable procurement are not observed in any DPA,18 while facilitation of safety standards is incorporated in only one agreement, that between the US and Korea (Box 14.9). Box 14.8. Example of provision on facilitation of safety standards: US-Korea agreement Article 17.7:Technical Specifications For greater certainty, a Party, including its procuring entities, may, in accordance with Article VI of the GPA, prepare, adopt, or apply technical specifications: (a) to promote the conservation of natural resources or protect the environment; or (b) to require a supplier to comply with generally applicable laws regarding i. fundamental principles and rights at work; and ii. acceptable conditions of work with respect to minimum wages, hours of work, and occupational safety and health, in the territory in which the good is produced or the service is performed. 14.6. CONCLUSIONS The proliferation of preferentialism in the last decade and a half, and the increasing use of PTAs to liberalize government procurement, warrants an analysis of procurement provisions in these agreements. This chapter builds on the existing literature to come up with a new methodology to classify procurement provisions in trade agreements and then presents stylized facts based on this classification. Our analysis suggests that 45 percent of the 283 WTO-notified PTAs in force as of March 2017 still do not include any provisions on government procurement, while 30 percent have 18 Note that principles of sustainable procurement for Australia and New Zealand are reflected in Australian and New Zealand Government Framework for Sustainable Procurement released in September 2007.The framework provides for the integration of sustainable development considerations in government procurement by the two countries. However, ANZCERTA, the PTA between Australia and New Zealand, does not have a detailed chapter on government procurement. 419 Handbook of Deep Trade Agreements deep provisions.These deep procurement agreements (DPAs) have been primarily negotiated among the developed and developing country trading partners of Canada, Chile, EFTA, the EU, and the US where at least one country is a GPA-signatory (with the exception of Chile). Most DPAs have come into effect since 2000. They are predominantly cross-regional and cover both goods and services trade. However, the coverage of government procurement can be classified as WTO+ in only three DPAs (US-Australia, US-Chile, and US-Peru), while the coverage was found to equal that in the WTO in the majority of other agreements. Significantly, more than 80 percent of the DPAs show high levels of enforceability marked by binding obligations with some form of dispute settlement. In terms of coverage of entities, the majority of the DPAs were found to cover procurement undertaken by entities listed under all three Annexes. Moreover, 27 DPAs were found to have lower-than-GPA threshold values in at least one area of goods, services, or construction services procurement, and 7 were found to have lower-than-GPA threshold values across all measured aspects - i.e., goods, services, and construction services under Annexes 1-3. Significantly, six of these seven agreements involve the US as a party: US-Morocco, US-Panama, US-Peru, US-Chile, US-Singapore, and CAFTA-Dominican Republic – highlighting the dominance of the US in being able to negotiate GPA+ provisions in its PTAs with both developed and developing country trading partners. We also found the following provisions to be covered in the majority of the DPAs: provisions on national treatment (68 DPAs); provisions on prohibition of offsets (64 DPAs); provisions on technical specifications (71 DPAs, of which 90 percent were found to be GPA consistent); and provisions on dispute settlement (70 DPAs). Among the new issues, provisions facilitating e-procurement were observed in 45 DPAs, followed by clauses on facilitation of SME participation in 39 DPAs. In contrast, the least-covered issues include MFN, transitional measures in the form of price preferences and offsets, extension of transitional periods, provisions on electronic auctions, provisions ensuring integrity in procurement practices, and provisions relating to the collection and reporting of statistics. Among the new issues, provisions on sustainable procurement were not observed in any agreement, while the provision on facilitation of safety standards is in only the US-Korea FTA. Finally, the primary objective of DPAs seems to be to offer trading partners preferential access to each other’s public markets by, inter alia, extending coverage of procurement to more entities; expanding procurement coverage to a larger set of goods and services; and lowering threshold values above which public markets can be contested by preferential partners. To that extent, DPAs lead to de jure and even de facto discrimination against third parties. 420 Public Procurement One way of monitoring actual implementation of these agreements would be to examine whether the number and value of government contracts awarded to preferential suppliers, relative to third parties, have risen since a DPA came into effect, using established empirical methodologies. 19 One reliable information source in this regard are the data submitted by GPA Contracting Parties to the WTO Committee on Government Procurement, which, at least for some GPA signatories, 20 include data over time on contract awards by procuring entity, sector, and nationality of the winning supplier. ACKNOWLEDGMENTS The authors would like to thank Aaditya Mattoo, Michele Ruta, and Nadia Rocha for valuable comments and feedback on the development of the methodology; and Julien Gourdon and participants in a World Bank conference, “The Evolution of Deep Trade Agreements,” for excellent comments.The usual disclaimer applies. 19 For example, Shingal 2015. 20 Despite the requirements of Article XIX: 5 of UR GPA and Article XVI 4: of RGPA, GPA signatories exhibit considerable heterogeneity in reporting procurement data to the WTO Committee on Government Procurement. For details and discussion of related issues see Shingal 2011, 2012, 2015. 421 Handbook of Deep Trade Agreements REFERENCES Anderson, R., C. Müller, K. Osei-Lah, J. Pardo de Leon, and Trionfetti, F. 2000. Discriminatory public procurement and P. Pelletier. 2011. Government procurement provisions in international trade. The World Economy 23: 57-76. regional trade agreements: A stepping stone to GPA accession? In The WTO regime on government procurement: Challenge and Ueno, A. 2013. Multilateralising regionalism on government reform, edited by S. Arrowsmith and R. Anderson Cambridge: procurement. OECD Trade Policy Paper 151, OECD Cambridge University Press. Publishing, Paris. Bourgeois, J., K. Dawar, and S. Evenett. 2007. A comparative UN Social and Economic Commission for Asia and the analysis of selected provisions in free trade agreements. Report for Pacific. Asia-Pacific trade and investment agreement database Directorate General for Trade, Chief Economist’s Office, (APTIAD). European Commission, Brussels. World Trade Organization. No date. Services sectoral classification Dür, A., L. Baccini, and M. Elsig. 2014. The design of list: MTN.GNS/W/120: Scheduling of commitments under the international trade agreements: Introducing a new dataset. GATS. http://tsdb.wto.org/Includes/docs/W120_E.doc. Review of International Organizations 9(3): 353-75. Gourdon, J., and J. Messent. 2017. How government procurement measures can Affect trade. OECD Trade Policy Papers 199, OECD Publishing, Paris. Heydon, K., and S. Woolcock. 2009. The rise of bilateralism: Comparing American, European, and Asian approaches to preferential trade agreements. Paris: United Nations University Press. Hoekman, B. 2015. International cooperation on public procurement regulation. EUI Working Paper RSCAS 2015/88, European University Institute, Fiesole. Horn H., P. Mavroidis, and A. Sapir. 2009. Beyond the WTO? An anatomy of EU and US preferential trade agreements. Brussels: Bruegel. OECD (Organisation for Economic Co-operation and Development). 2002. The size of government procurement markets. Journal of Budgeting 1(4). Rickard, S., and D. Kono, 2014. Think globally, buy locally: International agreements and government procurement. Review of International Organizations 9(3): 333-52. Shingal, A. 2009. Three essays on trade in services: Exploring tradability, regional trade and government procurement of services. PhD thesis, University of Sussex. ———. 2011. Services procurement under the WTO’s Agreement on Government Procurement: Whither market access? World Trade Review 10(4): 1-23. ———. 2012. Estimating market access in non-GPA countries:A suggested methodology. NCCR-Trade Working Paper 2012/06. ———. 2015. Econometric analyses of home-bias in government procurement. Review of International Economics 23(1): 188-219. 422 Public Procurement ANNEX Annex Table 14.A.1: List of PTAs with no, shallow, and deep provisions on government procurement PTAs with no provisions: APTA; APTA-Accession of China; ASEAN FTA; ASEAN- Australia New Zealand; ASEAN-India; ASEAN-Japan; ASEAN-Korea, Rep.; ASEAN- China; Agadir Agreement; Andean Community; Armenia-Kazakhstan; Armenia- Moldova; Armenia-Turkmenistan; Armenia-Ukraine; CACM; CARICOM; CEMAC; CEZ (Common Economic Zone); COMESA; Canada-Jordan; Chile-India; Chile- Malaysia; Chile-Mexico; Chile-Vietnam; China-Costa Rica; China-Hong Kong SAR, China; China-Macao SAR, China; China-New Zealand; China-Singapore; Colombia-Mexico; EAC; EAC-Burundi/Rwanda; EAEC; EAEU-Kyrgyz Republic; EAEU-Armenia; EC-10; EC-Enlargement-25; EC-Enlargement-27; EC Treaty; EC(12), Enlargement EC(9); ECO; ECOWAS; EU-Faroe Islands; EU-Syrian Arab Republic; EU-Albania; EU-Andorra; EU-Côte d’Ivoire; EU-Iceland; EU-Lebanon; EU-North Macedonia; EU-OCT; EU-Papua-New Guinea-Fiji; EU-San Marino; EU-Switzerland/Lichtenstein; El Salvador-Honduras-Taiwan, China; El-Salvador- Cuba; EU-Norway; GCC; GSTP Agreement; Georgia-Turkmenistan; Georgia- Armenia; Georgia-Azerbaijan; Georgia-Kazakhstan; Georgia-Russia; Georgia-Ukraine; Guatemala-Taiwan, China; India-Afghanistan; India-Singapore; India-Bhutan; India- Malaysia; India-Nepal; India-Sri Lanka; Japan-Malaysia; Japan-Indonesia; Korea, Rep.- Vietnam; Korea, Rep.-India; Korea, Rep.-Turkey; Kyrgyz Republic-Armenia; Kyrgyz Republic-Uzbekistan; Kyrgyz Republic-Kazakhstan; Kyrgyz Republic-Moldova; Kyrgyz Republic-Ukraine; Lao PDR-Thailand; Latin American Integration Association; MERCOSUR; MERCOSUR-India; Malaysia-Australia; Mauritius-Pakistan; Mexico- Panama; Mexico-Uruguay; New Zealand-Malaysia; Nicaragua-Taiwan, China; PAFTA; PATCRA; Pakistan-Malaysia; Pakistan-Sri Lanka; Panama-Chile; Panama-Taiwan, China; Panama-Dominican Republic; Panama-Nicaragua; Peru-Chile; Peru-China; Peru-Mexico; Russian Federation-Serbia; Russian Federation-Tajikistan; Russian Federation-Turkmenistan; Russian Federation-Uzbekistan; Russian Federation- Azerbaijan; Russian Federation-Belarus/Kazakhstan; SACU; SADC; SADC-Seychelles; SAFTA; SAFTA-Accession of Afghanistan; SAPTA; SPARTECA; Thailand-New Zealand; Turkey-Chile; Turkey-Albania; Turkey-Mauritius; Ukraine-Azerbaijan; Ukraine-Belarus; Ukraine-Kazakhstan; Ukraine-Montenegro; Ukraine-Tajikistan; Ukraine-Turkmenistan; Ukraine-Uzbekistan, Faroe Islands-Switzerland. PTAs with shallow provisions: ANZCERTA; Australia-China; Brunei Darussalam- Japan; CEFTA; CIS; Canada-Costa Rica; Chile-China; China-Korea, Rep.; China- Switzerland; EC Enlargement (15); EFTA-Albania; EFTA-Bosnia and Herzegovina; EFTA-Israel; EFTA-Jordan; EFTA-Lebanon; EFTA-Montenegro; EFTA-Morocco; 423 Handbook of Deep Trade Agreements EFTA-North Macedonia; EFTA-SACU; EFTA-Serbia; EFTA-Tunisia; EFTA-Turkey; EFTA-West Bank and Gaza; EU-Algeria; EU-Arab Republic of Egypt; EU-Bosnia and Herzegovina; EU-Cameroon; EU-Eastern and Southern Africa States Interim EPA; EU-Enlargement; EU-Israel; EU-Jordan; EU-Mexico; EU-Montenegro; EU-Morocco; EU-Serbia; EU-South Africa; EU-Turkey; EU-Tunisia; EU-West Bank and Gaza; Egypt-EFTA; Egypt-Turkey; Iceland-China; Iceland-Faroe Islands; India-Japan; Japan- Mongolia; Japan-Philippines; Japan-Thailand; Japan-Vietnam; Jordan-Singapore; MSG (Melanesian Spearhead Group); PICTA; Pakistan-China; Thailand-Australia; Turkey- Syrian Arab Republic; Turkey-Bosnia and Herzegovina; Turkey-Georgia; Turkey-Israel; Turkey-Jordan; Turkey-Montenegro; Turkey-Morocco; Turkey-North Macedonia; Turkey-Serbia; Turkey-Tunisia; Turkey-West Bank and Gaza; United States-Jordan; Ukraine-Moldova; Ukraine-North Macedonia; WAEMU; Faroe Islands-Norway. PTAs with deep provisions: Australia-Chile; CAFTA-Dominican Republic; Canada- Chile; Canada-Colombia; Canada-Honduras; Canada-Israel; Canada-Korea, Rep.; Canada-Panama; Canada-Peru; Chile-Nicaragua (Chile-Central America); Chile- Colombia; Chile-Costa Rica; Chile-El Salvador; Chile-Guatemala (Chile-Central America); Chile-Honduras; Chile-Japan; Colombia-Northern Triangle (El-Salvador, Guatemala-Honduras); Costa Rica-Colombia; Costa Rica-Peru; Costa Rica-Singapore; Dominican Republic-Central America; EAEU; EFTA; EFTA-Canada; EFTA-Central America (Costa Rica and Panama); EFTA-Chile; EFTA-Colombia; EFTA-Hong Kong SAR, China; EFTA-Korea, Rep.; EFTA-Mexico; EFTA-Peru; EFTA-Singapore; EFTA- Ukraine; EU-CARIFORUM; EU-Central America; EU-Chile; EU-Colombia/Peru; EU-Georgia; EU-Korea, Rep.; EU-Moldova; EU-Ukraine; EEA; Gulf Cooperation Council-Singapore; Hong Kong SAR, China-Chile; Israel-Mexico; Japan-Singapore; Japan-Australia; Japan-Mexico; Japan-Peru; Japan-Switzerland; Korea, Rep.-Chile; Korea, Rep.-Australia; Korea, Rep.-Colombia; Korea, Rep.-New Zealand; Korea, Rep.-Singapore; Korea, Rep.-US; Mexico-Central America; NAFTA; New Zealand- Taiwan, China; New Zealand-Hong Kong SAR, China; New Zealand-Singapore; Pacific Alliance; Panama-Costa Rica; Panama-El Salvador; Panama-Guatemala; Panama- Honduras; Panama-Peru; Panama-Singapore; Peru-Korea, Rep.; Peru-Singapore; Singapore-Australia; Singapore-Taiwan, China; TPP; Trans-Pacific Strategic Economic Partnership; US-Australia; US-Bahrain; US-Chile; US-Colombia; US-Israel; US- Morocco; US-Oman; US-Panama; US-Peru; US-Singapore. 424 Public Procurement Annex Figure 14.A.1. Distribution of provisions in PPAs under six different themes - non-discrimination, procedural issues, transparency, new issues, dispute settlement, and coverage Source: Deep Trade Agreements Database. 425 Handbook of Deep Trade Agreements 426 Subsidies CHAPTER 15 Subsidies L. Rubini 427 Handbook of Deep Trade Agreements CHAPTER 15 Subsidies L. Rubini1 University of Birmingham Law School, Birmingham, United Kingdom CONTENTS 15.1. INTRODUCTION 429 15.2. LITERATURE AND PREVIOUS DATASETS 430 15.3. NEW DATASET 431 15.3.1 Overview of the template 431 15.3.1.1 Categories and questions (rows) 431 15.3.1.2 Further information on the depth of PTAs (columns) 432 15.3.2 Selected coding issues 437 15.3.2.1 Objective of control of subsidies 437 15.3.2.2 Exemptions for legitimate subsidies, public services, and sectoral aid 439 15.4. ANALYSIS: DESCRIPTIVE STATISTICS 440 15.4.1 General overview of the evolution and distribution of subsidy provisions 440 15.4.1.1 Evolution of the number of PTAs with subsidy provisions 440 15.4.1.2 Distribution and evolution of PTAs with subsidy provisions by level of development of its members 443 15.4.1.3 Distribution and evolution of PTAs with subsidy provisions by geographic group 444 15.4.2 Analysis of scope 445 15.4.2.1 Most common subsidy provisions in PTAs 445 15.4.2.2 Most common provisions by level of development 447 15.4.2.3 Most common provisions by geographic group 448 15.4.2.4 Notable countries 449 15.4.3 Specific analysis of types of provisions 450 15.4.3.1 Coverage (sectors) 450 15.4.3.2 Export subsidies 452 15.4.3.3 Domestic subsidies 453 15.4.3.4 Transparency 455 15.4.3.5 Enforcement 457 15.5. CONCLUSIONS 458 ACKNOWLEDGMENTS 459 REFERENCES 460 1 Reader in International Economic Law, University of Birmingham Law School, Birmingham, UK. E-mail: l.rubini@bham.ac.uk. 428 Subsidies 15.1. INTRODUCTION Subsidies and their disciplines are one of the most ambiguous and controversial areas in international economic law. Governments have always subsidized sectors and industries in their economies. Crucially, however, the policy objectives pursued through subsidies (for example, supporting green energy while boosting local industry and jobs) may be mixed. So are their effects, with positives and negatives being produced, and, in an increasingly globalized economy, with spill-overs often crossing national borders. Hence, since the late 1800s (one can immediately think of the 1902 Brussels Sugar Convention, which created the first modern international trade institution2), the need to arbitrate these measures at the international level has emerged as a key priority for international law. Subsidies and state aid have always been included in trade agreements. From the international perspective, the biggest puzzle is how to create rules that manage to balance positives and negatives, especially where losers and winners of subsidization belong to different countries. How can a transnational trade-off be made and incorporated into a legal framework? This chapter makes a start in laying down the basis for a comprehensive analysis of how preferential trade agreements (PTAs) regulate subsidies and, in particular, how deep they go in regulating and balancing their negatives and positives. This chapter maps the provisions on subsidies in 283 PTAs signed between 1957 and early February 2016, using the World Bank’s Deep Integration database.3 This mapping is equivalent to assessing the vertical depth of these provisions, which is the ultimate purpose of the World Bank’s Deep Integration research agenda.4 This chapter outlines the methodology underlying the creation of a new dataset on subsidies and state aid in PTAs and includes an initial description of the main patterns that emerge from the dataset. The final section of the chapter offers some tentative conclusions on the vertical depth of the subsidy provisions in PTAs. 2 Fahkri 2014. 3 The dataset actually includes 282 PTAs. Normally amendments are not counted separately, but treaties for the accession of new parties are. For counting purposes, given the complexity of the successive iterations of the regulatory framework, there are 3 different sheets for (9) CARICOM, i.e., (9.1) CARIFTA, (9.2) CARICOM, and (9.3) CARICOM + CSME. For the sake of simplicity and consistency with the datasets used in the other chapters, only (9.2) CARICOM has been considered for the analysis of stylized facts. Also, for consistency reasons, (216) East African Community (EAC)-Accession of Burundi and Accession of Rwanda have been counted as two separate agreements, which brings the total to 283. 4 The research agenda aims at complementing the existent dataset on the “horizontal” depth of PTAs created by the World Bank. See Hofmann, Osnago, and Ruta 2017. 429 Handbook of Deep Trade Agreements 15.2. LITERATURE AND PREVIOUS DATASETS The literature on subsidy disciplines is vast, ranging from works analyzing subsidy rules in the GATT and in the specialized codes and agreements adopted to implement them,5 to the possibly even larger literature commenting on the specific regime of state aid control that has developed in Europe since the advent of the European Communities in the early 1950s.6 The typical approach one finds in this literature is that of commenting on rules that have been adopted or are about to be adopted, and assessing their ability to solve (or not) specific economic, political, or legal problems. The content of the disciplines is always under examination but, since most of this literature is legal, the focus is very often on how rules should or should not be interpreted. Equally broad is the economic literature on subsidies and state aids. Only a few works, however, focus on subsidy disciplines and their design, or directly engage with the actual disciplines of PTAs.7 Very interesting is the work carried out by the OECD, and in particular the Roundtables on Competition, Subsidies, and State Aid,8 which analyze subsidies and subsidy control - both the supra-national disciplines of the EU or the WTO, and those at the regional or even domestic level. Political and social scientists are paying increasing attention to subsidy regimes, especially in the EU.9 Despite the breadth of this literature, however, it is difficult to identify one single work that maps the content of subsidies disciplines in PTAs in a comprehensive and detailed manner or that expressly looks at the depths of these disciplines.10 The only example of a dataset that includes subsidies is DESTA (Design of Trade Agreements), in which state aid disciplines have been coded in the competition and trade defense chapters. However, that dataset includes far fewer questions than the one constructed for this chapter, which includes 36 questions on subsidy disciplines.11 5 GATT Tokyo Round Subsidy Code, WTO Agreement on Subsidies and Countervailing Measures, and WTO Agreement on Agriculture. On these rules and agreements, and referring only to monographs and edited collections, see Hufbauer and Shelton-Erb 1984; Wallace et al. 1984; Bourgeois 1991; Luengo 2007; Mavroidis et al. 2007; Rubini 2009; Bagwell et al. 2010; Coppens 2015; and Rubini and Hawkins 2016. 6 Referring again only to monographs and edited collections, one can cite Ehlermann and Everson 2001; Biondi et al. 2004; Dony 2007; Derenne and Merola 2007; Luengo 2007; Rubini 2009; De Cecco 2012; Hancher et al. 2012; Bacon 2013; Piernaz-Lopez 2015; Quigley 2015; Hofmann and Micheau 2016; and Rubini and Hawkins 2016. 7 See, e.g., Bagwell and Staiger 2006; Horn, Maggi, and Staiger 2010; Brou and Ruta 2012; Sykes 2010; and Friederiszick et al. 2007. 8 OECD 2001, 2010. 9 For a review of the literature concerning EU state aid, see Blauberger 2011. See also Shaffer,Wolfe, and Le (2015). 10 The Brexit process is raising interest in the various alternatives to regulate state aid. See, e.g., Biondi (2018). 11 The competition chapter includes only one question (“Is there a provision on state aid?”). The subsidies and countervailing duties chapter include the following questions: “Are there any subsidy provisions?”, “Is there a general reference to subsidies?”, “Is there a reference to GATT/WTO?”, “Do parties develop a common policy on subsidies?”, “Is there an explicit reference to allowing subsidies?”, “Is there an explicit reference to out-rule subsidies?”, “Are countervailing duties mentioned in the agreement?” For an introduction to the DESTA project, see Dür, Baccini, and Elsig 2014. 430 Subsidies 15.3. NEW DATASET This section of the chapter outlines the methodology used in the mapping exercise by offering an overview of the template and a specific examination of a few selected coding issues. 15.3.1 Overview of the template It is useful to distinguish the information included in the template according to its two main dimensions: rows and columns. 15.3.1.1 Categories and questions (rows) The mapping exercise is based on a template consisting of six categories which reflect the key types of provisions regulating subsidies in PTAs: • Objectives and coverage of disciplines • Substantive disciplines • Transparency • Enforcement • Special and differential treatment • Miscellaneous Each category is divided into questions, with a total of 36 questions. The template begins by asking whether the PTA specifically spells out the objective of regulating subsidies and reactions to them, and whether the PTA makes reference to GATT/WTO subsidy provisions. The template then focuses on whether the PTA includes, either explicitly or through reference, a definition of subsidy and whether it covers support granted by sub-central authorities or to state enterprises, and support in key sectors (services, agriculture, fisheries). The questions on substantive disciplines focus on (a) provisions that prohibit or regulate certain types of subsidies (export subsidies, local content subsidies, subsidies that distort trade or competition); (b) the presence of any ceiling to or de minimis threshold for permitted subsidies; and (c) whether there is any specific regulation for agricultural subsidies, fisheries subsidies, subsidies to pursue public services, or any other specific discipline for certain sectors or objectives (a catch-up clause used to cover, for example, special disciplines for steel subsidies). This section of the template concludes with questions on the existence of national treatment obligations applicable to subsidies in trade in goods, services, and investment. 431 Handbook of Deep Trade Agreements Next are questions that capture various levels of transparency (including notification requirements, deliberation and assessment, cross-notifications, and submissions by interested parties), followed by questions on enforcement, which capture those mechanisms that are specifically devoted to ensuring that the rules are respected, and possible breaches remedied. The relevant questions focus on (a) dispute settlement (DS); (b) the existence of a common institution dealing with transparency or enforcement; (c) the existence of a domestic authority dealing with subsidies; (c) the capacity for the enactment of secondary legislation; and (d) the obligation to withdraw illegal subsidies or rules on countervailing duties (CVD).The template ends with questions on the presence of special and differential treatment and cooperation provisions, and on the obligation for review and further negotiation of subsidy rules in the PTA. 15.3.1.2 Further information on the depth of PTAs (columns) The questions in the row are intersected with columns that convey further information about the depth of the PTA. There are columns on WTO coverage, enforceability, benefits to non- members, and sectoral coverage or exclusions, as well as a column for providing comments.The most relevant of the columns are discussed below. 15.3.1.2.1 WTO coverage This column indicates the relationship between the coverage of the disciplines on subsidies in the PTA and the corresponding regulation in the WTO. Essentially, it answers the question of whether the PTA adds to the WTO disciplines on subsidies. PTA disciplines on subsidies are coded “WTO =” (if the PTA essentially restates the WTO rule or provides a similar level of regulation); or “WTO +” (if the rules exceed WTO disciplines or commitments); or “WTO –” (if they provide for commitments that are more limited than WTO requirements). Since the column on WTO coverage provides key information that contributes to a preliminary understanding of the depth of the PTA, it is necessary to make a few comments on the coding approach. First, the kind of correspondence indicated by this coding is by necessity approximate, since it is often very difficult to make a definite or precise assessment. In agriculture, for example, a perfect and clear equivalence in product coverage between the WTO and a PTA is rare and is largely limited to cases where the PTA expressly refers to the WTO Agreement on Agriculture. In most of the cases, a PTA is coded “WTO =” when it covers a very large part of classified agriculture products in a similar way as the WTO. “WTO–” is used only if it is clear that the product coverage of the PTA is significantly lower than the WTO coverage. The same approach is used for the coding of the substantive disciplines.“WTO =” is used not only for those cases where the PTA provision replicates the corresponding provision in the GATT/ WTO, but also when, even though the legal details may change, there is a similar type and level of discipline, such as the prohibition of export subsidies or regulation of domestic support. 432 Subsidies Second, in some cases, the assessment is not straightforward. For example, what should be done when there is a conflict between the PTA and the WTO, as when the PTA permits something that WTO law would prohibit? Another example: With the exception of agriculture and, to a limited extent, civil aircraft, the WTO does not regulate subsidies by sector. What if a PTA has specific subsidy disciplines for certain sectors, say fisheries or steel? To deal with these cases, the coder has relied on the ultimate goal of the coding exercise; i.e., to determine the depth of integration pursued by the PTA. The assessment of the WTO coverage has thus focused on the depth of the commitment of the PTA relative to the WTO rule book.To be sure, this is not a compliance exercise.The goal is not to assess whether certain PTA provisions are WTO law compliant but whether they pursue a level of integration that is deeper (or otherwise) than the WTO. In some cases, it is clear that the PTA includes “WTO –” provisions; for example, when it authorizes parties to introduce measures on export (read: export subsidies) to compensate for the cost difference in agricultural raw materials. This provision is very common in the PTAs of the European Free Trade Association (EFTA); for example, (57) EFTA-North Macedonia (Protocol A). Equally, those rare provisions that authorize parties to maintain export subsidies have been coded “WTO –”; for example, (67) Chile-Mexico (Art. 3.13.3), which provides the possibility for a party to maintain an export subsidy if requested by the other party. One rather common example concerns the parties’ regulation of export subsidies granted by non- parties; for example, (108) US-Morocco (Art. 3.3): Where an exporting Party considers that a non-Party is exporting an agricultural good to the territory of the other Party with the benefit of export subsidies, the importing Party shall, on written request of the exporting Party, consult with the exporting Party with a view to agreeing on specific measures that the importing Party may adopt to counter the effect of such subsidized imports. If the importing Party adopts the agreed-on measures, the exporting Party shall refrain from applying any export subsidy to exports of such good to the territory of the importing Party. While the first part of the provision introduces a mechanism additional to WTO law and, hence, has been coded “WTO +”, the final provision indirectly hints at authorizing the adoption of an export subsidy and has been coded “WTO –.” In some cases, the language expressly authorizes the adoption of export subsidies in such circumstances - e.g., (111) CATFA-Dominican Republic (Art. 3.14.3) - which leads again to a “WTO –” coding. In other cases, the PTA hastens to add that any action should be consistent with WTO law; e.g., (116) US-Bahrain (Art. 2.11). This type of provision is coded “WTO =.” 433 Handbook of Deep Trade Agreements By contrast, other provisions that provide for the adoption of countermeasures, precautionary measures, or additional duties to counter certain measures or events have generally been coded “WTO =,” since the language is too general to conclude that a commitment is more limited than what is provided in the WTO rule book. See, for example, (34) EU-Faroe Islands (Art. 29.3(d)). Another case where the coding of WTO coverage has been connected to the depth of integration of the agreement relates to the question,“does the agreement provide for exemptions for legitimate subsidies?” With the exception of some agricultural subsidies, which are green- lighted (the so-called “Green Box”), WTO subsidy laws do not presently provide for specific exemptions for legitimate subsidies. When confronted with PTA provisions that expressly allow certain subsidies, the coder has assessed whether the authorization indicates a higher or lower level of commitments.12 This has meant distinguishing those provisions that allow subsidies that pursue horizontal or general objectives (such as environmental protection), public services or regional development (these partly mirror the now elapsed WTO categories of non-actionable subsidies), from those that permit sectoral aid (for example, to steel or coal).While the former have been coded “WTO +,” the latter have been coded “WTO –.” The difference between the two rests on the consideration that the former largely pursue a more general public interest and are more likely to target market failures rather than protecting specific sectors.13 It is important to highlight that sectoral rules are often coded under other entries as well, such as the question on “any other discipline for certain sectors or objectives” or “special and differential treatment.” These provisions have been coded in the same way as above. Thus, if sectoral rules are largely exceptions for subsidies or represent a case of special and differential treatment, the coding is “WTO –.” One good example where this coding was applied to sectoral (iron and steel) subsidy disciplines is (18) EC(12) Enlargement. The final provision of (195) EU Republic of Korea (Art. 11.11) is a good example of the coding approach that attempts to capture the depth of the PTA. This provision lists those subsidies that are prohibited, in particular states guarantees and support to insolvent or ailing companies. In the final part, it reads: This subparagraph [i.e., the prohibition of aid to insolvent or ailing companies] does not apply to subsidies granted as compensation for carrying out public service obligations and to the coal industry. 12 The provisions that allow certain agricultural subsidies through reference to the relevant WTO disciplines are coded “WTO =”. 13 Rodrik 2004, for example, suggests that subsidies should be as targeted as possible, with a preference for activities rather than sectors. 434 Subsidies Now, following the logic just explained that links the coding of the provision to the level of integration pursued, this exception for coal is “WTO –” and the exception for public service obligations is “WTO +.” Another good example is (2) EFTA,Annex Q (Art. 6), which essentially replicates the EU Treaty provisions on states aid that permit various forms of horizontal aid but are applicable only to the air transport sector. Here, given the clear horizontal nature of the legitimate aid, the coding has been “WTO+.” 15.3.1.2.2 (Legal) enforceability In general, the question of enforceability does not apply to definitions and coverage, but only to substantive disciplines, transparency, enforcement, and special and differential treatment provisions. The coders have used the scale established at the beginning of the Deep Integration project: - 0: non-binding - 1: non-binding provision with best efforts - 2: binding provision with no dispute settlement - 3: binding provision with state-to-state dispute settlement - 4: binding provision with only private-state dispute settlement - 5: binding provision with both state-to-state and private-state dispute settlement. The most common codes have been “3” when dispute settlement is available and “‘2” when it is not available. A coding of “4“is only limited to those rare cases where there is a national treatment obligation in the investment chapters for which investor-state dispute settlement (ISDS) is provided; for example, (199) India-Japan (Art. 96). A coding of “1” is even rarer since provisions are normally drafted in binding language. One clear example of non-binding language can be found in (10) APTA (Art. 12.e): The Participating States shall, as far as practicable, follow the provisions of relevant WTO Agreements including the Agreement on the Implementation of Article VI of the GATT 1994 and the Agreement on Subsidies and Countervailing Duties, and ensure that the provisions of this Agreement are harmoniously applied (emphasis added). Another example, which focuses on the possible outcome of consultations, is (77) Canada- Costa Rica, Art. III.13.3: Pending the elimination of trade-distorting domestic support measures, if either Party maintains such a measure which the other Party considers to be distortive of bilateral trade under this Agreement, the Party applying the measure shall, at the request of the other Party, consult with a view to making a best efforts endeavour to avoid nullification or impairment of concessions granted under this Agreement (emphasis added). 435 Handbook of Deep Trade Agreements Probably the most common example of a provision requiring best efforts concerns agricultural export subsidies. See, for example, (179) Peru–China (Art. 16, Agricultural export subsidies), which reads: 1. The Parties share the objective of the multilateral elimination of export subsidies for agricultural goods and shall work together toward an agreement in the WTO to eliminate those subsidies and avoid its reintroduction in any form. 2. No Party may maintain, introduce or reintroduce any export subsidy on any agricultural good destined for the territory of the other Party (emphasis added). While paragraph 2 is coded “2” or “3” depending on whether the PTA has a dispute settlement system applicable to agricultural subsidies, paragraph 1 is coded “1,” as “it is likely to be very difficult to prove that a party has not “cooperated.” This commitment must therefore be classified as “best endeavors.”14 The (179) Peru-China PTA (Art. 18), Domestic support measures for agricultural products, includes another example of best-endeavors language: In order to establish a fair and market-oriented agriculture trading system, the Parties agree to cooperate in the WTO agricultural negotiations on domestic support measures to provide for substantial progressive reduction in agriculture support and protection, resulting in correcting and preventing restrictions and distortions in world agricultural markets (emphasis added).) 15.3.1.2.3 Benefits to non-members This column focuses on the impact of the relevant provision, inquiring in particular whether the benefit (or, to use a different term, positive externality) of the provision de facto extends to non-members. There are two possible alternatives: “1” for a positive answer, or “0” for a negative answer. If the issue is not relevant (because it is not possible to determine the existence of an externality in the abstract), the cell has not been coded. In particular, the questions under objectives, definition, and coverage provisions have not been coded because either the question of any benefit to non-members is not relevant or the existence of any such benefit cannot be easily determined through the abstract coding of legal texts. By contrast, in most of the cases, the questions under disciplines, transparency, enforcement, and special and differential treatment could be be coded, with examples and rationales for the coding criteria indicated below. In general terms, only the cases where the benefit (or not) to non-members is clear have been coded. Thus, for example, (3) Central American Common Market (CACM) (Art. XIX) calls 14 Horn et al. 2010. 436 Subsidies for the harmonization of tax incentives.What impact this may have on non-members is unclear and is in any event something which cannot be inferred simply by the existence of a legal requirement. Hence, it was coded “0.” Do PTAs lead to discrimination against non-signatories? The answer depends on whether, for example, the prohibition (of export subsidies, local content subsidies, or subsidies causing distortions in trade or competition) is a general one or only concerns inter-party trade. Thus, if the PTA regulates subsidies through a simple reference to WTO disciplines, the benefits of constraining subsidies are general and impact non-members. The case is different for provisions that explicitly prohibit export subsidies (to agricultural goods or all goods) but restrict the prohibition to goods destined to the other party or within reciprocal trade. In these cases, there is no clear positive externality for non-members.This distinction has been reflected in the coding. National treatment obligations applicable to subsidies appear with frequency in PTAs. By definition, since the obligation normally applies only between the parties, there can be no benefit to non-members. If one moves to the enforcement part of the disciplines, a distinction can be drawn between the obligation to withdraw illegal subsidies and the right to introduce countervailing duty remedies. Here the externalities are asymmetric. The withdrawal of the subsidy may benefit everybody, while countervailing duty action only applies between the parties. 15.3.2 Selected coding issues After offering a general overview of the structure of the template and a few examples of how coding has been carried out, the chapter now specifically examines two questions which show how the coding has been useful for the overall goal of assessing the depth of subsidies provisions in PTAs. 15.3.2.1 Objective of control of subsidies PTAs rarely define the objectives of subsidy control, and when defined, they are usually very broadly phrased.While references in preambles to the pursuit of fair conditions for competition and trade (for example, (29) Turkey-Israel) are too general to be meaningful, somewhat more precise expressions of purpose in the context of subsidy rules can be occasionally found and have been coded “1.” Thus, for example, the Unfair Trade Practices chapter of (208) Peru- Mexico (Art. 9.2) outlines the general principles of subsidy control in the following way: The Parties recognize the need to eliminate export subsidies not permitted by the WTO and reject any unfair international trade and other domestic policies that cause distortions to trade between the Parties (translation by the author, emphasis added). 437 Handbook of Deep Trade Agreements More precise is the (251) EU-Georgia PTA, in which the chapter on competition under Art. 203 outlines the following principles: The Parties recognise the importance of free and undistorted competition in their trade relations. The Parties acknowledge that anti-competitive business practices and state interventions (including subsidies) have the potential to distort the proper functioning of markets and undermine the benefits of trade liberalisation. The rationale for subsidy control is extremely controversial, as evidenced, for example, by the lack of a preamble in the WTO Agreement on Subsidies and Countervailing Measures.15 This lack of agreement may explain the absence of express statements of objectives in PTAs. One therefore has to look for indirect hints. The location of subsidy disciplines (variously in the chapters on Trade in Goods, Trade in Agricultural Goods, Trade Remedies, Unfair Trade, Competition, Trade in Services or Investment) does not help to resolve the controversy but simply shows the pervasive presence of subsidy issues across the whole spectrum of PTA provisions. The very common presence of provisions on export subsidies (149 out of the 269 PTAs that include subsidy provisions),16 which largely consist of prohibitions, does indicate a relatively wide agreement on the negative welfare effects of these forms of support. At the same time, however, 120 of the 269 PTAs with subsidy provisions do not have specific provisions on export subsidies. The 55 percent share with export subsidy provisions vs. 45 percent share with no export subsidy provisions confirms that there is far from a universal consensus on the prohibition of export subsidies. One possible hint at a difference in objectives could be found in the distinction some PTAs make between provisions on subsidy control and provisions on countervailing duties. It is, in particular, in the former that some PTAs - especially those signed by the EU that set the terms for a close relationship with the EU Internal Market - introduce relatively comprehensive rules that attempt to distinguish “good” and “bad” subsidies and set up implementation mechanisms with supranational and domestic bodies. These stronger provisions on subsidy control may hint at different economic and political objectives pursued by these rules, as well as different welfare standards. In conclusion, the inquiry into the objectives of subsidies has not offered any meaningful result with respect to the depth of PTAs. 15 See, also, Sykes 2010, Spector 2009, and Buendia-Sierra 2006. 16 Only 269 out of the 283 sampled PTAs include subsidy provisions. 438 Subsidies 15.3.2.2 Exemptions for legitimate subsidies, public services, and sectoral aid There are 59 PTAs with provisions exempting legitimate subsidies, 24 with specific disciplines for certain sectors or objectives, and 14 that shelter support for public services. The coding of these questions about exemptions may offer some useful indication about the depth of subsidy disciplines. They largely concern the EU, some treaties concluded by the EU or by countries closely linked to the EU, and those PTAs that make reference to the WTO Agreement on Agriculture. Various types of exemptions are coded, including those for public service subsidies, horizontal objectives (e.g., environmental protection, research and development), regional aid, and special rules protecting public support in given sectors (agriculture, steel, coal, textiles). The significance of these provisions for the depth of integration has been discussed at length in the section above on WTO coverage. One good example of a public service subsidy exemption can be found in (56) EU-South Africa (Annex IX(a)), which states that the rules on public aid “should not obstruct the performance in law or in fact of the operation of services of general economic interest assigned to public undertakings.” The same PTA presents an interesting recognition that subsidies may be adopted to pursue public policy objectives (see Art. 41) and that this should be taken into consideration should a controversy between the parties arise. Art. 42 reads: If the Community or South Africa considers that a particular practice is incompatible with the terms of Article 41, and that such practice causes or threatens to cause serious prejudice to the interests of the other Party or material injury to its domestic industry, the Parties agree, where it is not adequately dealt with under existing rules and procedures, to enter into consultations with a view to finding a mutually satisfactory solution. Such consultations will be without prejudice to the Parties’ rights and obligations in terms of their respective laws and international commitments. Either Party may invite the Cooperation Council to examine, in the context of such consultation, the Parties’ public policy objectives justifying the grant of public aid referred to in Article 41. Following the logic in the section on WTO coverage, these and similarly worded subsidy provisions have been coded “WTO +.” 439 Handbook of Deep Trade Agreements 15.4. ANALYSIS: DESCRIPTIVE STATISTICS 15.4.1. General overview of the evolution and distribution of subsidy provisions This section begins with basic descriptive statistics on the evolution and distribution of subsidy provisions in the 283 PTAs examined. 15.4.1.1 Evolution of the number of PTAs with subsidy provisions The number of PTAs with subsidy provisions has increased dramatically over time (Figure 15.1). Even from a quick look, it is immediately clear that nearly all PTAs include subsidy provisions (269 out of 283) and that subsidy provisions have, from the beginning (1957), always been a feature of PTAs.17 Those PTAs that do not feature any subsidy provision amount to a mere 5.2 percent of the total. Figure 15.1: Evolution of the number of PTAs with subsidy provisions Source: Deep Trade Agreements Database. An initial breakdown introduces a distinction among three different types of subsidy provisions: (a) subsidy disciplines (which include prohibitions of export subsidies and other trade-distorting support measures, and more rarely, provisions exempting certain legitimate subsidies); (b) national treatment provisions; and (c) countervailing duty provisions (Figures 15.2 and 15.3).18 17 All figures and tables refer to the 2010-2017 period, mainly for the sake of consistency with the the other chapters. However, the most recent PTA coded in the context of the mapping exercise underlying this chapter dates back to February 2016. 18 Since the main focus of the dataset is on subsidy disciplines, the template simply indicates whether or not national treatment and countervailing duties are included in the PTA, with the important caveat that the rest of the entries pertain specifically to subsidy disciplines. 440 Subsidies Figure 15.2: Evolution of the number of PTAs with subsidy provisions Source: Deep Trade Agreements Database. National treatment provisions include general obligations that may well apply to subsidies. The number of national treatment provisions (Figure 15.2) has fluctuated (from 38, in the pre-1995 period, to 26 in the 1995-1999 and 2000-2004 periods, then rising again to 31 in 2005-2009, and then dramatically decreasing to 19 in 2010-2017).19 The full significance of these numbers comes out if considering their share in PTAs with subsidy provisions in the Figure 15.3: Evolution of the share of PTAs with different types of subsidy provisions Source: Deep Trade Agreements Database. 19 The count excludes those provisions that simply incorporate GATT Article III because of its sweeping exclusion of subsidies to domestic producers (paragraph 8(b)). 441 Handbook of Deep Trade Agreements same periods (Figure 15.3). What is clear is a consistent and sharp decrease (from 73 percent before 1995, to 63 percent in the 1995-1999 period, to 49 percent in the 2000-2004 period, to 39 percent in 2005-2009, and finally to 33 percent in 2010-2017). These basic statistics show that general national treatment provisions have been increasingly seen as not appropriate for regulating subsidies. This especially applies to goods. In absolute numbers, fewer and fewer PTAs have included national treatment provisions applicable to subsidies in the goods sector.20 The relative increase of these provisions in the services and investment sectors is largely due to the increasing presence of services and investment provisions in PTAs, but in absolute terms, exclusions of subsidies remain extremely common.21 In technical-legal terms, parties to PTAs shelter subsidies from national treatment obligations largely by incorporating GATT Article III (with its built-in exclusion of domestic subsidies), or by expressly excluding subsidies from services or investment commitments at large or, more specifically, from national treatment provisions. This general decline in the relevance of national treatment in subsidies corresponds to a rise of subsidy disciplines and CVD provisions. This already comes out from the absolute numbers of the relevant provisions (Figure 15.2),22 but it becomes clearer from the evolution of the share of PTAs with subsidy provisions (Figure 15.3). Period after period, there is an increase in both subsidy disciplines and CVD provisions. There is a moderate increase with a trend towards stabilization in subsidy disciplines (from 54 percent and 56 percent in the pre-1995 and 1995-1999 periods, to 70 percent in 2000- 2004, and 79 percent in both 2005-2009 and 2010-2017). There is an even more marked increase in CVD provisions (starting from a lower 37 percent in pre-1995, to equaling subsidies disciplines at 56 and 70 percent in the next two periods, and finally, climbing and surpassing subsidy disciplines in the last two periods - 86 percent in 2005-2009 and 89 percent in 2010-2017). In conclusion, subsidy disciplines and countervailing duties represent the two strongest features of the regulation of subsidies in PTAs, with a cumulative share rising from 37 percent in the pre-1995 period to 89 percent in more recent times. If one considers any subsidy provision (subsidy disciplines, national treatment obligations, and CVDs), the coverage is always above 90 percent and reaches 98 percent in 2000-2004, 95 percent in 2005-2009, and 96 percent in 2010-2017. 20 In the periods used in the figures (pre-1995, 1995-1999, 2000-2004, 2005-2009, and 2010-2017), there were 36, 25, 15, 16, and 5 PTAs with national treatment provisions for goods. 21 In the same periods, the numbers are, for services, 8, 8, 14, 13, and 13; and for investment, 4, 1, 5, 15, and 9. 22 The poor numbers in the 2010-2017 period are due mostly to the fact that only 57 PTAs were signed in that period (which contrasts with 80 signed in the 2005-2009 period but is comparable to the 53 signed in the 2000- 2004 period). 442 Subsidies 15.4.1.2 Distribution and evolution of PTAs with subsidy provisions by level of development of its members This section considers the distribution and evolution of PTAs with subsidy provisions by level of development of the member countries. Developed and developing countries are defined following the World Bank country classification of 2017.23 Subsidy provisions are part of the standard regulation of all PTAs concluded between developed countries (100 percent) and of virtually all PTAs (98 percent) concluded between developed and developing countries (Figure 15.4). If one considers those agreements entered into between developing countries, the share of those that include subsidy provisions is lower but still very significant (88 percent). Figure 15.4: Share of PTAs with subsidy provisions by level of development Source: Deep Trade Agreements Database. Considering the evolution of PTAs over time contributes two additional elements to the analysis (Figure 15.5). First, there is a decrease (from 95 percent in the pre-1995 period to 90 Figure 15.5: Evolution of the share of PTAs with subsidy provisions by level of development, over time Source: Deep Trade Agreements Database. 23 Developing countries are composed of low-income and lower-middle-income economies, whereas developed countries are upper-middle-income and high-income. Low-income economies are defined as those with a gross national income (GNI) per capita, calculated using the World Bank Atlas method, of USD 1,005 or less in 2016; lower-middle-income economies are those with a GNI per capita between USD 1,006 and USD 3,955; upper-middle-income economies are those with a GNI per capita between USD 3,956 and USD 12,235; and high-income economies are those with a GNI per capita of USD 12,236 or more. 443 Handbook of Deep Trade Agreements percent in 1995-1999) and then an increase of subsidy provisions in PTAs between developed and developing countries, reaching 100 percent coverage from the year 2000 onwards. Second, although the share of PTAs between developing countries with subsidy provisions is more or less constant, it is interesting to note a steady increase in the three first periods (87 percent, before 1995; 95 percent in 1995-1999; and 94 percent in 2000-2004), a fall to 81 percent in 2005-2009, and a return to the pre-WTO level of 87 percent in the most recent period (2010-2017). 15.4.1.3 Distribution and evolution of PTAs with subsidy provisions by geographic group The distribution and evolution of PTAs with subsidy provisions are also considered by geographic group. Geographic groups are defined following the World Bank classification (Figure 15.6).24 There is an extremely high presence of subsidy provisions in the PTAs of most of the geographic groups (from 96 to 100 percent), with lower levels in South Asia (82 percent) and Sub-Saharan Africa (72 percent). Figure15.6: Share of PTAs with subsidy provisions by geographic region/group Source: Deep Trade Agreements Database. Note: The figures listed at the bottom of each bar represent the number PTAs with subsidy provisions. If one, however, combines these data with the evolution of these provisions across trade agreements of the various geographic groups (Table 15.1), it is possible to notice a significant change toward a consistent and almost universal presence of subsidy provisions after 1999. While before that year there were indeed significant variations across the groups, after 1999 virtually all PTAs in all geographic areas and all periods included subsidy provisions. South Asia consistently scores high (hitting 100 percent in three of the relevant periods, including the most recent one), with its low score in Figure 15.6 (82 percent) due mainly to the low share (50 percent) in the pre-WTO 24 See https://blogs.worldbank.org/opendata/new-country-classifications-income-level-2018-2019. 444 Subsidies period. Even Sub-Saharan Africa reaches 100 percent share in the most recent period, with only Central Asia backsliding to 85 percent (from 100 percent in the previous two periods). For certain groups, the 100 percent share of new PTAs with subsidy provisions even precedes 1999 to stretch back to 1995 (EU, Latin America and Caribbean) and before 1995 (Central Asia, EFTA, Middle East and North Africa, North America, Sub- Saharan Africa). The EU would score 100 percent if (4) the Overseas Countries and Territories (OCTs) were not considered in this coding.25 Table 15.1: Share of new PTAs with subsidy provisions by geographic group, over time (%) Year of signature Before 1995 1995 to 1999 2000 to 2004 2005 to 2009 2010 to 2017 Central Asia 100 94 100 100 85 EFTA 100 100 100 100 100 East Asia & Pacific 88 0 100 100 100 European Union 93 100 100 100 100 Latin America & Caribbean 89 100 100 100 100 Middle East & North Africa 100 90 100 100 100 North America 100 100 100 100 100 South Asia 50 100 100 80 100 Sub-Saharan Africa 100 60 50 71 100 Source: Deep Trade Agreements Database. 15.4.2 Analysis of scope Following the analysis of the general traits of subsidy provisions in PTAs, this section focuses more specifically on the scope of these agreements. 15.4.2.1 Most common subsidy provisions in PTAs The most common subsidy provisions in PTAs include transparency, dispute settlement, countervailing duties, and prohibition of trade-distorting subsidies (Figure 15.7). Some of these provisions will be subject to deeper analysis in the next section. It may be useful to recall that of the 283 PTAs in the sample, 269 include subsidy provisions. Interestingly, the five most common provisions (ranking between 199 and 243) do not deal 25 This is a Decision of the Council of the EU which has not been coded because it does not regulate state subsidies of the parties, but only the use of EU structural or regional funds in overseas countries and territories. 445 Handbook of Deep Trade Agreements with disciplines or commitments. Two provisions are general and not specific to subsidies (existence of institutions dealing with transparency and enforcement, 243; dispute settlement mechanism, 203).The two most common provisions specifically designed to address subsidies are transparency provisions (209) and CVDs (199). Just as common (200) are those provisions that either expressly or by reference include definitions of subsidies (mostly general definitions, but also definitions of export agricultural subsidies). Figure 15.7: Most common types of subsidy provisions in PTAs Source: Deep Trade Agreements Database. The two most common provisions with commitments are those that prohibit or regulate export subsidies (149) or subsidies that distort trade or competition (128). The provisions on export and domestic subsidies are analyzed further below. Then come those PTAs (97) that feature a national treatment obligation applicable to subsidies in the goods sector. Similar obligations applicable to services or investment are less common (56 and 34). PTAs do not expressly regulate local content subsidies. The 69 PTAs that regulate local content do so through reference to WTO disciplines. The 73 PTAs that include specific provisions on agricultural subsidies concern regulation of domestic support, while legitimate agricultural subsidies are coded in a separate specific question. Fifty-nine PTAs have provisions exempting legitimate subsidies (mostly regional aid, agricultural subsidies, sectoral aid, and public service support). Provisions that carve out the financing of public services have been coded with a separate question and show a very low frequency: 14 PTAs. If, as noted, CVDs are the most common remedy to counter subsidies (199), it is also the case that 86 PTAs provide for the obligation to withdraw illegal subsidies. Other remedies, such as cooperation mechanisms to counter export subsidies granted by third parties, account for 53 provisions.With a score of 60, the latest significant (in terms of frequency) provision provides the capacity for enacting secondary legislation. 446 Subsidies The remaining provisions rarely feature in PTAs: special treatment or cooperation commitments (28); any other discipline for specific sectors or objectives (24, including support to the coal, steel, or textile sectors); the existence of national authorities to administer state aid (12); specific regulation of fisheries subsidies (8); and ceiling or de minimis threshold to permitted subsidies (2 and 1, respectively). In conclusion, if one wanted to single out the most relevant provisions and assess their frequency, transparency commitments are present in 78 percent of PTAs with subsidy provisions; and dispute settlement provisions applicable to subsidies are present in 75 percent. Crucially, the most common provisions disciplining subsidies are those on CVDs (74 percent) and those on export subsidies (55 percent). General national treatment obligations are included in only in 69.5 percent of PTAs with subsidy provisions. Provisions regulating other trade-distorting subsidies and domestic support to agriculture, respectively, account for 48 and 27 percent of subsidy provisions. Provisions expressly sheltering “good” subsidies amount to a mere 27 percent. As noted, a very high percentage of subsidy provisions are enforceable through dispute settlement mechanisms, which are included in 202 PTAs (75 percent). Most involve state- to-state dispute settlement and some also cover investor-state dispute settlement (ISDS), in particular to enforce national treatment claims in investment chapters. For those PTAs that do not provide for dispute settlement, this is mostly due to provisions that specifically exclude CVDs from dispute settlement and do not provide for dispute settlement for other subsidy provisions.26 The settlement of subsidy disputes is never specifically excluded. In some rare cases, PTAs do not provide for dispute settlement at all or, more frequently, simply provide for non-binding consultation or negotiation to settle disputes. This is particularly common in PTAs concluded by Central Asian countries. 15.4.2.2 Most common provisions by level of development The analysis shows that North-South (Developed-Developing) and North-North (Developed-Developed) PTAs have the higher number of subsidy provisions, while South- South (Developing–Developing) PTAs lag behind (Figure 15.8).Tellingly, South–South PTAs feature almost 40 percent fewer subsidy provisions than North–South PTAs. The outcome is similar if one focuses on subsidy provisions with dispute settlement. North–North and North–South PTAs have a similar percentage, with between 80 percent and 84 percent of subsidy provisions being legally enforceable through dispute settlement. The percentage falls dramatically, to 58 percent, with South–South PTAs. 26 It should be recalled that the exclusion of CVDs from dispute settlement has led to a “ 0” coding only if CVD provisions are the only subsidy provisions in the PTA. If other subsidy provisions are present and are justiciable through binding dispute settlement, the coding is “1.” 447 Handbook of Deep Trade Agreements Figure 15.8: Most common provisions by level of development Source: Deep Trade Agreements Database. 15.4.2.3 Most common provisions by geographic group The PTAs of EFTA, EU, and Middle East & North Africa have the higher number of subsidy provisions: between 9.3 and 11.4, with an average of 10.5 (Figure 15.9).The remaining groups lag behind, and it is perhaps possible to identify two different groupings.Those with almost 7 provisions (East Asia & Pacific, Central Asia, and Latin America & Caribbean), and those with 5.4 provisions, on average (North America, South Asia, and Sub-Saharan Africa).The PTAs of EFTA and EU have the highest number of subsidy provisions because they very often include disciplines on agricultural subsidies (through reference to WTO disciplines) and at times also on services and investment. Figure 15.9: Most common provisions by geographic group Source: Deep Trade Agreements Database. Looking at how many subsidy provisions are legally enforceable through dispute settlement (Figure 15.10), most of the geographic groups share a very high level of legal enforceability (between 80 percent and 92 percent). However, two groups score significantly lower: Central Asia and EFTA. Only 49 percent of Central Asia PTAs are legally enforceable, 448 Subsidies which is due to the very common stated intention of the parties to resolve disputes through negotiations. In EFTA, a significant share of PTAs - 43 percent - expressly excludes subsidies (or countervailing duties) from dispute settlement. Figure 15.10: Subsidy provisions with dispute settlement by geographic group Source: Deep Trade Agreements Database. 15.4.2.4 Notable countries Several countries or regional groups (EFTA, EU) have been selected for a deeper analysis, mainly because of their active role in signing PTAs (Figure 15.11).The examination of the most common provisions of these countries shows interesting differences. These countries can be divided into various groups. At the top are EFTA, the EU, and Turkey, with scores between 9.4 and 11.4.This commonality depends on the fact that all of these countries participate in an integration process which has the EU at its center. (EU PTAs have a slightly lower score, mostly because the coding results generated by accession treaties may focus only on specific subsidy issues.) These PTAs normally include various subsidy provisions: from definitions to provisions on subsidies that distort trade or competition; agricultural export subsidies and agriculture support; legitimate subsidies; public services; sectoral aid; transparency requirements; CVDs; dispute settlement; and significant institutional mechanisms dealing with transparency and enforcement. The second group, with scores between 7.5 and 7.8, includes China, Singapore, and Central America. In the third group are the large majority of other countries, including Canada, Korea, Mexico, and New Zealand, with scores between 6.2 and 6.8.The content of subsidy provisions of the second and third groups is comparable to but less rich than that of the first group, with fewer provisions on subsidies that distort trade and competition, legitimate subsidies or public services, transparency, or dispute settlement. At the bottom of the group, with scores below 6, are the US at 5.2, Chile at 5.7, and Japan at 5.9.What leads to the low scores of these countries is the minimal regulation of subsidies included in their PTAs, which normally focus mostly on agricultural export subsidies and CVDs. 449 Handbook of Deep Trade Agreements Figure 15.11: Most common provisions by notable countries Source: Deep Trade Agreements Database. 15.4.3 Specific analysis of types of provisions This section sketches a few descriptive statistics concerning specific types of subsidy provisions on coverage, export subsidies, domestic subsidies, transparency, and enforcement. 15.4.3.1 Coverage (sectors) A preliminary view of the coverage of subsidy provisions in PTAs shows that, while all PTAs with subsidy provisions regulate subsidies in the goods sector, only a small number regulate subsidies in services. By contrast, a large majority cover agriculture and fisheries (Figure 15.12). If one considers the subsidy coverage of PTAs by level of development (Figure 15.13), the only interesting difference emerging is that the largest number of service subsidy disciplines belongs to North-South PTAs: 40 PTAs, representing 27.6 percent of the total number of PTAs with subsidy provisions. North-North and especially South-South PTAs score significantly lower, with service subsidy disciplines representing respectively 17.5 percent and 9.5 percent of all relevant PTAs.27 While the interest in service subsidy disciplines seems to be evenly spread among developing countries, EFTA and Japan are the most active among developed countries. At the geographic level, the distribution of the provisions on service support reveals interesting variations (Table 15.2). Almost half the PTAs of the EU, EFTA, and Sub-Saharan Africa include service subsidy disciplines. The presence of these disciplines decreases in the 27 The 7 PTAs in the North-North group are (16) Australia New Zealand Closer Economic Agreement (ANZCERTA), (97) EFTA-Chile, (220) EFTA-Hong Kong SAR, China, (78) EFTA-Singapore, (115) EFTA- Korea, (75) Japan-Singapore, and (172) Japan-Switzerland. The 8 PTAs in the South-South group are (269) Agadir Agreement, (45) Commonwealth of Independent States (CIS), (41) Eurasian Economic Community (EAEC), (47) Economic and Monetary Community of Central Africa (CEMAC), (255) Eurasian Economic Union (EAEU), (63) Georgia-Russian Federation, (234) Ukraine- Montenegro, and (50) West African Economic and Monetary Union (WAEMU). 450 Subsidies other groups to reach the lowest level in North America and South Asia, with 1 agreement each for (71) US-Jordan and (199) India-Japan. By contrast, agriculture and fisheries subsidy disciplines are present in almost all PTAs of all geographic groups except for the EU and Central Asia, which still have a significant number of agreements that do not regulate subsidies in these sectors. Figure 15.12: Coverage (sectors) Source: Deep Trade Agreements Database. Figure 15.13: Coverage (sectors) by level of development Source: Deep Trade Agreements Database. Table 15.2: Coverage (sectors) by geographic groups Goods Services Agriculture Fisheries Central Asia 75 12 63 63 EFTA 35 14 34 32 East Asia & Pacific 85 17 82 83 European Union 46 20 32 30 Latin America & Caribbean 80 9 80 80 Middle East & North Africa 36 9 29 29 North America 25 1 25 25 South Asia 18 1 16 16 Sub-Saharan Africa 13 6 13 13 Source: Deep Trade Agreements Database. 451 Handbook of Deep Trade Agreements 15.4.3.2 Export subsidies As noted above, the most common subsidy provisions with commitments are those that prohibit or regulate export subsidies.These are present in 149 of the 269 PTAs in the sample, which means that 55 percent of all PTAs with subsidy provisions include export subsidy rules.The presence of export subsidy disciplines is, however, uneven by level of development. While export subsidy rules feature strongly in North-North and North-South PTAs (70 and 65 percent of PTAs, respectively), their frequency is low in South-South PTAs (26 percent). Prohibitions of export subsidies, either to all goods or only to agricultural goods, are very common in PTAs. Both instances have been coded under this question. Less frequent but certainly common are provisions that permit export subsidies under certain circumstances. For example, EFTA PTAs often authorize the parties to adopt “measures on export” in order to compensate for the cost difference in agricultural raw materials. Finally, it is very common for the parties to state they share the objective of the multilateral elimination of export subsidies and that they will work together in the WTO to reach an agreement in this respect. This information is enriched by looking at the frequency of export subsidy provisions by geographic group (Figure 15.14). Export subsidy disciplines are very common in PTAs concluded by Latin America & Caribbean (81 percent) and North American (80 percent) countries.They also feature strongly in EFTA PTAs (86 percent), mostly through reference to WTO subsidy disciplines. Export subsidy disciplines are also fairly common in the East Asia & Pacific (70 percent) and in the Middle East & North Africa (57 percent). They are very low in all remaining groups, with Central Asia being the lowest scorer (19 percent). Export subsidy provisions are rare in EU PTAs, mostly due to the specific regulatory technique used in these agreements, which generally include a broad prohibition of distorting state aid or subsidies, modeled on the definition of EU state aid. Figure 15.14: Export subsidies by geographic group Source: Deep Trade Agreements Database. Note: The figures listed at the bottom of each bar represent the number PTAs with export subsidy provisions. 452 Subsidies 15.4.3.3 Domestic subsidies A few questions in the template focus on domestic subsidy commitments (Figure 15.15). Many PTAs include specific provisions on domestic subsidies. In particular, 128 PTAs include provisions prohibiting subsidies that distort trade or competition; and 73 include provisions regulating support in agriculture (this number does not include agricultural export subsidies that have been coded under the general export subsidy question). Interestingly, 59 PTAs include provisions exempting legitimate subsidies (which includes both exceptions applying to various types of subsidies, mostly regional and sectoral aid, and exceptions for agricultural subsidies, mostly through reference to WTO disciplines). Another 14 PTAs specifically shelter public service support. Figure 15.15: Domestic subsidies Source: Deep Trade Agreements Database. The analysis of domestic subsidy provisions by level of development shows a similar pattern, with provisions on distorting subsidies being the most common, normally followed by provisions on agricultural support (Figure 15.16). Exemptions for legitimate subsidies (which may cover both horizontal and sectoral aid) are, on average, as frequent as those on agricultural support, with South-South and particularly North-North PTAs showing a significant variance and a distinctly higher presence of provisions on agricultural subsidies. Figure 15.16: Domestic subsidies by level of development Source: Deep Trade Agreements Database. 453 Handbook of Deep Trade Agreements The prevailing pattern, which shows provisions on distorting subsidies as the most common, followed by those on agricultural subsidies or on legitimate subsidies, is largely followed in the breakdown of PTAs by geographic groups (Table 15.3). The most significant variations pertain to the different frequency of agricultural and legitimate subsidies provisions, and also, to a minor extent, to the frequency of public service support provisions. One group (comprising Central Asia, the EU, and Middle East & North Africa) shows a clear prevalence of provisions on legitimate subsidies. This is particularly clear for the EU with a 2.2 (26/12) ratio of legitimate subsidies to agricultural provisions. These data are even more interesting if they are read together with the number of public service support provisions (12) in EU PTAs, since they show that these PTAs include several provisions on what are considered to be “good” subsidies, which slightly surpass those on “bad” (distorting) subsidies (38 vs. 34).28 The legitimate subsidies/agricultural provisions ratios for the Central Asia and Middle East & North Africa PTAs are, respectively, 1.8 (20/11) and 1.4 (12/9). These two groups follow the trend of EU PTAs, due to the regulatory pattern of PTAs concluded with the EU. In both these groups, legitimate subsidies provisions are less than half the number of provisions on distorting subsidies, and agricultural subsidies provisions are approximately one third.The next group encompasses those PTAs where the legitimate to agricultural subsidies ratio is one or close to one.This group includes South Asia, Sub-Saharan Africa, and EFTA.The ratio is exactly one for the first two groups (with, respectively, a 2/2 and 3/3). For EFTA, it is very close to one (0.85, 17/20), and it becomes one if public service support measures are factored in. Where the PTAs in this group vary is in the distorting subsidies/agricultural subsidies ratio, which is 4 (8/2) in South Asia PTAs, 2.3 (7/3) in Sub-Saharan Africa, and 1.6 (32/20) Table 15.3: Domestic subsidies by geographic groups Distorting Legitimate Agricultural Public service subsidies subsidies subsidies support Central Asia 41 20 11 2 EFTA 32 17 20 3 East Asia & Pacific 36 10 18 2 European Union 34 26 12 12 Latin America & Caribbean 22 11 34 1 Middle East & North Africa 27 13 9 2 North America 4 0 9 0 South Asia 8 2 2 0 Sub-Saharan Africa 7 3 3 1 Source: Deep Trade Agreements Database. 28 Clearly, the “bad” subsidies group is higher if provisions on export and local content subsidies are also considered. 454 Subsidies in EFTA PTAs. The third and final group (North America, Latin America & Caribbean, and East Asia & Pacific) includes very few provisions on legitimate subsidies (or public services). The most extreme example is North America PTAs, which do not feature any provision on “good” subsidies. Latin America & Caribbean and East Asia & Pacific PTAs provide a more nuanced picture with a negative ratio, though they include a non-insignificant number of provisions on legitimate subsidies. In particular, Latin America & Caribbean PTAs have a 0.3 (11/34) legitimate subsidies/agricultural subsidies ratio. For their part, East Asia & Pacific PTAs have a 0.55 (10/18) ratio. Interestingly, in both North America and Latin America & Caribbean PTAs agricultural subsidies provisions are more common than those on general distorting subsidies, with respectively a 2.25 (9/4) and 1.5 (34/22) ratios. 15.4.3.4 Transparency The pattern of transparency provisions in PTAs (Figure 15.17) shows a significant prevalence of notification requirements (128). Those provisions that provide for various types of deliberation follow with a lower frequency (54). Other transparency requirements, and cross- notification and third-party submissions, are rarer. Other transparency requirements, with a score of 19, include, for example, surveillance by national authorities and PTA bodies (2, EFTA); requirements to create repositories on laws and measures (25, ASEAN Free Trade Area, AFTA); an inventory of all aid schemes;29 notification of restructuring measures, including details of state aid granted;30 the permanent monitoring and analysis of agricultural support policies;31 and the requirement to carry out studies.32 Figure 15.17: Transparency provisions Source: Deep Trade Agreements Database. 29 (121) Albania, (132) EU-Montenegro, (142) EU-Bosnia and Herzegovina, (184) EU-Serbia, (250) EU-Ukraine. 30 (70) EU-North Macedonia, (121) EU-Albania, (132) EU-Montenegro, (142) EU-Bosnia. 31 (100) Panama-El Salvador, (164) Panama-Costa Rica, (176) Panama-Honduras. 32 (196) Guatemala-Taiwan, China. 455 Handbook of Deep Trade Agreements This general pattern is confirmed when PTAs are grouped by level of development (Figure 15.18). In all groups, there is a significant prevalence of notification requirements, as opposed to all other transparency provisions. This is particularly marked in North-South PTAs, while North-North and South-South PTAs show a higher balance among all the different transparency provisions. Figure 15.18: Transparency provisions by level of development Source: Deep Trade Agreements Database. The “other requirements” category features prominently in EU PTAs (Table 15.4). Central Asia and Latin America & Caribbean PTAs also include many additional transparency requirements, mainly in the agriculture sector.While East Asia & Pacific, EU and Latin America & Caribbean PTAs generally have high notification requirements, deliberation provisions are particularly common in the agreements signed by the EU (25) and Central Asia (21), and they refer to various forms of assessment of the impact of individual measures. The frequency Table 15.4: Transparency provisions by geographic groups Notification Deliberation Cross-notification Other requirements and submissions requirements Central Asia 42 21 2 7 EFTA 30 9 5 3 East Asia & Pacific 38 6 0 5 European Union 38 25 4 10 Latin America & Caribbean 26 13 1 7 Middle East & North Africa 27 12 1 0 North America 5 1 0 1 South Asia 7 0 0 0 Sub-Saharan Africa 4 1 0 0 Source: Deep Trade Agreements Database. 456 Subsidies of deliberation provisions in the PTAs of Central Asia is due mainly to the EU’s influence. Deliberation provisions are also fairly common in Latin America & Caribbean, Middle East & North Africa, and EFTA countries. As far as the two latter groups are concerned, one has to once again consider the influence of the EU model. East Asia & Pacific, North America, South Asia, and Sub-Saharan Africa PTAs largely rely on simple notification requirements. 15.4.3.5 Enforcement Finally, it is useful to briefly consider the key enforcement indicators: dispute settlement, the obligation to withdraw illegal subsidies, and CVDs. In general, dispute settlement and CVDs are present in very similar numbers (respectively, 203 and 199) in the PTAs with subsidy provisions. The obligation to withdraw illegal subsidies is rarer (86) and mostly refers to those cases where the PTA makes reference to WTO disciplines. The breakdown of these provisions by level of development largely follows the patterns highlighted above, with a similar prevalence of dispute settlement and CVD provisions. The frequency of withdrawal obligations is about half that of the other two enforcement measures in North-South PTAs, and approximately one third of the other measures in both South-South and North-North PTAs. Showing the distribution by geographic groups, however, gives more interesting results (Table 15.5). In general, most of the groups follow the pattern that has just been described. There are three groups, however, that show interesting variations. EFTA and Central Asia PTAs rely more on CVDs than dispute settlement. This is particularly marked in EFTA agreements, with a 2/3 ratio of dispute settlement to CVDs. By contrast, EFTA features many provisions providing for the withdrawal of illegal subsidies, mostly through the very frequent reference to WTO law subsidy disciplines. North America, for its part, seems to present a different ratio between dispute settlement and CVDs, with a prevalence of the former. Table 15.5: Enforcement provisions by geographic groups Dispute settlement Withdrawal Countervailing duties Central Asia 37 29 42 EFTA 20 27 31 East Asia & Pacific 75 30 72 European Union 37 17 33 Latin America & Caribbean 71 15 69 Middle East & North Africa 31 19 31 North America 23 4 16 South Asia 16 5 15 Sub-Saharan Africa 12 2 10 Source: Deep Trade Agreements Database. 457 Handbook of Deep Trade Agreements 15.5. CONCLUSIONS The goal of the mapping described in this chapter was to determine the depth of subsidy provisions in 283 PTAs. This chapter has made a start in providing a few descriptive statistics and some initial analysis. Few very general, macro-patterns can be noted. Given their very high frequency (in 95 percent of all PTAs), subsidy provisions represent one of the standard chapters of trade regulation. In terms of content, a high level of alignment with WTO subsidy disciplines can be noted. In most cases, PTAs simply refer to WTO subsidy disciplines. Additional commitments are largely of an ancillary or procedural - not substantive - nature. If one wanted to identify certain focal points, these are certainly export subsidies, agricultural support, and CVDs. Macro-variations within PTAs by level of development largely show a lower presence and extent of subsidy provisions in South-South PTAs. (North-North and North-South PTAs are largely comparable, though there are a few interesting variations.) The consideration of geographic areas gives the most interesting results. Two come out prominently. On the one hand, the most developed subsidy and state aid disciplines can be found in the PTAs signed by the EU or by those countries/blocs that, for various reasons, are close to the EU (for example, EFTA, Middle East & North Africa, and, partly, Central Asia). Even a quick perusal of these PTAs shows a different level of ambition and hence of regulation. On the other hand, the few PTAs signed by the US seem to adopt a completely different, minimal approach to subsidy regulation. With important variations, the other groups seem to represent various middle ways between these two poles. These are just the initial sketches of a very preliminary analysis. More could be done to determine the depth of these disciplines. In general terms, the depth of disciplines does not depend only on the commitments (obligations or prohibitions) in PTAs. Rather, it depends on a combination of three elements: coverage (sectors and entities covered), commitments (number, type, and enforceability), and institutional setting (governance, transparency, remedies, dispute settlement) and their variations. That being said, commitments do indeed constitute the central element to assess the depth of the relevant disciplines. One important aspect, in particular, is whether and how commitments balance the positive and negative spill-overs of subsidies. As noted in the introduction, this may well constitute a good proxy for the depth of the relevant disciplines, if only because the attempt to balance these effects signals the acceptance of significant trade-offs by and between the parties and, arguably, their stepping towards a more integrated path (at least with respect to the relevant provisions). This obviously applies to the EU and its very developed state aid regime but, with respect to agricultural products, is also present in the WTO disciplines on the non-distorting “green box” agricultural subsidies to which PTAs very often refer. 458 Subsidies ACKNOWLEDGMENTS I would like to thank all participants in the 2018 Washington, DC, workshop where a preliminary version of this chapter was presented. A special thank-you goes to Alvaro Raul Espitia Rueda for his painstaking technical support. All errors remain those of the author. 459 Handbook of Deep Trade Agreements REFERENCES Bacon, Kelyn. 2013. European Union law of state aid. Oxford: Ehlermann, Claus Dieter, and Michelle Everson. 2001. Oxford University Press. European Competition Law Annual 1999: Selected issues in the field of state aid. Oxford: Hart Publishing. Bagwell, Kyle, George A. Bermann, and Petros C. Mavroidis. 2010. Law and economics of contingent protection in international Fahkri, Michael. 2014. The institutionalization of free trade trade. Cambridge: Cambridge University Press. and empire: A study of the 1902 Brussels Convention. London Review of International Law, 49-76. Bagwell, Kyle, and Robert W. Staiger. 2006. Will international rules on subsidies disrupt the world trading system? American Friederiszick, H.W., L.W. Roller, and Vincent Verouden. 2007. Economic Review 96 (3): 877-95. European state aid control: An economic framework. In Handbook of antitrust economics, edited by Paolo Bucirossi. Cambridge, MA: Biondi, Andrea. 2018. The House of Lords EU Internal MIT Press. Market Sub-Committee’s inquiry on Brexit: State aid. Evidence submitted September 14, 2017. Hancher, Leigh, Tom Ottervanger, and Piet J. Slot. 2012. EC state aid. London: Sweet and Maxell. Biondi, Andrea, Piet Eeckhout, and James Flynn. 2004. The law of state aid in the European Union. Oxford: Oxford Hofmann, Claudia, C.H. Herwig, and Claire Micheau. 2016. State University Press Aid Law of the European Union. Oxford: Oxford University Press. Blauberger, Michael. 2011. State aid control from a political Hofmann, Claudia, Alberto Osnago, and Michele Ruta. science perspective. In Research handbook on European state aid 2017. Horizontal depth: A new database on the content of law, edited by Erika Sczysczack. Cheltenham, UK: Edward preferential trade agreements. Policy Research Working Paper Elgar. 7981, World Bank, Washington, DC. Bourgeois, Jacques H.J. 1991. Subsidies and international Horn, Henrik, Petros C. Mavroidis, and André Sapir. 2010. trade—a European lawyer’s perspective. Boston: Kluwer Law and Beyond the WTO? An anatomy of EU free trade agreements. Taxation Publishers. World Economy 33(11): 1565-88. Brou, Daniel, and Michele Ruta. 2012. A commitment Horn, Maggi, and Staiger 2010.Trade agreements as endogenously theory of subsidy agreements. CESifo Working Paper 3945, incomplete contracts. American Economic Review 10(1). CESifo GmbH, Munich. Hufbauer, Gary Clyde, and Joanna Shelton-Erb. 1984. Subsidies in Buendia-Sierra, Jose-Luis. 2006. Not like this: Some skeptical international trade. Cambridge, MA: MIT Press. remarks on the “refined economic approach” to state aid. In Proceedings of the 4th Experts’ Forum on New Developments in Luengo Hernández de Madrid, Gustavo. 2007. Regulation European State Aid Law. European State Aid Institute. Berlin: of subsidies and state aids in WTO and EC Law: Conflicts in Lexxion. international trade. The Hague: Kluwer. Coppens, Dominic. 2015. WTO disciplines on subsidies and Mavroidis, Petros C, Patrick Messerlin, and Jasper M. Wouters. countervailing measures—balancing policy space and legal constraints. 2007. The Law and economics of contingent protection in the WTO. Cambridge: Cambridge University Press. Cheltenham, UK: Edward Elgar. De Cecco, Francesco. 2012. State aid and the European Economic OECD (Organisation for Economic Co-operation and Constitution. Oxford: Hart Publishing. Development). 2001. Competition policy in subsidies and state aid. Competition Policy Roundtable. DAFFE/CLP (2001)24. Derenne, Jacques, and Massimo Merola. 2007. Economic analysis Paris: OECD. of state aid rules—contributions and limits. Berlin: Lexxion. ———. 2010. Competition, state aid and subsidies. AF/COMP/ Dony, Marianne. 2007. Contrôle des aides d’etat. Brussels: Les GF(2010)5. Roundtable on Competition, State Aid and Editions de l’Universite de Bruxelles. Subsidies. Paris: OECD. Dür, Andreas, Leonardo Baccini, and Manfred Elsig. 2014.The Piernaz-Lopez, Juan Jorge. 2015. The concept of state aid under design of international trade agreements: Introducing a new EU law: From internal market to competition and beyond. Oxford: dataset. Review of International Organizations 9: 353-73. Oxford University Press. 460 Subsidies Quigley, Conor. 2015. European state aid law and policy. Oxford: Hart Publishing. Rodrik, Dani. 2004. Industrial Policy for the 21st Century. Faculty Research Working Papers. JFK School of Government, Harvard University (September). Rubini, Luca. 2009. The definition of subsidy and state aid:WTO law and EC law in comparative perspective. Oxford: Oxford University Press. Rubini, Luca, and Jennifer K. Hawkins. 2016. What shapes the law? Reflections on the history, law, politics and economics of international and European subsidy disciplines. Florence: European University Institute. Shaffer, Gregory, Robert Wolfe, and Vinhcent Le. 2015. Can informal law discipline subsidies? Journal of International Economic Law (18):711-41. Spector, David. 2009. State aid: Economic analysis and practice in the European Union. In Xavier Vives (ed.), Competition policy in the EU: Fifty years on from the Treaty of Rome. Oxford: Oxford University Press. Sykes, Alan O. 2010. The questionable case for subsidies regulation: A comparative perspective. Journal of Legal Analysis (2):473-523. Wallace, Don, Frank J. Loftus, and Van Z. Krikorian. 1984. Interface three: Legal treatment of domestic subsidies. Washington, DC: International Law Institute. 461 Handbook of Deep Trade Agreements 462 State-Owned Enterprises CHAPTER 16 State-Owned Enterprises L. Rubini and T. Wang 463 Handbook of Deep Trade Agreements CHAPTER 16 State-Owned Enterprises L. Rubini* and T. Wang† * University of Birmingham Law School, Birmingham, United Kingdom † University of Bristol, Bristol, United Kingdom CONTENTS 16.1. INTRODUCTION 465 16.2. LITERATURE AND PREVIOUS DATASETS 465 16.3. NEW DATASET 466 16.3.1 Overview of the template 467 16.3.1.1 Categories and questions (rows) 467 16.3.1.2 Further information on PTAs depth (columns) 468 16.3.1.2.1 WTO coverage 468 16.3.1.2.2 (Legal) enforceability 469 16.3.1.2.3 Benefits to non-members 470 16.3.2 Selected coding issues 471 16.3.2.1 Objectives of PTAs regulating state enterprises 471 16.3.2.2 Exceptions 472 16.4. ANALYSIS: DESCRIPTIVE STATISTICS 474 16.4.1 General overview of the evolution and distribution of provisions on state enterprises 474 16.4.1.1 Evolution of the number of PTAs with provisions on state enterprises 474 16.4.1.2 Distribution and evolution of PTAs with state enterprise provisions by level of development of their members 477 16.4.1.3 Distribution and evolution of PTAs with state enterprise provisions by geographic group 478 16.4.2 Analysis of scope 480 16.4.2.1 Most common provisions 480 16.4.2.2 Most common provisions by level of development 482 16.4.2.3 Most common provisions by geographic group 483 16.4.2.4 Notable countries 485 16.4.3 Specific analysis of types of provisions 486 16.4.3.1 Definitions 486 16.4.3.2 Coverage (entities, sectors, ownership regimes) 487 16.4.3.2.1 Entities 488 16.4.3.2.2 Sectors 490 16.4.3.3 Disciplines 491 16.4.3.4 Exceptions 494 16.4.3.5 Transparency and corporate governance 496 16.4.3.6 Enforcement 499 16.5. CONCLUSIONS 501 ACKNOWLEDGMENTS 502 REFERENCES 503 464 State-Owned Enterprises 16.1. INTRODUCTION Across the ages and geographically, states have always intervened in the economy through various forms of state enterprises. Almost since the beginning, these enterprises have been regulated by trade rules and agreements. Notable examples are the General Agreement on Tariffs and Trade (GATT) of 1947 and the Treaty of Rome of 1957 setting up the European Economic Community. The authors have mapped the provisions on state enterprises in 283 preferential trade agreements (PTAs) signed between 1957 and early February 2016 and included in the World Bank’s Deep Integration database.1 This mapping, which uses as a benchmark the relevant disciplines in the WTO, is useful to assess the vertical depth of these provisions, which is the ultimate purpose of the World Bank’s Deep Integration research agenda.2 This chapter outlines the methodology underlying the creation of a new dataset on provisions regarding state enterprises, and includes an initial description of the main patterns that emerge.The final section of the chapter offers some tentative conclusions on the vertical depth of the state enterprise provisions in the examined PTAs. 16.2. LITERATURE AND PREVIOUS DATASETS Though states have always intervened in their economies through various forms of state entities, and both the GATT and EU have always had rules on state enterprises, the topic did not attract much attention in the literature until the 1990s, when the processes of liberalization and privatization in the 1980s started to produce their effects. Thus, for example, while the GATT rules on state enterprises received distinct treatment as early as 1969,3 it was not until the late 1990s that studies began to focus specifically on the regulation of state enterprises, mostly in the treaties that created the WTO and the EU.4 Despite these works, “issues surrounding the operation of state-owned enterprises (SOEs) in the international trading system [remained] an understudied area and yet one of increasing importance, particularly given the size and significance of Chinese state-owned enterprises (SOE).”5 It is only in the 1 The dataset actually includes 282 PTAs. Normally, amendments are not counted separately, but treaties for the accession of new parties are. For counting purposes, given the complexity of the successive iterations of the regulatory framework, there are 3 different sheets for (9) CARICOM; i.e., (9.1) CARIFTA, (9.2) CARICOM, and (9.3) CARICOM + CSME. For the sake of simplicity and consistency with the datasets of the other authors, only (9.2) CARICOM has been considered for the analysis of the stylized facts. Always for consistency reasons, (216) East African Community (EAC)-Accession of Burundi and Rwanda has been counted as two separate agreements (EAC-Accession of Burundi and EAC-Accession of Rwanda), thus raising the count to 283. 2 The research agenda aims at complementing the existent dataset on the ‘horizontal’ depth of PTAs created by the World Bank. See Hofmann, Osnago, and Ruta 2017. 3 Jackson 1969 was an early seminal study. 4 See, for example, Cottier and Mavroidis 1998, Blum and Logue 1998, and Buendia-Sierra 2000. 5 Marvroidis and Janow 2017. 465 Handbook of Deep Trade Agreements past few years that scholars have started to focus their attention on the ability of WTO rules to regulate state trading and state capitalism.6 Interesting research on state enterprises is being carried out by international institutions. The Organisation for Economic Co-operation and Development (OECD) and the World Bank have, for example, carried out work with a focus on SOEs from a competition and corporate governance perspective.7 For its part, the European Union Commission has recently completed a review of the situation of state-owned enterprises in the EU.8 There are also works that focus on specific issues or angles related to state enterprises. For example, the literature on subsidies regularly deals with issues related to the fact that the entity providing assistance is a state enterprise (e.g., imputability of conduct, benchmarking, and privatization).9 The analysis of the conduct of state entities in the market is also tackled in the competition or trade/internal market literature.10 Equally relevant is the literature on the regulation of liberalization and privatization processes11 or on the regulation of public services at both global and European levels.12 Against this background, it should be noted that the only mapping exercise that is comparable to the World Bank’s Deep Integration agenda is the Design of Trade Agreements (DESTA).13 It is, however, difficult to identify any works that map the content of state enterprise provisions in PTAs in a comprehensive and detailed manner and expressly look at the depth of the provisions.14 For example, DESTA has only one question on state enterprises in the competition chapter (“Is there a provision on state trading enterprises?”). 16.3. NEW DATASET The dataset of the Deep Integration project represents a new level of comprehensiveness. It examines a sample of 283 PTAs that include provisions on state enterprises. This section of the chapter outlines the methodology used in the mapping exercise by offering an overview of the template and a specific examination of a few selected coding issues. 6 See, for example, Wu 2016; Mavroidis and Janow 2017; Zhou et al. 2018; and Bown 2018. 7 See World Bank 2014; and OECD 2009, 2010, 2011, 2015. 8 See European Commission 2016. 9 See, for example, Rubini 2009 and Mavroidis et al. 2008. 10 See, for example, Fox and Crane 2010 and Oliver 2010. 11 Szyszczak 2007. 12 See, for example, Krajewski 2003, 2015. 13 See Dür, Baccini, and Elsig 2014. 14 There are recent works analyzing state enterprise provisions in PTAs, but they only consult specific PTAs, such as the Trans-Pacific Partnership. See Willemyns 2016 and Fleury and Marcoux 2016. 466 State-Owned Enterprises 16.3.1 Overview of the template It is useful to distinguish the information included in the template according to its two main dimensions: rows and columns. 16.3.1.1 Categories and questions (rows) The template is divided into six different categories, which reflect the key types of provisions regulating state enterprises in PTAs. These are: • Objectives and coverage of disciplines • Substantive disciplines • Transparency and corporate governance • Enforcement • Special and differential treatment (SDT) • Miscellaneous Each category is divided into various questions. The template begins by asking whether the PTA specifically spells out the objective of regulating state intervention in the economy via state-controlled or state-delegated entities, and whether the PTA makes reference to GATT/ WTO rules on state enterprises. It then maps the existence of specific definitions of a variety of public entities that may intervene in the market (state-owned enterprises, state enterprises or public undertakings, state trading enterprises, public or government monopolies, designated monopolies, entities with special or exclusive rights, sovereign wealth funds). The template also maps definitions of key concepts such as government control or influence, commercial activity, commercial considerations, and non-commercial assistance. The template then focuses on the coverage of the PTA, with questions on what entities or regulatory elements are present in the agreement and immediately signaling the most notable exclusions. There are questions on the presence of de minimis thresholds, on the coverage of specific activities and sectors, and on rules that impact the ownership of state enterprises. Questions on substantive disciplines focus on the provisions that prohibit or mandate certain types of behavior (e.g., prohibitions of discrimination, anti-competitive behavior, distortions of trade, subsidization; and obligations to act in accordance with commercial considerations, afford companies adequate opportunities to compete, accord fair and equitable treatment, and ensure the compliance of delegated entities with PTA obligations). There are questions capturing any requirement for proof of negative effects on the market, of any reference to the specific geographical market where the conduct or the effects need to take place, and exceptions specific to state enterprises. 467 Handbook of Deep Trade Agreements Finally, there are questions that capture various levels of transparency, deliberation, and assessment, and the existence of any corporate governance requirement. The section on enforcement aims to capture those mechanisms that, beyond transparency, are specifically devoted to ensuring that the rules are respected and possible breaches of the rules remedied. Questions on the presence of elements of special and differential treatment (SDT), and on the review of the PTA and further negotiation of rules in the area, close the template. Specific issues arising out of the coding of these questions are discussed below. 16.3.1.2 Further information on PTAs depth (columns) The questions in the row are intersected with columns that convey further information about the depth of the PTA. There are columns on WTO coverage, enforceability, benefit to non- members, and sectoral coverage and exclusions, as well as a column for providing comments. The most relevant for the purposes of this study are WTO coverage, enforceability, and benefits to non-members. 16.3.1.2.1 WTO coverage This column indicates the relationship between the coverage of the disciplines for state enterprises in the PTA and the corresponding regulation in the WTO. Essentially, it answers the question of whether the PTA adds to the WTO disciplines on state enterprises. State enterprises are regulated in the WTO. PTA disciplines on state enterprises can thus be “WTO =” (if the rules essentially restate or provide a level of regulation similar to WTO commitments); “WTO +” (if the rules exceed WTO disciplines or commitments), or “WTO –“ (if the commitments are more limited than WTO requirements). For example, the question on Competition Law may be coded “WTO +” for goods and “WTO =” for services (because of the presence of GATS Article IX). There are two important caveats when comparing the coverage of individual PTAs and corresponding WTO disciplines. First, WTO disciplines on state enterprises are quite varied according to the sector, the type of entity, and the type of prohibition or obligation. It may well be that there is not an exact correspondence between PTA and WTO coverage, with the result that the coding inevitably has to be approximate. For example, one PTA may expressly prohibit discriminatory behavior by monopoly suppliers of services without regulating similar conduct in trade in goods (though, via the GATT, the WTO does also prohibit discrimination by state trading enterprises). This type of provision is coded “WTO =.” Second, while the assessment of the PTA disciplines could be reasonably specific (by considering the relevant schedules and lists of reservations and coding some of the most interesting examples), the assessment in this study of the corresponding WTO discipline 468 State-Owned Enterprises has been carried out at a more preliminary level. The coding of the WTO coverage simply indicates whether there are disciplines in the PTA that can also be found in principle in the WTO. If, for example, the PTA regulates state enterprises in the service sector, the WTO coverage has been coded “WTO =” because state enterprises are in principle also regulated in the GATS. In other words, the coders have considered the coverage of WTO law in general, without examining the specific GATS schedules of the relevant parties. This approach necessarily leads to an approximation, but a specific perusal of the GATS schedules would have created too much complexity, especially when the PTA is multi-party. This is beyond the current mapping exercise. In some cases, the coders have specifically relied on the ultimate goal of the coding exercise; i.e., to determine the depth of integration pursued by the PTA. The assessment of WTO coverage has thus focused on the depth of the commitment of the PTA relative to the WTO rule-book.The goal is not to assess whether certain PTA provisions are WTO law compliant but whether they pursue a level of integration that is deeper (or otherwise) than the WTO. A few practical examples are discussed below in the section on specific coding issues and, in particular, in the exceptions section. 16.3.1.2.2 (Legal) enforceability In general, the question of enforceability does not apply to definitions or coverage (except for questions on ownership regimes of state entities and liberalization processes), but only to substantive disciplines, transparency, and special and differential treatment provisions. The coders have used the scale established at the beginning of the Deep Integration project: - 0: non-binding - 1: non-binding provision with best efforts - 2: binding provision with no dispute settlement (DS) - 3: binding provision with state-to-state dispute settlement - 4: binding provision with only private-state dispute settlement (ISDS) - 5: binding provision with both state-to-state and private-state dispute settlement. The most common codes have been “3” when dispute settlement mechanisms are available and “2” when they are not. For example, the coding of the Competition Law question is normally “2” because very often PTAs exclude competition provisions from dispute settlement. If, by contrast, dispute settlement provisions are applicable to the competition chapter at large or, in any event, to the specific competition provisions on state enterprises and monopolies, the coding is “3.” The use of “4” is limited to those rare cases where there is a national treatment obligation in investment chapters for which investor-state dispute settlement (ISDS) is provided. 469 Handbook of Deep Trade Agreements 16.3.1.2.3 Benefits to non-members This column focuses on the impact of providing de facto benefits or positive externalities to non-members of the PTA. There are two possible alternatives: “1” for a positive answer or “0” for a negative answer. If the issue is not relevant (because it is not possible to determine the existence of an externality in the abstract), the cell has not been coded. In particular, the questions under objectives, definitions and coverage, transparency and enforcement, and special and differential treatment provisions have not been coded because either the question of any benefit to non-members is not relevant or the existence of any such benefit cannot be easily determined through the abstract coding of legal texts. For similar reasons, the questions on the possible proof of negative effects or on the requirement to identify the geographical market affected were not coded. The question on whether the PTA requires the parties to ensure that state enterprises comply with the parties’ commitments and obligations has not been coded either. The other questions on disciplines were coded “1” if it was clear that the obligation/prohibition was not limited to the parties. For example, when the Competition Law question has been coded, a positive externality for non-members has been assumed, thus leading to coding “1.” Do PTAs lead to most-favored-nation (MFN) discrimination against non-signatories? The answer depends on whether, for example, the prohibition is a general one (for example, anti- competitive behavior), or rather it only concerns inter-party trade. The questions “Does the agreement require state enterprises to act in accordance with commercial considerations?,” “Does the agreement require affording companies adequate opportunities to compete?,” and “Does the agreement provide for an obligation to accord fair and equitable treatment?” trace their origins to GATT Article XVII and have accordingly all been coded “WTO =.” However, in terms of possible externalities to non-members, while the first two have been coded “1,” the last one has been coded “0.” In terms of WTO coverage, subsidy rules are considered to offer a benefit to non-members, thus leading to a “1” coding. If they are, however, expressly limited to the parties (for example, prohibition of export subsidies on goods destined to the territory of the other party), this is not considered as conferring any distinct benefit to non-members, and hence are coded “0.” The question “Does the agreement prohibit discrimination by state enterprises?” refers to any non-discrimination provision, mostly national treatment or MFN based. If the PTA has any such provision, the comments section of the template is used to highlight what type of discrimination is prohibited, and whether the prohibition applies to goods and/or services. If the prohibition of discrimination refers only to the other party or to reciprocal trade, this is coded “0” because there can be no benefit to non-members. 470 State-Owned Enterprises 16.3.2 Selected coding issues After offering a general overview of the template’s structure and a few examples of how coding has been carried out, the study now specifically examines how coding has contributed to the overall goal of determining the objectives of PTA provisions that regulate state enterprises, and assessing the depth of those provisions. 16.3.2.1 Objectives of PTAs regulating state enterprises The coders have looked for specific expressions of the objectives to regulate state intervention in the market through public entities. If present, these objectives are normally broadly phrased and may not have a huge practical relevance. It does not then come as a surprise that there are very few examples of these objectives, particularly in the opening provisions of Competition Law chapters which then include rules on state enterprises. One example is (160) US-Peru. The competition chapter of that PTA includes specific provisions on designated monopolies and state enterprises, and immediately sets out the objectives of those provisions:“Recognizing that the conduct subject to this Chapter has the potential to restrict bilateral trade and investment, the Parties believe that proscribing such conduct, implementing economically sound competition policies, and cooperating on matters covered by this Chapter will help secure the benefits of this Agreement” (Article 13.1). In conclusion, it is not normally possible to obtain any meaningful indication of the depth of the PTA by considering the explicit objectives the parties set for themselves.This goal can only be achieved through interpreting the content of the PTAs. In particular, the analysis of certain provisions may offer useful indications about the goals underlying the regulation of state enterprises. One good example comes from the coding of the question on ownership regimes, which may indeed indicate different objectives of the PTA disciplines on state enterprises. The question “Does the agreement expressly regulate ownership or property regimes, or liberalization processes?” refers to those provisions that may have an impact on government decisions about how to organize certain companies and sectors of the economy. Among the most common provisions are those that declare the neutrality of the parties with respect to the decision to set up or maintain public monopolies or state enterprises. An extremely common provision is, for example: “Nothing in this Chapter shall be construed to prevent a Party from establishing or maintaining state enterprises and/or designated monopolies” (276) Republic of Korea-Colombia (Article 13.9).The same PTA includes another provision with a similar content in the investment chapter: “For greater certainty, nothing in this Chapter shall be construed to impose an obligation on a Party to privatize any investment that it 471 Handbook of Deep Trade Agreements owns or controls or to prevent a Party from designating a monopoly” (Chapter 8, footnote 1). Another example of the “principle of neutrality” is (265) Mauritius-Pakistan, which reads at Article 16.1: “Nothing in this Agreement shall prevent a Contracting Party maintaining or establishing a state trading enterprise as provided for in Article XVII of GATT 1994.” There are also other provisions that have been coded under this question but that embody a different principle. One good example is (110) Korea-Singapore (Article 15.4): “Each Party shall take reasonable measures to ensure that its government does not provide any competitive advantage to any government-owned businesses in their business activities simply because they are government owned.” Along the same theme, many PTAs include provisions that call on the parties to progressively adjust their state monopolies of commercial character to ensure that no discrimination regarding the conditions under which goods are procured and marketed will exist between the parties. These provisions are particularly common in the PTAs signed by the EU and mirror the corresponding provision in the (1) EC Treaty (Article 37). See, for example, (30) EU-Turkey (Article 42). They can also be found in the PTAs of those countries closely linked to the EU. See, for example, (26) EFTA-Israel (Article 9.1) or (105) Turkey-Tunisia (Article 23). Importantly, these provisions do not embody the “principle of neutrality” (vis-à-vis state enterprises) but rather a concept of “level playing field” (in favor of private operators). They are not intended to protect state enterprises and, through them, the state prerogative in organizing the economy; rather, they aim to protect private operators from the possible advantages state enterprises may enjoy. Expanding a bit on this discussion of the possible objectives of PTA rules on state enterprises, and considering other provisions, it seems that the “level playing field” idea underlies most of the disciplines that have been coded (clearly competition Law and state enterprise provisions, but also rules and obligations to act in accordance with commercial considerations, and to accord fair and equitable treatment or adequate opportunities to compete). Different rationales support the various provisions that prohibit non-discrimination (national treatment, MFN, general prohibitions on discrimination). By contrast, the obligation to ensure that state enterprises comply with the commitments and obligations entered into by the PTA parties is general and open ended and its rationale fundamentally depends on the specific commitments or obligations. 16.3.2.2 Exceptions There are various questions that may lead to coding exceptions. Two questions are in the “coverage” part of the template (“Does the agreement expressly regulate/exclude state enterprises pursuing public services?” and “Does the agreement expressly regulate/exclude state enterprises in strategic sectors?”). Two more questions relate to “substantive disciplines” (“Does the agreement provide for exceptions specific to state enterprises?” and “Does the agreement include any other specific discipline for certain sectors or objectives?”). 472 State-Owned Enterprises Provisions that may be considered exceptions to the normal disciplines on state enterprises are common in PTAs. Not only was it important for the coders to determine where (i.e., under what questions) these provisions had to be coded, but, most crucially, how their relation to WTO disciplines should be assessed. This is because the coding, especially of the WTO coverage, is particularly important to determine vertical depth.Are these “WTO –” or “WTO +” provisions? The main rule followed is that exceptions for horizontal or general objectives, such as the performance of public services or the pursuit of public policy goals (e.g., environmental protection), have been coded “WTO +.” The thinking is that, while horizontal exceptions may depart from WTO commitments, their presence is typical of more integrated systems and implies that the parties accept the fact that certain economic or competitive benefits they would otherwise obtain from the PTA can (or indeed should) be limited by the pursuit of other, often non-economic, goals.This is considered a sign of more mature, deeper experiments with integration. The horizontal nature of these goals is a proxy for their non-protectionism. By contrast, sectoral carve-outs (exclusions or limitations for certain industries or products) for strategic sectors, and without any clear reference to a general public policy goal, are more likely to be protectionist and can be coded “WTO –.” See, for example, (217) Colombia- Northern Triangle (Annex III; Schedules), with various limitations on oil products and regional television; (239) Costa Rica-Peru (Schedules, Costa Rica), with Costa Rica’s express reservation on oil import and wholesale distribution as a State monopoly; or (277) Costa Rica-Colombia (Schedules), with various restrictions on oil products and national roads, docks, and airports (Costa Rica) and on regional television and financial services (Colombia).15 Similarly, there are a few PTAs that (except for sectors that can be defined as strategic) include limitations and reservations for state monopolies and for undertakings with special or exclusive rights in the schedules. These have been coded under the question on special rules for certain sectors. However, in the absence of specific information that could indicate the actual existence of horizontal public policy tasks (or otherwise), the “WTO coverage” has normally been left blank. See, for example, (195) EU-Korea (Schedules). One interesting example can be found in (48) EFTA-West Bank and Gaza (Article 9, Protocol C), in which the general obligation to adjust monopolies of commercial character shall apply to Liechtenstein and Switzerland with regard to State monopolies concerning salt, but only to the extent that these parties will have to fulfill corresponding obligations under their trade relations with the European Community and EFTA States. This reservation is common in many EFTA PTAs. In the absence of further information of the type indicated above, the “WTO coverage” question has been left blank. 15 To be sure, special treatment for certain sectors like regional television may be justified by horizontal or general goals like the protection of culture. 473 Handbook of Deep Trade Agreements Article 106(2) of the (1) EC Treaty is the model of provisions that can be frequently found in other PTAs. It reads: Undertakings entrusted with the operation of services of general economic interest or having the character of a revenue-producing monopoly shall be subject to the rules contained in the Treaties, in particular to the rules on competition, in so far as the application of such rules does not obstruct the performance, in law or in fact, of the particular tasks assigned to them. When present (and this is common in PTAs signed by the EU), this provision has been coded under two different questions and with different “WTO coverage” codings. The exception in favor of undertakings entrusted with the operation of services of general economic interest (read: public services) has been coded under the question on public services, with a “WTO +” coding. By contrast, the exception for revenue-producing monopolies has been coded under the general question on the existence of specific exceptions for state enterprises, and, given its generality, has not been coded under “WTO coverage.” (184) EU-Serbia is a good example on this point. Quite similarly in other cases, when the language of the provision was general, it was not possible to code either as “WTO –” or “WTO +” and, in the absence of more information, the relevant cell was left blank. It is, for example, very common to find provisions whereby the state enterprise is subject to certain obligations, such as the obligation to act in accordance with commercial considerations, or disciplines, such as those specified in competition law, but only in so far as those rules do not obstruct the performance of specific tasks assigned to it. These provisions were coded under the question on specific exceptions for state enterprises, but their WTO coverage has been left blank. Among several examples, see (270) China- Korea (Article 14.5.3(b), exception from competition laws); and (260) Canada-Korea (Article 15.2.3(b), exception from obligation to act in accordance with commercial considerations). 16.4. ANALYSIS: DESCRIPTIVE STATISTICS 16.4.1 General overview of the evolution and distribution of provisions on state enterprises This section begins with a few basic descriptive statistics on the evolution and distribution of state enterprise provisions in the 283 PTAs in the sample. 16.4.1.1 Evolution of the number of PTAs with provisions on state enterprises Figure 16.1 shows the evolution of the number of PTAs with state enterprises provisions vis- à-vis those without these provisions. Given their relevance the figure also shows those limited PTAs that regulate state enterprises via general competition laws only. While state enterprise provisions were fairly rare before 1991, featuring in only 11 out of 21 PTAs, their frequency steadily increased in the following decades, and the share of PTAs with such provisions has been 474 State-Owned Enterprises clearly and robustly dominant since 2012. In the 2012-2016 period, a total of 30 out of 33 PTAs signed included state enterprise provisions. Competition law appears as an important tool to regulate state enterprises in the 1992-2003 period, with 23 out of 102 PTAs signed including competition law provisions but no dedicated state enterprise provision.The most successful years for competition law as a tool to regulate state enterprises were 1992 to 1996, with an 18/24 ratio between competition provisions and specific state enterprise provisions. Cumulatively, in the full period covered by the template (1957-2016),16 218 PTAs have been concluded that regulate state enterprises, either through express state enterprise provisions (193) or through competition provisions (25). Only 65 PTAs have neither state enterprise nor competition provisions. Figure 16.1: Evolution of the number of PTAs with state enterprise provisions 20 300 18 16 250 Number of agreements 14 200 12 10 150 8 6 100 4 50 2 0 0 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 Year of Signature Without STE provisions With STE provisions Only Competition Law provisions Cumulative Agreements Source: Deep Trade Agreements Database. The steady increase of state enterprise provisions over time is confirmed by Figure 16.2, which shows the evolution of the average number of these provisions over time (the figure only includes those PTAs with state enterprise provisions). Figure 16.2: Number of state enterprise provisions (average) over time Source: Deep Trade Agreements Database. 16 All figures and tables refer to the 2010-2017 period, mainly for the sake of consistency with the other chapters. However, the most recent PTA coded in the context of the mapping exercise underlying this paper dates back to February 2016. 475 Handbook of Deep Trade Agreements Figures 16.3a and 16.3b introduce a distinction in state enterprise provisions among “disciplines” (provisions that tell parties what they can or cannot do with regard to state enterprises),17 “transparency” provisions, and “enforcement” provisions. Figure 16.3a: Evolution of the number of PTAs with state enterprise provisions Source: Deep Trade Agreements Database. Figure 16.3b: Evolution of the share of PTAs with state enterprise provisions Source: Deep Trade Agreements Database. Figure 16.3a shows that disciplines, transparency, and enforcement provisions all increase in the periods considered, reaching the apex in the 2005-2009 period, then decreasing in the 2010-2017 period. This decrease is mainly due to the lower number of PTAs signed and does not necessarily detract from the upward trend that has been noted. The sharpest increases occur for transparency provisions between the pre-WTO and 1995-1999 periods, and for disciplines and enforcement provisions between the 2000-2004 and 2005-2009 17 Disciplines in particular refer to the questions in rows 37-48 of the template (section on substantive disciplines). 476 State-Owned Enterprises periods. What is probably more significant, though, is the evolution of the share of PTAs with state enterprise provisions. The main indication coming from Figure 16.3b is one of stability. Except for the pre-1995 period, when they were present in only 8 percent of PTAs, transparency provisions are consistently present in the next periods with only small variations (46 percent in 1995-1999, 53 percent in 2000-2004, 46 percent in 2005-2009, and 49 percent in 2010-2017). A similar pattern can be noticed for enforcement provisions.They move from 58 percent (pre-1995) to 78 percent (1995-1999) and then fluctuate mildly (back to 66 percent in 2000-2004, to 74 percent in 2005-2009, and 77 percent in 2010-2017).The provisions on disciplines show a similar pattern, the only difference being that they were already strong in the pre-WTO period (71 percent), before peaking at 88 percent in the 1995-1999 period, decreasing to 72 percent in 2000-2004, and then progressively increasing to 75 percent in 2005-2009 and to 81 percent in 2010-2017. It goes without saying that these are very much macro categories and can only give very broad indications. A more detailed analysis of the trends within each category is presented below. 16.4.1.2 Distribution and evolution of PTAs with state enterprise provisions by level of development of their members This section considers the distribution and evolution of PTAs with state enterprise provisions by level of development of the country, as defined following the World Bank country classification of 2017.18 Figure 16.4 shows the distribution and the share of PTAs with state Figure 16.4: Share of PTAs with state enterprise provisions by level of development Source: Deep Trade Agreements Database. Note: The figures listed at the bottom (left) of each bar represent the number of North-North, North-South, and South- South PTAs with state enterprise provisions. 18 Developing countries are composed of low-income and lower-middle-income economies, whereas upper-middle-income and high-income countries are considered Developed. Low-income economies are defined as those with a gross national income (GNI) per capita, calculated using the World Bank Atlas method, of US$1,005 or less in 2016; lower-middle-income economies are those with a GNI per capita between US$1,006 and US$3,955; upper-middle-income economies are those with a GNI per capita between US$3,956 and US$12,235; high-income economies are those with a GNI per capita of US$12,236 or more. 477 Handbook of Deep Trade Agreements enterprise provisions by level of development. While state enterprise provisions are present in the large majority (90 percent) of trade agreements signed between developed (North) countries, they are less frequent although still common in those signed between developing (South) countries or between North and South countries (76 and 74 percent, respectively). Figure 16.5 shows an increase in state enterprise provisions in the PTAs between North countries from a 71 percent share before 1995 to 100 percent in all subsequent periods except 2005-2009, when it dropped to 83 percent.The share of state enterprise provisions in North-South PTAs has fluctuated (averaging 74 percent and with peaks at 85 and 88 percent, respectively, in 1995-1999 and 2010-2017). The data are more revealing for the PTAs between South countries, where the share has decreased consistently since the early 2000s. In particular, after an increase from 78 percent (before 1995) to 89 percent (between 1995 and 1999), there were clear decreases in almost every subsequent period: 76 percent (-14.6 percent compared to the previous period) in 2000-2004 and 2005-2009, down to 53 percent (-30.3 percent) in 2010-2017. Figure 16.5: Share of PTAs with state enterprise provisions by level of development, by year of signature Source: Deep Trade Agreements Database. 16.4.1.3 Distribution and evolution of PTAs with state enterprise provisions by geographic group We now consider the distribution and evolution of PTAs with state enterprise provisions by geographic group, defined following the World Bank classification.19 Figure 16.6 shows that almost all PTAs signed by EFTA (97 percent) include state enterprise provisions. A very high presence of these provisions (from 72 to 85 percent) can be found in the large majority of the geographic groups (from the highest to the lowest: Central Asia, North America, Latin America & Caribbean, Middle East & North Africa, the EU, and East Asia & Pacific). South Asia PTAs still score relatively high (64 percent). Sub-Saharan Africa lags behind (44 percent). 19 See https://blogs.worldbank.org/opendata/new-country-classifications-income-level-2018-2019. 478 State-Owned Enterprises Figure 16.6: Share of PTAs with state enterprise provisions by geographic group Source: Deep Trade Agreements Database. Note: The figures listed at the bottom of each bar represent the number of PTAs with state enterprise provisions. Looking now at these data together with those of Table 16.1, what emerges is a mixed picture. EFTA PTAs have consistently included state enterprise provisions, scoring 100 percent in all periods except for the period before 1995, when the share was 89 percent because of one agreement (with the Faroe Islands) that did not include state enterprise provisions. With the exception of the last period, Central Asia also has a high share of state enterprise provisions. East Asia & Pacific and South Asia both show a moderate increase in state enterprise provisions, reaching, respectively, 86 and 100 percent in the 2010-2017 period. North America features a strong increase from the pre-WTO to the post-WTO period, doubling the share of PTAs with state enterprise provisions (from 50 percent to 100 percent), then decreasing to almost the previous level (57 percent) in 2000-2004 and then increasing robustly in the next periods to reach 100 percent in 2010-2017. Latin America & Caribbean also shows a robust increase from the pre-WTO to the post-WTO period (from 56 to 89 percent), but then progressively decreases after that (to 85 percent in 2000-2004 and 78 percent in 2005-2009 and 2010-2017). The EU always scores well above 80 percent except for the pre-WTO period (71 percent) and the 2005-to-2009 period (64 percent). The share of PTAs with no state enterprise provisions during each period depends mainly on the special nature of the relevant PTAs (enlargement agreements, agreements with significantly smaller countries).20 South Asia shows a consistent increase, which is particularly noticeable in the latest three periods (after 100 percent in the post-WTO period, the shares were, respectively 60 percent, 70 percent, and 100 percent). The share of PTAs with state enterprise provisions for Sub-Saharan Africa is largely consistent, never scoring above 50 percent in any period. 20 (89E) U(25) Enlargement, (117) EU(27) Enlargement, (235) EU (28) Enlargement, (4) EU-OCT, (178) EU- San Marino, (38) EU-Andorra, (202) EU-Papua New Guinea/Fiji. See also (52) EC-Mexico, (12) EC-Syrian Arab Republic, (157) EU-Côte d”Ivoire, (207) EU-Eastern and Southern Africa States Interim EPA. 479 Handbook of Deep Trade Agreements Table 16.1: Share of PTAs with state enterprise provisions by geographic group, over time Year of signature Before 1995 1995 to 1999 2000 to 2004 2005 to 2009 2010 to 2017 Central Asia 80 94 92 93 62 EFTA 89 100 100 100 100 East Asia & Pacific 38 0 56 76 86 European Union 71 89 83 64 86 Latin America & Caribbean 56 89 85 78 78 Middle East & North Africa 60 90 73 83 100 North America 50 100 57 90 100 South Asia 25 100 60 70 100 Sub-Saharan Africa 50 40 50 43 50 Source: Deep Trade Agreements Database. 16.4.2 Analysis of scope After analyzing the general traits of state enterprise provisions in PTAs, this section focuses more specifically on the scope of these agreements. 16.4.2.1 Most common provisions The most common state enterprise provisions are shown in Figure 16.7. Some of these provisions will be subject to deeper analysis in Section 16.4.3. It may be useful to recall that, among the 283 PTAs coded, 218 include state enterprise provisions, divided into provisions specific to state enterprises (193) and general competition provisions (25). Figure 16.7 does not consider the questions on the types of covered entities (which are analyzed in detail below) or those provisions that scored less than 10.21 All provisions on definitions have been counted together.22 21 In particular, “Does the agreement expressly regulate/exclude sovereign wealth funds (SWFs)?” (9); “Does the agreement require domestic administrative bodies to act impartially?” (4); “Does the agreement require proof of negative effect on the market?” (3), “Does the agreement provide for any other special and differential treatment with respect to state enterprises?” (2); “Does the agreement provide for cooperation or technical assistance specific to state enterprises?” (2); “Does the agreement require domestic courts to have jurisdiction on covered entities?” (2); “Does the agreement include a de minimis threshold (i.e. state enterprises under it are not covered)?” (2); “Does the agreement indicate the geographical market where the objectionable conduct or the effect takes place?” (2); “Does the agreement provide for the supervision or review of the rules on state enterprises?” (2); “Does the agreement provide for further negotiations in the area of state intervention in the economy?” (2); “Does the agreement provide for any other requirement or mechanism to deal with transparency or corporate governance with respect to state intervention in the economy?” (1)”; “Does the agreement provide for any other enforcement mechanism?” (0). 22 The dataset has separate questions for the definitions of (a) state-owned enterprise (SOE), state enterprise (SE), or public undertaking; (b) state trading enterprise (STE); (c) public or government monopoly; (d) designated monopoly; (e) entity with special or exclusive privileges and rights; (f) government control or influence; (g) sovereign wealth fund (SWF); (h) commercial activity; (i) commercial considerations; and (j) non-commercial assistance. 480 State-Owned Enterprises Figure 16.7: Most common provisions Source: Deep Trade Agreements Database. The first information that can be derived from the figure is that the very large majority of PTAs with state enterprise provisions also regulate those in the agriculture sector (195, or 89 percent). The coverage of state enterprises operating in the service sector is significantly lower (56 percent). Two of the most common provisions do not deal with disciplines or commitments but are much more general (existence of institutions dealing with transparency and enforcement, 183; dispute settlement mechanism, 143). The most common provisions that include commitments are those that (a) prohibit the anti-competitive behavior of state enterprises (190, or 87 percent of PTAs); and (b) regulate subsidization of state enterprises (156, or 71.5 percent); and (c) prohibit discrimination by state enterprises (140, or 64 percent). Provisions that regulate ownership or property regimes or liberalization processes appear in 109 PTAs (exactly 50 percent of those with state enterprise provisions). Provisions that require parties to ensure that state enterprises or monopolies do not act inconsistently with the parties’ obligations or commitments appear in 83 of the PTAs (38 percent), and notification requirements23 occur in 78 (36 percent). There are 73 PTAs (33 percent) with provisions that provide for specific exceptions for state enterprises, and 64 PTAs (29 percent) that commit state enterprises to carry out any commerical activities in accordance with commercial considerations. 23 Notification requirements (for example, on granting a monopoly exclusive or special rights, the operations or conduct of state enterprises, and the goods/services covered) were the first type of transparency provisions included in PTAs. 481 Handbook of Deep Trade Agreements Much less frequent are PTAs with disciplines that oblige state enterprises to accord fair and equitable treatment to non-state enterprises (39), and to afford them adequate opportunities to compete (37). Nearly as frequent (36 PTAs) is a type of transparency commitment that requires transparency of ownership, governance, and financial information. By contrast, provisions providing for discussion of information on state enterprises, or deliberation and assessment of their operations or conduct, are significantly less common (10). Interestingly, only 24 of the PTAs that regulate state enterprises include corporate governance requirements (about “structure” or “behavior”). All transparency provisions are discussed in further detail below.While 35 PTAs specifically exclude the application of public procurement rules to state enterprises, 27 include the fourth-less-common type of discipline, which is the obligation of state enterprises not to distort trade. Only a handful of PTAs regulating state enterprises include specific rules on strategic sectors (19), public services (17), financial services (12), and other specific disciplines for certain sectors (15). The question on the existence of provisions with specific disciplines for certain sectors or objectives has been coded in 15 PTAs. In conclusion, four commitments emerge as the most relevant provisions in PTAs that regulate state enterprises: prohibitions of anti-competitive behavior (87 percent), rules on subsidies (71.5 percent), prohibitions of discrimination (64 percent), and provisions that directly impact ownership regimes (50 percent). Other disciplines are less common: the obligation to act in line with the parties’ obligations (38 percent) or in accordance with commercial considerations (29 percent), the obligation to accord fair and equitable treatment (18 percent) and to afford companies adequate opportunities to compete (17 percent), and finally the obligation not to distort trade (12 percent). Overall, transparency requirements show a modest recurrence. The most common commitment is the notification requirement, present in 36 percent of PTAs. Transparency of ownership, governance, and financial information are present in 16.5 percent of the agreements; discussion of information, or deliberation and assessment of operations or conduct, can be found in a mere 4.6 percent of PTAs; and requirements pertaining to corporate governance are in 11 percent of PTAs. A significant share of state enterprise provisions (34.4 percent) is not enforceable through dispute settlement mechanisms. The reason is that many state enterprise provisions are included in competition chapters, for which PTAs regularly exclude dispute settlement. 16.4.2.2 Most common provisions by level of development Figure 16.8 shows that North-North PTAs have the higher number of state enterprise provisions, followed by North-South PTAs. South-South agreements lag behind, with, respectively, 37.6 and 24.4 percent fewer state enterprise provisions than North-North and North-South PTAs. 482 State-Owned Enterprises Figure 16.8: Most common provisions by level of development Source: Deep Trade Agreements Database. This general pattern holds when looking at the categories of provisions in more detail, as shown in Figure 16.9.What comes out, however, is a higher frequency of disciplines and coverage provisions (from around 75 percent for South-South and North-South PTAs to 90 percent in North-North PTAs). The rules on enforcement also score high, with shares between 67 and 78 percent. By contrast, transparency provisions clearly score lower. Even the 50 percent share in North-North PTAs is not very high and decreases to 46 percent in North-South and to a mere 29 percent in South-South PTAs. Interestingly, special and differential treatment, cooperation, and miscellaneous (e.g., supervision of agreement and further negotiations) provisions are virtually non-existent. Figure 16.9: Most common provisions by level of development Source: Deep Trade Agreements Database. 16.4.2.3 Most common provisions by geographic group Figure 16.10 shows that North America and EFTA PTAs have the higher number of state enterprise provisions (respectively, 17.8 and 16.3, on average). Latin America and the 483 Handbook of Deep Trade Agreements Caribbean follows with 14.1. Then come two different groupings: PTAs with between 12.4 and 10.4 provisions (Middle East & North Africa, East Asia & Pacific, the EU, and Central Asia) and those with 8.8 and 6.0 (Sub-Saharan Africa, South Asia). Figure 16.10: Most common provisions by geographic group Source: Deep Trade Agreements Database. EFTA and North America PTAs have the highest average number of state enterprise provisions, as shown in Table 16.2. EFTA PTAs have a very strong frequency in coverage, disciplines, and enforcement provisions. North America PTAs consistently score very high in most of the categories, except for SDT and miscellaneous provisions. Interestingly, they have the highest score in definitions provisions. Table 16.2: Most common provisions by geographic group (percent) Central EFTA East Asia European Latin Middle North South Sub- Asia & Pacific Union America East America Asia Saharan & & North Africa Caribbean Africa Definition provisions 1 20 50 9 48 14 76 23 0 Coverage provisions 85 97 72 77 78 78 80 64 44 Discipline provisions 85 97 72 77 78 78 80 59 44 Transparency provisions 31 63 42 32 47 65 72 45 22 Enforcement provisions 76 89 66 68 74 78 80 59 44 SDT provisions 0 0 2 0 4 0 4 5 0 Miscellaneous provisions 0 3 1 2 2 0 4 0 0 Source: Deep Trade Agreements Database. 484 State-Owned Enterprises 16.4.2.4 Notable countries A few countries have been selected, mostly because of their active role in signing PTAs, for a deeper analysis. As shown in Figure 16.11, Canada has by far the highest score (23.1). EFTA is the second highest (16.3). Then there is a large group with scores between 13 and 15, which includes Korea (13), New Zealand (13.5), Australia (13.6), the United States (13.8), Singapore (14), Central America (14.4), and Mexico (15.1).The few other countries selected lag slightly behind: Japan and Chile (12.7), the EU and Turkey (11.5), and China (10.1). Figure 16.11: Most common provisions by notable countries Source: Deep Trade Agreements Database. These numbers are explained in Table 16.3. Generally speaking, there are high and very high scores for coverage, disciplines, and enforcement provisions, and lower scores for definitions, and transparency and corporate governance. As seen already, the recurrence of SDT and miscellaneous provisions is scarce. Looking now at the individual countries, Canada is extremely high in all scores (92-100 percent) and comparatively high in definitions and transparency. EFTA is very high in all categories (with a lower performance on transparency provisions) but is low on definitions. EU is very modest in transparency provisions (32) and definitions (only 9). Singapore, Central America, Mexico, and Korea show a similar pattern, with very high scores for coverage, disciplines, and enforcement and more modest (slightly below 60 percent) scores for definitions and transparency. Australia has good scores for coverage and disciplines (both 77) and relatively good scores (between 62 and 69) for definitions, transparency, and enforcement. New Zealand has very high scores except for transparency, and corporate governance, which score merely 50 percent. Similarly, Turkey has very high score in all categories, but its PTAs feature no definitions at all. The United States is consistent in not having high scores (50s-60s) in all categories. China scores less than 60 percent in all categories, while Chile scores relatively high (69) only for coverage and disciplines; definitions, transparency, and enforcement are below 60 percent. 485 Handbook of Deep Trade Agreements Table 16.3: Most common provisions by notable countries (percent) Korea, Rep. Singapore European Australia America Zealand Canada Central Mexico Turkey United States Union Japan China EFTA Chile New Definition provisions 9 64 20 46 0 55 92 69 67 63 57 57 59 43 Coverage provisions 77 64 97 69 84 82 100 77 83 81 81 71 76 57 Discipline provisions 77 64 97 69 84 82 100 77 83 81 81 71 76 57 Transparency provisions 32 57 63 46 68 50 92 62 50 38 54 50 41 43 Enforcement provisions 68 64 89 58 84 73 100 62 75 81 81 71 71 57 SDT provisions 0 0 0 8 0 5 8 15 8 6 3 7 0 0 Miscellaneous provisions 2 0 3 4 0 5 8 8 8 6 0 7 0 0 Source: Deep Trade Agreements Database. 16.4.3 Specific analysis of types of provisions This section sketches a few descriptive statistics concerning specific types of state enterprise provisions, in particular, those on definitions, coverage, disciplines, exceptions, transparency and corporate governance, and enforcement. 16.4.3.1 Definitions The WTO does not provide an official definition of state trading enterprise but only a working definition24 and an illustrative list.25 The same applies to the concept of monopoly. Definitions are, however, important because through them the parties can make the scope of disciplines broader or narrower. Therefore, many PTAs include their own definitions agreed upon by the parties. The most common definitions concern the concept of monopoly and state enterprise, as shown in Figure 16.12. In particular, 65 PTAs include definitions of public or government monopoly, 35 feature definitions of designated monopolies, and 37 include definitions of state-owned enterprise, state enterprise, or public undertaking. The remaining definitions included in the dataset are scarcely coded with the exception of the notion of commercial considerations, which appears in 17 PTAs (while only 2 agreements include an explicit definition of commercial activity). 24 See the Memorandum of Understanding on the interpretation of Article XVII of the GATT 1994, para. 1. 25 Illustrative list of relationships between governments and state trading enterprises and the kinds of activities engaged in by these enterprises (G/STR/4). 486 State-Owned Enterprises Over time, there has been a progressive increase in PTAs that include the most common definitions; i.e., public monopoly, state enterprise, and designated monopoly, shown in Figure 16.12. Figure 16.12: Definitions Source: Deep Trade Agreements Database. Given its practical importance, one specific question focuses on the tests used to determine government control or influence (“does the agreement provide for a definition of government control or influence?”). However, only 7 PTAs include such a definition.26 One good example is (83) US-Singapore (Articles 12.8.1, 12.8.5), which provides for various tests and includes an annex (Annex 12A) with various examples of application of these tests. Another good example is the definition in (281) Trans-Pacific Partnership (Article 17.1). Finally, it is interesting to note that a few PTAs, in addition to (or replacing) common definitions, incorporate domestic definitions. See, for example, (27) NAFTA, Annex 1505, which contains each country’s definitions of state enterprise. Domestic definitions are present in 10 PTAs.27 16.4.3.2 Coverage (entities, sectors, ownership regimes) This section analyzes the PTAs in terms of entities and sectors covered, as well as the presence of any provision regulating ownership and liberalization processes. 26 (232) Canada-Jordan, (231) Canada-Panama, (256) Korea-Australia, (276) Korea-Colombia, (211) Korea-US (281) Trans-Pacific Partnership, (83) US-Singapore. 27 (27) NAFTA; (33) Canada-Israel; (37) Canada-Chile; (83) US-Singapore; (169) Canada-Peru; (201) Canada- Colombia; (231) Canada-Panama; (232) Canada-Jordan; (260) Canada-Republic of Korea; (261) Canada- Honduras. Interestingly, Canada is almost always a signatory to these PTAs. 487 Handbook of Deep Trade Agreements 16.4.3.2.1 Entities It is necessary to begin with a brief note on methodology. The dataset includes questions on the coverage of various types of state entities (state-owned enterprises, state enterprises, public undertakings, state trading enterprises, government and designated monopolies, and entities with special or exclusive privileges or rights).When coding these questions, two approaches can be followed: either (a) explicit reference, or (b) inclusion through a reasonable interpretation of the legal text (which takes into account the significant overlap among these definitions).28 While the first approach is less prone to errors, it may not give an accurate representation of the entities actually covered. In other words, it may de facto be under-inclusive.To give a more accurate picture of the coverage of the various PTAs, the coding follows a more comprehensive approach, with the comments section of the dataset including the rationale for the coding decisions. This should reduce the susceptibility to over-inclusion of this approach. Since this chapter focuses mostly on the quantitative analysis of what is “coded” (“1”) and “not coded” (“0”), which may lead to losing important granularity of information, in this section we juxtapose and analyze the data coming out of the two approaches. Figure 16.13 depicts the statistics generated by a comprehensive approach to coding PTAs. If one bears in mind that 218 PTAs include state enterprise provisions (of which 193 include specific rules while 25 include general competition provisions), the main finding is that virtually all PTAs are apt to cover and regulate virtually all types of entities considered. The only important finding from this approach is not quantitative but qualitative, in so far as it signals a potentially broad coverage of all PTAs (always with respect to the question of the entities that are covered). It is now interesting to juxtapose these data with those coming out of the stricter coding, which detects only explicit references. We call this coding approach “limited” due to the Figure 16.13: Coverage (entities) – comprehensive coding approach Source: Deep Trade Agreements Database. 28 For example, if one PTAs explicitly regulates public monopolies this does not exclude that these monopolies may also act as a trading enterprise or may have special rights or privileges. 488 State-Owned Enterprises possibility of under-inclusion. In particular, it may be too easily concluded that, since the PTA does not explicitly make reference to a given type of entity, this entity cannot be captured and regulated. The template on which Figure 16.14 is based is similar to that of the previous figure with very few minor tweaks. On the one hand, references to state and delegated monopolies have been merged. On the other hand, the figure maps specific references to state entities that the dataset shows under different provisions (i.e., monopolies of commercial character, undertakings entrusted with services of general economic interest, and monopoly suppliers of services/exclusive service suppliers). Figure 16.14: Coverage of state entities, limited coding approach Source: Deep Trade Agreements Database. With this methodological premise in mind, the data coming out of the two figures can be compared. In particular, the two different coding exercises seem to bear similar results. PTA provisions on state enterprises seem to share a specific focus on monopolies, with a cumulative score of 196. Adding to this number the category of monopoly suppliers of services/exclusive service suppliers (32) brings the score significantly higher. References to state enterprises still feature second (99), and are followed by undertakings with special or exclusive rights (85) and state trading enterprises (63). Incidentally, it should be noted that the mission of state monopolies of commercial character (language derived from (1) EC Treaty) is essentially the same (in GATT parlance) as that of state trading enterprises; it would therefore make some sense to put these two together, which would lead to a score of 112. Finally, there are two categories which refer mostly to the PTAs signed by the EU and countries close to it: state monopolies of commercial character (49) and undertakings entrusted with services of general economic interest (17). The distribution of express references to state entities by income group is shown in Figure 16.15. Generally speaking, all groups seem to follow the general pattern indicated above, with a prevalence of references to monopolies, state enterprises, and undertakings with special or exclusive rights. These data are reinforced by considering the categories of state monopolies of commercial character and the specific group of state entities in services 489 Handbook of Deep Trade Agreements (monopoly suppliers of services/exclusive service suppliers). Specific references to state trading enterprises, mostly through references to GATT Article XVII, and to undertakings entrusted with services of general economic interest, lag behind. North-North and South- South PTAs are characterized by a relatively even distribution of references (and hence disciplines). This evenness is less marked for North-South PTAs, in particular for the very low frequency of provisions on enterprises that provide services of general economic interest. Figure 16.15: Coverage of state entities by income group, limited coding approach Source: Deep Trade Agreements Database. 16.4.3.2.2 Sectors Statistics on sectoral coverage, combined with the recurrence of special provisions regulating ownership and liberalization, are shown in Figure 16.16. Most of the PTAs with state enterprise provisions include regulation of state entities in the agriculture sector (89 percent). By contrast, only 61 percent of PTAs also cover service state enterprises. The provisions having an impact on ownership regimes are quite recurrent, featuring in half (that is 109) PTAs with state enterprise provisions. As seen above, these provisions belong to a varied group and are particularly useful in giving indications about the objectives of the rules. Figure 16.16: Coverage (sectors) Source: Deep Trade Agreements Database. 490 State-Owned Enterprises If one looks at the evolution of sector provisions over time, the coverage of agriculture is stable, while provisions on state enterprises in services and on ownership and liberalization are increasingly included. The analysis of these provisions by level of development, shown in Figure 16.17, offers interesting statistics. The general pattern analyzed above can be found only in North- South PTAs. Quite expectedly, while North-North PTAs have a similar coverage of state enterprises in the agriculture and services sectors, South-South PTAs cover state entities in agriculture more than three times as often as they cover state entities in services (66 versus 20). Interestingly, provisions on ownership regimes are very frequent in all PTA groups. Figure 16.17: Coverage (sectors) by income groups Source: Deep Trade Agreements Database. 16.4.3.3 Disciplines Figure 16.18 reproduces part of Figure 16.7 above.29 This section of the paper analyzes these known statistics to offer a little bit more detail on the relevant provisions. Figure 16.18: Disciplines Source: Deep Trade Agreements Database. 29 In referring to disciplines, Figure 16.7 focused on all the questions in rows 37-48 of the template (section on substantive disciplines). This part of the chapter distinguishes among these provisions. Section 16.4.3.3 thus only refers to the substantive commitments in rows 37-44. Other questions included in the substantive disciplines section of the template, and in particular “exceptions specific to state enterprises” and “any other specific discipline,” are analyzed in Section 16.4.3.4 below. 491 Handbook of Deep Trade Agreements The two most common disciplines on state enterprises are in fact generic provisions: competition laws (190) and subsidy rules (156). What has been coded under the subsidy law question is simply the existence of provisions that provide substantive obligations or prohibitions on subsidies to state enterprises (excluding non-substantive and countervailing duty [CVD] provisions). Importantly, if the only provisions applicable to state enterprises in the PTA are subsidy rules, they have not been coded. Given their importance, a different approach is adopted for competition provisions. We have coded also those agreements where the only coded disciplines are competition provisions. Many disciplines specifically dedicated to state enterprises are expressly located within competition chapters, signaling that the parties often consider state enterprises a competition policy/law issue.30 For example, in (33) Canada-Israel, state enterprises and government monopolies are regulated under separate provisions, both included in the chapter on competition policy. In other cases, the PTA does not include separate provisions on state enterprises in competition chapters, but it does expressly subject public undertakings to general competition provisions. This is the case, for example, in (221) EFTA-Montenegro (Article 17.2). At the same time, there are many PTAs that do not feature any specific discipline for state enterprises but include general competition law norms. These norms are broad enough to cover the conduct of state enterprises. In particular, the notions of enterprise, undertaking, or business have been taken by the coders to also include state or public enterprises, undertakings, or businesses. Considering the relevance of competition laws for governing the conduct of state enterprises, these general provisions have been coded as well. The data confirm the importance of competition provisions. In 25 PTAs, competition provisions are applicable to state enterprises on their own; in 165 PTAs, they are applicable in combination with specific state enterprise disciplines. As seen already in Figure 16.3, the third-most-common (140) disciplines are those that prohibit discrimination, which is mostly based on national treatment or MFN. It is interesting to note that, from the perspective of evolution over time, provisions on competition and subsidies for state enterprises are firmly stable, but many others show a progressive increase. This applies, for example, to the obligations to act in accordance with commercial considerations, accord fair and equitable treatment, and afford companies adequate opportunities to compete. The same evolution is seen with prohibition of discrimination. 30 What is coded is only anti-competitive (or antitrust) behavior proper. Thus, the question does not cover provisions such as GATS Article VIII:2, which is coded under the question on delegated entities’ compliance with the party’s commitments. 492 State-Owned Enterprises The distribution of disciplines by income group is shown in Figure 16.19. Considering the distribution of commitments within each group, this figure largely confirms the general pattern that emerges from Figure 16.18, with the same predominance and sequence as the first three types of disciplines (competition, subsidies, non-discrimination). There are, however, a few minor variations with respect to the remaining commitments. In the PTAs concluded between North countries, the fourth-most-common discipline concerns the obligation to comply with commitments, which is always more frequent than the obligation to comply with commercial considertations. In South-South PTAs, by contrast, the sequence is different, though all disciplines have very close scores. The fourth-most-common provision is the obligation to comply with commercial considerations (10), followed by the obligations to afford adequate opportunities to compete (9), to comply with commitments (8), and to accord fair and equitable treatment (8).The less common discipline for South-South and North-North PTAs is the obligation not to distort trade (featuring in respectively only 1 and 2 PTAs). For North-South PTAs the less common discipline is the obligation to accord adequate opportunities to compete (included in 22 PTAs). In North-North PTAs, there are very few agreements that include obligations to afford adequate opportunities to compete (6) and fair and equitable treatment (5). In North- South PTAs, by contrast, non-distortion of trade, adequate opportunities to compete, and commercial considerations all have significant and very close scores (26, 24, 22). Figure 16.19: Disciplines by income group Source: Deep Trade Agreements Database. From another perspective, in North-North PTAs most of the disciplines (in particular, the top four) are evenly spread (with scores between 24 and 34), showing that the regulation of state enterprises is fairly regular. This also occurs in North-South PTAs but only for the first three obligations (with scores between 83 and 96). There is then a significant fall (to 51, obligation to accord fair and equitable treatment, and 40, compliance with commitments) and a stabilization (in the 20s) for the remaining obligations. Interestingly, the difference in recurrence between the top four commitments is very marked for South-South PTAs. The fall from competition provisions (60) to subsidy provisions (38) and non-discrimination (30) is significant, and so is the decrease in compliance with commercial considerations (10). 493 Handbook of Deep Trade Agreements The distribution of various disciplines within each geographic region is shown in Table 16.4. Most of the regions follow the general pattern outlined in Figure 16.18, with competition, subsidy and non-discrimination provisions as the three most common disciplines regulating state enterprises. In Central Asia the predominance of this triad of provisions over the others is very marked. Four regions depart slightly from this pattern. In Middle East & North Africa and North America PTAs, competition and non-discrimination disciplines share the first position, with subsidy provisions following in the former and, interestingly, commercial considerations and compliance with commitments in the latter. In East Asia & Pacific, the third-most- common disciplines are compliance with commitments, followed by non-discrimination and commercial considerations. In South Asia PTAs, the most common disciplines are compliance with commitments, followed by two of the usual top three: non-discrimination comes first, followed by subsidies and then, with the same score, commercial considerations, adequate opportunities to compete, and fair and equitable treatment, which all precede competition rules. In some regions (EFTA, East Asia & Pacific, Latin America & Caribbean, Sub-Saharan Africa), the various disciplines are evenly spread, with very close scores for most of the disciplines. Interestingly, the EU and Middle East & North Africa PTAs both feature many non-distortion provisions, which thus come fourth in the list of the most common provisions. Table 16.4: Disciplines by geographic region (Number of PTAs) Central EFTA East Asia European Latin Middle North South Sub- Asia & Pacific Union America East America Asia Saharan & & North Africa Caribbean Africa Non-discrimination 34 30 35 30 46 27 18 9 5 Commercial considerations 7 21 24 1 28 8 17 7 2 Subsidization 38 33 45 29 51 21 12 8 7 Competition 64 34 53 32 53 27 18 6 7 Non-distortion of trade 6 4 2 18 6 11 3 0 0 Compliance with 5 12 43 9 25 9 17 11 5 commitments Adequate opportunities 6 15 13 1 14 2 4 7 2 to compete Fair and equitable 6 17 15 1 16 2 4 7 2 treatment Source: Deep Trade Agreements Database. 16.4.3.4 Exceptions This section discusses the exceptions to the normal application of PTA provisions to state enterprises.31 These types of provisions, and their importance for detecting the depth of 31 These provisions are spread in two parts of the template. Those on “public services” and “strategic sectors” in the section on Coverage (notably, in rows 32 and 33); and those on “exceptions” and “any other specific disciplines” in the Substantive Disciplines section (rows 47 and 48). 494 State-Owned Enterprises integration pursued by the state enterprises chapter of the PTAs, has been outlined above (section 16.3.2.2). In particular, as Figure 16.20 shows, exceptions are by far the most common provision (73), while the three other categories (public services, strategic sectors, any other specific disciplines) lag behind (with scores between 15 and 19). Figure 16.20: Exceptions Source: Deep Trade Agreements Database. Figure 16.21 shows the income group dynamics of these raw data. The large majority of all these provisions can be found in North-South PTAs. Some provisions (public services, any other specific disciplines) score very low or are even non-existent in the other groups. South- South agreements mostly include provisions on strategic sectors, while North-North PTAs have a preference for exceptions provisions (followed by those on strategic sectors). North- North PTAs do not have any provisions on public services or any other specific disciplines. Figure 16.21: Exceptions by income group Source: Deep Trade Agreements Database. The dynamics behind exceptions provisions are futher untangled in Table 16.5. While all geographic groups make use of exceptions and (partly) of any other specific disciplines, it is the EU (12), Central Asia (8, once again mostly because of EU links), and Latin America & 495 Handbook of Deep Trade Agreements Caribbean (6) that include specific provisions on public services.32 While the EU leads for public service provisions, Latin America & Caribbean is ahead in strategic sectors provisions (15). Other regions have these provisions but with a significantly lower occurrence. Table 16.5: Exceptions by geographic region/group Public Strategic Exceptions Any other Services Sectors Specific Discipline Central Asia 8 0 17 3 EFTA 1 1 22 5 East Asia & Pacific 0 3 18 1 European Union 12 0 25 4 Latin America & Caribbean 6 15 23 7 Middle East & North Africa 0 1 16 3 North America 0 4 14 0 South Asia 0 0 0 0 Sub-Saharan Africa 0 0 1 0 Source: Deep Trade Agreements Database. 16.4.3.5 Transparency and corporate governance The questions on transparency and corporate governance are always specific to state enterprises. The three template questions on transparency relate to the key categories of transparency broadly intended as a tool of trade policy.33 Figure 16.22 shows a high prevalence of basic notification requirements. The relevant template question outlines few examples: notification of the grant of monopoly, of exclusive or special rights, of the operations or conduct of state enterprises, and of goods or services supplied or procured by the state enterprise. These requirements often refer to obligations deriving from GATT/GATS provisions. See, for example, (82) US- Chile (Article 16.3.2(b), (169) Canada-Peru (Article 1305.2), or (163) Australia-Chile (Article 14.4.4(b) and 14.6), which provide for the obligation to notify the designation of monopolies. Notification requirements are present in 78 PTAs with state enterprise provisions, amounting to a 36 percent share of the total. 32 (188) Colombia-Mexico, (217) Colombia-Northern Triangle, (230) EU-Central America, (229) EU- Colombia and Peru, (242) Mexico-Uruguay, (204) Peru-Chile. 33 See Collins-Williams and Wolfe 2010. 496 State-Owned Enterprises The second-most-common commitment pertains to any mechanism to ensure transparency of ownership, governance, and financial information.This mechanism includes the obligation to inform PTA bodies, such as Joint Committees, of the measures adopted to comply with the obligation to adjust state monopolies of commercial character. See, for example, (39) Turkey- Israel (Article 13.2) or (40) EU-Tunisia (Article 37). With a score of 36, this commitment is present in only 16.5 percent of the relevant PTAs. Only 10 PTAs, or 4.6 percent, provide for any form of collaborative transparency, such as discussion of information, or deliberation and assessment of operations or of the conduct of state enterprises. Figure 16.22: Transparency and corporate governance Source: Deep Trade Agreements Database. Corporate governance requirements may relate broadly to the structure of state enterprises or their behavior.34 These commitments feature in only 24 PTAs, or 11 percent of the sample. There are two main types of corporate governance requirements. First, certain PTAs include specific behavioral and structural requirements in telecommunications chapters. For example, (241) Chile-Nicaragua (Article 13.7.2) provides that: Each Party shall endeavour to adopt or maintain effective measures to prevent anti- competitive behavior, such as: (a) accounting requirements; (b) structural separation requirements, (c) rules for the monopoly, main supplier or dominant operator to grant its competitors access to and use of their networks or public telecommunications services on terms and conditions no less favourable than those granted to themselves or their subsidiaries; or (d) rules for timely disclosure of the technical changes of public telecommunications networks and their interfaces. [original language: Spanish, translation into English by the author] 34 See World Bank 2014; OECD 2010, 2015. 497 Handbook of Deep Trade Agreements Second, certain PTAs provide for actions to ameliorate any negative impact when a state designates a monopoly. Thus, for example, (167) Panama-Taiwan, China (Article 15.03.2(b)) reads: If a Party’s law does permit it, where the Party intends to designate a monopoly or a state enterprise, and the designation may affect the interests of persons of the other Party, the Party shall: […] (b) endeavour to introduce at the time of designation such conditions on the operation of the monopoly as will minimize or eliminate any nullification or impairment of benefits under this Agreement. (original language: Spanish; translation into English by authors) Considering now the evolution over time of transparency and corporate governance requirements, the coding has shown a steady increase of notification requirements, while the remaining transparency and corporate governance provisions are stable. Figure 16.23 and Table 16.6 show the usual breakdown of the data by income group and geographic group. Starting with Figure 16.23, a cross-commitment comparison generally shows a similar pattern with, from highest to lowest, more provisions in North-South, and then North-North, and South-South PTAs. More specifically, the data show a significant variation for transparency of ownership, governance, and financial information, which are more common in North-South PTAs than in South-South agreements. Notification requirements are greatly prevalent among the various transparency requirements in North- South PTAs. A cross-group comparison generally follows the pattern of Figure 16.22. That being said, North-North PTAs have more corporate governance requirements (7) than transparency of ownership, governance, and financial information (5). Notification requirements and transparency of ownership, governance, and financial information have almost the same recurrence (13 and 12) in South-South PTAs. Figure 16.23: Transparency and corporate governance by income group Source: Deep Trade Agreements Database. 498 State-Owned Enterprises As shown in Table 16.6, notification requirements are the most common transparency provision in EFTA, East Asia & Pacific, Latin America & Caribbean, North America, and South Asia. By contrast, transparency of ownership, governance, and financial information are prevalent in Central Asia, the EU, and Middle East & North Africa. Quite interestingly, the large majority of corporate governance requirements (20 out of 34) are found in Latin America & Caribbean PTAs. This region also moderately leads for discussion and deliberation provisions, which are evenly spread among all groups (except for Central Asia and the EU, which both score “0”). Table 16.6: Transparency and corporate governance by geographic group Notification Transparency of ownership, Discussion and Corporate requirements governance, deliberation governance financial information Central Asia 9 15 0 0 EFTA 16 5 3 0 East Asia & Pacific 35 5 3 5 European Union 4 11 0 0 Latin America & Caribbean 30 6 5 20 Middle East & North Africa 6 18 1 3 North America 17 7 3 6 South Asia 10 0 1 0 Sub-Saharan Africa 2 0 2 0 Source: Deep Trade Agreements Database. 16.4.3.6 Enforcement The dataset includes a few questions on enforcement mechanisms that PTAs use for state enterprise provisions. The most important information coming out of the statistics is the relatively low legal enforceability of many state enterprise disciplines. We consider both the presence of bodies and committees tasked with the enforcement of the rules, including state enterprise provisions, and dispute settlement applicable to state enterprises. If one considers the number of PTAs with state enterprise provisions (South-South 72; North-North 36; North-South 110), bodies or committees tasked with the enforcement of the rules are a common features of most PTAs (85 percent in South-South PTAs, 72 percent in North-North PTAs, and 87 percent in North-SouthPTAs). However, despite the presence in 183 PTAs (84 percent of the total PTAs with state enterprise provisions) of bodies or committees tasked with the enforcement of the rules, only 143 (65.3 percent) have state enterprise provisions that are enforceable through dispute settlement mechanisms. This is largely due to the frequent exclusion of competition law provisions from the dispute settlement system of the PTA (and the already-mentioned 499 Handbook of Deep Trade Agreements close link that exists between state enterprise provisions and competition provisions). In a word, flexibility is achieved through exclusion of dispute settlement. But the picture is considered significantly differentiated according to the income group (Figure 16.24). Dispute settlement mechanisms applicable to state enterprises are present in 79 percent of PTAs that are North-South but only in 50 percent of South-South PTAs and 55.5 percent of North-North PTAs. Figure 16.24: Bodies or committees by income group Source: Deep Trade Agreements Database. Figure 16.25: Dispute settlement by income group Source: Deep Trade Agreements Database. As Figure 16.26 shows, the pattern of prevalence of institutional enforcement over dispute settlement is mostly followed in all geographical regions. In some cases, the data show close scores between the two systems of enforcement; in others the prevalence of institutional (i.e. bodies or committees) enforcement is more marked, Central Asia being the best example. Interestingly, Sub-Saharan Africa is the only region with a prevalence of dispute settlement. 500 State-Owned Enterprises Figure 16.26: Enforcement by geographic group Source: Deep Trade Agreements Database. 16.5. CONCLUSIONS The goal of the mapping described in this chapter has been to determine the depth of state enterprise provisions in 283 PTAs. This chapter has made a start by providing a few descriptive statistics and some initial analysis. Despite important variations among different levels of development, geographic groups, and countries, few very broad patterns can be noted. A very large share of PTAs (77 percent) include state enterprise provisions, which therefore represent a common chapter of most PTAs. If one looks, however, at the content of these chapters, some more granular findings emerge. First, the most common disciplines are not specific to state enterprises but refer to competition and subsidy laws. Even the third-most-common group of commitments is constituted of various non-discrimination provisions. Second, PTAs show a significant recurrence of rules that have an impact on ownership and property regimes, and these rules may embody opposite principles: neutrality (protecting state enterprises), and a level playing field (protecting private competitors). It is fair to conclude that most of the commitments can be subsumed under one or the other of these principles. Exceptions of various types, indicating different goals and integration dynamics, are common. Third, the picture that emerges is one of relative flexibility. Transparency and corporate governance requirements, which are key in regulating state enterprises, show a modest frequency. Legal enforceability is generally low, with access to dispute settlement in only two- thirds of the agreements and with a significant variation among income groups. Although the desire for flexibility is often a side-effect of the incorporation of state enterprises discipline 501 Handbook of Deep Trade Agreements within general competition disciplines, the fact that many state enterprises operating in cross-border trade may not be subject to enforceable rules raises concerns. As noted in Chapter 14 on the regulation of subsidies, these are just the initial sketches of a very preliminary analysis. More could be done to determine the depth of these disciplines, but the great wealth of data and descriptive analysis in this chapter certainly represent a solid starting point for further research. ACKNOWLEDGMENTS We would like to thank all participants in the 2018 Washington, DC, workshop where a preliminary version of this chapter was presented. A special thank-you goes to Alvaro Raul Espitia Rueda for his painstaking technical support. All errors remain those of the authors. 502 State-Owned Enterprises REFERENCES Blum, Françoise, and Anne Logue. 1998. State monopolies under Mavroidis, Petros C., and Merit E. Janow. 2017. Symposium EC law. London: Wiley. on state-owned enterprises in China. World Trade Review 16 (4). https://www.cambridge.org/core/journals/world-trade- Bown, Chad. 2018.Testimony before the US-China Economic review/issue/symposium-on-stateowned-enterprises-in- and Security Review Commission: Hearings on US tools to china/BEF7AE692CA146C272B3B4F7088C8E67. address Chinese market distortions (June 8). Mavroidis, Petros C., Patrick Messerlin, and Jasper Wauters. Buendia-Sierra, Jose-Luis. 2000. Exclusive rights and 2008. The law and economics of contingent protection in the WTO. monopolies under EC Law: Article 86 of the EC Treaty. Cheltenham, UK: Elgar. London: Sweet & Maxwell. OECD (Organisation for Economic Co-operation and Collins-Williams,Terry, and Robert Wolfe. 2010.Transparency Development). 2009. State-owned enterprises and the principle as a trade policy tool:The WTO’s cloudy windows. World Trade of competitive neutrality. Paris: OECD Publishing. Review 4:551-81. ———. 2010. Competition and corporate governance. Paris: Cottier, Thomas, and Petros C. Mavroidis. 1998. State trading OECD Publishing. in the twenty-first century. Ann Arbor: University of Michigan Press. ———. 2011. Competitive neutrality and state-owned enterprises: Challenges and policy options. Paris: OECD Publishing. Dür, Andreas, Leonardo Baccini, and Manfred Elsig. 2014.The design of international trade agreements: Introducing a new ———. 2015. Guidelines on corporate governance of state-owned dataset. Review of International Organization 9:353-73. enterprises. Paris: OECD Publishing. European Commission. 2016. State-owned enterprises in the ———. 2016. State-owned enterprises as global competitors: A EU: Lessons learnt and ways forward in a post-crisis context. challenge or an opportunity? Paris: OECD Publishing. Institutional Paper 031 (July), European Commission, Brussels. Oliver, Peter. 2010. Oliver on free movement of goods in the Fleury, Julien Sylvestre, and Jean-Michel Marcoux 2016. The European Union. Oxford, UK: Hart. US shaping of state-owned enterprise disciplines in the Trans- Pacific Partnership. Journal of International Economic Law 19 (2): Rubini, Luca. 2009. The definition of subsidy and state aid:WTO 445-65. and EC law in comparative perspective. Oxford: Oxford University Press. Fox, Eleanor, and David Crane. 2010. Global Issues in antitrust and competition law. Eagan, MN: West Publishing. Szyszczak, Erika. 2007. The regulation of the state in competitive markets in the EU. Oxford: Hart. Hofmann, Claudia, Alberto Osnago, and Michele Ruta. 2017. Horizontal depth: A new database on the content of Willemyns, Ines. 2016. Disciplines on state-owned enterprises preferential trade agreements. Policy Research Working Paper in international economic law: Are we moving in the right 7981, World Bank, Washington, DC. direction? Journal of International Economic Law 19 (3): 657-80. Jackson, John H. 1969. World trade and the law of GATT. World Bank. 2014. Corporate governance of state-owned enterprises: Indianapolis: Bobbs-Merrill. A toolkit. Washington, DC: World Bank. Krajewski, Markus. 2015. Services of general interest beyond the Wu, Mark. 2016. The “China Inc.” challenge to global trade single market: External and international law dimensions. The governance. Harvard International Law Journal 2:261-324 (Spring). Hague: T.M.C. Asser. Zhou, Weihuan, Henry Gao, and Xue Bai. 2018. China’s SOE Krajewski, Markus. 2003. National regulation and trade reform: Using WTO rules to build a market economy. Paper liberalization in services: The legal impact of the General Agreement presented at Sixth Biennial Global Conference, Society of on Trade in Services (GATS) on national regulatory autonomy. International Economic Law, July 1. Dordrecht: Kluwer. 503 Handbook of Deep Trade Agreements 504 Competition Policy CHAPTER 17 Competition Policy M. Licetti, G. Miralles, and R. Teh 505 Handbook of Deep Trade Agreements CHAPTER 17 Competition Policy M. Licetti*, G. Miralles*, and R. Teh† * World Bank, Washington, DC, United States † World Trade Organization, Geneva, Switzerland CONTENTS 17.1. INTRODUCTION: THE ROLE OF COMPETITION POLICY IN TRADE AGREEMENTS 507 17.2. PRIOR STUDIES MAPPING COMPETITION COMMITMENTS IN TRADE AGREEMENTS 509 17.3. METHODOLOGY 513 17.3.1 Economy-wide obligations 513 17.3.2 Sector-specific obligations 514 17.3.3 Enforceability 514 17.4. THE GROWING ROLE OF COMPETITION IN PTAs 515 17.4.1 Competition principles embedded in the general framework of PTAs 517 17.4.2 PTA obligations to foster effective competition law and antitrust enforcement 519 17.4.2.1 National competition requirements 519 17.4.2.2 Substantive provisions regulating competition policies 522 17.4.2.3 Procedural fairness commitments to support competition enforcement 526 17.4.2.4 Institutional arrangements 529 17.4.3 PTA obligations to foster competitive neutrality 532 17.4.4 PTA obligations to remove sector-specific anticompetitive regulations 535 17.4.4.1 Prohibitions on rules that reinforce dominance or limit entry 536 17.4.4.2 Prohibitions on rules that are conducive to collusive outcomes or increase the cost of competing in the market 540 17.4.4.3 Prohibitions on rules that discriminate or protect vested interests 541 17.4.5 Enforceability of provisions 543 17.4.6 Direct applicability 545 17.5. CONCLUSIONS 546 ACKNOWLEDGMENTS 546 REFERENCES 547 ANNEX 549 506 Competition Policy 17.1. INTRODUCTION: THE ROLE OF COMPETITION POLICY IN TRADE AGREEMENTS Even though many countries have opened to trade, markets in developing economies often underperform due to restrictive regulatory frameworks and anticompetitive behavior by a few dominant players. Anticompetitive practices are often permitted, supported or even created by public bodies themselves. These conditions affect not only the dynamics of internal markets but also the ability of these countries to compete internationally and reap the benefits of enhanced economic integration. Empirical evidence shows that when firms agree to fix prices, consumers pay on average 49 percent more, and even 80 percent more when key industries are cartelized.1 Competition is also impaired if regulations restrict the number of firms that may compete or limit private investment; or if they increase business risks, facilitate agreements among competitors, or discriminate against certain competitors, thereby affecting competitive neutrality. Effective competition policies, on the other hand, offer a tool to complement and support governments’ efforts to reduce barriers to trade. Active competition among market players has the potential to mitigate vested interests and facilitate the opening of markets to trade and investment. Greater competition within national markets reinforces the international competitiveness of potential exporters through increased incentives to foster productivity, innovation, and efficiency. Additionally, international trade reinforces competition in national markets by increasing contestability, entry, and rivalry through increased presence of foreign products, services, and investments. Empirical evidence suggests, for example, that (a) the elimination of entry barriers, increased rivalry, and leveling the playing field in upstream sectors contribute to export competitiveness in downstream manufacturing sectors; (b) pro-competition market regulation that reduces restrictions and promotes competition is an important determinant of trade; (c) enforcement of competition laws can enhance export performance and is complementary to trade reforms; and (d) industries with more intense domestic competition will export more.2 An effective competition policy framework can help accomplish several objectives: It facilitates entry to markets,3 ensures that all businesses interact on a level playing field, and discourages and penalizes anticompetitive behavior. Three complementary pillars provide the basis for achieving an effective competition framework: (a) fostering pro-competition regulations and government interventions; (b) developing the necessary measures to guarantee competitive neutrality in markets; and (c) promoting economy-wide enforcement of competition laws. These pillars rely on an effective institutional framework that can foster and guarantee healthy market conduct (Figure 17.1). 1 Conner 2014. 2 Goodwin and Pierola 2015. 3 See, for example, Kee and Hoekman 2007. 507 Handbook of Deep Trade Agreements Figure 17.1: A comprehensive competition policy framework FOSTERING COMPETITION IN MARKETS PROCOMPETITION REGULATIONS AND COMPETITIVE NEUTRALITY EFFECTIVE COMPETITION GOVERNMENT INTERVENTIONS: OPENING AND NON- DISTORTIVE LAW AND ANTITRUST MARKETS AND REMOVING ANTICOMPETITIVE PUBLIC AID SUPPORT ENFORCEMENT SECTORAL REGULATION Reform policies and regulations that Tackle cartel agreements that strengthen dominance: restrictions on the Control state aid to avoid raise the costs of key inputs number of firms, statutory monopolies, favoritism and minimize and final products and reduce bans on private investment, lack of access distortions on competition. access to a broader variety of regulation for essential facilities. products. Eliminate government interventions that are conducive to collusive outcomes or Ensure competitive neutrality, increase the costs of competing: controls including vis-à-vis state-owned Prevent anticompetitive on prices and other market variables that entities. mergers. increase business risk. Strengthen the general antitrust Reform government interventions that discriminate and harm competition on and institutional framework the merits: frameworks that distort the level playing field or grant high levels of to combat anticompetitive discretion. conduct and abuse of dominance. Source: WBG-OECD 2017; adapted from Kitzmuller and Licetti 2012. Based on this evidence, pro-competition policy obligations are being included in an increasing number of trade agreements. A recent study found that as of 2015,4 more than 200 preferential trade agreements (PTAs) include some reference to competition policy. The recognition of competition as a fundamental tool for trade, which was already tacitly embedded under the World Trade Organization (WTO) agreements by the inclusion of the concepts of most-favored-nation (MFN) treatment,5 national treatment,6 and transparency,7 is now explicitly incorporated in many PTAs. As tariffs have been progressively whittled away by trade agreements, the most important remaining barriers to integration are non- tariff - or behind-the-border - regulations. Countries that want to achieve greater economic integration have to go beyond trade in goods and make commitments in areas such as services, investment, and intellectual property. The barriers to entry in these areas are linked to domestic regulations, subsidies provided to domestic industry, the presence of state-owned enterprises (SOEs), market power, and anticompetitive practices by private enterprises. As 4 Hofmann, Osnago, and Ruta 2017. 5 Article 1 of the GATT, Article II of the GATS, Article 4 of the Trade-Related Intellectual Property Rights (TRIPS) agreement. 6 Article III of the GATT, Article II of the GATS, and Article 3 of the TRIPs. 7 Article X of the GATT, Article III of the GATS, and Article 63 of the TRIPs. 508 Competition Policy many of these issues can be dealt with effectively only through the instrument of competition policy, parties wanting to achieve deeper integration may have no other recourse but to include competition policy in their agreements. While traditional PTAs have tended to tackle competition-related issues through shallow trade-related obligations such as a general reduction of non-tariff barriers and of discriminatory customs regulations,8 more recent PTAs have recognized the need for countries to promote domestic competition in order to achieve the benefits of opening markets through trade.9 17.2. PRIOR STUDIES MAPPING COMPETITION COMMITMENTS IN TRADE AGREEMENTS During the past two decades, a number of international organizations have led the way in analyzing competition commitments in PTAs. An ECLAC study10 from 2014 reviewed 18 bilateral trade agreements with Latin America and Caribbean (LAC) countries that contain competition chapters promoting cooperation and coordination. The study found that such provisions especially benefit less developed economies paired with more developed partners. This confirmed the findings of a UNCTAD study11 from 2005, which found that competition provisions adopted at the regional level enable developing economies to reap the benefits from trade agreements.The following year,an OECD review12 of 86 regional trade agreements concluded that the main driver of competition provisions is the need to tackle anticompetitive practices that could otherwise undermine trade commitments. Another study,13 however, criticized the OECD mapping exercise for having too narrow a definition of competition provisions, and for neglecting sector-specific provisions and horizontal principles related to competition that were equally important, especially in the context of services. 8 For example, under the Uruguay Round Agreements, several WTO agreements contain competition-related provisions but do not treat competition law in a comprehensive way.The importance of incorporating competition provisions to facilitate trade and reduce entry barriers is, in any case, strongly acknowledged in several WTO agreements. Examples include the General Agreement on Trade in Services (GATS), TRIPS, the Agreement on Trade-Related Investment Measures (TRIMs), the Anti-Dumping Agreement, the Agreement on Technical Barriers to Trade (TBT), and the Agreement on Safeguards. See Article 1 of the General Agreement on Tariffs and Trade (GATT); Article II of the GATS; Article 4 of the TRIPS on MNF treatment; Article X of the GATT; Article III of the GATS and Article 63 of the TRIPS on National Treatment; Article X of the GATT 1994; Article III of the GATS; and Article 63 of the TRIPS on the Principle of Transparency. 9 See Article 17.2.1 of EU-Canada Comprehensive Economic and Trade Agreement (CETA), available at: http://ec.europa.eu/trade/policy/in-focus/ceta/ceta-chapter-by-chapter/. This same comprehensive treatment of competition policy is found in the European text for the Transatlantic Trade and Investment Partnership (EU Proposal TTIP), which includes both general competition law principles as well as a description of conduct that should be considered anti-competitive. See Article X.2. 10 Silva 2004. 11 Brusick, Alvarez, and Cernat 2005. 12 Solano and Sennekamp 2006. 13 Anderson and Simon 2006. 509 Handbook of Deep Trade Agreements That study emphasized the need to explore the treatment of monopolies, either public or private, as well as state aid as potential sources of distortion from government intervention in the markets. This inspired new research to address competition-related commitments from a broader perspective. A 2009 study14 expanded the realm of economy-wide and sector specific-competition provisions in PTAs beyond those included in the competition chapters. The study showed that focusing only on the competition policy chapter of a PTA would leave out important competition-related disciplines found in other chapters dealing with such relevant areas as services, government procurement, and intellectual property. The key provisions covered by that study (Box 17.1) form the basis of the mapping exercise in this chapter. Box 17.1: Key competition-related provisions of PTAs (i) the overall objectives of the PTA; (ii) horizontal principles of non-discrimination, transparency, and procedural fairness; (iii) sector-specific competition provisions involving investment, services, government procurement, and intellectual property; and (iv) the competition policy chapter of the agreement, including provisions on state aid, government enterprises, and other similar provisions, even though these may not appear in the competition chapter of the agreement. Source: Teh 2009. Further, in contrast with prior research, the 2009 study pointed toward a nuanced relationship between competition and trade. While competition principles are embedded in almost three- quarters or all trade agreements, they are not necessarily subordinated to trade objectives. Instead, parties appear to place an intrinsic value on the promotion of competition, as evidenced by the fact that nearly 46 percent of the sample of 74 PTAs include the promotion and advancement of market competition and cooperation in the field of competition as their main objectives. Horizontal principles such as transparency, non-discrimination, and procedural fairness are covered at some level in 31 percent of multilateral agreements and 83 percent of bilateral agreements. On a more granular level, 56 percent of the bilateral PTAs account for at least two of the three principles, compared to only 4 percent of the multilateral PTAs; and sectoral competition is covered in almost 70 percent of the bilateral agreements but fewer than a quarter of multilateral agreements. The greater number of competition provisions in bilateral PTAs may be explained by close coordination between the two negotiating parties, whereas such coordination decreases when more parties are involved in a negotiation (Table 17.1). This research showed that of the 40 percent of PTAs with specific anticompetitive provisions, most tend to focus on anticompetitive conduct rather than on merger control or state-related market distortions. Nearly three-quarters of the PTAs surveyed include provisions targeting at 14 Teh 2009. 510 Competition Policy Table 17.1: General competition provisions embedded in PTAs signed by 2009 Total Bilateral Multilateral Multilateral Multilateral agreements agreements agreements only group with an individual Total agreements by 2009 100.0% 31.1% 68.9% 29.7% 39.2% General objectives of RTA 45.9% 52.2% 43.1% 36.4% 48.3% Horizontal principles 47.3% 82.6% 31.4% 40.9% 24.1% Sectoral competition provisions 37.8% 69.6% 23.5% 36.4% 13.8% Investment 20.3% 47.8% 7.8% 18.2% 0.0% Government procurement 23.0% 47.8% 11.8% 13.6% 10.3% Intellectual property 14.9% 26.1% 9.8% 22.7% 0.0% Services 17.6% 21.7% 15.7% 18.2% 13.8% Financial services 6.8% 4.3% 7.8% 9.1% 6.9% Telecommunications 27.0% 56.5% 13.7% 18.2% 10.3% Maritime trasnport 8.1% 0.0% 11.8% 13.6% 10.3% Competition policy 74.3% 69.6% 76.5% 54.5% 93.1% Source: World Bank Group 2017, adapted from Teh 2009. Note: Of the 74 PTAs analyzed, 30 percent of the bilateral agreements involve the US and 59 percent of the multilateral agreements involve the European Union. least one of the following six specific competition issues: (a) concerted practices (agreements between competitors); (b) abuse of dominance (unilateral anticompetitive conduct); (c) monopolies; (d) state aid (selective incentives granted to certain market participants); (e) SOEs and undertakings with exclusive rights; and (f) mergers and acquisitions. Of these, concerted practices and abuse of market dominance, by either public or private sector firms, are the most common. Most multilateral agreements refer to at least five of these six topics, but merger control is seldom included. Bilateral agreements are primarily concerned with the regulation of designated monopolies and SOEs (Figure 17.2). Figure 17.2: PTAs and coverage of anticompetitive behavior Source: World Bank Group 2017, adapted from Teh 2009. 511 Handbook of Deep Trade Agreements Multilateral and bilateral PTAs also showed significant differences in terms of sectors covered (investment, government procurement, intellectual property, and services).Almost half the bilateral PTAs address investment and government procurement, around a quarter address intellectual property and services, and 57 percent explicitly cover competition in the telecommunications sector. In contrast, multilateral agreements include more competition provisions related to the services and government procurement sectors (16 and 12 percent, respectively), particularly as they relate to the telecommunications and maritime transport industries (Figure 17.3). Figure 17.3: PTAs and coverage of sectoral competition provisions Source: World Bank Group 2017, adapted from Teh 2009. Other studies since 2009 have continued to concentrate on the competition chapters of PTAs while ignoring their economy-wide and sector-specific competition provisions. For instance, a 2011 study15 examines the competition chapters of a set of “representative” PTAs but does not undertake a detailed examination of the provisions in those chapters. Instead, the study attempts to answer a number of broad questions such as the economic rationale for including competition provisions in PTAs, the benefits and costs of such agreements, whether competition policy in PTAs leads to third- party discrimination and trade diversion, and how to assess the implementation of the agreements. Further, the authors contend that PTAs can address market failures that national competition laws cannot because, in the absence of a multilateral competition framework, competition provisions in PTAs create an incentive for implementing and locking in national competition policy regimes, thereby promoting technical assistance and learning by doing.They argue that competition laws are unlikely to discriminate against third parties or to have any significant trade-diverting effects.They also claim that implementation is likely to be more successful in North-South than in South-South agreements, partly because of the ability of North countries to push South countries to create a more competitive playing field for its (North’s) firms. A more recent study16 is significant for its substantial mapping of some 216 agreements notified to the WTO, and for its identification of distinct approaches to addressing competition- 15 Dawar and Holmes 2011. 16 Laprévote et al. 2015. 512 Competition Policy related issues. The study calls these approaches the European, North America Free Trade Agreement (NAFTA), Oceania, and hybrid approaches. The study also proposes a model competition chapter that could serve as a basis for such chapters in future PTAs. Finally, it proposes setting up a “comprehensive, user-friendly database” to provide policymakers and trade negotiators with necessary information on the competition provisions in PTAs. The World Bank is creating such a database with its Deep Integration project. 17.3. METHODOLOGY This chapter reviews the competition provisions in more than 200 multilateral, regional, and bilateral trade agreements, and captures new dimensions of those provisions, including their existence in sections other than the competition chapters. The study divides these competition provisions into economy-wide and sector-specific obligations, and assesses the level of enforceability of both types of provisions. 17.3.1 Economy-wide obligations Economy-wide obligations are classified around four analytical categories: general objectives, horizontal principles, competition policy, and general exceptions. (a) The general objectives category captures the specific inclusion of competition as an overall objective of the trade agreement.The objective of increasing competition is given equal weight with increasing trade and opening markets. (b) Horizontal principles accounts for commitments that inform or complement competition policy, most importantly, transparency, non-discrimination and procedural fairness. (c) Competition policy captures: - the objectives of the competition chapter itself, as opposed to the general objectives of the treaty; - the existence of other treaties that may apply to the parties and result in multi-layered obligations; - the existence of a competition law and an enforcing body. Most PTAs limit the application of the competition law and the scope of the enforcing body, as well as potential exclusions for operators such as state-owned enterprises or designated monopolies; - cooperation obligations, including obligations for coordination, exchange of information, notifications, and technical assistance in the areas of competition; - procedural fairness obligations, which typically strengthen antitrust enforcement and the institutional framework to combat anticompetitive conduct throughout all sectors of the economy; - other areas: · regulated anticompetitive behavior (collusion, abuse, mergers, state aid); · unfair commercial practices; · consumer protection; · enforceability of competition policy provisions by domestic bodies; · provisions subject to dispute settlement (DS); · direct applicability of provisions; (d) general exceptions cover exceptions to the agreement as well as specific exceptions related to national security and the competition commitments. 513 Handbook of Deep Trade Agreements 17.3.2 Sector-specific obligations Sector-specific obligations follow the classification of the WBG Markets and Competition Policy Assessment Tool (MCPAT) to identify prohibitions against rules that (a) reinforce dominance or limit entry; (b) are conducive to collusive outcomes or increase the cost of competing in the market; and (c) discriminate or protect vested interests. The study captures these prohibitions against anticompetitive behavior in the investment, agriculture, electronic commerce, government procurement, and intellectual property sectors,17 and classifies them according to their effects. (a) Prohibitions against rules that reinforce dominance or limit entry encompass: - Monopoly or exclusive rights; absolute bans on entry; arbitrary refusals to grant concessions, licenses, or permits to enter a market; - Relative bans on entry or expansion of activities in order to limit the number of market players; numerical quotas for foreign providers covered by the PTA; - Rules favoring or protecting incumbent firms; - Excessive requirements for registry; - Impediments to customers switching suppliers. (b) Prohibitions against rules conducive to collusion or that increase the cost of competing include: - Facilitating agreements among competitors; self-regulation practices that determine entry, exit, pricing conditions, or industry standards; - Unreasonable restrictions, on the locations and commercial activities of new businesses; - Price controls by government authorities. (c) Prohibitions against discrimination or the protection of vested interests encompass: - Discriminatory application of rules regarding entry, exit, pricing, or marketing conditions; - Preferential treatment of certain operators, resulting in an uneven playing field; - State aid or incentives conferred on a selective basis to certain market players. 17.3.3 Enforceability The competition provisions in PTAs have different levels of enforceability.18 These levels are (a) non-binding; (b) best effort; (c) binding with no dispute settlement mechanism; (d) binding with state-to-state DS; (e) binding with private-state DS; and (f) binding with both state-state and private-state DS. The weighted enforceability level (WEL) of a provision is the average of scores given to each of these enforceability categories: 0 = non-binding, 1 = best effort, 2 = binding with no dispute settlement mechanism, 3 = binding with state-state dispute settlement, 4 = binding with private-state dispute settlement, and 5 = binding with both private-state and state-state dispute settlement. 17 Competition obligations in the service sectors are covered in a separate chapter of this volume. 18 The analysis maps the level of enforceability of both economy-wide and sector-specific commitments based on the approach developed in Horn, Mavroidiss, and Sapir 2010 and Hofmann, Osnago, and Ruta 2017. 514 Competition Policy 17.4. THE GROWING ROLE OF COMPETITION IN PTAs The results of this analysis confirm the critical role of competition commitments as a tool to further trade objectives. More than four-fifths of the PTAs studied19 (239 of the 285 in the sample) have competition-related provisions, defined as any kind of national or regional competition requirement, whether regulatory or institutional.20 PTAs with competition-related provisions are not a recent phenomenon; they have been in existence since the 1950s and have increased in proportion to the increase in the number of PTAs over the decades (Figure 17.4). Most of the 238 in-force PTAs with Figure 17.4: Evolution of in-force PTAs with competition-related provisions, 1958-2016 Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Figure 17.5: Evolution of in-force PTAs with competition-related provisions, by type of PTA Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 19 The sample includes one multilateral agreement (Trans-Pacific Partnership, TPP) that was in negotiation at the date of the analysis. All others in the sample have been notified to the WTO and were in force at the date of the analysis. Of these, 59 percent are bilateral and 41 percent are multilateral agreements. 20 There are three types of national competition requirements in PTAs: (a) the country must have or establish a competition law framework applicable to all economic agents sectors and exclusions; (b) the country must have a functional and independent competition authority; or (c) the country must have a broader competition law framework with principles that apply across the economy. 515 Handbook of Deep Trade Agreements competition-related obligations are bilateral (Figure 17.5), followed by agreements between a group of countries (plurilateral/multilateral) and a single country (e.g., EU-Mexico), and to a lesser extent among countries in plurilateral agreements (e.g., the European Free Trade Area, EFTA). Most signatories to PTAs with competition-related provisions are high- or middle-income countries (Figure 17.6). In the case of bilateral PTAs with such provisions, slightly more than 40 percent are between middle-income countries; another 40 percent are between middle- and high-income countries; and about 20 percent are between high-income countries. The same pattern obtains for multilateral agreements; more than 70 percent of PTAs with competition-related provisions are arrangements made up of middle-income countries.Another quarter involve high-income countries. The bulk of bilateral PTAs with competition-related provisions are in East Asia and the Pacific (EAP), Europe and Central Asia (ECA), and Latin America and the Caribbean (LAC), which together account for 95 percent of all bilateral PTAs. Countries from EAP and LAC are more likely than countries from other regions to form PTAs with partners from other parts of the world (Table 17.2). Figure 17.6: Bilateral PTAs with competition-related provisions, by income level Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Low-income countries, by contrast, show a lack of systematic participation in PTAs with competition-related provisions. This might represent a lost opportunity for those countries, given that competition principles and rules are a critical complement to trade liberalization. This trend may be related to the perceived limited additionality of core competition provisions in PTAs (e.g., enacting a competition law or creating a competition authority)particularly if the country already has a competition framework. Even so, countries that include these types of commitments in their PTAs can expect to build and improve upon their existing competition frameworks,21 which tend to be weaker in lower-income economies.22 21 The EC-Albania PTA of 2006, for example, provides for the establishment of a functional independent competition authority (Article 71.4). Albania had established a competition authority in 2004 (Law No. 9121 of 2003), but recommendations by the EU following signing of the 2006 PTA led to an increase in the budget and authority of the competition authority. http://www.caa.gov.al/uploads/publications/POLITIKA_eng%5B1%5D.pdf. 22 Silva 2004. 516 Competition Policy Table 17.2: Bilateral PTAs with competition-related provisions, by geographical region EAP ECA LAC MENA NA SA SSA Total East Asia & Pacific EAP 28 4 14 1 4 6 0 57 Europe & Central Asia ECA 38 0 7 0 0 0 45 Latin America & Caribbean LAC 18 1 10 1 0 30 Middle East & North Africa MENA 0 4 0 0 4 North America NA 0 0 0 0 South Asia SA 3 1 4 Sub-Saharan Africa SSA 0 0 Total 28 42 32 9 18 10 1 140 Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. In the absence of an international standard for competition-related obligations in trade agreements, these obligations vary by agreement.23 The European Union, for instance, advanced its goal of regional integration by promoting the establishment and enforcement of national competition policies that were in harmony with the EU’s overall objective of reducing market distortions toward a more integrated economy. Following the European example, the Caribbean Community and Common Market (CARICOM), the Eurasian Economic Union (EAEU), the West African Economic and Monetary Union (WAEMU), and the Common Market for Eastern and Southern Africa (COMESA) incorporate more or less comprehensive competition frameworks as a tool for economic integration. The competition provisions included in these agreements serve the dual purpose of creating national and supranational competition legal/institutional frameworks, while harmonizing sector regulations and eliminating technical barriers to entry. Building on these examples, the competition-related obligations included in PTAs can be classified around four conceptual blocks: the three pillars that form a comprehensive competition policy (pro-competition regulations and interventions, competitive neutrality and non-distortive state aid, and effective competition law and antitrust enforcement, as shown in Figure 17.1), along with the competition principles embedded in the general reasoning or goals of the parties. To account for how different PTAs incorporate these dimensions, the analysis is divided into four sections: Competition principles embedded in the general framework of PTAs, Competition law and policy, Competitive neutrality, and Pro-competitive economic regulation. 17.4.1 Competition principles embedded in the general framework of PTAs While the primary objective of trade agreements is to promote or expand trade, a surprisingly large number of PTAs also have the promotion of competition as a major goal. Out of the 23 Matsushita 2004. Even in the WTO agreements, the treatment of competition policy is scattered and not comprehensive. The General Agreement on Trade in Services (GATS), and the agreements on Trade-Related Intellectual Property Rights (TRIPS), Trade-Related Investment Measures (TRIMs), Anti-Dumping, Technical Barriers to Trade (TBT), and Safeguards, all have different approaches to competition obligations. 517 Handbook of Deep Trade Agreements 285 agreements analyzed for this study, 132 (46.3 percent) include among their objectives the promotion of (a) competition between the parties, and (b) cooperation on competition-related issues, as well as (c) open markets, both sector specific and economy wide.24 Bilateral agreements tend to include these objectives more often.25 Typically, the language of these provisions includes broad statements acknowledging the potential of anticompetitive practices to restrict trade and investment, and the importance of cooperating on competition enforcement.26 Specific objectives related to competition are included in 69.5 percent of PTAs.27 Most of these are included along with broader competition commitments that emphasize the role of the agreement in preventing gains in market access from being eroded by anticompetitive behavior.28 To a lesser extent, PTAs also include the objective of promoting consumer welfare or economic efficiency. For instance, CARICOM expressly refers to enhancing economic efficiency and protecting consumer welfare and consumers’ interests.29 The broad and specific commitments are often linked: PTAs that include competition as a general objective tend to deepen that goal through competition-specific objectives. Of the 238 PTAs that include the general objective of promoting competition conditions, 188 also include the specific objectives of promoting fair competition and curbing anticompetitive practices. Further, 32.5 percent of these 188 PTAs also include the objective of promoting consumer welfare or economic efficiency (Figure 17.7). Finally, most agreements with competition-related commitments include the horizontal principles of transparency, non-discrimination, and procedural fairness.These principles, which are critical for promoting competition by ensuring the even-handed treatment of economic actors, are often found in the administrative, institutional, or final provisions of the agreement rather than the competition chapter.30 For example, Chapter 16 of the US-Republic of Korea (KORUS) PTA on competition-related matters contains specific obligations focusing both on 24 PTAs including only one or two of these objectives are less prevalent: 58 agreements include the first and third broad objectives (promote competition and open markets), and only 13 only include the first objective (promote the conditions for competition). 25 Out of all 163 bilateral agreements, 47.2 percent (77 PTAs) include all three broad objectives, while out of the 122 multilateral agreements, 45.1 percent (55 PTAs) include all three objectives. 26 See, for example, Chile-Australia, Article 14.2; Canada-Colombia, Article 1301. 27 There are 198 PTAs that cover at least one of the two specific objectives related to competition, and 61 PTAs that cover both objectives: promoting fair competition and curbing anti-competitive practices, and promoting consumer welfare or economic efficiency. 28 For example, the PTAs signed between the European Union or EFTA and individual countries; or all but one of the PTAs signed with Japan (Japan-ASEAN is the exception). 29 CARICOM Article 30. 30 Of the 239 PTAs with competition provisions, 90.1 percent include at least one of these horizontal principles, and 30 percent include all three horizontal principles. 518 Competition Policy transparency (the obligation to share all public information on competition law enforcement activities) and on procedural fairness (the obligation to provide final decisions on violations of competition law in writing, including findings of fact and the legal reasoning on which the decision is based).31 Figure 17.7: General objectives and horizontal principles of PTAs, including - or not - competition provisions Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 17.4.2 PTA obligations to foster effective competition law and antitrust enforcement 17.4.2.1 National competition requirements Among the most significant competition-related commitments of PTAs are the obligation to establish or maintain competition laws and to create an institution to enforce them. Out of the 239 PTAs with competition provisions, 123 require each of the parties to adopt a competition law, and 74 require the parties to create a competition authority (Figure 17.8). These commitments are telling indicators of the agreement’s vocation to curtail domestic anticompetitive behavior. A third type of national commitment - the establishment of a competition policy or principles - is observed in fewer PTAs.32 Many PTAs that require a national competition law offer little guidance as to its content, so its commitments might be somewhat shallow. The EU-Canada (CETA) agreement, for example, says only that rules tackling anticompetitive conduct “shall be consistent with the principles 31 Article 16, sections 5.1.and 5.3, respectively. 32 Seventy-two PTAs require both the adoption of a competition law and establishment of a competition authority, while only 33 require all three commitments—competition law, authority and policy. 519 Handbook of Deep Trade Agreements Figure 17.8: Distribution of PTAs by specific stipulation within each competition national requirement Sources: WBG dataset on competition policy provisions in PTAs; WTO Regional Trade Agreements Information System. of transparency, non-discrimination and procedural fairness, exclusions from the application of competition law shall be transparent, and each Party shall make available to the other Party public information concerning such exclusions provided under its competition laws.”33 To determine the effectiveness of these commitments, therefore, it is important to look at the scope of the law and at the independence, functionality, and resources of the competition authority. The data show that a significant number (43.2 percent) of the treaties that include a competition law require that the law applies to all sectors and economic agents (Figure 17.9). These obligations are often weakened, however, through the ability of parties to introduce or maintain exclusions as long as they are transparent, non-discriminatory and/or taken on the grounds of public policy or public interest (e.g., Chile-Australia, Canada-Colombia). Among the agreements with exclusions, six apply to specific economic activities or sectors. The Trans- Pacific Strategic Economic Partnership (TPSEP), for example, excludes export arrangements as well as the meat industry for New Zealand, and also expressly excludes certain operators such as agricultural producer boards and Pharmac, an SOE. The TPSEP also excludes certain services, including some postal services and types of public transport, for Singapore.34 The East African Community (EAC) excludes acts by consumers, collective industrial bargaining, and “specific sectors or industries to the extent that the anticompetitive conduct is required by such regulation within their own Jurisdictions.” CARICOM excludes collective bargaining and approved professional standards. Some association agreements with the EU and EFTA (e.g., EC-North Macedonia) allow a transition period for full application of the competition law across activities and sectors. Finally, the Eurasian Economic Union (EAEU) establishes specific rules on the application of antitrust and the regulation of natural monopolies. 33 See CETA, Articles 17.1 and 17.2.4, available at: http://ec.europa.eu/trade/policy/in-focus/ceta/ceta-chapter- by-chapter/.s2xvscv1. 34 Annex 9.A of the TPSEP agreement lists exemptions from the application of Article 9.2 on Competition Law and Enforcement.The exclusions apply to a wide variety of commercial activities in New Zealand and Singapore. 520 Competition Policy Figure 17.9: Distribution of PTAs with requirement of competition law, by features of the law Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: The numbers displayed in the figure correspond to the number of PTAs with such specific requirements. The figure shows all combinations observed in the PTAs. In terms of institutional obligations, 64 percent of the PTAs that require a competition authority give more attention to the need for a functional institution than to independence or sufficient resources. Nine PTAs require only functionality,35 while one (Canada-Costa Rica) requires only independence. None of the PTAs requires only a sufficiently resourced competition agency.Three PTAs, all involving the East African Community (EAC), require both functionality and sufficient resources (Figure 17.10), but none of the three requires an independent competition authority. The 14 PTAs that include both independence and functionality, or all three characteristics, are association agreements with the EU. Figure 17.10: Distribution of PTAs requiring a competition authority, and which features they require Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: The numbers displayed in the figure correspond to the number of PTAs with the specific requirement.The figure shows all combinations observed in the PTAs. 35 These agreements are the EAEU treaty; PTAs of the EAEU with Armenia and with the Kyrgyz Republic; the European Union’s PTAs with Central America, the overseas territories, and Georgia; the European Economic Area (EEA) PTA; and the agreements signed by Korea with Turkey and Australia. 521 Handbook of Deep Trade Agreements The third aspect of national competition commitments relates to the promotion of competition and the principles that guide the implementation of competition law. Competition policy provisions range from a broad acknowledgment of the importance of undistorted competition for trade relationships (e.g., EU-Moldova) to more concrete obligations to promote competition under the principles of transparency, enhancement of economic efficiency, and/or cooperation (e.g., Hong Kong SAR, China-Chile, Canada-Honduras). Further, some PTAs (e.g., Andean Community) expressly include the obligation to apply competition policy principles to the adoption and application of market policies and regulatory measures (Figure 17.11). Figure 17.11: Distribution of PTAs with requirement for competition policy/principles, by features of the policy Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: The numbers displayed in the figure correspond at the number of PTAs with the specific requirement.The figure shows all combinations observed in the PTAs. It is notable that competition requirements are more common in multiparty PTAs than in bilateral agreements. Multilateral PTAs also have a higher level of enforceability than bilateral PTAs; more than 40 percent require state-to-state dispute settlement and a significant share also require private dispute settlement. By contrast, more than 90 percent of bilateral agreements require the agreement to be binding but do not have dispute settlement. 17.4.2.2 Substantive provisions regulating competition policies The requirement to have a competition law and/or an enforcing institution does not necessary imply that the PTA takes a position on how competition should be regulated or what situations the national law should cover. Therefore, to identify how PTAs tackle anticompetitive behavior through substantive commitments, the analysis differentiates between (a) those agreements that refer to other international instruments; and (b) those that include substantive provisions within the text of the PTA itself.36 Of the 126 PTAs covering at least one of the national competition requirements, 114 include substantive commitments; while 44 out of those 126 PTAs refer to 36 Substantive commitments are defined as those that (a) prohibit or regulate cartels or concerted practices; (b) prohibit or regulate abuse of market dominance; (c) regulate undertakings with exclusive rights; (d) regulate monopolies; (e) regulate anti-competitive behavior of SOEs; (f) regulate state aid; and (g) regulate mergers and acquisitions. 522 Competition Policy other instruments.37 Interestingly, 19 PTAs which lack any national competition requirement refer nonetheless to other instruments that have competition commitments.38 17.4.2.2.1 PTAs connected to other treaties regulating competition Other competition provisions that have not drawn the attention of earlier studies are those that either (a) make reference to the competition policy provisions in other international agreements (such as the GATT); or (b) acknowledge the competition obligations of other plurilateral and bilateral agreements (TFEU; Commonwealth of Independent States Free Trade Area, CISFTA); or (c) call for the enactment of a separate set of regional competition rules such as those adopted in the framework of ECOWAS (Figure 17.12). References to other international agreements are often used to (a) add another level of commitment to the competition-related provisions of the PTA; (b) further clarify the PTA obligations; or, in some cases, (c) fill a transitory void of competition rules. In the latter case, these international agreements can be sources of rules or disciplines on competition while parties to the PTA are in the process of implementing their competition commitments.39  Figure 17.12: References to other international agreements in PTAs, by type of PTA Sources: WBG dataset on competition policy provisions in PTAs; WTO Regional Trade Agreements Information System. 37 Of these 44 PTAs, 43 have at least one national competition requirement, at least mention another instrument, and at least include a substantive provision (those that regulate anticompetitive behavior); and only one PTA (Japan-Indonesia) does not include substantive provisions. 38 Of these 19 PTAs, 3 do not cover any substantive anti-competitive provisions, while the other 16 do not have a national competition requirement per se, but nevertheless include substantive provisions and refer to other international instruments. 39 For example, in the case of the agreement between Ukraine and Tajikistan, the PTA includes provisions laying the foundation for the application of other regional competition instruments; see Article 6. This provision is complemented by Article 17(a) of the CISFTA. 523 Handbook of Deep Trade Agreements Of the 63 PTAs that refer to international instruments, 62 percent involve a European multilateral party.40 In agreements involving the EU, competition articles either directly reference (EU-Moldova) or mirror (EC-Faroe Islands, EU-Norway) the text of EU agreement’s competition rules. In the case of the Common Market of South America (MERCOSUR), a protocol to the PTA deals with harmonization of competition policy within the region, as well as with the attributes of MERCOSUR’s Commerce Commission and Committee for the Defense of Competition. In the Economic Community of West African States (ECOWAS), regional competition rules were enacted based on the provisions of the earlier regional treaty for harmonization and coordination of national policies in trade, as well as measures for maintaining and enhancing economic stability in the region.41 The treaty for the East African Community included a framework for the future enactment of regional competition rules and harmonization of national rules and policies, which led to the enactment of the East African Community Competition Act.42 17.4.2.2.2 PTAs regulating substantive antitrust obligations and merger control A significant number of the PTAs with competition provisions include substantive commitments related to antitrust enforcement and merger control. Out of the 138 PTAs with core antitrust provisions, 113 cover cartels and abuse of dominance, 28 include only merger control provisions, and 24 PTAs cover all three types of anticompetitive conduct. Cartels, abuse of dominance, and anticompetitive mergers are typically prohibited as business or commercial conduct incompatible with the purposes of the PTA and tend to be bundled together.43 Anticompetitive practices are prohibited in 20 multilateral instruments, which include mainly EU/EC enlargement agreements, regional agreements (COMESA), and country accession instruments to regional PTAs (EAEU Accession-Armenia). They are also prohibited in four bilateral agreements (Singapore-Taiwan, China, Singapore-Costa Rica, Singapore-Korea, and Canada-Costa Rica). Multilateral agreements are typically characterized by a leading economic party that incorporates essential aspects of its competition regulatory framework into the agreement. For example, agreements subscribed by the EU tend to replicate, in general terms, definitions of anticompetitive conduct included in the Treaty on the Functioning of the European Union (TFEU), particularly 40 This number includes PTAs with regional competition instruments that directly bind the parties to the agreement, as well as other bilateral and multilateral agreements (such as GATT) that complement or clarify the commitments of the agreement. 41 The ECOWAS treaty adopts regional competition rules, and provides for the harmonization and coordination of national trade policies as a means of maintaining and enhancing economic stability in the region. 42 EAC also establishes a customs union and lists competition as one of the aspects to be included in the customs union protocol on monetary and fiscal policy harmonization. 43 Some PTAs, such as EFTA-Mexico, list anti-competitive conduct as incompatible with the objectives of the agreement without providing definitions. 524 Competition Policy when dealing with less-developed competition law jurisdictions. Agreements between jurisdictions which have a similar degree of development will strive to achieve convergence in specific areas of enforcement or recognize each other’s definitions; for example, the mutual recognition of competition definitions and the application of competition policy to enterprises in the CETA PTA. The higher enforceability of multilateral agreements is even more pronounced in the case of merger control obligations, while bilateral PTAs tend to put more emphasis on enforceability of provisions regarding dominance, collusion, and cartels (Figure 17.13). Figure 17.13: Regulated anticompetitive behavior in PTAs, by type of PTA Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Provisions against collusion generally consist of a standard prohibition against concerted practices that have the object or the effect of preventing competition.44 Abuses of dominance are either prohibited through general effects-based clauses (Georgia-Ukraine, Kyrgyz Republic-Armenia) or by providing specific examples of abusive conduct (CARICOM, EEA,Thailand-Australia). Merger control provisions are present in agreements among jurisdictions with well- established competition frameworks, as in the EU Proposal for the Transatlantic Trade and Investment Partnership (TTIP) between the EU and the US.45 Merger control provisions are also present in agreements aimed at creating supranational instruments to complement or clarify existing competition provisions (e.g., COMESA). Merger control obligations vary 44 PTAs with object and effect provisions include Turkey-North Macedonia, EU-Central America, and the CIS. PTAs with provisions on competition in entrepreneurial activity include Armenia-Turkmenistan. 45 The TTIP contains a general description of the EU treatment of anti-competitive conduct, including (a) horizontal and vertical agreements which have as their object or effect the prevention, restriction, or distortion of competition; (b) abuses of dominance by one or more enterprises; and (c) concentrations between enterprises which significantly impede effective competition, in particular as a result of the creation or strengthening of a dominant position. 525 Handbook of Deep Trade Agreements from a short reference to anticompetitive mergers in a list of anticompetitive behaviors (Costa Rica-Singapore, Singapore-Australia) to more detailed language linking the need for effective merger control with the prevention of abuse of dominance (EU-Georgia, EU- Republic of Moldova). 17.4.2.3 Procedural fairness commitments to support competition enforcement Instead of, or in addition to, focusing on substantive competition-related commitments, some PTAs emphasize commitments on procedural fairness, transparency, and collaboration among authorities, as minimum common denominators for a workable competition policy framework. The value of this approach is confirmed by the significant efforts of the International Competition Network (ICN) to embed and promote procedural fairness in antitrust investigations.46 Effective cooperation between two competition authorities within the framework of an international antitrust investigation would not be possible without minimum standards that ensure similar treatment of procedural parties or confidential information.47 The proposed Trans-Pacific Partnership (TPP) puts particular emphasis on procedural fairness, as do KORUS and the EU proposal for TTIP. There are 80 PTAs that include at least one procedural fairness provision, of which 55 percent are bilateral PTAs. Only 13 PTAs, all multilateral (involving the EU, EAEU, and COMESA), cover all three types of procedural fairness provisions (collusion, abuse of dominant position, and merger control). The enforceability of these provisions is significantly higher in PTAs with multilateral signatories. There are 20 PTAs that include at least one procedural fairness provision and also cover the three core substantive conducts,48 while only four PTAs with no substantive competition provisions include procedural fairness obligations (Peru-Singapore, Peru-Japan, Korea-New Zealand, Chile-Hong Kong SAR, China). The most frequent commitments are, in descending order, the right to defense, transparency, the right counsel, and predictability in proceedings (Figures 17.14 and 17.15). For obligations supporting the right of defense, the critical commitment seems to be the protection of confidential information, with less importance given to other core aspects of procedural fairness, such as ensuring that parties can be represented by counsel (Figure 17.15). 46 See the introductory statement of the ICN Guidance on Investigative Process, available at http://www. internationalcompetitionnetwork.org/uploads/library/doc1028.pdf. 47 On the importance of setting common rules on the exchange of confidential information to foster cooperation among antitrust enforcement agencies, see OECD (2014), Recommendation of the OECD Council Concerning International Co-operation on Competition Investigations and Proceedings, available at http://www.oecd.org/ daf/competition/2014-rec-internat-coop-competition.pdf. 48 These include 11 PTAs with the EU, 3 with the East African Community, 3 with EAEU, 1 with COMESA, and the PTAs of Costa Rica-Canada and of Singapore-Taiwan, China. 526 Competition Policy Figure 17.14: Procedural fairness in PTAs, by type of PTA Sources: WBG dataset on competition policy provisions in PTAs; WTO Regional Trade Agreements Information System. Figure 17.15: Distribution of PTAs with provisions that promote the right of defense, by the scope of such right Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: The numbers displayed in the figure correspond to the number of PTAs with such specific requirements. The figure shows all combinations observed in the PTAs. Transparency provisions tend to emphasize consultation among competition authorities, or the obligation to inform the parties of competition concerns or decisions (Peru- Singapore, Korea-Vietnam, Japan-Peru). Some PTAs also include obligations to avoid enforcement conflicts (Japan-Mongolia). However, the obligation to publish decisions, a central element of ensuring transparency, remains less common and tends to be found in combination with other transparency commitments (Figure 17.16). This is the case with the Chile-Australia and Korea-Vietnam PTAs as well as a number of association agreements with the EU. 527 Handbook of Deep Trade Agreements Figure 17.16: Distribution of PTAs with provisions that promote transparency, by specific obligations Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: The numbers displayed in the figure correspond at the number of PTAs with such specific requirements. The figure shows all combinations observed in the PTAs. Among the commitments focusing on enhanced predictability (Figure 17.17), the most common, typically found in a bundle, pertain to conducting procedures within a reasonable period of time and using written procedures to carry investigations, followed by the requirement to establish and maintain settlements or consent agreements. 49 Figure 17.17: Distribution of PTAs with provisions that promote predictability, by specific requirements Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: The numbers displayed in the figure correspond to the number of PTAs with such specific requirements. The figure shows all combinations observed in the PTAs. 49 These include a majority of EU accession agreements. Other agreements that also expressly refer to written procedures are EAEU, CARICOM, and TPP. 528 Competition Policy Among all PTAs reviewed, only the TPP, which has not entered into force, covers all procedural fairness obligations identified in this analysis (Box 17.2). Interestingly this instrument does not focus on promoting substantive convergence among the content of provisions.50 Box 17.2: The Trans-Pacific Partnership: Procedural fairness obligations under the competition chapter > General approach to competition law and policy -TPP parties shall adopt or maintain national competition laws that proscribe anticompetitive business conduct, with the objective of promoting economic efficiency and consumer welfare; and shall take appropriate action with respect to that conduct. -Each party shall endeavor to apply its national competition laws to all commercial activities in its territory. -Each party shall maintain an authority or authorities responsible for the enforcement of its national competition law. -Competition authorities shall act in accordance with the objectives of promoting economic efficiency and consumer welfare, and not discriminate on the basis of nationality. > Procedural fairness obligations - TPP parties shall observe procedural fairness in the enforcement of competition laws. - Each party shall adopt written procedures for conducting investigations, which shall be concluded within a reasonable period of time. - Before receiving a penalty, the party under investigation shall: (a) be informed by the authority of the competition concern; (b) have a reasonable opportunity to be represented by counsel, be heard and present evidence or testimony in their defense. In particular, this includes testimony of experts, cross-examination of testifying witness, and review and rebuttal of any evidence in the proceeding; (c) have the right of appeal, including review of alleged substantive or procedural errors, in a court or other independent tribunal, as well as the option to settle with the competition authority, which settlement can subject to independent or judicial review; - Competition authorities shall: (a) bear the burden of proof for establishing anticompetitive conduct; (b) respect the presumption of innocence in public notices of ongoing investigations; (c) protect confidential information; (d) provide the party under investigation with a reasonable opportunity to view the evidence against him or her; (e) provide the person under investigation with a reasonable opportunity to consult with the competition authority. > Enforceability - From an institutional perspective, the competition chapter of the TPP offers limited means of direct enforcement, since it is explicitly excluded from the mechanisms in the dispute settlement chapter. In addition, unlike other chapters that create ad hoc committees to oversee the implementation of commitments, the competition chapter only establishes the right to enter into consultations at request of another party but it fails to articulate the procedure. 17.4.2.4 Institutional arrangements 17.4.2.4.1 Institutional setup To better understand the potential impact of competition provisions in PTAs, it is important to consider the institutional arrangements under which they are to be implemented. Only 50 Regarding rules that promote predictability, the TPP requires procedural fairness, including the obligation to provide the investigated party with information regarding the competition concerns, a reasonable opportunity to be represented by counsel, and a reasonable opportunity to be heard and present evidence in the party’s defense (Article 16.2). The TPP also requires transparency (Article 16.7) and consultations (Article 16.8). 529 Handbook of Deep Trade Agreements 23 PTAs call for the creation of supranational competition frameworks, and all involve a multilateral party. These types of frameworks are useful for overcoming domestic constraints on institutionalizing competition policy. Figure 17.18: Institutional arrangement in PTAs, by type of PTA Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. The most prominent examples of supranational competition frameworks are regional PTAs. The process of enabling a CARICOM Single Market and Economy (CSME) included, for example, the creation of a regional competition authority, the CARICOM Competition Commission in 2008. Since then, several CARICOM members have begun a gradual process of drafting and implementing national or supranational competition laws51 and setting up competition authorities.The same is true for COMESA, which is actively seeking to deepen collaboration with Member States and change assistance in creating domestic competition institutions.52 Along the same lines, the WAEMU PTA has led to the creation of a regional legal framework and the WAEMU Competition Commission. 51 The Organization of Eastern Caribbean States (OECS) has agreed to establish a supranational competition agency to handle competition matters within its single market (Antigua and Barbuda, Dominica, Granada, Montserrat, Saint Lucia, Saint Kitts and Nevis, SaintVincent and the Grenadines).The Bahamas has a joint utilities and competition regulator, the Utilities Regulation and Competition Authority (URCA), acting under the Regulation and Competition Authority Act of 2009. In Barbados, the new competition authority established in 2001, evolved from the Public Utilities Board, which was responsible for regulating public utilities such as electricity and telephone services; the new authority is responsible for implementing the Fair Competition Act. Belize is in the process of establishing a national competition authority and has drafted a Fair Competition Bill. In Guyana, the Competition and Consumer Affairs Commission (CCAC) was established under the Competition and Fair-Trading Act of 2006, and is charged with administering and enforcing the Competition and Fair-Trading Act of 2006 and the Consumer Affairs Act of 2011. Suriname has drafted a competition bill and undertaken a series of stakeholder consultations to finalize the bill before presenting it to the Parliament.Trinidad and Tobago has established the Fair Trading Commission under its competition law, the Fair-Trading Act of 2006. 52 The COMESA Competition Commission was created in 2004 as part of the COMESA Treaty, and became operational in 2013. The Commission is responsible for investigating and sanctioning cases where the anti- competitive behavior of one member state affects another member state. 530 Competition Policy 17.4.2.4.2 Cooperation among competition authorities Another important element of procedural fairness is the commitment to cooperate. Given the prevalence of multi-jurisdictional anticompetitive behavior, the degree of cooperation among competition authorities will have an important bearing on the efficacy of any investigation or enforcement action. This study analyzes the existence of four types of cooperation: (a) coordination among bodies with a competition mandate (e.g., Japan-Mongolia); (b) exchange of information among bodies with a competition mandate (China-Korea); (c) notifications among bodies with a competition mandate (Trans-Pacific Strategic Economic Partnership,TPSEP); and (d) technical assistance on implementation among bodies with a competition mandate. 53 Of the 239 PTAs with competition provisions, 131 agreements include at least one of these forms of cooperation, and 99 of them explicitly establish cooperation as a general objective. However, only 31 PTAs, mostly multilateral, include all four forms of cooperation. The main forms of cooperation among competition authorities are exchange of information, followed closely by notification, coordination, and technical assistance.The latter has the highest enforceability level, followed by coordination among competition bodies (Figure 17.19). Figure 17.19: Forms of cooperation in PTAs, by type of PTA Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: All the PTAs in the figure have competition provisions. Obligations to coordinate, exchange information, and notify are usually found together, either in general terms (e.g., New Zealand-Singapore) or in the context of communicating and cooperating in enforcement activities that may be of interest to the other party due to cross-border effects (e.g., Korea-Vietnam). 53 Some PTAs that expressly mention technical assistance are Chile-Australia (strengthening systems); Korea- Vietnam (various areas of technical assistance); New Zealand-Taiwan, China (implementation of competition policy); Peru-Korea (exchange of experience and capacity building); Thailand-New Zealand (various areas); and Turkey-Jordan (technical assistance in the understanding of each other’s systems). 531 Handbook of Deep Trade Agreements In the case of technical assistance, most clauses tend to list the scope or types of technical assistance that parties may undertake (e.g., Korea-Vietnam; EU-Columbia-Peru). A few PTAs refer to the possibility of parties entering into additional cooperation and mutual assistance agreements (e.g., Canada-Costa Rica). The obligation to notify (Figure 17.20) is mostly related to investigations, sometimes in combination with the notification of decisions (EC-Mexico). Figure 17.20: Distribution of PTAs with notification requirements, by scope of such notifications Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: The numbers displayed in the figure correspond to the number of PTAs with such specific requirements. The figure shows all combinations observed in the PTAs. 17.4.3 PTA obligations to foster competitive neutrality Any sorts of measures that insulate firms from competition, whether in the form of regulation, state aid, or incentives, can have a significant impact on market outcomes. However, most governments often provide such treatment to SOEs and designated monopolies as well as to targeted private operators.This raises concerns as State assistance to specific firms can distort the normal competitive process, increase the likelihood of anticompetitive practices by enterprises that receive the favored treatment, and deter investments by non-favored enterprises. The risks of such policies grow exponentially when they target SOEs and firms that already enjoy a privileged position vis-à-vis less-connected private operators. If state support targets sectors with low levels of market competition, the effects in the markets are more pervasive. The Uruguay Round agreements initially disciplined SOEs and designated monopolies by submitting them to the non-discrimination obligations prescribed for state measures that affect imports or exports by private traders. This approach, based on the GATT, requires parties to ensure that in purchases or sales involving imports and exports, SOEs and designated monopolies should behave in accordance with commercial considerations, including price, quality, availability, marketability, and transportation. In addition, the SOEs and designated 532 Competition Policy monopolies should afford the other party’s enterprises adequate opportunity to compete for participation in such purchases or sales. This approach has been captured by PTAs that specifically refer to GATT commitments (ASEAN-New Zealand, Canada-Panama, EFTA-Egypt, EFTA-Mexico, EFTA-Ukraine) or follow a similar approach. For example, CARICOM prescribes the elimination of discriminatory measures enacted by SOEs and designated monopolies that limit market access or “distort competition or fair trade.” A comprehensive approach to leveling the playing field regarding SOEs is a rather recent trend in PTAs. The TPP is the first agreement that seeks to comprehensively address the commercial activities of SOEs competing with private companies in international trade and investment. Even though the commitments in the TPP’s competition chapter build on WTO principles and previous US PTAs (notably CETA), the TPP significantly expands the scope of commercial consideration and non-discrimination commitments by advancing the control of distortive public support and subsidies. In other words, the TPP works to promote competitive neutrality54 and non-distortive public aid support.55 More specifically, SOEs and designated monopolies should be bound to compete on quality and price rather than benefiting from discriminatory regulation and distortive subsidies. These obligations build on three main commitments by TPP parties: (a) parties must avoid discrimination and apply commercial considerations to SOEs, and limit the scope for designated monopolies to engage in anticompetitive practices; (b) parties must not provide non-commercial assistance56 capable of causing adverse effects or injury to the interests of another party; (c) and parties must offer an impartial regulatory and institutional framework for SOEs, and make them accountable for their actions in other TPP countries. 54 See OECD 2015, Roundtable on Competition Neutrality, Issues Paper by the Secretariat, referring to competitive neutrality as a principle according to which all enterprises, public or private, domestic or foreign, face the same set of rules, and where government’s contact, ownership, or involvement in the marketplace, in fact or in law, does not confer an undue competitive advantage on any actual or potential market participant. See also Roundtable on Competitive Neutrality, Note by the European Union, stating that while there is no universal definition of competitive neutrality, there are accepted interpretations of this principle. For instance, according to the European Union, competitive neutrality should be “broadly defined and cover all forms of direct and indirect public interventions of whatever nature, which may provide public or private undertakings with undue advantages over their actual or potential competitors, thereby distorting the competitive process.” 55 Other PTAs that provide comprehensive treatment of SOEs are KORUS and the EU Proposal for TTIP, both of which hold that while parties are free to create monopolies and SOEs, their conduct in the market should be carried out in accordance with competition rules and in a non-discriminatory manner. KORUS allows the parties to establish SOEs, subject to the following obligations: Each Party shall ensure that any state enterprise that it establishes or maintains (a) acts in a manner that is not inconsistent with the Party’s obligations under this Agreement wherever such enterprise exercises any regulatory, administrative, or other governmental authority that the Party has delegated to it, such as the power to expropriate, grant licenses, approve commercial transactions, or impose quotas, fees, or other charges; and (b) accords non-discriminatory treatment in the sale of its goods or services to covered investments. Along the same lines, the EU Proposal for TTIP authorizes parties to establish or maintain state enterprises and monopolies and to grant these enterprises special or exclusive rights. However, where an enterprise falls within the scope of this provision, the parties shall not require or encourage such an enterprise to act in a manner that is inconsistent with the PTA, and shall observe the principle of non-discrimination for covered investment. 56 Non-commercial assistance includes (a) direct transfers of funds, or potential direct transfers of funds or liabilities; and (b) goods or services other than general infrastructure on terms more favorable than those commercially available to that enterprise. The TPP excludes non-commercial assistance measures taken prior to the TPP entering into force and/or enacted within three years of it entering into force if based on a decision taken prior to that date. 533 Handbook of Deep Trade Agreements Although less systematized, a number of PTAs include provisions on SOEs to ensure non- discriminatory treatment (Korea-US, NAFTA), the applicability of competition law to SOEs (Korea-Vietnam, CAFTA-Dominican Republic-US, MERCOSUR), and obligations to restrain parties from granting competitive advantages to SOEs (Chile-Australia, Korea- Australia). Some PTAs (Korea-Singapore) go even further and expressly mention competitive neutrality in connection with the treatment of SOEs. The regulation of SOEs is often coupled with the regulation of statutory monopolies and/ or firms with exclusive rights. Most provisions dealing with statutory monopolies also focus on non-discrimination obligations (Turkey-Jordan, Ukraine-North Macedonia) and on the importance of treating statutory monopolies in accordance with the parties’ obligations under the agreement (NAFTA, New Zealand-Taiwan, China, ASEAN-Australia-New Zealand, Panama-El Salvador, Panama-Honduras).This is the same approach adopted toward firms with exclusive rights (New Zealand-China, Pakistan, Malaysia, Ukraine-Montenegro, China-Switzerland). Competitive neutrality also covers state aid provisions, which typically aim at limiting any negative effects of state aid on the economic conditions of the other party (Kyrgyz Republic- Armenia, Russian Federation-Tajikistan). Other commitments include transparency in the granting of state aid (Turkey-Albania,Turkey-Bosnia and Herzegovina), and the requirement that state aid not affect trade between parties (EC accession treaties,TFEU, EEA) or threaten to distort competition (EC-Faroe Islands, Turkey-Montenegro). PTAs with provisions to foster competitive neutrality tend to involve a multilateral party, except that statutory monopolies remain more prominent in bilateral agreements. Multilateral PTAs have higher levels of enforceability, in particular regarding the anticompetitive behavior of SOEs (Figure 17.21). Figure 17.21: Competitive neutrality provisions in PTAs, by type of PTA Sources:WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 534 Competition Policy 17.4.4 PTA obligations to remove sector-specific anticompetitive regulations Even though market regulation is a legitimate and necessary way to accomplish various policy objectives, in some cases it has the potential to negatively affect both trade and competition, harming market dynamics in an unintended and sometimes avoidable manner. Sector-specific obligations in trade agreements have typically focused on minimizing anticompetitive restrictions as an instrument to embed competition principles in markets open to trade and investment. Obligations for key industries are included to reduce technical barriers to trade (TBTs) and promote entry. However, this research also considers these sector-specific commitments from a competition angle. The sectors examined are agriculture, e-commerce, government procurement, intellectual property, and investments. Competition-related commitments affecting services are dealt with in separate chapters. All PTAs in the database consider competition commitments in at least one of these five sectors; 72 percent (206 agreements) include specific commitments in all five sectors, of which 96 percent (197 agreements) also include economy-wide competition policy provisions. Bilateral agreements include more sectoral-specific commitments than do multilateral PTAs, and also include more commitments in all five sectors—77 percent, compared to 66 percent of multilateral PTAs. More bilateral PTAs account for sectoral commitments specific to agriculture, followed by business investment, intellectual property, government procurement, and e-commerce. Multilateral PTAs follow the same order of coverage. Figure 17.22: PTAs and coverage of sector-specific competition provisions by type of PTA Sources:WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 535 Handbook of Deep Trade Agreements These obligations closely follow the Market and Competition Policy AssessmentTool (MCPAT) to identify rules that (a) reinforce dominance or limit entry - absolute entry restrictions, incumbents’ involvement in entry decisions; (b) facilitate collusive outcomes - regulations facilitating price fixing, self-regulation, or information exchange; (c) discriminate and protect vested interests - explicit discriminatory rules without justification, selective subsidies and incentives which distort the level playing field, and an explicit lack of competitive neutrality (Table 17.3). Table 17.3: PTA distribution according to provisions against anticompetitive behavior identified by MCPAT PTA that include provisions against specific rules as a share of total PTA with sectoral commitment (percent) Number of PTAs with Rules that Rules that are Rules that At least one of Sector sectoral reinforce conducive to discriminate or these rules commitments dominance collusive outcomes protect vested or limit costs to compete interests entry in the market Investment 250 81.2 10.8 66.4 86.8 Agriculture 285 92.6 4.2 93.3 99.3 e-commerce 213 30.5 0.9 22.5 33.3 Government procurement 226 53.5 4.4 57.5 69.5 Intellectual property 240 66.7 2.5 50.0 77.1 All five sectors included 206 21.8 0.5 8.7 22.8 Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. PTAs with sector-specific commitments tend to focus more on rules that reinforce dominance or limit entry, followed by rules that discriminate or protect vested interests. Except for investment-related commitments, few of the sectoral commitments in PTAs prohibit or limit rules that are conducive to collusive outcomes or increase costs to compete in the market. Many of the PTAs that include sectoral commitments also have economy-wide competition provisions. However, there are 46 PTAS with sectoral commitments that do not also have economy-wide competition provisions. These are mostly regional agreements, including the Pacific Alliance, the ASEAN Free Trade Area, and the Asia-Pacific Trade Agreement (APTA); agreements between regional blocs and specific countries such as the EU with Andorra, Côte d’Ivoire, or the Syrian Arab Republic; and bilateral agreements such as Chile-Colombia, Chile-Turkey, Chile-China, the US-Bahrain, and the US-Israel. 17.4.4.1 Prohibitions on rules that reinforce dominance or limit entry Rules that reinforce dominance or limit entry are undesirable because dominant market players can set prices above the competitive level, produce lower-quality products, or reduce the rate of innovation below what would exist in a competitive market. Provisions in the sectoral chapters that prohibit or try to limit these effects can be classified as those 536 Competition Policy that consider (a) monopoly rights and absolute bans on entry; (b) relative bans on entry or on expansion of activities; (c) incumbents’ rights regarding entry decisions; (d) registry requirements, including licenses and permits; and (e) impediments to switching providers. In general, sector-specific commitments against rules that reinforce dominance or limit entry have been included in PTAs since the 1970s, but it was not until the 1990s (Figure 17.23) that they became more prominent, especially in the agriculture and investment chapters (Figure 17.24). Figure 17.23: Evolution of PTAs in force with sector-specific provisions against rules that reinforce dominance or limit entry, by sector Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Figure 17.24: Sector-specific provisions against rules that reinforce dominance or limit entry Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 17.4.4.1.1 Provisions that limit or prohibit monopoly rights and absolute bans on entry Monopoly rights and absolute bans on entry occur when a regulatory instrument either limits or enables another body to limit the number of concessions, licenses, or permits to enter a market; when certain providers are prohibited from entering without justification; 537 Handbook of Deep Trade Agreements when regulation forbids the granting of permits or authorizations to enter a given market; or allows for the granting of exclusive rights. In investment chapters, competition provisions include prohibitions on establishment or residence obligations for service suppliers (Japan-Australia), as well as obligations to allow all transfers related to a covered investment to take place freely and without delay (Chile- Australia, Japan-Singapore). In agriculture, provisions include obligations to eliminate monopoly import arrangements (Australia-New Zealand). In the case of e-commerce, some general provisions are included in those chapters which provide for the elimination of barriers to e-commerce and the creation of an environment of trust and confidence for its use (Japan-Switzerland). Finally, in public procurement, provisions tend to focus on preventing those requirements that would restrict a public tender only to certain participants by, for example, requiring unnecessary technical specifications or specific trademarks (Chile-Australia). 17.4.4.1.2 Provisions that limit or prohibit relative bans on entry or expansion of activities Relative bans on entry and expansion of activities are a less restrictive or more indirect way of limiting entry, and generally occur when regulation enables another body to set the number of permits, licenses, or fees to enter the market. Potential negative effects from these types of restrictions may include lower incentives to enter the market, lower quality or higher prices for consumers, slower sector development, and less innovation in the market. Provisions against relative bans on entry and expansion of activities are the most common in all five sectors. In the investment chapter they typically relate to freedom of movement and establishment of persons (Canada-Honduras, Australia-China), free movement of capital (EFTA-Egypt, EC-Mexico), and creating a predictable framework for investment (Turkey- West Bank and Gaza. COMESA, ASEAN-China, EFTA-Mexico). In the agriculture chapter, clauses falling into this category are focused on prohibiting restrictions on import/ export (Korea-US, Panama-El Salvador), ensuring that sanitary and phytosanitary measures do not constitute a means of arbitrary or unjustifiable discrimination (Taiwan, China- Singapore, Turkey-Serbia) and are duly notified to promote transparency (Panama-Taiwan, China, Singapore-Australia). Similarly, in the case of e-commerce, provisions are aimed at achieving transparency regarding regulation and measures affecting e-commerce (ASEAN- Australia-New Zealand), adopting internationally approved standards for the industry (EC-CARIFORUM, EU-Georgia), and supporting cooperation to promote market access and reduce regulatory barriers (CEFTA). In the case of government procurement, these types of measures seek to achieve a better understanding of their respective government procurement systems (Singapore-Australia, Thailand-Australia, EFTA-Central America) and reduce barriers to procurement markets (Thailand-New Zealand). Finally, measures included in the intellectual property chapters focus on ensuring an adequate and effective 538 Competition Policy protection of intellectual, industrial, and commercial property rights, clear rules governing these industries (Korea-US, Nicaragua-Taiwan, China); and effective means to enforce those rights (EC-Mexico, Malaysia-Australia). 17.4.4.1.3 Provisions that limit or prohibit incumbents from participating in entry decisions These provisions are aimed at limiting measures that allow incumbents to participate in entry decisions by providing for equal and transparent conditions to access the market. Most of the provisions in this category are linked to government procurement and include prohibitions on offsets (EFTA-Mexico, Dominican Republic-Central America, EFTA-Ukraine), as well as clauses promoting open and transparent procurement procedures (Colombia-Korea, EU- Colombia-Peru) and clear, non-exclusionary technical specifications (Dominican Republic- Central America, Korea-New Zealand). Some agriculture chapters contain provisions ensuring that fees and charges related to import/export do not represent an indirect protection of domestic goods to the detriment of entrants, or a tax on imports/exports for fiscal purposes (US-Chile). 17.4.4.1.4 Provisions that limit or prohibit registry requirements, including licenses and permits Excessive or unjustified requirements for registry may restrict competition by creating unnecessary entry barriers. Relevant considerations include whether authorizations must be provided by different administrative authorities; conditions for granting authorizations, licenses, and permits; and conditions for their renewal. These obligations are especially common in investment chapters that provide for open and non-discriminatory special formalities and information requirements (India-Japan, US-Peru, ASEAN-India, Chile-Australia). In procurement, these obligations aim at ensuring open and equal conditions for participation and qualification (Canada-Honduras, EFTA-Central America, Israel-Mexico). In e-commerce, they prohibit parties from adopting or maintaining regulations for electronic signature that would limit or impede the determination of the electronic signature method for a transaction (Japan-Switzerland). Finally, in the case of IP rights, these provisions focus on establishing reasonable and publicly available protection and registration of trademarks (Japan-Australia, CARICOM). 17.4.4.1.5 Provisions that limit or prohibit impediments to switching providers Provisions that limit or prohibit impediments to switching providers are aimed at removing barriers to competition within a market. These provisions are usually present in telecommunications chapters (Trans-Pacific Partnership), but none were found in the sectors analyzed in this study. 539 Handbook of Deep Trade Agreements 17.4.4.2 Prohibitions on rules that are conducive to collusive outcomes or increase the cost of competing in the market Provisions limiting or prohibiting rules that are conducive to collusion or increase the cost of competing are aimed at preventing three types of anticompetitive effects: (a) facilitation of agreements among competitors; (b) restriction of the types and locations of products or services; and (c) establishment of price controls. Of all the competition-related sectoral commitments in PTAs, these are the least common (Figures 17.25 and 17.26). Figure 17.25: Evolution of PTAs with sector-specific provisions against rules that are conducive to collu- sion or increase the cost of competing in the market, by sector Sources: Wbg dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: Information is displayed for in-force PTAs. Figure 17.26: Sector-specific provisions against rules that are conducive to collusion or increase the cost of competing in the market Sources: Wbg dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: Information is displayed for in-force PTAs. 17.4.4.2.1 Provisions that limit or prohibit rules that facilitate agreements among competitors Rules that facilitate agreements among competitors are those that allow them to agree on key aspects that could restrict competition, such as prices of products and services offered 540 Competition Policy in the market, supply restrictions, dividing customers or territories, or coordinating their participation in public tenders. In investment chapters, these commitments limit excessive information requirements that would affect investors’ legitimate interests or otherwise distort competition (Chile-Australia, Chile-Japan, ASEAN-Australia-New Zealand). The treatment of information is also covered in government procurement provisions, with the objective of preventing anticompetitive conduct (Japan-Peru, Panama-Singapore). 17.4.4.2.2 Provisions that limit or prohibit restrictions on types and location of products or services Restrictions on types of products and services may occur when a regulatory instrument imposes unreasonable burdens or obstructs the activities of business in a way that may distort or prevent competition. Under this category, only one PTA has an e-commerce provision prohibiting limitations on the location of computing facilities (Japan-Mongolia). 17.4.4.2.3 Provisions that limit or prohibit rules establishing price controls Few PTAs include provisions that limit or prohibit price controls. When such provisions exist, they are as a general rule accompanied by a list of exceptions. However, some treaties include detailed regulations of price control methodologies for monopolies in all five sectors (EAEU-Accession of Kyrgyz Republic). In agriculture, price control provisions are more common. Some provisions regulate price measures at the community level (TFEU), and others regulate price band systems for agriculture (Canada-Colombia, Turkey-Chile). 17.4.4.3 Prohibitions on rules that discriminate or protect vested interests Discriminatory and protective rules favor firms that are already present in the market, either by granting them privileges or benefits that give them an advantage over their competitors, or by creating adverse or discriminatory conditions for non-favored entities. Prohibitions on these practices aim at limiting three types of market effects: (a) discriminatory application of rules or standards; (b) breach of competitive neutrality; and (c) distortive state aid or incentives. These types of provisions are most prevalent in agriculture chapters (265 PTAs), followed by investment chapters (165 PTAs). The latter have been on a rising trend since the early 2000s (Figure 17.27), with a stronger focus on the elimination of discriminatory application of rules and standards (Figure 17.28). 541 Handbook of Deep Trade Agreements Figure 17.27: Evolution of PTAs with sector-specific provisions against rules that discriminate or protect vested interests, by sector Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: Information is displayed for in-force PTAs. Figure 17.28: Sector-specific provisions against rules that discriminate or protect vested interests Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. Note: Information is displayed for in-force PTAs. 17.4.4.3.1 Provisions that limit or prohibit rules that allow for discriminatory application of rules or standards Provisions that limit discriminatory application of rules and standards are among the most prominent in all sectoral chapters, generally in the form of most-favored-nation or MFN (EFTA-Albania), national treatment or NT (EFTA-Albania), or other provisions prohibiting discriminatory treatment (TFEU and EC accession agreements) or providing for the adoption of non-discriminatory standards (Japan-Mongolia). 17.4.4.3.2 Provisions that limit or prohibit rules that breach competitive neutrality Breach of competitive neutrality occurs when a regulatory instrument allows another body to grant preferential treatment to public market players vis-à-vis private ones or to certain private operators. PTA provisions that limit this practice aim at leveling the playing field between foreign and domestic firms of the parties. These provisions appear across all sectoral chapters except for IP, and are prevalent in the agriculture sector. 542 Competition Policy In agriculture and e-commerce, competition neutrality commitments focus on reducing or eliminating customs duties and tariffs on imports of the other parties’ goods (Armenia- Ukraine for agriculture, Peru-Korea for e-commerce, US-Australia for customs duties). Clauses in investment chapters tend to provide for stable and transparent investment conditions (EFTA-Albania, EC-Mexico, EFTA-Bosnia and Herzegovina), no expropriation without due compensation (Japan-Mongolia), and rules extending the applicability of competition rules to government entities when they are competing in the market (Singapore-Taiwan, China). Finally, regarding government procurement, provisions are aimed at ensuring a fair and impartial review of tender procedures (US-Panama, Japan-Peru). 17.4.4.3.3 Provisions that limit or prohibit rules that create distortive state aid or incentives State support measures and incentives may create an uneven playing field when they benefit selected market players. These incentives may be explicit, in the form of direct subsidies, or implicit. Implicit incentives could be given through the provision of loans for capital expenditure or for operating costs at interest rates that are below market rates, or even interest-free as well as through tax exemptions. State aid clauses are typically found in agriculture and deal with domestic support measures and state subsidies (Canada-Colombia, Armenia-Russian Federation, Singapore-Australia). 17.4.5 Enforceability of provisions A key question about competition-related provisions in PTAs is the extent to which they are enforceable. Previous research on enforceability has failed to examine this question in terms of dispute settlement; that is, whether PTAs include or exclude competition provisions from dispute settlement processes. Based on the approach developed in two previous studies57 for judging the enforceability of competition-related provisions using the language of dispute settlement, we classify the provisions according to whether they are: • non-binding (level 0); • best effort (level 1); • binding but not subject to dispute settlement (level 2); • binding with state-to-state dispute settlement (level 3); • binding with private-state dispute settlement (level 4); • binding with both private-state and state-to-state dispute settlement (level 5). The levels of enforceability of each of the identified obligations in a PTA are averaged to obtain an enforceability score for the PTA, and these scores are then averaged to the obtain some insights about commonalities within groups of PTAs. The particular focus is on the degree of enforceability within the group of PTAs that include competition provisions. 57 Horn, Mavroidis, and Sapir 2010; Hofmann, Osnago, and Ruta 2017. 543 Handbook of Deep Trade Agreements The results of this exercise show a high degree of enforceability for the more substantive competition commitments (Annex Table 17.A.3 offers additional details on the degree of enforceability of specific obligations). To illustrate: • 154 PTAs (64.4 percent of all PTAs with competition-related provisions) have binding commitments regulating monopolies; • 131 PTAs (54.8 percent) have binding commitments prohibiting or regulating abuse of dominant position; • 117 PTAs (49 percent) have binding commitments regulating cartels and concerted practices; and • 112 PTAs (46.9 percent) have binding commitments regulating anticompetitive behavior of SOEs. Given that many PTAs also consider specific exceptions to each identified obligation, we constrained our analysis to this group of PTAs. Once again, the degree of enforceability was shown to be high in relation to substantive competition commitments, and very high in relation to the economy-wide provisions regarding institutional arrangements, direct applicability, and procedural fairness (Annex Table 17.A.4 shows the enforceability level of PTAs with specific exemptions). The degree of enforceability is also very high with regard to sectoral commitments, with more than 95 percent of the PTAs in the sample using binding language for the five sectors. Except for the provisions specific to investment, around 90 percent of the PTAs establish binding commitments with state-to-state dispute settlement. PTAs with provisions specific to investment tend to establish highly enforceable provisions, with 40 percent of PTAs establishing commitments with both private-state and state-to-state dispute settlement (Figure 17.29). Figure 17.29: Enforceability in PTAs with sectoral commitments Sources:WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 544 Competition Policy Sector commitments appear to be a major source of competition-enabling content in PTAs, since even PTAs without economy-wide commitments have a high level of enforceability (Table 17.4). However, it is possible that these findings may understate the competition-enhancing effects of sectoral commitments, since the analysis has not considered the service sectors. Table 17.4: Enforceability of sector-specific provisions Enforceability level Low High Economic sector Non- Best Binding, Binding, Binding, Binding, both binding effort with no DS with state- with private-state Total PTAs to-state DS private DS and state- to-state DS With competition Number (%) provisions Investment 0.9% 0.4% 17.8% 39.6% 0.9% 40.4% 100% 225 Agriculture 1.3% 1.7% 17.0% 70.2% 0.0% 9.8% 100% 235 e-commerce 2.0% 2.0% 20.2% 64.5% 0.0% 11.3% 100% 203 Government 1.9% 1.9% 20.9% 66.4% 0.0% 9.0% 100% 211 procurement Intellectual 0.9% 2.8% 18.5% 67.1% 0.0% 10.6% 100% 216 property Without competition provisions Investment 0.0% 0.0% 18.2% 50.0% 0.0% 31.8% 100% 22 Agriculture 0.0% 2.2% 17.8% 80.0% 0.0% 0.0% 100% 45 e-commerce 0.0% 0.0% 9.1% 90.9% 0.0% 0.0% 100% 11 Government 0.0% 0.0% 7.7% 92.3% 0.0% 0.0% 100% 13 procurement Intellectual 0.0% 0.0% 9.5% 90.5% 0.0% 0.0% 100% 21 property Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 17.4.6 Direct applicability Direct applicability of competition-related provisions - i.e., whether private firms can claim the application of competition provisions before the public bodies of the parties - increases the level of enforcement of competition related provisions. Direct applicability is more commonly included in multilateral PTAs than in bilateral ones. Moreover, enforceability does not seem to increase with direct applicability. The nine PTAs that expressly provide for direct applicability also include competition- related provisions. Out of these, seven are multilateral agreements or involve a multilateral party (COMESA, CARICOM, MERCOSUR, TPP, EAC, EAC-Burundi, EAC-Rwanda); and two are bilateral agreements (New Zealand-Taiwan, China and Canada-Costa Rica). 545 Handbook of Deep Trade Agreements 17.5. CONCLUSIONS A few core conclusions are apparent from the analysis. First, rules on competition are a prominent feature of the architecture of PTAs; they are found in four out of every five PTAs currently in force. Many PTAs also include the promotion of conditions for competition as a principal objective of the agreement. Second, the near absence of low-income countries in PTAs with strong competition provisions may be an important missed opportunity for making trade liberalization more effective and beneficial. Third, there appears to be a remarkable degree of enforceability in these provisions, based on the legal language used (in some cases including direct applicability), the applicability of the dispute settlement mechanism, and the possibility of bringing private claims to national bodies. Previous research on competition provisions in PTAs has not systematically examined the use of dispute settlement in competition-related provisions, so this result brings a new understanding to competition commitments in trade agreements. Fourth, the findings underline the importance of going beyond economy-wide competition obligations and examining competition-related commitments in the sectoral chapters of the agreement as a major source of the competition discipline in trade agreements. Finally, the dataset and mappings assembled for this study represent a valuable resource for researchers interested in measuring the effects of competition-related provisions on trade flows and other market outcomes. ACKNOWLEDGMENTS The authors would like to acknowledge Tilsa Ore Monago for her support with the coding and quantitative analysis, Nanna Matsson for her work in the interpretation and systematization of the treaties’ database, and Camila Ringeling for her valuable assistance in the analysis of treaties. We are also thankful to Bernard Hoekman for his valuable comments as well as Nadia Rocha, Michele Ruta, and Alvaro Espitia for their continuous support throughout this project. 546 Competition Policy REFERENCES Anderson, Robert, and Evenett, Simon. 2006. Incorporating Kee, Hiao Looi, and Bernard Hoekman. 2007. Imports, entry competition elements in regional trade agreements: and competition law as market disciplines. European Economic Characterization and empirical analysis. Paper presented Review 51 (4): 831-58. at the Seminar on Regional Rules in the Global Trading System,Washington, DC, July 26-27. Kitzmuller, Markus, and Marthus Licetti Martinez. 2012. Competition policy: Encouraging thriving markets for Brusick, Philippe, Ana Maria Alvarez, and Lucian Cernat. development. Financial and Private Sector Development Note 2005. Competition provisions in regional trade agreements: How to 331.Washington, DC:World Bank. http://documents.worldbank. assure development gains. Geneva: United Nations Conference org/curated/en/778181468328582034/Competition-policy- for Trade and Development. http://unctad.org/en/docs/ encouraging-thriving-markets-for-development. ditcclp20051_en.pdf. Khemani, R. Shyam, and Mark A. Dutz. 1995.The instruments of Buthe, Tim. 2015. The politics of market competition: Trade competition policy and their relevance for economic development. and antitrust in a global economy. In The Oxford handbook of In Regulatory policies and reform: A comparative perspective, edited by the political economy of international trade, edited by Martin Lisal. C.R. Frischtak.Washington, DC:World Bank. Oxford: Oxford University Press. Krakowski, Michael. 2005. Competition policy works: The effect Connor, John M. 2014. Price-fixing overcharges, 3rd edition. of competition policy on the intensity of competition—an Washington, DC: American Antitrust Institute of Purdue international cross-country comparison. Hamburg Institute of University. https://ssrn.com/abstract=2400780 or http:// International Economics, Hamburg. dx.doi.org/10.2139/ssrn.2400780. Kronthaler, Franz. 2007. Effectiveness of competition law: Dawar, Kamala, and Peter Holmes. 2011. Competition A panel data analysis. IWH Discussion Paper 7, IWH Halle policy. In Preferential trade agreement policies for development: A Institute for Economic Research, Halle, Germany. http:// handbook, edited by J.P. Chauffour and J.C. Maur.Washington, www.iwh-halle.de/e/publik/disc/7-07.pdf. DC: World Bank. Laprévote, François-Charles, Sven Frisch, and Burcu Can. 2015. Evenett, Simon 2003. Links between development and Competition policy within the context of free trade agreements. competition law in developing countries—case studies for E15 Initiative. International Centre for Trade and Sustainable World development report 2005: Investment Climate, Growth and Development and World Economic Forum, Geneva. Poverty. Commissioned by the United Kingdom’s Department for International Development (October). Matsushita, Mitsuo. 2004. Basic principles of the WTO and the role of competition policy. Washington University Global Goodwin, Tanja, and Martha Denisse Pierola. 2015. Export Studies Law Review 3 (2). competitiveness: Why domestic market competition matters. Viewpoint: Public Policy for the Private Sector OECD (Organisation for Economic Co-operation and Note 348. Trade and Competitiveness Global Practice, Development). 2015a. Roundtable on competition neutrality, World Bank. http://documents.worldbank.org/curated/ Issues Paper by the Secretariat. en/432141468189538318/pdf/97914-REPLACEMENT- FILE-VP-348-Export-Competitiveness-WEB.pdf. ———. 2015b. Roundtable on competitive neutrality, Note by the European Union. Gootiiz, Batshur, Joscelyn Magdeleine, Juan Marchetti, and Aaditya Mattoo. 2018. Trade in services. Unpublished Ore Monago, T. 2013. Assessing competition policy on manuscript. economic development. Revista Economia 36 (71) (Department of Economics-Pontificia Universidad Catolica del Peru). Hofmann, Claudia, Alberto Osnago, and Michele Ruta. 2017. Horizontal depth: A new database on the content of Rey, Patrick. 1997. Competition policy and economic preferential trade agreements. Policy Research Working Paper development. Institut d’Economie Industrielle. http://idei.fr/doc/ 7981, World Bank, Washington, DC. http://data.worldbank. by/rey/competition.pdf. org/data-catalog/deep-trade-agreements. Silva, Veronica. 2004. Cooperation on competition policy in Horn, Henrik, Petros C. Mavroidis, and Andre Sapir. 2010. Latin American and Caribbean bilateral trade agreements. Beyond the WTO? An anatomy of EU and US preferential trade Economic Commission for Latin America and the Caribbean. agreements. World Economy 33(11): 1565-88. http://repositorio.cepal.org/bitstream/handle/11362/3669/ S2004200_en.pdf?sequence=1&isAllowed=y. 547 Handbook of Deep Trade Agreements Solano, Oliver, and Andreas Sennekamp. 2006. Competition provisions in regional trade agreements. OECD Trade Policy Working Paper 31, OECD Publishing, Paris. http://dx.doi. org/10.1787/344843480185. Teh, Robert. 2009. Competition provisions in regional trade agreements. In Regional rules in the global trading system, edited by A. Estevadeordal, R. Teh, and K. Suominen. Cambridge: Cambridge University Press. https://www.researchgate.net/ publication/297476007_Competition_provisions_in_regional_ trade_agreements. World Bank. 2017. Deep versus shallow PTAs: Rules affecting competition dynamics – antitrust, competitive neutrality and pro-competition economic regulation. Unpublished paper. World Bank and Organisation for Economic Co-operation and Development. 2017. A step ahead: Competition policy for shared prosperity and inclusive growth. https://doi. org/10.1596/978-1-4648-0945-3. 548 Competition Policy ANNEX Annex Table 17.A.1: Economy-wide competition-related commitments in PTAs AREA SUB-AREA QUESTION General Does the agreement promote and advance the conditions of competition between parties? objectives Does the agreement establish cooperation in the field of competition? of the treaty Objectives Does the agreement promote open markets either (1) economy wide, (2) in specific sectors, or (3) both? Does the agreement promote the principle of transparency? Horizontal principles Does the agreement promote the principle of non-discrimination? Does the agreement promote the principle of procedural fairness? Does the agreement have as objective promoting fair competition and curbing anticompetitive practices? Does the agreement have as an objective promoting consumer welfare or economic efficiency? Treaties Are there existing regional competition instruments (e.g.,TFEU,WAEMU) binding the parties to the agreement? regulating Are there other bilateral or multilateral agreements that complement/clarify the provisions/commitments of the competition agreement? Are there other multilateral agreements (e.g., GATT) that complement/clarify the provisions/commitments of the agreement? National Does the agreement require the establishment/existence of competition law/measures? (if yes, (1) it applies competition to all commercial activities/to all economic agents; (2) it excludes specific activities/sectors; (3) it excludes requirements certain operators such as SOEs or designated monopolists.) Does the agreement require the establishment/existence of a competition authority? (if yes, (1) does it establish the need for the authority to be independent? (2) does it establish the need for the authority to be functional? (3) does it establish the need for the authority to have sufficient resources?) Does the agreement require the establishment/existence of competition policy/principles? (if yes, (1) it focuses both on economy-wide and sector-specific policies; (2) it focuses on economy-wide policies; (3) it focuses on sector-specific policies.) Regulated Does the agreement prohibit/regulate cartels/concerted practices? anticompetitive Does the agreement prohibits/regulate abuse of market dominance? behavior Does the agreement regulate undertakings with exclusive rights? Does the agreement regulate monopolies? Does the agreement regulate anticompetitive behavior of SOEs? Does the agreement regulate state aid? Competition Does the agreement regulate mergers and acquisitions? policy Procedural Does the agreement contain provisions that promote predictability? (if yes, (1) it requires written procedures fairness to conduct investigations; (2) it requires procedures to be conducted within a reasonable period of time; (3) it requires the establishment/existence of settlements or consent agreements.) Does the agreement contain provisions that promote transparency? (if yes, (1) parties should be informed of competition concerns; (2) there should be a reasonable opportunity to consult with the competition authority; (3) decisions should include legal reasoning; (4) decisions should be published.) Does the agreement contain provisions that promote the right of defense? (if yes, (1) it requires parties to have a reasonable opportunity to be represented by counsel; (2) it requires parties to have a reasonable opportunity to be heard and provide evidence or testimony in their defense.This includes testimony of experts, cross-examination of testifying witnesses, and access to review and opportunity to rebut any evidence in the proceeding; (3) it requires the presumption of innocence to be respected during investigation; (4) it requires the protection of confidential information; (5) it requires reasonable access to the case file; (6) it requires the right of appeal; (7) it requires the right to private enforcement for competition issues before domestic courts.) Other substantive Does the agreement regulate unfair commercial practices? aspects Does the agreement regulate consumer protection? Forms of Does the agreement provide for coordination among bodies with a competition mandate? cooperation Does the agreement provide for exchange of information among bodies with a competition mandate? Does the agreement provide for notifications among bodies with a competition mandate? (if yes, (1) investigations; (2) decisions.) Does the agreement provide for technical assistance on implementation among bodies with a competition mandate? Institutional Does the agreement provide for the creation of arrangements a regional/agreement-related competition authority? Direct Do the competition-related provisions of the agreement have direct applicability in the applicability sense that application can be claimed by private operators before the public bodies of the parties? What general exceptions are included in the agreement? (GATS Article XIV list) General Does the agreement allow for security exceptions? exceptions Do other exceptions apply to competition-related provisions? 549 Handbook of Deep Trade Agreements Annex Table 17.A.2: Sector-specific competition-related commitments in PTAs 1 Does the specific chapter include the submission to competition law/prohibition against anticompetitive behavior for covered parties? (if yes, (1) this applies to SOEs; (2) this applies to designated monopolies; or (3) both) 2 Rules that reinforce dominance or limit entry 2.1 Does the agreement limit/prohibit monopoly rights and absolute bans on entry? 2.2 Does the agreement limit/prohibit relative bans on entry or expansion of activities? 2.3 Does the agreement limit/prohibit entry to protect incumbents’ rights? 2.4 Does the agreement limit requirements for registry (licenses and permits)? 2.5 Does the agreement prohibit/limit impediments to switching providers? 3 Rules that are conducive to collusive outcomes or increase costs to compete in the market 3.1 Does the agreement prohibit/limit rules that facilitate agreements among competitors? 3.2 Does the agreement limit/prohibit restrictions on types of products or services and location? 3.3 Does the agreement limit/prohibit price controls? 4 Rules that discriminate or protect vested interests 4.1 Does the agreement limit/prohibit discriminatory application of rules or standards? 4.2 Does the agreement prohibit/limit rules that breach competitive neutrality? 4.3 Does the agreement limit/prohibit distortive state aid/incentives that distort a level playing field? 5 Enforceability (non-binding (0); best efforts (1); binding with no DS (2); binding with state-to-state DS; (3); binding with private DS (4); binding with both state-to-state and private DS (5). 6 Are there exemptions? 550 Competition Policy Annex Table 17.A.3: Enforceability of economy-wide provisions in PTAs, with competition provisions Total 239 PTAs with competition provisions Number Best Binding Binding Binding, of PTAs offorts with with both private Features of PTA that no DS state-to -state and aswered -state DS state-to yes -state DS Pomotes and advances the conditions General of competition between parties 221 2.7% 52.5% 34.8% 10.0% objectives of PTA Establishes cooperation in the field of competition 130 3.1% 65.4% 6.2% 5.4% Promotes open market 231 2.6% 19.9% 66.2% 11.3% Promotes the principle of transparency 163 1.8% 21.5% 63.2% 13.5% Horizontal principles Promotes the principle of non-discrimination 161 1.2% 35.0% 54.7% 8.1% Promotes the principle of procedural fairness 87 0.0% 34.5% 39.1% 26.4% Has as objective promoting fair competition and curbing anti-competitive practices 187 2.1% 56.7% 29.9% 11.2% Objectives Has as objective promoting consumer welfare or economic efficiency 68 2.9% 48.5% 35.3% 13.2 There exist regional competition instruments with direct Treaties applicability binding the parties to the agreement 51 0.0% 23.5% 58.8% 17.6% regulating competition There are other multilateral agreements (e.g., GATT) that complement/clarify the provisions/commitments of the agreement 21 0.0% 23.8% 71.4% 4.8% Requires the establishment/existence of competition law/measures 122 4.1% 69.7% 17.2% 9.0% National competition Requires the establishment/existence requirements of competition authority 74 2.7% 68.9% 17.6% 10.8% Requires the establishment/existence of competition policy/principles 43 7.0% 65.1% 9.3% 18.6% Prohibits/regulates cartels/concerted practices 117 0.9% 47.0% 43.6% 8.5% Prohibits/regulates abuse of market dominance 131 0.8% 48.1% 43.5% 7.6% Regulates undertakings with exclusive rights 97 0.0% 28.9% 61.9% 9.3% Regulated anticompetitive Regulates monopolies 154 1.9% 31.8% 53.6% 10.3% behaviour Regulates anticompetitive behavior of SOEs 112 0.0% 34.8% 51.8% 13.4% Regulates state aid 87 1.1% 29.9% 58.6% 10.3% Regulates mergers and acquisitions 28 10.7% 42.9% 17.9% 28.6% Contains provisions that promote predictability 17 5.9% 23.5% 17.6% 52.9% Procedural fairness Contains provisions that promote transparency 48 2.1% 54.2% 22.9% 20.8% Contains provisions that promote the right of defence 63 1.6% 50.8% 33.3% 14.3% Other Regulates unfair commercial practices 54 1.9% 50.0% 48.1% 0.0% substantive aspects Regulates consumer protection 60 1.7% 30.0% 51.7% 16.7% Provides for coordination among bodies with a competition mandate 69 7.2% 66.7% 14.5% 11.6% Provides for exchange of information among bodies Forms of with a competition mandate 115 5.2% 67.0% 20.9% 7.0% cooperation Provides for notifications among bodies with a competition mandate 76 5.3% 71.1% 13.2% 10.5% Provides for technical assistance on implementation among bodies with a competition mandate 55 5.5% 74.5% 5.5% 14.5% Institutional Provides for the creation of a regional/agreement-related arrangement competition authority 20 15.0% 20.0% 20.0% 45.0% Direct Competition-related provisions of the agreement have direct applicability applicability in the sense that application can be claimed by 9 0.0% 55.6% 44.4% 0.0% private operators before the public bodies of the parties General There are general exceptions included in the agreement 195 0.5% 13.3% 73.8% 12.3% exceptions Allows for security exceptions 217 0.5% 19.8% 68.2% 11.5% Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 551 Handbook of Deep Trade Agreements Annex Table 17.A.4: Enforceability of economy-wide provisions in PTAs, with specific exceptions With specific provision exemptions Number Non Best Binding Binding Binding, of PTAs binding efforts with with both private Features of PTA that no DS state-to -state and answered -state DS state-to yes -state DS Promotes and advances the conditions General of competition between parties 238 0.0% 2.5% 49.6% 38.7% 9.2% objectives of PTA Establishes cooperation in the field of competition 135 0.7% 3.0% 64.4% 17.0% 14.8% Promotes open markets 285 2.8% 2.1% 18.9% 67.0% 9.1% Promotes the principle of transparency 197 2.5% 1.5% 18.8% 66.0% 11.2% Horizontal Promotes the principle of non-discrimination 190 2.1% 1.1% 33.2% 56.8% 6.8% principles Promotes the principle of procedural fairness 127 4.7% 0.0% 26.8% 50.4% 18.1% Has as objective promoting fair competition and curbing anti-competitive practices 189 0.0% 2.1% 56.6% 30.2% 11.1% Objectives Has as objective promoting consumer welfare or economic efficiency 70 0.0% 2.9% 47.1% 37.1% 12.9% There exist regional competition instruments with direct applicability binding the parties to the agreement 50 0.0% 0.0% 24.0% 60.0% 16.0% National There are other bilateral or plurilateral agreements competition that complement/clarify the provisions/commitments 2 0.0% 0.0% 100.0% 0.0% 0.0% requirements of the agreement There are other multilateral agreements (e.g. GATT) that complement/clarify the provisions/commitments 22 0.0% 0.0% 22.7% 72.7% 4.5% of the agreement Requires the establishment/existence of competition law/measures 123 0.8% 4.1% 69.1% 17.1% 8.9% Treaties regulating Requires the establishment/existence competition of a competition authority 74 0.0% 2.7% 68.9% 17.6% 10.8% Requires the establishment/existence of a Competition policy/principles 43 0.0% 7.0% 65.1% 9.3% 18.6% Prohibits/regulates cartels/concerted practices 117 0.0% 0.9% 47.0% 43.6% 8.5% Prohibits/regulates abuse of market dominance 132 0.8% 0.8% 47.7% 43.2% 7.6% Regulated anti- Regulates undertakings with exclusive rights 99 2.0% 0.0% 28.3% 60.6% 9.1% competitive Competition policy behaviour Regulates monopolies 154 0.0% 1.9% 31.8% 53.9% 12.3% Regulates anticompetitive behaviour of SOEs 112 0.0% 0.0% 34.8% 51.8% 13.4% Regulates state aid 91 4.4% 1.1% 28.6% 56.0% 9.9% Regulates mergers and acquisitions 28 0.0% 10.7% 42.9% 17.9% 28.6% Contains provisions that promote predictability 17 0.0% 5.9% 23.5% 17.6% 52.9% Procedural fairness Contains provisions that promote transparency 48 0.0% 2.1% 54.2% 22.9% 20.8% Contains provisions that promote the right of defence 64 1.6% 1.6% 50.0% 32.8% 14.1% Other Regulates unfair commercial practices 59 0.0% 1.7% 45.8% 52.5% 0.0% substantive aspects Regulates consumer protection 65 3.1% 1.5% 27.7% 52,3% 15.4% Provides for coordination among bodies with a competition mandate 70 1.4% 7.1% 65.7% 14.3% 11.4% Provides for exchange of information among bodies with a competition mandate 116 0.9% 5.2% 66.4% 20.7% 6.9% Forms of cooperation Provides for notifications among bodies with a competition mandate 76 0.0% 5.3% 71.1% 13.2% 10.5% Provides for technical assistance on implementation among bodies with a competition mandate 55 0.0% 5.5% 74.5% 5.5% 14.5% Institutional Provides for the creation of a regional / agreement arrangement - related competition authority 22 9.1% 13.6% 18.2% 18.2% 40.9% Competition-related provisions of the agreement have Direct direct applicability in the sense that application can be applicability claimed by private operators before the public bodies 9 0.0% 0.0% 55.6% 44.4% 0.0% of the parties There are general exceptions included General in the agreement 236 1.7% 0.4% 13.6% 74.2% 10.2% exceptions Allows for security exceptions 255 0.8% 0.4% 18.8% 70.2% 9.8% Sources: WBG dataset on competition policy provisions in PTAs;WTO Regional Trade Agreements Information System. 552 Environmental Laws CHAPTER 18 Environmental Laws J.-A. Monteiro and J.P. Trachtman 553 Handbook of Deep Trade Agreements CHAPTER 18 Environmental Laws J.-A. Monteiro* and J.P. Trachtman† * World Trade Organization, Geneva, Switzerland † Tufts University, Medford, Massachusetts, United States CONTENTS 18.1. INTRODUCTION 555 18.2. DATABASES ON ENVIRONMENT-RELATED PROVISIONS IN PTAs 557 18.2.1 Literature review 557 18.2.2 New database 561 18.2.2.1 Mapping of five agreements 562 18.2.2.2 Coding process 563 18.3. PATTERNS OF ENVIRONMENT-RELATED PROVISIONS 563 18.3.1 Evolution of environment-related provisions 564 18.3.2 Dynamic nature of environment-related provisions 566 18.3.3 Scope of environment-related provisions in PTAs 568 18.3.3.1 Environmental goals and objectives 569 18.3.3.2 Balance between environmental and trade/investment goals 570 18.3.3.3 General environmental protection areas 571 18.3.3.4 Compliance with multilateral environmental agreements 572 18.3.3.5 External assistance 572 18.3.3.6 Participation, transparency, and cooperation in promoting environmental objectives 572 18.3.3.7 Enforcement mechanisms 574 18.3.4 Heterogeneity of environment-related provisions 575 18.4. CONCLUSIONS 577 ACKNOWLEDGMENTS 577 REFERENCES 578 ANNEX 579 554 Environmental Laws 18.1. INTRODUCTION An important part of the debate about trade and the environment lies in how to ensure that trade liberalization and environmental protection are mutually supportive. Although the relationship between trade liberalization and environmental protection has been discussed in detail in the World Trade Organization (WTO), none of the WTO agreements require any environmental protection measures. The WTO treaty does require reduction of some subsidies that may induce environmental harm, and provides exceptions that may permit environmental protection measures that otherwise violate another provision of the treaty. In parallel to the WTO dynamics, the last 25 years have witnessed a rapid rise in the number and regulatory coverage of preferential trade agreements (PTAs), and an increasing number have made explicit reference to environmental issues and policies. Different rationales for including environment-related provisions (ERPs) in PTAs have been discussed in the literature. One rationale is the “pollution haven hypothesis,” also known as “environmental arbitrage,” which posits that reduced trade barriers induce shifts of production to places of lower environmental regulation.1 There is some theoretical support for this form of leakage but little empirical evidence. Motivated by the possibility of leakage, both environmental activists and labor unions in the high-regulation states condition their support for PTAs or their investment chapters on the inclusion of provisions requiring low-regulation states to improve, or at least maintain, their environmental regulation. Another reason for including environmental provisions in PTAs is a political version of the environmental arbitrage effect: reduced trade barriers could produce a “race to the bottom” in environmental protection, so a state would not be able to increase its level of environmental protection without increasing costs and losing market share. Third, high-regulation states may use their superior bargaining power in PTA negotiations to achieve environmental goals in low-regulation states. One study suggests that agreements between developed and developing countries include, on average, a substantially higher number of ERPs, because of the bargaining power of developed countries to offer valuable market access in return for concessions regarding the inclusion of ERPs.2 Fourth, the intensification of production and increased transport that may accompany the establishment of PTAs may give rise to greater pollution concerns, causing environmental activists to condition their support on general improvement of environmental protection. 1 Copeland and Taylor 1994. 2 Berger, Brandi, and Bruhn 2016. 555 Handbook of Deep Trade Agreements Over the years, some governments have set policies calling for or requiring the inclusion of ERPs in their new PTAs. For instance, in 2001 New Zealand established the Framework for Integrating Environment Standards and Trade Agreements, requiring trade agreements to reflect sustainable development; promote coherence with multilateral environmental agreements (MEAs); and preserve the right of all parties to the agreement to regulate environmental concerns. More recently, in the United States, the Bipartisan Trade Deal of May 2007 calls for the inclusion of provisions on specific MEAs;3 a mandatory non- derogation provision prohibiting to reduction of environmental protections to promote trade or investment; and the requirement that ERP provisions be subject to the same dispute settlement (DS) mechanisms as trade provisions, including trade sanctions. This chapter has three objectives: (a) review existing literature on environment-related provisions in PTAs; (b) explain the methodology used to establish a template and coding of environmental provisions in PTAs in the World Bank’s PTA database; and (c) analyze selected aspects of the results. For purposes of this analysis, ERPs are those listed in the Environment-Related Provisions Template in Annex 18A of this chapter. The derivation of that list is explained in Section 18.2.2. Generally speaking, ERPs call for (a) specified levels of environmental protection; (b) specified political or legal mechanisms for enforcement of environmental protection; (c) acceptance of environmental protection obligations contained in other instruments, such as MEAs; or (d) maintenance of a national status quo in environmental protection. The scope of ERPs defined in this study does not include general exceptions that may be used to permit environmental protection where it may otherwise violate a PTA provision, because these environmental general exceptions do not establish an obligation of environmental protection. However, for completeness in coding, the mapping includes exceptions that may be used to defend national environmental measures, including but not limited to those for human, animal, or plant health and those relating to conservation of natural resources. This chapter examines the depth of the ERPs in the 280 PTAs in the Deep Integration database. An ERP is deemed to have depth in this context if it establishes a substantive environmental protection requirement.4 For each identified ERP, we also evaluate (a) whether it is aspirational or legally binding under international law; (b) whether it is applicable in DS proceedings brought by other parties to the agreement (state-to-state DS); and (c) whether it is applicable in DS brought by private persons (state-private DS). 3 The legislation calls for incorporation into US PTAs of (a) the Convention on International Trade in Endangered Species (CITES); (b) the Montreal Protocol on Ozone-Depleting Substances (Montreal Protocol); (c) the Convention on Marine Pollution; (d) the Inter-American Tropical Tuna Convention (ITTC); (e) the Ramsar Convention on Wetlands (Ramsar); (f) the International Whaling Convention (IWC); and (g) the Convention on the Conservation of Antarctic Marine Living Resources (CCAMLR). 4 This definition of depth is distinct from that utilized in the political science literature, in which depth is understood as the extent to which a provision requires a change in behavior from the counterfactual (Downs et al. 1996). 556 Environmental Laws The remainder of the chapter is structured as follows. Section 18.2 provides a literature review of the different existing databases on environment-related provisions in PTAs, as well as a description of the methodology used to create a new, comprehensive database. Section 18.3 provides a descriptive analysis and overview of the main types of ERPs. Section 18.4 concludes. 18.2. DATABASES ON ENVIRONMENT-RELATED PROVISIONS IN PTAs Although several studies discuss the various types of environment-related provisions included in PTAs, very few provide a comprehensive analysis of the scope and depth of these provisions. Before presenting the methodology used to build a new database on environment-related provisions in PTAs, this section reviews the relevant existing databases. 18.2.1 Literature review One study5 reports on the development of the Trade and Environment Database (TREND), which includes 308 different environmental provisions from about 630 PTAs signed between 1945 and 2016. The authors utilize a manual coding process to identify these ERPs and provide useful descriptive statistics. They then evaluate the following three hypotheses: Morin H1: Democratic countries include more environmental clauses in trade agreements than autocratic countries. Morin H2: Countries include more environmental clauses in trade agreements, the greater the agreements’ distributional effects are. Morin H3: Governments with higher levels of environmental protection include more environmental provisions in trade agreements than governments with weaker levels of protection. Another study6 develops a comprehensive typology of ERPs included in regional trade agreements (RTAs), and analyzes the 270 PTAs notified to the WTO through 2016 using that typology. The methodology uses a keywords-based textual analysis to identify ERPs. It then categorizes ERPs pursuant to a typology of 62 main types of provisions,7 and uses descriptive statistics to evaluate, among others, the following hypotheses: Monteiro H1: PTAs between developed and developing countries have greater numbers of ERPs. Monteiro H2: Over time, provisions related to domestic environmental law, policies, measures, or regulations are incorporated in an increasing number of PTAs. Monteiro H3: The number of PTAs with provisions related to MEAs has increased significantly since 2006. 5 Morin, Dür, and Lechner 2018. 6 Monteiro 2016. 7 The typology is similar to that used in the present chapter. 557 Handbook of Deep Trade Agreements A third study8 presents a dataset that codes 587 agreements for more than 100 items (e.g., elimination of tariffs, trade in services, investments, standards, public procurement, intellectual property rights). The study creates a depth indicator ranging from 0 to 7 to determine whether provisions relating to each items are deep or shallow. A deep provision covers the item in a substantive way, while a shallow one does not cover the item or mentions it only briefly. In that database, however, the coding for ERPs is quite limited, addressing only the following parameters: 1. Does the preamble to the PTA refer to environmental protection? 2. Does the agreement refer to environmental protection?9 3. Is there a reference to international environmental agreements? The authors state, however, that they cannot address the issue of depth in connection with environmental protection (or most other non-trade issues) because they do not determine the counterfactual or assess the counterfactual’s distance from the relevant treaty obligation. The present chapter also does not address depth in this sense. A fourth study10 relies on the same depth indicator11 to develop a database titled Sustainable Innovations in Trade Agreements (SITA) to begin to evaluate the depth of ERPs in PTAs. When completed, the SITA database will comprise detailed data on the design of environmental provisions along nine dimensions in all 400 or so PTAs and customs unions available in full text. These dimensions are: 1. Reference to environmental goals in the preamble or other chapters. 2. Environmental exceptions (general or specific to certain chapters). 3. Reference to MEAs. 4. Inclusion of a whole chapter on the environment or sustainable development. 5. Obligations to uphold environmental law. 6. Incorporation of the right to regulate in environmental matters. 7. Cooperation on environmental matters. 8. Transparency in environmental matters. 9. Public participation in environmental matters. The results rely on a double-blind coding procedure based on a detailed code book.While the SITA database represents an important advance in knowledge and specificity regarding ERPs 8 Dür, Baccini, and Elsig 2014. 9 The second question is a very general parameter, by its terms including environmental exceptions such as those contained in Article XX of GATT. 10 Burger, Brandi, and Bruhn 2016. 11 Dür, Baccini, and Elsig 2014. 558 Environmental Laws in PTAs, it is still somewhat general, as it does not disaggregate specific areas of environmental protection. Nor does the database make specific reference to or incorporate MEAs, distinguish between aspirational and mandatory provisions, or identify obligations subject to DS. A fifth study12 notes an upward trend in environmental provisions in trade agreements, with two basic provisions, one associated with GATT Article XX13 or GATS Article XIV, and the other a reference to the environment or sustainable development in the preamble to the PTA. These provisions are included in, respectively, about 80 percent and 50 percent of the PTAs reviewed. Moreover, the incidence of more substantive provisions has increased significantly in recent years, from around 30 percent of PTAs entering into force up to 2010, to greater than 50 percent in 2011 to close to 70 percent in 2012. The findings of the fifth study are based on the presence of following environmental content in PTAs: 1. Reference to environmental protection in the preamble. 2. GATT/GATS-type exceptions. 3. Obligations to uphold existing environmental law. 4. Substantive environmental protection requirements. 5. Requirement for environmental cooperation. 6. Requirement for public participation in environmental policy. 7. Dispute settlement. 8. Specific environmental issues. 9. Multilateral environmental agreements. 10. Implementation mechanism. Still another study14 seeks to assess the difference in environmental effects between PTAs that provide for harmonization in environmental standards and those that do not. The authors hypothesize that: Baghdadi H1: PTAs with environmental provisions cause a convergence of CO2 emissions and hold constant bilateral trade, openness, and income levels. Using the WTO database designation to identify whether an agreement contains environmental provisions, the study finds evidence that PTAs with environmental provisions statistically explain the convergence of pollution levels across pairs of countries. Moreover, 12 George 2014. 13 Article XX includes two exceptions of particular relevance to the protection of the environment: paragraph (b) provides that WTO members may adopt policy measures that are inconsistent with GATT disciplines if they are necessary to protect human, animal, or plant life or health; and paragraph (g) gives members the right to adopt or enforce measures related to the conservation of exhaustible natural resources. 14 Baghdadi, Martinez-Zarzoso, and Zituna 2013. 559 Handbook of Deep Trade Agreements agreements that specifically include provisions to ensure enforcement (e.g., NAFTA) are converging at a higher rate than are agreements that leave compliance measures to the legal system (e.g., EU).15 A seventh study16 sets out to develop a framework for evaluating the implementation of environmental provisions in PTAs. The working paper examines, by way of potentially helpful questions, how to make an ex-post assessment of environmental provisions. It complements the OECD’s Checklist for Negotiators for Environmental Provisions in Regional Trade Agreements,17 which includes the following environmental issues: 1. Institutional arrangements. 2. Cooperation. 3. Capacity building. 4. Public participation. 5. Resolution of differences (dispute settlement). 6. Monitoring and assessment. 7. Commitments. 8. Environmental laws and standards. 9.Voluntary and private action. 10. Environmental goods and services. Taking each of these issues individually, the authors define them, give them context, and provide examples of trade agreements that include them. More importantly, they also provide a list of questions to evaluate these issues. For example, under institutional arrangements, they offer the following:What provisions (if any) does the agreement include on institutional arrangements? What actions have the parties taken in relation to these provisions? For cooperation activities, the authors ask whether the parties have identified their respective priorities for cooperation and agreed on joint priorities or areas of common interest? The authors provide a wide array of PTA examples in their working paper, observing that most of these examples are North-South agreements that vary in their substantive commitment to environmental protection. The paper notes, for example, that some of the largest PTAs with the greatest number of environmental provisions create joint institutions and programs to promote and implement environmental cooperation. Examples are the Commission on Environmental Cooperation created under the North American Agreement on Environmental 15 The WTO database designation reflects agreements that contain commitments on the environment, often in the form of a separate chapter or side letter, or make reference to sustainable development, or include provisions on environmental measures in the investment chapter. Such provisions often refer to “best endeavor” rather than binding commitments. 16 Gallagher and Serret 2011. 17 OECD 2008. The checklist approach is also used by Tébar Less and Kim 2008. 560 Environmental Laws Cooperation (NAAEC); the joint commission created by the Canada-Chile Agreement on Environmental Cooperation (a side agreement of the Canada-Chile Free Trade Agreement); and the Environmental Cooperation Program under the United States-Central America- Dominican Republic Free Trade Agreement (US-CAFTA-Dominican Republic). In addition, the working paper suggests that the ambition and scope of environmental provisions can be assessed by looking at the legal and policy basis of the mandate for negotiations, the specificity of the mandate, the substantive nature of the obligations, and their enforceability. This approach could be the basis for a solid ex-post assessment of environmental issues in trade agreements. Yet another study18 examines a wide variety of PTA commitments, distinguishing between those that constitute intensification of existing WTO commitments and those that address issues not within the WTO domain. The authors also distinguish between aspirational and mandatory provisions, calling the former “legal inflation.” However, their coding for environment-related provisions is very general, focusing only on the categories into which they fall: development of environmental standards, enforcement of national environmental laws, establishment of sanctions for violating environmental laws, and publication of laws and regulations. In addition to these databases and these studies, a number of international organizations have published their own databases on PTAs, some of which specify the different issues, including environmental protection, covered in each agreement. For instance, the WTO database contains detailed information on all the PTAs notified to the GATT/WTO, including whether the agreements address environmental protection. Similarly, the Organization of American States (OAS) has created a database that tracks environmental provisions in PTAs to which OAS member states are party. 18.2.2 New database As part of the World Bank’s Deep Integration database project, this study develops a new database on ERPs based on the review of 295 PTAs signed between 1956 and November 2016. The template for coding the ERPs was created through an inductive, iterative process. It was initially developed by evaluating a group of PTAs likely to have a fairly representative and diverse group of ERPs, as well as by reviewing the secondary literature and some relevant government policies. During the coding process, additional types of ERPs were identified and added to the template. The goal was to create as inclusive and comprehensive a database as possible, while avoiding an excessive number of parameters. 18 Horn, Mavroidis, and Sapir 2010. 561 Handbook of Deep Trade Agreements While most ERPs are found in the main text of PTAs, some agreements include both internal ERPs and side agreements on the environment. The North-American Free Trade Agreement (NAFTA), for example, includes both environmental provisions and a side agreement - the North American Agreement on Environmental Cooperation (NAAEC) - that promotes sustainable development among the signatories based on mutually supportive environmental policies.These different types of environmental provisions raised an important methodological issue: should the template include (a) only the PTA itself; or (b) the PTA plus any side agreements or other agreements referenced in the PTA; or (c) in addition, other agreements that are not referenced in the PTA but shape the informal context for the PTA. Because agreements in the last group are not formally referenced in the PTA and identifying them would have been a formidable research challenge, such non- referenced agreements are not included in the template.Thus, the coding only includes PTAs and their side agreements. In order to identify these environmental side agreements, we performed an online search for each PTA along with the search term “environmental agreement.” We also checked national websites for any mention of PTA environmental side agreements. Provisions of PTAs may be expressed either as non-binding aspirational language (including “soft law”) or binding commitments. From a practical standpoint, the actual implementation and compliance of ERPs do not necessarily hinge on their formally binding nature. Nevertheless, the distinction between binding and non-binding commitments is important, since whether or not a commitment is binding may be the basis for a claim in DS. Therefore, in the template, the existence of a commitment, and whether it is binding or non-binding, are coded separately. If a provision is binding, it is coded as subject to state-to-state DS and/or private-state claims. In coding for the nature of environmental commitments in PTAs, we use the following coding convention: 0. No provision. 1. Non-binding provision. 2. Binding provision with no DS. 3. Binding provision with only state-to-state DS. 4. Binding provision with only private-state DS. 5. Binding provision with both state-to-state and private-state DS. The database also attempts to assess whether each ERP provides benefits to non-members of the PTA. The coding scale for this issue was developed as follows: 0 (benefit is excludible to non-members), 1 (benefit is non-excludible to non-members), and 2 (not applicable). 18.2.2.1 Mapping of five agreements For initial development of the template, five agreements were selected on the basis of their likelihood of containing an extensive set of ERPs, plus some degree of regional diversity. The 562 Environmental Laws Trans-Pacific Partnership (TPP), although not yet in force at the time, was selected because it represents a negotiated agreement containing an extensive set of US-requested ERPs. The European Union-Canada Comprehensive Economic and Trade Agreement (CETA), also not in force at the time, was included as an example of EU and Canadian practice in their latest trade agreement. For regional diversity, the PTAs of South America (MERCOSUR19), West Africa (ECOWAS), and South Asia (SAARC) were also selected. For each agreement, the coding protocol was as follows: • Obtain full agreement, using original English or English translation where available.20 • Check for news reports or government announcements of environmental side agreements. • Review the agreement plus any environmental side agreement against the template, assigning a code of 0-5 for each item on the template. • Where an issue listed in the template is addressed, specify the provision of the agreement where the issue appears. 18.2.2.2 Coding process Every PTA was coded independently by two coders. Coding was carried out agreement by agreement.21 The reports of the two coders were then compared, and the principal investigator resolved disparities in consultation with the coders. While this approach may have compromised coder independence and objectivity, it promoted a more analytical and considered disposition in circumstances of complex provisions and subtle coding issues. The coders recorded their deliberations, and dispositions by the principal investigator, in a coding memorandum in order to promote transparency and reproducibility. 18.3. PATTERNS OF ENVIRONMENT-RELATED PROVISIONS Provisions in PTAs are known to be heterogenous across agreements, and ERPs are no exception. Before reviewing the scope of each main type of ERPs and discussing the level of heterogeneity of ERPs, this section provides a descriptive analysis of the evolution and dynamic nature of ERPs. 19 MERCOSUR members adopted a Framework Agreement on Environment in 2001, ten years after the trade agreement; see http://www.medioambiente.gov.ar/mercosur. 20 There were some agreements in Spanish that did not have an English translation, so our Spanish-speaking coder explained each environmental provision of these agreements to the non-Spanish-speaking coder. 21 Following the method of Mitchell and Rothman 2006. 563 Handbook of Deep Trade Agreements 18.3.1 Evolution of environment-related provisions The inclusion of provisions referring explicitly to the environment is not a recent phenomenon. The 1957 Treaty of Rome establishing the European Economic Community was the first PTA to include a provision related to the environment - a general exception allowing a party to prohibit or restrict imports, exports, or goods in transit on grounds of protecting the health and life of animals or plants as long as these prohibitions or restrictions were not arbitrary or discriminatory. Of the 295 PTAs created between 1956 and 2016, 274 (93 percent) incorporate at least one ERP (Figure 18.1). Of these 274 PTAs, 143 were negotiated between developed countries (North-North PTAs), 108 between developed and developing countries (North-South PTAs), and only 23 between developing countries (South-South PTAs).22 The evolution of PTAs with ERPs can be further characterized by four different periods. Prior to the 1990s, the number of PTAs with ERPs was limited, namely, 15 agreements. Most ERPs in these agreements took the form of an environmental exception clause. For instance, the 1960 Convention establishing the European Free Trade Association (EFTA) includes environment-related preamble language as well as several exception clauses, including the right to unilaterally take appropriate safeguard measures if serious environmental difficulties of a sectoral or regional nature arise and are likely to persist. Between 1990 and 2005, the number of PTAs with ERPs increased, but the number of specific ERPs included in those PTAs remained limited (Figure 18.2). One of the few exceptions was NAFTA: the first PTA to include detailed ERPs in the main text as well as in a dedicated environmental cooperation agreement. The side agreement established commitments to effectively enforce environmental laws and standards and to not lower them in order to attract investment. It also set out a range of environment-related cooperation activities, institutional bodies, and specific DS procedures. Between 2005 and 2010, the creation of PTAs with ERPs accelerated, driven by the surge in North-North PTAs and North-South PTAs. In particular, the number of PTAs with ERPs involving high-income and middle-income countries increased significantly during this period, as did the types of ERPs (Figure 18.2). For instance, the Trade Promotion Agreement between the United States and Peru included environment-related language in the preamble; an environmental exception clause; an article on environmental measures in the investment chapter; a specific environment chapter with provisions on the enforcement of environmental laws; obligations to comply with MEAs; procedural guarantees of enforcement 22 Unless otherwise noted, developed countries are high income and upper middle income, while developing countries are lower middle and low income. 564 Environmental Laws of environmental laws; mechanisms to enhance environmental performance; and a detailed annex on forest sector governance. In addition, the US-Peru agreement included a side environmental cooperation agreement establishing a work program with various areas of cooperation, and a letter of understanding on biodiversity and traditional knowledge. Figure 18.1: Evolution of the number of PTAs with environment-related provisions Source: Deep Trade Agreements Database. Figure 18.2: Evolution of the number of types of environment-related provisions in PTAs Source: Deep Trade Agreements Database. 565 Handbook of Deep Trade Agreements Between 2011 and December 2017, the number and share of PTAs with ERPs continued to increase but at a significantly slower pace, due in part to a slowdown in the number of agreements being signed. Even with the slowing pace of PTA formulation, the content of ERPs in the more recent PTAs has, on average, continued to increase significantly, particularly in PTAs between developed and developing economies. The EU is party to several North- South PTAs that incorporate a high number of ERPs of different types. For example, the EU-Georgia Association Agreement included preamble language; environmental exceptions; commitments related to the enforcement of domestic environmental laws and MEA obligations, promotion of environmental goods and services, biological diversity, sustainable management of forests and fisheries, and trade in forest and fish products; and commitments related to transparency, cooperation, and review of sustainability impacts. The TPP, and more recently, the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP), signed in 2018, also included a large number of ERPs covering a broad range of issues, including wildlife trade, marine fisheries, environmental goods and services, biodiversity, transition to a lower-emission economy, MEAs, transparency, and access to remedies for environmental harm.23 While the more recent PTAs have, on average, started to close the gap in environmental content between North-North and North-South agreements, the gap between these and South-South agreements remains large. 18.3.2 Dynamic nature of environment-related provisions Most PTAs with several types of ERPs were initially negotiated between developed countries. Over time, some developing countries also began to include ERPs in their PTAs negotiated mostly with developed countries. In parallel, some countries that have in the past incorporated ERPs in some of their PTAs stopped doing so. This explains why the relationship between the number of signed PTAs with ERPs and the average number of ERPs is non-linear (see Annex Figure 18.A.1). For example, PTAs negotiated by the United States, Canada, EU, EFTA, and the Republic of Korea include a high average number of types of ERPs (Figure 18.3). The number of types of ERPs in PTAs tends also to be relatively higher in some Latin American countries, such as Chile, Colombia, and Peru, and in some East Asian and Pacific countries, such as Australia, New Zealand, and Singapore. Similarly, several countries in Africa and the Caribbean have negotiated one or more PTAs with a relatively high average number of ERPs. These countries include members of the Common Market for Eastern and Southern Africa (COMESA) and the Caribbean Community (CARICOM). By contrast, the average number of ERPs has decreased in some Latin American countries and in West and Central African countries. 23 Following the United States’ decision to withdraw from the TPP, the remaining 11 parties to the agreement forged ahead with a new version, known as the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP). Several provisions found in the original TPP have been suspended under the CPTPP. The environment chapter of the CPTPP is identical to the environment chapter of the original TPP, except for one provision found in the TPP requiring the parties to take action to address violations to the wildlife trafficking laws of non-parties to the TPP. 566 Environmental Laws Figure 18.3: Average number of types of environment-related provisions in PTAs by country Source: Deep Trade Agreements Database. The decision to include ERPs not only changes over time but also depends on the countries that take part in the negotiations. The relationship between the content of ERPs and the level of economic development of the parties to the PTA is highly non-linear (Figure 18.4). PTAs negotiated by parties at different levels of economic development tend to include a higher number of types of ERPs, as do PTAs between countries in different regions. The number of ERPs also tends to be greater in PTAs representing considerable market size. However, many PTAs do not follow these patterns, highlighting important idiosyncratic trends. Similarly, the content of ERPs tends to be more comprehensive in PTAs negotiated between countries experiencing large differences in environmental performance, suggesting that the inclusion of a large number of ERPs might aim at ensuring a level playing field between the parties. In particular, the number of types of ERPs tends to be higher when the parties to the PTA have different levels of environmental quality, measured by the Environmental Performance Index (EPI).The number of types of ERPs also tends to be greater when one of the parties to the PTA has signed more MEAs than the other party. However, the number of types of ERPs tends to be lower when the difference in environmental quality is extremely large. Similar patterns arise for differences in the level of CO2 emissions (in kilotons) and CO2 per capita emissions. 567 Handbook of Deep Trade Agreements Figure 18.4: Relationship between the number of types of environment-related provisions in PTAs and level of economic development and environmental quality Sources: World Development Indicators 2018; Yale University’s Environmental Quality Index 2017; International Environmental Agreements Database Project 2017; Deep Trade Agreements Database. Note: The similarity GDP index between countries i and j is defined as 1−[GDPi/(GDPj + GDPj)]2−[GDPj/ (GDPi + GDPj)]2 and ranges from 0 (dissimilar) to 0.5 (similar) (see Helpman 1987).The market size is measured by the sum of the parties’ real GDP (in US$ millions). Data on GDP and GDP per capita are of the PTA’s date of signature. The environmental performance index (EPI) ranks countries’ performance on protection of human health and protection of ecosystems, ranging from 0 (worst) to 100 (best). The number of MEAs includes agreements, amendments, and other modifications and protocols. 18.3.3 Scope of environment-related provisions in PTAs The scope, type, and number of ERPs have expanded considerably over the years. Provisions recognizing and/or establishing a balance between environmental and trade/investment goals are the most common, and are included in 270 PTAs (Figure 18.5). The second- and third-most-common types relate to enforcement mechanisms and environmental goals or objectives. A growing number of PTAs also have provisions related to external assistance, general areas of environmental protection, MEA compliance, and participation in promoting environmental objectives. North-North and North-South PTAs tend to incorporate a broader range of types of ERPs in addition to those related to balancing environmental and trade/investment goals. South-South PTAs tend to include mainly provisions related to environmental goals, the balance between environmental and trade goals, and enforcement mechanisms. 568 Environmental Laws Figure 18.5: Scope of types of environment-related provisions Source: Deep Trade Agreements Database. Note: The figures listed at the left and right of each bar represent, respectively, the number and percentage of PTAs with the stated types of ERPs. Most ERPs are by definition WTO-X (WTO-plus) provisions, as they set commitments that go beyond the WTO agreements. In fact, the environmental exception clause is the main type of WTO-like ERPs mirroring the general exceptions under Article XX of the GATT-1994 or Article XIV of the GATS. Another WTO-like ERPs is the right to regulate environmental issues, recognized for instance in the WTO agreement on Technical Barriers to Trade. In practice, most PTAs with EPRs include a combination of both WTO-like and WTO-plus ERPs. As discussed above, North-North and North-South PTAs are more likely to include a greater variety of ERPs, and these tend to be WTO-plus provisions. Conversely, South-South PTAs include fewer WTO-plus ERPs. With the exception of ERPs related to enforcement mechanisms, the benefits associated with many ERPs are not excludible to non-parties to the PTA. The next subsections discuss in greater detail the trends of the different types of ERPs, which often mask significant heterogeneity. 18.3.3.1 Environmental goals and objectives A relatively large and increasing number of PTAs include aspirational provisions that specify the agreement’s environmental objectives in the preamble or objectives section. An increasing number of PTAs also include provisions setting out general obligations of environmental cooperation to support achievement of those objectives. These objectives-related provisions are often complemented by provisions recognizing or reiterating the parties’ rights to regulate 569 Handbook of Deep Trade Agreements environmental matters. Other complementary provisions encourage or commit the parties to ensure that their domestic laws provide high levels of environmental protection and continue to improve them. A limited number of PTAs go farther and call for regulatory cooperation or harmonization in environmental regulations (Figure 18.6). Figure 18.6: Provisions on environmental goals or objectives Source: Deep Trade Agreements Database. Note: The figures listed at the left and right of each bar represent, respectively, the number and percentage of PTAs with the stated types of ERPs. 18.3.3.2 Balance between environmental and trade/investment goals The most common broad type of ERPs refers to the balance between environmental and trade/investment goals, and the most common form of this broad type is the exception clause (Figure 18.7).This provision enables the parties to derogate from the PTA’s obligations to fulfill environment-related policy objectives, including the protection of animal or plant life or health or the conservation of exhaustible natural resources. A similar provision is sometimes included explicitly in the investment chapter. An increasing number of North-North and North-South PTAs include provisions prohibiting the dilution of environmental protection to promote investment and/or trade. In particular, several PTAs specify that it is inappropriate to encourage investment and/or trade by relaxing, weakening, or reducing the protection afforded in domestic environmental law. This provision is often complemented by another provision stipulating that the parties should not, may not, or shall not waive or otherwise derogate from their environmental laws or offer to do so as an encouragement for trade and/or investment in 570 Environmental Laws Figure 18.7: Provisions on environmental goals or objectives Source: Deep Trade Agreements Database. Note: The figures listed at the left and right of each bar represent, respectively, the number and percentage of PTAs with the stated types of ERPs. their territory. A related provision requires that the parties not fail to effectively enforce their environmental laws in a manner that would affect trade or investment between the parties. A limited number of PTAs, often those negotiated between developed and developing countries, establish commitments requiring the parties to take science into account when preparing and implementing environmental policies and regulations, and to prepare an assessment of the PTA’s impacts on the environment. Other PTAs provide explicitly for differential and greater liberalization of trade in environmental goods. 18.3.3.3 General environmental protection areas A limited but increasing number of PTAs establish commitments on specific environmental issues, the language and scope of which differ widely across agreements. Some of the most common provisions are related to fisheries and require the parties to implement fisheries management. More recent but much less common provisions establish differential restrictions on fishing subsidies, or require the parties to take measures related to shipping pollution and protection of marine species. Other environmental issues covered in a small but increasing number of PTAs include commitments to promote and improve renewable energy, energy efficiency, water management, sustainable forestry management, and biodiversity, as well as to protect and prevent illegal trade of endangered wildlife. Less common provisions relate to the control of hazardous and toxic waste and ozone-depleting substances. 571 Handbook of Deep Trade Agreements 18.3.3.4 Compliance with multilateral environmental agreements An increasing number of North-North and North-South PTAs set out a broad range of provisions related to compliance with MEAs that address national, regional, or global environmental problems. Some provisions reaffirm the importance of MEAs, others reaffirm the parties’ obligations under MEAs, and still others call on the parties to adopt measures required to comply with MEA obligations. While some of these provisions refer to MEAs in general, others apply to specific MEAs.These include the Convention on International Trade in Endangered Species (CITES), the Montreal Protocol on ozone-depleting substances, and the Basel Convention on the transboundary movement of hazardous waste. Other MEAs, less often mentioned explicitly, include the UN Convention on the Law of the Sea, the UN Fish Stocks Agreement, the Inter-American Tropical Tuna Convention (ITTC), the International Convention on the Regulation of Whaling (IWC), the Convention on Conservation of Antarctic Marine Living Resources (CCAMLR), the International Convention for the Prevention of Pollution from Ships (MARPOL), and the Ramsar Convention on Wetlands. Another type of provision related to MEAs, found mostly in PTAs negotiated by Canada and the United States, clarifies the relationship between the PTA and any referenced MEAs in the event that the obligations set by those agreements are inconsistent. Such a provision generally specifies that the obligations under particular MEAs either are not affected, are covered by the exception clause, or would prevail in case of inconsistency with the PTA.The number of provisions on MEA compliance tends to be high when one of the parties to the PTA has signed more MEAs than the other party. 18.3.3.5 External assistance An increasing number of PTAs identify environmental cooperation as an objective; and those negotiated between developed and developing countries often foresee or provide for technical assistance, financial assistance, and/or capacity building on environmental matters. Environment-related cooperation may include information exchange, training activities, exchange of professionals, joint projects, and conferences. A few PTAs also specify transfer of technology as a possible form of cooperation.While some cooperation provisions refer to the environment in general without specifying any details, others set out technical cooperation on specific environmental issues. Several PTAs further clarify that any environment-related cooperative activities are subject to the availability of funds and/or human resources. 18.3.3.6 Participation, transparency, and cooperation in promoting environmental objectives A limited but increasing number of PTAs, mainly between high- and middle-income countries, include provisions aimed at fostering participation in promoting environmental objectives. In addition, some PTAs with a comprehensive environment chapter, cooperation chapter, and/or a side agreement on environmental cooperation establish specific mechanisms, including institutional bodies or committees, to discuss and oversee the implementation 572 Environmental Laws of environment-related commitments (Figure 18.8). The nature, structure, and functions of these mechanisms vary greatly across PTAs. One specific function, foreseen in several US and Canadian PTAs, is to determine whether a written submissions or questions filed by the parties’ citizens alleging that one of the parties is failing to effectively enforce its environmental laws and regulations merits a response from the concerned country. To help promote effective participation, several PTAs include provisions on transparency related to environmental laws and regulations and environmental management. One of the most common provisions requires the parties to ensure that environmental laws and regulations are promptly published. A complementary but less common provision further commits the parties to publish or make available the draft of any future environmental law or measure for comments. In some PTAs, the environment-related transparency provisions apply to specific environmental issues such as biodiversity and forestry. Another relatively common provision identifies improving access to environmental information as a potential area of cooperation with other parties to the PTA. In addition to transparency, many PTAs include provisions related to engaging the public in environment-related decision-making, environmental impact assessment, enforcement of environmental laws and policies, and/or implementation of environment-related activities and programs. Several PTAs include provisions aimed at promoting public participation through advisory committees comprising representatives of environmental, business, and civil society organizations to advise on the implementation of the PTA’s environment chapter or Figure 18.8: Provisions on participation in promoting environmental objectives Source: Deep Trade Agreements Database. Note: The figures listed at the left and right of each bar represent, respectively, the number and percentage of PTAs with the stated types of ERPs. 573 Handbook of Deep Trade Agreements environmental side agreement. Other provisions list improving, promoting, or strengthening mechanisms for public participation in environmental matters as potential cooperation issues. 18.3.3.7 Enforcement mechanisms In addition to transparency and public participation, a limited but increasing number of PTAs include provisions related to environmental governance, including enforcement of environmental commitments (Figure 18.9). Several PTAs commit the parties to ensuring that judicial, quasi-judicial, or administrative proceedings are available to sanction or remedy violations of domestic environmental laws. Beyond the availability of these proceedings, several PTAs, mainly those negotiated by the United States and Canada, require each party to ensure that its citizens have appropriate access to these proceedings, including the right to request that competent authorities investigate alleged violations of domestic environmental laws. A large share of these PTAs also commit the parties to providing appropriate and effective sanctions or remedies for these violations, such as penalties, fines, imprisonment, injunctions, suspension of activities, or compliance agreements. Figure 18.9: Provisions on enforcement mechanism Source: Deep Trade Agreements Database. Note: The figures listed at the left and right of each bar represent, respectively, the number and percentage of PTAs with the stated types of ERPs. Enforcement mechanisms are also provided in a number of agreements, mainly through DS procedures if consultations fail to achieve satisfactory results. These procedures enable the parties to identify, demonstrate, and retaliate against violations of an agreement within 574 Environmental Laws a specified framework. The language and enforceability of ERPs differ significantly across PTAs, making it particularly difficult to identify patterns. Very few types of ERPs are worded in non-binding terms. Similarly, only a small number - mostly those related to environmental goals or objectives - are worded in hortatory terms. Most other types of ERPs are formulated in binding terms, although some binding provisions in some PTAs are excluded from DS procedures. A few binding ERPs, particularly those found in investment chapters, such as the commitment to not derogate environmental policies as an encouragement for investment, are covered by state-private and state-state DS procedures. Some PTAs, mainly those involving Canada, the EU and the United States, further establish special environmental state-to-state DS procedures for some or all provisions of the environment chapter or side agreement. In some of these PTAs, environmental DS provisions require the establishment of a roster of individuals to serve on an arbitration panel; set out the qualifications for panelists, including circumstances when their participation on the panel is not allowed; and the timeframe for selecting a panel to hear a pending dispute. Some PTAs further include specific provisions on how to resolve a dispute involving an obligation under a covered MEA, such as deferring to the interpretative guidance under the relevant MEA. Enforcement mechanisms also differ in terms of the possibility of suspending trade concessions and imposing monetary sanctions. The environment chapters of most PTAs signed by the United States are covered by state-to-state DS procedures, and violations of the agreement are potentially subject to trade sanctions. Conversely, the DS proceedings established under the sustainable development chapter of most recent PTAs to which the EU is a party exclude the possibility of imposing monetary remedies or trade sanctions. Most other PTAs with an environment chapter explicitly exclude recourse to the DS procedures to address any matter arising under the environment chapter. 18.3.4 Heterogeneity of environment-related provisions As noted throughout this study, the scope and language of ERPs in PTAs vary significantly across agreements, including different agreements negotiated by the same country. Despite this heterogeneity, most PTAs with ERPs share at least one common type of provision with another agreement, often in the form of preamble language and/or an exception clause. The universe of PTAs with ERPs other than preamble language and/or an environmental exception clause and their level of similarity with each other is shown in Figure 18.10.24 Each 24 For visual convenience, PTAs with only preamble language and/or an environmental exception clause are excluded. In other words, any pairs of PTAs that only have preamble language and/or an environmental exception in common are not shown. 575 Handbook of Deep Trade Agreements symbol (square, diamond, or triangle) represents a PTA. The size of the symbol represents the number of types of ERPs in the PTA, while the width and opacity of the edges connecting two PTAs reflect, respectively, the number of common ERPs and their level of similarity, measured by the Jaccard index. For each possible pair of PTAs, the Jaccard index compares the ERPs of each PTA to see which ERPs are shared and which are distinct. The number of matching types of ERPs is then normalized into an index ranging from 0 to 1. The closer the Jaccard index is to one (or zero), the more (or less) both PTAs include the same type(s) of ERPs. Even though the level of similarity between PTAs is higher for some pairs of agreements, it tends to be much lower for most pairs of PTAs (see Annex Figure 18.A.2). Even in PTAs negotiated by the same country, the type of ERPs varies significantly. North-South PTAs tend to share more common types of ERPs than North-North and South-South PTAs. Similarly, South-South PTAs tend to share more similarities with North-South PTAs than with North-North PTAs. More generally, the level of similarity tends to increase with the number of ERPs included in the PTAs. This suggests a relative convergence of the main types of ERPs covered in PTAs negotiated in the last 10 years, although the language may differ significantly across agreements. Figure 18.10: Similarity of environment-related provisions in PTAs Source: Deep Trade Agreements Database. Note: The size of the shape associated with each PTA represents the number of types of ERPs.The width of each edge measures the number of similar types of EPRs between two PTAs, while the opacity of each edge captures the level of similarity (measured by the Jaccard index). For visual convenience, PTAs with only preamble language and/or an environmental exception clause are excluded. 576 Environmental Laws 18.4. CONCLUSIONS The last 30 years have witnessed a rapid increase in the number of PTAs and intensification of their coverage of the environment. Most PTAs include an environmental exception, similar to the general exceptions under Article XX of the GATT-1994. However, the scope, type, and number of ERPs have changed and grown considerably over the years. Most ERPs differ markedly from WTO agreements by covering environmental issues that go beyond the current WTO framework. PTAs negotiated between developed countries and between developed and developing countries tend to include the highest number of ERPs. High-income countries appear to be the primary proponents of including detailed ERPs in PTAs. Yet, several developing countries, in particular those that have already signed PTAs with high-income countries incorporating ERPs, have also increasingly incorporated ERPs into their trade agreements with other developing countries. The scope and level of commitments of these ERPs are, however, usually not as detailed as those found in PTAs negotiated between developed and developing countries. Overall, the evolution of the different types of ERPs incorporated in PTAs reflects a dynamic context in which PTAs often have a demonstration effect, enabling countries to negotiate and devise new ways to address emerging issues and challenges. ACKNOWLEDGMENTS We would like to thank Elizabeth Schere and Fatima Quraishi for their assistance with the construction of the database as well as Jennifer Hillman and the participants in the World Bank’s workshop on the Evolution of Deep Trade Agreements for helpful comments and suggestions. 577 Handbook of Deep Trade Agreements REFERENCES Anuradha, R.V. 2011. Environment. In Preferential trade agreement Gallagher, P. and Y. Serret. 2011. Implementing regional trade policies for development: A handbook, edited by J.P. Chauffour and agreements with environmental provisions: A framework for J.C. Maur. Washington, DC: World Bank. evaluation. OECD Trade and Environment Working Paper 2011/06, OECD Publishing, Paris. Baghdadi, L., I. Martinez-Zarzoso, and H. Zitouna. 2013. Are RTA agreements with environmental provisions reducing emissions? George, C. 2014. Environment and regional trade agreements: Journal of International Economics 90(2):378-90. Emerging trends and policy drivers. OECD Trade and Environment Working Paper 2014/02, OECD Publishing, Paris. Bastiaens, I., and E. Postnikov. 2015. Environmental provisions in EU and US trade agreements and regulatory change in the Government of New Zealand. 2001. Framework for developing world. Paper presented at 8th Annual Conference integrating environment standards and trade agreements. on the Political Economy of International Organizations, Berlin, February 12-14. http://wp.peio.me/wp-content/uploads/PEIO8/ Helpman, E. 1987. Imperfect competition and international Bastiaens,%20Postnikov%2029.09.2014.pdf. trade: Evidence from fourteen industrial countries. Journal of the Japanese and International Economies 1(1):62-81. Berger, A., C. Brandi, and D. Bruhn 2016. Environmental provisions in preferential trade agreements: Comparing the Horn, H., P. C. Mavroidis, and A. Sapir. 2010. Beyond European and emerging markets approach. https://www. the WTO? An anatomy of EU and US preferential trade oefse.at/fileadmin/content/Downloads/tradeconference/ agreements. World Economy 33(11):1565-88. BergerBrandiBruhn_Green_PTAs_Jan_16.pdf. Jinnah, S., and E. Morgera. 2013. Environmental provisions Berger, A., C. Brandi, D. Bruhn, and M. Chi. 2017. Towards in American and EU free trade agreements: A preliminary “greening” trade? Tracking environmental provisions in the comparison and research agenda. Review of European, Comparative preferential trade agreements of emerging markets. Discussion & International Environmental Law 22(3):324-39. Paper 2/2017, German Development Institute, Bonn. Johnson, T. 2015. Information revelation and structural Bourgeois, J., K. Dawar, and S. J. Evenett. 2007. A comparative supremacy: The World Trade Organization’s incorporation analysis of selected provisions in free trade agreements. Technical of environmental policy. Review of International Organizations Report, European Commission, Brussels. 10:2:1-23. Colyer, D. 2011. Green trade agreements. London: Palgrave Monteiro, J.-A. 2016. Typology of environment-related Macmillan. provisions in regional trade agreements. WTO Staff Working Paper ERSD-2016-13, World Trade Organization, Geneva. Copeland, B. R., and M. Scott Taylor. 1994. North-South trade and the environment. Quarterly Journal of Economics 109(3):755-87. Morin, J.-F., A. Dür, and L. Lechner. 2018. Mapping the trade and environment nexus: Insights from a new dataset. Cosbey, A. 2014. Inside CETA: Unpacking the EU-Canada free Global Environmental Politics 18(1): 122-39. trade deal. Biores 8(9). Organization of American States. Environmental Provisions Downs, G., D. Rocke, and P. Barsoom. 1996. Is the good news in Trade Agreements database. http://www.oas.org/en/ about compliance good news about cooperation? International sedi/dsd/ELPG/aboutELPG/database.asp. Organization 50(3): 379-406. Tébar Less, C., and J. A. Kim. 2008. Checklist for negotiators Dür, A., L. Baccini, and M. Elsig. 2014. The design of of environmental provisions in regional trade agreements. international trade agreements: Introducing a new dataset. OECD Trade and Environment Working Papers 2008/02. Review of International Organization 9:353-75. United States Trade Representative. 2007. May 2007 European Commission. 2016a. Environmental and trade and Bipartisan Agreement on Trade Policy. https://ustr.gov/ external relations. http://ec.europa.eu/environment/integration/ sites/default/files/uploads/factsheets/2007/asset_upload_ trade_en.htm. file127_11319.pdf. ———. 2016b. The environmental goods agreement: Liberalising World Trade Organization. 2011. The WTO and preferential trade in environmental goods and services. http://trade.ec.europa. trade agreements: From co-existence to coherence. Geneva: World eu/doclib/press/index.cfm?id=1116. Trade Organization. 578 Environmental Laws ANNEX Annex 18A: Environment-related provisions template (1/2) I. Environmental Goals/Objectives - Does the agreement specify an objective of environmental protection or sustainable development? - Does the agreement specify an objective of high levels of environmental protection? - Does the agreement specify a general obligation of environmental cooperation? - Does the agreement call for regulatory cooperation or harmonization in environmental regulation? - Does the agreement preserve the right to regulate in the environment? II. Balance between Environmental and Trade/Investment Goals - Does the agreement provide for a general exception to other obligations for environmental reasons? - Does the investment chapter provide for an environmental exception? - Does the agreement prohibit dilution of environmental protection to promote trade? - Does the agreement prohibit dilution of environmental protection to promote investment? - Does the agreement provide for differential and greater liberalization of trade in environmental goods? - Does the agreement require states to take science into account when preparing and implementing environmental regulation? - Does the agreement require states to prepare an environmental impact assessment of the PTA? III. Enforcement Mechanism - Does the agreement require states to maintain judicial or administrative proceedings for enforcement of environmental regulation? - Does the agreement subject environmental provisions to general state-to-state DS? - Does the agreement provide special environmental state-to-state DS? - Does the agreement provide international remedies of compensation or retaliation for violation of environmental provisions? IV. External Assistance - Does the agreement provide for technical assistance/financial assistance/capacity building specifically in the environmental area? V. General Environmental Protection Areas - Does the agreement require states to control ozone-depleting substances? - Does the agreement require states to prevent pollution by ships? - Does the agreement require states to implement fisheries management? - Does the agreement require states to take measures for conservation of specified marine species? - Does the agreement provide for differential restrictions of fishing subsidies? - Does the agreement require states to protect wild fauna and flora at risk and/or prevent illegal trade in these species? - Does the agreement require measures to prevent deforestation and/or require sustainable trade practices in forest products? - Does the agreement prohibit the dumping of hazardous and toxic wastes? - Does the agreement require states to promote and protect biodiversity? - Promotion of renewable energy and improving energy efficiency? - Does the agreement require states to implement water management? 579 Handbook of Deep Trade Agreements Annex 18A: Environment-related provisions template (2/2) VI. Multilateral Environmental Agreement (MEA) Compliance - Does the agreement require states to comply with MEAs generally? - Does the agreement specify supremacy of MEA obligations over PTA obligations? - Does the agreement require states to comply with CITES? - Does the agreement require states to comply with the Montreal Protocol on Ozone- Depleting Substances? - Does the agreement require states to comply with the Basel Convention on the Control Transboundary Movements of Hazardous Wastes and Their Disposal? of - Does the agreement require states to comply with the Inter-American Tropical Tuna Convention? - Does the agreement require states to comply with the International Convention for the Prevention of Pollution from Ships (MARPOL)? - Does the agreement require states to comply with the International Convention on the Regulation of Whaling? - Does the agreement require states to comply with the Ramsar Convention on Wetlands? - Does the agreement require states to comply with the Convention on Conservation of Antarctic Marine Living Resources? - Does the agreement require states to comply with the UN Fish Stocks Agreement, the FAO Code of Conduct for Responsible Fisheries, the 1993 FAO Agreement to Promote Compliance with International Conservation and Management Measures by Fishing Vessels on the High Seas (Compliance Agreement), and the 2001 IUU Fishing Plan of Action/IUU measures in general? - Does the agreement require states to comply with the 2005 Rome Declaration on IUU Fishing; the Agreement on Port State Measures to Prevent, Deter and Eliminate Illegal, Unreported and Unregulated Fishing of 2009; as well as instruments establishing and adopted by Regional Fisheries Management Organizations (RFMOs)? - Does the agreement require states to comply with the UN Convention on the Law of the Seas of 1982? - Does the agreement require states to comply with the UN Conference on Environment and Development? - Does the agreement require states to comply with the UN Environment Program? - Does the agreement require states to comply with the International Energy Program? VII. Participation in Promoting Environmental Objectives - Does the agreement establish an intergovernmental committee on environment? - Does the agreement require states to facilitate civil society involvement and/or establish a forum on trade and environment? - Does the agreement include special obligations of transparency in the environmental field? - Does the agreement include private rights to make submissions regarding environmental provisions? 580 Environmental Laws Annex Figure 18.A.1: Dynamic relationship between the number of PTAs and average number of environment-related provisions Source: Deep Trade Agreements Database. Annex Figure 18.A.2: Histogram of the Jaccard Similarity Index of environment-related provisions Source: Deep Trade Agreements Database. 581 Handbook of Deep Trade Agreements 582 Labor Market Regulations CHAPTER 19 Labor Market Regulations D. Raess and D. Sari 583 Handbook of Deep Trade Agreements CHAPTER 19 Labor Market Regulations D. Raess* and D. Sari† * World Trade Institute, University of Bern, Bern, Switzerland † University of Geneva, Geneva, Switzerland; and Harvard Law School, Cambridge, Massachusetts, United States CONTENTS 19.1. INTRODUCTION 585 19.2. CONCEPTUAL AND ANALYTICAL APPROACH TO MAPPING LABOR PROVISIONS IN PTAs 586 19.2.1 First generation of mapping efforts 586 19.2.2 Scope and structure of the coding scheme 587 19.2.3 The five main categories of labor provisions 589 19.2.3.1 Aspirational labor provisions 589 19.2.3.2 Substance-related labor provisions 590 19.2.3.3 Investment-related labor provisions 592 19.2.3.4 Cooperation-related labor provisions 592 19.2.3.5 Institution-related labor provisions 593 19.2.3.6 Cross-cutting category: enforceability 594 19.3. DESCRIPTIVE ANALYSIS OF LABOR PROVISIONS IN PTAs 596 19.3.1 The rise of LPs in PTAs 596 19.3.2 Types and enforceability of LPs in PTAs 599 19.3.3 LPs in PTAs by key players 602 19.3.4 LPs in South-South PTAs 604 19.4. CONCLUSIONS 605 ACKNOWLEDGMENTS 605 REFERENCES 606 The approach presented in this chapter builds on the Labor Provisions in Trade Agreements (LABPTA) dataset (Raess and Sari 2018). LABPTA with its original methodology was developed in the context of a two-year research project titled “A Social Clause through the Back Door: Labor Provisions in Preferential Trade Agreements,” funded by the Swiss Network for International Studies (call for projects 2014, project coordinator Damian Raess; see https://snis.ch/project/social-clause-through-back-door/). 584 Labor Market Regulations 19.1. INTRODUCTION The World Bank’s Deep Integration project aims to map the content of disciplines in different areas of trade regulation in order to generate a new database on deep trade agreements. This chapter covers labor market protections (or regulations, depending on one’s perspective) in trade agreements, which have gained unexpected ground in international labor regulation over the past three decades. Preferential trade agreements (PTAs) signed in recent decades have started to link the benefits of better market access to the recognition or enforcement of internationally recognized labor rights. The inclusion of labor provisions (LPs) under the international trade regime was first discussed during negotiations for the International Trade Organization (ITO), as part of the Bretton Woods negotiations, in 1945 and 1946. According to Article 7 of the Havana Charter for the International Trade Organization: The Members recognize that measures relating to employment must take fully into account the rights of workers under inter-governmental declarations, conventions and agreements. They recognize that all countries have a common interest in the achievement and maintenance of labour standards related to productivity, and thus in the improvement of wages and working conditions as productivity may permit.The Members recognize that unfair labour conditions, particularly in production for export, create difficulties in international trade, and, accordingly, each Member shall take whatever action may be appropriate and feasible to eliminate such conditions within its territory (cited in Alben 2001, p. 1431). Efforts to enact the ITO, however, were superseded by adoption of the General Agreement on Tariffs and Trade (GATT) in 1947. The GATT did not address workers’ rights except for a provision prohibiting the import of goods made with prison labor (Art. XX (e)). The debate on workers’ rights reemerged with the establishment of the World Trade Organization (WTO) in 1994. In spite of intense discussions, the linkage of international trade to basic labor rights was opposed by developing countries, which argued that such provisions were protectionist and a new form of discriminatory treatment of their products. The discussion ultimately led to the adoption of the Singapore Declaration in 1996 , which named the International Labour Organization (ILO) as the competent body for setting and dealing with labor standards, thus keeping those standards out of the WTO’s agenda. It was then that PTAs started to be seen as a way to protect workers’ basic rights in the context of international trade, resulting in a rapid increase in the inclusion of LPs in trade agreements. Governments in the North have used LPs, particularly with their Southern partners, to address unfair working conditions in the South and enable their own workers and businesses to compete on a level playing field. On a global scale, strong trade unions have been one of main driving forces for the inclusion of more and increasingly stringent labor provisions in PTAs (Raess, Dür, and Sari 2018). Section 19.2 introduces the methodological approach adopted to map LPs in PTAs. Section 19.3 provides a description of the main patterns and trends that emerge from the data. Section 19.4 concludes. 585 Handbook of Deep Trade Agreements 19.2. CONCEPTUAL AND ANALYTICAL APPROACH TO MAPPING LABOR PROVISIONS IN PTAs 19.2.1 First generation of mapping efforts Scholars have only recently started to focus their attention on the variation in the design of PTAs to explain the causes and consequences of such variation and to consider how various provisions in PTAs relate to multilateral rules (e.g., Dür and Elsig 2015; Acharya 2016).To date, however, there has been no systematic or detailed mapping of the content of labor provisions in PTAs. All too often, the same platitudes are being offered, such as US trade agreements contain enforceable labor standards whereas the EU agreements do not (e.g., Postnikov and Bastiaens 2014), which do not do justice to the broadening scope of enforceability in US PTAs or the recent evolution of EU PTAs, and omits the design developed by other key players, such as Canada, New Zealand, or the EFTA. Two recent projects have coded labor standards as part of a broader effort to map nontrade issues (NTIs) in PTAs; but because these projects covered several disciplines, the coding of the LPs lacked depth. First, Lechner (2016) constructs a novel dataset of NTI design for 474 PTAs signed since 1990. The coding includes provisions relating to civil and political rights, environmental protection, and economic and social rights, with the latter covering 72 items related to the right to work, rights at work, the right to education, the right to development, and the right to health. Of these 72 items, 25 are substantive, institutional, or cooperation commitments related to labor issues.Adopting the concept of legalization which distinguishes between the three dimensions of obligation, precision, and delegation (Abbott et al. 2000), one limitation is that extensive obligation and delegation are coded in relation to economic and social rights in general, so that we cannot know with certainty whether obligation and delegation applies effectively to LPs when they do for the overarching category economic and social rights. Second, Milewicz’s (2016; see also Milewicz et al. (2018) dataset on NTIs includes human rights, labor rights, environment, corruption, security, and democracy, covering 522 PTAs signed between 1951 and 2009. The broad definition of labor standards in this project includes not only labor rights and conditions but also social security rights dealing with sickness, invalidity, old age, industrial accidents, and unemployment. While the template distinguishes between references in the preamble and the main treaty text, the coding of the individual NTIs in the main text remains crude. Indeed, regarding labor issues, the template only captures the presence/absence of references to labor standards, on the one hand, and to ILO Conventions, on the other. A few projects have focused on LPs in PTAs. The ILO has been at the forefront of the analysis of the social dimension of trade (and investment) agreements (see ILO 2009, 2013, 2016). While detailed in its insights into the design and the effectiveness of LPs 586 Labor Market Regulations in PTAs, the ILO work has two main limitations; namely, it has failed to produce (a) a comprehensive template of LPs in PTAs and (b) a systematic coding of the content of LPs in all existing PTAs. Kamata (2016) investigates the impact of labor provisions on working conditions, based on a sample of 223 PTAs from 1995 to 2011. The study classified PTAs with labor clauses into two categories: (a) those with provisions that demand, urge, or expect the signatory countries to harmonize their domestic labor conditions and regulations with internationally recognized standards; and (b) those that stipulate issues on which the signatory countries will cooperate, along with dispute settlement procedures for labor issues. While this classification is an improvement over the binary coding of LPs, it remains too rudimentary to adequately capture the scope of LPs in PTAs. For instance, it does not provide any details about whether the commitments are specific to any particular standards; it conflates soft and hard mechanisms such as cooperation and enforceability; and it does not distinguish among different degrees of enforceability. Finally, under the aegis of the World Bank, Hofmann, Osnago, and Ruta (2017) constructed a database with a detailed assessment of the content of 279 PTAs signed between 1958 and 2015. The methodology covered 52 policy areas. Under the WTO-X policy areas (those not yet regulated by the WTO), the database contains a separate category for labor market regulation, referring to provisions that pertain to (a) national labor markets and affirm ILO commitments (a single item coded in a binary mode); and (b) the enforcement of these provisions (coded on a 0-2 scale). This approach, however, also lacks depth and detail. In short, a more precisely defined and fine-grained approach to the mapping of LPs in PTAs is needed to better understand trade-labor linkages on a global scale. Combined with readily available country-level data on the protection of labor rights (e.g., Labour Rights Indicators,1 the CIRI Human Rights Data Project,2 and the ILO’s ratification data3), a more comprehensive set of data on labor provisions in trade agreements would make it possible to test a variety of hypotheses regarding the consequences of such provisions. 19.2.2 Scope and structure of the coding scheme Labor provisions vary in terms of content and stringency as well as where in the PTA they are located (preamble, text, side agreement, Memorandum of Understanding). Some PTAs include, typically in their preambles, only an aspirational goal to improve working 1 Kucera and Sari (2019), http://labour-rights-indicators.la.psu.edu. 2 Cingranelli, Richards, and Clay (2014), http://www.humanrightsdata.com. 3 See http://www.ilo.org/dyn/normlex/en/f?p=NORMLEXPUB:11001:0::NO. 587 Handbook of Deep Trade Agreements conditions. Others specify commitments to internationally recognized labor standards, with or without a reference to the relevant ILO instruments. Still others detail procedures for consultation, dispute settlement, cooperation activities, and/or the institutions with oversight for monitoring and implementation of labor-related commitments. Given that there is no single understanding of LPs in PTAs, the first task is to define what constitutes LPs.This study uses a narrow definition of labor provisions, thereby drawing a distinction between labor and broader social provisions. In this context, labor provisions refer to rules and regulations that aim to protect and/or promote workers’ rights and working conditions, but do not cover social protection issues such as unemployment, old age, health, education, etc.With the exception of aspirational statements about employment creation, the definition also excludes provisions relating to employment and labor market policy, such as training, or “supply-side” measures aimed at better matching labor supply and demand. Finally, the definition used here does not cover the free movement of workers or the treatment of migrant workers. We consider LPs in the treaty texts, and in side agreements on labor and Memorandums of Understanding (MoUs), without distinction of any kind. Side agreements and MoUs, however, should have a clear relationship to the PTA in question – established by a concrete reference to the relevant instrument in the main treaty text – in order to be considered.4 In order to code the wide variety of LPs in PTAs, the coding scheme established for the purpose of this project is structured around five main categories together with the cross-cutting measure of enforceability. The five overarching categories are: (1) Aspirational labor goals/ objectives; (2) Substance-related LPs; (3) Substance-related LPs in relation to investment; (4) Cooperation over LPs; and (5) Institutions overseeing labor commitments. While substance- related LPs have a detailed list of items against which the PTAs are coded, LPs in relation to investment, cooperation, and institutions are captured with a lower level of granularity. LPs in the preamble and the objectives parts of the agreements constitute aspirational statements, which is why we code them separately. Under substantive commitments we list items related to fundamental rights at work, conditions of work, relevant international instruments, and domestic law-related commitments such as non-derogation and effective enforcement. Substance-related LPs in relation to investment include commitments to protect labor rights in the context of investment. Under Cooperation we code the presence of any commitments in relation to provisions on labor-related cooperation. Institution-related LPs are coded for attributes that can determine the effectiveness of monitoring and implementation of labor provisions. Whereas the Substance and Cooperation categories relate to what matters in terms of issues addressed in the relevant legal provisions in the PTAs, the enforcement measure addresses the question what 4 For example, in the China-Peru PTA (2009), under article 161 such relationship is established in the following manner: “The Parties shall enhance their communication and cooperation on labor, social security and environment issues through Memorandum of Understanding on Labor Cooperation between the Government of the People’s Republic of China and the Government of the Republic of Peru.” 588 Labor Market Regulations is the nature and scope of commitments taken in these legal provisions (see Bourgeois, Dawar, and Evenett 2007, p. 11). Based on the World Bank’s Deep Integration template, the enforcement measure is constructed to reflect on the enforceability provided for LPs. Being a cross-cutting issue, enforceability categories are added for each item listed under the five overarching categories. 19.2.3 The five main categories of labor provisions 19.2.3.1 Aspirational labor provisions We code LPs found in the agreement’s preambles and objectives. Although still subject of academic debate (Hulme 2016), LPs in preambles are predominantly considered to differ in terms of their legal effect from those found in other parts of the agreement, as they do not establish specific rights and obligations (and as such cannot directly be subject to dispute settlement), but hold an interpretive role.5 The aims to “improve working conditions” and to “create employment opportunities” are by far the most frequent references to labor standards in preambles and objectives, while other references to labor rights and working conditions remain rather rare.6 Because statements on improving working conditions and other labor rights and conditions refer to the quality of jobs, while statements on creating employment opportunities refer to the quantity of jobs, these references are coded separately. Thus, under this category we code two types of commitment under the following two headings: 1. Aspirational labor goals/protection or promotion of labor standards. Coding rule: 1 if reference to the protection or promotion of labor standards is in preambles and/or objectives parts of the agreement. 2. Aspirational labor goals/creation of employment opportunities. Coding rule: 1 if reference to employment creation is in preambles and/or objectives part of the agreement. Example 1: Switzerland-China (date of entry into force, 2014), MoU, Art.1.1 (Objectives and Scope). This we code under heading 1 because the references to labor standards appear in Objectives. Example 2: Asia Pacific Trade Agreement (2006), Preamble.This we code under heading 2 because the reference to new employment opportunities appears in the preamble. 5 According to Bourgeois et al. (2007, p. 13),“[T]he preamble to an FTA does not contain any binding obligations upon the parties. The statements contained in preambles are not intended to be operative provisions in the sense of creating specific rights or obligations. Rather, the preamble statements offer a context for the signatories’ overall objectives by introducing the agreement, setting out the motives of the contracting parties and the objectives to be accomplished by the provisions of the statutes.” 6 For example, Albania-EFTA PTA (2009), Preamble. 589 Handbook of Deep Trade Agreements 19.2.3.2 Substance-related labor provisions First, we list items related to relevant international labor standards commitments, that is, provisions derived from or related to internationally recognized labor commitments and provisions concerning international instruments containing such commitments. We code nine separate items, as follows: 3. ILO 1998 Declaration on Fundamental Principles and Rights at Work and Its Follow-Up. The Declaration stipulates that all ILO members have an obligation arising from their membership in the Organization to respect, promote and to realize the principles concerning the fundamental worker rights. Coding rule: 1 if reference to the ILO 1998 Declaration on Fundamental Principles and Rights at Work. 4. Freedom of association and collective bargaining rights. Coding rule: 1 if reference to one of the following: (a) freedom of association, right to organize, right to strike, trade union right(s); (b) ILO Convention No. 87 (“Freedom of Association and Protection of the Right to Organise Convention”; (c) collective bargaining including reference to (alternative) dispute resolution mechanism; (d) ILO Convention No. 98 (“Right to Organise and Collective Bargaining Convention”); (e) ILO fundamental labor/worker’s rights/Conventions. 5. Elimination of all forms of forced or compulsory labor. Coding rule: 1 if reference to one of the following: (a) elimination/abolition of forced labor; (b) ILO Convention No. 29 (“Forced Labour Convention”) and/or ILO Convention No. 105 (“Abolition of Forced Labour Convention”); (c) ILO fundamental labor/worker’s rights/Conventions. 6. Effective abolition of child labor. Coding rule: 1 if reference to one of the following: (a) (progressively raise) minimum age for admission to employment or work; (b) (abolition of) child labor; (c) ILO Convention No. 138 (“Minimum Age Convention”); (d) prohibition and elimination of worst forms of child labor; (e) ILO Convention No. 182 (“Worst Forms of Child Labour Convention”); (f) ILO fundamental labor/worker’s rights/Conventions. 7. Elimination of discrimination in respect of employment and occupation. Coding rule: 1 if reference to one of the following: (a) equal remuneration/pay for men and women for work of equal value; (b) ILO Convention No. 100 (“Equal Remuneration Convention”); (c) elimination of discrimination of any form in respect of employment and occupation/work; (d) ILO Convention No. 111 (“Discrimination (Employment and Occupation) Convention”); (e) ILO fundamental labor/worker’s rights/Conventions. 590 Labor Market Regulations 8. Working conditions and terms of employment. Coding rule: 1 if reference to one of the following: (a) conditions of work; (b) working time; (c) wages; (d) health and safety. 9. Other international instruments. We consider three international instruments, namely, the ILO 2008 Declaration on Social Justice for a Fair Globalization; the ILO’s Decent Work agenda; and the UN ECOSOC Ministerial Declaration on Generating Full and Productive Employment and Decent Work for All 2006. Coding rule: 1 if reference to one of the following: (a) ILO 2008 Declaration on Social Justice for a Fair Globalization; (b) Reference to ILO’s Decent Work Agenda; (c) decent work if clear from text it is understood in the way as ILO’s Decent Work Agenda; (d) UN ECOSOC Ministerial Declaration on Generating Full and Productive Employment and Decent Work for All 2006. 10. Internationally recognized labor standards. Given that internationally recognized labor standards can encompass a variety of standards, we consider the coding of reference to internationally recognized labor standards if no definition is provided for such reference in the agreement or if such definition is broader than the eight fundamental labor rights referred to under the 1998 ILO Declaration on the Fundamental Principles and Rights at Work and those considered under conditions of work. Coding rule: 1 if reference to either (a) internationally recognized labor standards or (b) ILO Conventions in general. 11. Corporate Social Responsibility. Coding rule: 1 if reference to one of the following: (a) good corporate governance and corporate social responsibility on labor issues; (b) internationally recognized guidelines and principles relating to good corporate governance and corporate social responsibility on labor issues. Note: labor-related corporate social responsibility references found in investment chapters are coded under substance-related labor provisions in relation to investment (see heading 14 and example 7 below). Example 3: Switzerland-China (2014), MoU, Art.2.1. This we code under heading 3. Example 4: EFTA-Montenegro (2012), CH. 6 on Trade and Sustainable Development, Art. 35(1). This we code under headings 3, 4, 5, 6, and 7. Second, we list items related to domestic laws commitments. Specifically, we code commitments related to non-derogation and to effective enforcement of domestic laws. 591 Handbook of Deep Trade Agreements 12. Non-derogation. Coding rule: 1 if reference to commitment not to encourage trade through the weakening of labor laws by ways of waiving or derogating from domestic labor law. 13. Effective enforcement of domestic laws. Coding rule: 1 if reference to commitment to effectively enforce domestic laws (to the extent as failing to do so would affect trade between the signatory countries). Example 5: New Zealand-Republic of Korea (2015): Ch. 15 on Labour, Art. 15.2. Paragraph 3 is coded under heading 12, whereas paragraph 4 is coded under heading 13. 19.2.3.3 Investment-related labor provisions Regarding Investment-related provisions, we code instances where the PTA requires investors to act in keeping with the protection and promotion of labor standards. Although such references have become more common provisions in PTAs, the extent of the commitments is not yet comparable to provisions agreed upon in relation to trade. The vast majority of investment-related labor provisions concern commitments related to non-derogation, that is, commitment not to encourage investment through the weakening of labor laws by ways of waiving or derogating from domestic labor law. There are only a handful of cases where investment-related labor provisions are referred to under the investment chapter of the PTA, while in the majority of cases such commitments are embedded in the labor chapter. 14. Requirement that investors act in accordance with the protection or promotion of labor standards. Coding rule: 1 if reference to any substance-related labor commitment (i.e., headings 3-13) if those are formulated in relation to investment. Example 6: Japan-Switzerland (2009), Article 101 (Health, Safety and Environmental Measures). Example 7: Canada-Panama (2013), Chapter 9 on Investment, Article 9.17 (Corporate Social Responsibility). 19.2.3.4 Cooperation-related labor provisions Under Cooperation we code the presence of any substantive commitments if those are agreed by the parties as issues over which they aim to cooperate. In praxis, cooperation typically takes the form of technical assistance and capacity building by means of the exchange of information, people, joint research, seminars, etc. This is heading number 15. 15. Reference to cooperation over labor provisions. Coding rule: 1 if reference to one of the following: (a) cooperation over any of the substance- related LPs referred listed under headings 3-11; (b) cooperation over labor laws; (c) cooperation over industrial relations and social dialogue; (d) cooperation over labor administration and inspection; (e) cooperation over gender equality. 592 Labor Market Regulations Example 8: US-Morocco (2006), Chapter 16 on Labor, Article 16.5.1-2 (Labor Cooperation) and Annex 16-A (Labor Cooperation Mechanism). 19.2.3.5 Institution-related labor provisions The final category depicts three institutional features that influence the effective monitoring and implementation of labor provisions. Such items are (a) a separate specialized committee responsible for the implementation and/or supervision of labor commitments (including contact points); (b) third-party consultation/involvement (either the social partners, the ILO, NGOs, or other third-party organizations); and (c) the requirement of the realization of labor-related impact assessment. These are the final headings in our template - numbers 16, 17, and 18, respectively. 16. Reference to a separate specialized committee or contact point for the monitoring and implementation of labor provisions. Coding rule: 1 if reference to a separate specialized committee or contact point established exclusively to monitor and implement labor provisions agreed upon under the PTA. Example 9: EU-Korea (2011), Ch. 13 on Trade and Sustainable Development, Article 13.12.1-3. 17. Reference to third-party (e.g., social partners, civil society organizations, ILO etc.) inclusion in the monitoring and implementation of labor provisions. Coding rule: 1 if PTA allows for the inclusion of third parties in the monitoring and implementation of LPs agreed upon under the PTA. Third-party reference typically includes (a) social partners or workers’ organizations, trade unions, and employers’ organizations; (b) civil society organizations; (c) ILO; (d) other bodies. Note: Inclusion of third parties is coded also in cases where their recommendation is not binding on the signatory parties. Example 10: EU-Korea (2011), Ch. 13 on Trade and Sustainable Development, Article 13.12.4. 18. Reference to ex-post assessment of the impact of labor provisions. Coding rule: 1 if reference to one of the following: (a) a periodic or one-time review of progress after the entry into force of the agreement; b) periodic or one-time ex-post review, monitoring, or assessment of the labor impact of the agreement. Example 11: Dominican Republic-Central America-US (2006), Chapter 16 (Labor), Article 16.4.2 (Institutional Arrangements). Example 12: EU-Colombia and Peru (2012),Title IX (Trade and Sustainable Development), Article 279 (Review of Sustainability Impacts). 593 Handbook of Deep Trade Agreements 19.2.3.6 Cross-cutting category: enforceability Under Enforcement, building on the WB’s template, we focus on the model of dispute settlement (DS) mechanism covering labor-related commitments. In the coding of enforceability, we equally code the DS mechanism that applies to the entire agreement (as long as LPs are covered by it) and labor-specific DS mechanisms that are set out to deal with disputes related exclusively to LPs. Enforceability is defined along a 0 to 3 scale, where the weakest degree covers non-binding and best-endeavor commitments (0); followed by provisions that are binding but not subject to DS (1); by provisions that are binding and subject to state-to-state DS (2); and then by provisions that are binding and subject to private-state DS (3). Evidently we find private-state dispute settlement to apply only to labor provisions in relation to investment found in the investment chapter of the trade agreement, with overall only very few instances of such an enforcement mechanism. Regarding the bindingness of LPs, following a strict legal analysis of the treaty texts, we carefully assess whether or not a substantive labor commitment is a legally binding commitment representing an obligation (e.g., the Parties shall ratify ILO fundamental conventions or shall strive to ratify ILO fundamental conventions). Binding obligations are indicated by the use of terms such as shall, will, agree, undertake, ensure, realize. The term shall is interpreted by courts to be stronger than the term should; and commitments expressed with words such as strive to ensure are weaker than those using ensure. Regarding the issue of state-to-state DS, we rely partially on a WTO taxonomy (Chase et al. 2016). Under the current method, enforceability entails that the parties can resort at a minimum to quasi-judicial arbitration-based DS over labor standards provisions. In other words, PTA members have an “automatic” right7 of access to (often ad hoc) third-party adjudication (or standing judicial courts) in case of a dispute. Last, private-state DS refers to investor-state dispute settlement (ISDS) established under the investment chapters of the PTAs. Given that LPs are rarely featured in the investment chapter and ISDS only concerns provisions agreed upon under the investment chapter, in only one case do we find private-state DS to be applicable to the relevant labor provision (Japan-Mongolia PTA, 2015). 7 That is, PTA members have no right to veto a referral. 594 Labor Market Regulations In short, headings 1-18 are repeated and coded for each of the four enforceability levels (0-3) according to the following coding rules: (0) Non-binding and best-endeavor provisions. Coding rule: 1 if labor-related commitment is not legally binding on the parties (by the use of wording such as may, might, etc.), including provisions that imply a higher degree, but non-binding, commitment in complying with agreed provisions (by the use of wording such as shall strive, shall endeavor, should, etc.). Note: Provisions coded under “0” are not subject either to the general or the labor- specific DS, as defined above (state-to-state or private-state DS). Example 13: Korea-Australia (2014), Chapter 17 on Labour, Article 17.1.1 (General Principles) read together with Article 17.6 (Dispute Settlement). (1) Provisions that are binding but not subject to dispute settlement. Coding rule: 1 if labor-related commitment is legally binding (by the use of wording such as shall, will, agree, etc.), but only when the provision is not subject either to the general or the labor-specific DS, as defined above (state-to-state or private-state DS). Example 14: New Zealand-Taiwan, China (2013), Chapter 16, Trade and Labour, Article 2.1-2 (Key Commitments) read with Article 5.9 (Consultations). (2) Provisions that are binding and subject to state-to-state dispute settlement. Coding rule: 1 if labor-related commitment is legally binding (by the use of wording such as shall, will, agree, etc.) and is subject to either the general or the labor-specific state-to- state DS. Example 15: Trans-Pacific Partnership (2017), Chapter 19 on Labour, Article 19.3.1-2 (Labour Rights) read together with Article 19.15.12 (Labour Consultations). (3) Provisions that are binding and include private-state dispute settlement. Coding rule: 1 if labor-related commitment is legally binding (by the use of wording such as shall, will, agree, etc.) and is subject to the private-state DS. Example 16: Japan-Mongolia (2016), Chapter 10 on Investment: Article 10.13, Settlement of Investment Disputes between a Party and an Investor of the Other Party, and its application to Article 10.17, Health, Safety and Environmental Measures and Labor Standards. 595 Handbook of Deep Trade Agreements 19.3. DESCRIPTIVE ANALYSIS OF LABOR PROVISIONS IN PTAs 19.3.1 The rise of LPs in PTAs Over the past 50 years, the number of PTAs has proliferated, as have the types of issues they cover (e.g., Dür and Elsig 2015). Important issues that intersect with trade and have been included in more recent PTAs include environment and investment, addressed in other chapters of this volume, and workers’ rights, which are the subject of this study.8 PTAs began to include provisions on workers’ rights and working conditions in the late 1980s,9 first as aspirational statements in the preamble or objectives sections (“shallow LPs”), and later through more substantive commitments found in the main text of the PTA (“comprehensive LPs”).While the total number of PTAs without LPs still surpasses the number of PTAs with comprehensive LPs, a steady increase in the pace of adoption of such PTAs can be observed in conjunction with the plateauing of shallow (and no LP) PTAs (Figure 19.1), resulting in a larger number of PTAs with comprehensive LPs compared to PTAs with shallow LPs for the first time in 2014.10 Figure 19.1: Cumulative number of PTAs with no LP, shallow LPs, and comprehensive LPs, 1980-2017 Source: Deep Trade Agreements Database. 8 Our sample includes only WTO-notified PTAs with the year of the entry into force of the agreement as the year of the agreement in the various calculations. The underlying data include 271 PTAs that entered into force during the period 1960-2017.The sample includes two interim PTAs (EU-Cameroon Interim EPA, 2009; and EU-Eastern and Southern Africa States Interim EPA, 2012), the coding of which would need to be verified once the final PTAs were adopted. The sample does not include sui generis PTAs that are a mixture of trade and political in their nature with LPs included not in direct relation to trade-related provisions but within the context of the broader political entity established by the given PTA. Such agreements are: Andean Community (CAN), 1988; East African Community (EAC), 2000; East African Community (EAC) – Accession of Burundi and Rwanda, 2007; EC Treaty, 1958; EC (9) Enlargement, 1973; EC (10) Enlargement, 1981; EC Enlargement (12), 1986; EC Enlargement (15), 1995; EC Enlargement (25), 2004; EC Enlargement (27), 2007; and EU (28) Enlargement, 2013. 9 Only one PTA included comprehensive LPs before 1980: the EU-Overseas Countries and Territories (OCT) agreement of 1971. While this PTA is part of the sample, it is included in the analysis below only in the figures that illustrate trends over the period 1960-2017. 10 In our sample, the only PTA that entered into force in 2017 is the Transpacific Partnership (TPP). 596 Labor Market Regulations In terms of the share of PTAs with comprehensive LPs in the total number of PTAs signed in a given year, the data indicate that compared to an average of 9 percent in the 1990s, the share of such PTAs rose to 24 percent during the first decade of the 2000s and reached 52 percent during the period of 2010-2017. It peaked at 100 percent in 2016-2017 (Figure 19.2). Figure 19.2: Share of PTAs with comprehensive LPs in total PTAs per year, 1980-2017 Source: Deep Trade Agreements Database. Looking at the breakdown of PTAs by the level of development of its members11 and by broad types of LPs (i.e., no LPs, shallow LPs, and comprehensive LPs), the data show that while agreements signed between countries from the North and the South (North-South PTAs) have the highest share of PTAs with any kind of LPs, PTAs signed between countries from the North (North-North PTAs) exhibit - although only by a fraction - the highest share of PTAs with comprehensive LPs (panel a of Figure 19.3). However, when looking at PTAs with comprehensive LPs only, most of those agreements are signed between North and South countries (panel b of Figure 19.3). While this result highlights potential protectionist motivations by Northern countries, introducing LPs into their PTAs also indicates a general concern about worker protection in the South by Northern countries. Interestingly, PTAs signed between North and South countries also have the largest share of PTAs with shallow LPs, followed by South-South and North-North PTAs (panel b of Figure 19.3).The majority of PTAs without any LPs are South-South PTAs, which also have the smallest share of agreements with comprehensive LPs.  11 The classification of countries is based on the World Bank’s 2017 Country and Lending Groups classification. High-income countries are classified as North, and middle- and low-income countries are classified as the Global South. Countries in the high-income group but not OECD member countries are Andorra; Antigua and Barbuda; the Bahamas; Bahrain; Barbados; Brunei Darussalam; Cyprus; Faroe Islands; Greenland; Hong Kong SAR, China; Kuwait; Liechtenstein; Lithuania; Macao SAR. China; Malta; Monaco; Oman; Palau; Qatar; San Marino; Saudi Arabia; Seychelles; Singapore; St. Kitts and Nevis;Taiwan, China;Trinidad and Tobago; United Arab Emirates; and Uruguay. 597 Handbook of Deep Trade Agreements Figure 19.3: Distribution of PTAs by level of development of its members and broad types of LPs, 1960-2017 a: By level of development b: By broad types of LPs Source: Deep Trade Agreements Database. Delving into country-level data over the 1960-2017 period, the countries with the highest number of PTAs with LPs are EU and EFTA (European Free Trade Association) members, followed predominantly by Latin American countries (such as Chile, Costa Rica, Peru, Panama, Colombia, Honduras, El Salvador, Guatemala, Mexico, and Nicaragua, in that order) together with the US, Canada, Korea, and New Zealand. Of the BRICS countries (Brazil, Russia, India, China, and South Africa), China stands out, with seven PTAs with LPs (four of which are with comprehensive LPs), compared to Brazil with no PTAs with LPs. India has only one PTA with shallow LPs, while Russia has one PTA with shallow and one with comprehensive LPs, and South Africa has two PTAs with comprehensive LPs and one with shallow LPs (Figure 19.4). Figure 19.4: Number of PTAs with LPs by country, 1960-2017 Source: Deep Trade Agreements Database. Note: Gray areas denote countries with no PTAs with LPs.The color scale is applied to countries that have at least one PTA with either shallow or comprehensive LPs and hence uses 1 as the lowest score on the scale. 598 Labor Market Regulations The picture is somewhat different when looking at the countries with the highest number of PTAs with shallow or comprehensive LPs (figures not shown). For shallow LPs, the leading countries are the EFTA member states, followed by Chile and the EU member states, and then other Latin American countries such as Costa Rica, Honduras, El Salvador, Guatemala, Nicaragua, Mexico, and Peru, with the US and New Zealand having no such PTAs, and Canada having only one agreement with shallow LPs (Canada-EFTA 2009). By contrast, the EU member states have the highest number of PTAs with comprehensive LPs, followed by the US, Chile, Canada, and New Zealand, and only after that Peru, Korea, and the EFTA member countries with other Latin American countries (e.g., Colombia, Costa Rica). 19.3.2 Types and enforceability of LPs in PTAs Most of the labor provisions included in comprehensive PTAs over the period of 1990-2017 are, in equal share, substance related, cooperation related, and provisions establishing the institutional framework for the monitoring and implementation of LPs. Investment-related provisions represent the fewest number of commitments (panel a of Figure 19.5). Most of the provisions are at the non-binding or best-effort level of enforceability, or binding but without the possibility of state-to-state or private-state DS. Only a small portion are subject to state-to-state DS, while only the PTA signed between Japan and Mongolia in 2016 allows for private-state DS regarding a labor commitment included in its investment chapter (panel b of Figure 19.5). The largest change in the shares of LPs of a given type took place in relation to the adoption of investment-related LPs (panel c of Figure 19.6), which increased from 0 percent during the first part of the period (1990-1995) to 47 percent during the last part (2011-2017).The shares of PTAs with institution- and substance-related provisions - despite the latter being included less often - increased over the same period by 31 and 39 percentage points, respectively (panels b and e of Figure 19.6). Figure 19.5: Average number of LPs per main category and enforceability in PTAs with comprehensive LPs, 1990-2017 a: Per main categories b: Per enforceability Source: Deep Trade Agreements Database. Note: average number of provisions per main category was calculated by generating dummy variables for each main category of LPs per PTA (where the category gets 1 if any of the items listed under the category is coded and 0 if nothing is coded under the main category) and then averaging those across all PTAs. 599 Handbook of Deep Trade Agreements By contrast, the share of cooperation-related LPs, some of the most frequently included, rose only by 24 percentage points, from 13 to 37 percent, between 1990-95 and 2011-2017 (panel d of Figure 19.6). Among the PTAs with comprehensive LPs, the share of PTAs with aspirational provisions rose by 34 points, from 8 to 42 percent (panel a of Figure 19.6). For PTAs with both shallow and comprehensive LPs, the share of PTAs with aspirational provisions increased by 46 points, from 21 to 67 percent (figures not shown). Figure 19.6: Share of LP types in total PTAs per five-year windows, 1990-2017 a: Share of PTAs with aspirational LPs b: Share of PTAs with substance-related LPs c: . Share of PTAs with investment-related LPs d: Share of PTAs with cooperation-related LPs e: Share of PTAs with institution-related LPs Source: Deep Trade Agreements Database. In terms of enforceability, while LPs have become more binding over time, exhibiting a 30 percentage point increase (from 17 percent to 47 percent) between 1990-1995 and 2011-2016, LPs that are also subject to state-to-state DS or private DS rose at a more modest rate, from 13 percent to 24 percent and from 0 percent to 1 percent, respectively. By contrast, the share of non- binding/best-effort provisions rose by 41 percentage points over the same period of time (from 8 percent to 49 percent), indicating that such provisions continue to remain the prevailing form of commitments (figures not shown). 600 Labor Market Regulations Among PTAs with comprehensive LPs, the provisions most frequently included are those related to cooperation, closely followed by those related to the institutional framework together with the commitment to effectively enforce domestic labor laws. Least common are provisions related to corporate social responsibility, references to other relevant international instruments and to internationally recognized labor standards. References to the ILO 1998 Declaration on Fundamental Principles and Rights at Work and related labor rights (such as freedom of association and collective bargaining rights, prohibitions against child and forced labor, and the elimination of discrimination in employment and occupation) fall between the most and least frequent provisions, with relatively minor variation between them (panel a of Figure 19.7). Interestingly, the LPs that are most frequently subject to state-to-state DS are by far those concerning effective enforcement of domestic labor laws, followed by references to fundamental labor rights defined under the ILO 1998 Declaration and the commitment of non- derogation from labor laws (panel b of Figure 19.7). By contrast, investment-related LPs are less often, and corporate social responsibility rarely, subject to DS (panel b of Figure 19.7). By their nature, institution-related provisions (impact assessment, third- party inclusion, separate specialized committee/contact points) and aspirational LPs are not subject to state-to-state DS (panel b of Figure 19.7). Figure 19.7: Frequency of types of LPs and their enforceability, 1990-2017 a: Frequency of types of LPs b: Frequency of enforceability (binding LPs with state-to-state DS) of types of LPs Source: Deep Trade Agreements Database. 601 Handbook of Deep Trade Agreements 19.3.3 LPs in PTAs by key players The two actors that have played the most pivotal role in influencing the design of LPs in PTAs are the US and the EU. While the EU has signed more PTAs with LPs than the US since the early 1990s, the US still has the higher average number of LPs included in its PTAs. Although this trend has started to change, particularly with the introduction of the EU’s latest generation of PTAs with the EU-Korea agreement in 2010 (panel a of Figure 19.8), the US continues to have, with the exception of cooperation-related LPs, more provisions under each of the main categories of LPs (panel b of Figure 19.8). Figure 19.8: LPs in EU- and US-signed PTAs, 1990-2017 a: Average number of LPs b: Average number of LPs per main category Source: Deep Trade Agreements Database. According to previous literature, the key difference between the PTAs of the US and the EU is that the US uses a hard-law approach while the EU uses a soft-law approach (Hafner-Burton 2005). While it is true that the US includes overall more enforceable LPs in its PTAs (Figure 19.9), the difference between the two approaches is more subtle than previously indicated. 12 Since 2006, the EU has had more binding provisions without DS than the US, while state-to-state DS provisions are not at all absent from the EU-signed agreements. 13 At the same time, US agreements include about the same number of cooperation-related provisions as those signed by the EU (panel b of Figure 19.8). 12 See, for instance, Hafner-Burton (2005); and Postnikov and Bastiaens (2014). 13 The EU-CARIFORUM Economic Partnership Agreement (2008), for example, allows parties to bring disputes related to LPs under DS procedure and to adopt – with the exception of suspension of trade concessions – appropriate measures in case the party complained against does not comply with the ruling handed down by the arbitration panel (Art. 213 (2)). 602 Labor Market Regulations Figure 19.9: Enforceability of LPs in EU- and US-signed PTAs, 1990-2017 a: Average number of binding LPs without DS b: Average number of binding LPs with state-to-state DS Source: Deep Trade Agreements Database. In addition to the US and the EU, Canada and New Zealand (the latter only since the early 2000s) are also key drivers of LPs in PTAs. Although Canada includes more LPs in its PTAs than New Zealand (panel a of Figure 19.10), the main difference concerns the number of substance-related provisions. Regarding aspirational, investment- cooperation-, and institution-related LPs, the designs adopted by the two countries do not show significant differences, except that New Zealand has slightly more investment- and institution-related LPs (panel b of Figure 19.10). Figure 19.10: LPs in Canada- and New Zealand–signed PTAs, 1990-2017 a: Average number of LPs b: Average number of LPs per main categories Source: Deep Trade Agreements Database. Looking at the enforceability of provisions, a more pronounced difference can be identified: while New Zealand, like Canada, uses binding language in its commitments to labor protection and promotion (panel a of Figure 19.11), it was only with the signing of the TPP that New Zealand included the possibility of subjecting LPs to state-to-state DS (panel b of Figure 19.11). 603 Handbook of Deep Trade Agreements Interestingly, comparing Canada with the EU and the US, the data indicate that Canada has, on average, a higher number of LPs than the other two key trade players, especially with respect to substance-related provisions and enforceability (Figures 19.8-19.11). Canada’s progressive approach was also outlined by its proposal for a more ambitious approach to advancing labor rights in the renegotiation of NAFTA in 2018.14 Figure 19.11: Enforceability of LPs in EU- and US-signed PTAs, 1990-2017 a: Average number of binding LPs without DS b: Average number of binding LPs with state-to-state DS Source: Deep Trade Agreements Database. 19.3.4 LPs in South-South PTAs Throughout the past decades, and particularly since the early 2000s, countries from the South have also begun to include LPs in the PTAs signed among themselves, 15 although these commitments remain rather limited both in terms of scope and stringency. The most noticeable increase in the average number of LPs among newly signed PTAs was during 2006-2010 (from 2 LPs in the 1990-1995 period to 5 LPs during 2006-2010), with the majority of those relating to cooperation or the institutional framework. Regarding enforceability, the South’s commitments lag far behind those made by the key players from the North. Only the Economic Community of West African States (ECOWAS) agreement allows for state-to-state DS for labor provisions, particularly in relation to cooperation regarding the harmonization of labor laws. 14 https://www.cigionline.org/articles/what-nafta-negotiators-can-do-about-worker-anxiety. 15 South-South PTAs with comprehensive LPs signed between 1990 and 2016 are: Economic Community of West African States (ECOWAS),1993; Common Market for Eastern and Southern Africa (COMESA), 1994; Eurasian Economic Community (EAEC), 1997; Southern African Development Community (SADC), 2000; Peru-China, 2010; Costa Rica-Colombia; 2016; and Pacific Alliance, 2016. 604 Labor Market Regulations 19.4. CONCLUSIONS This first comprehensive mapping of the content of LPs in PTAs reveals a twofold trend.16 On the one hand, LPs are in general becoming more binding and enforceable, mainly as a result of the design adopted by key trade players such as the US, Canada, and the EU. On the other hand, agreeing less stringent provisions (with few binding and even fewer enforceable provisions) continues, particularly by actors that have more recently started to incorporate LPs into their trade agreements (New Zealand, EFTA, and South-South trading partners). However, one key commonality between both groups, aside from the general increase in substance-related provisions, is the increased emphasis on the institutional framework responsible for the monitoring and implementation of LPs, resulting in more inclusive and specialized institutions. With the recent NAFTA re-negotiations and the EU’s renewed debate on the effective implementation and enforcement of Trade and Sustainable Development chapters,17 the possible design of LPs in PTAs has yet again come to the center of attention. In both cases, labor unions were pushing for the broadening of the scope of issues incorporated in PTAs and for more efficient dispute settlement mechanisms.18 Such demands have gained momentum in light of the recent ruling in the US-Guatemala case finding that no violations of the PTA could be determined, in part because of the lack of evidence for the non-compliance with LPs in a manner affecting trade. In both cases, while acknowledging some of the new provisions as improvements compared to earlier texts, trade unions expressed dissatisfaction with the overall result, particularly given the lack of convincing/sufficient progress regarding the enforcement of labor provisions.19 Given that labor provisions in trade agreements continue to evolve over time, it is likely that additional items to address new developments will be added to the coding template in the near future. ACKNOWLEDGMENTS An earlier version of this chapter was presented at “The Evolution of Deep Trade Agreements” workshop held at the World Bank, Washington, DC, on 1-2 February 2018. For helpful comments, we are grateful to Rohini Acharya, Aaditya Mattoo, Nadia Rocha, Michele Ruta, Liz Schéré, Daria Taglioni, Joel Trachtman, and Mark Wu. 16 See also Raess and Sari 2018. 17 http://trade.ec.europa.eu/doclib/press/index.cfm?id=1689. 18 https://www.etuc.org/en/document/etuc-resolution-eu-progressive-trade-and-investment-policy; https:// aflcio.org/statements/written-comments-how-make-nafta-work-working-people. 19 https://aflcio.org/pressreleases/new-nafta-deal-far-finished; https://www.etuc.org/en/document/etuc-assessment- commissions-non-paper-trade-and-sustainable-development-tsd-chapters-eu. 605 Handbook of Deep Trade Agreements REFERENCES Acharya, R. (ed.). 2016. Regional trade agreements and the multilateral Kamata, I. 2016. Labor clauses in regional trade agreements trading system. Cambridge: Cambridge University Press. and effects on labor conditions: An empirical analysis. IDE Discussion Paper 609, Institute of Developing Economies, Alben, E. 2001. GATT and the fair wage: A historical Tokyo. perspective on the labor-trade Link. Columbia Law Review 101(6):1410-47. Kucera, D., and D. Sari. 2019. New labour rights indicators: Method and trends for 2000–15. International Labour Review Abbott, K.W., R. O. Keohane, A. Moravcsik, A-M. Slaughter, 158(3):419-46. and D. Snidal. 2000. The concept of legalization. International Organization 54(3):401-19. Lechner, L. 2016. The domestic battle over the design of non-trade issues in preferential trade agreements. Review of Bourgeois, J., K. Dawar, and S. J. Evenett. 2007. A comparative International Political Economy 23(5):840-71. analysis of selected provisions in free trade agreements. Commissioned by DG Trade. https://pdfs.semanticscholar. Milewicz, K., J. Hollway, C. Peacock, and D. Snidal. 2018. org/cb3d/3b6bb04174906b95ad1dc9604b60df6c791d. Beyond trade:The expanding scope of the nontrade agenda in pdf?_ga=2.67848466.2007458718.1572435771- trade agreements. Journal of Conflict Resolution 62(4): 743–73. 1270183482.1572435771. Milewicz, K. 2016. Non-trade issues codebook. https:// Chase, C., A. Yanovich, J.-A. Crawford, and P. Ugaz. 2016. www.designoftradeagreements.org/media/filer_public/ Mapping of disputes settlement mechanisms in regional ef/35/ef351ccf-b698-4e6e-b8e7-686a1becaf7d/nti_ trade agreements – innovative or variations on a theme? In codebook.pdf. R. Acharya (ed.), Regional trade agreements and the multilateral trading system. Cambridge: Cambridge University Press. Postnikov, E., and I. Bastiaens 2014. Does dialogue work? The effectiveness of labor standards in EU preferential trade Cingranelli, D., D. L. Richards, and K. C. Clay. 2014. The agreements. Journal of European Public Policy, 21(6):923-40. CIRI human rights dataset. http://www.humanrightsdata. com.Version 2014.04.14. Raess, D., and D. Sari. 2018. Labor provisions in trade agreements (LABPTA): Introducing a new dataset. Global Dür, A., and M. Elsig (eds.). 2015. Trade cooperation: The Policy 9(4):451-66. purpose, design and effects of preferential trade agreements. Cambridge: Cambridge University Press. Raess, D., A. Dür, and D. Sari. 2018. Protecting labor rights in preferential trade agreements: The role of trade unions, Hafner-Burton, E. 2005. Trading human rights: How left governments, and skilled labor. Review of International preferential trade agreements influence government Organizations 13(2):143-62. repression. International Organization 59 (3):593-629. Hofmann, C., A. Osnago, and M. Ruta. 2017. Horizontal depth: A new database on the content of preferential trade agreements. Policy Research Working Paper 7981,Trade and Competitiveness Global Practice Group,World Bank,Washington, DC. Hulme, M. H. 2016. Preambles in treaty interpretation. University of Pennsylvania Law Review 164(5):1281-1343. ILO (International Labour Organization). 2009.World of work report 2009:The global jobs crisis and beyond. Geneva: ILO. ———. 2013. Social dimensions of free trade agreements. Geneva: ILO. ———. 2016. Assessment of labour provisions in trade and investment arrangements. Geneva: ILO. 606 Labor Market Regulations APPENDIX. LIST OF FIGURES, TABLES, AND BOXES FIGURES Figure O.1: Number of policy areas covered in PTAs, 1970-2017 6 Figure O.2: Number of policy areas covered in PTAs, by policy 8 Figure O.3: A classification of policy areas in DTAs 10 Figure O.4: Tariffs in PTAs and MFN tariffs 12 Figure O.5: Number of agreements over time vs. average coverage ratio 13 Figure O.6: Coverage ratios by policy area, over time 14 Figure O.7: Substantive provisions and a breakdown of non-substantive provisions in PTAs, over time 15 Figure O.8: Inclusion of substantive commitments in PTAs, by level of development 16 Figure O.9: Coverage ratio by type of policy and level of development 16 Figure O.10: Substantive provisions in PTAs by country 17 Figure O.11: Similarity of agreements 19 Figure O.12: Boxplot of scores calculated by the RF of variables/provisions in PTAs 22 Figure 1.1: Trade agreements have proliferated over time 47 Figure 1.2: Most recent tariff information 51 Figure 1.3: Applied MFN rates have steadily declined over time 52 Figure 1.4: Almost two-thirds of imports by value are subject to MFN rates of less than 5 percent 52 Figure 1.5: MFN rates vary significantly across the three economic sectors (value of imports) 53 Figure 1.6: MFN rates vary significantly across the three economic sectors 53 Figure 1.7: More than half of the value of global trade took place under an agreement in 2015 54 Figure 1.8: Preferential liberalization has reduced trade-weighted average tariff rates to less than 5 percent for more than two-thirds of countries 54 Figure 1.9: Preferential agreements have reduced protection across the board but less so where tariffs are high 55 Figure 1.10: Tariffs have been reduced across sectors but are still high for agricultural products 55 Figure 1.11: There is room for further liberalization 56 Figure 1.12: Although preferential liberalization has targeted highly protected sectors (MFN tariffs greater than 15 percent), agricultural products, textiles, and footwear remain pockets of protection 57 Figure 1.13: Distribution of preference margins 58 Figure 1.14: Most countries benefited from a competition-adjusted margin between -2 and 2 percent 59 Figure 1.15: Lower-income countries tend to benefit the most in terms of competitive-adjusted margin 60 Figure 1.16: EU imports by tariff regime and country group (US$ million) 61 Figure 1.17: Utilization of EU preferences by beneficiary countries 62 Figure 1.18: Utilization rates vs. preferential margin 62 Figure 1.A.1: Reduction in trade-weighted tariffs is uniform across sectors 66 Figure 1.A.2: On average countries had reduced tariffs by half 67 Figure 1.A.3: Share of sector tariff lines weighted by partner’s share of global trade 67 Figure 1.A.4: Competition-adjusted preference margin using import demand elasticities 68 Figure 2.1: Categorization of preferential imports by duty schemes 76 Figure 2.2: Total imports by duty categories used, 2016 79 Figure 2.3: Imports from least-developed countries by duty categories, 2016 (or latest available year) 80 607 LIST APPENDIX.of Handbook FIGURES, OF Trade Deep TABLES, AND BOXES Agreements Figure 2.4: Utilization of the LDC duty schemes in developed markets by product and duty categories, 2016 81 Figure 2.5: MFN duties paid on GSP-LDC preference-eligible imports by product category, 2016 82 Figure 3.1: Number of PTAs addressing export taxes and/or other export-related provisions 98 Figure 3.2: Percentage of PTAs addressing export tax and/or other export-related restrictions, 1991-2015 98 Figure 3.3: Percentage of PTAs concluded in developing countries addressing export taxes and/or other export restrictions, 1991-2015 99 Figure 4.1: Proportion of PTAs with services over time 122 Figure 4.2: PTAs with services by level of development 122 Figure 4.3: Number of PTAs with service, intra-, and extraregional groupings 123 Figure 4.4: Liberalization approaches 124 Figure 5.1: Number of PTAs that include investment provisions, 1958-2018 146 Figure 5.2: Definitions and scope provisions of the investment chapter, share of PTAs 156 Figure 5.3: Investment liberalization provisions, share of PTAs 159 Figure 5.4: Investment protection provisions, share of PTAs 161 Figure 5.5: Fair and equitable treatment provisions, share of PTAs 161 Figure 5.6: Provisions regarding protection of assets, share of PTAs 162 Figure 5.7: Provisions regarding protection of assets, share of PTAs 162 Figure 5.8: Selected social and regulatory goals in the investment chapter, share of PTAs (%) 165 Figure 5.9: Selected transparency provisions, share of PTAs 165 Figure 5.10: Dispute settlement mechanisms, share of PTAs 166 Figure 5.11: Boxplot of average number of Investment provisions in new PTAs over time 167 Figure 5.12: Average number of Investment provisions by country 167 Figure 5.13: Incidence of core investment provisions 168 Figure 5.14: Rising incidence of investment liberalization provisions over time, by level of development 169 Figure 5.15: Evolution of share of Investment protection provisions over time, by level of development 169 Figure 5.16: Evolution of share of ISDS provisions over time, by level of development 170 Figure 6.1: Free transfers commitments over time 195 Figure 6.2: Treaty commitment by income level 195 Figure 7.1: Percentage of PTAs with IPR-related provisions, 1991-2005 211 Figure 7.2: Percentage of PTAs concluded among developing countries with IPR-related provisions, 1991-2005 212 Figure 8.1: Number of PTAs that include migration provisions over time 233 Figure 8.2: Economies with “visa and asylum” PTAs (World Bank classification) 234 Figure 8.3: Economies with “visa and asylum” PTAs (WTO classification) 235 Figure 8.4: Types of movement covered by type of PTA 237 Figure 8.5: Methods of facilitation covered by type of PTA 239 Figure 8.6: Enforceability of “visa and migration” provisions 242 Figure 8.7: Do “visa and asylum” provisions benefit non-parties? 244 Figure 9.1: Distribution of product-specific rules of origin across regional PTAs 275 Figure 9.2: Heterogeneity of product-specific rules of origin over time, average across all six-digit products 279 Figure 9.3: Index of restrictiveness for regime wide rules of origin over time 288 Figure 10.1: Average coverage ratio of TF provisions in new PTAs over time 300 Figure 10.2: Trade facilitation provisions with highest frequency 301 608 LaborTABLES, APPENDIX. LIST OF FIGURES, AND BOXES Market Regulations Figure 10.3: Trade facilitation provisions with lowest frequency 302 Figure 10.4: Share of PTAs with more “complex” provisions 302 Figure 10.5: Average number of trade facilitation provisions in PTAs, by level of development 303 Figure 10.6: Key provisions other than customs, over time 304 Figure 10.7: Share of agreements including provisions on internet publication by level of development 307 Figure 12.1: TBT-related specific trade concerns submitted to WTO, 1995-2017 345 Figure 12.2: Number of PTAs that include TBT provisions, 1958-2017 348 Figure 12.3: Enforceability of TBT provisions over time 354 Figure 12.4: Exclusion of specific provisions in TBT agreements over time 355 Figure 12.5: Sector-specific coverage of TBT provisions 356 Figure 12.6: Average coverage ratio of TBT provisions in new PTAs over time 356 Figure 12.7: Average depth of TBT provisions in new PTAs over time 357 Figure 12.8: TBT depth across countries, 2017 358 Figure 12.9: Percentage of PTAs by TBT provision, 2017 358 Figure 12.10: Percentage of PTAs by preferred integration approach, 2017 359 Figure 12.11: Correlations of similarity indexes with average number of provisions, 2017 362 Figure 12.12: Number of TBT provisions, average over agreements in force during 2015 by level of development 362 Figure 12.13: Percentage of PTAs by TBT provision and development level, 2017 363 Figure 13.1: Number of agreements with SPS provisions over time 383 Figure 13.2: Number of agreements with SPS provisions, by income 383 Figure 13.3: Standards cited by method and level of enforcement 384 Figure 13.4: Standards cited by method and level of development 385 Figure 13.5: Distribution of level of enforcement across transparency provisions 386 Figure 13.6: Distribution of transparency provisions by income level 387 Figure 13.7: Provisions for technical cooperation/consultation by level of enforcement 388 Figure 14.1: Evolution of PTA groups by government procurement coverage over time 402 Figure 14.2: Breakdown of PTA membership by procurement coverage and income 402 Figure 14.3: Frequency distribution of leading provisions in DPAs by level of development 410 Figure 14.4: Distribution of non-discrimination procurement provisions in DPAs 411 Figure 14.5: Frequency distribution of non-discrimination provisions in DPAs 412 Figure 14.6: Distribution of procedural disciplines in DPAs 415 Figure 14.7: Frequency distribution of procedural disciplines in DPAs 415 Figure 14.8: Distribution of transparency-related provisions in DPAs 417 Figure 14.A.1: Distribution of provisions in PPAs under six different themes - non-discrimination, procedural issues, transparency, new issues, dispute settlement, and coverage 425 Figure 15.1: Evolution of the number of PTAs with subsidy provisions 440 Figure 15.2: Evolution of the number of PTAs with subsidy provisions 441 Figure 15.3: Evolution of the share of PTAs with different types of subsidy provisions 441 Figure 15.4: Share of PTAs with subsidy provisions by level of development 443 Figure 15.5: Evolution of the share of PTAs with subsidy provisions by level of development, over time 443 Figure 15.6: Share of PTAs with subsidy provisions by geographic region/group 444 Figure 15.7: Most common types of subsidy provisions in PTAs 446 609 LIST APPENDIX.of Handbook FIGURES, OF Trade Deep TABLES, AND BOXES Agreements Figure 15.8: Most common provisions by level of development 448 Figure 15.9: Most common provisions by geographic group 448 Figure 15.10: Subsidy provisions with dispute settlement by geographic group 449 Figure 15.11: Most common provisions by notable countries 450 Figure 15.12: Coverage (sectors) 451 Figure 15.13: Coverage (sectors) by level of development 451 Figure 15.14: Export subsidies by geographic group 452 Figure 15.15: Domestic subsidies 453 Figure 15.16: Domestic subsidies by level of development 453 Figure 15.17: Transparency provisions 455 Figure 15.18: Transparency provisions by level of development 456 Figure 16.1: Evolution of the number of PTAs with state enterprise provisions 475 Figure 16.2: Number of state enterprise provisions (average) over time 475 Figure 16.3a: Evolution of the number of PTAs with state enterprise provisions 476 Figure 16.3b: Evolution of the share of PTAs with state enterprise provisions 476 Figure 16.4: Share of PTAs with state enterprise provisions by level of development 477 Figure 16.5: Share of PTAs with state enterprise provisions by level of development, by year of signature 478 Figure 16.6: Share of PTAs with state enterprise provisions by geographic group 479 Figure 16.7: Most common provisions 481 Figure 16.8: Most common provisions by level of development 483 Figure 16.9: Most common provisions by level of development 483 Figure 16.10: Most common provisions by geographic group 484 Figure 16.11: Most common provisions by notable countries 485 Figure 16.12: Definitions 487 Figure 16.13: Coverage (entities) – comprehensive coding approach 488 Figure 16.14: Coverage of state entities, limited coding approach 489 Figure 16.15: Coverage of state entities by income group, limited coding approach 490 Figure 16.16: Coverage (sectors) 490 Figure 16.17: Coverage (sectors) by income groups 491 Figure 16.18: Disciplines 491 Figure 16.19: Disciplines by income group 493 Figure 16.20: Exceptions 495 Figure 16.21: Exceptions by income group 495 Figure 16.22: Transparency and corporate governance 497 Figure 16.23: Transparency and corporate governance by income group 498 Figure 16.24: Bodies or committees by income group 500 Figure 16.25: Dispute settlement by income group 500 Figure 16.26: Enforcement by geographic group 501 Figure 17.1: A comprehensive competition policy framework 508 Figure 17.2: PTAs and coverage of anticompetitive behavior 511 Figure 17.3: PTAs and coverage of sectoral competition provisions 512 Figure 17.4: Evolution of in-force PTAs with competition-related provisions, 1958-2016 515 610 LaborTABLES, APPENDIX. LIST OF FIGURES, AND BOXES Market Regulations Figure 17.5: Evolution of in-force PTAs with competition-related provisions, by type of PTA 515 Figure 17.6: Bilateral PTAs with competition-related provisions, by income level 516 Figure 17.7: General objectives and horizontal principles of PTAs, including - or not - competition provisions 519 Figure 17.8: Distribution of PTAs by specific stipulation within each competition national requirement 520 Figure 17.9: Distribution of PTAs with requirement of competition law, by features of the law 521 Figure 17.10: Distribution of PTAs requiring a competition authority, and which features they require 521 Figure 17.11: Distribution of PTAs with requirement for competition policy/principles, by features of the policy 522 Figure 17.12: References to other international agreements in PTAs, by type of PTA 523 Figure 17.13: Regulated anticompetitive behavior in PTAs, by type of PTA 525 Figure 17.14: Procedural fairness in PTAs, by type of PTA 527 Figure 17.15: Distribution of PTAs with provisions that promote the right of defense, by the scope of such right 527 Figure 17.16: Distribution of PTAs with provisions that promote transparency, by specific obligations 528 Figure 17.17: Distribution of PTAs with provisions that promote predictability, by specific requirements 528 Figure 17.18: Institutional arrangement in PTAs, by type of PTA 530 Figure 17.19: Forms of cooperation in PTAs, by type of PTA 531 Figure 17.20: Distribution of PTAs with notification requirements, by scope of such notifications 532 Figure 17.21: Competitive neutrality provisions in PTAs, by type of PTA 534 Figure 17.22: PTAs and coverage of sector-specific competition provisions by type of PTA 535 Figure 17.23: Evolution of PTAs in force with sector-specific provisions against rules that reinforce dominance or limit entry, by sector 537 Figure 17.24: Sector-specific provisions against rules that reinforce dominance or limit entry 537 Figure 17.25: Evolution of PTAs with sector-specific provisions against rules that are conducive to collusion or increase the cost of competing in the market, by sector 540 Figure 17.26: Sector-specific provisions against rules that are conducive to collusion or increase the cost of competing in the market 540 Figure 17.27: Evolution of PTAs with sector-specific provisions against rules that discriminate or protect vested interests, by sector 542 Figure 17.28: Sector-specific provisions against rules that discriminate or protect vested interests 542 Figure 17.29: Enforceability in PTAs with sectoral commitments 544 Figure 18.1: Evolution of the number of PTAs with environment-related provisions 565 Figure 18.2: Evolution of the number of types of environment-related provisions in PTAs 565 Figure 18.3: Average number of types of environment-related provisions in PTAs by country 567 Figure 18.4: Relationship between the number of types of environment-related provisions in PTAs and level of economic development and environmental quality 568 Figure 18.5: Scope of types of environment-related provisions 569 Figure 18.6: Provisions on environmental goals or objectives 570 Figure 18.7: Provisions on environmental goals or objectives 571 Figure 18.8: Provisions on participation in promoting environmental objectives 573 Figure 18.9: Provisions on enforcement mechanism 574 Figure 18.10: Similarity of environment-related provisions in PTAs 576 Figure 18.A.1: Dynamic relationship between the number of PTAs and average number of environment-related provisions 581 Figure 18.A.2: Histogram of the Jaccard Similarity Index of environment-related provisions 581 611 APPENDIX.of Handbook Deep LIST FIGURES, OF Trade TABLES, AND BOXES Agreements Figure 19.1: Cumulative number of PTAs with no LP, shallow LPs, and comprehensive LPs, 1980-2017 596 Figure 19.2: Share of PTAs with comprehensive LPs in total PTAs per year, 1980-2017 597 Figure 19.3: Distribution of PTAs by level of development of its members and broad types of LPs, 1960-2017 598 Figure 19.4: Number of PTAs with LPs by country, 1960-2017 598 Figure 19.5: Average number of LPs per main category and enforceability in PTAs with comprehensive LPs, 1990-2017 599 Figure 19.6: Share of LP types in total PTAs per five-year windows, 1990-2017 600 Figure 19.7: Frequency of types of LPs and their enforceability, 1990-2017 601 Figure 19.8: LPs in EU- and US-signed PTAs, 1990-2017 602 Figure 19.9: Enforceability of LPs in EU- and US-signed PTAs, 1990-2017 603 Figure 19.10: LPs in Canada- and New Zealand–signed PTAs, 1990-2017 603 Figure 19.11: Enforceability of LPs in EU- and US-signed PTAs, 1990-2017 604 TABLES Table O.1: Share of policy areas for different PTAs 7 Table O.2: Number of substantive and other provisions per policy area in all PTAs notified to the WTO 11 Table O.A.1: Number of provisions included and coverage ratio – by agreement 26 Table O.A.2: Average number of provisions included and coverage ratio – by country 33 Table O.A.3: Maximum number of provisions included and coverage ratio – by country 38 Table 1.1: Non–ad valorem tariffs and ad valorem equivalent (AVE) composition 48 Table 1.2: Examples of technical duties 50 Table 1.3: More than a quarter of world trade is subject to an average preference margin of 7.4 percent 58 Table 1.A.1: Agreements with partial information 68 Table 1.A.2: Comparison of available data and entry into force of last agreement 69 Table 1.A.3: Countries with MFN information in at least 15 years between 1995 and 2015 69 Table 1.A.4: Share of sectoral and development-level tariff lines (multilaterally sensitive preferentially free) 70 Table 1.A.5: Share of sectoral and development-level tariff lines (excluded) 70 Table 1.A.6: Share of preferential trade by preferential margin and MFN range 71 Table 1.A.7: Share of preferential trade by preferential margin and MFN range, by importing country group 71 Table 2.1: List of PTAs included in the analysis 78 Table 3.1: Most common provisions found in PTAs concerning export taxes and other export-related restrictions 100 Table 4.1: Previous datasets on services PTAs 117 Table 4.2: Design of PTAs: Coverage of issues in previous datasets / exercises 118 Table 4.3: Structure of services trade, number of PTAs 125 Table 4.4: Inclusion of provisions clarifying the hierarchy/relationship between the investment chapter/protocol and trade in services chapter 126 Table 4.5: Inclusion of separate sector-specific chapter 126 Table 4.6: Inclusion of services sector-specific annexes to a trade in services or invesment chapter 126 Table 4.7: Sectoral exclusions from services and/or investments sections 127 Table 4.8: Policy areas excluded from PTAs 128 Table 4.9: Obligation on market access 129 Table 4.10: Standstill and ratchet provisions 130 612 LaborTABLES, APPENDIX. LIST OF FIGURES, AND BOXES Market Regulations Table 4.11: Obligation not to have certain requirements concerning trade in services 131 Table 4.12: Number of PTAs with provisions on monopolies, additional commitments, and liberalization 131 Table 4.13: Number of PTAs containing provisions on qualification, licensing, technical standards, and a necessity test 132 Table 4.14: Number of PTAs with obligation to inform on status of application 132 Table 4.15: Number of PTAs with obligation to make decisions within a certain period of time 132 Table 4.16: Obligation to administer the measures / laws / regulations in a reasonable, objective, and impartial manner 132 Table 4.17: PTAs with mutual recognition provisions 133 Table 4.18: Obligation to publish 134 Table 4.19: Obligation to provide for prior comment on proposed regulation 134 Table 4.20: Obligation to set up an independent authority for appeal procedures 135 Table 4.21: Treatment of the movement of natural persons 135 Table 4.22: Rules of origin for juridical persons 136 Table 4.23: Rules of origin for natural persons 136 Table 4.24: Exceptions 136 Table 4.25: Safeguard Mechanisms 137 Table 4.26: Dispute settlement 137 Table 4.A.1: Template on architecture and design of services PTAs 140 Table 5.1: Patterns of investment provisions across regions, scope and definitions 171 Table 5.2: Patterns of investment liberalization 172 Table 5.3: Patterns of investment protection 172 Table 5.4: Patterns of investment provisions across regions – social and regulatory goals 173 Table 5.5: Patterns of institutional framework 174 Table 5.6: Dispute settlement provisions 174 Table 6.1: Free transfer commitments 181 Table 6.2: Free transfer provisions: examples 182 Table 6.3: Safeguards: sample triggers and disciplines 186 Table 6.4: Carve-outs and general exceptions 191 Table 6.5: Enforcement 192 Table 8.1: Migration questionnaire template 232 Table 8.2: Top 10 economies with visa and asylum PTAs 234 Table 8.3: Types of movement of persons covered in PTAs over time 236 Table 8.4: Methods of facilitating movement of persons in PTAs over time 238 Table 8.5: Exceptions or limitations regarding migration in PTAs over time 240 Table 8.6: Reference to international instruments on migration in PTAs over time 241 Table 9.1: Heterogeneity scores for some EAP countries 278 Table 9.2: Distribution of product-specific rules of origin across HS sections Share of HS 6 digit products covered by different types of PSR (in percentage) 281 Table 9.3: Provisions on certification aspects 282 Table 9.4: Provisions on verification process 283 Table 9.5: Provisions on cumulation 284 613 LIST APPENDIX.of Handbook FIGURES, OF Trade Deep TABLES, AND BOXES Agreements Table 9.6: Provisions on calculation of value content 286 Table 9.7: Provisions on other aspects 287 Table 9.A.1: List of codes and definitions 291 Table 9.A.2: Index for restrictiveness in regime-wide provisions 292 Table 10.1: Origin administration systems in free trade agreements to which Australia is a party 314 Table 11.1: Summary of AD/CVD rules in PTAs 326 Table 11.2: AD/CVD rules in PTAs 333 Table 11.3: AD/CVD rules over time 333 Table 11.4: AD/CVD rules by PTA type 334 Table 11.A.1: List of PTAs and AD/CVD rules 337 Table 12.1: Structure of the template 351 Table 12.2: Patterns of TBT integration across regions – only deep provisions, 2017 360 Table 12.3: Patterns of TBT integration in intra- and extra-regional agreements, 2017 361 Table 14.1: Summary of stylized facts on government procurement provisions in PTAs 403 Table 14.2: US-Bahrain agreements: Threshold values higher than GPA levels 413 Table 14.A.1: List of PTAs with no, shallow, and deep provisions on government procurement 423 Table 15.1: Share of new PTAs with subsidy provisions by geographic group, over time (%) 445 Table 15.2: Coverage (sectors) by geographic groups 451 Table 15.3: Domestic subsidies by geographic groups 454 Table 15.4: Transparency provisions by geographic groups 456 Table 15.5: Enforcement provisions by geographic groups 457 Table 16.1: Share of PTAs with state enterprise provisions by geographic group, over time 480 Table 16.2: Most common provisions by geographic group 484 Table 16.3: Most common provisions by notable countries 486 Table 16.4: Disciplines by geographic region 494 Table 16.5: Exceptions by geographic region/group 496 Table 16.6: Transparency and corporate governance by geographic group 499 Table 17.1: General competition provisions embedded in PTAs signed by 2009 511 Table 17.2: Bilateral PTAs with competition-related provisions, by geographical region 517 Table 17.3: PTA distribution according to provisions against anticompetitive behavior identified by MCPAT 536 Table 17.4: Enforceability of sector-specific provisions 545 Table 17.A.1: Economy-wide competition-related commitments in PTAs 549 Table 17.A.2: Sector-specific competition-related commitments in PTAs 550 Table 17.A.3: Enforceability of economy-wide provisions in PTAs, with competition provisions 551 Table 17.A.4: Enforceability of economy-wide provisions in PTAs, with specific exceptions 552 BOXES Box 4.1: Market access limitations in the GATS 129 Box 9.1: CTC type rules versus RVC type rules 277 Box 9.2: Case study 1 - CTC rules in textiles 280 Box 9.3: Case Study 2 - The facilitating role of cumulation schemes in Mexico and Central America 284 Box 9.4: Case study 3 - Net cost method and traced material list in automotive goods 286 614 LaborTABLES, APPENDIX. LIST OF FIGURES, AND BOXES Market Regulations Box 14.1: Turkey-Israel FTA: An example of a shallow procurement agreement 404 Box 14.2: Incorporation of GPA provisions in the Canada-Republic of Korea FTA 405 Box 14.3: Example of a procurement provision that goes beyond the WTO: US-Chile FTA 407 Box 14.4: Non-applicability of DS chapter: Japan-Switzerland agreement 408 Box 14.5: Example of high levels of enforceability and DS provisions: Japan-Chile FTA 409 Box 14.6: Example of a fully GPA-consistent provision on technical specification: NAFTA 416 Box 14.7: Free trade agreement between Korea and Canada 418 Box 14.8: Example of provision on facilitation of safety standards: US-Korea agreement 419 Box 17.1: Key competition-related provisions of PTAs 510 Box 17.2: The Trans-Pacific Partnership: Procedural fairness obligations under the competition chapter 529 615 Handbook of Deep Trade Agreements ECO-AUDIT Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, we leverage electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. We follow the recommended standards for paper use set by the Green Press Initiative. The majority of our books are printed on Forest Stewardship Council (FSC)–certified paper, with nearly all containing 50–100 percent recycled content. The recycled fiber in our book paper is either unbleached or bleached using totally chlorine-free (TCF), processed chlorine–free (PCF), or enhanced elemental chlorine–free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://www.worldbank.org/corporateresponsibility. 616 583