Global Investment Competitiveness Report 2017/2018 Global Investment Competitiveness Report 2017/2018 Foreign Investor Perspectives and Policy Implications © 2018 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 20 19 18 17 This work is a product of the staff of The World Bank with external contributions. The fi ndings, inter- pretations, 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: World Bank. 2018. Global Investment Competitiveness Report 2017/2018: Foreign Investor Perspectives and Policy Implications. Washington, DC: World Bank. doi:10.1596/978-1-4648-1175-3. 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 con- tained 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): 978-1-4648-1175-3 ISBN (electronic): 978-1-4648-1185-2 DOI: 10.1596/978-1-4648-1175-3 Cover design: Bill Pragluski, Critical Stages. Library of Congress Cataloging-in-Publication Data has been requested. Contents Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Anabel Gonzalez, Christine Zhenwei Qiang, and Peter Kusek 1 What Matters to Investors in Developing Countries: Findings from the Global Investment Competitiveness Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Peter Kusek and Andrea Silva 2 Effects of FDI on High-Growth Firms in Developing Countries. . . . . . . . . . . . . . . . . . . . . . . 51 José-Daniel Reyes 3 Corporate Tax Incentives and FDI in Developing Countries . . . . . . . . . . . . . . . . . . . . . . . . . 73 Maria R. Andersen, Benjamin R. Kett, and Erik von Uexkull 4 Outward FDI from Developing Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 Jose Ramon Perea and Matthew Stephenson 5 FDI in Fragile and Conflict-Affected Situations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 Alexandros Ragoussis and Heba Shams Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 v vi CONTENTS Boxes O.1 Global Investment Competitiveness Survey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 1.1 Top Five Findings of the Global Investment Competitiveness Survey. . . . . . . . . . . . . . . . 20 1.2 Investor Motivation Framework According to Dunning and Lundan . . . . . . . . . . . . . . . 22 1.3 MNCs Involved in Efficiency-Seeking Investments Tend to Be More Selective . . . . . . . . 26 2.1 Factors Influencing High-Growth Firms: The Four-Layer Onion Framework . . . . . . . . . 53 2.2 AAA Growers: A High-Growth Firm in Kenya . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 2.3 Chile’s Supplier Development Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.1 The Developing Country Tax Incentives Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 3.2 Methodologies and Results from Country-Level Cost-Benefit Analysis of Incentives . . . 83 3.3 Examples of Transparency-Enhancing Reforms of Tax Incentives . . . . . . . . . . . . . . . . . . 87 4.1 The Evolving Role of OFDI in China’s Economy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106 4.2 Developing Country MNCs Use OFDI to Boost Innovation and Exports . . . . . . . . . . . 117 4.3 Absorptive Capacity Matters at Both Firm and Economy Levels . . . . . . . . . . . . . . . . . . 118 5.1 FCS Are Highly Heterogeneous in Terms of Risks and Opportunities for Investment . 137 5.2 Postconflict Growth in Construction and FDI Opportunities . . . . . . . . . . . . . . . . . . . . 142 5.3 Prioritizing Economic Reforms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Figures O.1 FDI Inflows, Global and by Development Group, 2005–16 . . . . . . . . . . . . . . . . . . . . . . . . 2 O.2 High-Growth Firms Benefit from the Presence of Foreign Firms . . . . . . . . . . . . . . . . . . . . 3 O.3 Factors Affecting Investment Decisions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 O.4 Prevalence of Incentives and FDI Concentration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 O.5 Political Risks Are Prevalent and Discourage FDI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 O.6 Regulatory Predictability and Efficiency Are Critical . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 O.7 Regional Investment Occurs on a Large Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 B1.2.1 Investor Motivation Framework According to Dunning and Lundan . . . . . . . . . . . . . . . 22 1.1 Most Investors Have Multiple Motivations and Are Market-Seeking . . . . . . . . . . . . . . . 23 1.2 Respondents Represent Firms across Various Sectors. . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 1.3 Business-Friendly Legal and Regulatory Environment Is Important for Investors . . . . . . 25 1.4 MNCs Involved in Efficiency-Seeking FDI Are More Selective . . . . . . . . . . . . . . . . . . . . 25 1.5 Investors Seek Predictable, Transparent, and Efficient Conduct of Public Agencies . . . . . . 27 1.6 MNCs Involved in Efficiency-Seeking FDI Value Incentives, Trade Agreements, and Ease of Entry More than Other Investors . . . . . . . . . . . . . . . . . . . . . . . 28 1.7 Duty-Free Imports, Tax Holidays, and VAT Exemptions Are the Most Attractive Investment Incentives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 1.8 Investors Strongly Value Business-Friendly Policies and Procedural Efficiency of Entry and Establishment of Affiliates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 1.9 Wait Times for Investment Approvals Vary but Typically Take Three Months . . . . . . . . 31 1.10 Nearly Half of Material Inputs, Supplies, and Services Are Sourced Locally. . . . . . . . . . 32 1.11 Capacity and Skills of Suppliers Are Critical Linkages-Related Features . . . . . . . . . . . . . 32 1.12 Corporate Programs to Promote Linkages Are Not Very Widespread . . . . . . . . . . . . . . . 33 1.13 More than a Third of Investors Reinvest All Their Affiliate-Generated Profits Back into the Affiliate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 1.14 Severe Political Risks Are Infrequent but Can Have Highly Negative Effects on FDI . . . . . 35 1.15 More than a Quarter of Respondents Had Shut Down an Affiliate in a Developing Country . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 CONTENTS vii 1.16 Reasons for Exiting an Investment Are Mixed, Some Controllable and Others Not . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 1.17 Investors Value IPA Help in Resolving Problems and Setting Up More than Promotion Efforts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 1B.1 MNCs Come from Various Regions and Levels of Development . . . . . . . . . . . . . . . . . . . 40 1B.2 Size of MNCs by Number of Employees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 1C.1 Importance of Country Characteristics by Manufacturing versus Services Firms . . . . . . 45 1C.2 Importance of Country Characteristics by Developed versus Developing Source Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 1C.3 Importance of Country Characteristics by Parent Company versus Affiliate . . . . . . . . . . 47 1C.4 Importance of Investment Climate Factors by Manufacturing versus Services Firms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 1C.5 Importance of Investment Climate Factors by Developed versus Developing Source Economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48 1C.6 Importance of Investment Climate Factors by Parent Company versus Affiliate . . . . . . . 48 B2.1.1 The Four-Layer Onion Framework of Growth Factors . . . . . . . . . . . . . . . . . . . . . . . . . . 53 2.1 High-Growth Firms Create the Most Jobs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 2.2 High-Growth Firms Tend to Be Small... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 2.3 … and Young . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 2.4 High-Growth Firms Benefit from the Presence of Foreign Firms . . . . . . . . . . . . . . . . . . . 59 2.5 The Linkages Channel More Efficiently Transmits FDI Benefits in Nearly All Regions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 2.6 High-Growth Firms Benefit from FDI Mainly through the Linkages Channel, Both in Services and Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 3.1 Tax Incentives Are Widespread in Developing Countries, Especially in Construction and Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 3.2 Policy Makers Continue Cutting Corporate Income Tax (CIT) Rates in Most Regions . . 78 3.3 Nearly Half of Developing Countries Have Introduced New Tax Incentives or Increased the Generosity of Existing Ones . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 3.4 Incentives Are Used Most in Sectors with Heavy Competition for Efficiency-Seeking Investment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83 4.1 Developing Country OFDI Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 4.2 Developing Country OFDI Stocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 4.3 East Asia and Pacific Leads in Developing Country OFDI . . . . . . . . . . . . . . . . . . . . . . . 104 4.4 Top Developing Country Outward Investors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.5 Developing Countries Most Internationalized through OFDI . . . . . . . . . . . . . . . . . . . . 109 4.6 The Location of Developing Country OFDI Varies across Regions . . . . . . . . . . . . . . . . 111 4.7 Developing Country Manufacturing MNCs Prefer Investing via M&A . . . . . . . . . . . . 113 4.8 Developing Countries Have a Mixed Record on OFDI Restriction . . . . . . . . . . . . . . . . 120 4A.1 Distribution of Developing Country OFDI across Income Categories . . . . . . . . . . . . . . 122 4C.1 Distribution of Developing Country OFDI across Industries . . . . . . . . . . . . . . . . . . . . . 125 5.1 FCS Depend on Income from ODA, the Diaspora, and Foreign Investors, 2015 . . . . . . 137 5.2 Agriculture Dominates Highly Fragile Economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 5.3 Postconflict Growth Clocks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 5.4 Foreign Investors Concentrate in Natural Resources and a Few Other Capital-Intensive Activities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 5.5 Outside of Natural Resource Sectors, Investors Are Cautious . . . . . . . . . . . . . . . . . . . . 143 5.6 Regional Investment Occurs on a Large Scale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 5.7 Perceptions on Severity and FCS-Specificity of Challenges, 2016 . . . . . . . . . . . . . . . . . 146 viii CONTENTS 5.8 Senior Management Spent Less Time Dealing with Government Regulations in FCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 5.9 Varying Levels of Fragility and Government Effectiveness among FCS . . . . . . . . . . . . . 148 5B.1 Expected Inward Investment Varies across FCS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156 Maps O.1 FDI Flows to FCS Remain below Potential, 2008–Present. . . . . . . . . . . . . . . . . . . . . . . . . 2 O.2 Growth of OFDI in Most Developing Countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 4.1 More Developing Countries Engage in OFDI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 4.2 Exposure to Developing Country OFDI Rises for Many Developing Host Economies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110 5.1 FDI Flows to FCS Remain below Potential, 2008–14 . . . . . . . . . . . . . . . . . . . . . . . . . . 138 Tables 1B.1 Location of Headquarters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 1B.2 Composition of Respondents Compared with Global FDI Stock . . . . . . . . . . . . . . . . . . . 42 1B.3 Position of Respondents in the Company . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 1B.4 Sectoral Distribution of Respondents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 1B.5 Sectoral Distribution of Respondents Compared with Global FDI Flows . . . . . . . . . . . . 43 1B.6 Number of Motivations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45 2.1 High-Growth Firms Appear in All Economic Sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 2.2 Linkages Are More Important in Manufacturing while Demonstration Effects Are Balanced across Sectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 2D.1 Role of FDI Spillovers on Firm Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 2D.2 Role of FDI Spillovers on Firm Performance, by Regions and Sectors . . . . . . . . . . . . . . . 68 3.1 Pros and Cons of Various Tax Incentives Instruments . . . . . . . . . . . . . . . . . . . . . . . . . . . 77 3.2 Efficiency-Seeking FDI Is Clustered in Few Locations While Natural Resource– and Market-Seeking FDI Is Geographically Dispersed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81 3A.1 Countries in Developing Country Tax Incentives Database . . . . . . . . . . . . . . . . . . . . . . . 90 3A.2 Global Use of Tax Holidays, 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 3A.3 Global Use of Preferential Tax Rates, 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 3A.4 Global Use of Tax Allowances and Credits, 2015 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 3A.5 Changes in Tax Incentives, 2009–15 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 3A.6 Regression Results on Tax Incentives and Foreign Firms’ Perceptions of Tax Rates as a Business Obstacle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 4B.1 Variables and Data Sources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 4B.2 Poisson Pseudo-Maximum Likelihood Estimation Results . . . . . . . . . . . . . . . . . . . . . . 124 5B.1 Regression Coefficients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 Foreword This inaugural issue of the World Bank into their respective economies. Enhancing Group’s Global Investment Competitiveness investment competitiveness thus requires Report presents novel analytical insights and establishing a business environment in which empirical evidence on foreign direct invest- both domestic and foreign companies can ment’s (FDI) drivers and contributions to efficiently enter the market, expand opera- economic transformation. The report tions, and develop more and better linkages focuses on developing countries, given their with local, regional, and global economies. growing role as both sources and recipients This report examines the key dimensions of of FDI, and explores how policy makers and investment competitiveness and highlights local companies can best harness FDI’s those that most commonly influence compa- potential benefits for inclusive and sustain- nies’ investment decisions. able development. The report’s groundbreaking survey of Three key features distinguish this report 754 executives of multinational corpora- from other leading FDI studies. First, its tions investing in developing countries finds insights come from a variety of sources, includ- that—in addition to political stability, secu- ing a new survey of investor perspectives, rity, and macroeconomic conditions—a extensive analysis of available data and evi- business-friendly legal and regulatory envi- dence, and a thorough review of international ronment is the key driver of investment deci- best practices in investment policy design and sions. The report also explores the potential implementation. Second, the report provides of FDI to create new growth opportunities targeted, in-depth analysis of FDI differentiated for local firms; assesses the effectiveness of by motivation, sector, and geographic origin fiscal incentives in attracting FDI; analyzes and destination of investment. Third, the the characteristics of FDI originating in report offers practical and actionable recom- developing countries—so-called South– mendations to developing country South and South–North FDI—and examines governments. the experience of foreign investors in coun- The report introduces a new concept of tries afflicted by conflict and fragility. Future investment competitiveness, defined by the editions of this biennial Global Investment ability of countries to not only attract but Competitiveness Report will present findings also retain and integrate private investment on new sets of investment competitiveness ix x FOREWORD topics high on the agendas of reform- and geographies. For academic audiences, oriented governments, complemented by an the report’s new datasets on investment update of the survey. incentives and FDI motivations offer scope We are confident this new report will for additional research and analysis. Last, for bring value and a fresh perspective to a vari- development assistance providers, the report ety of audiences. For policy makers, the highlights approaches for harnessing FDI’s report offers clear insights into the role of potential development benefits. policy and the decision-making processes of Above all, we recommend this report to all investors. For foreign investors and site loca- audiences interested in the central role that pri- tion consultants, the report discusses relevant vate investment can and must play in furthering FDI developments and drivers across sectors sustainable and inclusive development. Anabel Gonzalez Ted H. Chu Senior Director Chief Economist Trade and Competitiveness Global Practice International Finance Corporation World Bank Group World Bank Group Acknowledgments This inaugural flagship report on global The team is also grateful to the many inter- investment competitiveness was developed nal and external reviewers who provided as a joint initiative of the Investment Climate thoughtful insights and guidance throughout team in the World Bank Group’s Trade and the process, including Cecile Fruman, Neil Competitiveness Global Practice and the Gregory, Mary Hallward-Driemeier, Theodore Economics and Private Sector Development Moran, Richard Newfarmer, Emanuel Salinas, Vice Presidency of the International Finance and Pierre Sauvé. In April 2017, a consultative Corporation (IFC). The report’s prepara- authors, workshop provided added feedback tion was managed by Christine Zhenwei from Nabila Assaf, Sebastien Bradley, Marcio Qiang and Peter Kusek, under the general Cruz, Jan Loeprick, Ernesto Lopez-Cordova, guidance of Anabel Gonzalez, World Bank Denis Medvedev, Sebastien Miroudot, and Group Senior Director for Trade and Gonzalo Varela. The authors are very grateful Competitiveness, and Ted Chu, IFC Chief for the generous time and advice given at vari- Economist. The report’s authors comprised ous stages of this report by external research- Maria R. Andersen, Benjamin R. Kett, ers, including Fritz Foley, Beata Javorcik, Peter Kusek, Jose Ramon Perea, Alexandros Michael Overesch, Karl Sauvant, and Charles Ragoussis, José-Daniel Reyes, Heba Shams, Udomsaph. Andrea Silva, Matthew Stephenson, and The report’s authors are grateful for the Erik von Uexkull. The authors particularly excellent research assistance and overall appreciate the useful advice of Roberto support of Laura Ardila, Abdullah Aswat, Echandi. Angelina Yue Ben, Kunxiang Diao, Zhi Gan, The team would like to thank the follow- Jingyu Gao, Daisy Claire Homolka, Xinyuan ing donors for making this report possible Huang, Salima Madhany, Jordan Pace, through their financial contributions: Martin Schmidt, and Xiaoxu Zhang. Prosperity Fund of the United Kingdom, the The team would also like to recognize Department of Foreign Affairs and Trade various World Bank Group and other col- (DFAT) of the Government of Australia, and leagues for their helpful guidance and the Federal Ministry of Finance of the assistance—without them this report would Government of Austria. not have been complete. These include xi xii ACKNOWLEDGMENTS Daniela Gomez Altamirano, Gerlin Catangui, The Global Investment Competitiveness Laura Dachner, Wim Douw, Persephone Survey analyzed in chapter 1 was undertaken Economou, Amr El Afifi, Francis Gagnon, on behalf of the World Bank Group by Ulla Heher, Armando Heilbron, Sebastian Kantar Public, a global survey firm. We are James, Priyanka Kher, Kathy Khuu, Barbara particularly grateful for the contributions of Kotschwar, Jana Krajcovicova, Hania the following Kantar Public colleagues: Kronfol, Veselin Kuntchev, Wim Naude, Ivan Jamie Burnett, Lavinia Deaconu, Christopher Nimac, Ngan Thuy Nguyen, Nadia Piffaretti, Hanley, and Marco Pelucchi. Yassin Sabha, Patricia Steele, Trang Tran, and The team acknowledges the many formal Robert Whyte. and informal contributions of individuals, David M. Cheney was the principal edi- groups, and organizations that provided tor of the report. Andrea Silva, Amanda L. meaningful comments and inputs to enable Tan, and Edward Atkinson provided edito- the successful publication of this report. rial assistance. Production and logistics Keeping in mind these contributions, the support was provided by Aziz Gökdemir team apologizes if it has inadvertently omit- and Jewel McFadden. The communica- ted formally acknowledging any such valu- tions efforts were led by John Diamond able contributions. and included Egidio Germanetti, Amelia Kelly, Kristina Nwazota, and Madelynne Wager. Abbreviations BRICS Brazil, the Russian Federation, India, China, South Africa BIT bilateral investment treaty CATI computer-assisted telephone interviews CBA cost–benefit analysis CCSD Center on Conflict, Security and Development CIT corporate income tax CORFO Chilean Economic Development Agency (Corporación de Fomento de la Producción de Chile) CPIA Country Policy and Institutional Assessment EAP East Asia and Pacific ECA Europe and Central Asia EU European Union FCS fragile and conflict-affected situations FDI foreign direct investment FFP The Fund for Peace GDP gross domestic product GIC Global Investment Competitiveness GVC global value chain ICT information and communications technology IEG Independent Evaluation Group IFC International Finance Corporation IPA investment promotion agency ISIC International Standard Industrial Classification IT information technology LAC Latin America and the Caribbean M&A mergers and acquisitions xiii xiv ABBRE VIATIONS MENA Middle East and North Africa MNC multinational corporation ODA Official Development Assistance OECD Organisation for Economic Co-operation and Development OFDI outward foreign direct investment POEs privately owned enterprises PPML Poisson Pseudo-Maximum Likelihood R&D research and development SAR South Asia Region SDGs Sustainable Development Goals SDP Supplier Development Program SEZ special economic zone SMEs small and medium-size enterprises SOEs state-owned enterprises SSA Sub-Saharan Africa UCC user cost of capital UNCTAD United Nations Conference on Trade and Development UNSD United Nations Statistics Division USAID United States Agency for International Development VAT value added tax WBG The World Bank Group WEF World Economic Forum Overview Anabel Gonzalez, Christine Zhenwei Qiang, and Peter Kusek Foreign Investment Is a Major foreign markets. Furthermore, FDI has a sig- nificant potential to transform economies Contributor to Development through innovation, enhancing productivity, For many developing countries, 1 foreign and creating better-paying and more stable direct investment (FDI) has become the larg- jobs in host countries, in sectors attracting est source of external finance, surpassing FDI as well as in the supportive industries official development assistance (ODA), (Arnold, Javorcik, and Mattoo 2011; remittances, or portfolio investment flows. Bijsterbosch and Kolasa 2009; Echandi, In 2016, more than 40 percent of the nearly Krajcovicova, and Qiang 2015; Rizvi and $1.75 trillion of global FDI flows was Nishat 2009; WEF 2013). Importantly, for- directed to developing countries, providing eign investors are becoming increasingly much-needed private capital (figure O.1). prominent players in delivering global public Yet the financing required to achieve the goods, addressing climate change, improving Sustainable Development Goals (SDGs) 2 labor conditions, setting global industry stan- remains prohibitively large and largely unmet dards, and delivering infrastructure to local by current FDI inflows—especially in frag- communities (IFC, forthcoming). This report ile and conflict-affected situations (FCS) builds on the literature in highlighting the role (map O.1). To maximize the development of FDI in upgrading growth and adding value impact of FDI and thus help meet the SDGs, to domestic firms, in filling the investment private investment will have to expand into void in FCS, and more generally, in increasing areas where it has not yet ventured, notwith- competitiveness and stability. standing the associated risks. FDI can accelerate productivity gains in The benefits of FDI extend well beyond host countries. It brings foreign technology attracting needed capital. Foreign investment and frontier knowledge that, if successfully also confers technical know-how, managerial absorbed by local firms, can improve their and organizational skills, and access to productivity directly. FDI can also increase 1 2 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE O.1 FDI Inflows, Global and by Development Group, 2005–16 2.0 1.8 1.6 1.4 1.2 US$, trillions 1.0 0.8 0.6 0.4 0.2 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Developed economies Developing economies World Source: Statistics and World Investment Report 2017, United Nations Conference on Trade and Development (UNCTAD). Note: FDI = foreign direct investment. MAP O.1 FDI Flows to FCS Remain below Potential, 2008–Present Investment BOSNIA & Actual HERZEGOVINA Potential IRAQ AFGHANISTAN LEBANON Investment size 70,490 60,000 NEPAL 40,000 MALI 20,000 THE GAMBIA CHAD HAITI GUINEA ERITREA GUINEA-BISSAU SUDAN LIBERIA CENTRAL AFRICAN REPUBLIC 7 TOGO (Numbers in million $US) CÔTE D'IVOIRE SOMALIA D. R. OF CONGO BURUNDI MALAWI COMOROS SOLOMON MADAGASCAR ISLANDS ZIMBABWE IBRD 43085 | SEPTEMBER 2017 Source: Computation based on Investment Map Database, International Trade Centre; World Development Indicators, the World Bank; CEPII Database; Fragile States Index (2014), the Fund for Peace. Note: Investment potential is calculated as foreign direct investment (FDI) inflow estimates without the negative effect of fragility. They are calculated for selected fragile and conflict-affected situations (FCS) based on countries’ economic fundamentals (market size, growth, trade openness, savings), geo- graphical remoteness, and abundance of natural resources, where the negative effect of fragility is removed. competition among firms in the local market transmission of foreign firms’ technology, by leading to a reallocation of resources away knowledge, and practices, as well as from less productive to more productive requirements that may help domestic firms, thereby increasing aggregate productiv- suppliers upgrade their technical and ity over the long run. FDI can benefit domes- quality standards (Du, Harrison, and tic firms mainly through linkages and Jefferson 2011; Farole and Winkler 2014; demonstration channels: Javorcik and Spatareanu 2009). A recent study in Turkey suggests that interactions • Linkages between foreign firms and local between multinational corporations partners or suppliers can promote (MNCs) and their Turkish suppliers OVERVIEW 3 facilitate an upgrading of Turkish prod- two years for the average high-growth firm. For ucts (Javorcik, Lo Turco, and Maggioni the demonstration channel, an increase of 2017). Firm-level analyses from Lithuania 1 percentage point in the share of foreign out- and Vietnam present evidence that there put in the sector is correlated with a 0.1 unit are positive productivity spillovers from gain in output growth of high-growth firms, or FDI through linkages between foreign 12 percent increase in sales over the two years affiliates and their local suppliers in the for the average high-growth firm (figure O.2). upstream sectors (Javorcik 2004; While high-growth firms usually account Newman and others 2015). for only a small part of the private sector, • The demonstration effect , in which they have a disproportionately large role in domestic fi rms imitate foreign technolo- job creation and productivity gains. They are gies and managerial practices either better able to maximize the benefits from through observation or by hiring work- FDI because of their higher absorptive ers trained by foreign companies (Alfaro capacities—their ability to recognize the and Chen forthcoming; A lfaro and value of, assimilate, and apply new informa- Rodriguez-Claire 2004; Alfaro and oth- tion. Such abilities allow these firms to inter- ers 2006; Barba Navaretti and Venables nalize foreign technologies and processes to 2004; Lipsey 2004), is another key chan- improve their productivity, thereby dampen- nel benefitting firms in host countries. ing the competitive impact of rivalry with For example, the contribution of work- foreign-established firms. Furthermore, the ers’ mobility from MNCs to domestic demands of global brands, and their firms in the Ghanaian manufacturing sectors has had a positive impact on the productivity of domestic enterprises. In Norway, workers with prior experience FIGURE O.2 High-Growth Firms Benefit from the Presence of in MNCs contribute 20 to 25 percent Foreign Firms Average impact of FDI spillovers on firm growth, by firm type more to productivity than workers with- out such experience (Balsvik 2006; Görg 0.80 and Strobl 2005). 0.70 0.62 0.60 Point estimates and 90% confidence intervals High-Growth Firms in Host 0.50 Countries Benefit Most from FDI 0.40 This report analyzes the ability of domestic firms to benefit from the presence of MNCs, 0.30 drawing on firm-level information across 50 manufacturing and services sectors and 121 0.20 economies in the developing world from the 0.12 0.10 0.05 World Bank’s Enterprise Surveys. It finds that 0.01 local high-growth firms (defined as the subset of 0 enterprises with the highest job creation rates) –0.10 are most able to internalize FDI spillovers— through both linkages and demonstration chan- –0.20 nels. For the linkages channel, an increase of High-growth Rest High-growth Rest 1 percentage point in the share of inputs firms firms sourced domestically by foreign firms is corre- Linkages channel Demonstration channel lated with a 0.6 unit rise in the measure of out- Source: Computation based on data from Enterprise Surveys, the World Bank. Note: This figure shows the estimated coefficient and 90% confidence interval of the linkages and put growth of domestic high-growth firms. This demonstration channels on high-growth firms and the rest of businesses in a sample of 121 econo- result implies a 58 percent increase in sales over mies. Vertical lines capture 90% confidence intervals. FDI = foreign direct investment. 4 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 commitment to their suppliers, create a technology network is now emerging, with strong incentive and impetus for suppliers to growing South–South and South–North adopt new practices and invest in new tech- innovation-oriented interaction and collabo- nologies. From a policy perspective, identify- ration (Nepelski and De Prato 2015). This ing and targeting these firms, analyzing the may be partly because knowledge originating constraints on their emergence, and deepen- in developing countries may be better suited ing their absorptive capacities are all key to to other developing country settings, and unleashing their full potential. The empirical because the level of complexity of that evidence presented in this report indicates knowledge may be more easily absorbed by that policies that encourage FDI linkages as a other economies at similar levels of develop- way for high-potential indigenous firms to ment. This report highlights how the grow will enhance knowledge transmission increased absorptive capacity of firms in between MNCs and domestic firms, and source markets can promote a wide disper- deliver strong development results. sion of outward FDI benefits in the home economy. Despite abundant evidence on the devel- Outward FDI Also Benefits opment benefits of FDI, the global economic outlook remains uncertain, clouded by Source Economies risks of trade and investment protectionism FDI brings benefits not only to destination and geopolitical risk. While globalization markets but also to source economies (“home brings aggregate productivity and economic country effects”). MNCs from developing growth, it may also bring hardship for low- countries use outward investment to productivity firms and low-skill workers. strengthen their capabilities and competitive- Slow public policy responses to rapidly ness by entering new markets, importing evolving patterns of investment and eco- intermediate inputs from foreign affiliates at nomic activities contribute to misconceptions lower prices, producing a larger volume of and oversimplification of features and poten- final goods and services abroad at lower cost, tial effects of FDI. In certain countries, oppo- and accessing foreign technology (Herzer nents of FDI-led integration further contend 2012). Some developing countries, instead of that its effects are often limited and, in some exploiting existing technological assets, aim cases, detrimental—as it crowds out local to acquire new ones through outward FDI. competition, results in enclave production Case studies of leading MNCs from BRICS with limited linkages, and engenders a “race countries (Brazil, the Russian Federation, to the bottom” in labor or environment stan- India, China, and South Africa) show that dards or in their enforcement.3 Not surpris- they are disadvantaged in terms of patents, ingly, policy discussions increasingly management know-how, or cutting-edge pro- distinguish between “good” and “bad” FDI. cesses, which prompt them to acquire Some argue that a foreign presence can lead companies abroad to permit “late-comer to political grievances through its adverse catch-up” (Holtbrügge and Kreppel 2012; effects on the distribution of income and Rodriguez-Arango and Gonzalez-Perez opportunities, particularly concerning FDI in 2016; UNCTAD 2005). extractive industries (International Dialogue Outward FDI by developing countries can for Peace-Building and State-Building 2016). bring significant economic advantages back Others, however, find that trade and FDI to source economies, especially enhanced complement each other in reducing the risk innovation. While developed countries were of conflict (Polachek and Sevastianova once seen as the prime source of knowl- 2012). The truth is that there are different edge and technology—thus imparting a types of FDI, each with different potential North–North or North–South bias to cross- social, economic, and environmental effects. border investment—a multipolar global Further, evidence shows that there is not OVERVIEW 5 intrinsic “good” or “bad” FDI. Rather, there host countries. Some countries may attract are good or bad policies that can or cannot FDI yet not enable its entry and establish- lead countries to fully reap the potential ben- ment, or enable its establishment yet not its efits of FDI for development (Echandi, expansion and “rooting” in the host economy Krajcovicova, and Qiang 2015). through linkages and other spillovers. These On balance, the bulk of the research and point to the need for a more nuanced analysis empirical evidence finds that FDI helps to of FDI impacts. foster development in recipient economies. Though some of the above criticisms are warranted and the distributional effects of the Investment Decisions Are different types of FDI merit closer study, evi- Influenced by Risk–Return dence for such claims is often anecdotal and applicable to only a narrow subset of indus- Calculations tries and economies. As this report shows, the Investors consider a broad range of factors benefits of FDI can be strongly magnified in in their decision to invest, including domes- economies with good governance, well- tic market size, macroeconomic stability and functioning institutions, and transparent and a favorable exchange rate, labor force talent predictable legal environments. Moreover, and skills, and physical infrastructure. not all types of FDI nor all stages in the According to the Global Investment investment life cycle4 exert the same effects on Competitiveness (GIC) survey (box O.1), BOX O.1 Global Investment Competitiveness Survey The Global Investment Competitiveness (GIC) survey 2. Importance of factors in investing in a developing was commissioned by the World Bank Group as a com- country, where respondents rate the importance panion piece of the GIC report to bring data and infor- of country characteristics and investment policy mation on the views and behavior of global investors factors on a scale from 1 to 4 from “not at all that goes beyond anecdotal evidence. Phone interviews important” to “critically important.” “Critically were conducted between February and June 2017 with important” means it is a deal-breaker—by itself, it 754 international business executives involved with could change the company’s decision about whether the operations of their multinational corporation in to invest or not in a country. developing countries. Respondents come from both developed and developing countries and represent a 3. Political risks and investment exit, where respon- wide range of sectors. dents identify experiences of political risks and the company’s course of action. They were also asked The GIC survey captures perceptions of these about experience of shutting down a foreign affili- investors on the role of investment climate factors ate in a developing country and their reasons for in guiding their FDI decisions. It complements other doing so. existing investor surveys by focusing on variables such as administrative and legal barriers rather than 4. Investment in a specific developing country, where broader economy-wide factors. These specific invest- respondents select a specific developing country ment climate variables are areas that are actionable where they are most familiar with the operations for policy makers. of the affiliate. Questions on the specific invest- ment included sector, activity, motivation, rein- The survey is composed of four sections: vested earnings, efficiency of government agencies, 1. General information on the company and respon- availing services of investment promotion agencies, dent, including sector, number of employees, and incentives received, sources of inputs, and corpo- position of the respondent in the company. rate programs for suppliers. 6 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 political stability and a business-friendly reg- Yet investment incentives become relevant ulatory environment are most important in only when investors waver between similar investors’ decision making (figure O.3). locations. Where FDI is motivated by access to Macroeconomic, political, and regulatory domestic markets or natural resources, incen- risks—whether actual or perceived—deter tives are generally of limited value. However, in investors by raising their risk calculations. sectors where FDI is mainly efficiency-seeking De-risking, or reducing project or country in nature (for example, manufacturing of risk, can lead to the right risk–return profile information technology [IT] and electronics, and help attract private investment. machinery and equipment, automotive, air- Otherwise investments that are commercially and spacecraft, and biotechnology and phar- profitable and economically attractive may maceuticals), competition for FDI is high and simply not materialize. developing countries frequently offer incen- Governments in both developing and tives to compete. In these sectors, most FDI developed countries use tax and other invest- projects are clustered in a limited number of ment incentives to reduce the relative cost or successful host countries. At the same time, risks to foreign investment so as to attract the use of incentives is particularly prevalent in more FDI, often not distinguishing among the these sectors (figure O.4, upper right quad- different types of FDI.5 Given that most coun- rant). This suggests that developing countries tries offer incentives, investment promotion use incentives strategically in sectors with high agencies face pressure to match or even sur- shares of efficiency-seeking FDI where loca- pass offers by competing countries to com- tional competition for FDI is particularly pensate for adverse geography, small size, or intense. It also reveals that, while incentives distance to markets, in order to remain attrac- may be a more important part of the value tive for foreign investors. proposition to efficiency-seeking investors, FIGURE O.3 Factors Affecting Investment Decisions Share of respondents (percent) Importance of country characteristics Political stability and security 50 37 9 2 Legal and regulatory environment 40 46 12 2 Large domestic market size 42 38 14 4 Macroeconomic stability and favorable 34 44 16 5 exchange rate Available talent and skill of labor 28 45 22 5 Good physical infrastructure 25 46 24 5 Low tax rates 19 39 31 9 Low cost of labor and inputs 18 35 35 11 Access to land or real estate 14 31 32 22 Financing in the domestic market 16 28 31 24 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: Multinational corporation executives were asked how important these characteristics were in their decision to invest in developing countries. OVERVIEW 7 FIGURE O.4 Prevalence of Incentives and FDI Concentration (Incentives are used most in sectors with high competition for efficiency-seeking FDI) 0.8 Air- and spacecraft Automotive industry and other transport equipment Machinery and equipment 0.7 Low to high geographic clustering of FDI projects Other manufacturing IT services HHI of geographic concentration of FDI IT and electronics Apparel, textile, and footwear 0.6 Trade and retail Biotechnology, pharmaceuticals, and Transport and medical products logistics services Power, utilities, and Agriculture and fishing Construction and 0.5 telecomm u nic a tions building materials Business services Education and health Food and beverages Entertainment Tourism and hospitality 0.4 Extractive industries Financial services 0.3 45 50 55 60 65 70 75 Share of countries offering incentives or CIT rates ≤ 15% Low to high degree of tax competition Mostly efficiency-seeking FDI Mostly natural resource–seeking FDI Mostly market-seeking FDI Source: Developing Country Tax Incentives database and FDI data from fDi Markets database, the Financial Times. Note: The size of each bubble represents the number of FDI projects within the sector in developing countries. This was constructed based on information from the fDi Markets database. CIT = corporate income tax; FDI = foreign direct investment; HHI = Herfindahl-Hirschman Index; IT = information technology. they are not a sufficient condition for FDI entry, improvements in the design, transparency, as efficiency-seeking FDI tends to concentrate and administration of incentives can help geographically in relatively few locations reduce indirect costs and unintended conse- despite the broad availability of incentives. quences including economic distortions, red More targeted, transparent, and cost- tape, and corruption. Such policy reforms effective use of investment incentives can can greatly improve the cost–benefit ratio of improve their impact. By targeting incen- incentives. tives toward those investors most likely to respond to them, developing countries can reduce unnecessary tax losses resulting from Governments Play a Key Role in incentives granted to firms that would have invested anyway. This requires a thorough De-Risking Private Investment understanding of the type and motivation Reducing the risks of private investment at for FDI in the country, as well as measur- the project level does not compensate for able policy objectives. At the same time, failing to de-risk regulations and institutions 8 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 at the country level. Investment incentives or Political risks are wide-ranging and include investment guarantees are frequently used to expropriation, transfer and convertibility bolster locational competitiveness or invest- restrictions, breach of contracts, unpredictable ment viability for specific projects or sectors, and arbitrary actions, discrimination, and the but investment climate weaknesses must be absence of regulatory transparency. Loss of addressed first. If fundamental elements at investment and the associated damage to long- the country level are lacking, investors are term harmonious relations with a promising unlikely to respond to even the most gener- investor can have a debilitating impact on a ous incentive packages or such incentives developing country. Political risk related to may only attract unviable investments. government conduct also sends negative sig- Governments can reduce risks to private nals to prospective investors, creating strong investors through a policy and institutional ripple effects. framework that supports an enabling busi- More than three-quarters of investors sur- ness climate and ensures good governance. veyed in this report encountered some type of Since reliable regulations and institutions are political risk in their investment projects in key to de-risking private investment at the developing countries. In severe cases, such as country level, they are an increasingly impor- expropriation, about half of the investors can- tant element on the Maximizing Finance for celed a planned investment or withdrew an Development agenda. existing one (figure O.5). Legal protection to In this report, de-risking involves removing investors against such risk is usually provided or reducing political and regulatory risks by “investor protection guarantees” typically caused by government action, building on included in a country’s domestic legal frame- macroeconomic stability and good infrastruc- work and its international investment agree- ture in order to attract private investment. ments (IIAs). In this report’s survey, 81 percent FIGURE O.5 Political Risks Are Prevalent and Discourage FDI Share of respondents (percent) Lack of transparency and predictability 23 24 27 14 11 in dealing with public agencies (50%) Sudden change in the laws and regulations 27 25 25 11 11 with a negative impact on the company (49%) Delays in obtaining necessary government permits 20 17 37 13 12 and approvals to start or operate a business (47%) Restrictions in the ability to transfer 26 20 29 11 11 and convert currency (42%) Breach of contract by the government (13%) 14 23 26 20 15 Expropriation or taking of property 33 5 10 13 36 or assets by the government (5%) Don’t know None Consider delay or Significantly delay Cancel planned Withdraw existing cancellation investment investment investment Source: Computation based on the GIC Survey. Note: FDI = foreign direct investment. OVERVIEW 9 FIGURE O.6 Regulatory Predictability and Efficiency Are Critical Share of respondents (percent) Importance of investment climate factors Transparency and predictability in 37 45 15 3 the conduct of public agencies Investment protection guarantees 45 36 14 4 provided in the country’s laws Ease of obtaining government approvals to start 36 41 18 5 a business and to own all equity in the company Investment incentives such as tax holidays 21 35 32 10 Having a preferential trade agreement 14 40 32 11 Having a bilateral investment treaty 15 36 33 13 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey of investors rate country legal protections Developing Country MNCs and 51 percent rate bilateral investment trea- ties as important or critically important in Are Today an Increasing their investment decisions (figure O.6). Such Source of FDI findings echo the literature documenting FDI from developing countries has increased the generally positive impacts of IIAs on twentyfold in the last two decades, account- FDI inflows (Echandi, Krajcovicova, and ing for nearly one-fifth of global FDI flows in Qiang 2015). 2015. As such, contribution of Southern Investors also seek predictability and effi- MNCs to economic development of emerg- ciency in the implementation of laws and reg- ing markets is significant, especially given ulations (figure O.6). About four out of five low investor confidence prevailing today investors surveyed rate the transparency and among traditional Northern MNCs. Despite predictability of public agency conduct—and a fall in FDI from Organisation for Economic the ease of doing business—as important Co-operation and Development (OECD) determinants of their locational decisions. countries by 57 percent below 2007 levels in This is not surprising, since many developing 2012, FDI from developing countries rose by countries have inefficient bureaucracies, 19 percent (OECD 2014). While larger opaque regulations, complex procedures, and developing countries, especially the BRICS, high transaction costs—all of which can are driving this phenomenon, about 90 per- undermine their competitiveness. More than cent of developing countries of all sizes and one-third of investors rate these as critically income levels are now undertaking outward important factors or potential deal-breakers. foreign direct investment (OFDI) (map O.2). Predictability and efficiency are essential Both domestic policy choices in developing ingredients of sound and sustained interaction countries and global economic conditions between MNCs and host governments, com- have shaped changes in the investment land- prising both regulations themselves and their scape. Firms in Singapore and other high- implementation. growth economies embraced OFDI in the 10 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 MAP O.2 Growth of OFDI in Most Developing Countries 1995 OFDI Stock/GDP 0% 5% 10% 15% >20% 2015 OFDI Stock/GDP 0% 5% 10% 15% >20% IBRD 43087 | AUGUST 2017 Source: UNCTAD and World Development Indicators, World Bank. Note: GDP = gross domestic product; OFDI = outward foreign direct investment. OVERVIEW 11 late 1990s and early 2000s as a development Developing Country and strategy to “achieve efficiency in resource allocation and diversify risks from economic Regional Investors Target shocks in any one region” (Lee, Lee, and Yeo Higher-Risk Markets 2016). Developing country investors may also be Firms in other developing economies soon more willing to target higher-risk markets emulated such efforts, with OFDI increas- in host economies with weaker institutional ingly seen as a means to access markets, quality.6 In 2001, only 11 countries in the capital, technology, and knowledge in inter- developing world (5 in Sub-Saharan Africa, national markets—and thus boost firm-level 5 in Europe and Central Asia, and 1 in and national competitiveness (Luo, Xu, and Latin America and the Caribbean) had half Han 2010). Global economic conditions or more of their inward FDI stock coming also “pulled” developing-market firms into from investors from other developing coun- OFDI. First, rapid and sustained growth in tries. In 2012, that number had risen to 55 much of the developing world during the countries. Developing countries are a par- last two decades helped firms to grow and ticularly key source of FDI for countries in prosper and, consequently, to international- Sub-Saharan Africa, Europe and Central ize. Second, the commodity super-cycle Asia, and South Asia. With many of these (until recently) gave some developing coun- host economies characterized by low levels try exporters large windfalls, creating sub- of economic development, such trends stantial liquidity that was used partly to accord with the literature, which finds finance OFDI. developing country OFDI to be less discour- The emergence of developing countries as aged by weak institutional and economic a key source of FDI begs the question of host-country environments (Cuervo- whether they differ from developed countries Cazurra 2008; Dollar 2016; Ma and Assche in terms of the drivers and risk tolerance of 2011) owing to the “institutional advan- their OFDI. Both the report’s investor survey tage” argument (Cuervo-Cazurra and Genc and data analysis suggest that developing 2008). This argument suggests that manag- country OFDI reacts to standard host econ- ers of developing country MNCs are more omy locational determinants (for example, accustomed to uncertainty and may be more market size, income level, distance, common adept in dealing with unpredictable regula- language, colonial links) in much the same tory practices and less transparent adminis- way as developed country OFDI. Both are trative procedures. Several studies support attracted to large and growing economies that this argument, finding that developing are geographically close and culturally country OFDI investors are relatively more similar. present in least developed countries. Some Developing country investors are more demonstrate an inverse relationship between willing to target smaller and closer economies host country political risk and, for example, (Arita 2013) in a “stepping-stone” strategy. Chinese OFDI (Cui and Jiang 2009; Evidence suggests that some of these firms Duanmu and Guney 2009; Kang and Jiang find it difficult to compete in larger, more 2012; Quer, Claver, and Rienda 2015). competitive, and more distant markets, not Risks in FCS range from security and least because they often lack the networks value-chain disruptions to regulatory, finan- and experience of developed country firms. cial, and reputational uncertainty, all of Studies from Asia and Latin America find that which make foreign investors reluctant to regional investors usually expand into larger engage. In many cases, governments lack the and more complex markets only after first capacity and revenue base to perform basic successfully expanding in smaller, lower- functions. Often, informal and noninclusive income economies in the same region institutions fill the governance vacuum, and (Cuervo-Cazurra 2008; Gao 2005). 12 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 their interaction with businesses is frequently resource-rich economies. Such FDI targets a motivated by rent-extraction. Firms also face handful of sectors, all of which are capital an array of adverse market conditions simi- intensive and sustained mostly by foreign lar to those in other low-income countries, demand. Investors are more cautious when such as weak macroeconomic and regulatory they enter FCS markets: they tend to commit environments, infrastructure bottlenecks, to smaller projects that produce fewer jobs and a limited supply of skilled labor, com- for every dollar invested and tend to concen- pounded by low demand. However, unlike in trate their investment spatially in the most many developing countries, the destruction stable regions or cities in FCS. of physical and human capital and dimin- Regional investors may have a compara- ished state control result in highly risky busi- tive advantage in FCS contexts relative to ness environments. As a result, FDI in FCS global firms. A considerable amount of green- represents a mere 1 percent of global flows, field investment in FCS comes from regional more than five times lower than the world firms (figure O.7). The investment footprint average. Despite having increased tenfold of France and the United Kingdom remains over the last two decades, the distribution of large in Africa and the Middle East, but FDI directed to FCS is still mostly concen- greenfield investments (for example, from trated in a handful of middle-income or Russia to Uzbekistan, Malaysia to Cambodia, South Africa to Nigeria, Japan and Thailand to Myanmar, and the United Arab Emirates FIGURE O.7 Regional Investment Occurs on a Large Scale Origins of greenfield FDI project announcements in FCS (2008–16) to Iraq) confirm that intraregional investment takes place in FCS on a large scale. Other FCS recipients regional investors include, for example, com- panies from Lebanon investing in neighboring Middle Eastern countries, companies from Pap Congo, Rep. Morocco expanding into markets in North Zimbabwe ua Cambo n Cam Ne Africa, and Nigerian firms expanding into e wG Franc eroo s Uz te West Africa. These firms leverage their supe- dia uin ta be dS k ea rior knowledge of the local context and their ist ite an Un An affinity with their target markets. As a result, go la an such investors show greater resilience, take Jap more risks (for example, committing to larger Arab United projects), and accept lower returns. This trend a te s Emir highlights once more the importance of Myanmar Russian Federation regional sources of investment, and of Thail regional integration schemes, in transitioning and Kor out of fragility. q ea, Ira Un Rep First movers willing to make pioneer ite . dK investments in challenging environments in M ing au do FCS are critical for signaling the viability of Ca h Afri rit m Sou f OECD na ria ius Rest business opportunities in these markets. Malaysia India da Netherlands Nige t MNCs operating in FCS often make strategic o ca choices in terms of scale, staffing, and loca- tion that seek to address multiple challenges Origin of greenfield investments and risks simultaneously. Some of the Intraregional Interregional response strategies documented by interviews Source: Computation based on fDi Markets database, Financial Times. with investors (IFC 2017) include integrated Note: Origins (on the right side of the chord diagram, in orange) and FCS destinations (on the left management and due diligence systems; stra- side, in red) of greenfield projects exceeding US$3 billion since 2008. Blue chords indicate intra- regional investment. FCS = fragile and conflict-affected situations; FDI = foreign direct investment; tegically locating warehouses and production OECD = Organisation for Economic Co=operation and Development. sites; staged investments; striving to meet OVERVIEW 13 international standards; and flexibility in to occur—it took the fastest-reforming coun- scale, supply, and business plans. Pioneering tries in the 20th century two decades to investments can help host-country govern- achieve a functioning governance quality— ments develop regulations and support ser- and the scope and speed of reforms are them- vices, establish business and consumer selves risk factors (World Bank Group 2011). markets, and generate positive externalities. But strengthening public institutions that pro- They also offer a demonstration effect to vide citizens with security, justice, and jobs is other investors that the target countries and crucial to breaking the cycle of violence. markets are open to financially viable invest- ments despite high risk perceptions. The rest of this report is organized around five thematic chapters, each exploring a dif- ferent dimension of FDI in developing Investment Climate Reforms countries: Reduce Uncertainty and The discussion of the findings from the GIC survey (chapter 1) aims to help policy Unpredictability makers design policies and prioritize Investment climate reforms are necessary reforms valued by foreign investors. for markets to move from conflict to peace Through some 750 interviews of executives and from fragility to resilience. Firm-level of MNCs with investments in developing responses are limited in what they can countries, the survey measures the role of achieve—investors may strive to keep their investment climate variables (for example, companies out of harm’s way, but they can investment incentives, investment promo- only go so far in coping with them and can- tion activities, FDI regulations, and admin- not address these risks in a holistic and sys- istrative processes) in influencing FDI temic way. Investment climate reforms decisions. By identifying factors that are tailored to the context of FCS, however, important to investors, reform-minded gov- can go a long way toward reducing inves- ernments can leverage policy instruments tors’ risks and creating markets for viable that can most effectively attract, retain, and investment. The limited capacity of many leverage FDI for development. Recognizing governments in FCS, combined with the resource constraints faced by most govern- urgency of positive returns on reform ments, the authors of this chapter suggest efforts, require the proper sequencing and where policy makers can focus their efforts prioritization of interventions. An invest- to maximize impact. ment climate engagement must be imple- FDI in developing countries benefits local mented in a balanced way by securing high-growth firms the most (chapter 2). short-term gains while building the momen- This is due likely to their higher absorptive tum for deeper institutional transformation capacity—that is, their ability to recognize the over the longer term. value of, assimilate, and apply new informa- Regulatory simplification, removing barri- tion to improve production processes. High- ers to investment entry, and addressing infra- growth firms account for a sizable share of structure constraints (for example, access to job creation and productivity gains in devel- electricity and transit) rank among the most oping countries. The distinctive characteris- important confidence-building signals that tics of these firms have been the subject of can produce early results and trigger a private study from the perspective of both individual sector response. Value chain development firms interested in sales and revenue growth through skills building, access to finance and and policy makers interested in job creation technology, and connecting producers to and economic growth. The findings discussed markets can be second-stage interventions in this chapter have strong implications for suitable for FCS (World Bank Group 2011). programs aimed at facilitating the connection Deeper institutional reforms may take longer of domestic firms to established MNCs 14 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 through government-supported linkage domestic investment, employment, and eco- programs. nomic growth—is still nascent. Tax incentives play a role in FDI in devel- The discussion of FDI in fragile and oping countries (chapter 3), and the authors conflict-affected situations (chapter 5) fills a of this chapter offer practical evidence to help gap in understanding the potential, patterns, developing country policy makers design and and constraints of FDI in such states and implement more effective incentives. Using a explores ways to support investments that new dataset on tax incentives in developing have a positive effect on peace and stability. countries compiled by the World Bank Group, The authors draw on original data and analy- the authors provide sector- and firm-level evi- sis of investment in high-risk environments to dence to guide policy makers on how to tar- explain investment decision-making and cop- get investment incentives more efficiently. The ing mechanisms in such contexts. They pro- analysis assesses how developing countries pose an approach to investment climate use tax incentives by sector and over time, reforms that aims at securing short-term gains links the effectiveness of incentives to a simple while building the momentum for deep insti- framework of investor motivation to guide tutional transformation. Key elements of that policy makers in these targeting questions, strategy focus on reducing risks to investors and presents new evidence on the relevance of as well as maximizing investment opportuni- tax incentives for investors. Tax incentives are ties and rewards. found to be commonly used by developing countries, with some variation across sectors and regions, and tend to be more effective in Notes attracting efficiency-seeking FDI. The authors 1. “Developing countries” in this report refer to also identify priorities for the design, trans- low- and middle-income countries as defined parency, and administration of incentive by the World Bank. The full list of countries reforms. appears in the glossary. The list is based on Developing country OFDI has increased income categories in fiscal year 2017 at http:// considerably in recent years (chapter 4), and databank.worldbank.org/data/ download the authors of this chapter explore its main /site-content/OGHIST.xls. drivers and offer policy proposals to maxi- 2. The 17 SDGs of the 2030 Agenda for mize its development impact. They use several Sustainable Development were adopted by global data sources to assess changes over world leaders at a United Nations summit in September 2015 and are listed in http://www time in the investment patterns of developing .un.org/sustainabledevelopment/sustainable country MNCs, particularly with regard to -development-goals/. source and destination economies, target sec- 3. According to work by Theodore Moran tors, and modes of entry. The authors com- (2014), the evidence actually indicates that plement this information with findings from a the entry of foreigners and their first-tier sup- gravity model to explain the influence on pliers introduces “Schumpeterian winds of developing country OFDI behavior of several creative destruction” that may lead to ben- FDI location determinants, such as relative eficial restructuring of the entire industry, market size, geographical distance, common including opportunities for better-performing cultural and institutional features, and the local companies in the same industry, and for existence of bilateral investment agreements. suppliers in the vertical industries to emerge over time. Moran notes that the outcome They also consider whether OFDI can foster to observe is the changing economic per- the development of source economies and formance of the entire sector, as opposed review the relevant literature. They offer evi- to arbitrary measurement of the absolute dence that OFDI increases home firm innova- amount of capital invested at any particular tion and exports, but the literature on other moment in the sector (as is often highlighted aggregate benefits—such as productivity, in the debates on crowding-in and crowd- ing-out). The Czech Republic is a good OVERVIEW 15 example of how acquisition of a dilapidated Arita, S. 2013. “Do Emerging Multinational local carmaker Skoda by Volkswagen, one Enterprises Possess South-South FDI of the leading global firms, led to a success- Advantages?” International Journal of ful transformation of the entire automotive Emerging Markets 8 (4): 329–53. industry in the country. Arnold, J., B. S. Javorcik, and A. Mattoo. 4. The FDI life cycle considers the relationship 2011. “Does Services Liberalization Benefit among foreign and domestic investors, gov- Manufacturing Firms? Evidence from the Czech ernments, and civil society in various stages Republic.” Journal of International Economics of investment. Based on the FDI vision 85 (1): 136–46. and objectives of the host economy, the Balsvik, R. 2006. “Is Mobility of Labour cycle begins with attraction of FDI into a a Channel for Spillovers from Multinationals country. It then moves into enabling inves- to Local Domestic Firms?” Mimeo. Norwegian tors to enter and establish presence in the School of Economics. domestic economy. Once operational, FDI is Barba Navaretti, G., and A. Venables. 2004. encouraged to stay in the long term, lead- Multinational Firms in the World Economy. ing to expansion and “rooting” the FDI into Princeton, NJ: Princeton University Press. the domestic economy through linkages and Barnard, R., M. de Bruyn, N. Kempson, and other spillover effects with the domestic pri- P. McLaren. 2014. “Sector Case Study: vate sector. Mining.” In Making Foreign Direct Investment 5. Other ways that help to de-risk private Work for Sub-Saharan Africa , edited by investment projects include blending grants T. Farole and D. Winkler, 117–62. Washington, or concessional funds with private finance DC: World Bank. to improve risk–return ratios; guarantees to, Bijsterbosch, M., and M. Kolasa. 2009. “FDI for example, large infrastructure projects to and Productivity Convergence in Central cover key noncommercial risks or sharing and Eastern Europe: An Industry-Level risks; partial credit guarantees that enhance Investigation.” ECB Working Paper 992, the terms of commercial debt by extending European Central Bank, Frankfurt. maturity, lowering interest rates, or enabling Cuervo-Cazurra, A. 2008. “The access to financial markets. Multinationalization of Developing Country 6. The World Bank’s World Governance Indicators MNEs: The Case of Multilatinas.” Journal of (WGI) decompose institutional quality into six International Management 14 (2): 138–54. dimensions: Voice and Accountability; Political Cuervo-Cazurra, A., and M. Genc. 2008. Stability and Absence of Violence; Government “Transforming Disadvantages into Advantages: Effectiveness; Regulatory Quality; Rule of Developing-Country MNEs in the Least Law; and Control of Corruption. See http:// Developed Countries.” Journal of International info.worldbank.org/governance/wgi/#home. Business Studies 39 (6): 957–79. Cui, L., and F. Jiang. 2009. “FDI Entry Mode Choice of Chinese Firms: A Strategic Behavior Bibliography Perspective.” Journal of World Business 44 (4): Alfaro, L., A. Chanda, S. Kalemli-Ozcan, and 434–44. S. Sayek. 2006. “How Does Foreign Direct Dollar, D. 2016. “China as a Global Investor.” In Investment Promote Economic Growth? China’s New Sources of Economic Growth: Exploring the Effects of Financial Markets Vol. 1: Reform, Resources and Climate Change, on Linkages.” NBER Working Paper 12522, edited by Ligang Song, Ross Garnaut, Cai National Bureau of Economic Research, Fang and Lauren Johnston. Australia National Cambridge, MA. University Press, Acton. Alfaro, L., and A. Rodriguez-Clare. 2004. Duanmu, J.-L., and Y. Guney. 2009. “A Panel “Multinationals and Linkages: Evidence from Data Analysis of Locational Determinants of Latin America.” Economia 4: 113–70. Chinese and Indian Outward Foreign Direct Alfaro, L., and M. X. Chen. Forthcoming. Investment.” Journal of Asia Business Studies “Selection and Market Reallocation: 3 (2): 1–15. Productivity Gains from Multinational Du, L., A. Harrison, and G. Jefferson. 2011. “Do Production.” American Economic Journal: Institutions Matter for FDI Spillovers? The Economic Policy. Implications of China’s ‘Special Characteristics.’” 16 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 NBER Working Paper 16767, National Bureau Kang, Y., and F. Jiang. 2012. “FDI Location of Economic Research, Cambridge, MA. Choice of Chinese Multinationals in East and Echandi, R., J. Krajcovicova, and C. Z. W. Qiang. Southeast Asia: Traditional Economic Factors 2015. “The Impact of Investment Policy in a and Institutional perspective.” Journal of World Changing Global Economy: A Review of the Business 47: 45–53. Literature.” Policy Research Working Paper Lee, C., C. G. Lee, and M. Yeo. 2016. “Determinants 7437, World Bank, Washington, DC. of Singapore’s Outward FDI.” Journal of Farole, T., and D. Winkler, eds. 2014. Making Southeast Asian Economies 33 (1): 23–40. Foreign Direct Investment Work for Sub-Saharan Lipsey, R. E. 2004. “Home- and Host-Country Africa: Local Spillovers and Competitiveness in Effects of Foreign Direct Investment.” In Global Value Chains. Directions in Development. Challenges to Globalization: Analyzing the Washington, DC: World Bank. Economics , edited by Robert E. Baldwin Gao, T. 2005. “Foreign Direct Investment from and L. Alan Winters, 333–82. University of Developing Asia: Some Distinctive Features.” Chicago Press. Economics Letters 86 (1): 29–35. Luo, Y., Q. Z. Xu, and B. J. Han. 2010. “How Görg, H., and E. Strobl. 2005. “Spillovers from Emerging Market Governments Promote Foreign Firms through Worker Mobility: An Outward FDI: Experience from China.” Empirical Investigation.” Scandinavian Journal Journal of World Business 45 (1): 68–79. of Economics 107 (4): 693–709. Ma, A. C., and A. V. Assche. 2011. “Product Herzer, D. 2012. “How Does Foreign Direct Distance, Institutional Distance and FDI.” Investment Really Affect Developing Countries’ Mimeo. University of San Diego, School of Growth?” Review of International Economics. Business Administration. 20 (2): 396–414. Moran, T. 2014. “Foreign Investment and Supply Holtbrügge, D., and H. Kreppel. 2012. Chains from Emerging Markets: Recurring “Determinants of Outward Foreign Direct Problems and Demonstrated Solutions.” Investment from BRIC Countries: An Working Paper Series 14-12, Peterson Institute Explorative Study.” International Journal of for International Economics, Washington, DC. Emerging Markets 7 (1): 4–30. https://doi. Nepelski, D., and G. De Prato. 2015. org/10.1108/17468801211197897. “International Technology Sourcing between IFC (International Finance Corporation). 2017. a Developing Country and the Rest of the Private Enterprise in Fragile and Conflict World. A Case Study of China.” Technovation Situations. Washington, DC. 35: 12–21. ———. Forthcoming. Multinational Corporations: Newman, C., J. Rand, T. Talbot, and F. Tarp. 2015. Important Partners of IFC to Foster Global “Technology Transfers, Foreign Investment and Economic Integration and Value Addition. Productivity Spillovers.” European Economic Washington, DC. Review 76: 168–87. International Dialogue for Peace-Building and OECD (Organisation for Economic Co-Operation State-Building. 2016. International Standards and Development). 2014. Development for Responsible Business in Conflict-Affected Co-operation Report 2014: Mobilising and Fragile Environment. Paris: OECD. Resources for Sustainable Development. Paris: Javorcik, B. S. 2004. “Does FDI Increase the OECD. Productivity of Domestic Firms? In Search Polachek, S. W., and D. Sevastianova. 2012. of Spillovers through Backward Linkages.” “Does Conflict Disrupt Growth? Evidence of American Economic Review 94 (3): the Relationship between Political Instability 605–27. and National Economic Performance.” The Javorcik, B. S., A. Lo Turco, and D. Maggioni. Journal of International Trade and Economic 2017. “New and Improved: Does FDI Boost Development 21 (3): 361–88. Production Complexity in Host Countries?” Quer, D., E. Claver, and L. Rienda. 2015. “Chinese CEPR Discussion Paper DP11942. Outward Foreign Direct Investment: A Review Javorcik, B. S., and M. Spatareanu. 2009. of Empirical Research.” Frontiers of Business “Tough Love: Do Czech Suppliers Learn from Research in China 9 (3): 326–70. Their Relationships with Multinationals?” Rizvi, S. Z. A., and M. Nishat. 2009. “The Impact Scandinavian Journal of Economics 111 (4): of Foreign Direct Investment on Employment 811–33. Opportunities: Panel Data Analysis: Empirical OVERVIEW 17 Evidence from Pakistan, India and China.” The Outward Foreign Direct Investment by South Pakistan Development Review 48 (4): 841–51. African Enterprises.” Geneva. http://unctad Rodriguez-Arango, L., and M. A. Gonzalez-Perez. .org/en/Docs/c3em26d2a5_en.pdf. 2016. “Giants from Emerging Markets: The World Bank Group. 2011. World Development Internationalization of BRIC Multinationals,” Report 2011: Conflict, Security and Development: in The Challenge of BRIC Multinationals , 2011. Washington, DC: World Bank. edited by Rob Van Tulder, Alain Verbeke, Jorge W E F ( Wo r l d E c o n o m i c F o r u m ) . 2 0 1 3 . Carneiro, and Maria Alejandra Gonzalez-Perez, Manufacturing for Growth: Strategies for 195–226. Progress in International Business Driving Growth and Employment. Geneva: Research, Vol. 11. Emerald Group Publishing WEF. http://www3.weforum.org/docs/WEF Limited. http://www.emeraldinsight.com/doi _ManufacturingForGrowth_ReportVol1_2013 /full/10.1108/S1745-886220160000011011. .pdf. UNCTAD (United Nations Conference on Trade and Development). 2005. “Case Study on 1 What Matters to Investors in Developing Countries: Findings from the Global Investment Competitiveness Survey Peter Kusek and Andrea Silva D eveloping countries compete to investments in developing countries, the survey attract foreign direct investment measures the role in influencing FDI decisions (FDI) because of its potential benefits of such investment climate variables as invest- for the local economy, which include technol- ment incentives, promotion, FDI regulations, ogy transfer, stronger managerial and organi- and administrative processes (see box 1.1 for zational skills, increased access to foreign key findings, annex 1A for survey methodol- markets, and export diversification. FDI can ogy, and annex 1B for profile of respondents). enhance productivity, increase investment in By identifying variables that are most valued research and development, and create better- by investors, this chapter provides practical paying and more stable jobs in host countries. guidance to where policy makers in host coun- But these benefits are not guaranteed, nor do tries can focus their efforts to attract and retain all types of FDI have the same potential FDI, and maximize its gains for development. impact. Thus, host governments must adopt Policy reform initiatives must consider that the right policies to maximize their gains FDI is heterogeneous, driven by different from different types of FDI. motivations and having different economic, The Global Investment Competitiveness environmental, and social impact. MNCs Survey (GIC Survey) offers practical evidence possess different characteristics that influence to help policy makers design policies and pri- their perspectives and decisions. This report is oritize reforms that investors value. Through based on an FDI typology that builds on a interviews with 754 executives of multina- framework proposed by Dunning and tional corporations (MNCs) that have Lundan (2008) (see box 1.2). The framework 19 20 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 BOX 1.1 Top Five Findings of the Global Investment Competitiveness Survey Through interviews with 754 executives of multina- in and out of the country, and existence of legal tional corporations with investments in developing protections against expropriation, against breach of countries, the GIC survey finds the following: contract, and against nontransparent or arbitrary government conduct. 1. Investors involved in export-oriented efficiency- 4. Investors strongly value the existing capacity and seeking FDI that look for internationally cost- skills of local suppliers, but also find that govern- competitive destinations and potential export ment support, such as providing information on the platforms value linkages, incentives, trade agree- availability of local suppliers, matters. With foreign ments, and investment promotion agency (IPA) ser- investors sourcing about 43 percent of their produc- vices more than other investors. Incentives such as tion inputs locally, supplier contracts and linkages tax holidays are important for 64 percent of inves- with local businesses have the potential to create tors involved in efficiency-seeking FDI, compared significant benefits for the local private sector. to only 47 percent of their counterparts involved in 5. For close to 30 percent of investors that have expe- other types of FDI. IPA services are rated important rienced shutting down an affiliate in a developing by about half of investors involved in efficiency- country, some reasons for exiting the investment seeking FDI but by only about a third of those could have been avoided, such as unstable macro- involved in other types of FDI. economic conditions and increased policy and regu- 2. More than a third of investors reinvest all of their latory uncertainty. Three-quarters of investors have profits into the host country. Investors value poli- experienced disruptions in their operations due to cies that help them expand their business more than political risk forces and events. A quarter of inves- just policies used by governments to attract them. tors that did experience disruptions canceled or 3. Investment protection guarantees are critical for withdrew their investment. Severe cases occur fairly retaining and expanding investments in the long infrequently—about 13 percent for breach of con- term across all types of FDI. Over 90 percent of all tract and 5 percent for expropriation—but when investors rate various types of legal protections as they do, the negative impact is strong. In cases of important or critically important, the highest rat- breach of contract, over a third of investors can- ing among all factors included in the survey. These cel or withdraw investments, and for expropriation guarantees include the ability to transfer currency almost half do so. contends that MNCs are lured to a particular This chapter provides a corporate perspec- location with a predominant motivation in tive on the investment decision making of mind: accessing domestic markets, seeking MNCs across the stages of the investment increased efficiencies of production, taking cycle: attraction, entry and establishment, advantage of natural resources, and acquiring operations and expansion, linkages with the strategic assets. This report extends the use of local economy, and in some cases, divestment this typology to explore how various policy and exit. The survey reveals how MNCs instruments influence investors differently decide on FDI and how they identify and depending on their FDI motivation, and how select a country for investment. It also looks the impact of investment on the host economy at MNCs’ operational, reinvestment, and varies by type of FDI. As a result, different expansion experiences, as well as their types of FDI are based not only on investors’ encounters with political risks and their deci- subjective motivation for cross-border invest- sions to shut down foreign affiliates. ment, but also on the inherent objective char- While host-country policy makers listen to acteristics of various investment projects, and investor preferences, they must also consider their implications for developing countries.1 the public interest. Although the survey focuses WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 21 on MNC perspectives and preferences, this In addition to the subjective motivation of report does not necessarily recommend that investors, the FDI typology considers FDI’s governments simply yield to investors’ wishes. objective impact on the host country—for Addressing investor concerns should be bal- example, increase in exports brought about anced with the public interest. For instance, by efficiency-seeking investments. The GIC low tax rates and incentives may be desirable survey focuses on the subjective motivation from the perspective of MNCs, but govern- by asking investors to self-identify their com- ments should not simply lower tax rates and pany’s motivations in a specific investment give more investment incentives, especially if project in a developing country. these limit the country’s gains from FDI. This In this survey, close to 90 percent of inves- chapter offers practical evidence on the relative tors said that accessing new markets or new importance of investment policies to guide pol- customers was one of their motivations icy makers in formulating and prioritizing (figure 1.1). About half of respondents are reforms. motivated by lowering production costs or The following sections discuss the hetero- establishing a new base for exports. The moti- geneity of FDI and how it affects MNCs’ per- vation to coordinate a value chain occurs for ceived importance of the legal and regulatory two-fifths of respondents. For those investors environment relative to other country charac- that want to coordinate their companies’ teristics, and of various investment policy– value chain, 70 percent are also motivated to related factors. The chapter is organized cut production costs. Few respondents iden- according to the life cycle of investments— tify with the motivation to acquire strategic selecting a location, entering a country and assets (15 percent) or access natural resources establishing an investment, running and and raw materials (12 percent). Critically, expanding operations, and considering almost two-thirds of investors selected multi- divestment. ple motivations and when asked about which motivation prevails, most investors (71 percent) say they are market-seeking. Foreign Investors Are Survey respondents represent a range of Heterogeneous with Multiple sectors with a combination of investor moti- vations (figure 1.2). They are in primary Motivations sectors (6 percent), manufacturing (47 percent), Investors with different motivations con- and services (45 percent), and other nonspeci- sider different factors in their decision to fied sectors (2 percent). Although some sec- invest (box 1.2). MNCs that primarily seek tors are naturally linked with specific access to natural resources—as in extractive motivations (for example, the primary sector industries—care more about such variables being natural resource–seeking), motivations as access to land and resources they wish to do not correlate strongly with sectors. While exploit than other variables. Market-seeking about 80 percent of services firms tend to be FDI tends to prioritize the size of and pur- primarily market-seeking, some are also effi- chasing power in the domestic market. ciency-seeking, such as services enabled by Efficiency-seeking2 FDI values policies that information technology (IT). Manufacturing facilitate the import and export of goods firms are also mainly market-seeking but and services, and lower production costs. include a large concentration of efficiency- Efficiency-seeking FDI also includes firms seeking firms and a handful of natural that participate in global value chains resource–seeking ones. (GVCs), an important way for developing Investors involved in efficiency-seeking countries to integrate into the global econ- FDI, relative to investors involved in other omy. MNCs that seek strategic assets pri- types of FDI, are more sensitive to various marily pursue technologies and brands that host market characteristics, including invest- can enhance their operations. ment climate factors. These host market 22 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 BOX 1.2 Investor Motivation Framework According to Dunning and Lundan A well-known framework proposed by Dunning and technology transfer, and integration of a country into Lundan (2008) differentiates four sources of foreign global value chains. The levels of benefits vary, and direct investment (FDI) motivation: natural resources in some carry more risks than others. the host country, access to the host country market, stra- From an investment policy and promotion per- tegic assets of firms in the host market, or cost savings spective, it is important to note that the four types through higher production efficiency (figure B1.2.1). of investment can respond differently to policy mea- The last type of investment is typically associated with sures and the overall investment climate. Effi ciency- offshoring production stages to the host country, and is seeking investors—whose investment decisions are thus export-oriented. driven largely by the motive to save costs—tend to All four types of investment can have important, be highly sensitive to any variables that raise their though varying, benefits for the host economy. For cost of operation or hinder their free exchange of example, natural resource–seeking investment often goods and services with the rest of the world as part generates sizable government revenues. Market-seek- of global production networks. Natural resource–, ing FDI can be associated with availability of better strategic asset–, and market-seeking investments tend and cheaper goods and services consumed by the to be less sensitive to investment climate variables if population or used as inputs by other fi rms. Strate- either the resource to be exploited or the fi rm that gic asset–seeking investment allows domestic fi rms possesses competitive advantages can be found in the to expand their global networks. Effi ciency-seeking country or if the domestic market offers attractive investment is often seen as a means of job creation, opportunities. FIGURE B1.2.1 Investor Motivation Framework According to Dunning and Lundan Natural resource–seeking FDI enters the country It leads to exporting of natural resources or to exploit locally available natural resuorces resource-based products Market-seeking FDI enters the country to gain It leads to domestic sales of final products access to the domestic markets to consumers or intermediates to firms Strategic asset–seeking FDI enters the country to enhance the capabilities of the investing firm by It leads to sales of final goods in the home acquiring a firm with technology and brands that country and third countries have competitive advantage Efficiency-seeking FDI enters the country to save costs in international production networks (offshoring) ... and to exporting of final products or intermediates It leads to importing of intermediate products... Source: Based on Dunning and Lundan 2008. WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 23 FIGURE 1.1 Most Investors Have Multiple Motivations and Are Market-Seeking Share of respondents (percent) a. Motivation b. Most important motivation Market- seeking, Access new markets or new customers 87 71 Lower production costs or 51 establish a new base for exports Coordinate company’s value chain, None of 39 the above, such as being closer to suppliers 5 Acquire another firm that will provide the Strategic 15 company new technologies or brands asset-seeking, 1 Efficiency- Natural Access natural resources and raw materials, seeking, 12 resource-seeking, Coordinate such as oil, gas, or agricultural products 13 4 GVC, 6 Source: Computation based on the GIC Survey. Note: The numbers on the left do not add up to 100 percent because respondents are permitted to select multiple motivations: 62 percent of respondents selected two or more motivations. Many respondents may have understood the motivation to access new markets or new customers to apply not only to the domestic market in which they were investing, but also to the regional market. In fact, this motivation was commonly selected for investments in many small developing countries with an extensive network of trade and investment agreements with other economies, suggesting that the respondents were interested in accessing new regional markets or regional consumers, rather than just the small domestic market of the host country. characteristics include macroeconomic sta- the results of this survey are based on bility and favorable exchange rate, labor investors’ responses mostly for middle- pool, physical infrastructure, tax rates, access income developing countries, although they to land, and domestic financing sources. are likely relevant to low-income countries Among investment climate variables, MNCs as well. involved in efficiency-seeking FDI assign a higher importance to investment protection Investment Exploration and guarantees, ease of entry, local suppliers, incentives, trade agreements, and bilateral Location Decision: First Phase investment treaties, compared with other in the Investment Life Cycle investors. This suggests that firms involved What Variables Determine MNC in efficiency- seeking FDI may be more Investment Decisions? responsive to policies and reforms aimed at improving the business environment. This Investors consider a broad range of factors chapter thus explores the differences between in deciding to invest, the most important MNCs involved in efficiency-seeking FDI being political stability and security, as well and those that are involved in other types of as a business-friendly legal and regulatory FDI (box 1.3). environment. These top other variables such Host countries are also heterogeneous. as infrastructure, labor talent and skill, and A vast majority of survey respondents have low costs of labor and inputs. Among survey operations in upper-middle-income countries respondents, 86 percent find the legal and (87 percent), about a third in lower-middle- regulatory environment important or criti- income countries, and very few have foreign cally important, suggesting that it weighs affiliates in low-income countries (8 percent). heavily in investors’ decision to invest Thus, policy implications emanating from (figure 1.3). 24 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 1.2 Respondents Represent Firms across Various Sectors Share of respondents per sector (percent) Primary Manufacturing Services Wholesale and Logistics, 6.36 46.83 44.56 retail trade transport, Metals and 5.70 and storage Other metal 4.64 manufacturing products Mining, 4.77 Automobiles, 5.17 quarrying, and other petroleum motor 3.45 vehicles, and transport equipment Alternative Business Chemicals Textiles, energy services 8.89 Pharmaceuticals, and apparel, Construction 2.52 2.39 Agriculture, biotechnology, chemical and Professional, 7.03 hunting, and medical products leather scientific, and devices technical forestry, 3.18 3.05 3.45 Other and 4.24 Computer travel and fishing and Other tourism- 2.92 Other software related machinery Refined Paper and services services petroleum services and paper Financial 1.99 1.33 1.06 products, coke, electronic products and nuclear 0.80 services equipment fuel 0.93 Media including Health and Agroprocessing, Plastic Information Printing Non- insurance services and technology metal entertain- components food products, products and tele- and publish- mineral 5.84 Electricity, 1.06 Other 1.86 communications prod- ment 8.49 and beverages ing gas, and Telecom- 2.25 equipment 0.53 ucts 0.93 3.18 0.80 0.40 water munications Wood and 2.65 Hotels and wood 1.72 Real Rubber products Furni- restaurants (other than ture estate 0.66 0.27 0.93 furniture) 0.53 0.40 Primary Manufacturing Services Other Source: Computation based on the GIC Survey. Note: Respondents were asked to identify the main sector of their company globally, which may or may not reflect the sector of the affiliates in developing countries. About 10 percent of respondents noted that the sector in the foreign affiliate they are most familiar with is different from the main sector of the global company. See table 1B.4 for the complete list of sectors, distributional shares across respondents, and comparison with global FDI flows. WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 25 FIGURE 1.3 Business-Friendly Legal and Regulatory Environment Is Important for Investors Share of respondents (percent) Importance of country characteristics Political stability and security 50 37 9 2 Legal and regulatory environment 40 46 12 2 Large domestic market size 42 38 14 4 Macroeconomic stability and favorable 34 44 16 5 exchange rate Available talent and skill of labor 28 45 22 5 Good physical infrastructure 25 46 24 5 Low tax rates 19 39 31 9 Low cost of labor and inputs 18 35 35 11 Access to land or real estate 14 31 32 22 Financing in the domestic market 16 28 31 24 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: Respondents were asked, “How important are the following characteristics to your company’s decision to invest in developing countries?” Factors were asked in random order. They are listed in the graph in descending order of importance, based on the combination of “critically important” and “important” in dark green and light green bars. Critically important means it is a deal-breaker; by itself this factor could change a company’s decision to invest or not in a country. FIGURE 1.4 MNCs Involved in Efficiency-Seeking FDI Are More Selective Share of respondents (percent) Importance of country characteristics Efficiency-skg 52 35 11 2 Political stability and security Non-eff-skg 49 39 8 3 Efficiency-skg 39 46 13 1 Legal and regulatory environment 41 45 11 2 Non-eff-skg Efficiency-skg 43 36 16 5 Large domestic market size 41 41 12 4 *** Non-eff-skg Macroeconomic stability and Efficiency-skg 39 43 15 3 30 44 18 7 *** favorable exchange rate Non-eff-skg Efficiency-skg 31 47 18 3 Available talent and skill of labor Non-eff-skg *** 24 44 25 6 Efficiency-skg 30 46 19 4 Good physical infrastructure *** Non-eff-skg 19 45 28 6 Efficiency-skg 22 42 29 6 Low tax rates Non-eff-skg *** 16 36 34 13 Efficiency-skg 26 40 27 6 Low cost of labor and inputs *** Non-eff-skg 9 30 42 17 Access to land or real estate Efficiency-skg 16 36 32 16 *** Non-eff-skg 12 25 32 28 Efficiency-skg 17 32 30 21 Financing in the domestic market *** Non-eff-skg 15 23 31 28 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: Country characteristics that have statistically significant differences between investors involved in efficiency-seeking FDI and investors involved in other types of FDI are marked on the right side of the graph. The differences between the two groups are significant at ***p<0.01, **p<0.05 and *p<0.1. 26 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 BOX 1.3 MNCs Involved in Efficiency-Seeking Investments Tend to Be More Selective Investors’ preferences and behavior differ depend- also received incentives more often in a typical ing on their motivation for investing in developing investment. countries. In this survey, about half of respondents 4. In terms of ease of entry, MNCs involved in efficiency- said that at least one of their motivations is to lower seeking FDI view efficiency of obtaining approvals, production costs or establish a new base for exports. owning all equity, easily bringing in expatriate staff, Relative to investors with other motivations, these and importing production inputs as more important efficiency-seeking firms differ in the following ways: compared with investors involved in other types of FDI. For firms with an efficiency-seeking motiva- 1. MNCs involved in efficiency-seeking investments tion, the ability to import production inputs is rated view most characteristics of host countries as more slightly more important (73 percent) than the abil- important than investors involved in other types ity to bring in expatriate staff (71 percent) while of FDI. These characteristics include stable macro- the reverse is true for firms with other motivations economic conditions and favorable exchange rate, (61 and 65 percent, respectively). available talent and skill of labor, good physical 5. Capacity and skills of local suppliers are impor- infrastructure, low tax rates, low cost of labor and tant or critically important for 77 percent of inputs, access to land or real estate, and available MNCs involved in effi ciency-seeking FDI, com- financing in the domestic market. Among these, pared with 70 percent of investors with other the difference is largest for low cost of labor and motivations. Government initiatives including inputs, which 66 percent of firms involved in effi- information about availability of local suppli- ciency-seeking investment find important or criti- ers, upgrading potential suppliers, and incentives cally important compared with only 39 percent of to invest in supplier upgrading are rated more investors with other motivations. important by about 8 to 12 percentage points 2. Investors involved in efficiency-seeking FDI also more by firms involved in efficiency-seeking FDI rate most investment policy factors as more impor- relative to firms involved in other types of FDI. To tant than investors involved in other types of FDI. promote linkages, 55 percent of MNCs involved These include investment protection guarantees, in efficiency-seeking FDI have internal “talent ease of obtaining approvals, investment incen- scouts” to find local suppliers, compared with tives, preferential trade agreements, and bilateral only 45 percent of investors involved in other investment treaties. The difference is notable for types of FDI. preferential trade agreements, which 65 percent of 6. MNCs involved in efficiency-seeking FDI value firms involved in efficiency-seeking investment find the services of investment promotion agencies important or critically important compared with only (IPAs) more highly, with 52 percent of respondents 45 percent of investors with other motivations. identifying IPA services as important or critically 3. Incentives also matter more for firms with efficiency- important, compared with 37 percent of investors seeking investments. In this group, 63 percent find involved in other types of FDI. Specifically, meet- incentives important or critically important, in con- ings with agency officers to discuss investment trast with only 43 percent of investors with other opportunities, information and assistance in setting motivations. Firms with efficiency-seeking invest- up an affiliate, and assistance in problem resolution ments rated eight different incentive instruments are valued more by firms with efficiency-seeking more highly than other investors with a difference investments, by about 9 to 12 percentage points, of about 13 percentage points on average. They than by other investors. Firms involved in efficiency-seeking FDI are infrastructure, low tax rates, low costs of labor more sensitive to a broad range of factors. and input, access to land, and domestic financ- MNCs seeking cost-competitive locations for ing more than other investors. Because these their mostly export-oriented production value investors are more sensitive to costs, they more macroeconomic stability, labor skills, reliable carefully consider factors that directly affect WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 27 their cost structure and productivity. important, or potential deal-breakers. The size of the domestic market is valued Transparency and predictability may be inter- slightly more by investors without an effi- preted as a reflection of the overall interaction ciency-seeking motivation, which are predomi- between MNCs and host governments— nantly motivated by accessing new markets. comprising both regulations themselves and The two most important factors—political sta- their implementation. bility and security, and the legal and regulatory Investors value policies that help them environment—are consistently valued highly expand their business more than policies to across all types of investors (figure 1.4). (See attract them. Forty-five percent of respon- figures 1C.1, 1C.2, and 1C.3 for differences in dents rate investment protection guarantees as importance rating by manufacturing versus critically important or deal-breakers, highest services firms, developed versus developing among all investment climate factors. Over 90 source countries, and parent company versus percent of investors rate various types of legal affiliate.) protections as critical, including the ability to Investors seek both strong legal protections transfer currency in and out of the country as and predictability and efficiency in implement- well as legal protections against expropria- ing laws and regulations (figure 1.5). Many tion, against breach of contract, and against developing countries have inefficient bureau- nontransparent or arbitrary government con- cracies, opaque regulations, complex proce- duct. All investors—regardless of sector, dures, and high transaction costs that source country, or FDI motivation—find these undermine their competitiveness. Not surpris- guarantees of greatest value. These policies ingly, four out of five surveyed investors rate are bigger deal-breakers than investment transparency and predictability in the conduct incentives, preferential trade agreements, and of public agencies, investment protection guar- bilateral investment treaties. These results sug- antees provided in the country’s laws, and the gest that host countries need to pay as much ease of starting a business as important in their attention to investor aftercare as they do to decision on where to invest. Moreover, about a attracting investors to their country. Given third of investors rate these as critically that respondents are investors that already FIGURE 1.5 Investors Seek Predictable, Transparent, and Efficient Conduct of Public Agencies Share of respondents (percent) Importance of investment climate factors Transparency and predictability in 37 45 15 3 the conduct of public agencies Investment protection guarantees 45 36 14 4 provided in the country’s laws Ease of obtaining government approvals to start 36 41 18 5 a business and to own all equity in the company Investment incentives such as tax holidays 21 35 32 10 Having a preferential trade agreement 14 40 32 11 Having a bilateral investment treaty 15 36 33 13 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. 28 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 have ongoing operations in developing coun- How Critical Are Incentives in tries and not prospective investors, this partly Attracting FDI? explains the emphasis on aftercare. Investment incentives to attract FDI are MNCs involved in efficiency-seeking FDI widespread and used by governments in place more importance on investment climate both high-income and developing countries. factors compared to firms involved in other Developing country policy makers often types of FDI. Except for transparency and pre- view incentives as necessary for their coun- dictability in the conduct of public agencies, tries to compete for FDI. As discussed later which firms find most important regardless in this report, incentives impose sizable costs of motivation, firms involved in efficiency- on host countries through fiscal losses from seeking FDI value most investment policies non-collection of taxes, rent-seeking by more highly (figure 1.6). This suggests that firms, and associated tax evasion. Countries MNCs involved in efficiency-seeking FDI may must thus walk a fine line between remaining be more sensitive to these factors when decid- competitive by offering incentives and ensur- ing to invest. Such results are not surprising, ing that benefits outweigh their costs. given that most efficiency-seeking investment Investment incentives rank only fourth in is export oriented and highly selective in where importance to investors out of six investment it locates, hence the importance of trade agree- climate characteristics listed in the GIC survey. ments and investment incentives. As such, They rank lower than transparent government policy makers in host countries should target conduct, investment protection guarantees, their initiatives to attract these investors. (See and ease of establishing a business (figure 1.5). figures 1C.4, 1C.5, and 1C.6 for differences Overall only one in five investors finds the in importance rating by manufacturing versus absence of investment incentives as deal- services firms, developed versus developing breakers in deciding to invest. Another third source economies, and parent company versus of respondents find incentives to be important affiliate.) FIGURE 1.6 MNCs Involved in Efficiency-Seeking FDI Value Incentives, Trade Agreements, and Ease of Entry More than Other Investors Share of respondents (percent) Importance of investment climate factors Transparency and predictability in the conduct of Efficiency-skg 37 45 16 2 public agencies Non-eff-skg 37 44 14 4 Investment protection guarantees provided in Efficiency-skg 47 37 13 3 the country’s laws Non-eff-skg 42 36 16 5 * Ease of obtaining government approvals to start a Efficiency-skg 41 38 16 4 business and to own all equity in the company Non-eff-skg 30 44 19 5 *** Efficiency-skg 23 41 28 8 Investment incentives such as tax holidays *** Non-eff-skg 18 29 37 13 Efficiency-skg 17 46 26 8 Having a preferential trade agreement *** Non-eff-skg 12 33 37 14 Efficiency-skg 17 38 31 11 Having a bilateral investment treaty ** Non-eff-skg 12 34 34 15 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: Most investment climate factors in this graph have statistically significant differences between investors involved in efficiency-seeking FDI and investors involved in other types of FDI. The differences between the two groups are significant at ***p<0.01, **p<0.05 and *p<0.1. WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 29 but not deal-breakers. This does not necessar- Duty-free imports, tax holidays, and VAT ily suggest that incentives can be completely exemptions are the top three most important eliminated but that, by themselves, they are incentives for investors (figure 1.7). About unlikely to convince investors to shift the loca- two-thirds of investors who said that incen- tion of their investment. The policy fundamen- tives are at least somewhat important find tals of the investment climate must be these three instruments to be important or addressed before policy makers resort to critically important. MNCs involved in effi- incentives as a means of attracting investors. ciency-seeking FDI rated all types of incen- MNCs involved in efficiency-seeking FDI, tives more highly compared with investors however, value incentives more than investors involved in other types of FDI, with a differ- with other motivations. Among investors ence of about 13 percentage points on aver- motivated by cutting production costs and age. They also received incentives more often finding new export platforms, 64 percent find in a typical investment. When asked about incentives important or critically important, the specific incentives that their companies in contrast with only 47 percent of investors have received, respondents identified the same with other motivations (figure 1.6). Investors three types of instruments—duty-free imports, involved in efficiency-seeking FDI are tax holidays, and VAT exemption—as most also granted certain incentives—duty-free frequently received. This suggests that the imports, subsidized loans, and value respondents’ high rating of these types may added tax (VAT) exemption—more often than owe to their familiarity with the specific other investors. This suggests that they may instruments. be more responsive to incentives than inves- Obtaining fiscal and financial incentives tors with other motivations such as accessing typically takes three months but varies from new markets and natural resources. about a week to over a year, depending on the FIGURE 1.7 Duty-Free Imports, Tax Holidays, and VAT Exemptions Are the Most Attractive Investment Incentives Share of respondents (percent) Importance of incentives Duty-free imports 34 40 16 9 Tax holidays 25 45 22 7 VAT exemption 29 36 26 9 Technical or business support incentives 16 42 31 9 Direct subsidies 16 37 32 14 Subsidized loans 14 35 31 19 Accelerated depreciation 11 36 34 15 Subsidized land 12 30 37 20 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The question on incentives was answered by 663 respondents. These respondents answered somewhat important, important, or critically important on incentives in the question in figure 1.5. VAT = value added tax. 30 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 country and type of incentive. About one FDI, the ability to import production inputs quarter of surveyed investors said obtaining is rated slightly more important (73 per- incentives took less than one month, while cent) than the ability to bring in expatriate about 6 percent noted it took more than staff (71 percent) while the reverse is true for a year. firms involved in other types of FDI (61 and 65 percent respectively). Although efficiency in obtaining per- Investment Entry and mits is most important overall, restrictions on foreign equity ownership appear to be Establishment: Second Phase the biggest deal-breaker. Forty percent of in the Investment Life Cycle respondents claim that owning all equity in their affiliate and not being required How Do Policies and Administrative to share ownership with local firms or the Procedures for Business Establishment government is critically important, highest Affect FDI Decisions? among all policy factors considered. This Investors strongly value business-friendly result is significant in the context of foreign policies and efficient procedures related ownership restrictions still being relatively to business establishment. About four out prevalent across developing countries, of five respondents say that the ease of especially in services. obtaining approvals for their investment Obtaining investment approvals and per- is important or critically important, while mits to start a business typically takes three only 2 percent say it is not at all important months, but varies by country and type of (figure 1.8). In fact, the speed of obtaining investment (figure 1.9). The variation is quite approvals and permits ranks even higher wide: on one end of the spectrum, about than investors’ ability to own all equity in 10 percent of respondents say they waited a project, to easily bring in expatriate staff, less than a month while on the other end, and to import production inputs. For MNCs another 10 percent of investors waited a year involved in efficiency-seeking FDI, all these or longer. Respondents who value efficiency characteristics are rated as more important of government approvals encountered some- relative to investors involved in other types of what shorter waits. For this group, only FDI. For firms involved in efficiency-seeking 12 percent had processing times exceeding FIGURE 1.8 Investors Strongly Value Business-Friendly Policies and Procedural Efficiency of Entry and Establishment of Affiliates Share of respondents (percent) Importance of ease of entry factors Quickly obtain investment approvals and permits 34 48 16 2 to start a business Own all equity in the foreign affiliate 40 37 16 6 Easily bring in expatriate staff and get visas and 30 38 22 9 work permits Import production inputs instead of being 30 36 23 9 required to purchase them from local suppliers Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The questions on ease of entry were answered by 709 respondents. These respondents answered somewhat important, important, or critically important on ease of entry in the question in figure 1.5. WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 31 six months compared with 25 percent other- FIGURE 1.9 Wait Times for Investment Approvals Vary but wise. This confirms that investors who value Typically Take Three Months efficiency tend to favor destinations where Obtain investment approval approvals are quicker to obtain. and permits to start a business 3 The median length of time for obtaining a land lease is two months, and for obtaining Obtain a land lease 2 work permits is about 1.5 months. The dis- Obtain work permits for persion of responses for both of these formal- 1.5 expatriate staff ities also appears tighter than for obtaining 0 5 10 15 20 initial investment approvals. Fewer respon- Months dents also experience wait times longer than six months—9 percent of respondents when Source: Computation based on the GIC Survey. Note: The boxplot shows the median point (with data label) as the middle bar. The ends of the obtaining a land lease and only 6 percent boxes represent the 25th and 75th percentiles. The ends of the black lines show the 5th and when obtaining work permits. 95th percentiles. Investment Operations and production inputs they need. On average, 43 percent of material inputs, supplies, and Growth: Third Phase in the s e r v i c e s a r e s o u r c e d l o c a l l y, v e r s u s Investment Life Cycle 34 percent of inputs sourced from another unit of the company and 23 percent of What Role Do Local Suppliers Play in inputs imported (figure 1.10). The percent- MNCs’ Operations? age of inputs sourced locally varies widely: FDI brings potential benefits to the host about 13 percent of surveyed companies do country through a variety of channels not source any inputs locally, another including linkages with the local private 13 percent source all their inputs locally, sector. Linkages between foreign firms and and the rest of the firms (about 74 percent) local suppliers enable knowledge and tech- source some portion of their inputs locally. nology transfer, including know-how and Linkages are more prevalent for MNCs in practices that allow domestic suppliers to service sectors compared with manufactur- upgrade the quality and efficiency of their ing firms. production. Linkages also expand the mul- Overall, 61 percent of MNCs consider tiplier effect in the local economy. When linkages as important or critically impor- foreign investors source inputs locally tant in their location decisions. Among instead of importing them, they boost pro- those investors who identified linkages as duction of local firms and create jobs in at least somewhat important, 74 percent the local economy. As such, policy makers find that capacity and skills of local sup- try to promote linkages through various pliers are important or critically important policies and programs. One such policy is (figure 1.11). Local skills and capacity are local content requirements, where a cer- valued even more by MNCs involved in tain percentage or absolute amount of efficiency-seeking FDI (77 percent). This local input is required of foreign firms. suggests that government initiatives to pro- Research finds, however, that local con- mote linkages will only be effective if local tent requirements and similar measures companies can offer the capacity and skills have a largely negative effect and discour- expected by MNCs. At the same time, gov- age FDI.3 ernments of host countries have the scope While investors resist being mandated to to facilitate linkages. Investors value infor- source their inputs locally, many of them mation on the availability of local suppliers, prefer to do so if they are able to find in the rated as important or critically important by local market the quality and quantity of the 68 percent of respondents. About 61 percent 32 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 of respondents also rate supplier upgrad- matchmaking events with suppliers. These ing as important, whether in the form of government initiatives are rated as important direct financial incentives for companies to by about 8 to 12 percentage points more by invest in supplier development or govern- firms involved in efficiency-seeking FDI rela- ments’ own initiatives to upgrade suppliers. tive to other investors. Only 42 percent of respondents value When capacity and quality constraints in the local market prevent investors from find- ing appropriate suppliers, investors value FIGURE 1.10 Nearly Half of Material Inputs, being able to import inputs instead of being Supplies, and Services Are Sourced Locally required to source them locally. This is espe- Share of respondents (percent) cially true for MNCs involved in efficiency- Imported seeking FDI and manufacturing firms. Many 23 manufacturing MNCs invest in developing Sourced locally countries to reduce their cost of production. 43 At the same time, to maintain a high quality of final products, which are often intended for export, foreign manufacturers appreciate the flexibility of importing their own inputs for production rather than sourcing them locally. Of the surveyed manufacturing firms, 68 per- cent rate the ability to import inputs as impor- tant or critically important, as opposed to only 56 percent of services companies. Among Sourced within firms involved in efficiency-seeking FDI, 73 company percent find this attribute important or criti- 34 cally important while only 61 percent of firms Source: Computation based on the GIC Survey. involved in other types of FDI consider it Note: The number of respondents for each source vary and are fewer than 754 because some respondents answered “don’t know.” important. FIGURE 1.11 Capacity and Skills of Suppliers Are Critical Linkages-Related Features Share of respondents (percent) Importance of factors related to linkages Capacity and skills of local suppliers 24 50 23 3 Information about the availability 21 47 28 4 of local suppliers Proactive government role in 19 42 32 6 upgrading potential suppliers Incentives from government to 20 40 31 9 invest in supplier upgrading Government-organized matchmaking 10 32 42 15 events with potential suppliers Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The questions on linkages were answered by 679 respondents who answered somewhat important, important, or critically important on the question, “How important are the capabilities of local firms to act as suppliers in your decision to invest in developing countries?” WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 33 Foreign investors themselves also have an FIGURE 1.12 Corporate Programs to Promote Linkages Are Not interest in promoting linkages, but company- Very Widespread Share of respondents (percent) initiated programs are uncommon. Sourcing inputs, supplies, and services locally instead Prevalence of corporate programs to promote linkages of importing them can reduce costs for foreign-owned firms. Some MNCs have their Internal “talent scouts” to 51 46 own programs to promote linkages, but these seek out local suppliers are not widespread. The survey finds that, among the foreign firms that do source Vocational or training programs 31 67 locally, half use internal “talent scouts” to to upgrade local suppliers find local suppliers. Firms involved in efficiency-seeking FDI tend to have talent Equipment-financing programs scouts more often (55 percent) than investors 11 87 for local suppliers involved in other types of FDI (45 percent). Over 30 percent have vocational or training Yes No Don't know programs to upgrade local suppliers, and 11 Source: Computation based on the GIC Survey. percent have equipment-financing programs Note: These questions on corporate programs to promote linkages were answered by 454 respondents. These respondents answered somewhat important, important, or critically important on the question for local suppliers (figure 1.12). Among firms “How important are the capabilities of local firms to act as suppliers in your decision to invest in that have vocational or training programs, developing countries?” and source some or all of their inputs locally. about a third sponsor certification programs and partner with local technical colleges and universities. and expanding existing investments in addi- tion to attracting new ones. How Much Do MNCs Reinvest in Host Countries? How Do Investors Respond to Host countries not only need to attract and Political Risks? retain FDI but also need to facilitate its growth to motivate investors to reinvest Among survey respondents, 76 percent expe- their earnings in the host country. Many rienced political risks in their investment variables may influence investors in decid- projects. Political risk is the probability of ing on the share of their profits to repatriate disruption of business operations by political as dividends versus reinvest in growing their forces or events, and especially by govern- operations in the host country. These vari- ment actions. About half of respondents ables include taxation systems, transfer experienced lack of transparency and pre- costs, investment opportunities in the ongo- dictability in dealing with developing coun- ing business and elsewhere, relative costs of try public agencies. Almost half encountered shifting financial resources out of the host adverse regulatory changes and delays in country, and need to expand the ongoing obtaining necessary government permits and business. Reinvested earnings are becoming approvals to start or operate a business. an increasingly important source of FDI, Over 40 percent encountered restrictions in growing from less than 30 percent of FDI transferring and converting currency. In flows in 2007 to about 50 percent in 2015 these cases, about one in four investors can- (UNCTAD 2016). This trend is confirmed celed a planned investment or withdrew an by the survey results, where over a third of existing investment owing to political risks respondents say that they reinvest all their (figure 1.14). profits in the host country, and another More severe cases of political risk occur 14 percent reinvests more than half less frequently but with far worse impact. (figure 1.13). This trend highlights the Only 13 percent of respondents experienced importance for host economies of retaining breach of contract by the government but 34 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 1.13 More than a Third of Investors Reinvest All Their Affiliate-Generated Profits Back into the Affiliate Reinvested earnings 40 34.7 35 Share of respondents (percent) 30 25 21.9 20 15.6 13.7 15 10 7.7 6.4 5 0 0 1–25 26–50 51–75 76–99 100 Share of profits reinvested (percent) Source: Computation based on the GIC Survey. Note: The question on reinvested earnings was answered by 597 respondents. The remaining 158 either refused, did not know the answer, or made the investment within the year. Respondents were asked about reinvested earnings in a specific developing country of their choice. the impact was much greater—35 percent sectors report more frequent experiences of of those investors canceled a planned breach of contract by the government and investment or withdrew an existing one. lack of transparency and predictability in Expropriation was even more extreme: while dealing with public agencies. only 5 percent of respondents experienced Governments should more adequately it, almost half of them canceled or withdrew manage investor grievances. According to an investment. the survey, governments often do not effec- Investments in services tend to be more tively address grievances related to political affected by political risk than manufac- risks. Only about one in five affected inves- turing. Firms in the services sector experi- tors felt that their grievances were promptly enced more disruptions related to political resolved by the government, that the process risk, particularly restrictions in transfer- of complaint was clear and efficient, or that ring and converting currency, breach of the government introduced a systematic contract by the government, and expro- solution to address or prevent such griev- priation. Services—such as energy, telecom- ances in the future. munications, or finance—are more tightly regulated than manufacturing, and thus more exposed to potential political inter- ference. In particular, according to survey Divestment: Fourth Phase in the results, companies in the utilities sector— Investment Life Cycle including electricity, gas, alternative energy, Why Do MNCs Divest from Developing and telecommunications—experience more Countries? frequent adverse regulatory changes and expropriation and more delays in obtaining Some 29 percent of investors surveyed had permits. Construction and business services shut down at least one of their company’s WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 35 FIGURE 1.14 Severe Political Risks Are Infrequent but Can Have Highly Negative Effects on FDI Share of respondents (percent) Lack of transparency and predictability 23 24 27 14 11 in dealing with public agencies (50%) Sudden change in the laws and regulations 27 25 25 11 11 with a negative impact on the company (49%) Delays in obtaining necessary government permits 20 17 37 13 12 and approvals to start or operate a business (47%) Restrictions in the ability to transfer 26 20 29 11 11 and convert currency (42%) Breach of contract by the government (13%) 14 23 26 20 15 Expropriation or taking of property 33 5 10 13 36 or assets by the government (5%) Don’t know None Consider delay or Significantly delay Cancel planned Withdraw existing cancellation investment investment investment Source: Computation based on the GIC Survey. Note: The height of the bars reflects the percentage of respondents that experienced disruption in any of their investments owing to the political risk identified. The risks are arranged in descending order from most frequently experienced at the top, to least frequently experienced at the bottom. The numbers across rows do not add up to 100 percent because respondents could select multiple types of disruptions that their companies had experienced. The horizontal bars show the responses of companies, with the darker red bars reflecting more severe reactions. The bars reveal the most severe reactions of companies after experiencing the particular disruption. If, for example, a company experienced withdrawing an existing investment in one country, but only delaying in another, the most severe reaction was considered and the company was included in the withdraw bar. affiliates in a developing country (figure more highly regulated and thus vulnerable 1.15). The most common reasons were to political interference. Among the surveyed changes in the company’s strategy and services companies, 35 percent had shut unstable macroeconomic conditions, includ- down an affiliate, versus just 23 percent of ing an unfavorable exchange rate. Increased manufacturing firms. policy or regulatory uncertainty was the Although some reasons for exiting invest- third most common reason, which occurred ments are beyond the control of governments in about a third of the divestment cases of host countries, many are avoidable. While (figure 1.16). Arbitrary government conduct, governments cannot do much about changes sudden restrictions on currency transfer, in investor firms’ corporate strategies or and breach of contract by governments are about global economic conditions, they reported as factors by more than 20 percent can influence factors in their own countries. of investors. These results confirm that com- In particular, maintaining an appropriately panies value transparency and predictability valued exchange rate, managing macroeco- in the conduct of public agencies, as well as nomic stability, and ensuring transparent, investment protections. Foreign investors in consistent, and predictable policies and regu- services divest more frequently than manu- lations are critical in keeping investors from facturing MNCs, possibly because they are exiting. 36 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 1.15 More than a Quarter of What Role Do Investment Respondents Had Shut Down an Affiliate in a Developing Country Promotion Agencies Play across Share of respondents (percent) the Investment Life Cycle? Shut down an affiliate in any developing country Although MNCs have their own strategic motivations for selecting specific investment No locations, the quality of services provided by 65 Don’t know the host economies can play a key role in 6 MNCs’ corporate decisions. The role of investment promotion agencies (IPAs) in facilitating investments can be particularly important in countries with larger physical or cultural distance from the home economies of investors. IPAs complement rather than substitute for a good investment climate and ecosystem for Yes investment projects. Only 43 percent of the 29 surveyed investors say that IPAs are impor- Source: Computation based on the GIC Survey. tant or critically important in their decision to invest, the lowest among investment climate variables queried in the survey. Only 12 per- cent consider quality of IPA services to be FIGURE 1.16 Reasons for Exiting an Investment Are Mixed, deal-breakers, while 14 percent rate IPAs as Some Controllable and Others Not not at all important. These results suggest Share of respondents (percent) that other factors play a more prominent role in firms’ decision making. Sound economic Reasons for exiting an investment fundamentals need to prevail before the Change in company strategy 45 services delivered by IPAs become critical Unstable macroeconomic conditions for investors. 41 and unfavorable exchange rate IPA services thus have great scope for Increased policy or 32 improvement. The relatively low rating of regulatory uncertainty the importance of IPAs does not necessarily Global economic downturn 29 suggest that host countries should not strengthen them. The reverse could actually Arbitrary government conduct 23 be true—that host countries currently offer Breach of contract by the poor-quality IPA services for investors, 17 government which is why investors’ perceptions are not Sudden restrictions on transferring very positive. Only 11 percent of respon- 16 and converting currency dents use IPA services in their typical invest- Cost increase of labor and materials 10 ment, despite 43 percent saying they are important. The proportion of users is some- Expropriation 8 what greater for investments in low-income Sudden restrictions on countries than in middle-income countries, 5 hiring expatriate staff suggesting that IPAs provide more value Withdrawal of tax incentives 5 when the business environment is more dif- ficult and information harder to obtain, as is Source: Computation based on the GIC Survey. often the case in low-income countries. Note: Results are based on 219 respondents that were aware that their companies had shut down an affiliate in a developing country. Shares do not add up to 100 because respondents could select up to MNCs involved in efficiency-seeking FDI five of the most relevant reasons. value IPA services more highly, with 52 WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 37 percent of respondents identifying IPA ser- frequently. Investors used IPA services for vices as important or critically important, assistance in registering and obtaining per- compared with 37 percent of investors mits for a new investment (76 percent), involved in other types of FDI. expanding investment (59 percent), explor- Among investors who do find IPAs to be ing locations for a new investment (46 per- important or somewhat important, two- cent), helping address operational issues or thirds highly value help in handling issues problems (41 percent), and finding domes- and resolving grievances with government, tic suppliers (28 percent). These results information and assistance in setting up, likely reflect the availability of services and business advocacy efforts to improve offered by IPAs in the first place rather the business environment. These services are than investors’ needs. IPAs often dedicate rated more important than investment pro- resources for investment promotion and motion activities (figure 1.17). Promotion facilitation, but not many offer additional efforts to attract investors—advertising services after the investment becomes online and in media, and exhibitions at operational. A potential mismatch is trade shows, investment conferences, and apparent—while investors would appreci- events—are rated as relatively less impor- ate assistance with their operations (for tant. Only about a third of investors find example, in resolving issues or grievances these services important or critically impor- with the government), the services they tant, the lowest rated among the various fac- typically receive from IPAs are more tors considered. focused on the start-up phase. Among the 11 percent of investors that Some investors value IPA services more did engage with IPAs, their services during than others. In particular, investment pro- entry and establishment were used most motion efforts—exhibitions, advertising, FIGURE 1.17 Investors Value IPA Help in Resolving Problems and Setting Up More than Promotion Efforts Share of respondents (percent) Importance of IPA services Assistance to handle issues and problems, 28 47 21 4 and resolve grievances with the government Information and assistance in setting up 26 47 21 5 Efforts to improve the business environment 23 46 25 5 in the country Meetings with agency officers to 17 37 35 10 discuss investment opportunities Exhibitions about the country at trade 11 31 42 16 shows and other events Advertising about investment opportunities 9 30 41 20 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The questions on IPA services were answered by 632 respondents. These respondents answered somewhat important, important, or critically important on the question, “How important are high-quality services and support from the country’s IPA in your decision to invest in developing countries?”. IPA = investment promotion agency. 38 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 and meetings with agency officers— arbitrary government conduct. Addressing resonate with investors from developing these weaknesses can not only attract new countries more than those from developed investments but also prevent divestments by countries, and with investors in the services existing investors. sector more than those in manufacturing. Addressing policy reforms to attract FDI Meetings with agency officers to discuss and offering aftercare services are equally investment opportunities, information and important. Policy makers tend to focus on assistance in setting up an affiliate, and attracting FDI through investment incentives, assistance in problem resolution are valued facilitation, and proactive investment promo- more by fi rms involved in efficiency-seek- tion. While these are important, investors say ing FDI. In general, the ratings remain rela- that investment protection is even more criti- tively low, but this suggests that IPA cal to them, suggesting that government services in attracting investments can be efforts should also aim to encourage investors better targeted to those companies that to stay in the country and expand their opera- may be more responsive, whenever they tions. Policy initiatives should include align with the country’s target sectors and strengthening investor protection guarantees, target markets. IPAs often focus on tradi- providing proactive investor aftercare, man- tional investors from industrialized econo- aging grievances, and promoting linkages. mies, but as FDI increasingly originates Targeting policies and reforms to relevant from developing countries, IPAs may well types of investors can maximize effectiveness benefit from redirecting their activities and cost efficiency. While most investors accordingly. value some characteristics across the board—such as investment protection guar- antees and transparency and predictability— Policy Implications some policy variables are more important To maximize the gains from foreign invest- for certain investor types. Firms involved in ments, developing country governments efficiency-seeking FDI seem more responsive must adopt effective reform strategies, to incentives. Manufacturing firms may be champion reform at the highest politi- more responsive to business-friendly policies cal levels, and strengthen interagency on importing inputs, while services firms are coordination. They must also balance the more sensitive to adverse government con- public interest with investor preferences duct. MNCs from developing countries to ensure that the host country truly ben- value IPA services and some types of incen- efits from FDI. The results of the survey of tives more than firms from developed econo- MNC executives highlight several priori- mies. These results reinforce the need for ties for policy makers in developing coun- targeted policy approaches by governments, tries seeking to create a conducive business keeping in mind the specific types of FDI climate for FDI: they wish to attract, retain, and harness for Predictable government conduct is at development. least as important to MNCs as countries’ laws and regulations. Investors cited the importance of transparency and predictabil- Annex 1A. Survey Methodology ity in the conduct of government agencies as and Characteristics the most important among investment cli- Methodology mate factors. Investors look not only at poli- cies on paper but also at implementation and The GIC Survey captures perceptions of administration of those policies. international business executives on the Implementation weaknesses can include role of investment climate factors in their bureaucratic inefficiencies, complex regula- FDI decisions. The survey respondents were tions and procedures, and unpredictable or 754 business executives involved with the WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 39 operations of their MNC in developing rate the importance of country character- countries. The sample frame consisted of istics and investment policy factors on a nearly 8,000 eligible companies in the com- scale from 1 to 4 from “not at all impor- mercially available Dunn and Bradstreet tant” to “critically important.” “Critically database. The 754 respondents were execu- important” means it is a deal-breaker—by tives who were reached by telephone and itself, it could change the company’s deci- agreed to participate in the survey. The sion about whether or not to invest in a sample included investors with existing country. investments in at least one developing coun- try. Respondents were a combination of 3. Political risks and investment exit, where executives at the global headquarters and respondents identify experiences of politi- executives at a foreign affiliate. The charac- cal risks and the company’s course of teristics of their firms are discussed below. action. They were also asked whether they The World Bank Group commis- had shut down a foreign affiliate in a sioned a survey firm to conduct 30-minute developing country and their reasons for computer-assisted telephone interviews. doing so. The interviews were conducted in 13 lan- guages: Arabic, Bahasa Indonesia, Chinese, 4. Investment in a specific developing coun- English, French, German, Italian, Japanese, try, where respondents select a specific Korean, Portuguese, Russian, Spanish, and developing country where they are most Turkish. The interviews consisted of a screener familiar with the process of establishing phase, to ensure the eligibility of respondents. an affiliate. Questions on the specific The interviews were conducted between investment included sector, activity, moti- February and June 2017. vation, reinvested earnings, efficiency of government agencies, IPA services used, incentives received, sources of inputs, and Characteristics corporate programs for suppliers. The survey complemented other existing investor surveys by focusing on investment climate variables, such as administrative and Annex 1B. Respondent Profile legal barriers rather than broader economy- 1. Location of company headquarters. wide factors. These specific investment cli- Among 754 respondents, 73 percent were mate variables were areas actionable for headquartered in high-income countries policy makers. and 27 percent in developing countries. The survey was intended to provide a Over half of respondents had headquarters broad understanding of corporate perspec- in Western Europe (figure 1B.1). tives and investor behavior and is not The respondents were stratified by source intended as a benchmarking tool to compare economy of FDI. The sampling method con- countries. sidered whether the source economy was The survey was composed of four developed or developing but did not aim to sections: make the composition of respondents repre- sentative at a country level. Practical consid- 1. General information on the company and erations such as sample size and translations respondent, including sector, number of to local languages precluded the survey employees, and position of the respondent methodology from obtaining a representa- in the company. tive sample of companies globally. The sam- pling method also considered that 2. Importance of factors in investing in a respondents should comprise a large enough developing country, where respondents sample of developing economies as source 40 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 1B.1 MNCs Come from Various Regions and the Chief Executive Officer (CEO) or Chief Levels of Development Finance Officer (CFO), or their equivalent Share of respondents (percent) (table 1B.3). Home economies of investors by region and income level 4. Sectoral Distribution. Some 47 percent of Developed Developing respondents were executives of manufactur- 73.1 26.7 Europe ing firms, 45 percent were from services, 6 East Asia and and Central percent were from extractives, and 2 percent Pacific, 9.8 Latin Asia, were from “other” noncategorized sectors 6.9 America (table 1B.4). and the Caribbean, Table 1B.5 compares the composi- 9.2 tion of survey respondents with global Sub- FDI flows for greenfield investments and North Saharan America, Africa, mergers and acquisitions (M&A). Data 8.8 South Asia 2.4 6.1 on greenfield investments and M&A are Europe and Central Asia, 54.2 East Asia and Pacific 2.0 based on data from UNCTAD’s World Investment Report, based on the total Developed Developing number of investment projects (not value Source: Computation based on the GIC Survey. of investments) over the last five years Note: Respondents were asked to identify the location of their global headquarters. The classifications of developing and developed are based on the World Bank Group’s income level classifications. (2012–16). During this period, there were High-income economies are considered developed economies, while low-, lower-middle-, and 15,692 greenfield investment projects and upper-middle-income economies are considered developing economies. The analysis for this report is unable to disaggregate into income groups owing to the small sample size. MNC = multinational 51,283 M&A purchases. corporation. 5. Number of employees. Large companies countries of FDI. Table 1B.2 compares the with 1,000+ employees constituted composition of respondents from developed 40 percent of the sample. About one-third and developing economies with global out- (32 percent) of the interviewed companies ward FDI stock in 2016. had fewer than 250 employees, and 26 percent had between 251 and 1,000 2. Location of respondent. Of respondents, employees (figure 1B.2). 401 (53 percent) were executives located at the global headquarters while 353 (47 per- 6. Motivation. Only about a third of compa- cent) were executives of an MNC affiliate nies (33 percent) had one dominant moti- in a developing country. vation for an investment in a specific developing country. A significant majority 3. Position of respondent in the company. (62 percent) had two or more FDI motiva- A large majority of respondents were either tions (table 1B.6). WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 41 TABLE 1B.1 Location of Headquarters No. of Percentage of No. of Percentage of Developed economies respondents respondents Developing economies respondents respondents Germany 111 14.72 South Africa 35 4.64 Spain 80 10.61 Argentina 23 3.05 United States 60 7.96 Turkey 20 2.65 Italy 53 7.03 India 16 2.12 Korea, Rep. 37 4.91 Mexico 14 1.86 Austria 36 4.77 Bulgaria 10 1.33 Japan 32 4.24 Brazil 9 1.19 France 30 3.98 China 8 1.06 United Kingdom 28 3.71 Malaysia 6 0.80 Netherlands 22 2.92 Russian Federation 6 0.80 Sweden 20 2.65 Nigeria 4 0.53 Switzerland 20 2.65 Colombia 4 0.53 Canada 6 0.80 Peru 4 0.53 Belgium 5 0.66 Venezuela, RB 4 0.53 Australia 4 0.53 Belarus 3 0.40 United Arab Emirates 1 0.13 Bosnia and Herzegovina 3 0.40 Uruguay 1 0.13 Guatemala 3 0.40 Chile 1 0.13 Romania 3 0.40 Taiwan, China 1 0.13 Serbia 3 0.40 Iceland 1 0.13 Ukraine 3 0.40 Finland 1 0.13 Kenya 2 0.27 Estonia 1 0.13 Costa Rica 2 0.27 Denmark 1 0.13 Panama 2 0.27 Egypt, Arab Rep. 1 0.13 Bolivia 1 0.13 Botswana 1 0.13 Ecuador 1 0.13 El Salvador 1 0.13 Pakistan 1 0.13 Saint Lucia 1 0.13 Sri Lanka 1 0.13 Swaziland 1 0.13 Thailand 1 0.13 Uzbekistan 1 0.13 Djibouti 1 0.13 Ghana 1 0.13 Zambia 1 0.13 Cameroon 1 0.13 42 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 TABLE 1B.2 Composition of Respondents Compared with Global FDI Stock Percent Location of headquarters Percentage of respondents Share of global FDI stock Developed economies 73.21 76.31 Developing economies 26.79 23.69 Source: Computation based on the GIC Survey and UNCTAD. Note: FDI = foreign direct investment. TABLE 1B.3 Position of Respondents in the Company Position No. of respondents Percentage of respondents CFO/Finance director/Treasurer/Comptroller 336 44.6 CEO/President/Managing director 146 19.4 Head of business unit/Head of department 126 16.7 Other C-level executive 61 8.1 SVP/VP/Director 26 3.4 Board member 24 3.2 Director of global operations or global manufacturing 18 2.4 Other 12 1.6 CIO/Technology director 5 0.7 Total 754 100.0 TABLE 1B.4 Sectoral Distribution of Respondents Percentage of Sector No. of respondents respondents Primary Agriculture, hunting, forestry, and fishing 22 2.92 Mining, quarrying, and petroleum 26 3.45 Manufacturing Refined petroleum products, coke, and nuclear fuel 7 0.93 Agroprocessing, food products, and beverages 24 3.18 Textiles, apparel, and leather 23 3.05 Chemicals and chemical products 24 3.18 Rubber 5 0.66 Plastic products 14 1.86 Pharmaceuticals, biotechnology, and medical devices 26 3.45 Metals and metal products 39 5.17 Nonmetal mineral products 3 0.40 Wood and wood products (other than furniture) 3 0.40 Furniture 2 0.27 Paper and paper products 6 0.80 Printing and publishing 4 0.53 Automobiles, other motor vehicles, and transport equipment 67 8.89 Information technology and telecommunications equipment 6 0.80 Machinery, and electrical and electronic equipment and components 64 8.49 Other manufacturing 36 4.77 table continues next page WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 43 TABLE 1B.4 Sectoral Distribution of Respondents (continued) Percentage of Sector No. of respondents respondents Services Electricity, gas, and water 20 2.65 Alternative energy 19 2.52 Construction 53 7.03 Wholesale and retail trade 43 5.70 Hotels and restaurants 7 0.93 Other travel and tourism-related services 8 1.06 Logistics, transport, and storage 35 4.64 Telecommunications 13 1.72 Computer and software services 10 1.33 Financial services including insurance 44 5.84 Real estate 4 0.53 Business services 18 2.39 Professional, scientific, and technical services 32 4.24 Health services 8 1.06 Media and entertainment 7 0.93 Other services 15 1.99 Other 17 2.25 Total 754 100.00 TABLE 1B.5 Sectoral Distribution of Respondents Compared with Global FDI Flows Percent Share of Share of global FDI flows global FDI flows Percentage of Sector for greenfield for M&A respondents Primary 0.5 4.7 6.4 Agriculture, hunting, forestry and fisheries 0.0 0.5 2.9 Mining, quarrying and petroleum 0.5 4.1 3.5 Manufacturing 47.3 22.2 46.8 Food, beverages, and tobacco 3.6 2.8 3.2 Textiles, clothing, and leather 8.8 0.6 3.1 Wood and wood products 0.9 0.2 0.4 Paper and paper products — 0.7 0.8 Publishing and printing 0.1 0.2 0.5 Coke, petroleum products, and nuclear fuel 0.5 0.2 0.9 Chemicals and chemical products 5.0 2.7 3.2 Pharmaceuticals, biotechnology, medical devices — 1.6 3.5 Rubber and plastic products 2.5 0.5 2.5 Nonmetallic mineral products 1.0 0.8 0.4 Metals and metal products 2.2 1.9 5.2 Machinery and equipment, electrical and electronic equipment 12.4 7.0 8.5 Precision instruments 1.1 — — Motor vehicles and other transport equipment 6.7 1.5 8.9 Manufacturing of furniture — 0.2 0.3 Other manufacturing 2.4 1.2 4.8 table continues next page 44 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 TABLE 1B.5 Sectoral Distribution of Respondents Compared with Global FDI Flows (continued) Share of Share of global FDI flows global FDI flows Percentage of Sector for greenfield for M&A respondents Services 52.2 73.2 44.6 Electricity, gas, and water 2.2 1.9 5.1 Construction and real estate 1.6 1.1 7.6 Trade 5.3 4.5 5.7 Hotels and restaurants, travel and tourism-related 0.8 3.0 2.0 Transport, storage, and communications 6.4 49.1 6.4 Finance 7.2 11.1 5.8 Business services 26.2 — 2.4 Public administration and defense — 0.7 — Education 0.7 0.2 — Health and social services 0.5 0.7 1.1 Arts, entertainment, and recreation 1.2 0.4 0.9 Other services 0.2 0.3 2.0 Other — — 2.3 Source: Computation based on UNCTAD World Investment Report 2017, which sourced its data from UNCTAD M&A database and fDi Markets database, the Financial Times, and based on the GIC Survey. Note: Sector categories have been slightly adapted to harmonize across the three data sources. Sectors marked with “—” are not in the list of sectors from their original source. FDI = foreign direct investment; M&A = mergers and acquisitions. FIGURE 1B.2 Size of MNCs by Number of Employees Share of respondents (percent) Less than 100 Don't employees, know/refused, 17 2 More than 10,000 employees, 13 100 to 250 employees, 15 251 to 1,000 employees, 26 1,001 to 10,000 employees, 27 Source: Computation based on the GIC Survey. Note: MNC = multinational corporation. WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 45 TABLE 1B.6 Number of Motivations No. of motivations No. of respondents Percentage of respondents 0 34 4.51 1 249 33.02 2 227 30.11 3 159 21.09 4 64 8.49 5 21 2.79 Total 754 100.00 Annex 1C. Differences by Group than services firms, probably because the effi- The importance of country characteristics var- ciency-seeking motivation is more common in ies by sector and source of FDI. Manufacturing the manufacturing sector than in services. firms find cost of labor and other inputs, and Services firms, on the other hand, are more access to land or real estate, more important sensitive to political stability and security, FIGURE 1C.1 Importance of Country Characteristics by Manufacturing versus Services Firms Share of respondents (percent) Political stability and security Services 52 39 7 3 Manufacturing 48 35 13 2 Legal and regulatory environment Services 47 41 10 1 Manufacturing 33 49 14 2 *** Large domestic market size Services 43 38 14 4 Manufacturing 43 37 14 5 Macroeconomic stability Services 36 46 14 4 and favorable exchange rate Manufacturing 31 42 19 6 Available talent and skill of labor Services 28 44 24 4 Manufacturing 27 46 20 5 Good physical infrastructure Services 25 43 24 7 Manufacturing 24 50 22 3 Low tax rates Services 21 38 31 8 Manufacturing 16 40 33 9 Low cost of labor and inputs Services 15 29 41 14 Manufacturing 21 41 27 9 *** Access to land or real estate Services 13 26 31 29 Manufacturing 14 34 32 17 *** Financing in the domestic market Services 18 29 28 22 Manufacturing 12 25 34 26 *** Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The differences between the two groups are significant at ***p<0.01, **p<0.05 and *p<0.1. 46 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 the legal and regulatory environment, macro- countries, value more highly factors such as a economic stability, and financing in the business-friendly legal and regulatory environ- domestic market (figure 1C.1). Many of these ment; indeed, almost half said that the absence services firms offer financial services, retail of such an environment was a deal-breaker, trade, energy, and telecommunications that versus only 32 percent of respondents in par- are more highly regulated. Investors from ent companies. developing countries also tend to value many The importance of investment climate fac- of these factors highly, compared with their tors also varies by sector. Services firms are counterparts from developed economies— more sensitive to transparency and predict- these characteristics include macroeconomic ability in the conduct of public agencies, stability, low cost of labor and inputs, low tax investment protection guarantees, and ease of rates, and availability of domestic financing starting a business, likely owing to these (figure 1C.2). Respondents from affiliates industries being more highly regulated located in developing countries tend to rate (figure 1C.4). Investors from developing most characteristics as important compared countries also seem to value investment cli- with respondents based at the companies’ mate factors more highly than those from global headquarters (figure 1C.3). This sug- developed economies, but the differences are gests that executives on the ground, who are not statistically significant (figure 1C.5). more aware of the challenges in setting up Respondents from affiliates located in devel- and operating MNC affiliates in developing oping countries tend to rate investment FIGURE 1C.2 Importance of Country Characteristics by Developed versus Developing Source Countries Share of respondents (percent) Developed 49 37 10 3 Political stability and security Developing 54 37 7 2 Developed 39 46 13 2 Legal and regulatory environment Developing 43 45 10 1 Developed 43 38 14 4 Large domestic market size Developing 41 40 14 4 Macroeconomic stability Developed 32 44 18 5 and favorable exchange rate Developing 41 44 10 4 *** Developed 27 47 20 5 Available talent and skill of labor Developing 30 41 26 2 Developed 24 45 25 6 Good physical infrastructure Developing 26 49 20 4 Developed 18 37 34 9 Low tax rates *** Developing 23 44 23 9 Developed 16 34 36 12 Low cost of labor and inputs ** Developing 23 38 30 9 Developed 15 30 32 22 Access to land or real estate Developing 12 34 32 22 Developed 13 28 30 28 Financing in the domestic market *** Developing 24 28 32 15 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The differences between the two groups are significant at ***p<0.01, **p<0.05 and *p<0.1. WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 47 FIGURE 1C.3 Importance of Country Characteristics by Parent Company versus Affiliate Share of respondents (percent) Affiliate 55 35 7 3 *** Political stability and security Parent 46 39 12 2 Affiliate 49 39 10 1 *** Legal and regulatory environment Parent 32 52 14 2 Affiliate 44 37 11 5 Large domestic market size Parent 40 39 17 3 A Macroeconomic stability and ffiliate 40 39 15 5 ** favorable exchange rate Parent 30 48 17 5 Affiliate 33 47 15 4 *** Available talent and skill of labor Parent 22 44 27 5 Affiliate 33 42 18 6 *** Good physical infrastructure Parent 18 49 28 5 Affiliate 23 44 23 8 Low tax rates *** Parent 16 35 39 10 Affiliate 20 37 33 9 Low cost of labor and inputs ** Parent 17 33 36 13 Affiliate 14 34 29 20 Access to land or real estate Parent 14 27 34 23 Affiliate 20 28 28 21 *** Financing in the domestic market Parent 12 27 33 27 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The differences between the two groups are significant at ***p<0.01, **p<0.05 and *p<0.1. FIGURE 1C.4 Importance of Investment Climate Factors by Manufacturing versus Services Firms Share of respondents (percent) Transparency and predictability in the conduct of Services 39 46 12 2 public agencies Manufacturing 33 45 16 4 *** Investment protection guarantees provided in Services 47 37 12 3 the country’s laws Manufacturing 41 36 18 4 ** Ease of obtaining government approvals to start a Services 39 40 16 4 business and to own all equity in the company Manufacturing 32 42 19 5 *** Services 23 32 31 12 Investment incentives such as tax holidays Manufacturing 18 39 33 9 Services 15 37 31 13 Having a preferential trade agreement Manufacturing 13 42 32 9 Services 18 35 31 12 Having a bilateral investment treaty * Manufacturing 12 36 35 14 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The differences between the two groups are significant at ***p<0.01, **p<0.05 and *p<0.1. 48 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 1C.5 Importance of Investment Climate Factors by Developed versus Developing Source Economies Share of respondents (percent) Transparency and predictability in the conduct of Developed 36 45 14 4 public agencies Developing 39 44 16 1 Investment protection guarantees provided in Developed 43 36 15 4 the country’s laws Developing 49 35 13 2 Ease of obtaining government approvals to start a Developed 35 41 18 5 business and to own all equity in the company Developing 38 41 16 3 Developed 18 36 34 10 Investment incentives such as tax holidays Developing 26 34 28 11 Developed 14 40 33 11 Having a preferential trade agreement Developing 17 39 29 12 Developed 13 37 33 14 Having a bilateral investment treaty Developing 18 34 32 11 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: None of the differences is statistically significant. FIGURE 1C.6 Importance of Investment Climate Factors by Parent Company versus Affiliate Share of respondents (percent) Transparency and predictability in the conduct of Affiliate 46 38 12 3 public agencies Parent 29 51 17 3 *** Investment protection guarantees provided in Affiliate 51 33 10 4 the country’s laws Parent 39 39 18 4 *** Ease of obtaining government approvals to start a Affiliate 36 39 18 5 business and to own all equity in the company Parent 36 42 17 4 Affiliate 26 36 28 8 Investment incentives such as tax holidays *** Parent 16 34 36 12 Affiliate 17 39 28 12 Having a preferential trade agreement Parent 12 40 34 10 Affiliate 18 37 27 14 Having a bilateral investment treaty ** Parent 11 36 38 12 Critically important Important Somewhat important Not at all important Don’t know Source: Computation based on the GIC Survey. Note: The differences between the two groups are significant at ***p<0.01, **p<0.05 and *p<0.1. WHAT MAT TERS TO INVESTORS IN DE VELOPING COUNTRIES 49 climate factors as important compared with World Bank Studies. Washington, DC: World respondents based at the companies’ global Bank. doi:10.1596/978-1-4648-0371-0. headquarters (figure 1C.6). In particular, Hufbauer, G.C., J.J. Schott, C. Cimino, M. these are transparency and predictability in Vieiro, and E. Wada. 2013. “Local Content Requirements: A Global Problem.” Policy the conduct of public agencies, investment Analyses 102. Peterson Institute for protection guarantees, investment incentives, International Economics, Washington, DC. and having bilateral investment treaties. MIGA (Multilateral Investment Guarantee Agency). 2013. World Investment and Political Risk 2013. Washington, DC: MIGA. Notes Moran, T. H. 1998. Foreign Direct Investment and Development: The New Policy Agenda 1. This broader definition and use of FDI typol- for Developing Countries and Economies in ogy will be further elaborated in a forthcom- Transition. Washington, DC: Peterson Institute ing World Bank Group publication. for International Economics. 2. In this chapter, “efficiency-seeking” investors ———. 2006. Harnessing Foreign Direct are those respondents who said that lowering Investment for Development: Policies for production costs and establishing a new base for Developed and Developing Countries. exports was one of their motivations for setting Wa s h i n g t o n , D C : C e n t e r f o r G l o b a l up an affiliate in a developing country. Development. 3. Echandi, Krajcovicova, and Qiang (2015) ———. 2011. Foreign Direct Investment provide a literature review of local content and Development: Launching a Second requirements including studies from UNCTAD Generation of Policy Research: Avoiding the (2007), Moran (1998, 2006, 2011), Hufbauer Mistakes of the First, Re-Evaluating Policies and others (2013). for Developed and Developing Countries. Washington, DC: Peterson Institute for International Economics. UNCTAD (United Nations Conference on Trade Bibliography and Development). 2007. Elimination of AON Centre of Innovation and Analytics. 2015. TRIMS: The Experience of Selected Developing Global Risk Management Survey. U.K.: AON Countries. New York and Geneva: UNCTAD. Centre of Innovation and Analytics. ———. 2014. World Investment Prospects Survey: A.T. Kearney. 2017. Foreign Direct Investment 2013–2015. New York: UNCTAD. Confidence Index: Glass Half Full. Washington, ———. 2016. World Investment Report 2016: DC: A.T. Kearney. Investor Nationality: Policy Challenges. Dunning, J.H. 1993. Multinational Enterprises Geneva: UNCTAD. and the Global Economy. Addison Wesley. ———. 2017. World Investment Report 2017: Dunning, J.H., and S.M. Lundan. 2008. Investment and the Digital Economy. Geneva: Multinational Enterprises and the Global UNCTAD. Economy. Cheltenham, U.K.: Edward Elgar WEF (World Economic Forum). 2017. The Global Publishing. Risks Report 2017, 12th Edition . Geneva: Echandi, R., J. Krajcovicova, and C.Z.W. Qiang. WEF. 2015. “The Impact of Investment Policy World Bank. 2014. “Conceptual Framework.” in a Changing Global Economy.” Policy In Making Foreign Direct Investment Work Research Working Paper 7437, World Bank, for Sub-Saharan Africa: Local Spillovers and Washington, DC. Competitiveness in Global Value Chains, edited Ernst & Young Emerging Markets Center. 2015. by Farole, Thomas, and Deborah Winkler, EY’s Attractiveness Survey Africa 2015: 23–55. Washington, DC: World Bank. Making Choices . London: Ernst & Young World Bank Group/MIGA. 2002. Foreign Direct Global Limited. Investment Survey: A Study Conducted by the Gómez-Mera, L., T. Kenyon, Y. Margalit, J.G. Reis, Multilateral Investment Guarantee Agency and G. Varela. 2015. New Voices in Investment: with the Assistance of Deloitte & Touche LLP. A Survey of Investors from Emerging Countries. Washington, DC: World Bank Group. 2 Effects of FDI on High-Growth Firms in Developing Countries José-Daniel Reyes F oreign direct investment (FDI) pro- successfully transmitted to local firms, motes economic growth, job creation, improves their productivity. At the same time, and poverty reduction. Countries more FDI can exert a negative effect by increasing open to trade and investment tend to be the competition in local input and output more productive and grow faster (Dollar markets, thereby undermining the perfor- 1992; Harrison 1996; Frankel and Romer mance of local firms. The balance between 1999). Policy makers seek to attract FDI to these two forces determine the overall effect create jobs, bring in cutting-edge knowledge of foreign firms on local enterprises. At the and technology, connect to global value sectoral level, greater competition in prod- chains, and diversify and upgrade their econ- uct and factor markets results in the effi- omies’ production capabilities.1 The poten- cient reallocation of resources from less tial transmission of knowledge between productive to more productive firms, thereby foreign firms and local enterprises is an increasing sectoral productivity over the added benefit of FDI, one that can improve long run.2 the productivity of domestic enterprises and FDI can benefit domestic firms through thus make economic growth more inclusive. two main channels:3 The effects of foreign investment on the host economy are therefore a crucial element • Contractual linkages between foreign in a country’s development strategy. These firms and local suppliers that promote the FDI effects—or spillovers—on domestic firms formal transmission of foreign firms’ can be positive or negative, depending on knowledge and practices, which may help whether local firms improve or worsen their domestic suppliers upgrade their technical performance as a result of FDI. It can and quality standards.4 have positive effects if it brings foreign tech- • The demonstration effect, in which domes- nology and frontier knowledge that, if tic firms imitate foreign technologies or 51 52 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 managerial practices either through obser- The evidence presented here shows that link- vation or by hiring workers trained by the ages programs to connect local suppliers foreign company.5 with foreign firms can help achieve this goal. Considering the different absorptive capaci- This chapter explores the role of these ties of indigenous firms and the various two transmission channels of FDI spillovers potential market failures is fundamental for on the performance of firms across 50 sec- evidence-based policy making. Particularly tors and 121 economies in the developing important is the design of programs that tar- world. 6 Employing data from the World get high-potential suppliers and tackle Bank’s Enterprise Surveys, it constructs specific failures, such as information asym- sectoral measures of the linkages and dem- metries, and scale and quality constraints of onstration channels and examines their role domestic suppliers. Linkages programs in the ability of domestic firms in the sector should include a comprehensive set of inter- to benefit from FDI. The analysis reveals a ventions aimed at the supply side, the large variation of FDI spillovers across local demand side, and market exchange. firms. In line with the literature, an average Compulsory local content requirements may firm in the developing world does not neces- cause more harm than good because they sarily benefit from these FDI effects (Damijan may discourage FDI from entering the coun- and others 2013; Fons-Rosen and others try, thereby shutting down any channel of 2017). It is primarily the local high-growth positive spillover effects. A comprehensive firms that are able to internalize FDI spill- policy intervention aimed at reducing search overs through both linkages and demonstra- costs and tackling constraints of both buyers tion channels.7 For the linkages channel, an and sellers is more effective than a piecemeal increase of 1 percentage point in the share of approach. inputs sourced domestically by foreign firms is correlated with a 0.6 unit rise in the mea- sure of output growth of domestic high- High-Growth Firms Are growth firms. For the demonstration Important for Job Creation, channel, an increase of 1 percentage point in the share of foreign output in the sector is and Are Small and Young correlated with a 0.1 unit gain in output While the private sector is the main engine of growth of high-growth firms. countries’ economic growth, only a small This chapter therefore focuses on domestic part of the private sector—the “high- high-growth firms, which the analysis shows growth” firms—plays a disproportionately benefit from FDI more than other firms. This large role in job creation (Coad and others is likely due to their higher absorptive 2014; Haltiwanger, Jarmin, and Miranda capacity—their ability to recognize the value 2016; Hsieh and Klenow 2014). Identifying of new information, assimilate it, and apply it them and assessing the constraints that hin- to improve production processes. 8 High- der the emergence and performance of these growth firms account for a sizable share of high-growth firms is critical to realize their job creation and productivity gains in devel- full potential (box 2.1). oping countries. The distinctive characteris- The identification of high-growth firms in tics of these firms have been the subject of this dataset focuses exclusively on domesti- study from the perspective of both individual cally owned enterprises to highlight the ability firms interested in sales and revenue growth of these firms to benefit from the presence of and policy makers interested in job creation foreign firms. The analysis uses the rate of and economic growth. firm-level job creation to characterize firm From a policy perspective, developing growth.9,10 In each country, high-growth firms countries are interested in spreading the are located in the top fifth percentile of the benefits of FDI to the local economy. distribution of firm-level job growth rates EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 53 BOX 2.1 Factors Influencing High-Growth Firms: The Four-Layer Onion Framework Firm performance, and hence the potential emergence • Personal and professional networks. of high-growth firms, is influenced by a variety of • The overall business environment in which firms factors: operate. • Individual characteristics of entrepreneurs such as The “four-layer onion” provides a representation of age, education, experience, and motivation. these factors (see figure B2.1.1). • Firm-level attributes such as firm age, size, location, sector, and absorptive capacity. FIGURE B2.1.1 The Four-Layer Onion Framework of Growth Factors Business environment Personal and professional networks Enterprise characteristics Entrepreneur characteristics Source: Hampel-Milagrosa, Loewe, and Reeg 2015. over two years. The key advantage of this FIGURE 2.1 High-Growth Firms Create the Most Jobs method is that it establishes country-specific Distribution of firm-level growth rates in Indonesia, 2012–14 minimum growth rates required for firms to 4 be classified as high-growth, thereby taking into account characteristics that support or hinder the performance of the private sector 3 in each economy (annex 2A provides the com- plete list of economies and the years in which each Enterprise Survey was conducted).11 2 The case of Indonesia—where the Enterprise Survey was conducted in 2015— illustrates the identification of high-growth 1 firms. According to the chosen criterion, high-growth firms increased employment by High-growth firms at least 35.3 percent between 2012 and 0 2014.12 In figure 2.1, these firms are shown Firm-level growth rates (employment) in the shaded right tail of the firm growth Source: Computation based on data from Enterprise Surveys, the World Bank. distribution. Note: This figure shows the distribution of firm-level mid-point growth rates for Indonesia between 2012 and 2014. The survey was conducted in 2015 and firms were asked about the total number of Applying the criterion to the sample of full-time employees the year before (2014) and three years ago (2012). The dotted line indicates the countries, two common characteristics of 95th percentile. The shaded area of the distribution indicates the presence of high-growth firms. 54 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 high-growth firms in the developing world the median size of these firms is less than emerge: they tend to be small and young. 10 employees (annex 2C). High-growth Across countries, they represent 7.9 percent firms are also more common among young of small firms, relative to 2.3 percent of enterprises; 6.9 percent of firms younger large firms (figure 2.2). In 89 countries, than 10 years are high-growth while only 2.3 percent older than 50 years are high- growth (figure 2.3). The median age of FIGURE 2.2 High-Growth Firms Tend to Be Small... high-growth firms is lower than the median Share of high-growth firms, by size bins age of the rest of firms in 105 countries in 7.9 the sample (annex 2C).13 8 High-growth businesses in the developing 7 world exist in all economic sectors but are Share of high-growth firms (percent) more common in services than in manufac- 6 turing (table 2.1). Information and commu- nications technology (ICT) and the 5 construction sector show the highest shares of high-growth firms; these firms account for 4 3.2 8.1 percent of all firms in the ICT sector and 3 6.6 percent of all firms in the construction 2.3 sector.14 In terms of output and employment 2 growth, high-growth firms in services out- perform those in manufacturing. Overall, 1 high-growth firms in services grew in terms Small Medium Large Size of firm of employment by 133 percent (versus 127 percent in manufacturing) and increased Source: Computation based on data from Enterprise Surveys, the World Bank. sales by 40 percent over the previous two Note: This figure shows the number of high-growth firms as a share of the total number of firms, by size bins. Small firms have fewer than 20 employees, medium firms have 20–100 employees, and years (versus 38 percent in manufacturing). large firms have more than 100 employees. Many variables determine the presence of high-growth firms across sectors in developing economies. As noted above, FIGURE 2.3 … and Young these elements range from the characteris- Share of high-growth firms, by age bins tics of the entrepreneur to the regulatory and institutional framework in which the 8 firm operates. Key determinants also vary 6.9 7 across the life cycle of the firm, but the Share of high-growth firms (percent) process of internationalization is usually a 6 5.0 central element in the firms’ success 5 (box 2.2). 4.0 In sum, high-growth firms are few in 4 3.6 3.0 number but critical for job creation. The 3 2.3 evidence indicates that they are young, small, present across all economic activities, 2 and diverse in terms of the factors that 1 determine their performance. Their strong performance indicates their superior capa- 0 1–9 10–19 20–29 30–39 40–49 ≥50 bilities relative to other firms in the economy Age of firm (years) facing the same constraints on operations and growth, which enable them to benefit Source: Computation based data from Enterprise Surveys, the World Bank. Note: This figure shows the number of high-growth firms as a share of the total number of firms, from the presence of multinational corpora- by age bins. tions (MNCs). EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 55 TABLE 2.1 High-Growth Firms Appear in All Economic Sectors Firm-level employment and output growth across sectors Share of high- High-growth firms Rest of firms growth firms Rest of in the sector Employment Output Employment Output High-growth firms [3] = [1]/([1] + growth growth growth growth ISIC codes—sector firms [1] [2] [2]), percent (percent) (percent) (percent) (percent) Manufacturing 1,608 27,188 5.6 127 38 0 14 17—Textiles 158 2,414 6.1 124 43 0 13 29—Machinery and 123 1,972 5.9 100 33 0 15 equipment 18—Apparel and fur 226 3,574 5.9 141 40 0 14 28—Metal products 180 2,938 5.8 150 47 0 17 15—Food products 393 6,508 5.7 133 35 0 14 and beverages 36—Furniture 114 1,927 5.6 150 52 0 15 24—Chemicals 150 2,789 5.1 132 35 0 17 26—Nonmetallic 144 2,684 5.1 130 34 0 11 mineral products 25—Rubber and 120 2,382 4.8 100 33 0 14 plastic Services 1,479 24,446 5.7 133 40 0 13 64 & 72—ICT 116 1,319 8.1 115 53 9 17 45—Construction 173 2,463 6.6 115 53 0 12 50–52—Wholesales 929 14,845 5.9 133 39 0 13 and retail trade 60–63—Transport 109 2,251 4.6 150 34 0 11 and storage 55—Hotels and 152 3,568 4.1 130 33 0 9 restaurants Source: Computation based on data from Enterprise Surveys, the World Bank. Note: This table shows the total number of firms by type and their associated employment and output growth across economic sectors. To reduce clutter, sectors with fewer than 100 enterprises have been dropped. Sectors are ranked by the presence of high-growth firms (column [3]). The data use the revision 3.1 of the International Standard Industrial Classification (ISIC). For output and employment, the table presents the median standard growth rate within each cell. ICT = Information and communications technology. BOX 2.2 AAA Growers: A High-Growth Firm in Kenya AAA Growers is a company that produces vegetables The management team cited three elements as central and flowers in Kenya—and is the largest commercial to the company’s success: grower and exporter of chilies in the nation. The com- pany started with 50 employees in 2000 and now owns • Family support to set up the business. Family five farms that employ some 4,000 during peak seasons. capital was used to set up and maintain low-scale The workforce consists of rural workers, 60 percent of operations during the company’s first three years. whom are women. The main objective of AAA Growers This period did not generate positive margins but since its inception is to produce vegetables to export, was central to learning about the dynamics of primarily to the U.K. market. Currently, about 98 per- different crops, the requirements to export, and the cent of its production is sold in international markets. need to build professional networks. box continues next page 56 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 BOX 2.2 AAA Growers: A High-Growth Firm in Kenya (continued) • Connection with foreign buyers. Establishing a allowed AAA Growers to invest in state-of-the-art commercial presence in international markets was equipment and installations, which helped not only challenging. The owners employed family connections to expand output but also to comply with stringent to identify potential buyers in the U.K. market and production and agricultural standards in the Euro- secured small orders with the goal of building long- pean market. term professional relationships. The first three to five years of operations of the company were dedicated After growing at a high rate over the last 10 years, mostly to identifying and securing international buyers. the company is now consolidating. The top priority for • IFC funding to set up large-scale operations. An management is to stabilize the company’s operations International Finance Corporation (IFC) loan to ensure sustainable expansion. High-Growth Firms Benefit from increasing sectoral productivity over the FDI Mainly through the Linkages long term. Because all firms in the same sector face Channel the same degree of competitive pressures This section looks at the two channels posed by the presence of the foreign firm, through which FDI affects domestic enter- their ability to ultimately benefit from FDI prises, with a focus on high-growth firms. hinges on whether they can capture positive The linkages channel is characterized by spillovers through the linkages and demon- direct contractual arrangements in which stration channels. 15 This ability, in turn, domestic firms become suppliers of foreign depends on their absorptive capacity—the firms. The demonstration channel enables ability to recognize the value of new infor- domestic firms to replicate foreign technolo- mation, assimilate it, and apply it to improve gies or management practices either through production processes. By virtue of their fast observation or by hiring workers trained by growth trajectory, which may reflect high foreign firms. Thus, the stronger the presence absorptive capacity and productivity, high- of FDI in the sector, the more opportunities growth firms may be better able to capture for the demonstration channel to positively positive spillovers than other local firms. affect local firms. Using information from the World Bank’s But while foreign firms bring technology Enterprise Surveys, the team constructed indi- and frontier knowledge that can improve cators of the linkages and demonstration the performance of indigenous firms, they channels across 50 sectors and 121 develop- may also increase competitive pressures in ing countries. Following the literature, the the host economy, which could hurt some linkages channel is captured by the average local businesses (Alfaro and Chen, forth- share of inputs of domestic origin that foreign coming; Fons-Rosen and others 2017). The firms acquire in the host country.16 The dem- relative magnitude of these two forces deter- onstration channel is measured as the share of mines the ultimate effect on domestic firms. foreign output in total output (see Blalock At the sectoral level, however, more compe- and Gertler 2009; Farole and Winkler 2015; tition promotes the efficient reallocation of and annex 2D). These measures represent the factors of production from low-productivity importance of the FDI spillover channels to high-productivity firms, thereby within country-sector observations and, EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 57 therefore, capture the potential for intra- explain this: First, the competition that for- industry spillover effects.17 eign firms bring to the domestic market out- The relevance of the transmission channels weighs the FDI benefits that the average firm of FDI spillovers varies across sectors and internalizes. Second, the low absorptive countries. On average, linkages are more capacity of the average firm prevents it from apparent in manufacturing than in services capturing more FDI benefits.18 (table 2.2). In manufacturing, Asia shows the Contrary to these aggregate results, the highest prevalence of linkages. In East Asia, analysis finds that high-growth firms are able for example, foreign manufacturing firms to capture the benefit of FDI in their markets source 70 percent of the inputs locally, rela- through both channels. The results are par- tive to the average for the rest of the world of ticularly significant for the linkages channel, about 60 percent. Demonstration effects are where an increase of 1 percentage point in the relatively balanced between manufacturing share of inputs that are sourced domestically and services; foreign firms account broadly by foreign firms is associated with a 0.6 unit for 20 percent to 30 percent of sectoral out- gain in the measure of output growth of high- put across sectors and regions. growth firms (figure 2.4).19 The demonstra- The sole presence of linkages and demon- tion effect is also positively related to the stration channels does not guarantee that performance of high-growth firms, albeit its domestic enterprises benefit from FDI. impact is lower: an increase of 1 percentage Domestic firms can become suppliers of for- point in the share of foreign output in the sec- eign enterprises but may be incapable of using tor is associated with a 0.1 unit rise in the the information acquired to improve their output growth of high-growth firms.20 The production techniques. Arguably, the trans- impact of these channels on the performance mission of FDI benefits to local firms via the of non-high-growth firms is also positive but demonstration channel can be even more statistically insignificant.21 challenging, given the absence of a direct link High-growth firms are better able to inter- between the foreign and domestic enterprises. nalize foreign technologies and processes to To examine the relationship between domestic improve their productivity and counterbal- firms’ performance and the two channels of ance FDI’s competitive effect. From a policy FDI spillover, the analysis employs a regres- perspective, increasing absorptive capacity in sion framework to investigate whether firms domestic enterprises is therefore key to maxi- operating in sectors with high potential for mizing the benefits of FDI for job creation. FDI spillover effects—as indicated by the The importance of FDI for the perfor- presence and importance of the linkages and mance of local high-growth firms varies demonstration channels—display a higher across regions. Employing the same empirical rate of output growth. The analysis differenti- framework, the analysis estimates the role of ates between high-growth firms and others. the linkages and the demonstration channels The regression controls for other variables rel- across six regions of the world (figure 2.5 and evant to firm growth, specifically, age, export annex 2D). The analysis yields three key status, and labor productivity (annex 2D). messages: The results indicate that, on average across firms and countries, FDI benefits are • High-growth firms in Sub-Saharan Africa not uniformly transmitted to local firms. do not internalize FDI spillovers. Since the While both linkages and demonstration lion’s share of FDI going to Africa is directed channels are positively correlated with out- to natural resources, this result may indicate put growth at the firm level, they are not sta- that the potential of this type of investment tistically different from zero. In other words, to generate positive spillovers is limited. the average firm in the developing world is • Europe and Central Asia is an outlier unable to benefit from the presence of foreign because the demonstration channel out- companies. Two self-enforcing mechanisms weighs the linkages channel. In fact, TABLE 2.2 Linkages Are More Important in Manufacturing while Demonstration Effects Are Balanced across Sectors Average size of linkages and demonstration channels across sectors and regions Latin America and Middle East and East Asia and Pacific Europe and Central Asia the Caribbean North Africa South Asia Sub-Saharan Africa ISIC codes—Sector Linkages Demonstration Linkages Demonstration Linkages Demonstration Linkages Demonstration Linkages Demonstration Linkages Demonstration Manufacturing 0.7 0.3 0.6 0.2 0.6 0.3 0.6 0.2 0.8 0.1 0.6 0.4 15—Food 0.8 0.3 0.7 0.2 0.7 0.4 0.7 0.2 0.8 0.1 0.7 0.5 products and beverages 17—Textiles 0.8 0.3 0.7 0.2 0.6 0.3 0.7 0.1 0.8 0.1 0.5 0.5 18—Apparel 0.6 0.4 0.5 0.2 0.7 0.2 0.4 0.3 0.6 0.0 0.6 0.2 and fur 24—Chemicals 0.7 0.3 0.6 0.4 0.5 0.4 0.5 0.2 0.8 0.1 0.4 0.4 25—Rubber and 0.7 0.1 0.5 0.2 0.4 0.5 0.6 0.1 0.9 0.0 0.4 0.5 plastic 26—Nonmetallic 0.8 0.4 0.7 0.3 0.7 0.2 0.8 0.2 1.0 0.2 0.7 0.4 mineral products 28—Metal 0.6 0.2 0.6 0.1 0.5 0.5 0.6 0.2 0.9 0.0 0.5 0.4 products 29—Machinery 0.8 0.2 0.6 0.2 0.7 0.5 0.6 0.4 0.9 0.1 0.6 0.3 and equipment 36—Furniture 0.7 0.3 0.4 0.1 0.8 0.2 0.8 0.1 0.6 0.0 0.6 0.2 Services 0.3 0.2 0.0 0.1 0.3 0.3 0.7 0.2 0.0 0.0 0.4 0.4 45—Construction — 0.1 0.0 0.1 0.7 0.2 0.7 0.2 — 0.0 0.7 0.4 50–52—Wholesale 0.7 0.2 0.0 0.2 0.7 0.3 0.3 0.2 0.0 0.0 0.5 0.3 and retail trade 55—Hotels and — 0.2 — 0.2 — 0.3 0.9 0.1 0.0 0.1 0.4 0.5 restaurants 60–63—Transport 0.7 0.2 0.0 0.1 — 0.3 1.0 0.2 0.0 0.0 0.6 0.3 and storage 64 & 72—IT and 0.2 0.1 0.0 0.2 — 0.4 0.6 0.2 — 0.1 — 0.3 communications Source: Computation based on data from Enterprise Surveys, the World Bank. Note: This table shows the average value of the linkages and the demonstration effects across economic sectors and world regions. For the linkages channel, each figure shows the average share of domestically sourced input for foreign firms within the sector. For the demonstration channel, each figure shows the average share of foreign output as a percentage of total sectoral output. — = data unavailable. EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 59 the role of the demonstration channel is FIGURE 2.4 High-Growth Firms Benefit from the Presence of much larger than in other regions. Foreign Firms Average impact of FDI spillovers on firm growth, by firm type • The linkages channel is the key engine for FDI spillovers to high-growth fi rms 0.8 in Latin America and the Caribbean, East Asia and Pacific, South Asia, and the 0.7 Middle East and North Africa. 0.62 0.6 Point estimates and 90% confidence intervals The linkages channel is more important 0.5 than the demonstration channel for leverag- ing FDI spillovers in both manufacturing and 0.4 services for high-growth firms. In services, 0.3 both linkages and demonstration channels transmit FDI benefits (figure 2.6 and annex 0.2 2D). High-growth firms in manufacturing, 0.12 however, benefit from FDI only through link- 0.1 0.05 ages. The findings indicate that a 1 percentage 0.01 0 point increase in linkages is associated with a 0.7 unit rise in the measure of sales growth of –0.1 high-growth firms in manufacturing, and a –0.2 0.5 unit gain in service sectors. In services, an High-growth Rest of firms High-growth Rest of firms increase of 1 percentage point in the demon- firms firms stration channel is associated with a 0.2 unit Linkages channel Demonstration channel increase in the measure of sales growth of Source: Computation based on data from Enterprise Surveys, the World Bank. domestic high-growth firms. The demonstra- Note: This figure shows the estimated coefficient of the linkages and demonstration channels on tion channel is not statistically significant for high-growth firms and the rest of businesses in a sample of 121 countries. Vertical lines capture 90 percent confidence intervals. The estimates correspond to the estimation shown in column 7 in the manufacturing sector’s high-growth firms. table 2D.1 in annex 2D. FDI = foreign direct investment. FIGURE 2.5 The Linkages Channel More Efficiently Transmits FDI Benefits in Nearly All Regions FDI spillovers to high-growth firms, by region a. Linkages channel b. Demonstration channel 3.0 3.0 Point estimates and 90% confidence intervals Point estimates and 90% confidence intervals 2.5 2.5 2.25 2.0 2.0 1.5 1.5 1.11 1.0 1.0 0.69 0.5 0.50 0.5 0.38 0.10 0.02 0.06 0.08 0.08 0 0 –0.27 –0.03 –0.5 –0.5 –1.0 –1.0 SSA ECA LAC EAP SAR MENA EAP MENA SSA SAR LAC ECA Source: Computation based on data from Enterprise Surveys, the World Bank. Note: These figures show the estimated coefficient of the role of the channels for foreign direct investment (FDI) spillover effects on high-growth firms, by region. Vertical lines capture 90 percent confidence intervals. Regression results are presented in annex 2D. The regions are EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa. 60 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 2.6 High-Growth Firms Benefit from FDI Mainly through Policy interventions aimed at strengthening the Linkages Channel, Both in Services and Manufacturing linkages programs should tackle specific mar- Average impact of spillover effect on high-growth firms, by sector ket failures. The most common failure pre- 0.9 venting the development of linkages is 0.8 asymmetry of information, which increases Point estimates and 90% confidence intervals 0.7 search costs for both buyers and sellers 0.7 (Monge-González and Rodríguez-Álvarez 0.6 2013). Foreign companies may find it difficult 0.5 to identify potential local suppliers while 0.5 0.4 domestic enterprises may struggle to identify potential contracting opportunities with for- 0.3 eign enterprises. Another common failure is 0.2 0.2 the small size and scale of domestic suppliers 0.1 0.1 who may find it impossible to respond to large 0 orders from MNCs, yet find it too risky to invest in production expansion. Additionally, –0.1 Linkages Demonstration Linkages Demonstration domestic enterprises may not have the produc- tion standards required to meet minimum Services Manufacturing quality requirements of foreign enterprises. Source: Computation based on data from Enterprise Surveys, the World Bank. Linkages programs should consider a mech- Note: This figure shows the estimated coefficient of the role of foreign direct investment (FDI) spillovers on high-growth firms in manufacturing and services. Vertical lines capture 90 percent anism to identify and support high-potential confidence intervals. Regression results are presented in annex 2D. local firms in becoming suppliers of foreign firms. The evidence presented in this chapter shows that not all domestic firms are equal in Policies to Maximize the Gains terms of their ability to benefit from FDI. from FDI: Promoting Linkages for High-growth firms have the absorptive capac- ity allowing them to internalize FDI spillovers Development that, in turn, improve their productivity and While FDI may result in negative distribu- support job creation. There is no empirical tional effects for particular groups of local basis for setting up linkages programs to sup- firms, countries open to multinational pro- port the broad group of small- and medium- duction experience aggregate productivity sized enterprises (SMEs), yet such programs gains. These benefits can be accrued via posi- are rather common in many economies.22 tive spillovers for domestic enterprises and Linkages programs should include a com- the reallocation of factors of production prehensive set of interventions to tackle con- from less productive firms to more produc- straints at the market level, the supply side tive firms. From a policy perspective, govern- (domestic firms), and the demand side (for- ments are advised to design policies to reduce eign company). A comprehensive approach is the adjustment cost of the reallocation pro- more effective than a piecemeal one. cess and increase the ability of local enter- Interventions at the market level seek to prises to internalize positive spillovers from reduce search costs and facilitate matchmak- the presence of foreign firms. Policies and ing. For example, they usually tackle informa- programs that can increase the absorptive tion asymmetries by providing information capacities of local enterprises—mainly by about business opportunities for both buyers promoting domestic linkages—will maxi- and sellers. The most common intervention is mize the potential gains from FDI. While the the creation of supplier databases that contain analysis focuses on connecting domestic contact details and the menu of goods and firms to MNCs, these policies can be tailored services offered by potential domestic suppli- to also support linkages between domestic ers. Matchmaking services among buyers and suppliers and large buyers. sellers are another common practice for EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 61 promoting business-to-business transactions. policies may also include more broad-based They include organizing trade fairs, support- entrepreneurship development and training ing supplier audits, and organizing site visits, programs to support local skills development missions, exhibitions, and seminars. as well as help in obtaining international qual- Supply-level interventions seek to improve ity certifications (box 2.3). The second set of the capacity of domestic firms to meet mini- policies leverages and encourages existing busi- mum scale and quality standards expected by ness groups (business clusters, for instance) to MNCs. Two common policy instruments are tackle their capacity constraints. The rationale relevant. The first set of policies is aimed at is that domestic firms acting in a coordinated improving firm-level productivity via the cre- manner can provide the scale expected by for- ation of supplier development programs that eign firms, thus allowing the cluster to jointly provide targeted training support aligned with fulfill large orders from foreign companies. buyer requirements. These programs are often Demand-level interventions are directed at based on a partnership between domestic foreign firms with the goal of encouraging their enterprises and foreign firms, and may also increased use of domestic inputs. Financial and include local educational institutions (for tax incentives are the most common policy example, universities and colleges) and consult- tool. Examples include exempting foreign firms ing firms in the form of joint research projects, from value added taxes to encourage their use customized training programs, and help in of local rather than imported inputs, treating identifying local strategic partners. These expenditures incurred in the creation of BOX 2.3 Chile’s Supplier Development Program Chile’s Supplier Development Program (SDP) was stage. The diagnostic stage lasts up to six months and launched by the Chilean Economic Development identifies areas of intervention that the sponsor (that Agency (Corporación de Fomento de la Producción is, the large fi rm) wishes to develop with its suppli- de Chile; CORFO) in 1998. The program is aimed ers. The result is a development plan designed by a at improving and stabilizing the commercial linkages consultant or consulting firm. CORFO pays for up to already existing between domestic suppliers and their 50 percent of its cost with a ceiling of $16,000. The large-firm customers to achieve higher levels of adapt- development stage is the implementation of the devel- ability and guarantee the quality of goods and services opment plan and can last up to three years. CORFO at different stages of production. By requiring that a pays for up to 50 percent of the cost of this stage with commercial relationship be already established among an annual ceiling of $110,000 (or $5,000 per supplier firms, the program sought to ensure that suppliers fi rm). CORFO assesses annually the renewal of the were local firms with high potential. The SDP provides project fi nancing depending on the implementation partial funding to strengthen the management of local progress. The implementation of the development businesses through specialized services, professional plan is the responsibility of the sponsor fi rm and can advice, training, and technology transfers. be carried out by a consultant or consulting fi rm or by For a large fi rm to be eligible to participate in the the sponsor’s in-house staff. program and receive a subsidy to train the local fi rms A rigorous impact evaluation has shown that the that make up its supply chain, its net annual sales had SDP not only increased sales, employment, and the to be greater than or equal to $42.6 million in August sustainability of suppliers, but also improved the sales 2010. Each project must include at least 20 domestic of large fi rms and raised their probability of becoming fi rms (in the agriculture and forestry sector), or a min- exporters (Arráiz, Henríquez, and Stucchi 2013). The imum of 10 in other economic activity sectors. After positive effect on suppliers appeared one year after the project is approved, the program is implemented the fi rms enrolled in the program while the impact on in two stages: a diagnostic stage and a development large fi rms appeared after two years. 62 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 linkages (for example, training or research and by indigenous firms (the demonstration chan- development activities) as tax- deductible nel). These workers can also launch successful expenses, incentives for training programs of business ventures on their own. local suppliers, and co-financing skills develop- Lack of access to finance often limits the ment activities (see box 2.3). While compul- ability of local firms to strengthen their pro- sory local content requirements are sometimes ductivity and, therefore, improve their chances used, they can discourage the entry of foreign of becoming suppliers of foreign enterprises. investors and, more important, preclude alto- Various types of incentives can financially and gether the entry of foreign technology that legally support local firms in obtaining finance could itself be the source of positive spillover to fund further investment in their human cap- effects. In practice, these requirements are also ital, technological and managerial capacities; often circumvented by foreign companies, ren- and lower the risks of linkages. Some of these dering them largely ineffective (Echandi, interventions include loans, grants, guaran- Krajcovicova, and Qiang 2015; Hufbauer, tees, and legal protection against unfair con- Schott, and Cimino-Isaacs 2013). tractual arrangements. Key elements for creating an enabling policy environment for linkages also include a suitable lead agency, proper coordination mechanisms Conclusion across institutions, and strong stakeholder Foreign investors bring a wide range of engagement. The lead agency should have knowledge and know-how with the potential political clout and a clear mandate to coordi- to bring positive spillovers for the host econ- nate among the different agencies involved in omy. These benefits, however, are not guaran- private sector development. The lead agency, teed. Indeed, foreign firms may also generate typically associated with the Ministry of competitive pressures in the local economy to Commerce and Industry, should organize the the detriment of some local firms. The bal- representatives of different agencies involved in ance between these potentially positive and supporting linkages programs, which can negative impacts determines the overall effect include the investment promotion agency, the of foreign firms on local enterprises. regulators of special economic zones, private Over the long run, competitive pressures sector representatives, the agency to support encourage the efficient reallocation of factors small and medium enterprises and industry within sectors, thereby increasing sectoral pro- associations. Clear operating rules should gov- ductivity. Foreign knowledge and technology ern the coordination mechanism among them can be transferred to domestic firms through to ensure policy coherence along the different two main channels. The first is linkages parts of the program. A key component of the between foreign firms and domestic suppliers. linkages program is also the constant interac- The second is the demonstration channel tion with the private sector, including feedback through which domestic firms imitate and rep- mechanisms. Designing and implementing rig- licate foreign technologies and management orous impact evaluation of the linkages pro- practices in their own production processes. grams are useful for basing policy design and Employing a firm-level dataset for 121 decisions on evidence. developing economies, this chapter evaluates Last, flexible labor markets and better the role of these channels in supporting the access to finance remain key constraints on performance of local firms across the develop- the ability of domestic firms to internalize FDI ing world. It finds that high-growth firms spillovers. A flexible labor market facilitates internalize positive spillover effects mainly the movement of managers and skilled work- through the linkages channel. This points to ers between foreign and local companies. The their superior absorptive capacities, which shift of experienced workers from foreign make them ideal targets for policy interven- firms to domestic enterprises could be the tions aimed at maximizing the benefits of FDI channel through which spillovers are accrued in the local economy. EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 63 Governments can implement an array of More policy-oriented research is needed to policies to maximize the potential for FDI better understand the role of FDI on the spillover. Government interventions should performance of domestic high-growth firms. seek to offset specific market failures, which One aspect is the long-term consequences of usually take the form of information asym- FDI spillovers, particularly on productivity, metries or the small scale and low quality of innovation, and the ability of high-growth domestic firms. Linkages programs should firms to create better jobs. Another aspect that consider a mechanism to identify domestic merits further study is the different potential of suppliers with potential to connect with FDI spillovers that different types of FDI bring foreign firms. Supplier databases and to developing countries. Reyes (forthcoming) matchmaking services are the most com- finds that FDI embedded in global value chains mon tools to promote linkages. Supplier may have higher potential to generate FDI development programs and support for spillovers than FDI that seeks to serve mostly cluster development are interventions that the domestic market. seek to improve the capacity of domestic But regardless of the distributional impact firms to connect to foreign firms. Financial of foreign firms on domestic businesses, coun- and tax incentives are also used to encour- tries open to multinational production tend to age the use of domestic inputs by foreign experience aggregate productivity gains. enterprises. A flexible labor market can go a These benefits are accrued through spillover long way to support workers’ movement effects on domestic firms and the reallocation from foreign firms into the domestic of factors of production from less productive economy. firms to more productive firms. Annex 2A. Economies Included in the Analysis East Asia and Pacific Cambodia (2016), China (2012), Fiji (2009), Indonesia (2015), Lao PDR (2016), Malaysia (2015), Micronesia, Fed. Sts. (2009), Mongolia (2013), Myanmar (2014), Papua New Guinea (2015), Philippines (2015), Samoa (2009), Solomon Islands (2015), Thailand (2016), Timor-Leste (2015), Tonga (2009), Vanuatu (2009), Vietnam (2015) Europe and Central Asia Albania (2013), Armenia (2013), Azerbaijan (2013), Belarus (2013), Bosnia and Herzegovina (2013), Bulgaria (2013), Macedonia, FYR (2013), Georgia (2013), Hungary (2013), Kazakhstan (2013), Kosovo (2013), Kyrgyz Republic (2013), Moldova (2013), Montenegro (2013), Romania (2013), Serbia (2013), Tajikistan (2013), Turkey (2013), Ukraine (2013), Uzbekistan (2013) Latin America and the Argentina (2010), Belize (2010), Bolivia (2010), Brazil (2009), Colombia (2010), Costa Rica (2010), Caribbean Dominica (2010), Dominican Republic (2010), Ecuador (2010), El Salvador (2010), Grenada (2010), Guatemala (2010), Guyana (2010), Honduras (2010), Jamaica (2010), Mexico (2010), Nicaragua (2010), Panama (2010), Paraguay (2010), Peru (2010), St. Lucia (2010), St. Vincent and the Grenadines (2010), Suriname (2010), Venezuela, RB (2010) Middle East and North Djibouti (2013), Egypt, Arab Rep. (2013), Iraq (2011), Jordan (2013), Lebanon (2013), Morocco (2013), Africa Tunisia (2013), West Bank and Gaza (2013), Yemen, Rep. (2013) South Asia Afghanistan (2014), Bangladesh (2013), Bhutan (2015), India (2014), Nepal (2013), Pakistan (2013), Sri Lanka (2011) Sub-Saharan Africa Angola (2010), Benin (2009), Botswana (2010), Burkina Faso (2009), Burundi (2014), Cameroon (2009), Cabo Verde (2009), Central African Republic (2011), Chad (2009), Congo, Dem. Rep. (2013), Congo, Rep. (2009), Côte d’Ivoire (2009), Eritrea (2009), Ethiopia (2015), Gabon (2009), Gambia, The (2006), Ghana (2013), Guinea (2006), Guinea-Bissau (2006), Kenya (2013), Lesotho (2009), Liberia (2009), Madagascar (2013), Malawi (2014), Mali (2010), Mauritania (2014), Mauritius (2009), Mozambique (2007), Namibia (2014), Niger (2009), Nigeria (2014), Rwanda (2011), Senegal (2014), Sierra Leone (2009), South Africa (2007), South Sudan (2014), Sudan (2014), Swaziland (2006), Tanzania (2013), Togo (2009), Uganda (2013), Zambia (2013), Zimbabwe (2011) Source: World Bank Enterprise Survey. Note: This table presents the economies included in the analysis using the World Bank Enterprise Survey data. The year in which the survey was imple- mented in each country is in parentheses. The information was accessed on September 8, 2016. 64 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 Annex 2B. Measuring Firm the midpoint growth rate, a measure pro- posed by Davis, Haltiwanger, and Schuh Growth (1998) that uses absolute changes relative to When characterizing firm performance, there the average size of the firm across the period are at least three issues that need to be con- of time considered in the study. This measure sidered: the indicator of growth, the measure is formally defined as follows: of growth, and the study period. empi ,t − empi ,t − 2 The indicator of growth refers to the vari- g i ,t = , 1 able over which growth is observed. The most 2 ( empi ,t − 2 + empi ,t ) commonly used indicators in the high-growth firm literature are sales and number of where empi,t refers to total number of perma- employees (Daunfeldt, Elert, and Johansson nent, full-time employees that firm i reports in 2014). Because we are interested in the role of year t. By construction, this growth rate is high-growth businesses in job creation, we symmetric around zero and bounded between use the number of permanent, full-time −2 and 2. It is also monotonically related to employees of the firm as our growth the conventional growth rate measure (Gi,t), indicator. and it approximates the latter for small The number of possible indicators for mea- growth rates. Both growth measures are suring firm-level employment growth is 2 g i ,t linked by the following identity: Gi ,t ≈ ample. The two most basic approaches are the (2 − g i ,t ) absolute and relative changes in the indicator The underlying statistical properties of this of growth. The first one examines the simple growth rate are discussed in detail in difference in employment between two points Törnqvist, Vartia, and Vartia (1985). in time while the second presents this differ- The time period of study of our analysis is ence relative to the initial size of the firm. two years. The surveys ask firms about total These two measures can lead to different employment during the last fiscal year and in results. Almus (2002) and Daunfeldt, Elert, the three previous fiscal years. Three- or four- and Johansson (2014) show that measures of year periods are used in most studies examin- absolute growth are biased toward larger ing high-growth firms, although some studies firms, while measures of relative growth favor have used shorter periods (Coad and others small firms. To reduce these biases, we employ 2014; Reyes, Roberts, and Xu 2017). Annex 2C. Median Size and Age of High-Growth Firms and Rest of Businesses High-growth firms Rest of firms High-growth firms Rest of firms Employment Age Employment Age Employment Age Employment Age Afghanistan 6 7 12 9 Liberia 3 8 6.5 7 Albania 3 10 9 12 Macedonia, 5.5 8 9 16.5 FYR Angola 9 9 15 10 Madagascar 7.5 11 12 12 Argentina 10 15 36 28 Malawi 6 14.5 15 16 Armenia 6.5 8 18 13 Malaysia 13.5 18 32 17 Azerbaijan 10 15 16 12 Mali 4 12 10 10 Bangladesh 20 17 26 18 Mauritania 7 16 19.5 14 Belarus 8 8 17 15 Mauritius 5 5 15 16 Belize 9.5 20 16 15 Mexico 6.5 12 44 20 table continues next page EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 65 High-growth firms Rest of firms High-growth firms Rest of firms Employment Age Employment Age Employment Age Employment Age Benin 3 6 7 14 Micronesia, 2.5 3.5 10 16 Fed. Sts. Bhutan 5.5 7 13 15 Moldova 8 13 15 13 Bolivia 8 15 35 23 Mongolia 10 10.5 15 12 Bosnia and 12.5 13 15 16 Montenegro 7 12 10 15 Herzegovina Botswana 6 7 20 14 Morocco 7.5 15 30 18 Brazil 5 16 25 18 Mozambique 3.5 7 10 12 Bulgaria 5 11 15 17 Myanmar 10 10 11 14 Burkina Faso 8 6 10.5 12 Namibia 3 6 12 9 Burundi 10 4.5 16 12 Nepal 3.5 10.5 12 15 Cabo Verde 4.5 6.5 19.5 13 Nicaragua 6 18 24 19 Cambodia 3.5 14 15 13 Niger 4 6 14 11 Cameroon 10 12 20 16 Nigeria 4 14 9 14 Central 3 12 10 10 Pakistan 10 15 20 20 African Republic Chad 4 11 12 14 Panama 20 18 28.5 17 China 20 10 56 11 Papua New 79.5 41.5 44 25 Guinea Colombia 9 15 30 20 Paraguay 4 7 25 18 Congo, Dem. 4 6 9 9 Peru 9 11 30 16 Rep. Congo, Rep. 2.5 7.5 14 11 Philippines 20 14.5 35 19 Costa Rica 20 12 26.5 21 Romania 5 9 15 17 Côte d’Ivoire 3 6 7.5 9 Rwanda 6 5 16 9 Djibouti 5 10 12 14 Samoa 4 9 12 16 Dominica 3 9 13.5 10 Senegal 3.5 10 10 14 Dominican 5 11 35 17 Serbia 8 11 18 17 Republic Ecuador 12 11 30 22 Sierra Leone 2.5 14.5 10 14 Egypt, Arab 11 13 28 18 Solomon 8.5 5.5 19 18.5 Rep. Islands El Salvador 15 12 35 20 South Africa 6 9 25 15 Eritrea 15 8 16 13 South Sudan 3 5 7 6 Ethiopia 5.5 9 16 12 Sri Lanka 5 13 18 19 Fiji 9 13 15 23 St. Lucia 4.5 9 18 13 Gabon 5 7 10 12 St. Vincent 3 11.5 9 18 and the Grenadines Gambia, The 8 6 9 9 Sudan 10 11 15 11 Georgia 3 4.5 11 10 Suriname 34 17.5 20 18.5 Ghana 2 8 10 13 Swaziland 2 8 10 10 Grenada 2 24 13.5 20 Tajikistan 6.5 9.5 17 10 Guatemala 7 13 32 21 Tanzania 2 15 10 13 Guinea 2 6.5 6 8 Thailand 15 16 27 19 Guinea- 2.5 10.5 7 10 Timor-Leste 6 9 10 11 Bissau Guyana 12.5 17.5 30 19 Togo 3 6 13 11 Honduras 4 17.5 20 20 Tonga 3 4 7 10 Hungary 6.5 11 13 16 Tunisia 10 10.5 35 20 table continues next page 66 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 High-growth firms Rest of firms High-growth firms Rest of firms Employment Age Employment Age Employment Age Employment Age India 15 13 30 16 Turkey 9 10 22 16 Indonesia 20 15 30 19 Uganda 6 10 10 13 Iraq 3 12 9 10 Ukraine 20 12 20 14 Jamaica 10 10 24 20 Uzbekistan 6 7 25 14 Jordan 7.5 9 22 15 Vanuatu 7 6 12 19 Kazakhstan 10 8 17 12 Venezuela, 6 11.5 16 13 RB Kenya 9.5 13.5 20 18 Vietnam 10 8 28 11 Kosovo 5 7 15 13 West Bank 3 11 10 16 and Gaza Kyrgyz 20 10 22.5 15 Yemen, Rep. 9 16 14 20 Republic Lao PDR 5 12.5 13 16 Zambia 7 8 12 12 Lebanon 7 7 19 22 Zimbabwe 13 19 40 31 Lesotho 4 11 15 10 Source: Computation based on data from Enterprise Surveys, the World Bank. Note: This table reports the median employment and age of firms at the beginning of the period under consideration (that is, two years before the implementation of the survey). Annex 2D. Identifying the Role as firms with at least 10 percent foreign ownership. Specifically, this variable is con- of FDI Spillover in High-Growth structed as Firms23 dom 1 input ijc To capture the role of FDI spillovers on the linkages jc = ∑ in=1 tot (2) performance of domestic enterprises, we n input ijc regress measures of linkages and demonstra- dom tion effects on the growth rate of domestic where input ijc represents the value of inputs firms’ output as follows: of domestic origin used by the foreign firm, tot and input ijc corresponds to total value g ijc = β1linkages jc + β 2demonstration jc inputs, regardless of their origin. The total + BXijc + γ c + γ j + ε ijc (1) number of foreign firms in the sector is n. The demonstration channel (demonstra- where the subscript i stands for firm, j for sec- tionjc) is defined by the share of foreign output tor, and c for country. g c represents country as a percentage of total output at the sectoral fixed effects and gj sector fixed effects, intro- level. This measure is standard in the literature duced to the specification in order to account to measure intra-industry spillover effects. See for unobserved heterogeneity within each one Farole and Winkler (2015) and references of these dimensions. Sector fixed effects are therein. defined at the two-digit International Standard Industrial Classification (ISIC) level. gijc is the ∑ i output ijc fgn demonstration jc = (3) sales midpoint growth rate of firm i over the ∑ i output ijc all last two years in which the survey was imple- fgn mented in each country (see annex 2B). where output ijc represents the sales of for- all The linkages channel (linkagesjc) is defined eign firms exclusively, while output ijc by the average share of inputs of domestic accounts for the sales of all firms in each sec- origin that foreign firms in sector j (two-digit tor, country, and year. ISIC codes) acquire in each country. In line The model controls for firm-specific attri- with the literature, foreign firms are identified butes contained in the matrix Xijc, including EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 67 a log transformation of the firm age (defined different impact that these effects have on as the years between the beginning of opera- high-growth firms, we modify equation [1] tions of the firm and the application of the to include a dummy variable indicating if survey), a log transformation of the labor the firm is a high-growth business and inter- productivity (US$ sales per worker), and a act this term with the FDI spillover chan- dummy variable to capture exporter status, nels. A high-growth firm is defined as an taking a value of one if direct exports enterprise located in the top fifth percentile accounted for more than 5 percent of the of the distribution of employment growth in local firm’s total sales. We retained country- each country. The results of these estima- sector cells with presence of foreign firms. tions are presented in table 2D.1. The final sample of the regressions included To examine how FDI spillovers vary across about 33,000 domestic firms in 121 countries, we run the specification separately economies. for six regions of the world, following the The coefficients b 1 and b 2 provide the World Bank Group country classification. We average impact of linkages and demonstra- also separate the sample between manufac- tion effects on domestic firms’ sales growth turing and services sectors. The results are across countries and sectors. To test the presented in tables 2D.1 and 2D.2. TABLE 2D.1 Role of FDI Spillovers on Firm Performance Variables (1) (2) (3) (4) (5) (6) (7) Linkages channel 0.023 0.028 0.053 0.053 (0.051) (0.052) (0.050) (0.051) Demonstration channel 0.010 0.011 0.002 0.006 (0.015) (0.019) (0.019) (0.019) High-growth firm 0.207*** 0.207*** 0.190*** (0.019) (0.021) (0.021) X Linkages channel 0.678*** 0.568*** (0.190) (0.201) X Demonstration 0.171*** 0.109* channel (0.061) (0.063) Log age − 0.068*** − 0.068*** − 0.068*** − 0.068*** − 0.061*** − 0.061*** − 0.061*** (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) Exporter − 0.001 − 0.001 − 0.001 − 0.001 − 0.003 − 0.003 − 0.003 (0.010) (0.010) (0.009) (0.010) (0.010) (0.010) (0.010) Log labor productivity 0.082*** 0.082*** 0.082*** 0.082*** 0.083*** 0.083*** 0.083*** (0.005) (0.005) (0.002) (0.005) (0.005) (0.005) (0.005) Constant − 0.783*** − 0.784*** − 0.785 − 0.786*** − 0.813*** − 0.816*** − 0.816*** (0.059) (0.059) (0.551) (0.059) (0.060) (0.059) (0.059) Country fixed effects Yes Yes Yes Yes Yes Yes Yes Sector fixed effects Yes Yes Yes Yes Yes Yes Yes Observations 33,305 33,305 33,305 33,305 33,305 33,305 33,305 R-squared 0.165 0.165 0.165 0.165 0.174 0.174 0.174 Source: Computation based on data from Enterprise Surveys, the World Bank. Note: Standard errors (in parentheses) are clustered at the country-sector level. FDI = foreign direct investment. ***p<0.01; **p<0.05; *p<0.1. 68 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 TABLE 2D.2 Role of FDI Spillovers on Firm Performance, by Regions and Sectors World Bank regions Economic sectors East Europe Latin America Middle East Sub- Asia and and Central and the and North South Saharan Variables Pacific Asia Caribbean Africa Asia Africa Manufacturing Services Linkages channel − 0.123 0.299 0.009 − 0.092 0.349 − 0.059 0.092 − 0.111* (0.103) (0.265) (0.108) (0.164) (0.647) (0.098) (0.108) (0.064) Demonstration channel 0.047 − 0.051 0.042 − 0.003 − 0.086 − 0.001 − 0.003 − 0.009 (0.050) (0.047) (0.036) (0.043) (0.101) (0.034) (0.023) (0.037) High-growth firm 0.238*** 0.171*** 0.212*** 0.241*** 0.127*** 0.294*** 0.181*** 0.235*** (0.067) (0.065) (0.056) (0.066) (0.025) (0.071) (0.021) (0.061) X Linkages channel 0.814*** − 0.202 0.491 2.342*** 0.764*** − 0.207 0.570* 0.581** (0.309) (0.892) (0.320) (0.700) (0.240) (0.439) (0.307) (0.284) X Demonstration − 0.076 0.427** 0.042 0.019 0.163 0.059 0.078 0.211** (0.118) (0.192) (0.127) (0.208) (0.201) (0.147) (0.077) (0.115) Log age − 0.056*** − 0.086*** − 0.059*** − 0.059*** − 0.045*** − 0.088*** − 0.059*** − 0.069*** (0.015) (0.016) (0.011) (0.012) (0.006) (0.017) (0.006) (0.012) Exporter − 0.000 0.017 − 0.065*** 0.015 − 0.009 0.064 − 0.013 0.083* (0.023) (0.029) (0.019) (0.021) (0.014) (0.045) (0.010) (0.043) Log labor productivity 0.062*** 0.086*** 0.107*** 0.099*** 0.038*** 0.122*** 0.075*** 0.108*** (0.009) (0.011) (0.007) (0.009) (0.006) (0.011) (0.005) (0.012) Constant − 0.579*** − 0.718*** − 0.679*** − 0.983*** 0.062 − 0.641*** − 0.444*** − 0.734*** (0.100) (0.171) (0.089) (0.128) (0.104) (0.131) (0.078) (0.119) Country fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Sector fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Observations 5,876 2,749 5,557 4,086 9,155 5,882 26,398 6,893 R-squared 0.103 0.171 0.116 0.306 0.050 0.184 0.175 0.190 Source: Computation based on data from Enterprise Surveys, the World Bank. Note: Standard errors (in parentheses) are clustered at the country-sector level. FDI = foreign direct investment. ***p<0.01; **p<0.05; *p<0.1. Notes duced by domestic suppliers can increase potential for scale economies. Second, 1. Moran (2011, 2015) provides a comprehens- domestic suppliers may face incentives to ive overview of the challenges countries face improve product quality and increase effi- when using FDI to reach these policy goals. ciency, owing to more stringent requirements 2. Alfaro and Chen (forthcoming) provide from the foreign firms. Third, competition empirical evidence on the positive impact of for other local firms for foreign consumers FDI spillovers and the reallocation of fac- may also spur productivity upgrading. The tors on aggregate productivity using a rich analysis in this chapter focuses on the knowl- cross-country database. edge diffusion impact of linkages. 3. See Alfaro and others (2006), Alfaro and 5. Some studies such as Morrissey, López, and Chen (forthcoming), Lipsey (2004), Barba Sharma (2015), separate the learning process Navaretti and Venables (2004), and Alfaro from observation from the labor turnover and Rodriguez-Claire (2004) for an overview effect. Given data limitations, this chap- of the empirical literature about the channels ter compounds these two channels into the of FDI spillovers. demonstration effect. 4. Linkages can increase the productivity of 6. The dataset covers a broad range of busi- domestic firms in at least three other ways: ness environment topics including access to First, greater demand for intermediates pro- finance, corruption, infrastructure, crime, EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 69 competition, and firm-level performance of identifying high-growth businesses. The measures. The raw data include information Organisation for Economic Co-operation for various waves of surveys for 139 coun- and Development (OECD) defines them tries. This analysis retains the latest survey as firms with 10 or more employees that conducted in each country and economies have an average annualized growth higher classified as low- and middle-income coun- than 20 percent for three consecutive years tries by the World Bank Group. In total, (Ahmad 2008; OECD 2008, 2010). But this information for about 63,000 firms in 121 definition is overly restrictive for developing developing economies is analyzed. countries where 95 percent of businesses have 7. These findings are in line with Damijan and nine or fewer workers (McKenzie 2017). others (2013), which employs 10 transition 12. This threshold changes in every country. On economies to find positive effects of hori- average, across countries in the database, zontal spillovers only on large and high- firms need to double the number of employ- productivity domestic enterprises. ees in two years to be considered high-growth 8. A growing body of literature aims to under- firms (Reyes, forthcoming). stand the conditions under which the benefits 13. The fact that high-growth firms are an impor- of FDI materialize at the firm level. Some tant source of job creation and tend to be firm characteristics have been linked to their young is a well-established empirical fact in absorptive capacity, including elements such the literature (Coad and others 2014). When as the size of their technology gap (Wang and firm growth is computed in relative terms—as Blomström 1992), their share of skilled labor in this analysis—small firms are also overrep- (Blalock and Gertler 2009), and their size resented among high-growth firms (Delmar, (Meyer and Sinani 2004). Davidsson, and Gartner 2003). 9. Although the Enterprise Surveys allow the 14. This finding is in line with Henrekson and systematic study of firm performance across Johansson (2010), who find that high-growth a broad range of developing countries, some businesses exist in all industries but tend to important caveats are in order. First, firm be overrepresented in services. performance outcomes are available for just 15. Jiménez-Barrionuevo, García-Morales, and two points in time, separated by only two Molina (2011) propose a scale of 18 items years. Second, the surveys are representative to measure the absorptive capacity of firms. only of the broad manufacturing and services They are grouped under four categories: sectors, not at the detailed two-digit ISIC They are grouped under four categories: codes. Third, the data include only firms that acquisition (interaction, trust, friendship, and survived between the two points of time, not reciprocity); assimilation (common language, those that exited. Fourth, there may be some complementarity, similarity, and organization differences across countries in the minimum culture and management style); transforma- size of firms included in the surveys. tion (communications, meetings, documents, 10. Because standard growth rates are relative transmission, time, and flows); and exploita- to the initial size of the firm and, therefore, tion (responsibility and application). biased toward smaller firms, the analysis 16. This approach on backward linkages, which uses midpoint growth rates representing the focuses on the demand for inputs from for- change in employment relative to the average eign companies, is also used in Sánchez- size of the firm between the fiscal year before Martín, De Piniés, and Antoine (2015) and the survey was administered and three fiscal complements that in Javorcik (2004) and years prior. Annex 2B discusses the character- Blalock and Gertler (2008), who adopt the ization of firm growth adopted in this study. perspective of the local supplying sector and 11. While this methodology is based on pre- look for foreign presence downstream in the vious literature, there is no general agree- supply chain. Forward linkages, which focus ment on the definition of high-growth on the relationship with upstream sectors, firms. Growth rate thresholds have been can also be important, particularly in the ser- employed by Schreyer (2000) and Davidsson vices sector. Hoekman and Shepherd (2017) and Henrekson (2002), among others. find strong impacts of services efficiency and Henrekson and Johansson (2010) provide the productivity of downstream manufactur- a meta-analysis of the empirical literature ing firms. 70 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 17 . Owing to limitations with the level of sectoral Almus, M. 2002. “What Characterizes a Fast- disaggregation of the World Bank’s Enterprise Growing Firm?” Applied Economics 34 (12): Surveys data, the channels for FDI spillovers 1497–1508. are defined at a broader sectoral classification Arráiz, I., F. Henríquez, and R. Stucchi. 2013. (two-digit ISIC codes). Consequently, in addi- “Supplier Development Programs and Firm tion to horizontal spillovers, the measures are Performance: Evidence from Chile.” Small likely to capture some vertical spillovers. For Business Economics 41 (1): 277–93. example, manufacture of leather and related Barba Navaretti, G., and A. Venables. 2004. products (classified under ISIC 15) includes Multinational Firms in the World Economy. both final footwear and the tanning and Princeton, NJ: Princeton University Press. dressing of leather—an input for footwear. Blalock, G., and P. J. Gertler. 2008. “Welfare Thus, FDI in this sector could affect domestic Gains from Foreign Direct Investment through final producers of footwear as well as domes- Technology Transfer to Local Suppliers.” tic suppliers of footwear production. Journal of International Economics 74 (2): 18. The finding that intra-industry spillover 402–21. effects are rarely accrued by domestic firms is ———. 2009. “How Firm Capabilities Affect Who standard in the literature. Meyer and Sinani Benefits from Foreign Technology.” Journal of (2009) and Görg and Strobl (2001) provide Development Economics 90 (2): 192–99. two meta-analyses reviewing this literature. Coad, A., S. Daunfeldt, W. Hölzl, D. Johansson, 19. This effect is statistically significant at the 1 and P. Nightingale. 2014. “High-Growth Firms: percent level. Introduction to the Special Section.” Industrial 20. This effect is statistically significant at the 10 and Corporate Change 23 (1): 91–112. percent level. Damijan, J. P., M. Rojec, B. Majcen, and M. Knell. 21. These findings are robust to 80 percent and 2013. “Impact of Firm Heterogeneity on Direct 90 percent thresholds to identify high-growth and Spillover Effects of FDI: Micro-Evidence firms. See Reyes (forthcoming). from Ten Transition Countries.” Journal of 22. The argument that medium-sized and large Comparative Economics 41 (3): 895–922. indigenous firms are usually better candi- Daunfeldt, S., N. Elert, and D. Johansson. 2014. dates to qualify as suppliers of MNCs is also “The Economic Contribution of High-Growth made by Freund and Moran (2017). Firms: Do Policy Implications Depend on 23. For a discussion of the identification strategy the Choice of Growth Indicator?” Journal of see Reyes (forthcoming). Industry, Competition and Trade 14 (3): 337–65. Davis, S. J., J. C. Haltiwanger, and S. Schuh. 1998. Job Creation and Destruction. MIT Press Books. Bibliography Davidsson, P., and M. Henrekson. 2002. “Determinants of the Prevalence of Start-Ups Ahmad, N. 2008. “A Proposed Framework and High-Growth Firms.” Small Business for Business Demography Statistics.” In Economics 19 (2): 81–104. Measuring Entrepreneurship, edited by Emilio Delmar, F., P. Davidsson, and W. B. Gartner. 2003. Congregado, 113–74. Boston, MA: Springer. “Arriving at the High-Growth Firm.” Journal Alfaro, L., A. Chanda, S. Kalemli-Ozcan, and of Business Venturing 18 (2): 189–216. S. Sayek. 2006. “How Does Foreign Direct Dollar, D. 1992. “Outward-Oriented Developing Investment Promote Economic Growth? Economies Really Do Grow More Rapidly: Exploring the Effects of Financial Markets Evidence from 95 LDCs, 1976–1985.” on Linkages.” NBER Working Paper 12522, Economic Development and Cultural Change National Bureau of Economic Research, 40 (3): 523–44. Cambridge, MA. Echandi, R., J. Krajcovicova, and C. Z. W. Qiang. Alfaro, L., and A. Rodriguez-Clare. 2004. 2015. “The Impact of Investment Policy in a “Multinationals and Linkages: Evidence from Changing Global Economy: A Review of the Latin America.” Economia 4: 113–70. Literature.” Policy Research Working Paper Alfaro, L., and M. X. Chen. Forthcoming. “Selection 7437, World Bank, Washington, DC. and Market Reallocation: Productivity Gains Farole, T., and D. Winkler. 2015. “The Role from Multinational Production.” American of Foreign Firm Characteristics, Absorptive Economic Journal: Economic Policy. Capacity and the Institutional Framework EFFECTS OF FDI ON HIGH-GROWTH FIRMS IN DEVELOPING COUNTRIES 71 for FDI Spillovers.” Journal of Banking and Instrument to Measure Absorptive Capacity.” Financial Economics 1 (3): 77–112. Technovation 31 (5): 190–202. Fons-Rosen C., S. Kalemli-Ozcan, B. T. Sorensen, Lipsey, R. E. 2004. “Home- and Host-Country C. Villegas-Sanchez, and V. Volosovych. Effects of Foreign Direct Investment.” In 2017. “Foreign Investment and Domestic Challenges to Globalization: Analyzing the Productivity: Identifying Knowledge Spillovers Economics, edited by Robert E. Baldwin and and Competition Effects.” NBER Working L. Alan Winters, 333–82. University of Chicago Paper 22643, National Bureau of Economic Press. Research, Cambridge, MA, August. McKenzie, D. 2017. “Identifying and Spurring Frankel, J. A., and D. Romer. 1999. “Does Trade High-Growth Entrepreneurship: Experimental Cause Growth?” American Economic Review Evidence from a Business Plan Competition.” 89 (3): 379–99. American Economic Review 107 (8): 2278–2307 Freund, C., and T. Moran. 2017. “Multinational Meyer, K. E., and E. Sinani. 2004. “Spillovers of Investors as Export Superstars: How Technology Transfer from FDI: The Case of Emerging-Market Governments Can Reshape Estonia.” Journal of Comparative Economics Comparative Advantage.” Working Paper 17-1, 32 (3): 445–66. Peterson Institute for International Economics, ———. 2009. “When and Where Does Foreign Washington, DC. Direct Investment Generate Positive Spillovers? Görg, H., and E. Strobl. 2001. “Multinational A Meta-Analysis.” Journal of International Companies and Productivity Spillovers: A Business Studies 40 (7): 1075–94. Meta Analysis.” The Economic Journal 111 Monge-González, R., and J. A. Rodríguez-Álvarez. (475): 723–39. 2013. “Impact Evaluation of Innovation and Haltiwanger, J., R. S. Jarmin, and J. Miranda. Linkage Development Programs in Costa Rica: 2013. “Who Creates Jobs? Small versus Large The Cases of PROPYME and CR Provee.” versus Young.” Review of Economics and IDB Working Paper Series IDB-WP-461, Inter- Statistics 95 (2): 347–61. American Development Bank, Washington, DC. Hampel-Milagrosa, A., M. Loewe, and C. Reeg. Moran, T. 2011. Foreign Direct Investment and 2015. “The Entrepreneur Makes a Difference: Development: Launching a Second Generation Evidence on MSE Upgrading Factors from of Policy Research: Avoiding the Mistakes of the Egypt, India, and the Philippines.” World First, Reevaluating Policies for Developed and Development 66: 118–30. Developing Countries. Columbia University Press. Harrison, A. 1996. “Openness and Growth: Moran, T. H. 2015. “The Role of Industrial Policy A Time-Series, Cross-Country Analysis as a Development Tool: New Evidence from the for Developing Countries.” Journal of Globalization of Trade-and-Investment.” Policy Development Economics 48 (2): 419–47. Paper 071, Center for Global Development, Henrekson, M., and D. Johansson. 2010. “Gazelles Washington, DC. as Job Creators: A Survey and Interpretation of Morrissey, O., R. López, and K. Sharma, eds. the Evidence.” Small Business Economics 35 2015. Handbook on Trade and Development. (2): 227–44. Cheltenham, U.K.: Edward Elgar Publishing. Hsieh, C., and P. J. Klenow. 2014. “The Life Cycle OECD (Organisation for Economic Co-Operation of Plants in India and Mexico.” The Quarterly and Development). 2008. OECD Journal of Economics 129 (3): 1035–84. Framework for the Evaluation of SME and Hoekman, B., and B. Shepherd. 2017. “Services Entrepreneurship Policies and Programmes. Productivity, Trade Policy and Manufacturing Paris: OECD Publishing. Exports.” The World Economy 40 (3): 499–516. ———. 2010. What Governments Can Do to Hufbauer, G. C., J. Schott, and C. Cimino-Isaacs. Make a Difference. OECD Studies on SMEs and 2013. Local Content Requirements: A Global Entrepreneurship High-Growth Enterprises. Problem. Columbia University Press. Paris: OECD Publishing. Javorcik, B. 2004. “Does Foreign Direct Investment Reyes, J-D. Forthcoming. “FDI Spillovers and Increase the Productivity of Domestic Firms? In High-Growth Firms in Developing Countries.” Search of Spillovers through Backward Linkages.” Mimeo, World Bank, Washington, DC. The American Economic Review 94 (3): 605–27. Reyes, J-D., M. Roberts, and L. C. Xu. 2017. Jiménez-Barrionuevo, M. M., V. J. García-Morales, “The Heterogeneous Growth Effects of the and L. M. Molina. 2011. “Validation of an Business Environment: Firm-Level Evidence 72 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 for a Global Sample of Cities.” Policy Industry Working Paper 2000/3, Organisation Research Working Paper 8114, World Bank, for Economic Co-operation and Development, Washington, DC. Paris. Sánchez-Martín, M. E., J. De Piniés, and K. Törnqvist, Leo, Pentti Vartia, and Yrjö O. Vartia. Antoine. 2015 “Measuring the Determinants 1985. “How Should Relative Changes Be of Backward Linkages from FDI in Developing Measured?” The American Statistician 39 (1): Economies: Is It a Matter of Size?” Policy 43–46. Research Working Paper 7185, World Bank, Wang, J., and M. Blomström. 1992. “Foreign Washington, DC. Investment and Technology Transfer. A Simple Schreyer, P. 2000. “High-Growth Firms and Model.” European Economic Review 36 (1): Employment.” OECD Science, Technology and 137–55. 3 Corporate Tax Incentives and FDI in Developing Countries Maria R. Andersen, Benjamin R. Kett, and Erik von Uexkull P olicy makers in developing countries show how to target incentives more effi- often find themselves in a dilemma ciently, based on a new dataset on tax incen- over the use of tax incentives to attract tives in developing countries compiled by the foreign direct investment (FDI). They would World Bank Group. The analysis considers likely prefer that no country offer tax incen- whether and how developing countries use tives and that all firms contribute equitably tax incentives by sector and over time, links to public coffers. But given that most other the effectiveness of incentives to a simple countries—including high-income ones— framework of investor motivation, and pres- offer incentives, investment promotion prac- ents new evidence on the relevance of tax titioners often feel obliged to match, or even incentives for investors. The chapter also surpass, the competition to attract FDI. 1 reviews priorities for design, transparency, Binding international coordination could and administration reforms of incentives resolve this dilemma, but such a solution regimes. does not appear to be on the horizon. Tax incentives are more effective in Although efforts to increase international attracting efficiency-seeking FDI motivated coordination are under way at both the by lowering production costs than for other regional and global levels,2 and countries are types of investment. Yet many developing well advised to continue these, the process is countries offer incentives to all investors, slow and often leaves gaps.3 In the mean- including those motivated by access to natu- time, developing countries continue to make ral resources or the domestic market, who heavy and increasing use of tax incentives. are less likely to respond to incentives. While While general principles for incentives some developing countries target their incen- reform are well documented, this chapter tives at efficiency-seeking FDI, many also contributes practical evidence to help devel- offer incentives to market- and natural oping country policy makers design and resource–seeking FDI. In most cases, this is implement reforms to make their incentives not because incentives are deliberately target- regimes more effective for FDI attraction. It ing these investors but rather because they provides sector- and firm-level evidence to are offered indiscriminately. At the same 73 74 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 time, efficiency-seeking FDI also requires that the design, transparency, and administration host countries have a more favorable overall of incentives can help reduce unintended investment climate than natural resource– or effects and costs, such as economic distor- market-seeking FDI. Incentives do not com- tions, red tape, and corruption. Although pensate for such shortcomings and are likely these policy reforms do not obviate the need to succeed only if they are part of a broader for regional and global solutions, they can strategy to address investment climate substantially improve the cost–benefit ratio of constraints. incentives. Tax incentive regimes in developing coun- tries often suffer from weak design, lack of transparency, and cumbersome administra- Developing Countries Make Wide tion. Tax holidays and preferential tax rates remain by far the most widely used incentive Use of Tax Incentives instruments in developing countries, despite A “Developing Country Tax Incentives their well-documented shortcomings. Lack of Database”4 compiled for this report provides transparency and high administrative costs data on the use of tax incentives in the devel- also diminish the attractiveness of incentives oping world. Information on tax incentives and raise their indirect costs in terms of is often freely available to the public, in par- economic distortions and potential for ticular through the tax summaries published corruption. by global accounting firms. In many cases, Even in the short run, developing countries information is also available from a coun- can undertake unilateral reforms to make tax try’s investment promotion agency (IPA), but incentives better targeted and more cost- this information is typically provided in efficient. By focusing incentives on those types qualitative form and does not lend itself to of investors most likely to respond, develop- quantitative research. The new tax incentives ing countries can reduce the unnecessary loss database compiled for this report quantifies of tax revenue resulting from incentives information from publicly available sources granted to firms that would have invested on a number of frequently used incentive anyway. At the same time, reforms to improve instruments (box 3.1). BOX 3.1 The Developing Country Tax Incentives Database The Developing Country Tax Incentives Database investment expenses from taxable income or credit provides information on 107 countries for the period them against payable taxes. Information on the 2009–15 (table 3A.1). Data are broken down by 22 eco- magnitude of these instruments was not collected nomic sectors to the extent that incentives explicitly target owing to methodological challenges. a specific sector. The following information is covered: The database also contains information on three con- ditions for receiving incentives, tracked by type of • The standard corporate income tax (CIT) rate. incentive and by sector: • The availability and maximum duration of tax holidays. • Investment location, including requirements for • The availability and level of preferential rates below establishment in a certain region of the country or a the standard CIT rate for a specific sector or type of special economic zone (SEZ). investment. • Company exporting status, including requirements • The availability of investment tax allowances to sell a certain share of output to other exporting or credits that grant investors the right to deduct companies. box continues next page CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 75 BOX 3.1 The Developing Country Tax Incentives Database (continued) • Other conditions, such as requirements to under- it did not corroborate the tax and incentives infor- take research and development (R&D) or incen- mation reported by the sources mentioned above. tives specific to income from intellectual property. In addition, many countries provide tax incentives at the subnational level and these are not covered The data were collected through desk research of pub- by the data sources used. Moreover, some countries lic sources for country-level tax information in July negotiate ad hoc tax incentives and other discretion- and August of 2016. As a default, Ernst and Young’s ary deals with potential investors, and these are also “Global Tax Guides” and PricewaterhouseCoopers’ not captured by the database. Finally, the database “Worldwide Tax Summaries” for the years 2009–15 focuses on corporate tax incentives, excluding infor- were consulted and compared. In cases of missing mation on incentives through indirect taxes such as information or discrepancies, other publicly available customs duties and VAT exemptions, or other types data sources were consulted, such as the website of a of incentives such as subsidies or regulatory advan- country’s investment promotion agency (IPA) or rel- tages. Many countries make incentives available to evant country reports. both domestic and foreign investors. The database A few caveats bear mention: While the World Bank registers all such incentives, unless foreign investors Group made signifi cant efforts to ensure accuracy, are explicitly excluded. While tax incentives are common in most widely used instrument (table 3A.2). developing countries, they vary at the sector, More than half of the developing countries in regional, and income levels. Across sectors, the database offer tax holidays in at least one 49–72 percent of all developing countries offer sector. Across regions, the highest incidence tax holidays, preferential or very low general of tax holidays is in construction and manu- tax rates, or tax allowances. Tax incentives are facturing sectors, where up to 46 percent of most common for construction, information developing countries use them. Their applica- technology (IT) and electronics, machinery tion is less common in services and natural and equipment, and other manufacturing sec- resource sectors, with retail showing the low- tors. The share of countries offering incentives est use (23 percent). The median duration in services sectors is lower but the majority do of tax holidays across regions and sectors is offer incentives for most services sectors. 10 years. Some developing countries deliberately tar- Most developing countries that grant tax get incentives to manufacturing sectors and holidays condition them on location require- construction to attract investors, but most ments within the country (77 percent), which apply incentives across the board. While about mostly consist of either special economic 30 percent of developing countries have incen- zone (SEZ) locations or requirements to tives that specifically target certain manufac- establish in a designated region of the turing sectors (figure 3.1, blue bar), targeting is country. Thirty percent of developing coun- less common for services and natural resource tries also condition tax holidays on a require- sectors. Forty percent of developing countries ment to export or sell to exporting firms, have incentive systems that grant either incen- which raises concerns about compliance tives or low general corporate income tax with World Trade Organization (WTO) (CIT) rates across all or most sectors. rules.5 Forty percent of developing countries Countries deliver tax incentives through have additional requirements in place, such a number of different instruments. Among as spending on research and development developing countries, tax holidays are the (R&D). 76 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 3.1 Tax Incentives Are Widespread in Developing Countries, Especially in Construction and Manufacturing Construction and building materials IT and electronics Machinery and equipment Air- and spacecraft Apparel, textiles, and footwear Automotive industry and other transport equipment Food and beverages Biotechnology, pharmaceuticals, and medical products Other manufacturing Agriculture and fishing Transport and logistics services Extractive industries Tourism and hospitality Education and health Renewable energy Power, utilities, and telecommunications Entertainment IT services Financial services Business services Trade and retail 0 10 20 30 40 50 60 70 80 Share of countries offering incentives or CIT rates lower than 15 percent by sector (percent) Share of countries offering any incentives applicable to this sector or a general CIT rate below 15% Share of countries offering targeted incentives specifically to this sector Source: Developing Country Tax Incentives Database. Note: Incentives specifically targeted to a sector, shown by the dark blue bar, are those applicable to less than 15 out of 22 sectors. This is to account for a small share of incentives generally available across the board, but excluding a limited number of sectors. Some countries also exclude specific sectors from overall low CIT rates. Such exclusions explain the slight variation in the green graph, showing the share of countries offering general incentives applicable to this sector or a general CIT rate below 15 percent. CIT = corporate income tax; IT = information technology. Preferential tax rates below the standard against payable taxes are much less common CIT rate for specific sectors or investors are in developing countries; just 16 percent of also common, with 40 percent of countries in countries offer them in at least one sector the database offering them for at least one (table 3A.4). Tax allowances and credits also sector (table 3A.3). The median preferential mainly target the manufacturing sector. margin6 is 13 percentage points. Conditions Almost all tax allowances and credits come on location (45 percent), exporting (32 per- with conditions, which is consistent with the cent), and other investment project character- performance-based character of this instru- istics (46 percent) are also common albeit ment. Receiving the allowance or credit is with significant regional variation. As with typically linked to making specific investments, tax holidays, preferential rates are most such as R&D or the purchase and installation widely used in the manufacturing sector (led of new machinery or technology. by food and beverages) and IT and electron- Profit-based incentives, such as tax holi- ics, where 31 percent of developing countries days and preferential rates, have serious limi- offer preferential tax rates. tations. They lower the tax rate for any Tax allowances and credits that grant amount of profit earned by the firm, including investors the right to deduct investment setting the tax rate to zero for a limited period expenses from taxable income or credit them during a tax holiday. The value of the CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 77 incentive for such an instrument is thus a is undertaken. Such instruments have various direct function of the company’s profits. As a advantages: They do not suffer from the bias of result, the incentive heavily favors firms with profit-based incentives in favor of highly profit- high profits, which least need government able firms and are thus less likely to be biased support. This can lead to high redundancy of toward firms that would have invested anyway. expenditure on incentives since an investor They are also less prone to abuse through profit anticipating high profits would likely have shifting, and their magnitude is directly linked proceeded anyway. Also, host governments to the policy outcome on which they are face the risk of losing substantial revenue conditioned. Still, only a few developing coun- when a firm earns extraordinary profits in a tries currently use these more advanced instru- given year. The risk of tax evasion through ments in granting corporate tax incentives. Part profit shifting is high for profit-based incen- of the reason may be insufficient tax adminis- tives as firms can artificially allocate profits tration capacity. Table 3.1 provides a more within the firm to a plant or subsidiary enjoy- detailed overview of the respective strengths ing preferential tax treatment (UNCTAD and weaknesses of these instruments. 2015). The widespread use of these incentive Policy makers have continued to reduce CIT instruments in developing countries is a sig- rates across developing countries. In the Middle nificant shortcoming in the design of tax East and North Africa, East Asia and Pacific, incentives. Latin America and the Caribbean, Sub-Saharan Cost-based instruments, such as tax allow- Africa, and South Asia, average CIT rates fell ances and credits, offer superior design fea- between 2009 and 2015; in contrast, Europe tures. Unlike profit-based incentives, cost-based and Central Asia showed a small increase ones lower the cost of a specific input or pro- in average CIT rates (figure 3.2). Variation in duction factor. In the case of investment allow- average tax rates across regions is substantial, ances or credits, the government may grant a ranging from 38 percent in South Asia to firm the right to deduct a certain share of the 15 percent in Europe and Central Asia. investment value from its taxable income. The At the same time, developing countries magnitude of the benefit to the company is also continued to implement new tax incen- independent of its profit level and instead tives and to make existing ones more gener- depends on the size of the investment that ous. More specifically, 46 percent of countries TABLE 3.1 Pros and Cons of Various Tax Incentives Instruments Profit-based instruments • Tax holidays: Time-bound exemption of new firms or investments from taxes (typically CIT) • Concessionary/preferential tax rates: Reduced tax rates that act as a partial exemption of the standard CIT rate Pros Cons • Strong signaling effect to investors, easy to • Disproportionately favors investments with high profit communicate and advertise. margins that would have likely occurred anyway and investment with short time horizons (in the case of time- bound holidays and concessions). • Typically granted against up-front assurances from the investor rather than actual performance in terms of expected outcomes such as investment or jobs generated. • Prone to abuse through profit shifting within firms. • High fiscal risk owing to little predictability of actual fiscal cost. • Tax holidays only: Investors may appreciate • Tax holidays only: Liberating investors from tax filing complete liberation from interaction with tax requirements makes it impossible to monitor costs of authorities for the duration of the holiday. incentives in terms of forgone revenue. table continues next page 78 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 TABLE 3.1 Pros and Cons of Various Tax Incentives Instruments (continued) Cost-based instruments • Tax allowance: Deduction of a share of the cost of investment from taxable income. • Tax credit: Deduction of a share of the cost of investment from taxes owed. • Accelerated depreciation: Depreciation of fixed assets for tax purposes at a faster schedule than what is normally applied. Pros Cons • Amount of benefit to investor is directly linked to • More challenging to administer. amount invested. • May bias production technology toward more capital- • Tax revenue loss is more predictable than under intensive investment. profit-based instruments. • Less prone to abuse through profit shifting than profit-based instruments. • Does not liberate firms from filing taxes, which makes the process more transparent and allows tracking of costs in terms of foregone revenue. • Accelerated depreciation only: Nominal tax burden is not actually reduced, but payment is merely deferred to a later stage of the investment. introduced new tax incentives or increased abolished tax incentives or made them less the generosity of existing ones in at least one generous in at least one sector over the same sector during the period covered by the data- period (figure 3.3 and table 3A.5). set (2009–15). At the median, developing In the Middle East and North Africa, the countries that made incentives more generous shares of countries introducing new tax or introduced new ones expanded tax holi- incentives and of countries abolishing exist- days by seven years or dropped concessionary ing ones during this period are high—at tax rates by five percentage points. In con- 50 percent each—reflecting reforms under- trast, only 24 percent of developing countries taken in both directions. The strongest growth in incentives was in Sub-Saharan FIGURE 3.2 Policy Makers Continue Cutting Corporate Income Tax Africa, where 65 percent of countries intro- (CIT) Rates in Most Regions duced new or more generous incentives, while only 21 percent removed existing Sub-Saharan Africa 29 incentives or made them less generous. South 27 Asia is the only region in which more coun- South Asia 38 tries reduced the use of tax incentives relative 38 to countries that increased them. Middle East and North Africa 24 These trends in CIT rates and changes 22 in incentives are consistent with a global Latin America and the Caribbean 29 pattern of lower taxation of geographically 27 mobile capital, as governments around the Europe and Central Asia 14 world strive to attract investment and jobs 15 (Klemm and Van Parys 2012; OECD 1998). East Asia and Pacific 26 This underscores the risk of tax competition 23 when a country that introduces lower taxes 10 15 20 25 30 35 40 or new incentives triggers a similar action Average CIT rates by region, 2009–15 (percent) by a competing country. Such retaliation 2009 2015 diminishes the intended effect of incentives Source: Developing Country Tax Incentives Database. to attract more FDI and also reduces both CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 79 FIGURE 3.3 Nearly Half of Developing Countries Have Introduced New Tax Incentives or Increased the Generosity of Existing Ones Share of countries with changes in use of tax incentives, 2009–15 (percent) East Asia and Pacific 36 14 Europe and Central Asia 39 22 Latin America and the Caribbean 35 22 50 Middle East and North Africa 50 South Asia 17 50 65 Sub-Saharan Africa 21 Low Income 44 22 Lower-middle-income 43 16 Upper-middle-income 48 31 Developing countries average 46 24 0 10 20 30 40 50 60 70 Share of countries that made incentives more generous in at least one sector Share of countriesthat made incentives less generous in at least one sector Source: Developing Country Tax Incentives Database. Note: Making a tax incentive more generous refers to either extending the maximum duration of a tax holiday or reducing the preferential tax rate offered. countries’ fiscal revenues. Global and regional to be as high as 5.9 (Cambodia), 5.2 approaches to reducing harmful tax competi- (Ghana), and 3.9 (Dominican Republic) tion are thus warranted to reach a sustainable percent of GDP. Such expenditure through equilibrium of corporate taxation. forgone revenue often does not undergo the same scrutiny and public control as regular government spending, and in Tax Incentives Are Generally Not many developing countries tax expendi- Cost-Effective ture is not even systematically measured or published. Tax incentives impose significant costs on the • Rent-seeking by fi rms engaging in non- countries using them, though these costs are productive behavior to obtain an incen- not always easily visible: tive, or outright corruption where deci- • Fiscal losses resulting from the non- sion makers are bribed to grant incentives collection of taxes that would otherwise (James 2009). Such costs are often ampli- be due, also referred to as tax expenditure. fied by a lack of transparency in the Such expenditure can be very significant, design and administration of incentives. especially in developing countries. While • Tax planning and evasion by the private data limitations are often severe, recent sector, for example, through shifting of World Bank technical assistance has esti- profits from nonexempted to exempted mated tax expenditures from incentives affi liates in the same fi rm by manipulat- 80 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 ing internal transfer prices (Heckemeyer incentives, which is consistent with previous and Overesch 2013; UNCTAD 2015). survey results on the subject (UNIDO 2011). • Administrative costs for both fi rms and Nonetheless, incentives often play a role in the government due to cumbersome pro- the final stage of negotiations between inves- cedures for granting and monitoring tors and governments of the shortlisted incentives. investment locations (Freund and Moran • Economic distortions resulting from real- 2017). One reason countries offer incentives locating resources to activities benefitting is precisely because they can make a differ- from incentives, including a “status quo ence among similar countries on the inves- bias,” in that already-established fi rms or tor’s shortlist. Incentives, by themselves, will sectors tend to be more successful than not get a country on the list. But when sev- newcomers in lobbying to extend incen- eral countries are on the shortlist, with simi- tives (Zolt 2013). lar conditions, incentives can be decisive. In • Retaliation against new or more gener- other words, the effectiveness of incentives is ous incentives by competing investment likely conditional upon other factors that locations (Klemm and Van Parys 2012; determine whether a country “makes the OECD 1998). shortlist” in the first place. This underscores the importance of taking Evidence on the benefits of incentives for a closer look at investor motivation and firm FDI attraction is mixed and that for develop- and country characteristics to understand the ing countries is particularly limited. While effectiveness of tax incentives for FDI promo- high corporate tax rates clearly have a nega- tion. Even where incentives are able to influ- tive effect on FDI entry (Bénassy-Quéré, ence an investor’s location decision, the Fontagne, and Lahreche-Revil 2005; Bellak, benefits do not always justify the costs. Rather Leibrecht, and Damijan 2009; Desai, Fritz than judging the success of an incentive by the Foley, and Hines 2006; Djankoff and others absolute amount of FDI it has attracted, 2010; Egger and others 2008; Hebous, Ruf, countries should weigh the benefits of this and Weichenrieder 2010; Overesch and FDI in terms of its contribution to such devel- Wamser 2008), evidence on the impact of tax opment outcomes as job creation, technology incentives is much more mixed. Several studies transfer, or other positive externalities, against (Allen and others 2001; James 2009; James the above described costs. and van Parys 2010; Klemm and van Parys 2012; van Parys 2012) find them to be of lim- ited effectiveness at the aggregate level. But the The Effectiveness of Incentives research base for a targeted approach to incen- tives in developing countries remains small as Varies by FDI Motivation most existing studies focus on OECD coun- Not all FDI is the same; it differs, among tries and often do not allow sector- or investor- other things, in terms of the motivation of the type-specific conclusions on the effectiveness investor (see box 1.2 in chapter 1). Investor of incentives. motivation is difficult to observe in available Incentives are rarely among the top char- global FDI data, and a one-to-one categoriza- acteristics that multinational corporations tion of sectors by FDI motivation is not pos- (MNCs) initially consider in their location sible. In fact, FDI in the same sector can be decisions, but they can play an important driven by different motives across countries or role in the final decision among shortlisted even within the same country.7 But for illus- locations. The Global Investment trative purposes, a basic distinction between Competitiveness (GIC) survey results in the predominantly market-seeking relative to effi- first chapter confirm that such variables as ciency-seeking FDI sectors can be made on the political stability, regulatory quality, and basis of the share of revenue that is derived market size are generally considered more from exports versus domestic sales. The third important by investors than tax rates and type of motivation—natural resource–seeking CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 81 FDI—can be broadly identified with the well-informed decision making on the tar- extractive and agricultural sectors. geting of incentives, this analysis must be Table 3.2 shows the underlying data and conducted more thoroughly according to approach for this approximate classifica- firm-level data on the activities of foreign tion for the purposes of this chapter. For affiliates in a specific host country. FDI that TABLE 3.2 Efficiency-Seeking FDI Is Clustered in Few Locations While Natural Resource– and Market-Seeking FDI Is Geographically Dispersed FDI in developing countries by sector and likely primary FDI motivation Within Herfindahl-Hirschman Export firm sales Number Index of geographic share (percentage of FDI concentration of FDI (percentage of total sales projects in projectsc of total to affiliated developing sales) of parties) of countries in All U.S. foreign U.S. foreign fDi Markets developing Excl. China Sector affiliatesa affiliatesb database countries and India Mainly natural resource–seeking Agriculture and fishing 36 47 555 0.05 0.03 Extractive industries 58 31 1,112 0.03 0.03 Renewable energy n.a. n.a. 45 0.05 0.05 Total mainly natural resource–seeking 47 39 1,712 0.04 0.04 Mainly market-seeking Business services n.a. 19 3,690 0.07 0.04 Construction and building materials 11 8 1,840 0.07 0.03 Education and health n.a. n.a. 546 0.11 0.03 Entertainment n.a. n.a. 179 0.07 0.04 Financial services 1 0 4,082 0.04 0.02 Food and beverages 28 27 1,150 0.06 0.04 Power, utilities, and telecommunications n.a. n.a. 1,878 0.04 0.03 Tourism and hospitality n.a. n.a. 872 0.08 0.03 Trade and retail 13 5 3,902 0.07 0.05 Total mainly market-seeking 13 12 18,139 0.07 0.04 Mainly efficiency-seeking Air- and spacecraft n.a. n.a. 371 0.12 0.13 Apparel, textiles, and footwear 52 24 544 0.07 0.07 Automotive industry and other transport 50 46 2,867 0.12 0.10 Biotechnology, pharmaceuticals, and medical products 43 48 640 0.11 0.04 IT and electronics 60 45 2,167 0.13 0.06 IT services n.a. 33 3,275 0.10 0.07 Machinery and equipment 51 36 2,657 0.13 0.07 Other manufacturing 46 25 2,164 0.09 0.06 Transport and logistics services 59 11 2,909 0.07 0.04 Total mainly efficiency-seeking 51 34 17,594 0.10 0.07 Total all sectors 37,445 0.08 0.05 Source: Computation based on data from Bureau of Economic Analysis (BEA) Statistics on activities of US foreign affiliates (Table II.E 11. Goods Supplied by Affiliates, Industry of Affiliate by Destination, 2014), and fDi Markets database (2009–15), the Financial Times. Note: FDI = foreign direct investment; IT = information technology; n.a. = not applicable. a. The export share by sector is calculated as non-host country sales divided by total sales based on the BEA data. Sectors are classified as natural resource– seeking if the sector description clearly indicates a direct link with natural resources. Remaining sectors are classified as efficiency-seeking if the share of exported sales exceeds 40 percent, and as market-seeking otherwise. Sectors with no BEA data availability are classified based on authors’ intuition. b. Because of data limitations in more recent years, this indicator is based on the 2008 BEA data on US foreign affiliates. c. The Herfindahl-Hirschman Index (HHI) of geographic concentration is defined as the sum of the squares of all developing countries’ shares in the total number of FDI projects for a given sector. It would hence take the value of 1 in a hypothetical case where all FDI projects in a given sector went to one country and approach zero the more dispersed FDI projects are across countries. China and India are excluded in the last column as a robustness check owing to their high share in the overall number of investment projects. 82 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 is primarily in natural resource– or efficiency- these sectors show the highest prevalence of seeking sectors is associated with a large share incentives (figure 3.4, upper right quadrant). of exports, while market-seeking investment, The IT services sector is somewhat of an out- by definition, leads mainly to domestic sales. lier in that, while it is highly geographically On the basis of this categorization, FDI in concentrated and mainly efficiency-seeking, mainly market-seeking sectors accounts for fewer developing countries offer incentives for 48 percent of projects in developing countries, this sector than for other mainly efficiency- followed by projects that are efficiency-seeking seeking sectors. (47 percent) and natural resource–seeking This suggests that some developing coun- (5 percent). FDI projects in natural resources, tries use incentives strategically in sectors with however, tend to be large in terms of the size high shares of efficiency-seeking FDI where of capital investment, and thus account for a competition is particularly intense. It also higher share of overall FDI value than their shows that, while incentives may be an impor- share in the number of projects. tant part of the value proposition to investors, Natural resource– and efficiency-seeking they are not a sufficient condition for FDI FDI tends to exhibit much higher shares of in these sectors as FDI is concentrated in rela- intrafirm sales than market-seeking FDI tively few locations despite the widespread (table 3.2). In the case of efficiency-seeking availability of incentives for these sectors. FDI, this finding reflects firms’ attempts to On the other hand, FDI in mainly market- organize and control their global value chains and natural resource–seeking sectors also (GVCs) across different production locations. flows to less competitive locations; and, while Being able to attract efficiency-seeking FDI is incentives remain common, they may not be therefore often a prerequisite for countries to necessary. FDI projects in extractives, power integrate with GVCs and to export to the mar- and utilities, and financial services, for exam- kets they serve. ple, are among the most dispersed geographi- Efficiency-seeking FDI tends to cluster in cally. Incentives are less common in these relatively few successful host countries while sectors yet are still offered by about 50 per- market- and natural resource–seeking FDI cent of developing countries ( figure 3.4, are more geographically dispersed (table 3.2). lower left quadrant). As competition for FDI Such a pattern of clustering is consistent with is more limited in these sectors, and location efficiency-seeking FDI being highly mobile decisions are likely dominated by questions and driven by firms strategically organizing of market demand and availability of natural their value chains by locating in cost- resources, such incentives are good candi- competitive host countries. Depending on the dates for further study and possible elimina- industry, this means that countries must com- tion as they may well be redundant. pete for efficiency-seeking FDI and that not In the GIC survey results, the share of all of them win. On the other hand, market- respondents rating incentives such as tax holi- and natural resource–seeking FDI, by defini- days as important or critically important for tion, must go where the market or natural their investment decision is considerably lower resource is located, and are thus more geo- for market- and natural resource–seeking graphically dispersed. investors (47 percent) than for efficiency- In sectors where FDI is predominantly effi- seeking investors (64 percent). The GIC survey ciency-seeking, competition for FDI is high also finds that developing country–based and incentives are commonly offered by devel- efficiency-seeking investors care more about oping countries. For FDI in such efficiency- incentives, relative to efficiency-seeking compa- seeking sectors as IT and electronics, machinery nies of developed countries. But country- and equipment, automotive, air- and space- specific analysis of FDI motivation and costs craft, and biotechnology and pharmaceuticals, and benefits of incentives is an important step most FDI projects are clustered in a limited in confirming these broad trends before reform- number of host countries; at the same time, ing a country’s incentives regime (box 3.2). CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 83 FIGURE 3.4 Incentives Are Used Most in Sectors with Heavy Competition for Efficiency-Seeking Investment Prevalence of incentives and FDI concentration 0.8 Air - and spacecraft Automotive industry and other transport equipment Machinery and equipment 0.7 Low to high geographic clustering of FDI projects HHI of geographic concentration of FDI (index) Other manufacturing IT services IT and electronics Apparel, textile, and footwear 0.6 Trade and retail Biotechnology, pharmaceuticals, and Transport and medical products logistics services Power, utilities, Agriculture and fishing 0.5 Business services and telecommunications Food and Education and health Construction and beverages building materials Entertainment Tourism and hospitality 0.4 Extractive industries Financial services 0.3 45 50 55 60 65 70 75 Share of countries offering incentives or CIT rates <= 15%, (percent) Low to high degree of tax competition Mostly efficiency-seeking FDI Mostly natural resource-seeking FDI Mostly market-seeking FDI Source: Computation based on Developing Country Tax Incentives Database and FDI data from fDi Markets database, the Financial Times. Note: The size of each bubble represents the number of FDI projects within the sector in developing countries. This was constructed based on information from the fDi Markets database. CIT = corporate income tax; FDI = foreign direct investment; IT = information technology. BOX 3.2 Methodologies and Results from Country-Level Cost-Benefit Analysis of Incentives This box summarizes recent work on cost–benefit To analyze the costs of incentives, a minimum analysis (CBA) of tax incentives for FDI attraction in requirement is to collect a list of fi rms, by sector, ben- developing countries where data availability is often efitting from incentives. While not explicitly covering limited. Even in low data environments, basic analyti- costs, such information can be a useful starting cal steps can help promote a more informed policy dia- point to see which sectors enjoy the most incentives. logue on tax incentives. It can also highlight distortions to competition if box continues next page 84 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 BOX 3.2 Methodologies and Results from Country-Level Cost-Benefit Analysis of Incentives (continued) incentives benefi t only a few fi rms within a sector. with and without the incentive. While this approach A sector-level analysis to motivate further data col- involves judgment in defining a credible minimum lection and research can be done by merging data on return for an investment to proceed, it can lead to the prevalence of incentives with outcome variables intuitive yet highly policy-relevant insights. For (for example, employment and investment) from sec- example, the above-mentioned analysis of Sri Lanka ondary sources such as an Enterprise Census or Labor also revealed that fi rms in the communication sector Force Survey. While falling short of a proper CBA, averaged high returns on investment, and that these this basic approach can help a country identify sectors returns would have remained above the country aver- with an obvious disproportion between the grant- age even without the incentives they received. Such a ing of incentives and the benefits of doing so. For fi nding suggests that incentives granted to this sector example, a recent study on Côte d’Ivoire (World Bank were likely redundant and that the investment would 2016b) found that while almost 15 percent of com- have been undertaken in any case. panies receiving incentives were in the construction A formal quantitative assessment of the tax incen- sector, this sector accounted for only about 5 percent tive’s costs and benefits is offered by the user cost of of total investment and 2 percent of employment in capital (UCC) methodology. This approach is more the country. data intensive as it requires fi rm-level data from bal- A much better starting point for understanding the ance sheets and/or tax returns over a period of several costs of incentives is a tax expenditure analysis. This years. It can produce an econometrically solid esti- entails assessing the corporate and indirect taxes that mate of the tax-investment relationship in a country would have been due from a given company in the by isolating the marginal investment effect of a given absence of incentives. Such information can be pro- tax concession. The UCC can be regarded as the pre- duced by the tax authorities using individual com- tax minimum rate of return required for an invest- panies’ tax returns. Collecting and publishing this ment to be considered profitable. By construction, the data on a regular basis increases the transparency of investment elasticities to UCC will vary across time incentives and enables policy makers and other stake- and fi rm (or group of fi rms); thus, comparing these holders to better assess their cost. Countries such as trends with what the UCC would have been without Colombia,a Morocco,b and South Africac follow this tax incentives permits an estimation of the change practice; but many others neither track nor publish in fi xed assets that is due to existing tax incentives. tax expenditure. Recent analytical work based on this methodology Confidentiality concerns often limit the ability of has produced rigorous measures of the net fi scal costs the tax administration to share fi rm-level tax expen- per job created, or unit of investment, for different diture data for analytical purposes. In such a case, sectors and incentive instruments in the Dominican aggregate results at the sector level can nevertheless Republic, Malaysia, and South Africa. But its heavy provide useful policy guidance by identifying dispro- data needs make this approach difficult to replicate in portions between tax expenditure and benefits gener- many lower-middle-income countries. ated by a sector. Research in Sri Lanka (World Bank A more easily replicable approach to shed light on 2016a), for example, shows that, although the com- the question of attribution of benefits to tax incentives munication sector absorbed 27 percent of total tax is an investor motivation survey. Such surveys ask expenditure, it accounted for only 1 percent of total fi rms a series of questions about the role of incentives employment. and other characteristics in their location decisions. A more rigorous assessment of costs and benefits Firms are classified as marginal investors if attracted is possible when firm-level data are available. One by an incentive versus nonmarginal investors that possibility is to analyze a fi rm’s return on investment would have come anyway based on their responses. box continues next page CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 85 BOX 3.2 Methodologies and Results from Country-Level Cost-Benefit Analysis of Incentives (continued) While this classification by survey responses requires combined with information on tax expenditure and some nontrivial judgment, the approach has been used benefits in terms of jobs, investment, and other vari- widely across developing countries. ables to calculate cost–benefit ratios. At the aggregate level, the share of investors who would have invested without an incentive (redun- a. “Article 87 of Act 788 (2002) established the Colombian government’s obligation dancy rate) is often high, ranging from 32 percent in to present a detailed report in which the fiscal impact of benefits must be evalu- ated and made explicit. The Oficina de Estudios Económicos de la Dirección de El Salvador to 92 percent in Guinea and 98 percent in Impuestos y Aduanas Nacionales (DIAN) (National Customs and Tax Directorate’s Rwanda, based on a recent series of investor motiva- Economic Research Office) has systematically published Colombia’s tax expenditure tion surveys (James 2013). However, because of sig- estimates since 2003 and presents the principal categories of preferential treat- ments for the last 10 years, making the distinction between those treatments to nificant variation by sector and investor motivation, individuals and companies.” Villela, Lemgruber, and Jorratt (2010). aggregate results are insufficient to derive credible b. Morocco publishes a detailed account of tax expenditure as part of its annual budget. Expenditure is presented by tax instrument, by type of beneficiary, and cost–benefit results. Thus, the survey sample size must by industrial sector. The detailed report also contains information on the types be large enough to disaggregate the resulting redun- of incentives granted, their legal basis, the intended objectives, and the eligible dancy rates by sector and motivation of the investor, beneficiaries. The full document for 2015 is available at http://www.finances.gov .ma/Docs/2014/DB/dep_fisc_fr.pdf. which is costly. If such a detailed breakdown were c. South Africa publishes supplementary information to the National Budget that available, sector-specific redundancy ratios could be provides some detail on tax expenditure. Firm- and Country-Level also been shown to be more effective in Variables Influence the countries with better infrastructure (Bellak and others 2009) and investment climates Impact of Incentives (James 2009). The effectiveness of tax incentives in attract- Linking the Developing Country ing FDI also depends on several firm- and Tax Incentives Database to data from the country-level variables. Previous research World Bank’s Enterprise Surveys sheds light has differentiated between greenfield invest- on the role of incentives, and firm and coun- ment versus mergers and acquisitions try characteristics in developing countries. (Hebous, Ruf, and Weichenrieder 2010), The Enterprise Surveys systematically collect export- versus domestic-oriented FDI firms’ perceptions concerning a number of (James 2009), and horizontal versus vertical obstacles they face in their operations, FDI (Overesch and Wamser 2008). It has including the tax burden. While there is some found a stronger effect of tax incentives on evidence linking this indicator to actual FDI greenfield FDI (Hebous, Ruf, and inflows (Kinda 2010), the observed effect of Weichenrieder 2010), export-oriented FDI incentives on investors’ perceptions of the (James 2009), and vertical FDI (Overesch tax system is more reasonably interpreted as and Wamser 2008). As all these firm charac- a necessary, but not sufficient, condition for teristics are consistent with efficiency-seek- the incentive to lead to more FDI. Companies ing FDI, the findings generally confirm that may be facing other obstacles and thus still efficiency-seeking FDI is more responsive to not invest, even if their perception of the tax tax incentives. Investment incentives have system improves owing to an incentive. But if 86 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 companies do not even see an incentive as for incentives. It could also suggest a prob- an improvement in the tax system they face, lem with high up-front costs of obtain- it is logical to conclude that this incentive is ing incentives—such as determining the not effective. Merging the Developing requirements to qualify for them and going Country Tax Incentives Database with infor- through cumbersome application pro- mation on perceptions of foreign firms from cesses—that make incentives worthwhile the Enterprise Surveys yields useful insights only for larger firms. This raises serious effi- (table 3A.6): ciency and equity concerns. Transparency- Not surprisingly, the CIT rate is posi- enhancing reforms (box 3.3) can mitigate tively associated with the likelihood up-front costs of incentives and also help that firms will rank taxes as an obstacle. avoid indirect costs attributable to corrup- A 10-percentage-point drop in the CIT rate tion and economic distortions. is associated with a 3.6 to 4 percentage point The link between tax holidays and the per- fall in the probability of foreign firms perceiv- ceptions of old versus new firms does not ing the tax rate as an obstacle. seem to differ. This should be reason for con- A tax holiday offered in the firm’s sector of cern because tax holidays are typically operation is associated with a 3.3 to 6.9 per- intended to promote new investments rather centage point drop in the likelihood of rank- than sustain existing ones. In practice, exist- ing the tax rate as an obstacle. This average ing investors often use rent-seeking behavior, finding masks significant variation of the including lobbying and strategic reinvest- effect depending on firm and country charac- ments, to extend tax holidays beyond their teristics. For example, the link between tax intended duration, which may explain this holidays and a firm’s perception of the tax finding in the data. These types of targeting rate is much stronger for exporting firms. problems seriously limit the effectiveness of Among exporters, the probability of ranking tax incentives for FDI promotion. A predeter- tax rates as an obstacle declines by 12 per- mined sunset clause for incentives can help centage points if a country offers tax holidays better shield policy decision making from in their sector of operation. The correspond- such pressures. ing figure for nonexporters is 3.8 percentage The positive link between tax holidays points. and firms’ perceptions of the tax rate does This finding is in line with results recorded not hold in countries with poor transport or by the GIC survey, suggesting that incentives investment climates. This is consistent with matter more for efficiency-seeking investors: literature showing that incentives are ineffec- 29 percent of efficiency-seeking firms reported tive in promoting FDI in such environments that tax holidays were critical when deciding (Bellak, Leibrecht, and Damijan 2009; James to invest or expand in developing countries. 2009). Tax holidays thus apparently cannot The Enterprise Surveys include only manufac- compensate for shortcomings in these areas turing and services firms and no natural and may be benefiting mainly firms that resource–seeking firms, so export-oriented would have invested anyway. Efficiency- firms can be equated with efficiency-seeking seeking FDI, the most likely to respond to investors in this dataset, confirming the previ- incentives, is particularly sensitive to the ous finding that incentives matter more for quality of the investment climate and trans- this type of FDI.8 port costs, and prone to clustering in the Similarly, the link between the exis- most competitive locations. This finding may tence of tax holidays and firms’ percep- thus result from efficiency-seeking investors tions of taxes as a barrier appears to be avoiding countries with weak investment cli- stronger for large firms (9.8 percentage mates regardless of incentives, while market- points) than for small ones (3.3 percentage and natural resource–seeking investors are points). This may reflect the widespread less responsive and operate in these countries use of minimum investment requirements regardless of the investment climate. CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 87 BOX 3.3 Examples of Transparency-Enhancing Reforms of Tax Incentives Incentives Inventories legislature reviews the incentives as part of the annual Publishing up-to-date information on the types of budget process. Furthermore, it supports the ability incentives offered, their legal basis, granted amounts, of the tax administration to keep track of, and moni- eligibility criteria, administration process, and other tor, incentives effectively. At the same time, granting relevant information is an important first step toward incentives based only on tax law avoids the discretion- increasing transparency. Often, this information is not ary practice of concluding individual agreements with available in developing countries in a comprehensive investors and thus limits the scope for rent-seeking manner and needs to be compiled by reviewing laws and corruption. In Tunisia, the new Investment Code and regulations that may include incentives—a pro- approved in 2016, instead of providing for incentives, cess that also yields important insights into incentives refers to a Fiscal Incentives Decree connected to the design. For investors, the inventory can be used to pub- Tax Code. In Sri Lanka, the new Inland Revenue Act licize relevant information and create a more level play- being considered would move all existing tax incen- ing field. A good example of an incentives inventory in tives into the tax code, no longer allowing the Sri a developing country is Jordan. The Jordan Investment Lanka Bureau of Investment to grant incentives under Commission publishes on its website, in a user-friendly its own authority. format, a list of incentives available to investors across Minimizing Discretion and Establishing Clear, all laws, as well as administrative procedures for Objective Eligibility Criteria for Granting Incentives applying for incentives. This inventory is underpinned Reducing the discretion of agencies administering by an internal IT system and is updated annually by a or awarding incentives enhances predictability for dedicated team. Another recent example is Pakistan, investors and reduces opportunities for rent-seeking where the Federal Board of Investment publishes on its and corruption. For tax incentives, a good prac- website all tax and customs duty incentives available to tice is awarding incentives to qualified investors investors through federal-level legislation. based on the criteria set out in the law, rather than Consolidating All Incentive Provisions in the Tax Law through a separate approval process. Costa Rica, for Keeping incentives in the tax law avoids scattering example, has established clear eligibility criteria for them through a country’s legislation (often including incentives through its Free Trade Zone Law, which the investment code, mining code, agricultural code, identifies the thresholds and practices for granting or special economic zone law). It also ensures that the incentives. Conclusion and the costs and benefits of existing incentives. While efforts to reduce harmful tax competi- • Improving the design, transparency, and tion remain a priority on the multilateral administration of incentives to reduce agenda, developing country governments indirect costs and avoid unintended can take unilateral steps to use tax incentives consequences. in a more targeted and cost-efficient manner. They can do this by implementing tailored An important starting point for any incen- reform strategies based on two pillars: tives system is to achieve clarity and consen- sus among stakeholders as to the specific and • Targeting incentives at those investors measurable policy goals to be pursued whose decision to invest is most likely through the incentives. Leaving objectives swayed by incentives. This requires a undefined, or trying to accomplish too many thorough understanding of the type of or vaguely defined goals, makes it impossible and motivation for FDI in the country to assess the success of incentives and is 88 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 bound to lead to failure. A robust monitoring incentives have a higher likelihood of being and evaluation framework to track progress redundant in that the investments they sup- toward such objectives is indispensable to port may have proceeded anyway. justify the public cost of tax incentives, and Country-specific cost–benefit analysis of detect and adjust redundant or inefficient incentives, including an assessment of redun- expenses. dancy by analyzing the return on investment Tax incentives should be targeted at effi- with or without an incentive, is important in ciency-seeking investors, but fundamentals of further tailoring this recommendation to the investment climate must be addressed country-specific circumstances. first. Getting a “piece of the cake” of globally Developing countries can improve the mobile efficiency-seeking FDI requires more design of incentives by moving away from effort in terms of proactive government profit-based to cost-based instruments involvement. Tax competition for efficiency- linked to clear policy goals. Most developing seeking FDI is intense; for some sectors with countries continue to rely heavily on tax the highest shares of efficiency-seeking FDI, holidays and preferential tax rates. The almost all developing countries offer some shortcomings of such profit-based instru- sort of corporate tax incentives. But effi- ments have been well established in that ciency-seeking FDI is also considerably more they are more attractive for firms with demanding than other forms of FDI in that it already high profits and short time horizons, requires a higher-quality investment climate, as opposed to cost-based instruments, such basic infrastructure, reasonable transport as tax allowances and credits, that directly costs, and a policy framework favoring lower the cost of investment. Profit-based investment. If these elements are lacking, incentives are also more prone to abuse investors are unlikely to respond to even the through tax planning and profit shifting. most generous incentives. Thus, for develop- As cost-based incentives can be tailored ing countries with poor performance along more closely to policy goals, host countries these dimensions, the most promising strat- should identify a realistic set of policy goals egy is to avoid the use of incentives and and design instruments accordingly. instead protect their revenue base to support Monitoring and evaluation systems should investment in infrastructure and improve- be put in place to track progress against ment of the investment climate while formu- the intended results. Finally, throughout lating a medium-term strategy to become this experiential process, policy makers more competitive for efficiency-seeking FDI. should be taking steps to learn and adapt On the other hand, countries that already accordingly.9 have the attributes to attract efficiency- By enhancing transparency and adminis- seeking FDI may in some cases find targeted tration practices, developing countries can incentives for this type of FDI useful to bol- reduce the indirect costs of incentives result- ster their locational competitiveness. ing from rent-seeking and corruption, and Tax incentives for natural resource– and avoid excessive administrative costs. This market-seeking investors are often redun- includes avoiding the use of discretionary or dant and should be primary targets for ad hoc incentives by mandating that all incen- further evaluation and potential removal. tives be clearly laid out in the relevant law. Countries across geographic regions and Consolidating the legal basis for incentives in income groups continue to offer investment the tax law can also help enhance transpar- incentives to market- and natural resource– ency and facilitate control by the tax adminis- seeking FDI. In most cases, these investors tration. On the administration side, reducing are not explicitly targeted by incentives but discretion in awarding incentives and, ideally, benefit from incentives offered to all or most awarding them automatically to any investors investors in a country. For these investors, qualified under the law can reduce up-front CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 89 costs that can render incentives unattractive, limitations of existing data and methodolo- especially for smaller investors. Finally, to gies to systematically explore causal effects avoid capture and perpetual renewal of incen- between incentives and FDI, a key priority is tives by established firms that in practice often to collect longer-term time series data on make tax incentives ineffective in terms of incentives and FDI, by sector, for developing generating new investment, incentives should countries. always be temporary in nature, including Another avenue of research could focus on through a pre-announced sunset clause. globally comparable firm-level data and look The evidence on the use of tax incentives at the micro effects of incentives (for example, in developing countries clearly needs to be returns on investment and firm expansion). developed further. The current version of the Such micro-based research could also move Developing Country Tax Incentives Database beyond the focus on FDI entry and consider covers only CIT incentives; an extension, in the role of incentives for FDI retention, particular to customs and value added tax linkages between foreign and domestic firms, incentives, would be desirable, as would be employment, or other behavioral characteris- the inclusion of subnational data.10 Given the tics of firms receiving incentives. 90 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 Annex 3A TABLE 3A.1 Countries in Developing Country Tax Incentives Database Low-income countries Lower-middle-income countries Upper-middle-income countries Haiti LAC Cambodia EAP China EAP Afghanistan SAR Indonesia EAP Fiji EAP Nepal SAR Lao PDR EAP Malaysia EAP Burundi SSA Mongolia EAP Thailand EAP Chad SSA Myanmar EAP Albania ECA Congo, Dem. Rep. SSA Papua New Guinea EAP Azerbaijan ECA Ethiopia SSA Philippines EAP Belarus ECA Gambia, The SSA Samoa EAP Bosnia and Herzegovina ECA Guinea SSA Sri Lanka EAP Bulgaria ECA Liberia SSA Timor-Leste EAP Georgia ECA Madagascar SSA Vietnam EAP Kazakhstan ECA Malawi SSA Armenia ECA Macedonia, FYR ECA Mozambique SSA Moldova ECA Montenegro ECA Rwanda SSA Tajikistan ECA Romania ECA Senegal SSA Ukraine ECA Serbia ECA Sierra Leone SSA Uzbekistan ECA Turkey ECA South Sudan SSA Bolivia LAC Turkmenistan ECA Tanzania SSA Guatemala LAC Argentina LAC Uganda SSA Honduras LAC Belize LAC Zimbabwe SSA Nicaragua LAC Brazil LAC Egypt, Arab Rep. MENA Colombia LAC Morocco MENA Costa Rica LAC Tunisia MENA Dominica LAC Bangladesh SAR Dominican Republic LAC India SAR Ecuador LAC Pakistan SAR Grenada LAC Cameroon SSA Guyana LAC Cabo Verde SSA Jamaica LAC Congo, Rep. SSA Mexico LAC Côte d’Ivoire SSA Panama LAC Ghana SSA Paraguay LAC Kenya SSA Peru LAC Lesotho SSA Saint Lucia LAC Mauritania SSA Suriname LAC Nigeria SSA Venezuela, RB LAC São Tomé and Príncipe SSA Algeria MENA Sudan SSA Iraq MENA Swaziland SSA Jordan MENA Zambia SSA Lebanon MENA Libya MENA Maldives SAR Angola SSA Botswana SSA Equatorial Guinea SSA Gabon SSA Mauritius SSA Namibia SSA South Africa SSA Source: Developing Country Tax Incentives Database. Note: EAP = East Asia and Pacific: ECA = Europe and Central Asia; LAC = Latin America and Caribbean; MENA = Middle East and North Africa; SAR = South Asia; SSA = Sub-Saharan Africa. CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 91 TABLE 3A.2 Global Use of Tax Holidays, 2015 Europe Latin Middle East and America East and Sub- Lower- Upper- Asia and Central and the North South Saharan Low- middle- middle- Pacific Asia Caribbean Africa Asia Africa income income income Total Number of countries covered in database 15 18 23 8 6 37 20 39 48 107 Prevalence of tax holidays by sector and region (share of countries offering tax holidays in a given sector, percent) Construction and building materials 71 56 48 50 50 32 35 47 52 47 Machinery and equipment 71 56 48 50 50 30 30 47 52 46 Air- and spacecraft 64 56 48 50 50 30 30 47 50 45 Automotive industry and other transport 64 56 48 50 50 30 30 47 50 45 IT and electronics 71 56 48 50 33 30 30 45 52 45 Apparel, textiles, and footwear 64 56 48 50 33 30 30 45 50 44 Food and beverages 64 56 48 50 33 30 30 45 50 44 Other manufacturing 64 56 48 50 33 30 30 45 50 44 Biotechnology, pharmaceuticals, and medical products 57 56 48 50 33 30 30 42 50 43 Agriculture and fishing 64 39 30 13 33 32 30 42 33 36 Tourism and hospitality 50 33 35 38 33 24 25 37 33 33 Extractive industries 29 39 26 25 33 24 20 32 29 28 Transport and logistics services 43 33 22 13 33 24 20 29 29 27 Education and health 50 28 22 13 50 19 15 32 27 26 IT services 50 39 22 13 17 19 15 26 31 26 Financial services 29 39 26 13 17 19 15 21 31 25 Power, utilities, and telecommunications 36 28 22 13 50 19 15 29 25 25 Renewable energy 29 33 26 13 33 19 15 26 27 25 Business services 43 28 26 13 17 16 15 24 27 24 Entertainment 43 28 22 13 17 19 15 24 27 24 Recycling 29 28 22 13 17 22 15 24 25 23 Trade and retail 29 33 22 13 17 19 15 21 27 23 Total (countries with tax holidays in at least one sector) 71 61 48 50 50 41 40 55 52 51 Median duration of tax holidays by sector and region, years Air- and spacecraft 10.0 6.0 15.0 10.0 5.0 10.0 8.5 9.5 10.0 10.0 Apparel, textiles, and footwear 9.0 6.0 15.0 10.0 7.5 10.0 8.5 9.0 10.0 10.0 Automotive industry and other transport 9.0 6.0 15.0 10.0 5.0 10.0 8.5 8.5 10.0 10.0 Biotechnology, pharmaceuticals, and medical products 9.0 6.0 15.0 10.0 7.5 10.0 8.5 9.0 10.0 10.0 Business services 9.0 5.0 12.0 10.0 10.0 10.0 10.0 10.0 9.0 10.0 Construction and building materials 8.5 6.0 15.0 10.0 10.0 10.0 7.0 10.0 10.0 10.0 Financial services 9.0 5.0 12.0 10.0 10.0 10.0 10.0 10.0 9.0 10.0 Food and beverages 9.0 6.0 15.0 10.0 7.5 10.0 8.5 9.0 10.0 10.0 IT and electronics 8.5 6.0 15.0 10.0 7.5 10.0 8.5 9.0 10.0 10.0 Machinery and equipment 8.5 6.0 15.0 10.0 5.0 10.0 8.5 8.5 10.0 10.0 Other manufacturing 9.0 6.0 15.0 10.0 7.5 10.0 8.5 9.0 10.0 10.0 Power, utilities, and telecommunications 10.0 5.0 9.0 10.0 10.0 10.0 10.0 10.0 8.5 10.0 Tourism and hospitality 8.0 6.5 10.0 10.0 6.5 10.0 5.0 7.5 10.0 10.0 table continues next page 92 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 TABLE 3A.2 Global Use of Tax Holidays, 2015 (continued) Europe Latin Middle East and America East and Sub- Lower- Upper- Asia and Central and the North South Saharan Low- middle- middle- Pacific Asia Caribbean Africa Asia Africa income income income Total Trade and retail 9.0 6.5 9.0 10.0 10.0 10.0 10.0 10.0 9.0 10.0 Transport and logistics services 9.5 6.5 9.0 10.0 7.5 10.0 7.5 10.0 9.5 10.0 Extractive industries 10.0 5.0 12.0 10.0 8.5 10.0 7.5 10.0 8.5 9.5 Recycling 9.0 5.0 9.0 10.0 10.0 10.0 10.0 10.0 8.5 9.5 Renewable energy 9.0 5.0 12.0 10.0 7.5 10.0 10.0 7.5 9.0 9.5 Entertainment 8.5 5.0 9.0 10.0 10.0 10.0 10.0 10.0 8.0 9.0 Agriculture and fishing 9.0 5.0 15.0 10.0 7.5 8.5 6.0 9.5 8.5 8.5 Education and health 8.0 5.0 9.0 10.0 5.0 10.0 10.0 7.0 8.0 8.5 IT services 8.0 5.0 9.0 10.0 10.0 10.0 10.0 9.0 8.0 8.5 Total 9.0 5.0 15.0 10.0 10.0 10.0 10.0 10.0 10.0 10.0 Prevalence of conditions for getting the tax holiday by type, percent Conditional on location (province or SEZ) 92 68 69 100 91 72 57 83 78 77 Conditional on exporting or selling to exporters 40 16 25 81 30 24 38 49 33 30 Subject to other conditions (for example, R&D, use of advanced machinery) 65 24 32 48 34 38 24 40 23 40 Source: Developing Country Tax Incentives Database. Note: IT = information technology; R&D = research and development; SEZ = special economic zone. TABLE 3A.3 Global Use of Preferential Tax Rates, 2015 Europe Latin Middle East and America East and Sub- Lower- Upper- Asia and Central and the North South Saharan Low- middle- middle- Pacific Asia Caribbean Africa Asia Africa income income income Total Number of countries covered in database 15 18 23 8 6 37 20 39 48 107 Prevalence of preferential rates by sector and region (share of countries offering concessions in a given sector, percent) Food and beverages 40 33 22 25 67 27 25 46 21 31 IT and electronics 33 39 22 25 67 27 25 46 21 31 Air- and spacecraft 33 33 22 25 67 27 25 44 21 30 Automotive industry and other transport 33 33 22 25 67 27 25 44 21 30 Biotechnology, pharmaceuticals, and medical products 33 33 22 25 67 27 25 44 21 30 Machinery and equipment 33 33 22 25 67 27 25 44 21 30 Construction and building materials 40 33 22 13 50 27 25 38 23 29 Other manufacturing 33 33 22 25 50 27 25 41 21 29 Apparel, textiles, and footwear 27 33 22 25 50 27 25 38 21 28 Agriculture and fishing 33 33 13 0 67 14 20 31 15 21 Power, utilities, and telecommunications 27 28 17 0 50 19 20 28 17 21 Tourism and hospitality 27 28 17 13 50 16 20 28 17 21 Education and health 20 28 17 13 50 16 20 23 19 21 table continues next page CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 93 TABLE 3A.3 Global Use of Preferential Tax Rates, 2015 (continued) Europe Latin Middle East and America East and Sub- Lower- Upper- Asia and Central and the North South Saharan Low- middle- middle- Pacific Asia Caribbean Africa Asia Africa income income income Total Financial services 27 28 17 0 33 19 20 26 17 21 IT services 33 28 17 0 50 14 15 26 19 21 Transport and logistics services 20 28 17 13 50 16 20 23 19 21 Entertainment 27 28 17 0 50 14 15 23 19 20 Extractive industries 13 22 13 13 50 22 30 23 13 20 Recycling 20 28 17 0 50 16 20 23 17 20 Renewable energy 20 28 17 0 50 16 20 23 17 20 Business services 20 28 17 0 50 14 15 23 17 19 Trade and retail 13 28 17 0 50 14 15 21 17 18 Total 60 39 26 38 67 38 40 56 27 40 Median preferential margin (standard CIT rate—preferential rate by sector and region, percent) Air- and spacecraft 16.0 10.0 25.0 18.0 16.0 13.5 11.0 15.0 15.0 15.0 Apparel, textiles, and footwear 14.5 10.0 25.0 18.0 13.0 13.5 11.0 15.0 15.0 15.0 Automotive industry and other transport 16.0 10.0 25.0 18.0 16.0 13.5 11.0 15.0 15.0 15.0 Biotechnology, pharmaceuticals, and medical products 16.0 10.0 25.0 18.0 16.0 13.5 11.0 15.0 15.0 15.0 Machinery and equipment 15.5 10.0 24.5 0.0 15.0 12.0 11.5 15.5 12.5 15.0 Construction and building materials 14.5 10.0 25.0 18.0 16.0 13.5 11.0 15.0 15.0 15.0 Other manufacturing 16.0 10.0 25.0 18.0 16.0 13.5 11.0 15.0 15.0 15.0 Apparel, textiles, and footwear 16.0 10.0 25.0 18.0 16.0 13.5 11.0 15.0 15.0 15.0 Agriculture and fishing 16.0 10.0 25.0 18.0 15.0 13.5 12.0 15.0 15.0 15.0 Power, utilities, and telecommunications 16.5 10.0 24.5 0.0 10.0 15.0 11.5 25.0 12.5 15.0 Construction and building materials 12.0 10.0 25.0 15.0 13.0 13.5 11.0 15.0 15.0 13.0 Entertainment 15.5 10.0 24.5 0.0 13.0 12.0 12.0 13.0 15.0 13.0 Transport and logistics services 16.0 10.0 24.5 15.0 13.0 13.5 11.5 13.0 15.0 13.0 Business services 16.0 10.0 24.5 0.0 13.0 12.0 12.0 13.0 12.5 12.5 IT services 16.0 10.0 24.5 0.0 13.0 12.0 11.5 18.0 10.0 12.5 Agriculture and fishing 13.0 9.5 26.0 0.0 13.0 11.0 12.0 13.0 10.0 12.0 Education and health 12.0 10.0 24.5 12.0 13.0 12.0 12.0 13.0 12.0 12.0 Extractive industries 22.5 9.5 26.0 12.0 13.0 11.0 11.0 15.0 15.0 12.0 Recycling 16.0 10.0 24.5 0.0 13.0 12.0 12.0 13.0 12.5 12.0 Renewable energy 16.0 10.0 24.5 0.0 13.0 12.0 12.0 13.0 12.5 12.0 Tourism and hospitality 14.5 10.0 24.5 12.0 13.0 12.0 12.0 13.0 16.0 12.0 Trade and retail 17.0 10.0 24.5 0.0 13.0 12.0 12.0 13.0 12.5 12.0 Total 16.0 10.0 25.0 15.0 13.0 12.0 12.0 15.0 15.0 13.0 Prevalence of conditions for getting the tax allowance by type, percent Conditional on location (province or SEZ) 53 31 54 95 43 41 54 34 57 45 Conditional on exporting or selling to exporters 34 18 9 95 56 35 44 47 5 32 Subject to other conditions (for 46 10 23 14 67 81 67 47 35 46 example, R&D, use of advanced machinery) Source: Developing Country Tax Incentives Database. Note: CIT = corporate income tax; IT = information technology; R&D = research and development; SEZ = special economic zone. 94 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 TABLE 3A.4 Global Use of Tax Allowances and Credits, 2015 Europe Latin Middle East and America East and Sub- Lower- Upper- Asia and Central and the North South Saharan Low- middle- middle- Pacific Asia Caribbean Africa Asia Africa income income income Total Number of countries covered in database 15 18 23 8 6 37 20 39 48 107 Prevalence of tax allowances by sector and region (share of countries offering tax allowance in a given sector, percent) Machinery and equipment 20 11 4 13 17 14 20 10 10 12 Apparel, textiles, and footwear 20 11 4 13 0 14 20 10 8 11 Automotive industry and other transport 13 11 4 13 0 14 20 8 8 10 Biotechnology, pharmaceuticals, and medical products 20 11 4 13 0 11 20 10 6 10 Construction and building materials 20 11 4 13 0 11 20 8 8 10 Food and beverages 20 11 4 13 0 11 20 10 6 10 IT and electronics 13 11 4 13 0 14 20 8 8 10 Air- and spacecraft 13 11 4 13 0 11 20 8 6 9 Other manufacturing 13 11 4 13 0 11 20 8 6 9 Tourism and hospitality 13 11 0 13 0 14 20 8 6 9 Renewable energy 13 11 4 13 0 8 15 5 8 8 Education and health 13 11 0 13 0 8 10 5 8 7 Entertainment 13 11 0 13 0 8 5 8 8 7 Power, utilities, and telecommunications 13 11 4 13 0 5 10 5 8 7 Agriculture and fishing 13 11 0 13 0 5 10 5 6 7 IT services 20 11 0 13 0 3 5 5 8 7 Recycling 13 11 0 13 0 5 10 5 6 7 Trade and retail 20 11 0 13 0 3 5 8 6 7 Transport and logistics services 13 11 0 13 0 5 10 5 6 7 Business services 13 11 0 13 0 3 5 5 6 6 Financial services 13 11 0 13 0 3 5 5 6 6 Extractive industries 13 11 0 0 0 3 5 3 6 5 Total 33 11 9 13 17 16 25 13 15 16 Prevalence of conditions for getting the tax allowance by type, percent Conditional on location (province or SEZ) 20 0 18 100 100 76 58 61 19 44 Conditional on exporting or selling to exporters 14 0 100 100 100 60 58 7 16 41 Subject to other conditions 96 100 100 0 100 84 100 95 99 83 (for example, R&D, use of advanced machinery) Source: Developing Country Tax Incentives Database. Note: IT = information technology; R&D = research and development; SEZ = special economic zone. CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 95 TABLE 3A.5 Changes in Tax Incentives, 2009–15 Europe Latin Middle East and America East and Sub- Lower- Upper- Asia and Central and the North South Saharan Low- middle- middle- Pacific Asia Caribbean Africa Asia Africa income income income Total Number of countries covered in database 14 18 23 8 6 34 18 37 48 103 Share of countries introducing new tax incentives between 2009 and 2015 or making existing incentives more generous, percent Agriculture and fishing 36 33 22 25 17 44 28 32 35 33 Air- and spacecraft 36 39 22 25 17 35 22 30 35 31 Apparel, textiles, and footwear 36 39 22 25 17 32 17 30 35 30 Automotive industry and other transport 36 39 22 25 17 32 17 30 35 30 Biotechnology, pharmaceuticals, and medical products 36 39 22 25 17 32 17 30 35 30 Business services 36 33 30 50 17 35 22 30 42 34 Construction and building materials 36 39 22 25 17 32 17 30 35 30 Education and health 36 33 26 50 17 41 22 32 42 35 Entertainment 36 33 30 50 17 44 22 35 44 37 Extractive industries 36 33 22 13 17 32 28 32 25 28 Financial services 36 33 26 38 17 35 22 24 42 32 Food and beverages 36 39 22 25 17 32 17 30 35 30 IT services 36 39 26 50 0 38 22 30 42 34 IT and electronics 36 39 22 25 17 32 17 30 35 30 Machinery and equipment 36 39 22 25 17 32 17 30 35 30 Other manufacturing 36 39 22 25 17 32 17 30 35 30 Power, utilities, and telecommunications 36 33 26 38 0 38 17 30 40 32 Recycling 36 33 26 50 17 41 22 35 40 35 Renewable energy 36 33 26 50 17 38 22 32 40 34 Tourism and hospitality 36 39 26 50 17 41 22 35 42 36 Trade and retail 36 33 26 50 17 35 22 32 38 33 Transport and logistics services 36 33 26 50 17 38 17 32 42 34 Total (countries with more generous incentives in at least one sector) 36 39 35 50 17 65 44 43 48 46 Share of countries removing tax incentives between 2009 and 2015 or making them less generous, percent Air- and spacecraft 7 22 17 38 33 12 17 14 21 17 Apparel, textiles, and footwear 0 17 17 38 33 12 17 11 19 16 Automotive industry and other transport 7 17 17 38 33 15 17 14 21 17 Biotechnology, pharmaceuticals, and medical products 0 17 17 38 33 12 17 11 19 16 Business services 0 17 17 38 33 12 17 11 19 16 Construction and building materials 7 17 9 13 33 12 17 11 13 13 Financial services 0 17 17 25 33 12 17 8 19 15 Food and beverages 7 17 9 13 33 12 17 11 13 13 IT and electronics 0 17 9 13 33 12 17 8 13 12 Machinery and equipment 0 17 17 38 33 15 22 8 21 17 Other manufacturing 0 17 9 13 17 12 17 8 10 11 table continues next page 96 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 TABLE 3A.5 Changes in Tax Incentives, 2009–15 (continued) Europe Latin Middle East and America East and Sub- Lower- Upper- Asia and Central and the North South Saharan Low- middle- middle- Pacific Asia Caribbean Africa Asia Africa income income income Total Power, utilities, and telecommunications 0 17 17 38 33 12 17 11 19 16 Tourism and hospitality 7 17 9 13 33 15 17 11 15 14 Trade and retail 0 17 17 38 33 12 17 11 19 16 Transport and logistics services 0 17 17 38 33 12 17 11 19 16 Extractive industries 0 17 17 38 33 12 17 11 19 16 Recycling 0 17 9 13 50 15 17 11 15 14 Renewable energy 0 17 9 13 33 12 17 8 13 12 Entertainment 0 17 4 13 33 12 17 8 10 11 Agriculture and fishing 7 17 13 13 33 12 17 8 17 14 Education and health 0 17 9 13 33 15 17 8 15 13 IT services 7 17 9 13 33 12 17 11 13 13 Total (countries with less generous incentives in at least one sector) 14 22 22 50 50 21 22 16 31 24 Source: Developing Country Tax Incentives Database. Note: IT = information technology. TABLE 3A.6 Regression Results on Tax Incentives and Foreign Firms’ Perceptions of Tax Rates as a Business Obstacle (1) (2) (3) (4) (5) (6) (7) Variables Margins Margins Margins Margins Margins Margins Margins CIT Corporate income 0.0037*** 0.0036*** 0.0037*** 0.0037*** 0.0036*** 0.0040*** 0.0036*** tax rate (0.0008) (0.0009) (0.0009) (0.0009) (0.0009) (0.0010) (0.0009) HOLIDAY 1 = Availability of tax −0.0686*** −0.0600*** −0.0384* −0.0327** −0.0608*** 0.0099 −0.0262 holiday in country and (0.0163) (0.0172) (0.0220) (0.0163) (0.0167) (0.0195) (0.0175) sector of operation EXPORTER 1 = Exporting firm −0.0633*** −0.0268 −0.0642*** −0.0634*** −0.0631*** −0.0646*** (>50% of sales) (0.0169) (0.0201) (0.0169) (0.0169) (0.0186) (0.0173) LARGE 1 = Large firm (>50% 0.0081 0.0069 0.0332** 0.0081 0.0096 0.0111 of sales) (0.0153) (0.0151) (0.0162) (0.0153) (0.0160) (0.0148) NEW 1 = New firm (10 years −0.0175 −0.0169 −0.0175 −0.0184 −0.0157 −0.0172 or younger) (0.0160) (0.0160) (0.0161) (0.0175) (0.0165) (0.0159) LPI 1 = Above median 0.0055 0.0038 0.0090 0.0055 0.0586*** 0.0119 Logistics Performance (0.0181) (0.0165) (0.0181) (0.0181) (0.0158) (0.0172) Index Score DB 1 = Above median −0.0322*** −0.0291*** −0.0295** −0.0322*** −0.0231* −0.0104 Doing Business DTF (0.0122) (0.0110) (0.0123) (0.0121) (0.0138) (0.01435) (excl. “paying taxes”) Interaction −0.0819*** HOLIDAY*EXPORTER (0.0310) Interaction −0.0655*** HOLIDAY*LARGE (0.0204) table continues next page CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 97 TABLE 3A.6 Regression Results on Tax Incentives and Foreign Firms’ Perceptions of Tax Rates as a Business Obstacle (continued) (1) (2) (3) (4) (5) (6) (7) Variables Margins Margins Margins Margins Margins Margins Margins Interaction HOLIDAY*NEW 0.0024 (0.0173) Interaction HOLIDAY*LPI −0.1330*** (0.0320) Interaction HOLIDAY*DB −0.0700** (0.0333) Other controls GDP, GDP per capita Fixed effects Sector, year Observations 5,396 5,191 5,191 5,191 5,191 5,191 5,191 Source: Computation based on data from World Bank Developing Country Tax Incentives Database and Enterprise Surveys, the World Bank. The Doing Business variable excludes the tax component of this indicator to avoid collinearity with the tax variables. Note: The table shows marginal effects from a logit regression linking foreign-owned firms’ responses to the World Bank Enterprise Surveys question on how severe a business obstacle tax rates represent for them to the CIT rate, availability of a tax holiday in their sector of operation in the year the survey was taken, and a number of firm- and country- specific control variables. Coefficients can be interpreted as the estimated change in probability of thinking tax rates present a “major” or “very severe” business obstacle given a change in the value of the relevant explanatory variables and holding all other explanatory variables at their mean value. The Doing Business variable, DB, measures the “Distance to the Frontier” (DTF) of highest-performing countries, with tax-related components removed to avoid endogeneity issues. The sample contains 5,396 manufacturing and services firms with a foreign ownership share of at least 10 percent, distributed across 81 developing countries with Enterprise Surveys available between 2009 and 2015. The results have a potential selection bias, meaning that the extension to marginal investors warrants caution. Specifically, the set of survey respondents are either firms that have opted to invest despite whatever weakness in the investment environment or, conversely, firms that have invested because of the availability of special tax incentives. Similar selection effects may also be reflected in firm size and export status. Similar results to those for tax holidays can be obtained for the availability of preferential tax rates. Tax allowances and credits are not used widely enough in developing countries to replicate results. The results for tax holidays are also robust to including controls for preferential taxes rate and tax allowances/credits. Standard errors in parentheses, clustered at sector level. ***p<0.01; **p<0.05; *p<0.1. Results at 10 percent or greater significance are shown in bold. Notes organizations such as the West African Economic and Monetary Union (WAEMU) 1. Following James (2009), investment incen- and the East African Community (EAC), tives can be defined as “measurable eco- which adopted a code of conduct for member nomic advantages that governments provide countries’ use of tax incentives. to specific enterprises or groups of enter- 3. For example, the WAEMU Treaty seeks prises, with the goal of steering investment to reduce distortions to intracommunity into favored sectors or regions, or of influ- trade and mobilize domestic tax revenue. encing the character of such investments. To this end, member countries have agreed These benefits can be fiscal (as with tax on an advanced mechanism for tax coordi- concessions) or non-fiscal (as with grants, nation that has led to some convergence in loans, or rebates to support business devel- members’ corporate tax rates. But regional opment or enhance competitiveness).” This coordination rules allow exemptions for definition raises two important distinctions: incentives granted under member countries’ locational incentives (intended to influence investment codes, creating what Mansour the location decision of investors) versus and Rota-Graziosi (2013) characterize as the behavioral incentives (intended to influence “Achilles heel” of the agreement. The same the character of the investment) and fiscal authors present evidence of the proliferation (through tax concession) versus nonfiscal. of investment incentives under various legal This chapter focuses on locational fiscal bases in member countries that have, in fact, incentives. undermined the purpose of the tax harmoni- 2. Examples include the base erosion and profit zation mechanism. sharing (BEPS) process under the umbrella 4. The database is available on request for of the OECD and EU rules on state aid. research purposes. Interested researchers can Developing countries have also made some contact the author of this chapter, Erik von progress in this regard, including regional Uexkull, at jvonuexkull@worldbank.org. 98 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 5. The WTO Subsidies and Countervailing Corporate Taxation?” International Tax and Measures Agreement (SCM) prohibits export Public Finance 12 (5): 583–603. subsidies for most products and defines meas- Desai, M. H., C. Fritz Foley, and J. R. Hines, Jr. ures against such subsidies (for example, 2006. “Taxation and Multinational Activity: requiring companies to export a certain share New Evidence, New Interpretations.” Survey of of production to be eligible for an incentive, Current Business 86 (2): 16–22. as well as requirements to buy local over Djankoff, S., T. Ganser, C. McLiesh, R. Ramalho, imported inputs). Certain exceptions apply and A. Schleifer. 2010. “The Effect of Corporate for low-income countries. Taxes on Investment and Entrepreneurship.” 6. The preferential margin refers to the differ- American Economic Journal: Macroeconomics ence between the standard CIT rate and the 2 (3): 31–64. preferential rate granted as an incentive. Dunning, J. H. 1980. “Toward an Eclectic Theory 7. For example, one car manufacturer may set of International Production: Some Empirical up a plant in a country to serve the domestic Tests.” Journal of International Business market while another may do so as part of a Studies 11 (1): 9–31. global offshoring strategy to export. Dunning, J. H. 1993. Multinational Enterprises 8. Incentives conditional on firms’ exporting and the Global Economy. Addison Wesley. status were removed from these estimations Egger, P. H., S. Loretz, M. Pfaffermayr, and in order to isolate how different types of H. Winner. 2008. “Bilateral Effective Tax firms react to the same type of incentives. Rates and Foreign Direct Investment.” Oxford This would not be a valid conclusion if incen- University Centre for Business Taxation tives available only to exporters were left in Working Papers 0802. Oxford, UK. the database. F r e u n d , C . , a n d T. H . M o r a n . 2 0 1 7 . 9. Andrews, Pritchett, and Woolcock (2017): “Multinational Investors as Export Superstars: The Problem Driven Iterative Adaptation How Emerging-Market Governments Can approach emphasizes the importance of exper- Reshape Comparative Advantage.” Working imenting, learning, iterating, and adapting in Paper 17-1, Peterson Institute for International order to address a problem. Economics, Washington, DC. 10. As the current database includes only infor- Hebous, S., M. Ruf, and A. Weichenrieder. 2010. mation on locational incentives, the evidence “The Effects of Taxation on the Location for behavioral incentives in investor deci- Decision of Multinational Firms: M&A vs. sions for developing countries is not explored Greenfield Investments.” CESifo Working in this chapter. Paper 3076. Munich, Germany. Heckemeyer, J., and M. Overesch. 2013. “Multinationals’ Profit Response to Tax Differentials: Effect Size and Shifting Channels.” Bibliography ZEW Discussion Paper 13-045, Zentrum für Allen, N. J., J. Morisset, N. Pirnia, and L. T. Wells, Europäische Wirtschaftsforschung, Mannheim. Jr. 2001. “Using Tax Incentives to Compete for Hunady, J., and M. Orviska. 2014. “Determinants Foreign Investment: Are They Worth the Costs?” of Foreign Direct Investment in EU Countries— Foreign Investment Advisory Service Occasional Do Corporate Taxes Really Matter?” Procedia Paper 15, World Bank, Washington, DC. Economics and Finance 12: 243–50. Andrews, M., L. Pritchett, and M. Woolcock. IMF (International Monetary Fund), OECD 2017. Building State Capability: Evidence, (Organisation for Economic Co-operation Analysis, Action . Oxford, U.K.: Oxford and Development), UN (United Nations), University Press. and World Bank. 2015. “Options for Low Bellak, C., M. Leibrecht, and J. P. Damijan. 2009. Income Countries’ Effective and Efficient “Infrastructure Endowment and Corporate Use of Tax Incentives for Investment.” In Income Taxes as Determinants of Foreign International Monetary Fund Report to the Direct Investment in Central and Eastern G-20 Development Working Group. European Countries.” The World Economy James, S. 2009. “Tax and Non-Tax Incentives and 32 (2): 267–90. Investments: Evidence and Policy Implications.” Bénassy-Quéré, A., L. Fontagne, and A. Lahreche- Foreign Investment Advisory Service, World Revil. 2005. “How Does FDI React to Bank, Washington, DC. CORPORATE TAX INCENTIVES AND FDI IN DE VELOPING COUNTRIES 99 ———. 2013. “Tax and Non-Tax Incentives and UNCTAD (United Nations Conference on Trade Investments: Evidence and Policy Implications.” and Development). 2015. World Investment Investment Climate Advisory Services, World Report: Reforming International Investment Bank, Washington, DC. Governance. Geneva: UNCTAD. James, S., and S. Van Parys. 2010. “The UNIDO (United Nations Industrial Development Effectiveness of Tax Incentives in Attracting Organization). 2011. Africa Investor Investment: Panel Data Evidence from the CFA Report: Towards Evidence-Based Investment Franc Zone.” International Tax and Public Promotion Strategies. Geneva: UNIDO. Finance 17 (4): 400–29. Van Parys, S. 2012. “The Effectiveness of Tax Kinda, T. 2010. “Investment Climate and FDI in Incentives in Attracting Investment: Evidence Developing Countries: Firm-Level Evidence.” from Developing Countries.” Reflets et World Development 38 (4): 498–513. perspectives de la vie économique 2012/3: Klemm, A., and S. Van Parys. 2012. “Empirical 129–41. Evidence on the Effects of Tax Incentives.” Villela, L., A. Lemgruber, and M. Jorratt. 2010. International Tax and Public Finance 19 (3): “Tax Expenditure Budgets: Concepts and 393–423. Challenges for Implementation.” IDB Working Mansour, M., and G. Rota-Graziosi. 2013. “Tax Paper 131, Inter-American Development Bank, Coordination, Tax Competition, and Revenue Washington, DC. Mobilization in the West African Economic and World Bank. 2016a. “Sri Lanka: Moving toward Monetary Union.” IMF Working Paper 13/163, a New Investment Incentives Framework.” International Monetary Fund, Washington, DC. Mimeo, World Bank, Washington, DC. OECD (Organisation for Economic Co-operation ——— . 2 0 1 6 b . “ E x a m e n d u r é g i m e and Development). 1998. “Harmful Tax d’incitations à l’investissement en République Competition: An Emerging Global Issue.” de Côte d’Ivoire.” Mimeo, World Bank, OECD, Paris. Washington, DC. Overesch, M., and G. Wamser. 2008. “Who Zolt, E. M. 2013. “Tax Incentives and Tax Base Cares about Corporate Taxation? Asymmetric Protection Issues.” Papers on Selected Topics Tax Effects on Outbound FDI.” IFO Working in Protecting the Tax Base of Developing Papers 59, IFO Institute for Economic Countries Draft Paper 3, United Nations, Research, University of Munich. New York, NY. 4 Outward FDI from Developing Countries Jose Ramon Perea and Matthew Stephenson O utward foreign direct investment impact, and policy implications. It draws on (OFDI) by fi rms from developing several global data sources to assess changes countries1 has grown dramatically over time in the investment decisions of devel- in recent years, accounting for nearly one- oping country multinational corporations fifth of global foreign direct investment (MNCs). The chapter also looks at findings (FDI) flows in 2015, up from just 4 percent from a gravity model on FDI flows and in 1995. While larger developing countries, qualitative evidence on developing country especially the BRICS (Brazil, the Russian MNC investments across several industries— Federation, India, China, and South including pharmaceuticals, wind turbines, Africa), are driving this phenomenon, many household appliances, and automobiles. developing countries are now engaged in The analysis answers three questions, OFDI, regardless of their size or level of whose answers have important implications development. The increasing importance of for policy makers, firms, and development such OFDI calls for a better understanding practitioners: of it and its implications. OFDI has economic effects not only in recipient economies, as 1. What are the salient features of develop- research shows, but also in source econo- ing country OFDI, especially with respect mies (“home effects”). Growing OFDI may to trends, destinations, sectors, and entry thus require that developing country gov- modes? ernments adopt new investment policy 2. Does OFDI benefit the source economy, reforms and investment promotion efforts and if it does, what are the facilitating or to maximize the benefits for both the home mediating factors? economy and its firms. 3. What role does OFDI-related policy play This chapter describes the rise of OFDI by and what further research is needed to developing country firms, its development better understand and shape it? 101 102 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 Several key findings emerge: restrict the entry of FDI and the potential OFDI from developing countries has emergence of OFDI, as developing countries boomed in recent years, leading to a greater aimed to nurture domestic industries and relative share of total OFDI, across both keep capital at home (Cuervo-Cazurra 2008; flows and stocks. In absolute terms, BRICS Gammeltoft, Barnard, and Madhok 2010). investors are the key drivers of developing Protectionist measures reduced incentives for country OFDI, accounting for 62 percent of domestic firms to become internationally total developing country OFDI stock in competitive, limiting their ability to expand 2015—with China alone accounting for outside their home markets. The small 36 percent. amount of developing country OFDI that did Developing country governments have take place generally went to other developing moved gradually from restricting to support- countries in the same region and was mostly a ing OFDI, although some form of restriction combination of natural resource–seeking2 (as remains in half of all developing countries— developing countries sought primary inputs especially lower-income countries. In some they lacked) and market-seeking (as a few cases, developing country governments have developing countries sought to expand sales even begun to provide incentives to target in culturally and geographically close neigh- strategic sectors. One reason is the increasing bors) (Dunning, Kim, and Park 2008; evidence that OFDI can boost innovation and Ramamurti 2009; Wells 2009). exports in the home economy. However, lim- The second wave, during the 1980s and ited absorptive capacity in developing econo- 1990s, saw investment patterns shift signifi- mies, vis-à-vis developed economies, is a key cantly. Structural reforms and export-oriented constraint on positive home effects from out- industrialization opened developing countries ward investment. to FDI, with countries seeking to attract the These findings suggest several policy con- foreign capital, knowledge, and skills needed siderations. Investment promotion agencies to make their exports competitive. With trade (IPAs) may wish to target not only traditional and investment liberalization progressing rap- sources of FDI but also new sources such as idly, developing country OFDI also began to developing country OFDI. At the same time, grow. About two-thirds of OFDI flows went policy makers may wish to review their coun- to developed economies, while the remaining tries’ OFDI regulatory frameworks, given third went to developing countries, mostly that restrictions may be undermining the neighbors (Aykut and Ratha 2004). It became positive effects on the home economy. increasingly efficiency-seeking, as developing Policy makers may also wish to consider countries began to plug into global value measures that expand firm-level and econ- chains (GVCs) by locating some manufactur- omy-level absorptive capacity to realize the ing activities in lower-cost locations and full positive effects of OFDI in home econo- integrating into international production mies. More policy-oriented research is clearly networks (UNCTAD 2013). needed to help developing country officials The third wave, from the early 2000s to better tailor and target future policy the present, is witnessing a fresh rise in devel- interventions. oping country OFDI, across both flows and stocks. While OFDI from both developed and developing economies has been dynamic, the The Rise of Developing relative share of developing country OFDI flows in total FDI (figure 4.1) surged from Country OFDI 4 percent in 1995 to 27 percent in 2014, The rise of developing country OFDI has equivalent to $315 billion. Developing coun- occurred in three “waves” (Gammeltoft try OFDI stocks (figure 4.2) have also 2008). The first, during the 1960s and 1970s, increased as a share of total FDI stocks, saw import-substitution industrialization although at a slower pace. Between 1995 and OUT WARD FDI FROM DEVELOPING COUNTRIES 103 FIGURE 4.1 Developing Country OFDI Flows 1,800 100 90 Share of total OFDI flow (percent) 1,600 80 Developed countries 1,400 OFDI flow (US$, billion) Developed countries 70 1,200 60 1,000 50 800 40 600 30 400 Developing countries 20 Developing countries 200 10 0 0 1995 2000 2005 2010 2015 1995 2000 2005 2010 2015 Year Year Source: Computation based on United Nations Conference on Trade and Development (UNCTAD). Note: OFDI = outward foreign direct investment. FIGURE 4.2 Developing Country OFDI Stocks 20,000 100 Share of total OFDI stock (percent) 18,000 90 Developed countries Developed countries OFDI stock (US$, billion) 16,000 80 14,000 70 12,000 60 10,000 50 8,000 40 6,000 30 4,000 Developing countries 20 Developing countries 2,000 10 0 0 1995 2000 2005 2010 2015 1995 2000 2005 2010 2015 Year Year Source: Computation based on UNCTAD. Note: OFDI = outward foreign direct investment. 2015, developing countries tripled their share resource allocation and diversify risks from in global FDI stocks, increasing from 4 per- economic shocks in any one region” (Lee, cent to 12 percent, equal to $2.8 trillion. Lee, and Yeo 2016). Firms in other develop- Both domestic policy choices in develop- ing countries soon followed, with OFDI ing countries and global economic condi- increasingly seen as a means to access mar- tions helped shape these changes in the kets, capital, technology, and knowledge in investment landscape. In terms of domestic international markets—and thus boost policy, liberalization and deregulation national competitiveness (Luo, Xu, and Han reforms embraced in the second wave (the 2010). Supportive policy measures, in the 1980s–90s) raised competitive pressures in form of generous financing and incentives, many developing countries, eventually helped. “pushing” firms out of their home markets Global economic conditions also “pulled” (Sauvant 2008). At the same time, firms in developing market firms into OFDI. First, Singapore and other high-growth economies rapid and sustained growth in much of the embraced OFDI in the late 1990s as a devel- developing world during this decade facili- opment strategy to “achieve efficiency in tated firms to grow and prosper and, 104 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 consequently, internationalize. Second, the main geographic origins and destinations of boom in commodity prices during the same these flows, principal modes of entry (green- decade gave commodity exporters in the field versus mergers and acquisitions developing world large windfalls, creating [M&A]), and sectoral distribution, among substantial liquidity that they used in part to other patterns. finance OFDI. Who? Sources of Developing Zooming In: Who, Where, What, Country OFDI and How East Asia and Pacific has gradually become the major source of OFDI among developing This section looks at trends in OFDI by regions (figure 4.3). It generated 22 percent of developing country firms as revealed by vari- total OFDI from developing country firms ous global datasets (UNCTAD, fDi Markets, during 2000–04, surging to 49 percent in and Thomson Reuters) and identifies the 2010–15.3 In contrast, Europe and Central Asia, and Latin America and the Caribbean FIGURE 4.3 East Asia and Pacific Leads in Developing Country OFDI have reduced their relative shares over time. Latin America and the Caribbean held a share 100 of 37 percent of developing country OFDI 3 6 4 3 during the second half of the 1990s, falling to 12 15 percent during 2010–15.4 And Europe 90 6 4 9 and Central Asia’s share fell to 25 percent 4 15 in 2010–15 from a peak of 36 percent in 80 2000–04.5 Finally, outward flows from Sub- 17 Saharan Africa, the Middle East and North Share of total developing country OFDI (percent) 30 70 Africa, and South Asia maintained more mar- ginal shares across all periods. 37 25 As noted earlier, the BRICS are a key 60 source of developing country OFDI (figure 4.4). These five countries generated 50 35 62 percent of such OFDI in 1995, a share that remained essentially unchanged in 2015. 40 36 These numbers, however, largely align with 15 other aspects of these countries’ participation in the global economy. 6 Aside from the 30 BRICS, other large or relatively higher-income 49 developing countries (for example, Chile, 20 Malaysia, and Mexico) are also top investors 31 32 among developing countries. In fact, when 22 classified across income thresholds (annex 10 4A), developing country OFDI is driven largely by higher-income developing coun- 0 tries. During 1995–99, 78.8 percent of FDI 1995–99 2000–04 2005–09 2010–15 flows from the developing world originated Region in upper-middle-income countries, with Sub–Saharan Africa Middle East and North Africa 13.8 percent from developing high-income South Asia Latin America and the Caribbean countries, 7.1 percent from lower-middle- Europe and Central Asia East Asia and Pacific income, and only 0.3 percent from low- Source: Calculation based on UNCTAD. income countries. Such relative shares did not Note: OFDI = outward foreign direct investment. change much during 2010–15 when OUT WARD FDI FROM DEVELOPING COUNTRIES 105 upper-middle-income countries accounted for China in particular has become the main 79.9 percent of total developing country driver of developing country OFDI, accounting OFDI stocks, high-income countries for for 36 percent of the total (figure 4.4). When 11 percent, lower-middle-income for measured across flows, Chinese OFDI sus- 8.7 percent, and low-income countries for tained a steady upward trend since 2004— 0.3 percent. In this way, upper-middle-income moving from 10 percent of total developing and high-income countries have consistently country OFDI flows to 49 percent in 2015. accounted for the vast majority of developing China is also the main reason for the country OFDI. rise of East Asia and Pacific as the leading FIGURE 4.4 Top Developing Country Outward Investors Year Country 2015 China 36.0 Russian Federation 9.0 Brazil 6.5 South Africa 5.8 Mexico 5.4 India 4.9 Malaysia 4.9 Chile 3.1 Thailand 2.4 Saudi Arabia 2.3 Colombia 1.7 Turkey 1.6 Philippines 1.5 Hungary 1.4 Argentina 1.3 Indonesia 1.1 Poland 1.0 Greece 0.9 Venezuela, RB 0.9 Kazakhstan 0.8 1995 Brazil 30.7 South Africa 16.1 China 12.3 Argentina 7.4 Indonesia 4.1 Malaysia 3.5 Mexico 2.9 Venezuela, RB 2.4 Russian Federation 2.3 Greece 2.0 Saudi Arabia 2.0 Chile 1.9 Thailand 1.6 Turkey 1.0 Colombia 0.7 Poland 0.4 India 0.3 Philippines 0.2 Hungary 0.2 Kazakhstan 0.0 Share of total developing country OFDI stock (percent) Region Sub–Saharan Africa Middle East and North Africa South Asia Latin America and the Caribbean Europe and Central Asia East Asia and Pacific Source: Computation based on UNCTAD. Note: OFDI = outward foreign direct investment. 106 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 developing region generating OFDI (figure 4.3). countries are internationalized through The country has gone from accounting for OFDI. This ratio shows that developing 40 percent of East Asia and Pacific OFDI flows country OFDI is a relatively recent phenome- during 1995–99 to 75 percent in 2010–15. The non: in 1995, 87 out of 135 developing coun- dynamism of Chinese OFDI reflects a unique tries had a positive OFDI stock. Yet virtually institutional and regulatory framework that all developing countries had very low ratios supports firm internationalization (box 4.1). of OFDI to GDP with only three economies A different set of countries emerges if (Botswana, Nigeria, and South Africa, all in OFDI activity is assessed relative to the size Sub-Saharan Africa) having stocks above of the national economy. The ratio of OFDI 10 percent of GDP. A more diverse picture stock to gross domestic product (GDP) 7 emerges in 2015, with 109 developing (map 4.1) reveals the extent to which countries having positive OFDI stocks and, BOX 4.1 The Evolving Role of OFDI in China’s Economy OFDI from China accounts for more than a third of the services sector. A decade later, these distributions all developing country OFDI stock, and the country fl ipped: during 2013–15, 26 percent of Chinese OFDI has been at the vanguard of OFDI policy reform. fl ows targeted the primary sector while 47 percent Trends in Chinese OFDI are remarkable. From 2000 targeted the service sector. This reversal can partly be to 2015, its OFDI flows on average more than dou- explained by the evolution in Chinese OFDI motiva- bled each year (UNCTADstat) so that, by 2016, it had tions, moving from initially natural resource–seeking attained two milestones: OFDI overtook inward FDI to increasingly market-seeking, efficiency-seeking, for the first time, and Chinese OFDI flows were the and finally strategic asset–seeking. Chinese firms second highest in the world after the United States. increasingly see OFDI as a means for opening new This meant that China generated the sixth-largest markets for excess domestic capacity and for acquir- OFDI stock (UNCTAD 2017). Nevertheless, in terms ing hard-to-develop capabilities faster and more of the ratio of OFDI to GDP, China’s OFDI exposure cheaply than developing these indigenously. The goal is still below some of the most outwardly invested is to continue domestic upgrading and increase inter- developing economies in the world (map 4.1 and fig- national competitiveness. ure 4.5). This change in OFDI distribution can also be What accounts for this dramatic growth? Chi- explained partly by differences in OFDI behav- nese OFDI has been driven by both push and pull ior between state-owned enterprises (SOEs) and forces. On the one hand, macroeconomic conditions privately owned enterprises (POEs), and the increas- pushed fi rms out of the domestic market—initially ingly important role of POEs in OFDI. Evidence balance-of-payment surpluses and later domestic shows Chinese SOEs are willing to invest in politi- overcapacity—making investment abroad a policy cally risky host economies to acquire assets in line priority. On the other hand, key inputs to sustain with national priorities (for example, securing natu- domestic growth pulled firms abroad—initially ral resources) (Amighini, Rabellotti, and Sanfi lippo securing essential commodities and later procuring 2013). In contrast, Chinese POEs behave as private knowledge and technology—as China’s development firms do in other countries—seeking to maximize strategy sought to move the country from a manufac- profits and minimize risk—and avoid risky invest- turing-driven to an innovation-driven economy. ment climates. Reflecting a growing domestic pri- The sector breakdown of Chinese OFDI has, as vate sector in China, POEs are becoming increas- a result, undergone major transformation. During ingly important as drivers of OFDI, contributing to 2003– 05, 65 percent of Chinese OFDI flows tar- growing market and strategic asset–seeking OFDI in geted the primary sector while 18 percent targeted developed economies (Dollar 2016; Lardy 2014). In box continues next page OUT WARD FDI FROM DEVELOPING COUNTRIES 107 BOX 4.1 The Evolving Role of OFDI in China’s Economy (continued) 2006, SOEs held 81 percent of China’s OFDI stock, 2016 but averaged less than US$8 billion a month while POEs held only 19 percent; 10 years later, Chi- during January–June 2017 (Hanemann, Lysenko, na’s OFDI stock was almost evenly divided between and Gao 2017). While POE OFDI had been rising as SOEs (50.4 percent of nonfi nancial assets) and POEs a share of total OFDI, the tightening in regulations (49.6 percent) (Wang 2017). Looking specifi cally at seems to favor SOEs, perhaps because they are bet- Chinese OFDI into the United States (the largest des- ter able to navigate the changing political context: tination market for Chinese OFDI), POEs accounted in the fi rst half of 2017, there were virtually no large for nearly 80 percent of OFDI in both 2015 and 2016, private sector M&A deals, and state-related compa- even as Chinese OFDI into the United States tripled in nies accounted for 60 percent of total deals by value, this single year (Rosen and Hanemann 2017). a reversal of the 2016 pattern (Hanemann, Lysenko, T hese pat terns of Chinese OF DI should be and Gao 2017). While M&A OFDI has fallen in most understood in the context of an evolving and increas- sectors, OFDI into the primary sector, high-tech ingly sophisticated OFDI regulatory framework. industries, and modern services (telecom, media, and Between 2001 and 2014, China gradually liberalized computing) has proven most resilient, reflecting the OFDI regulations, moving from a restrictive to a strategic importance of these three areas in China’s supportive framework (Sauvant and Chen 2014). In development strategy. 2014, the regulatory framework matured to embrace China’s increasing use of OFDI to source advanced corporate social responsibility when investing knowledge and technology has also generated grow- abroad, such as the environmental and social impact ing political economy tensions with some developed on host economies. Then, at the end of 2016, the economies, notably the United States and European government announced plans to tighten the inspec- Union. To give a sense of these growing pressures, tion and supervision of Chinese OFDI, especially in only the fi rst half of 2016, China invested more when not related to the core business of the investing in Europe than in the previous three years com- fi rms, or in areas with limited economic value for the bined and often targeted cutting-edge technology. home economy (for example, OFDI in fi lm studios or This sparked European concerns over the long-term sports clubs). This also includes plans for identify- impact on host economies. The lack of market-access ing industries in which Chinese SOEs cannot invest reciprocity for investment—with developed econo- (a “negative list”), such as heavily polluting indus- mies much more open to Chinese OFDI than vice tries (China Daily 2017a). Similar to the changes in versa—has prompted calls for a more level playing 2014, which added a quality dimension to the way fi eld. In February 2017, Germany, France, and Italy that Chinese OFDI was carried out, Chinese policy presented the European Commission with a common has recently added a quality dimension to the sectors position on screening foreign investments, implicitly to which OFDI is targeted. targeting Chinese OFDI and drawing on practices This recent regulatory tightening has had a large in Australia, Canada, Japan, and the United States effect on Chinese OFDI. Chinese mergers and acqui- (Grieger 2017). In early 2017 China decided to open sitions (M&A) transactions fell by 20 percent in the more sectors to FDI (for example, automation, digiti- fi rst six months of 2017 relative to the same period zation, fi nancial services, transportation, and renew- a year earlier (Hanemann, Lysenko, and Gao 2017). able energy) (China Daily 2017b). Then, in August By the middle of 2017, the number of transactions 2017, China started requiring that state groups assess had returned to almost the same level as in the pre- political risks to OFDI before proceeding with any tightening period, yet the average deal size had fallen deal (FT 2017). It is too soon to tell whether these dramatically owing to greater scrutiny of large trans- measures, coupled with implementation of any poten- actions. The value of announced OFDI acquisitions tial new screening mechanisms, will alleviate politi- averaged more than US$15 billion a month during cal economy tensions. 108 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 MAP 4.1 More Developing Countries Engage in OFDI 1995 OFDI Stock/GDP 0% 5% 10% 15% >20% 2015 OFDI Stock/GDP 0% 5% 10% 15% >20% IBRD 43087 | AUGUST 2017 Source: Computation based on UNCTAD and World Development Indicators, World Bank. Note: The five color thresholds correspond to shares of OFDI stock over GDP that are 0–5 percent, 5–10 percent,10–15 percent,15–20 percent, and greater than 20 percent. GDP = gross domestic product; OFDI = outward foreign direct investment. OUT WARD FDI FROM DEVELOPING COUNTRIES 109 FIGURE 4.5 Developing Countries Most Internationalized through OFDI South Africa 51.8 Malaysia 46.2 Togo 43.1 Trinidad and Tobago 38.8 Chile 36.3 Hungary 31.6 Azerbaijan 28.9 Estonia 27.0 Lebanon 26.8 Tonga 24.3 Angola 22.6 Russian Federation 18.9 Thailand 17.2 Colombia 16.2 Philippines 14.1 Greece 13.6 Mexico 13.3 Kazakhstan 12.9 Slovenia 12.8 Mauritius 12.4 0 5 10 15 20 25 30 35 40 45 50 55 OFDI stock over GDP (percent) Region East Asia and Pacific Europe and Central Asia Latin America and Caribbean Middle East and North Africa Sub–Saharan Africa Source: Computation based on UNCTAD and World Development Indicators, World Bank. Note: GDP = gross domestic product; OFDI = outward foreign direct investment. more important, with 26 of these countries inward FDI stock from developing coun- having an OFDI-to-GDP ratio of 10 percent tries held by other developing countries or greater. The list of countries with the high- (map 4.2)8 has risen for many economies. In est values of this ratio (figure 4.5) includes 2001, only 11 developing countries (5 in Sub- low-, lower-middle-, and upper-middle-income Saharan Africa, 5 in Europe and Central Asia, economies, suggesting greater heterogeneity 1 in Latin America and the Caribbean) had across countries’ economic size or develop- half or more of their inward FDI stock owned ment levels. In all, this relative measure by other developing countries. In 2012, that reveals a set of economies actively engaged in number reached 55 countries. Developing outward investment that are generally absent countries are a particularly key source of FDI from the debate on OFDI, owing to their for countries in Sub-Saharan Africa, Europe marginal role in aggregate FDI. and Central Asia, and South Asia. With many of these host economies characterized by low economic development,9 these trends seem to Where? Source–Host FDI Relationships conform with the literature that finds devel- The rise of OFDI by developing country oping country OFDI to be less discouraged by MNCs has also expanded the number of weak institutional and economic environ- countries increasingly dependent on this ments in host countries (Cuervo-Cazurra source of external capital. The share of 2008; Ma and Van Assche 2011). 110 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 MAP 4.2 Exposure to Developing Country OFDI Rises for Many Developing Host Economies 2001 0% 20% 40% 60% 80% 100% 2012 0% 20% 40% 60% 80% 100% IBRD 43088 | AUGUST 2017 Source: Computation based on UNCTAD. Note: The five-color thresholds correspond to ratios of inward FDI from developing countries over total inward FDI stocks that are less than 20 percent, 20–40 percent, 40–60 percent, 60–80 percent, and 80–100 percent. OFDI = outward foreign direct investment. OUT WARD FDI FROM DEVELOPING COUNTRIES 111 FIGURE 4.6 The Location of Developing Country OFDI Varies across Regions 100 4 12 90 23 29 23 33 80 70 Share of total OFDI stock (percent) 23 60 53 72 50 40 75 69 30 65 54 20 34 10 23 0 East Asia Europe Latin America Middle East South Asia Sub–Saharan and and and and Africa Pacific Central Asia Caribbean North Africa Home developing region Other developing region Developed countries Source: Computation based UNCTAD. Note: OFDI = outward foreign direct investment. The geographical distribution of develop- and Latin America and the Caribbean, ing country OFDI across regions (figure 4.6) the share of OFDI remaining in the same suggests the trade-off that developing country region is also relevant. This “regional bias” multinationals face when deciding where to owes to the preference of such regional MNCs locate their investments. For example, OFDI for the lower transaction costs of operating in from South Asia, Europe and Central Asia, markets characterized by cultural ties, geo- and Latin America and the Caribbean is rela- graphical proximity, or prior trade relations10 tively concentrated in developed economies. (Aykut and Goldstein 2006). In all, the geo- For South Asia, developed economies account graphical distribution of developing country for 75 percent of its total 2012 outward stock; OFDI suggests the trade-off that developing for Europe and Central Asia, 69 percent; and country multinationals face when deciding on for Latin America and the Caribbean, 65 per- a location for their subsidiaries—that is, cent. The importance of developed economies weighing the benefits of investing in close, as destinations for developing country MNC familiar markets against the cost of weak con- investments can be attributed to the size and sumer demand or an inefficient institutional strength of these host markets, a key FDI loca- environment. tion determinant (Assunção, Forte, and Is OFDI by developing country firms influ- Teixeira 2011). For Europe and Central Asia enced by this trade-off between market size 112 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 and strength, and physical and cultural dis- embraced by much of the developing world in tance? Our econometric analysis (annex 4B) previous decades, which attracted FDI into extends the analysis in Gómez-Mera and these sectors (Sader 1993). More recently, others (2015), a study that explains the OFDI OFDI into knowledge-intensive industries, patterns of four emerging economies (Brazil, both in manufacturing and services (for India, the Republic of Korea, and South example, pharmaceuticals, software, and Africa), to a sample of 133 developing information technology [IT] services) has countries.11 Our results show that OFDI by gained traction (Gammeltoft 2008). OFDI is developing country MNCs seeks to balance thus a tool to acquire superior technology and market attractiveness with the transaction contribute to firms’ international competitive- costs associated with distant and unfamiliar ness. All things considered, the rich sectoral markets. On the one hand, measures of host distribution of developing country OFDI sug- country market size (population, per capita gests an equally rich set of investment motiva- GDP) are significant predictors for the loca- tions, with all developing regions participating tion of OFDI. On the other hand, transaction to some degree in outward natural resource– costs associated with geographical distance seeking, efficiency-seeking, market-seeking, and the lack of a shared language or colonial and strategic asset–seeking investments. experience between source and host economy Based on the number of FDI projects dur- limit the prospects of cross-border invest- ing 2003–15, companies from most develop- ments by developing country MNCs. ing regions show a slight preference for greenfield FDI rather than for acquisitions.14 This confirms the same bias found in previous What and How? Sector and studies (Davies, Desbordes, and Ray 2015; Mode of Entry UNCTAD 2015). Yet the pro-greenfield bias The sector distribution suggests an increas- is stronger for OFDI from developed econo- ingly rich set of investment motivations mies (figure 4.7): out of 39 industries, OFDI guiding OFDI patterns. The cumulative from developed countries accounts for a OFDI value between 2003 and 2015 12 majority share of greenfield operations in 25 (annex 4C) is relatively evenly distributed of them, with a median share of 58 percent. across broad sectors (primary, manufactur- On the other hand, developing country OFDI ing, and services). But service sectors account is biased toward greenfield in only 20 indus- for a large share of OFDI stock in almost all tries, with a median share of 50 percent. regions, ranging from 36 percent (Europe The relative preference for M&A in devel- and Central Asia) to 41 percent (East Asia oping country OFDI—when compared to and Pacific). Europe and Central Asia and that of advanced economies—is more evident Sub-Saharan Africa also strongly favor in knowledge-intensive manufacturing indus- extractive industries, which account for tries 15 (figure 4.7): of the nine industries about 40 percent of outward stocks. Thus, where developing country OFDI shows a pro- manufacturing industries13 tend to be under- M&A difference of 15 percentage points or represented in these two regions. more (relative to OFDI from developed econ- The relatively balanced sectoral distribu- omies), seven are technology- and knowledge- tion suggests that developing country OFDI is intensive16 (automotive components, business increasingly complex. Previous attempts to machines and equipment, engines, transporta- d i s e n t a n g l e O F D I ’s s e c t o r p a t t e r n s tion original equipment manufacturer, space (Gammeltoft 2008) found a particularly high and defense, and semiconductors). preference for service sectors over manufac- The previous trends suggest the impor- turing or natural resources. Such a bias tance of OFDI as a mechanism for toward services was partly attributed to the upgrading in manufacturing by develop- wave of privatization of public services ing country MNCs. A crucial aspect of OUT WARD FDI FROM DEVELOPING COUNTRIES 113 FIGURE 4.7 Developing Country Manufacturing MNCs Prefer Investing via M&A Developed countries Developing countries Developing country OFDI entry mode bias Primary Coal, oil, and natural gas 63 37 64 36 1 pp Metals 54 46 50 50 4 pp Minerals 60 40 38 62 –22 pp Manufacturing Aerospace 62 38 84 16 21 pp Automotive components 61 39 44 56 –17 pp Automotive OEM 100 100 0 pp Beverages 13 87 15 85 3 pp Biotechnology 12 88 31 69 20 pp Building and construction materials 39 61 50 50 11 pp Bussiness machines and equipment 66 34 57 43 –9 pp Ceramics and glass 47 53 76 24 29 pp Chemicals 64 36 80 20 16 pp Consumer electronics 65 35 77 23 12 pp Consumer products 53 47 54 46 1 pp Electronic components 67 33 65 35 –2 pp Engines and turbines 50 50 55 45 5 pp Food and tobacco 38 62 44 56 6 pp Industrial machinery, equipment, and tools 34 66 53 47 19 pp Medical devices 17 83 40 60 22 pp Nonautomotive transport OEM 64 36 65 35 1 pp Paper, printing, and packaging 38 62 69 31 31 pp Pharmaceuticals 17 83 30 70 13 pp Plastics 64 36 52 48 –12 pp Rubber 58 42 71 29 14 pp Semiconductors 70 30 19 81 –51 pp Space and defence 31 69 66 34 35 pp Textiles 73 27 71 29 –2 pp Wood products 36 64 72 28 36 pp Services Alternative/renewable energy 95 5 5 91 9 –4 pp Business services 34 66 43 57 8 pp Communications 31 69 40 60 8 pp Financial services 20 80 34 66 14 pp Health care 13 87 22 78 9 pp Hotels and tourism 69 31 75 25 5 pp Leisure and entertainment 38 62 34 66 –4 pp Real estate 53 47 70 30 17 pp Software and IT services 34 66 52 48 18 pp Transportation 20 80 32 68 12 pp Warehousing and storage 88 12 78 22 –10 pp M&A Greenfield Source: Computation based on fDi Markets database, the Financial Times; and Thomson Reuters. Note: For both developed and developing countries, the figure shows the mode of entry distribution of the cumulative number of OFDI projects between 2003 and 2015. The last column shows the deviation in percentage points of developing country OFDI modes of entry, with positive (negative) values identifying a greenfield (M&A) bias for the OFDI of developing countries relative to developed economies. IT = information technology; M&A = mergers and acquisitions; MNC = multinational corporation; OEM = original equipment manufacturer; OFDI = outward foreign direct investment. knowledge-intensive industries is their reliance In sum, our data analysis reveals the fol- on intangible assets, involving largely tacit and lowing main trends: experiential knowledge in such areas as research and development (R&D), branding, • OFDI by developing country firms is an or organizational software. These features increasingly important source of global make intangible assets difficult to replicate investment flows and stocks. (OECD 2013). M&A is therefore the only • The main source of developing coun- means of acquiring the type of knowledge or try OFDI across developing regions is intangible asset that is inherent to the target East Asia and Pacific. In absolute terms, firm (Slangen and Hennart 2007). BRICS investors are the key drivers of 114 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 developing country OFDI, accounting for The rest of this chapter will address these 62 percent of total developing country possibilities by reviewing the literature on OFDI stock in 2015—with China alone OFDI home effects. accounting for 36 percent. • The countries with a high OFDI-to-GDP ratio are far more heterogeneous, both Does Development Level Affect across countries’ economic sizes and development levels. OFDI Behavior? • As for regional differences in the geo- Both the investor survey and the gravity graphical location of developing coun- model estimation (annex 4B) suggest that try OFDI: South Asia and Europe and OFDI by developing country MNCs reacts Central Asia channel more than two- to standard host economy location determi- thirds of their OFDI stock to developed nants (for example, market size, income economies, while the Middle East and level, distance, common language, colonial North Africa and Sub-Saharan Africa links) in much the same way as developed concentrate, respectively, 76 percent and country OFDI: both are attracted to large 65 percent of outward stock in develop- and growing economies that are geographi- ing countries. In general, the geographi- cally close and culturally similar. However, cal distribution of developing country evidence suggests that developing country OFDI suggests that developing coun- investors are relatively more willing to target try MNCs balance the importance of smaller and closer economies (Arita 2013) in market size with physical and cultural a “stepping-stone” strategy. Some of these proximity. firms find it difficult to compete in larger, • Relative to OFDI from developed coun- more competitive markets farther away, tries, developing country OFDI shows lacking the networks and experience of greater reliance on M&A when targeting developed country firms. Studies of Asia and manufacturing industries. This is especially Latin America find that investors usually true for knowledge-intensive industries, as expand into large and complex markets only developing country MNCs resort increas- after first successfully expanding in smaller, ingly to OFDI to augment capabilities and lower-income economies in the same region competitiveness. (Cuervo-Cazurra 2008; Gao 2005; • Finally, developing country OFDI is dis- Hiratsuka 2006). tributed across a rich set of industries, Differences between developed and devel- including manufacturing, extractives, oping country outward investment behavior and services. It thus covers the full range also arise with regard to the role of technol- of investment types (natural resource– ogy. Developed countries generally exploit seeking, efficiency-seeking, market- existing technological assets in undertaking seeking, and strategic asset–seeking). OFDI. But some developing country MNCs use OFDI to acquire new technological assets. As more developing countries continue to Case studies of leading BRICS firms provide internationalize through OFDI, a pertinent examples (Holtbrugge and Kreppel 2012; question is the role that OFDI can play in sup- Rodriguez-Arango and Gonzalez-Perez 2016; porting domestic development. Developing UNCTAD 2005). The reason is that most countries may be able to leverage OFDI to BRICS multinationals face disadvantages in source technology, increase domestic capacity, terms of patents, management know-how, or upgrade production processes, boost competi- cutting-edge processes, and thus seek to tiveness, augment managerial skills, and access acquire these abroad as part of a strategy of distribution networks (Amann and Virmani late-comer catch-up. Looking at the econo- 2014; Driffield and Love 2003, 2007). metric evidence, however, this seems to apply OUT WARD FDI FROM DEVELOPING COUNTRIES 115 mostly to China. Across many studies, a con- only the MNC will directly experience the sensus has emerged that Chinese MNCs use impact of investing abroad (first-order effect). OFDI to acquire the knowledge, skills, and Later, the firm’s enhanced knowledge, capac- technology they lack (Dong and Guo 2013; ity, and behavior may affect other domestic Huang and Wang 2011; Kang and Jiang firms that are not themselves foreign investors 2012; Ramasamy, Yeung, and Laforet 2012; (second-order effect). Finally, the impact may Zhang and Roelfsema 2014). be spread throughout the home economy over Developing country investors may also be time. relatively more willing to target host econo- OFDI can impact the home economy in at mies with weaker institutional quality,17 in least three ways: view of the “institutional advantage” argu- ment (Cuervo-Cazurra and Genc 2008). This 1. Scale effects: OFDI allows a firm to grow theory suggests that managers of developing larger than it would have if limited to country MNCs are more used to uncertainty operating in its home market. This and may be more flexible in dealing with growth may yield traditional gains based unpredictable regulatory agencies and corrupt on economies of scale and scope,19 low- government officials. Several studies support ering costs of production and operation. this argument, finding that developing coun- 2. Competition effects: Competition with try MNCs are relatively more present in least fi rms in foreign markets where develop- developed countries (Cuervo-Cazurra and ing country fi rms invest may force them Genc 2008) or by demonstrating an inverse to improve efficiency and upgrade pro- relationship between host political risk and, duction processes. Competition in host specifically, Chinese OFDI (Buckley and oth- markets can thus bring efficiencies and ers 2007; Cui and Jiang 2009; Duanmu and expansion of developing country firm Guney 2009; Kang and Jiang 2012; Quer, activities at home. Claver, and Rienda 2012). 3. K n owledge ef fec t s: OF DI enables fi rms to acquire knowledge directly, as through M&A, joint ventures, or other Does OFDI Matter for forms of partnership. Knowledge can Development? Identifying take the form of technology, production techniques, or management skills. Such OFDI Home Effects knowledge transfer initially benefits only Developing country OFDI can affect the the foreign subsidiary. For it to benefit home economy of investors through differ- the home economy, it needs to be trans- ent transmission channels. This section first ferred back to the parent firm—so-called considers these channels and then presents reverse knowledge transfer (for example, evidence of these effects across two vari- through personnel exchanges, produc- ables: innovation and exports. tion shifting, or management rotation). A developing country can use OFDI as a At the same time, indirect knowledge catch-up strategy to source technology, transfer may occur through knowledge increase domestic capacity, upgrade produc- spillovers to other firms in the home tion processes, boost competitiveness, aug- economy. ment managerial skills and access distribution networks (Amann and Virmani 2014; These transmission channels can, however, Driffield and Love 2003, 2007). As a result, lead to diverse effects on developing countries’ OFDI can play a major role in a develop- MNCs, as well as on local firms in home ing country’s developmental strategy. 18 markets. Scale and competition effects may The effects of OFDI on the home economy push less competitive firms to exit can show up at three different levels. Initially, the home market. Knowledge effects may only 116 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 accrue to those firms with the capacity to inte- R&D spending by developing country parent grate such knowledge, causing outward companies.22 Host market R&D intensity investment to contribute to skills-based therefore seems to be a key element in deter- inequalities. Rigid factor markets for labor mining the potential for overseas investment and capital may exacerbate adjustment costs, by developing country MNCs to generate while undeveloped factor markets may limit innovation spillovers in the home economy the potential benefits of outward investment (box 4.2). for the home economy (for example, unskilled The evidence also suggests that the effect labor unable to integrate OFDI-generated of outward investment on home innovation is knowledge and innovation or capital market more pronounced in knowledge-intensive sec- imperfections causing OFDI to crowd out tors.23 In the auto and chemical and pharma- domestic investment in the home economy). ceuticals industries, evidence reveals that Appropriate policies are needed to maximize OFDI firms generate reverse technology the benefits of outward investment while min- spillovers to domestic firms that did not invest imizing its costs. abroad.24 The positive effect of OFDI on home R&D is apparent for investments in both developed and developing host coun- OFDI Impact on Innovation tries, although it is stronger for developed countries.25 and Exports South–South OFDI is also showing signs The following review focuses on two key of increasingly becoming a source for home economic benefits where the existing litera- innovation. Whereas previous paradigms ture provides the most evidence of OFDI considered developed countries as the repos- impact on the home economy: driving inno- itory of knowledge and technology, and thus vation and expanding exports. focused on North–North or North–South investment flows, a multipolar global tech- nology network is emerging, with growing OFDI by Developing Country MNCs Can South–South innovation-oriented interac- Spur Innovation at Home tions and collaboration.26 Part of the reason OFDI’s ability to increase innovation in the is that knowledge created in developing home economy is well-documented.20 The key countries may be more adapted to the needs transmission channels are competition effects of other developing countries, and that the that encourage innovation and direct and level of complexity of that knowledge may indirect knowledge effects. Knowledge can be more easily absorbed by other economies take the form of technology, production tech- at similar levels of development. Evidence niques, or management skills. Disaggregating from Africa shows that, when the knowl- outward investment by type is especially edge gap between firms is too great, interac- important, as one particular type of OFDI— tions between firms are less likely to lead to knowledge-seeking, which is part of strategic knowledge transfer or spillovers because asset–seeking investment 21 —is likely to firms are unable to absorb the knowledge have the greatest positive effect on home (Boly and others 2014 in Moran, Gorg, innovation. and Seric 2016; Deng 2010; Farole and Developing country MNCs seem to be Winkler 2014). Using outward investment to using outward investment in innovation- target highly sophisticated knowledge so as intensive economies to spur home innovation. to leapfrog to the knowledge frontier may One study examines OFDI from 20 develop- therefore not be an effective strategy until a ing countries into developed countries from firm has first increased its absorptive capac- 2000 to 2008 (Chen, Li, and Shapiro 2012). ity. Different levels of development may thus It finds that both R&D employment and call for different OFDI knowledge acquisi- R&D expenditure in host economies increase tion and innovation strategies depending on OUT WARD FDI FROM DEVELOPING COUNTRIES 117 BOX 4.2 Developing Country MNCs Use OFDI to Boost Innovation and Exports Across the developing world, firms are using outward Hikma now has manufacturing facilities approved by investment to improve their capabilities and perfor- the U.S. Food and Drug Administration in Germany, mance. Particularly noteworthy is the breadth of dif- Italy, Jordan, Portugal, Saudi Arabia, and the United ferent industries involved. Three industries in three States; it also has R&D centers in Algeria, the Arab different countries illustrate how outward investment Republic of Egypt, Jordan, Saudi Arabia, Tunisia, and can boost home-firm innovation, exports, and firm the United States. Hikma has thus become the third growth. largest generic injectable supplier to the U.S. market. In Turkey, two of the leading household appliance According to the Jordanian Association of Pharma- fi rms have used outward investment to locate R&D ceutical Manufacturers, about 80 percent of Jordanian activities in foreign markets to increase parent-fi rm production is destined for export to more than innovation. The leading firm, Arcelik, has seven sixty countries, with most exports heading to other R&D centers around the world. This emphasis on Arab countries. R&D means that in 2015 the fi rm had by far the most China’s wind turbine industry illustrates how out- World Intellectual Property Organization (WIPO) ward investment can drive innovation in the home patent applications among all Turkish fi rms—a stag- market and the key role that supportive policies gering eight times more than the second highest Turk- can play. China’s wind power capacity in 2005 was ish fi rm—placing Arcelik in the 78th position glob- 1,260 megawatts; by the end of 2016, it had grown ally. Another of the top Turkish fi rms, Vestel, is also more than 100-fold to 168,690 megawatts (Global using outward investment to tap into foreign technol- Wind Energy Council 2016). The International Energy ogy and boost innovation. It devotes 2 percent of sales Agency estimates that China builds two wind turbines revenue to R&D spending, with foreign R&D centers every hour. As a result, China now has more installed in the United Kingdom and China. As a result, Vestel wind power capacity than all of the European Union has also been listed as one of the three Turkish com- combined, and more than double the capacity of the panies among the top 1,000 companies in the world United States. OFDI has played a key role in facilitat- by R&D spending. ing this remarkable growth by helping to access tech- Jordan’s pharmaceutical sector provides an excellent nology. From 2009 to 2014 China made 44 outward example of how a relatively smaller developing coun- investments in the wind energy industry. The Chinese try can use outward investment to develop a domes- state guided and facilitated this process through pol- tic industry’s capacity and competitiveness. Al Hikma icy instruments such as subsidies, tax incentives, R&D Pharmaceuticals, Jordan’s largest pharmaceutical spending, technical partnership, and outward invest- fi rm, has led a series of M&A and greenfield invest- ment fi nancial incentives and support. This represents ments across the world, in both developed and devel- a dramatic example of a developing country using pol- oping countries, to access technology and markets. icy measures to leapfrog developed economies. the economy’s absorptive capacity (Criscuolo for increased export-oriented production of and Narula 2008). either intermediate or finished goods. Outward investment may also bring back to the home economy knowledge and technology Overseas Investment by Developing that boost export competitiveness. OFDI may Country MNCs Can Expand Home also be used to plug into GVCs through Exports backward and forward supply-chain integra- Empirical evidence confirms that outward tion, stimulating exports of intermediate investment increases home country exports. inputs. Yet negative effects may arise if relo- The key transmission channels are scale effects cating production abroad lowers exports of and knowledge effects: outward investment final goods and services since foreign markets may open new markets, creating opportunities are now being served by local production. 118 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 The net effect is therefore theoretically ambig- developed countries on job growth and eco- uous, depending on the relative strength of nomic activity, the literature on developing these different effects. country OFDI is more nascent and still offers In practice, however, empirical evidence only tentative conclusions, the review of overwhelmingly confirms that outward invest- which is beyond the scope of this report. ment and home exports are complements and not substitutes, and that OFDI increases home exports (box 4.2). For example, looking at Malaysia, the Philippines, Singapore, and Absorptive Capacity Is Key Thailand from 1981 to 2013, a recent study While OFDI can generate benefits for home finds that in all cases OFDI increases rather economies, limitations on firm-level and than substitutes home country exports.27 In this economy-wide absorptive capacity in devel- study, a 1 percent increase in OFDI leads to a oping countries may limit OFDI home effects $750 million rise in exports for the Philippines, (box 4.3). $72 million for Singapore, $41 million for Absorptive capacity can affect the home Thailand, and $31 million for Malaysia. effects of OFDI in two divergent ways. One Time horizon may be an important dimen- view is that firms farthest from the technology sion in determining the effect of OFDI on frontier may benefit most from spillovers as home-country exports. A longer time horizon they are starting from a low base. Another may allow more time for adjustments through view suggests that these firms may not have the the different transmission channels, and capacity to make the best use of new technolo- thereby have larger effects. Evidence for this is gies. Rather, it argues that firms closest to the provided by European Union exports, where technology frontier are best placed to adopt growth in outward investment caused small, cutting-edge technologies available through positive effects on exports in the short term OFDI.29 Empirical evidence supports both but with long-run effects that were consistently views, indicating a U-shape function in the greater than their short-run equivalents.28 relationship between absorptive capacity and When it comes to other potential home OFDI home effects, with simple knowledge at country benefits—such as productivity, the low range and complex knowledge at the domestic investment, employment and, ulti- high range being more likely to facilitate these mately, economic growth—the literature effects (Girma 2005; Girma and Gorg 2007). is still inconclusive. While research has The key to positive home effects is a match found a mostly positive effect of OFDI from between the firm’s level of absorptive capacity BOX 4.3 Absorptive Capacity Matters at Both Firm and Economy Levels Absorptive capacity is defined as the “ability to iden- At the economy level, absorptive capacity depends tify, assimilate, and exploit knowledge from the envi- on whether frameworks and mechanisms exist to help ronment” (Cohen and Levinthal 1989). It applies at fi rms integrate knowledge resources and develop link- both the level of the individual firm and the level of the ages and learning between fi rms. Measures to boost overall economy. At the firm level, absorptive capacity economy-wide absorptive capacity can include estab- is a function of how effectively a firm can productively lishing institutional partnerships, helping to diffuse integrate knowledge resources. Measures to boost firm- information, promoting fi rm linkages, and designing level absorptive capacity can include instituting training school curricula. These measures will largely depend programs, increasing R&D spending, and/or develop- on decisions by policy makers. ing knowledge management tools. These measures will largely depend on decisions by individual firms. OUT WARD FDI FROM DEVELOPING COUNTRIES 119 and the knowledge it seeks to target through Economies Are Gradually OFDI. Firms starting from a more basic level of knowledge can benefit most from exposure to Liberalizing OFDI Regulations simpler knowledge, giving them potentially a In many developing countries, OFDI policy bigger boost to their innovation than if they has shifted gradually from restrictive to more were to target knowledge at the frontier and not supportive, although restrictions persist be able to absorb it. In contrast, firms already (figure 4.8). In 2015, almost half of develop- enjoying more sophisticated knowledge can ing countries (49 percent, or 77 out of benefit most from exposure to more complex 156 countries) had some OFDI restrictions in knowledge at the frontier, giving them a bigger place. Low-income developing countries31 boost to innovation than if they were to target were more likely to restrict OFDI than other knowledge they already have. In both cases, the developing countries. In 2015, 60 percent of ability of the home firm to absorb knowledge low-income developing countries had and the kind of knowledge being targeted must OFDI restrictions (36 of 60 countries); in match. This match will change over time as contrast, only 43 percent of non-low-income knowledge is gained and absorptive capacity developing countries had OFDI restrictions increases. At some point, the developing country (41 of 96).32 This original finding that OFDI firm should have sufficient absorptive capacity restrictions vary with development level to invest in acquiring knowledge at the frontier. accords with earlier work on foreign Governments may therefore wish to ensure that exchange restrictions to FDI across econo- their efforts to boost absorptive capacity take mies, which found that all high-income coun- into account different needs at both ends of the tries maintain unrestricted foreign exchange spectrum of the private sector.30 regimes for FDI.33 Therefore as countries Absorptive capacity may be measured at raise their development level, restrictions on both the level of the firm and of the economy. outward investment seem to fall, although When undertaking OFDI decisions, the firm’s the direction of this relationship requires absorptive capacity is key to determining the further study. appropriate match with target knowledge and Restrictive regulatory frameworks regard- technology. But policy interventions to boost ing OFDI stem from concern that capital out- absorptive capacity should be considered at flows can worsen the balance of payments and the economy level. Officials can adopt mea- capital availability in the home economy. sures that boost the absorptive capacity of Measures to restrict OFDI can take the form of whole sectors—such as training programs, approval requirements, reporting require- infrastructure provision, and network ments, foreign exchange controls, ceilings on creation—rather than try and boost the investment amounts, or limits on destination absorptive capacity of individual firms through sectors or destination economies (Kuz ´min´ska- subsidies or protectionist measures in order to Haberla 2012). A snapshot of OFDI restric- create national champions (Moran 2015). tions in 2011 in 84 developing countries reveals Finally, differences in absorptive capacity great variation in OFDI restrictiveness, even between developed and developing countries for countries at similar levels of development are not caused by structural variables but (Sauvant and others 2014). may simply reflect different stages of develop- The BRICS provide a representational ment (Ramamurti 2012). Developed coun- picture of variation in OFDI regulation: tries have been building their absorptive capacity for longer, whether through training, • China , over the course of 2000–14, R&D, linkages, or institutional partner- moved from restricting to encouraging ships—all of which can be fostered through OFDI, although it tightened restrictions policy intervention. Government policies can again at the end of 2016 (box 4.1). thus help developing countries catch up by • Brazil has generally favored OFDI, and boosting their absorptive capacity to maxi- in 2007 adopted fi nancial incentives to mize the positive effects of OFDI (box 4.3). encourage it in specific sectors in which 120 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 4.8 Developing Countries Have a Mixed Record on OFDI Restriction 100 90 80 Share of countries with OFDI restrictions (percent) 70 60 50 40 30 20 10 0 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Year Emerging market and developing countries Low–income developing countries Source: Computation based on IMF Annual Report on Exchange Arrangements and Exchange Restrictions (IMF 2016) Note: This figure uses IMF country category definitions. The IMF defines low-income developing countries (LIDCs) as those with a level of per capita Gross National Income (GNI) less than the Poverty Reduction and Growth Trust (PRGT) income graduation level for non-small states (that is, twice the Interna- tional Development Association (IDA) operational threshold, or 2 x IDA-OT) (see IMF 2014). Emerging market and developing countries are all developing countries that are not LIDCs (see IMF 2016). OFDI = outward foreign direct investment. the Brazilian economy had a comparative manufacturing, education, and hospitals advantage (for example, mining, petro- requires prior approval by the Reserve leum, pulp and paper, and beef) (Nunes Bank of India. Restrictions also apply on de Alcântara and others 2016). how OFDI is carried out in neighboring • Russia has also generally welcomed countries (for example, Bhutan, Nepal, OFDI, mostly in the energy sector, but it and Pakistan).35 Quantitative restrictions has also blocked individual deals (Fortes- are also set by the net worth of the Indian cue and Hanson 2015). fi rm. If OFDI is approved, the fi rm must • India maintains a relatively restrictive submit annual performance reports on OFDI framework, despite recent liberal- each OFDI deal.36 ization. OFDI in real estate34 is forbidden, • South Afric a also rest ric ts OF DI , in financial services is quite restricted, although with its own particular reg- and in energy and natural resources, ulatory conditions. Firms face a limit OUT WARD FDI FROM DEVELOPING COUNTRIES 121 of 1 billion rand per calendar year for (scale effects, competition effects, and knowl- OFDI, above which they must formally edge effects) may play out, augmenting and apply to the South African Reserve Bank accentuating effects on the home economy. To and ensure that at least 10 percent of the understand OFDI, we need to move beyond target entity’s voting rights are obtained thinking of it as having simply a positive or through the investment. Even for deals negative impact on home economies and under the 1 billion rand limit, restric- disaggregate its effects across different tions remain, such as the net sale pro- dimensions. ceeds being repatriated to South Africa OFDI policy should therefore adopt a and South African-owned intellectual holistic approach. It should consider both the property not being sold without prior effects on single variables and on the set of approval.37 variables that policy makers care about. Just as with trade, OFDI will create winners and Given the potential benefits of OFDI to losers, but overall the positive effects on the home economies, developing country govern- home economy may outweigh the negative ments with OFDI restrictions may wish to effects. Concretely, our study suggests the fol- carefully weigh their costs and benefits. lowing policy considerations: Given the growing importance of develop- ing country OFDI, governments can target investment promotion activities not only to Conclusion traditional sources of FDI from developed From the empirical evidence, developing coun- economies, but also to new sources from try OFDI clearly has the potential to contrib- developing economies. South–South and intra- ute substantially to development in home regional developing country OFDI represent a markets. Evidence suggests that OFDI sizable share of total FDI flows. IPAs may increases home innovation and exports, but therefore wish to court developing country conclusive evidence is not yet available regard- OFDI from regional neighbors and develop- ing productivity, domestic investment, employ- ing economies in other regions as a potential ment, and economic growth. One reason may source of investment. This source holds con- be that it is relatively easier to detect effects for siderable promise but has been largely under- variables at the firm or sector level and more emphasized in many investment promotion difficult to do so at the economy level. strategies. Even within a single variable, the effect of Governments may also want to review any outward investment can vary across sectors, restrictions on OFDI, weigh their costs and factors of production, investment types, and benefits, and ensure that these are based on over time. OFDI may, in fact, simultaneously sound policy goals.38 Several of the largest exhibit positive and negative effects across source markets of developing country OFDI these different dimensions. For example, it have recently eased restrictions on OFDI, may benefit high-skilled labor while hurting although restrictions do remain. These con- low-skilled labor; or it may force less com- trols may be based on macroeconomic objec- petitive home firms to exit the market, while tives such as securing financial stability or boosting the productivity and profits of more promoting domestic investment. But the evi- competitive home firms that seize opportuni- dence suggests source countries can also ben- ties or adjust to new realities. Differences may efit from OFDI, and restrictions may only be also arise concerning the time horizon. In the constraining positive home effects. short term, the impact of outward investment Governments can maximize the potential on the home economy may be more limited positive home effects from OFDI by adopting but over time different transmission channels measures that strengthen economy-wide 122 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 absorptive capacity. Given that empirical evi- whether natural resource–seeking, efficiency- dence indicates that absorptive capacity is a seeking, market-seeking, or strategic asset– U-shape function—with simple knowledge at seeking. The effect on the home economy is one end and complex knowledge at the likely to depend on the motivation for under- other—governments may wish to first identify taking OFDI, but no work has yet disentan- the size of the technology gap to tailor the type gled these dynamics. of policy intervention accordingly. Measures In addition, more evidence is needed to consider include boosting R&D expendi- regarding developing country OFDI’s effect ture, providing training programs, promoting on home economy productivity, employment, firm linkages, establishing institutional part- growth, and domestic investment. nerships, helping to disseminate information, Finally, developing country governments and redesigning school curricula. need to better understand how investment Given that OFDI by developing country incentives and other policies affect their firms’ firms has only boomed in the last decade, cur- OFDI decisions. A clearer understanding of rent research is fairly limited and many ques- the dynamics in these three areas would allow tions remain. More work is needed regarding policy makers to better design and implement how home effects vary across OFDI type, OFDI policy interventions. Annex 4A. Developing Country OFDI by Income Category FIGURE 4A.1 Distribution of Developing Country OFDI across Income Categories 0.3 0.1 0.1 0.3 100 7.1 10.6 8.7 14.1 Share of developing country OFDI flow (percent) 80 60 68.6 78.8 72.8 79.9 40 20 20.8 13.8 13.1 11.0 0 1995–99 2000–04 2005–09 2010–15 Income group Low-income Lower-middle-income Upper-middle-income High-income Source: Computation based on UNCTAD and World Development Indicators, World Bank. Note: OFDI = outward foreign direct investment. OUT WARD FDI FROM DEVELOPING COUNTRIES 123 Annex 4B. Estimation of a Gravity The present analysis departs from Gómez- Mera and others (2015) in two main ways. Model on Developing Country First, the use of the United Nations FDI Determinants Conference on Trade and Development FDI This annex presents the details and results of bilateral dataset allows for creation of a panel a gravity model that evaluates the strength of dataset that covers developing countries standard FDI location determinants in guid- engaged in OFDI between 2001 and 2012. ing developing country OFDI. Gravity mod- Second, having a panel dataset influences the els have become a widely used framework choice for the Poisson Pseudo-Maximum for explaining economic relations between Likelihood method39 (PPML), which offers countries. Early empirical applications, dat- several advantages for estimating panel data- ing back to the decade of the 1960s, largely sets with gravity variables (Santos Silva and focused on explaining patterns of bilateral Tenreyro 2006). The following equation illus- trade (Linneman 1966). One of the most trates the baseline econometric specification. robust findings of this research strand is the significance of relative market size, geo- FDIijt = α + β1 GDPPC jt + β 2 POPjt graphical distance, and common cultural + β3 DISTCAPij + β 4Contig ij and institutional features, such as language, colonial experience, or trade agreements, as + β 5Commlang ij + β6Colonyij predictors of trade between two countries. + β 7 BITijt + β8 DISTij * BITijt Taking advantage of the increasing availabil- + β 9 Xijt + β10 Di + eijt , ity of bilateral economic data, the gravity specification has eventually been applied to the study of capital flows, and FDI in partic- where the dependent variable is the flow of ular (Bevan and Estrin 2004; Talamo 2007). FDI between source i and host j in year t. The This gravity exercise follows the empirical specification model includes a categorical inquiry of Gómez-Mera and others (2015), a variable controlling for fixed effects of the study that explains OFDI patterns of four source country40 (D). The host market attrac- emerging economies (Brazil, India, Korea, tiveness variables include per capita GDP in and South Africa) through a gravity specifica- purchasing power parity in current interna- tion. Such specification includes standard tional dollars ( GDPPC ) and population location determinants on host market size (POP). The standard gravity variables are the (GDP per capita, population) and some of the distance between source and host country standard bilateral variables (for example, dis- capitals ( DISTCAP ), a dummy variable tance, common language, colonial links). for source and host country sharing the Thus, it arrives at the following conclusions: same border (Contig), the same language First, the market size of the host economy is a (Commlang), and the same colonial history significant predictor of the outward invest- (Colony). In line with Gómez-Mera and oth- ments of these emerging economies. Second, ers (2015), exports from source to host (X) the lower transaction costs derived from shar- are included to control for the complemen- ing the same language or colonial heritage are tarities between trade and FDI. In addition, a significant determinants of the probability of dummy for a ratified BIT is included, both investing. Third, physical distance between independently and interacted with distance countries reduces the probability of investing. (data definitions and sources are included in Fourth, the existence of bilateral investment table 4B.1). The use of these variables and treaties (BITs) between source and host econ- data sources allows for the creation of a omy is a predictor of OFDI for these coun- panel for 133 developing source countries tries, reducing also the cost derived from and 147 host countries (developed and devel- geographical distance. oping), across the 2001–12 interval. 124 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 TABLE 4B.1 Variables and Data Sources Variable Definition Source FDI Bilateral flow of FDI ($) UNCTAD FDI bilateral dataset GDPPC GDP per capita, PPP (current international $) World Development Indicators POP Total population (million) World Development Indicators DISTCAP Capital-to-capital distance CEPII Contig Dummy variable for source and host country sharing a common border. CEPII Commlang Dummy variable for source and host country sharing the same official CEPII language Colony Dummy variable for source and host country sharing the same CEPII colonial history. BIT Existence of a ratified bilateral investment treaty between two countries. ICSID (2017) X Export value from origin to destination, 3-year moving average of t-2, t-1, t UN Comtrade TABLE 4B.2 Poisson Pseudo-Maximum Likelihood Estimation Results Variables PPML coefficients (1) PPML coefficients (2) GDPPC 0.0220*** 0.0219*** (9.98) (9.72) POP 0.00084*** 0.00083*** (3.17) (3.14) DISTCAP −0.00019*** −0.00021*** (−7.49) (−6.93) Contig 0.3014 0.2967 (1.42) (1.40) Commlang 0.9148*** 0.9069*** (4.04) (4.01) Colony 0.7181** 0.7063** (2.25) (2.27) BIT 0.9210*** 0.7700*** (4.48) (2.82) X 0.0166*** 0.0169*** (6.14) (5.98) DISTCAP*BIT 0.0000348 (0.88) Year 0.2058*** 0.2059*** (12.45) (12.45) Constant −421.40*** −421.51*** (−12.70) (−12.70) Observations 216,009 216,009 Source: Computation using data sources in table 4B.1. Note: ***p<0.01; **p<0.05; * p<0.1. Z statistics in parentheses. The Poisson Pseudo-Maximum Likelihood estimation specifies robust standard errors and origin-destination clustering to allow for clustered standard errors within pairs. The results of the PPML estimation, with largely corroborate the ones found in and without interaction term (table 4B.2), Gómez-Mera and others (2015): both host show that the trade-off between host market attractiveness variables ( GDPPC, market strength and physical and cultural POP ) and the reduced transaction costs proximity remains when the analysis is derived from shared cultural links extended to a comprehensive sample of (Commlang, Colony) are significant predic- developing country FDI sources. The results tors of FDI flows from developing OUT WARD FDI FROM DEVELOPING COUNTRIES 125 countries. Distance, on the other hand, acts in Gómez Mera and others (2015) is the as a significant inhibitor of these flows. role of BITs in reducing the deleterious Thus, BITs are found to be an enabler of effect of distance over FDI flows, with the FDI flows. All things considered, the only interaction term between both variables not result that is in dissonance with those found being significant across any specification.41 Annex 4C. Developing Country OFDI by Industry FIGURE 4C.1 Distribution of Developing Country OFDI across Industries Coal, oil, and natural gas 24.0 Primary Metals 10.6 Other primary 0.4 East Asia and Pacific Other manufacturing 15.8 Manufacturing Food and tobacco 5.2 Automotive OEM 3.2 Real estate 11.9 Other services 10.6 Services Financial services 8.2 Transportation 5.5 Alternative or renewable energy 4.5 Coal, oil, and natural gas 24.5 Primary Metals 11.3 Other primary 0.7 Europe and Central Asia Other manufacturing 18.0 Manufacturing Food and tobacco 3.4 Transportation 11.8 Financial services 9.0 Communications 7.5 Services Real estate 6.9 Other services 4.4 Alternative or renewable energy 2.6 Metals 12.6 Primary Coal, oil, and natural gas 11.9 Latin America and Caribbean Other primary 0.3 Other manufacturing 13.5 Beverages 9.0 Manufacturing Food and tobacco 8.6 Building and construction materials 6.0 Communications 12.1 Financial services 9.0 Services Transportation 8.8 Other services 8.1 0 5 10 15 20 25 30 Share of cumulative FDI outflows (percent) figure continues next page 126 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 4C.1 Distribution of Developing Country OFDI across Industries (continued) Coal, oil, and natural gas 20.7 Primary Other primary 4.5 Middle East and North Africa Chemicals 9.2 Other manufacturing 7.8 Manufacturing Plastics 7.8 Building and construction materials 7.2 Food and tobacco 4.5 Other services 11.0 Communications 9.3 Services Financial services 9.1 Real estate 8.9 Coal, oil, and natural gas 24.4 Primary Metals 8.3 Other primary 0.4 Other manufacturing 19.0 South Asia Manufacturing Chemicals 7.3 Automotive OEM 4.4 Other services 12.1 Communications 7.3 Services Financial services 5.9 Software and IT services 5.9 Real estate 5.0 Coal, oil, and natural gas 27.8 Primary Metals 11.6 Other primary 2.0 Sub–Saharan Africa Other manufacturing 11.2 Manufacturing Building and construction materials 4.8 Food and tobacco 2.9 Communications 11.8 Financial services 10.1 Services Other services 9.6 Real estate 5.7 Healthcare 2.4 0 5 10 15 20 25 30 Share of cumulative FDI outflows (percent) Source: Computation based on fDi Markets database, the Financial Times; and Thomson Reuters. Note: For each developing region, the figure shows the relative share of the cumulative value of OFDI projects across sectors between 2003 and 2015. IT = information technology; OEM original equipment manufacturer; OFDI = outward foreign direct investment. Notes exceed the high-income threshold. These economies include Argentina (2014–15), 1. A country is considered developed or devel- Chile (2012–15), Croatia (2008–15), the Czech oping according to the World Bank income Republic (2006–15), Estonia (2006–15), classification. With 1995 as the initial year Equatorial Guinea (2007–14), Greece (1996– for the descriptive analysis, the group of 2015), Hungary (2007–15), Latvia (2009; developing countries includes all that were 2012–15), Lithuania (2012–15), Mauritius low- or middle-income that year. For con- (1998–2015), Oman (2007–15), Poland sistency, these countries remain in the devel- (2009–15), the Russian Federation (2012–14), oping country group even if they eventually Saudi Arabia (2004–15), the Slovak Republic OUT WARD FDI FROM DEVELOPING COUNTRIES 127 (2007–15), Slovenia (1997–2015), Trinidad (Mercosur), 22 percent in the West African and Tobago (2006–15), Uruguay (2012–15), Economic and Monetary Union (WAEMU), and Republíca Bolivariana de Venezuela 36 percent in the Association of Southeast (2014). Two additional adjustments have Asian Nations (ASEAN), and 76 percent in been made. First, the sample disregards 38 tax the East African Community (EAC). and financial havens, following an OECD list 11. The sample of developing countries is the available at http://www.oecd.org/countries same used for the descriptive analysis in this /monaco /jurisdictionscommittedtoimprovin chapter (for example, the countries below the gtransparencyandestablishingeffectiveexcha high-income threshold in 1995). ngeofinformationintaxmatters.htm. Second, 12. To illustrate FDI trends across sectors, the Hong Kong SAR, China, is also removed, analysis relies on information from two given its intermediary role for Chinese OFDI. transaction-level data sources on FDI pro- 2. The literature uses a four-part typology to dis- jects. First, the fDi Markets dataset tracks aggregate FDI flows by investor motivation: media announcements of firm-level green- natural resource–seeking, efficiency-seeking, field FDI projects. Second, the Thomson market-seeking, or strategic asset–seeking Reuters dataset provides the value of individ- (Dunning 2000). ual M&A transactions. The matching meth- 3. In value terms, the region’s OFDI increases odology detailed in Kierkegaard (2013) is from $32 billion in 2000–04 to $788 billion used to merge both datasets. In all, the infor- in 2010–15. mation from these two datasets allows us to 4. These shares correspond to $33 billion and create a single FDI dataset covering the years $242 billion, respectively. 2003–15 and including both greenfield and 5. Equivalent to $51 billion during 2000–04 M&A. and $399 billion in 2010–15. 13. The groups “Other Manufacturing” and 6. An example is the contribution of these “Other Services” are residual categories countries to global GDP growth. During that include manufacturing and service 1995–2000, the BRICS accounted for half industries with shares smaller than the sec- the GDP growth generated in the develop- tors included for each region. For exam- ing world. This contribution has increased ple, all the industries included in “Other further in recent years, reaching 60 percent Manufacturing” in East Asia and Pacific during 2010–15. hold an OFDI share that is smaller than 7. The heat map establishes discrete thresh- Communications (3.7 percent). Therefore, olds at lower than 5 percent, 5–10 percent, the industries included in these residual cat- 10–15 percent, 15–20 percent, and greater egories may differ by region. The manufac- than 20 percent. turing industries that are always included 8. This figure is based on the UNCTAD FDI in “Other Manufacturing” are aerospace, bilateral dataset, which maps investment automotive components, biotechnology, flows and stocks across both home and host business machines and equipment, consumer economy, between the years 2001 and 2012. electronics, consumer products, electronic 9. The correlation between the share of inward components, engines and turbines, industrial developing country OFDI and per capita machinery, medical devices, nonautomotive GDP (PPP) is -0.51. transport original equipment manufacturer, 10. Another dimension by which to study this and pharmaceuticals. Service industries that regional bias is that across regional trade are marginal enough to be included in the blocks. In this regard, the relative importance “Other Services” category are business ser- of FDI between members varies markedly vices, leisure and entertainment, and ware- across different customs unions and regional housing and storage. trade agreements. As an example, and rely- 14. The preference for greenfield is most evident ing on the UNCTAD bilateral FDI dataset, in South Asia and Europe and Central Asia, the share of the intra-agreement stock of with 60 percent of South Asia and 55 percent OFDI (for example, the share of OFDI stock of Europe and Central Asia projects chan- from member countries located in another neled this way. The only exception is Latin member of the agreement) is virtually zero America and the Caribbean, where 54 per- in the Caribbean Community (CARICOM), cent of projects are through M&A. When 10 percent in the Mercado Común del Sur measured by cumulative project value (versus 128 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 the number of projects), the share of green- recognition is strategic asset–seeking but field projects in total OFDI is 68 percent in not knowledge-seeking. Knowledge-seeking South Asia, 62 percent in Middle East and OFDI aims to augment firm-specific advan- North Africa, 60 percent in East Asia and tage owned by the firm to improve its compet- Pacific, 59 percent in Europe and Central itiveness by acquiring new knowledge (Chen, Asia, and 58 percent in Sub-Saharan Africa. Li, and Shapiro 2012). This chapter is mostly Again, the only exception is Latin America concerned with knowledge-seeking OFDI and the Caribbean, where the share of green- and not other forms of strategic asset–seeking field projects in total OFDI is 33 percent. as this type of investment is more likely to 15. This information is conveyed in the last generate home effects. In this chapter, the column of figure 4.7, which shows the devi- term “knowledge” is used to subsume differ- ation in percentage points of the greenfield ent forms of knowledge, including technology share of developing country OFDI relative to and management know-how. developed economies. 22. Chen, Li, and Shapiro (2012) investigate the 16. Eurostat (2017) identifies those indus- explanatory power of three host economy tries with high or medium-high technology knowledge-related independent variables intensity. (R&D employment, R&D expenditures, and 17. Institutional quality embraces several attrib- patents) for variation in home technological utes. The World Bank’s World Governance ability (proxied by home economy firm-level Indicators (WGI) give it six dimensions: R&D expenditure). Voice and Accountability, Political Stability 23. See earlier section “What and How? Sector and Absence of Violence, Government and Mode of Entry” for a discussion of Effectiveness, Regulatory Quality, Rule of knowledge-intensive industries. Law, and Control of Corruption. 24. For the chemical and pharmaceutical sectors, 18. Countries can also play a significant role in see Criscuolo (2009). For the auto industry, giving their firms incentives to undertake see Mani (2013). OFDI, through what has alternatively been 25. Looking at the Indian automotive industry, called “home country measures” or “home Pradhan and Singh (2008) examine OFDI determinants.” For a comprehensive dis- from 1988 to 2008 and find a positive effect cussion of the measures that economies can on home in-house R&D intensity for invest- enact in support of their firms undertaking ments in both developed and developing host OFDI, see Sauvant and others (2014). These economies, although it is stronger for OFDI measures can take the form of information, in developed economies. support services, financial measures, and 26. For a discussion of the growing importance fiscal measures. The relationship between of South–South technology networks, see home determinants and home effects is a Nepelski and De Prato (2015). rich and unexplored area that merits future 27. The coefficients of OFDI are positive and sta- investigation. tistically significant at the 5 percent level for 19. While economies of scale arise from lower all countries, indicating that the OFDI and average costs attributable to an increase in exports are complementary (Ahmad, Draz, the size of the operation, economies of scope and Yang 2016). arise from lower average costs owing to pro- 28. The study looked at the effect of outward duction of similar goods or services. investment stocks on bilateral exports among 20. Innovation is generally examined through the 15 countries of the European Union from R&D measures (expenditures, employment) 1986 to 1996 (Egger 2001). and patent measures (registration, citation). 29. For a discussion of the implications of dif- 21. Dunning’s classic typology for FDI motiva- ferent levels of absorptive capacity, see Tang tions includes strategic asset–seeking FDI and Altshuler (2015). (Dunning 2000); more recently, scholars 30. Other studies have suggested that the export have used the term knowledge-seeking FDI intensity of a firm, its size, governance struc- (Meyer 2015). The former type is broader tures, and R&D spending all may affect than the latter: all knowledge-seeking FDI absorptive capacity. First, firms that are is strategic asset–seeking, but not all strate- exporters have more knowledge of, and gic asset–seeking is knowledge-seeking. For experience with, foreign markets, which may example, acquiring a brand for brand-name make them more capable of understanding OUT WARD FDI FROM DEVELOPING COUNTRIES 129 and absorbing foreign technologies (Tang and Overseas Direct Investments, Reserve Bank Altshuler 2015). Second, small firms may enjoy of India (updated April 12, 2017), available more spillovers as they are less bureaucratic, at https://www.rbi.org.in/Scripts/FAQView. making it easier to adjust to new technologies aspx?Id=32. (Sinani and Meyer 2004); nonetheless, small 35. See Question 12 “Are overseas investments firms may not be able to compete as effect- freely allowed in all the countries and are ively with foreign firms (Aitken and Harrison there any restrictions regarding the cur- 1999). Third, large, family-owned conglom- rency of investment?” in Frequently Asked erates have emerged in many developing Questions, Overseas Direct Investments, countries to address market failures linked to Reserve Bank of India (updated April 12, weak property rights, contract enforcement, 2017), op cit. and widespread corruption. Yet studies have 36. See full list of 61 Frequently Asked Questions, found such relation-based governance to be Overseas Direct Investments, Reserve Bank associated with lower levels of innovation— of India (updated April 12, 2017), op cit. as innovation makes the sunk costs invested 37. See Guidelines to Authorised Dealers in respect in relationships less valuable—suggesting of genuine new foreign direct investments of lower levels of absorptive capacity (Li, Park, up to R1 billion per company per calendar and Li 2003). Fourth, R&D spending may year (2016-05-10), published by the Financial improve recipients’ absorptive capacity, while Surveillance Department of the South African also helping transform pure knowledge into Reserve Bank. Available at https://www inputs for productive innovation (Chen, Li, . resbank.co.za /RegulationAndSupervision and Shapiro 2012). / FinancialSurveillanceAndExchangeCon 31. The International Monetary Fund defines trol / Guidelines/Guidelines%20and%20 low-income developing countries as those public%20awareness/Guidelines%20-%20 with a level of per capita gross national FDI.pdf. income less than the Poverty Reduction and 38. For discussion of how developing econo- Growth Trust (PRGT) income graduation mies in Asia have successfully reformed their level for non-small states (IMF 2014). OFDI regulatory frameworks, see Rasiah, 32. Looking at the share of countries at differ- Gammeltoft, and Jiang (2010). ent income levels that maintain some form of 39. Gómez-Mera and others (2015) devise a OFDI restrictions does not, however, capture cross-sectional econometric specification the relative intensity of restrictions. On the with two steps: a logit model to deter- basis of individual country examples, OFDI mine the probability of investment, and a restrictions seem to be getting less restrictive zero-truncated negative binomial model to over time even if some form of OFDI restric- determine the drivers of the positive count tion today remains in place in many coun- of investments. With our dependent variable tries. Future work will explore the relative being the flow of FDI between two countries intensity of OFDI restrictions across coun- at a given year, our analysis adopts a Poisson tries at different levels of development, and Pseudo-Maximum Likelihood Estimator across time. (PPML). Under weak assumptions, Santos 33. See Anderson (2013) for a World Bank Group Silva and Tenreyro (2006) find that the report on “Converting and Transferring PPML provides consistent estimates, cir- Currency: Benchmarking Foreign Exchange cumventing the problem of heteroscedas- Restrictions to Foreign Direct Investment ticity in standard nonlinear gravity specifi- across Economies.” cations. Thus, the PPML estimator is also 34. The Government of India specifies that real consistent in the presence of fixed effects. It estate is the “buying and selling of real estate is also better suited to include zero observa- or trading in Transferable Development tions, eliminating the possibility of sample Rights (TDRs) but does not include develop- selection bias. ment of townships, construction of residen- 40. The gravity equation under PPML does not tial/commercial premises, roads or bridges” specify bilateral country-pair fixed effects See Question 4 “Can overseas direct invest- controlling for unobserved time-invariant ment be made in any activity? What are heterogeneity, due to problems of collinearity the prohibited activities for overseas direct with explanatory variables. Instead, the spec- investment?” in Frequently Asked Questions, ification includes single source fixed effects. 130 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 41. The CEPII dataset includes alternative vari- CEPR Discussion Paper 4284, Centre for ables to test for shared colonial history and Economic Policy Research, London. language. Specifically, a dummy variable for Bevan, A.A. and S. Estrin. 2004. “The source and host having the same colonizer Determinants of Foreign Direct Investment into after 1945 (comcol) and a dummy that takes European Transition Economies.” Journal of the value of 1 if a language is spoken by at Comparative Economics 32 (4): 775–87. least 9 percent of the population in both Bitzer, J., and H. Görg. 2009. “Foreign Direct countries (comlang-ethno). The use of these Investment, Competition and Industry alternatives did not change any of the results, Performance.” The World Economy 32 (2): except for comcol, which is insignificant as a 221–33. proxy for shared colonial history. Boly, A., N. D. Coniglio, F. Prota, and A. Seric. 2014. “Diaspora Investments and Firm Export Performance in Selected Sub-Saharan Bibliography African Countries.” World Development Ahmad, F., M. U. Draz, and S. C. Yang. 2016. 59: 422–33. “A Novel Study on OFDI and Home Country ———. 2015. “Which Domestic Firms Benefit Exports: Implications for the ASEAN Region.” from FDI? Evidence from Selected African Journal of Chinese Economic and Foreign Countries.” Development Policy Review 33 (5): Trade Studies 9 (2): 131–45. 615–36. Aitken, B. J., and A. E. Harrison. 1999. “Do Braconier, H., K. Ekholm, and K. H. M. Knarvik. Domestic Firms Benefit from Direct Foreign 2001. “In Search of FDI-Transmitted R&D Investment? Evidence from Venezuela.” Spillovers: A Study Based on Swedish American Economic Review 89 (3): 605–18. Data.” Review of World Economics 137 (4): Amann, E., and S. Virmani. 2014. “Foreign 644–65. Direct Investment and Reverse Technology Buckley, P. J., L. J. Clegg, A. R. Cross, X. Liu, H. Spillovers.” OECD Journal: Economic Studies Voss, and P. Zheng. 2007. “The Determinants 3 (1): 129–53. of Chinese Outward Foreign Direct Amighini, A. A., R. Rabellotti, and M. Sanfilippo. Investment.” Journal of International Business 2013. “Do Chinese State-Owned and Private Studies 38 (4): 499–518. Enterprises Differ in Their Internationalization Chen, V. Z., J. Li, and D. M. Shapiro. 2012. Strategies?” China Economic Review 27: “International Reverse Spillover Effects on 312–25. Parent Firms: Evidences from Emerging Market Anderson, J. 2013. “Converting and Transferring MNEs in Developed Markets.” European Currency: Benchmarking Foreign Exchange Management Journal 30 (3): 204–18. Restrictions to Foreign Direct Investment across Chen, W., and H. Tang. 2015. Chinese Investment Economies.” World Bank, Washington, DC. in Africa Is More Diverse and Welcome than You Arita, S. 2013. “Do Emerging Multinational Think. Quartz Africa. https://qz.com/488589 Enterprises Possess South-South FDI /chinese-investment-in-africa-is-more-diverse Advantages?” International Journal of -and-welcomed-than-we-give-it-credit/. Emerging Markets 8 (4): 329–53. China Daily. 2017a. “SOEs Face ‘Red Line’ on Assunção, S., R. Forte, and A. Teixeira. 2011. Investment.” China Daily, January 19. http:// “Location Determinants of FDI: A Literature english.gov.cn/state_council/ministries/2017/01/19 Review.” FEP Working Papers 433, Faculdade /content_281475545487067.htm. de Economia, Universidade do Porto, Portugal. ———. 2017b. “Making China FDI-friendly Aykut, D., and A. Goldstein. 2006. “Developing again.” China Daily , February 20. http:// Country Multinationals: South–South english.gov.cn/news/top_news/2017/02/20 Investment Comes of Age.” Working Paper /content_281475572940524.htm. 257, OECD Development Centre, Paris, France. Cohen, W. M., and D. A. Levinthal. 1989. Aykut, D., and D. Ratha. 2004. “South-South FDI “Innovation and Learning: The Two Faces Flows: How Big Are They?” Transnational of R&D.” The Economic Journal 99 (397): Corporations 13 (1): 149–76. UNCTAD, 569–96. Geneva. Criscuolo, C. 2009. “Innovation and Productivity: Barba Navaretti, G., and D. Castellani. 2004. Estimating the Core Model across 18 Countries.” “Investments Abroad and Performance at In OECD, Innovation in Firms: A Microeconomic Home: Evidence from Italian Multinationals.” Perspective, 111–38. Paris: OECD Publishing. OUT WARD FDI FROM DEVELOPING COUNTRIES 131 Criscuolo, P., and R. Narula. 2008. “A Novel Reallocation, Employment and Productivity A p p r o a c h t o N a t i o n a l Te c h n o l o g i c a l in Taiwan.” International Journal of the Accumulation and Absorptive Capacity: Economics of Business 16 (1): 19–38. Aggregating Cohen and Levinthal.” Duanmu, J.-L. and Y. Guney. 2009. “A Panel The European Journal of Development Data Analysis of Locational Determinants of Research 30 (1): 56–73. Chinese and Indian Outward Foreign Direct Cuervo-Cazurra, A. 2008. “The Multination- Investment.” Journal of Asia Business Studies alization of Developing Country MNEs: The 3 (2): 1-15. Case of Multilatinas.” Journal of International Dunning, J. H. 2000. “The Eclectic Paradigm as an Management 14 (2): 138–54. Envelope for Economic and Business Theories Cuervo-Cazurra, A., and M. Genc. 2008. of MNE Activity.” International Business “Transforming Disadvantages into Advantages: Review 9 (2): 163–90. Developing-Country MNEs in the Least Dunning, J. H., C. S. Kim, and D. H. Park. 2008. Developed Countries.” Journal of International “Old Wine in New bottles: A Comparison of Business Studies 39 (6): 957–79. Emerging-Market TNCs Today and Developed- Cui, L. and F. Jiang. 2009. “FDI Entry Mode Country TNCs Thirty Years Ago.” In The Rise Choice of Chinese Firms: A Strategic Behavior of Transnational Corporations from Emerging Perspective.” Journal of World Business 44 (4): Markets: Threat or Opportunity? , edited 434–44. by K. P. Sauvant, 158–78. Cheltenham, UK: Deng, P. 2009. “Why Do Chinese Firms Tend Edward Elgar Publishing. to Acquire Strategic Assets in International Echandi, R., J. Krajcovicova, and C. Z. W. Qiang. Expansion?” Journal of World Business 44 (1): 2015. “The Impact of Investment Policy in a 74–84. Changing Global Economy.” Policy Research ———. 2010. “Absorptive Capacity and A Failed Working Paper 7437, World Bank, Washington, Cross-Border M&A.” Management Research DC. Review 33 (7): 673–82. Egger, P. 2001. “European Exports and Outward Davies, R., R. Desbordes, and A. Ray. 2015. Foreign Direct Investment: A Dynamic Panel “Greenfield versus Merger & Acquisition FDI: Data Approach.” Weltwirtschaftliches Archiv Same Wine, Different Bottles?” Working Paper 137 (3): 427–49. Series, UCD Centre for Economic Research, No Eurostat. 2017. “Statistics Explained.” http:// 15/03. ec.europa.eu/eurostat/statistics-explained Dollar, D. 2016. “China as a Global Investor.” /index.php/Main_Page. China’s New Sources of Economic Growth: Farole, T., and D. Winkler, eds. 2014. Making Vol. 1: Reform, Resources and Climate Change. Foreign Direct Investment Work for Sub-Saharan Edited by Ligang Song, Ross Garnaut, Cai Africa: Local Spillovers and Competitiveness in Fang and Lauren Johnston. Australia National Global Value Chains. Directions in Development. University Press, Acton. World Bank, Washington, DC. Dong, B. and G. Guo. 2013. “A Model of China’s FT (Financial Times). 2017. “China Tightens Export Strengthening Outward FDI.” China Rules on State Groups’ Foreign Investments.” Economic Review 27: 208–26. August 3. Financial Times . https://www.ft Driffield, N., and J. H. Love. 2003. “Foreign .com/content/3251987c-7806-11e7-90c0 Direct Investment, Technology Sourcing and -90a9d1bc9691. Reverse Spillovers.” The Manchester School Fortescue, S., and P. Hanson. 2015. “What Drives 71 (6): 659–72. Russian Outward Foreign Direct Investment? ———. 2007. “Linking FDI Motivation and Host Some Observations on the Steel Industry.” Post- Economy Productivity Effects: Conceptual and Communist Economies 27 (3): 283–305. Empirical Analysis.” Journal of International Gammeltoft, P. 2008. “Emerging Multinationals: Business Studies 38: 460–73. Outward FDI from the BRICS Countries.” Driffield, N., J. H. Love, and K. Taylor. 2009. International Journal of Technology and “Productivity and Labour Demand Effects of Globalisation 4 (1): 5–22. Inward and Outward Foreign Direct Investment Gammeltoft, P., H. Barnard, and A. Madhok. on UK Industry.” Manchester School 77 (2): 2010. “Emerging Multinationals, Emerging 171–203. Theory: Macro- and Micro-Level Perspectives.” Driffield, N., and P. C. (Michelle) Chiang. Journal of International Management 16 (2): 2009. “The Effects of Offshoring to China: 95–101. 132 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 Gao, T. 2005. “Foreign Direct Investment Huang, Y. and B. Wang. 2011. “Chinese Outward from Developing Asia: Some Distinctive Direct Investment: Is There a China Model?” Features.” Economics Letters 86 (1): 29–35. China & World Economy 19 (4): 1–21. Garcia-Herrero, A., L. Xia, and C. Casanova. ICSID (International Centre for Settlement of 2015. “Chinese Outbound Foreign Direct Investment Disputes). 2016. Annual Report Investment: How Much Goes Where after on Exchange Arrangements and Exchange Roundtripping and Offshoring?” BBVA Restrictions. Washington, DC. Working Paper 15/17, Hong Kong SAR, China. ———. 2017. Database of Bilateral Investment Girma, S. 2005. “Absorptive Capacity and Treaties. https://icsid.worldbank.org/en /Pages Productivity Spillovers from FDI: A Threshold /resources / Bilateral-Investment-Treaties Regression Analysis.” Oxford Bulletin of -Database.aspx. Economics and Statistics 67 (3): 281–306. IMF (International Monetary Fund). 2014. Girma, S., and H. Gorg. 2007. “Evaluating “Proposed New Grouping in WEO Country the Foreign Ownership Wage Premium Classifications: Low-Income Developing using A Difference-in-Differences Matching Countries.” IMF Policy Paper, Washington, DC. Approach.” Journal of International Economics ———. 2016. Annual Report on Exchange 72 (1): 97–112. Arrangements and Exchange Restrictions . Global Wind Energy Council. 2016. “Global Wind Washington, DC: IMF. Report: Annual Market Update.” http://www Kalotay, K., and A. Sulstarova. 2010. “Modelling .gwec.net/publications/global-wind-report-2 Russian Outward FDI.” Journal of /global-wind-report-2016/. International Management 16 (2): 131–42. Gómez-Mera, L., T. Kenyon, Y. Margalit, Kang, Y. and F. Jiang. 2012. “FDI Location J. G. Reis, and G. Varela. 2015. New Voices Choice of Chinese Multinationals in East and in Investment: A Survey of Investors from Southeast Asia: Traditional Economic Factors Emerging Countries . World Bank Studies. and Institutional perspective.” Journal of World Washington, DC: World Bank. Business 47: 45–53. Grieger, G. 2017. “Foreign Direct Investment Kierkegaard, J. 2013. “New Avenues for Empirical Screening: A Debate in Light of China-EU FDI Analysis of Cross-Border Investments: An flows.” EPRS (European Parliamentary Research Application for the ASEAN Members and Service) Document PE 603.941. http://www Middle and Low Income Country Outward .europarl.europa.eu/thinktank /en/document Investments.” PhD dissertation, Johns Hopkins .html?reference=EPRS_BRI%282017%29603941 University, Washington, DC. Hanemann, T., A. Lysenko, and C. Gao. 2017. Kimura, F., and K. Kiyota. 2006. “Exports, FDI, “Tectonic Shifts: Chinese Outbound M&A and Productivity: Dynamic Evidence from in 1H 2017.” Rhodium Group Report. June Japanese Firms.” Review of World Economics 2017. http://rhg.com/notes/tectonic-shifts (Weltwirtschaftliches Archiv) 142 (4): 695–719. -chinese-outbound-ma-in-1h-2017. Kuz ´min´ska-Haberla, A. 2012. “The Promotion of Herzer, D. 2011. “The Long-Run Relationship Outward Foreign Direct Investment—Solutions between Outward FDI and Total Factor from Emerging Economies.” Working Paper, Productivity: Evidence for Developing Institute of International Business, University of Countries.” Journal of Development Studies Gdansk no. 31, Poland. 47 (5): 767–85. Lardy, N. R. 2014. Markets over Mao. Peterson ———. 2012. “How does Foreign Direct Institute for International Economics, Investment Really Affect Developing Countries’ Washington, DC. Growth?” Review of International Economics. Lee, C., C. G. Lee, and M. Yeo. 2016. 20 (2): 396–414 “Determinants of Singapore’s Outward FDI.” Hiratsuka, D. 2006. “Outward FDI from and Journal of Southeast Asian Economies 33 (1): Intraregional FDI in ASEAN: Trends and 23–40. Drivers.” Discussion Paper 77, Institute of Li, S., S. H. Park, and S. H. Li. 2003. “The Great Developing Economies, JETRO, Chiba Leap Forward: The Transition from Relation- Holtbrügge, D. and H. Kreppel. 2012. Based Governance to Rule-Based Governance.” “Determinants of Outward Foreign Direct Organizational Dynamics 33 (1): 63–78. Investment from BRIC Countries: An Linneman, H. 1966. “An Econometric Study of Explorative Study.” International Journal of International Trade Flow.” North-Holland, Emerging Markets 7 (1): 4–30. Amsterdam, 77. OUT WARD FDI FROM DEVELOPING COUNTRIES 133 Luo, Y., Q. Z. Xu, and B. J. Han. 2010. “How Automotive Sector.” International Journal of Emerging Market Governments Promote Institutions and Economies 1 (1): 155–86. Outward FDI: Experience from China.” Quer, D., E. Claver, and L. Rienda. 2012. “Political Journal of World Business 45 (1): 68–79. Risk, Cultural Distance, and Outward Foreign Ma, A. C., and A. Van Assche. 2011. “Product Direct Investment: Empirical Evidence from Distance, Institutional Distance and FDI.” Large Chinese Firms.” Asia Pacific Journal of Mimeo, University of San Diego, School of Management 29 (4): 1089–1104. Business Administration. ———. 2015. “Chinese Outward Foreign Direct Mani, S. 2013. “Outward Foreign Direct Investment: A Review of Empirical Research.” Investment from India and Knowledge Flows, Frontiers of Business Research in China 9 (3): the Case of Three Automotive Firms.” Asian 326–70. Journal of Technology Innovation 21: 25–38. Ramamurti, R. 2009. “What Have We Learned Meyer, K. E. 2015. “What Is ‘Strategic Asset- about Emerging-Market MNEs?” In Emerging Seeking FDI’?” Multinational Business Review Multinationals in Emerging Markets, edited by 23 (1): 57–66. Ravi Ramamurti and Jitendra V. Singh, 399–426. MOFCOM. 2015. Belt and Road Helps China Cambridge, U.K.: Cambridge University Press. Become Net Capital Exporter: Report ———. 2012. “What is Really Different About (accessed May 5, 2017), http://english. Emerging Market Multinationals?” Global m o f c o m . g o v. c n / a r t i c l e / z t _ b e l t a n d r o a d Strategy Journal 2 (1): 41 –7. /news/201511/20151101156468.shtml. Ramasamy, B., M. Yeung, and S. Laforet. 2012. Moran, T. 2015. “The Role of Industrial Policy “China’s Outward Foreign Direct Investment: as a Development Tool: New Evidence from Location Choice and Firm Ownership.” Journal the Globalization of Trade-and-Investment.” of World Business 47: 17–25. CGD Policy Paper 071. Center for Global Rasiah, R., P. Gammeltoft, and Y. Jiang. 2010. Development, Washington, DC. http://www “Home Government Policies for Outward FDI .cgdev.org/publication/role-industrial-policy from Emerging Economies: Lessons from Asia.” -development-tool-new-evidenceglobalization International Journal of Emerging Markets -trade-and-investment 5 (3/4): 333–57. Moran, T., H. Gorg., and A. Seric. 2016. “Quality RBI (Reserve Bank of India). “Frequently Asked FDI and Supply-Chains in Manufacturing.” Kiel Questions, Overseas Direct Investments” Centre for Globalization Policy Paper 1, Kiel. (accessed April 12, 2017), https://www.rbi.org Narula, R. 2004. “Understanding Absorptive .in/Scripts/FAQView.aspx?Id=32. Capacity in an ‘Innovation System’ Context: Rodriguez-Arango, Liliana and Maria Alejandra Consequences for Economic and Employment Gonzalez-Perez. 2016. “Giants from Emerging Growth.” MERIT-Infonomics Research Markets: The Internationalization of BRIC Memorandum Series 3, Maastricht. Multinationals.” In The Challenge of BRIC Nepelski, D., and G. De Prato. 2015. “International Multinationals , edited by Rob Van Tulder, Technology Sourcing between a Developing Alain Verbeke, Jorge Carneiro, and Maria Country and the Rest of the World. A Case Alejandra Gonzalez-Perez, 195–226. Progress Study of China.” Technovation 35: 12–21. in International Business Research, Vol. 11. Nguyen, H. T., G. Duysters, J. H. Patterson, and Emerald Group Publishing Limited. H. Sander. 2009. “Foreign Direct Investment Rosen, D. H., and T. Hanemann. 2017. “New Absorptive Capacity Theory.” Paper presented Neighbors 2017 Update: Chinese FDI in the at GLOBELICS 2009, 7th International United States by Congressional District.” Conference, Dakar, Senegal, October 6–8. Rhodium Group Report. Nunes de Alcântara, J., C. M. P. Paiva, N. C. P. Bruhn, Sader, F. 1993. “Privatization and Foreign Investment H. R. de Carvalho, and C. L. L. Calegario. 2016. in the Developing World. 1988–92.” World Bank “Brazilian OFDI Determinants.” Latin American Policy Research 1202, Washington, DC. Business Review 17 (3): 177–205. Santos Silva, J. M. C., and S. Tenreyro. 2006. “The OECD (Organisation for Economic Co-Operation Log of Gravity.” The Review of Economics and and Development). 2013. Interconnected Statistics 88 (4): 641–58. Economies: Benefiting from Global Value Sauvant, K. P. 2008. “The Rise of TNCs from Chains. Paris. Emerging Markets: The Issues.” In The Rise Pradhan, J. P., and N. Singh. 2008. “Outward FDI of Transnational Corporations from Emerging and Knowledge Flows: A Study of the Indian Markets: Threat or Opportunity? edited by 134 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 K. P. Sauvant, 3–14. Cheltenham, U.K.: Edward UNCTAD (United Nations Conference on Elgar Publishing. Trade and Development). 2005. “Case Study Sauvant, K. P., and V. Z. Chen. 2014. “China’s on Outward Foreign Direct Investment by Regulatory Framework for Outward Foreign South African Enterprises.” TD/B/COM.3 Direct Investment.” China Economic Journal /EM.26/2/Add.5. Geneva. http://unctad.org/en 7 (1): 141–63. /Docs/c3em26d2a5_en.pdf. Sauvant, K. P., P. Economou, K. Gal, S. W. Lim, ———. 2013. World Investment Report 2013: and W. Wilinski. 2014. “Trends in FDI, Global Value Chains: Investment and Trade for Home Country Measures and Competitive Development. Geneva. Neutrality.” In Yearbook on International ———. 2015. World Investment Report Investment Law & Policy 2012–2013, edited 2015: Reforming International Investment by Andrea Bjorklund, 3–107. New York: Governance. Geneva. Oxford University Press. ———. 2017. World Investment Report 2017: Sinani, E., and K. E. Meyer. 2004. “Spillovers of Investment and the Digital Economy. Technology Transfer from FDI: The Case of Geneva. Estonia.” Journal of Comparative Economics Vahter, P., and J. Masso. 2007. “Home versus Host 32 (3): 445–66. Country Effects of FDI: Searching for New Slangen, A., and J. F. Hennart. 2007. “Greenfield Evidence of Productivity Spillovers.” Applied or Acquisition Entry: A Review of the Empirical Economics Quarterly 53 (2): 165–96. Foreign Establishment Mode Literature.” Wang, M. 2017. “Chinese Overseas Investment.” Journal of International Management. 13 (4): Unpublished (3/19/17). 403–29. Wang, P., and Z. H. Yu. 2014. “China’s Outward South African Reserve Bank. 2016. “Guidelines to Foreign Direct Investment: The Role of Natural Authorised Dealers in Respect of Genuine New Resources and Technology.” Economic and Foreign Direct Investments of Up to R1 billion Political Studies 2 (2). per Company per Calendar Year (2016-05-10).” Wells, L. T., Jr. 2009. “Third World Multinationals: South African Reserve Bank, Pretoria. A Look Back.” In Emerging Multinationals in Talamo, G. 2007. “Institutions, FDI, and the Emerging Markets, edited by R. Ramamurti Gravity Model.” In Workshop PRIN 2005 SU, and J. V. Singh, 23–41. Cambridge, UK: Economic Growth; Institutional and Social Cambridge University Press. Dynamics, 25-27. Zhang, Y. and H. Roelfsema. 2014. “Unravelling Tang, J. T., and R. Altshuler. 2015. “The Spillover The Complex Motivations Behind China’s Effects of Outward Foreign Direct Investment Outward FDI.” Journal of the Asia Pacific on Home Countries: Evidence from the Economy 19 (1): 89–100. United States.” Unpublished. http://dx.doi Zheng, N., Y. Wei, Y. Zhang, and J. Yang. 2016. .org/10.2139/ssrn.2545129. “In Search of Strategic Assets through Cross- Turkish Household Appliances Suppliers. 2017. Border Merger and Acquisitions: Evidence from “Turkish Household Appliances Supply Chinese Multinational Enterprises in Developed Industry Sector in Turkey.” http://www.turkhas Economies.” International Business Review .org/upload/statistics.pdf. 25 (1): 177–86. 5 FDI in Fragile and Conflict- Affected Situations Alexandros Ragoussis and Heba Shams F oreign direct investment (FDI) in frag- cost of war). Yet, while the argument makes ile and conflict-affected situations intuitive sense, the empirical evidence on the (FCS) 1 represents just 1 percent of direct relationship between foreign invest- global flows, more than five times less per ment and conflict remains inconclusive. Some capita than the world average, according to argue that a foreign presence can generate latest estimates. Despite increasing tenfold grievances by adversely affecting income dis- over the last two decades, FDI is still mostly tribution and worsening political unrest in concentrated in a handful of fragile coun- low-income countries (Gissinger and tries, all middle-income or resource-rich or Gleditsch 1999), while others contend that both. Furthermore, differences in FDI poten- trade and FDI complement each other in tial and dependence within the FCS group reducing conflict risk (Polachek and are also stark: FDI inflow as a share of gross Sevastianova 2012). More nuanced effects national income (GNI) ranges from more have been acknowledged, too, such as that than 40 percent in Liberia to virtually zero in FDI reduces the duration of civil wars but not South Sudan. the likelihood of their onset (Barbieri and In response to the proliferation of conflicts Reuveny 2005). and forced displacements in this decade to Clearly, not all FDI has the same effects date, the development community has com- on host countries. The sectoral distribution mitted itself to doing more for fragile coun- of FDI, especially amid distorted conditions, tries. Foreign investment is a central part of can potentially reinforce opposite trends. This that commitment, yet consensus on the facts, is partly why policy discussions have focused drivers, and imperatives surrounding it has on dilemmas surrounding “good” and not yet been achieved. Better understanding is “bad” FDI in fragile contexts, often related key for the development community to design to the exploitation of natural resources the right interventions. (International Dialogue for Peace-Building But can FDI support stabilization and pre- and State-Building 2016). Recognizing the vent violent conflicts? FDI can create jobs, limitations of econometrics in addressing generate wealth and tax income, and thereby the question is also critical. The pro-cyclical affect what fragile societies risk losing by movement of foreign investment,2 and the engaging in conflict (that is, the opportunity indirect channels through which it affects 135 136 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 the opportunity cost of conflict, complicate The Where, Who, and How of the identification of its effect on peace and stability. Foreign Investment in FCS The purpose of this chapter is to take a Foreign sources of income sustain a large step back from this discussion and fill in a gap part of economic activity in fragile and in understanding FDI across these sensitive conflict-affected situations. Yet international environments. The discussion rests on a fun- investors do not typically consider FCS as damental notion of FDI’s potential to gener- hosts, owing to economic fundamentals ate jobs, increase wealth, and improve public and fragility, which are mutually reinforc- goods—all of which are essential for a stable ing. While fragile situations are remarkably and prosperous society. The chapter considers heterogeneous (box 5.1), commonalities the where, who, and how of foreign invest- do exist: investment opportunities arise in ment in FCS before delving into difficult capital-intensive activities sustained by for- questions of why and of ways to support eign demand, particularly during transitions investment through policies that, in principle, from conflict to peace. But investors are would also enhance stability. cautious in how they leverage these oppor- Foreign investment in fragile situations has tunities. Those who understand the context the potential to deliver good results. Apart do better. from resource-seeking investment, the struc- ture of economic activity in these countries FCS Depend Heavily on Foreign reveals a strong potential for FDI-driven value Sources of Income creation in sectors with low domestic competi- tion or others experiencing growth attribut- Foreign investment, along with other able to postconflict reconstruction. But sources of income sustains a large part of investors are cautious: outside natural resource economic activity in fragile and conflict- sectors, they concentrate their investment in a affected situations. The combination of limited number of capital-intensive activities. remittances from the diaspora, official devel- They also tend to commit to smaller projects, opment assistance (ODA), official aid, and create fewer jobs, and avoid geographical foreign investment often exceed a third of exposure to security risks. FCS pose unique national income at varying degrees of depen- conditions and risks at both the operational dence (figure 5.1). and institutional levels where investment cli- Diaspora income, ODA, and FDI interact in mate reforms could make a difference. a variety of complementary ways in fragile Investment climate reforms that unlock states. For example, although remittances opportunities for the private sector and create are largely used for consumption, they are jobs are necessary to consolidate peace and increasingly seen as a resource for investment.3 move from fragility to resilience. Broad and And while a lot of debate has been had on the deep changes to the rules of the game are relationship between ODA and FDI, conven- essential. An investment climate reform strat- tional wisdom and empirical evidence point to egy requires proper sequencing and prioritiz- the catalyzing effects of ODA on FDI. The ing and must take into account the country’s composition of ODA matters in this respect: conflict dynamics, economic opportunity, Assistance used to finance complementary institutional capacity, and willingness to inputs, such as public infrastructure and reform. The strategy must be implemented in human capital investments, has been shown to a balanced way to secure short-term gains draw in FDI, while assistance in the form of while building the momentum for deep insti- pure physical capital transfers may crowd out tutional transformation. The key elements investment (Selaya and Sunesen 2012). of the strategy should be reducing risks to The prevalence of FDI among foreign investors while maximizing investment sources of income reflects to a large extent opportunities and rewards. heterogeneous conditions among FCS. FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 137 BOX 5.1 FCS Are Highly Heterogeneous in Terms of Risks and Opportunities for Investment Although encountered mainly in low-income countries, Countries at risk of conflict suffer substantial fragile and conflict-affected situations (FCS) persist economic marginalization, political polarization, also in middle-income countries where opportunities and external stresses, which heighten uncertainty for investment are markedly different. Abundant natu- and point to the need for prevention of conflict- ral resources explain the middle-income status of such related situations, including through foreign direct FCS as Iraq or Libya, but not exclusively. The former investment (FDI). Yugoslavia and Lebanon are examples of middle- Subnational conflict within otherwise stable coun- income countries with a history of violent conflict. tries means that foreign investment can take place on Investment opportunities in this group are associated a large scale in stable parts of the territory, and that not only with greater purchasing power of the popula- government capacity for reform exists. tion and market growth, but also with existing indus- Active large-scale conflict and crisis situations trial structures, skills, and government capacity that are distinct in that investment no longer takes place allow working toward more ambitious targets in terms and priority is given to political solutions and basic of investment climate. stabilization. The risks facing investors also vary across FCS, Finally, states in postconfl ict or frequent con- affecting investor decisions as well as the scope and fl ict-to-peace transitions typically suffer from weak depth of necessary reforms. Along the so-called fragil- institutions with poor governance but offer the ity chain, confl ict-affected situations include territo- greatest opportunities for economic transforma- ries under severe risk of confl ict, others experiencing tion through investment, as well as momentum for active confl icts, and states in postconfl ict transitions. reform. FIGURE 5.1 FCS Depend on Income from ODA, the Diaspora, and Foreign Investors, 2015 See note 50 37 62 40 Share of GNI (percent) 30 20 41 10 0 Sie go, eria Le p. Ge one Ca rita a m nia An dia m a M alia oP i M TogR ya o M C ar De an r So T m. R n m ki . Zi Isla n ba s er e ud n an in go q -B a Uz uru au kis i Cô G tan d’ ea Ha e m M ti a, i Af Com The an ros Ni tan n N ria ut ub l Ye h Su lic en an . ng L gas d lo aji ep ep La law be nd bi al So Rep epa o, eb ca m nd au gi So gol ea vin am bw ir D o a oo Gu rze Ira i a a rra Re nm te uin B ss on st m d ge Ivo ad h gh o M or bo ,R is Co Lib a i S n C Ga He ca fri d an lA Co ra ia sn nt Bo Ce FDI ODA and Official Aid Remittances Source: Computation based on World Development Indicators, World Bank. Note: The figures for ODA and official aid and remittances for Liberia have been scaled, and the actual figures are noted in the graph. The Central African Republic and South Sudan have missing values for remittances. FCS = fragile and conflict-affected situations; FDI = foreign direct investment; GNI = gross national income; ODA = official development assistance. 138 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 Remittances, for example, are important for a that trade less with the rest of the world than few fragile states with large diaspora popula- other emerging economies. tions (for example, Haiti, Lebanon, Liberia, Fragility also takes a heavy toll: the dis- and Nepal). Foreign investment represents a tance between expected and actual investment substantial share in such resource-rich coun- is considerable (map 5.1), highlighting the tries as the Republic of Congo and Sierra extent to which investment opportunities Leone, while smaller low-income states (for remain unleveraged. Considering that fragil- example, island territories) rely far more on ity affects existing economic fundamentals ODA and official aid. Unstable territories that are used to form expectations, the dis- in transition fall under this category too: tance between actual investment and what Afghanistan, the Central African Republic, would likely have taken place under stability Libya, Somalia, and South Sudan depend heav- and peace is likely even greater. Only a few ily on aid for reconstruction and less on FDI countries had high levels of both expected despite their wealth of natural resources. and actual investment in recent years: Iraq, Lebanon, Bosnia and Herzegovina, Sudan, Côte d’Ivoire, and others. Countries currently Investment in Many FCS Is experiencing high levels of violence, such as below Potential Libya, the Syrian Arab Republic, and the Economic fundamentals alone—such as Republic of Yemen, also presented high current size of the market, growth, and expected values at points of their latest avail- savings—in addition to remoteness, trade able data (that is, before conflicts escalated). openness, and natural resources would sug- All the countries with higher expected val- gest lower levels of expected investment in ues within the group are either middle-income FCS than in the rest of the world (annex 5B). with developed local markets that can attract Among emerging economies, the bottom market-seeking investment or countries pos- quartile of investment predicted by economic sessing natural resources with high potential fundamentals is populated mostly by FCS for investment. Remoteness to large industrial (figure 5B.1). This result is not surprising economies makes even fewer of them attrac- because fragile and conflict-affected situa- tive for export-oriented, efficiency-seeking tions represent small and remote markets FDI; notable examples being Bosnia and MAP 5.1 FDI Flows to FCS Remain below Potential, 2008–14 Investment BOSNIA & Actual HERZEGOVINA Potential IRAQ AFGHANISTAN LEBANON Investment size 70,490 60,000 NEPAL 40,000 MALI 20,000 THE GAMBIA CHAD HAITI GUINEA ERITREA GUINEA-BISSAU SUDAN LIBERIA CEN T RAL A FRICAN REPUBLIC 7 TOGO CÔTE D'IVOIRE SOMALIA D. R. OF CONGO BURUNDI MALAWI COMOROS SOLOMON MADAGASCAR ISLANDS ZIMBABWE IBRD 43085 | SEPTEMBER 2017 Source: Computation based on Investment Map Database, International Trade Centre; World Development Indicators, World Bank. CEPII Database; Fragile States Index (2014), the Fund for Peace. Note: Investment expectations based on economic fundamentals, represented with green circles, shed light on how much FDI can be expected nett of the effect of fragility. Separating the negative impact of fragility from the predicted inflow (that is, fitted value) from a regression on FDI determinants yields this estimate. Actual Investment flows are represented with a blue circle for comparison. Data are only presented for selected countries officially designated as FCS (in millions of US dollars). Countries with latest data from before 2012, or significantly changed circumstances since the latest data point, are excluded. FCS = fragile and conflict-affected situations; FDI = foreign direct investment. FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 139 Herzegovina, Lebanon, and to a lesser extent, Furthermore, economies in active conflict Haiti. The gap between expected and actual present dynamics that differ sharply from investment is the highest in Iraq, a country what is experienced in pre- or postconflict suffering from protracted instability and vio- cases (see box 5.1). lence for more than two decades. Most FCS However imperfect, 4 the available evi- have many disadvantages—as in market size, dence broadly confirms the scarcity of growth, connectivity, openness, and natural capital-intensive activities in FCS, except resources—making them less appealing to for mining, and oil and gas in some investors. Burundi is typical of FCS, with low resource-rich states. One can argue that it expectations for market-seeking, efficiency- is precisely in these capital-intensive sec- seeking, and resource-seeking investments, all tors where large business opportunities of which are almost entirely unrealized. lie, given the lack of local competition. Yet this is not always so because fragility discourages financial market development, Investment Is Concentrated in which in turn creates barriers to growth. Capital-Intensive Activities Labor-intensive activities, especially ser- Opportunities for investment vary consider- vices and agriculture, are essential for the ably from one fragile situation to another survival of much of the population and since the group is highly variable in eco- hence dominate the economy. However, nomic development terms. While most are variation within the group is pronounced: low-income, middle-income countries in for example, the share of services ranges the group (for example, Angola, Bosnia from 17 percent in Liberia to 90 percent in and Herzegovina, and Iraq) offer different Lebanon, and in agriculture, from 1 percent opportunities and development prospects. in Libya to 64 percent in Liberia (figure 5.2). FIGURE 5.2 Agriculture Dominates Highly Fragile Economies Share of agriculture in GDP of FCS, 2014 or latest year Agriculture 80 Higher degree of fragility LBR SOM Share of agriculture in GDP (percent) 60 SLE CAF TGO GNB 40 MLI BDI UZBCOMNPL TCD PNG MWI SDN KH TJK SLB M MMR CIV MDG AFG LAO CMR NGA 20 GIN COD GMB MRT SYR GEO ZWE BIH AGO LBN COG IRQ 0 LBY 20 40 60 80 100 120 Fragile States Index (FFP, 2014) Quadratic fit Non-FCS FCS Source: Computation based on the United Nations Statistics Division database on gross value added across sectors; Fragile States Index, the Fund for Peace. Note: Not all countries at high degree of fragility, according to FFP, are officially classified as FCS by the World Bank. The latter group is designated with red and labeled. FCS = fragile and conflict-affected situations; GDP = gross domestic product. 140 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 The agricultural sector itself is highly frag- Opportunities Grow during Transitions mented. The bulk of employment in FCS is from Conflict to Peace in the small farmer and household enter- Within the group of FCS, postconflict econo- prise5 sectors, driven by necessity and resil- mies offer significant new business oppor- ience rather than growth. tunities. The reestablishment of peace is Whether a country is at a high risk of associated with renewed investment confi- conflict, is in conflict, or is postconflict mat- dence and growth. In fact, evidence points ters for how prevalent different economic to distinct episodes of high growth in the activities are, explaining at least partly the wake of conflicts and many opportunities variation within the group. For example, for investment. Recent evidence shows that, construction accounts for a large share of a year after the end of conflict, FDI increases economic activity in such FCS as Lebanon, dramatically, and, three years after the end of which are not in full-blown conflict, or conflict, inflows about double relative to the countries where large reconstruction efforts last years of conflict (Mueller, Piemontese, are taking place, such as Afghanistan or and Tapsoba 2017). By sector, construction Angola. The weight of the sector in coun- and services experience high growth and tries with deep fragility and frequent peace- pull labor out of agriculture in postconflict to-conflict transitions like Somalia or Sudan years. An illustration of the average share is significantly smaller. More capital-inten- each activity gains or loses over a 12-year sive activities, such as manufacturing, period after peace is established (figure 5.3) exhibit reverse linear relationships with the suggests common trends across postconflict levels of fragility—specifically because of the countries6 and time periods. For example, capital flight in the face of fragility the weight of agriculture in gross domestic (IFC 2017). FIGURE 5.3 Postconflict Growth Clocks Median change in shares of GDP by sector 1–12 years postconflict, 1990–2014 a. Mining and quarrying b. Construction % % 0.4 0.4 0.3 Yea 0.3 Yea rs a rs a 12 fte 12 fte 0.2 rc 0.2 rc 11 1 o 11 1 o nfl nfl 0.1 0.1 ict ict has has 10 0 2 10 0 2 ceas ceas –0.1 –0.1 ed ed –0.2 –0.2 9 –0.3 3 9 –0.3 3 8 4 8 4 7 5 7 5 6 6 figure continues next page FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 141 FIGURE 5.3 Postconflict Growth Clocks (continued) c. Manufacturing d. Transport, storage, and communications % % 0.4 0.4 0.3 Yea 0.3 Yea rs a rs a 12 fte 12 fte 0.2 rc 0.2 rc 11 1 o 11 1 o nfl nfl 0.1 0.1 ict ict has has 10 0 2 10 0 2 ceas ceas –0.1 –0.1 ed ed –0.2 –0.2 9 –0.3 3 9 –0.3 3 8 4 8 4 7 5 7 5 6 6 Source: Computation based on United Nations Statistics Division database on gross value added across sectors on selected postconflict economies; Uppsala Conflict Dataset (1990–2014). Note: “Growth clocks” present for each sector the median year-to-year change in shares of gross domestic product (GDP) across economies that have recently transitioned from conflict to peace. The bars for each of the 1–12 years postconflict are illustrated at the positions of hours in a hypothetical clock. The inner circle represents zero growth, the blue bars represent a positive change, and the green bars a negative change. The exact year the conflict has ceased is identified using the Uppsala Conflict dataset, and the sample covers postconflict economies for the period 1990–2014. product (GDP) gradually declines after the resilience during conflict, which translates cessation of hostilities.7 into little transformation in the aftermath Of all economic sectors, construction of conflicts. shows the most pronounced growth in the aftermath of conflicts. The sector grows Foreign Investors Are Cautious in the short run in response to recon- struction efforts and fluctuates around a Investment opportunities exist in fragile and steady state over the medium term. Much postconflict situations but are generally hard of this growth represents an opportunity for foreign investors to exploit. Multinational for foreign firms (box 5.2). Higher rates corporations (MNCs) will choose to do busi- of growth in telecommunications and ness in FCS only when the reward outweighs, transport are apparent over the medium by a sufficiently large margin, the risk. In term— infrastructural weaknesses possi- addition, MNCs will tend to concentrate in bly explaining the time lag in growth. The activities where there is limited domestic necessary conditions for diversification competition, owing to advantages enjoyed only materialize after a substantial period. by domestic firms in markets where the polit- Manufacturing, for example, tends to ical economy is distorted. exhibit slower growth in postconflict econ- High rewards and low competition occur omies, specifi cally because conditions for simultaneously only for selected natural its growth take more time to materialize.8 resource and other capital-intensive activities, In contrast, mining and other sectors that which depend on high demand outside FCS. rely on natural resources remain stable This exact pattern is confirmed by comparing throughout, possibly because of the sectors’ the distribution of sectoral shares in aggregate 142 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 BOX 5.2 Postconflict Growth in Construction and FDI Opportunities Construction opportunities abound in postconflict and they do not have prior experience with such con- countries, where sizable funds are available from tracts or how to bid for them. A snapshot of recon- donors. For example, in the fi rst decade after con- struction efforts in Haiti in 2012 shows that, of the fl ict ended in Lebanon, the country received about billions spent by the U.S. Agency for International $10 billion for reconstruction, while Bosnia and Development, more than 99 percent went to for- Herzegovina received $5.4 billion in the same period. eign firms. The extent to which the local private sec- How much of this activity actually benefits foreign tor benefits from the presence of multinational firms firms and investors? A disproportionally large part. depends on supply linkages, the development of which Local firms are at a disadvantage in seizing these remains a priority in many postconflict contexts. opportunities for several reasons: they lack the capac- ity and skills to carry out large, complex projects, Sources: Bray 2005; Porter Peschka 2011; Ramachandran and Walz 2012. investment inflows across FCS and non-FCS industries coupled with the difficulty of bring- low-income countries. While countries in these ing in skilled expatriate staff. Finally, inves- two groups show significant variation, FCS tors tend to concentrate their investment exhibit systematically different shares in four spatially in the most stable territory of the broad industries: extractives (mining, petro- fragile countries. leum, mineral products), construction, forestry and fishing, and food and beverages. Of those, Understanding the Context Helps in only construction, and food and beverages rely Seizing Opportunities largely on local demand, supplemented in some cases by foreign aid. The opportunity is From capturing local demand and mitigat- presumably generated by the absence of down- ing operational risks to avoiding uninten- stream value-chain development and capital tional consequences, a deep understanding scarcity for large-scale production. All of the local context is necessary for success- these sectors are relatively capital-intensive ful foreign investment. While this applies (figure 5.4). for international business in any context, it But investors are more cautious when is particularly relevant for investment in they enter FCS markets, as revealed by green- FCS. Firms employ many strategies in field investment patterns across countries doing so. (figure 5.5). In natural resource sectors, the Engaging with the local private sector in range of their choices on scale and location domestic supply chains features prominently are bound by the location and volume of among strategies of foreign firms. Local firms reserves. By contrast, in sectors other than tend to have a higher risk tolerance, know the extractives, the more fragile a country, the less local market and political economy, and have investors will tend to commit to large proj- contacts with the authorities that mitigate ects. Avoiding financial exposure at the begin- risks faced by MNCs (USAID 2016). Some of ning makes sense where there is significant the risk borne by these entrants can be shared uncertainty. Investors also tend to commit to with local suppliers, for example, through fewer jobs for every dollar they invest in FCS. license agreements or “contract manufactur- These patterns are probably due to the con- ing,” both of which are safer for MNCs than centration of projects in capital-intensive joint ventures (Campbell 2002). FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 143 Operating in a so-called conflict- FIGURE 5.4 Foreign Investors Concentrate in Natural Resources sensitive manner is another strategy deeply and a Few Other Capital-Intensive Activities Distribution of sector shares in inward FDI flows across FCS, 2008–14 rooted in understanding the local context. Firms in a fragile context stand to aggravate Food,beverages, and tobacco local tensions unintentionally by dispropor- Textiles, clothing, and leather tionally employing staff from one community Hotels and restaurants or another, providing revenue for authorities Agriculture and hunting that engage in human rights violations, or Transport and communications training security forces that can later be Construction deployed in conflicts. To avoid such pitfalls, Chemicals and chemical products and the associated risks to their businesses, Electricity, gas, and water large MNCs increasingly add to their operational Mining and quarrying policy such concepts as “do-no-harm” or Forestry and fishing “conflict sensitivity,” which originated in the Finance development and humanitarian community. Nonmetallic mineral products Adopting a conflict-sensitive approach means Petrole um that a company invests in understanding the context in which it operates, becomes aware 0 0.2 0.4 0.6 0.8 1.0 of potential positive and negative effects it FCS Low-income non-FCS may have on a conflict environment, and Source: Computation based on Investment Map Database, International Trade Centre; World takes all the necessary steps to avoid causing, Development Indicators, World Bank. or worsening, conflict. Note: The distribution of shares of sectors in total FDI inflows across all FCS (in blue) is compared with the same distribution across all low-income and lower-middle-income non-FCS countries On all these accounts, regional MNCs may (in green) for which data exist after 2008. Each horizontal box illustrates the median of the distribu- have a comparative advantage in these chal- tion across the two groups with a black line; the box delimits the 25th percentile (left) and 75th percentile (right) of each distribution—i.e., the top and bottom quartile; and the lines extend- lenging contexts relative to global firms. This ing from the box illustrate the full range of shares. FCS = fragile and conflict-affected situations; category includes, for example, companies FDI = foreign direct investment. FIGURE 5.5 Outside of Natural Resource Sectors, Investors Are Cautious Characteristics of greenfield FDI project announcements, nonextractives, 2008–16 Size Jobs Spatial concentration Median project size (US$, millions) 80 10 1.0 Jobs created per million US$ Ratio of highest sub-national 8 0.8 60 entity to aggregate 6 40 0.6 4 0.4 20 2 0.2 0 0 20 40 60 80 100 120 20 40 60 80 100 120 20 40 60 80 100 120 Fragile States Index (2014) Fragile States Index (2014) Fragile States Index (2014) Quadratic fit FCS Non-FCS Source: Computation based on fDi Markets database, the Financial Times; Fragile States Index (2014), the Fund for Peace. Note: The sample does not include extractive industries. Not all countries that are highly fragile according to FFP are officially classified as FCS by the World Bank. The latter group is designated with red. Green represents all other countries. FCS = fragile and conflict-affected situations; FDI = foreign direct investment. 144 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 from Lebanon investing in other countries in suggests that it represents a considerable the Middle East and companies from amount (figure 5.6). While the footprint of Morocco and Nigeria expanding into places France and the United Kingdom remains large in West Africa. These firms leverage the fact in Africa and the Middle East, greenfield that they are “local” in a particular area of investment, for example, from the Russian the world, and have sufficient affinity with Federation to Uzbekistan, Malaysia to their target markets because of the similarity Cambodia, South Africa to Nigeria, Japan with their home market. Lower risks resulting and Thailand to Myanmar, and the United from this affinity would tend to make inves- Arab Emirates to Iraq confirm that intrare- tors, in principle, commit to larger projects, gional investment takes place in FCS on a and lower information asymmetries for local large scale. recruitment that can make a greater difference in terms of stability and in catalyzing further investment. These firms deserve special atten- Barriers to Investment: Risks and tion from the development community. But how much of greenfield investment Obstacles comes from regional firms? The evidence The previous sections have painted a clear picture of the investment potential for most FIGURE 5.6 Regional Investment Occurs on a Large Scale fragile and conflict-affected situations. Even Origins of greenfield FDI project announcements in FCS, 2008–16 at its lowest benchmark, this potential FCS recipients remains unfulfilled. The data also show that investors are cautious and keep a small foot- print, creating fewer jobs relative to similar investments in less risky environments and Pap Congo, Rep. Zimbabwe ua concentrating in capital-intensive sectors. Cambo Cam Ne e Understanding the reasons for these trends wG Franc eroo s Uz te dia will help create deeper and more inclusive uin ta be dS n ea kis markets in FCS and expand their investment ite tan Un An opportunities. go la an Several global data sources document what Jap investors and businesses perceive to be the irates Arab Em biggest obstacles hindering their ability to United expand their investment in a given market. Myanmar Russian Federation Among these sources are the World Bank Group’s Enterprise Surveys9 and the World Thail and Economic Forum’s (WEF) Executive Opinion Kor q ea, Surveys.10 Surveys of business executives such Ira Un Rep . ite dK as these are frequently used to measure M ing perceptions about problems whose severity au do Ca rit m can be compared across countries and over Sou na ius ria Rest Malaysia da India Netherlands Nige th A time. These surveys are particularly appealing of O f when quantitative data are either unavailable rica ECD or difficult to gather, as in many FCS. This section looks into the recent findings Origin of greenfield investments of the Executive Opinion Survey and the Intraregional Interregional Enterprise Surveys to examine what they tell Source: Computation based on Financial Times FDI Markets database, the Financial Times. us about the obstacles and risks that limit the Note: Origins (on the right side of the chord diagram, in orange) and FCS destinations (on the willingness to invest in FCS. Building on these left side, in red) of greenfield projects exceeding $3 billion since 2008. Blue chords indicate intraregional investment. FCS = fragile and conflict-affected situations; FDI = foreign direct findings, the section analyzes the institutional investment; OECD = Organisation for Economic Co-operation and Development. realities in FCS to understand the scope for FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 145 government action to promote foreign well as shortages in the supply of water, are investment. more common in FCS than in non-FCS, according to several surveys (Speakman and Rysova 2015). In the Republic of Yemen, for Charting the Obstacles Facing Investors example, three out of four firms surveyed by The fundamental questions that arise from the World Bank in 2013 reported power the analysis of market conditions and risks outages as a major constraint on their oper- facing businesses in fragile situations are: ations. Because of similar power grid fail- ures, in South Sudan two-thirds of all power • How pervasive the challenges are (that is, consumed by firms in 2014 was produced by how disruptive risks and market condi- privately owned generators, imposing added tions can be for business), and costs of operations, an upper limit to their • How specific they are to FCS (that is, scale, and narrower returns to investment how distinct they are to fragility and (Speakman and Rysova 2015). The numbers conflict rather than a specific level of are equally striking in other domains and development). countries. Banking penetration in Guinea- However imperfect, the WEF Executive Bissau, for example, remained below Opinion Survey offers answers to both 1 percent of the population in 2013, and questions. Average perceptions on the inten- access to finance was cited by three out of sity of constraints across fragile states shed four businesses as an important constraint light on the major challenges. And the dif- for business operations, on par with electric- ference in averages between FCS and non- ity (Arvanitis 2014). The constraints identi- FCS low-income countries determines how fied by executives are also interrelated. Low FCS-specific a problem is. By charting these local demand, for example, boils down to two variables on a scatter plot (figure 5.7, widespread poverty, limiting the volume of severity on the horizontal axis and FCS business activity that the local population specificity on the vertical axis), four groups can sustain, while foreign markets often of challenges can be distinguished: those remain out of reach because of the poor that are both severe and FCS-specific (top- quality of transport infrastructure. right hand corner of the panel); those that are severe but similar to what is experienced Institutional Constraints Are in other low-income countries (bottom Severe and Diverse right); those that are FCS-specific but not severe (top left); and the remaining vari- Business executives also identify a number of ables, which are less relevant on both institutional constraints that hinder business dimensions (bottom left). expansion in FCS. Two clusters of institu- tional concerns can be identified: one relates to property rights and the means for their Operational Constraints Are enforcement, and the second concerns the Most Pervasive quality of public governance. Executives Operational constraints are high on the responding to the survey identified weak- mind of surveyed businesses as an obstacle nesses in intellectual property rights, judicial to growth, affecting the opportunity for independence, and hence dispute settlement investment in fragile environments as severe obstacles in FCS. Weakness in (figure 5.7). The quality of electricity is at property right regimes was given a score of the top of this list, followed by constraints severity below the median, which may be related to the size of markets (domestic and more reflective of the small footprint that foreign), transport infrastructure, and access investors keep in FCS and the coping mecha- to finance. The results are hardly surprising: nisms they deploy (including political risk frequent and prolonged power outages, as insurance). 146 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 5.7 Perceptions on Severity and FCS-Specificity of Challenges, 2016 –1.3 Quality of electricity supply –1.2 Less severe but specific to FCS Severe and specific to fragile and –1.1 conflict-affected situations –1.0 –0.9 Foreign market size index Domestic market size index –0.8 FCS–specificity (lower is more specific) Quality of overall infrastructure –0.7 Internet access in schools Quality of air transport infrastructure Production process sophistication –0.6 Ease of access to loans Quality of port infrastructure Regulation of securities exchanges Quality of roads Quality of railroad infrastructure –0.5 Affordability of financial services Financing through local equity market Soundness of banks State of cluster development Financial services meeting business needs Intellectual property protection Reliability of police services Judicial independence –0.4 Property rights Buyer sophistication Median Transparency of government policy making Irregular payments and bribes Protection of minority shareholders’ interests Ethical behavior of firms Venture capital availability –0.3 Business costs of terrorism Diversion of public funds Wastefulness of government spending Business costs of crime and violence –0.2 Efficacy of corporate boards Efficiency of legal framework in challenging regs. Organized crime Favoritism in decisions of government officials –0.1 Efficiency of legal framework in settling disputes Public trust in politicians 0 Burden of government regulation 0.1 Less severe and less specific to FCS Severe but less specific to FCS 0.2 Median 5.0 4.8 4.6 4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 Severity in FCS (lower is more severe) Pillar Financial market development Infrastructure Institutions Other categories Source: Computation based on Global Competitiveness Index database, World Economic Forum. Note: FCS = fragile and conflict-affected situations. The quality of public governance is also a approach to private sector development in major obstacle to private investment in FCS. FCS. Executives operating in FCS rank the Issues of irregular payments and bribes, weak burden of government regulations below public trust in politicians, and favoritism in average in terms of severity. It also ranks very decision making by government officials were low in terms of FCS-specificity. This is cor- perceived as severe obstacles. roborated by the findings of the Enterprise The analysis of the WEF Executive Surveys, which show that the amount of time Opinion Survey Data points to another ele- that senior management spent on dealing ment that strongly influences the choice of with government regulations is lowest on FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 147 average in FCS, at 8.6 percent, rising to FIGURE 5.8 Senior Management Spent Less Time Dealing with 11 percent in past-FCS (figure 5.8). Thus, the Government Regulations in FCS Average share of senior management time spent dealing with the requirements of problem in FCS may be less one of regulatory government regulation (percent) burden and more the absence of needed market regulation. 12 10 The Link between Operational and Institutional Constraints 8 The analysis now turns to the state of public Percent 6 institutions in FCS using global indicators. It aims to illustrate that, while institutional weaknesses are a defining feature of FCS, 4 cross-country variations are significant for the design of private sector development 2 approaches in FCS. Institutional weaknesses in FCS are partly the cause of operational constraints that FCS Past-FCS Other low-income Other lower- worry foreign investors. Government capac- countries middle-income ity, regulatory effectiveness, and institutional countries quality are fundamentally interconnected Source: Computation based on Enterprise Surveys, World Bank. with market conditions that constrain busi- nesses. Infrastructure projects, for example, require a minimum government capacity to deliver 11 but also basic regulation and workers—and may not develop a retail market enforcement to protect investor property for several years after the end of the conflict, rights. until such regulatory conditions are met. In Examples of the relationship between insti- addition, a key reason that foreign market tutional and operational constraints abound. access is prohibitively expensive for firms, Successful power projects in Afghanistan, including MNCs, in fragile countries is the Côte d’Ivoire, Guinea, Iraq, Mali, Myanmar, logistical burden of certification require- and Nepal all involved extensive work devel- ments, corruption in customs authorities, oping regulations and government sector and other failures directly related to institu- plans, building the relevant government tional and governance weaknesses (Hoeffler capacity, providing up-front advisory 2012). resources in project development, and sup- To conclude, investors and businesses porting complementary government invest- face severe challenges in FCS. The obstacles ments (for example, electricity distribution range from market characteristics to infra- and provision of co-financing support and structure and access to finance constraints risk guarantees) (Mills and Fan 2006, 29; combined with a myriad of institutional USAID 2009, 45, 48). constraints. Institutions in FCS are weak The same applies to development of and the weakness has persisted over the financial services to address the scarcity of years. There are, however, significant varia- capital. Banks avoid setting up operations in tions in weaknesses among these countries territories without viable banking laws and (figure 5.9). These variations matter in foreign exchange regulations (Bray 2005). determining the best approach to facilitating Initially, they tend to concentrate on interna- and attracting investment in a particular tional customers—such as diplomats and aid country. 148 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 5.9 Varying Levels of Fragility and Government As the challenges in FCS mount and the Effectiveness among FCS pressure for short-term returns on reforms increases, policy makers tend to de-emphasize 2 broad-based deep reforms in favor of inter- Government effectiveness score ventions with quicker yields in terms of job 1 creation and investment flows. This tendency, however, poses some risks. The transforma- 0 tion from a conflict economy to a peace econ- omy, and from a fragile market to a resilient –1 and inclusive one, is not possible without changing the rules of the game. A targeted –2 approach that strengthens the institutional foundation for the market is therefore the 20 40 60 80 100 120 way to go. Fragile States Index This section analyzes the limits of tradi- Past-FCS Non-FCS FCS tional approaches to investment climate reform and outlines a path forward to suc- Source: Worldwide Governance Indicators, World Bank; Fragile States Index, the Fund for Peace. Note: Government effectiveness is an aggregate indicator that reflects perceptions of the quality cessfully de-risk investment and expand of public services, the quality of the civil service and its independence from political pressures, the investment opportunity through reforms that quality of policy formulation and implementation, and the credibility of the government’s commit- ment to such policies. This is part of the Worldwide Governance Indicators dataset. take into account countries’ institutional capacity, the investment opportunities that they offer, and the nature of conflict and Investment Climate Reforms Can instability. Create Markets and Maximize Investment Traditional Approaches to Investment climate reforms are essentially Investment Climate Reforms legal, regulatory, procedural, and institu- tional reforms that enhance a country’s Have Their Limits investment competitiveness. Such reforms Private investment attracted attention as an can affect, in different ways, the risk–return area of focus in FCS development only in the equation that investors use to make their early 2000s. A 2013 review conducted by investment decisions. Some reforms reduce the World Bank Group’s Independent risk to investors by improving the transpar- Evaluation Group (IEG) of the investment ency and predictability of investment policy climate portfolio identified some 120 proj- making. Other investment climate reforms ects implemented in FCS with an average of contribute to increasing investment oppor- 12 active projects a year. This indicates sig- tunity and maximizing the return on invest- nificant attention paid to private sector ment by facilitating access to the market issues in FCS in the past decade. or by encouraging clustering and interfirm Traditional investment climate reforms in linkages. FCS have tended to focus mostly on business Analysts agree that investment climate licensing, permitting, and administrative bar- reforms are necessary but insufficient for pri- riers to the growth of the private sector, as vate sector development in fragile countries. well as investment promotion and public– Where they diverge is on the appropriate tim- private dialogue. Many FCS have used ing and sequencing of reforms (box 5.3) and Doing Business 12 indicators to map and the balance between broad-based interven- frame their strategy for business environ- tions and direct interventions that benefit spe- ment reform, focusing initially on simplify- cific communities, firms, sectors, economic ing burdensome administrative processes spaces, or identified value chains. (IEG 2015). FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 149 BOX 5.3 Prioritizing Economic Reforms The World Development Report 2011 on Conflict, Regulatory simplifi cation and removing barriers Security and Development (World Bank Group 2011) to investment entry were identifi ed by the Report as identifies legitimate institutions as the common “miss- good confidence-building signals that can yield early ing factor” in countries affected by violence relative to results. In the same vein, addressing infrastructure those that do not slide into violence despite compa- constraints, such as access to electricity and tran- rable threats and stresses. The 2011 Report found that sit, were also identified as good early confidence- countries with good governance indicators have 30 to building interventions that can stimulate the private 40 percent lower risk of civil war than their peers with sector. The Report also highlighted the importance weaker governance indicators. Therefore, the path to of value chain development through skills building, resilience must be through institutional transforma- access to finance and technology, and connecting tion. Yet institutional reform is difficult, even more so producers to markets as a second stage of inter- in countries starting from a low base. Prioritization, vention suitable for fragile environments. Deeper thus, is a central theme of the path out of violence as institutional reform such as, privatization, may envisioned by the WDR. take longer and may not be suitable for early-stage The Report advocates prioritizing “ending and pre- interventions. venting violence” as the main impact that all interven- The Report stresses, however, that prioritiza- tions in fragile and confl ict-affected situations should tion should be based on the local context. Priorities aim to achieve. Armed by research and analysis, the should be identifi ed not based on a prototypical pre- Report identifi ed three key outcomes as essential to scription but rather on the basis of assessment of the achieving this ultimate objective: security, justice, reality in each country. The Report indicates that and jobs. It also showed how the three outcomes are countries with a long tradition of strong institu- interlinked. In Kosovo, for example, creating jobs by tions, such as some of the middle-income countries encouraging regional trade depended on securing the affected by confl ict, may be able to take on more main road connecting Kosovo to neighboring coun- ambitious institutional transformations at an early tries. In Mozambique, providing livelihood oppor- stage that other countries affected by confl ict may tunities to ex-combatants was essential to achieving be unable to do. security. The central message of the Report is that strength- As far as economic reform is concerned, the Report ening legitimate institutions that can provide citizens defi nes job creation as the priority outcome that all with security, justice, and jobs is crucial for break- efforts should be geared toward. As such, this out- ing the cycle of violence. This institutional transfor- come determines the sequencing path identified by the mation, however, should adopt “best-fi t” not “best Report. It argues for starting the process by building practices” approaches. Institutional transformation confi dence through signaling change and achieving takes time. It took the fastest reforming countries in short-term results and moving from that to transform- the 20th century 20 years to achieve a functioning ing institutions. It also stresses that this process is a bureaucratic quality. So, proceed with realism and repeated one with transition being an ever-expanding recognize that the “scope and speed of reforms are spiral of change. themselves risk factors.” Evaluations of these reforms did not find investment climate reforms must go well clear evidence of the relationship between beyond simplifying procedures and must simplification and investment flows or job respond more clearly to the challenges and creation. They also questioned the realism of characteristics of FCS. such reform efforts considering the low levels There is value in prioritizing the simplifica- of institutional capacity and political commit- tion of business regulations over revamping ment found in many FCS. It is now clear that and expansion at the early stages of reform 150 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 in FCS. The rationale for this approach is that sectors, should get priority. Targeted such reforms give the necessary signal of approaches to reform, thus, can be seen to friendliness toward business and mark a have a greater influence on reforms in FCS. departure from the past. They have also been A market-creation approach to investment seen to produce short-term results needed to climate reform would focus on reducing the build confidence in the reforms (World Bank risk to investment in fragile countries, and on Group 2011, 157–66). expanding the investment opportunity and It therefore bears noting, before outlining maximizing its rewards. Moreover, the risk– an approach that takes into account the limits return equation differs by type of investor. of traditional approaches to reform, the con- Investors with affinity for the jurisdiction— tinued relevance of such approaches as part of such as local investors, diaspora investors, or a more targeted package of reforms: investors from neighboring countries with cul- tural ties to fragile countries—are equipped 1. Improving the business environment with local knowledge that may offset some of with the guidance of the Doing Business the risks precluding other investors. Noting indicators gives reformers in FCS the quick this distinction is important for designing poli- wins needed to sustain the momentum for cies that remove obstacles to investment faced reform. by this amenable group of investors. 2. Doing Business reforms cut across govern- ment agencies and, when implemented De-Risking: Reducing Risks Faced by effectively, they can be an opportunity for Investors building a coalition of reformers. 3. Simplifying regulations and removing The defining risks of a fragile country for obsolete rules is a key step toward freeing investors, domestic or foreign, are: the capacity of government to regulate • Security risks arising from political con- effectively and reduce opportunities for flict or private criminal violence, and rent-seeking. • Political risk arising from institutional fragility. New Approaches to Investment Investment climate reforms that better protect investments, improve transparency, Climate Reforms Create Markets and encourage rule-based decision making The agenda for investment climate reform in reduce the perception of political risk fragile countries is long, yet institutional among investors. Spatial solutions that capacity is typically low and patience for create secure zones for investors to operate results limited. Thus, such reforms must be also contribute to reducing the security designed for the long-term goal of institution risk and make investment opportunities building with an eye on the short-term goal more accessible. of creating jobs and attracting investment. Existing investors are the first category of The long-term effort of building institutions investors that should be targeted by de-risking and developing regulatory capacity should interventions. As noted earlier, FCS attract be combined with faster-yielding reforms their own pioneer investors, albeit at a lower that target priority sectors and support value rate. When the return to investment exceeds chains. the cost of risk, investors come. They often Sectors offering the most immediate prom- initially invest mainly in the extractive sector, ise should have priority in long-term interven- but also in telecommunications, finance, and tions. For example, in postconflict construction. The first step for investment cli- construction booms, construction permit mate reform in fragile countries is to identify reforms and removal of entry barriers that the pool of existing investors and to set up benefit the construction sector, among other systems for investor aftercare and grievance FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 151 redress so that these investors remain. This value chains. Variations of these approaches approach works for all FCS regardless of their have been tried in fragile states with mixed level of institutional capacity. results. A key difficulty is the requirement for Government services that seek to retain sufficient state capacity in formulating and investors should also target domestic investors. implementing a coherent, responsive, and rea- These investors, especially high-growth sonable SEZ package, without either failing ones, also leave the country if the risks to do so or being captured by vested interests exceed rewards. So investor retention inter- (AfDB 2015). ventions that reduce the risks to investment In Iraq, where more than 50 percent of the can also be used to prevent this type of capi- population were affected by conflict in 2016, tal flight and protect domestic private sector private investment flowed to more stable capacity. regions, such as Basra in the South and the Recent unpublished investor surveys con- Kurdish region in the north. Institutional ducted as part of the World Bank Group’s reforms to encourage private sector develop- engagement in FCS have revealed that inves- ment were undertaken at the subnational tors in these economies are well acclimated to level. In Iraq, with natural resources, large the risks of violence and terrorism but are less population, high GDP per capita, and long willing or able to handle the challenges posed institutional tradition, severe and widespread by adverse regulations or cumbersome pro- conflict did not preclude opportunities for cesses. In one case, investors indicated that both investment and reform. the number one reason for considering divest- ment and relocation out of a particular Maximizing Investment Opportunities market was regulatory and procedural constraints.13 Encouraging Formalization and Supporting Investor aftercare systems and grievance Firms with High Growth Potential redress mechanisms should take into account One of the key effects of conflict and insecu- the government’s institutional capacity. They rity is excessive business informality. In should also reflect the political economy of response to conflict and fragility, high-potential the country. In FCS, investors’ grievances are domestic firms tend to flee the country while as likely to arise from formal government small firms tend to be informal and “go action as they are from informal rules and under the radar” to avoid harassment or institutions such as customary laws and tribal extortion by public authorities. These trends authorities. Any mechanism set up to identify reduce the size and productivity of economic and address such grievances should be able to activities, and undermine market develop- influence formal and informal decision ment. They increase the cost of operations as making (Echandi 2013). they mandate reliance on foreign input. In Targeting investment climate reforms to some cases, where needed inputs cannot be subnational regions that demonstrate higher secured, they may render the investment levels of security and stability is another way opportunity unrealizable. Encouraging for- of lowering the risk to investors and creating malization and supporting domestic firms safer spaces for economic activity. This with high growth potential is therefore a key approach can be combined with special eco- component of a private sector development nomic zones (SEZs) or other types of spatial strategy in fragile states. solutions to reassure investors in FCS. In While not all economic activities have to addition to minimizing geographical exposure be formalized, a high degree of formality is to conflict, SEZs can help address several necessary for markets to be created and for other problems, such as infrastructural, regu- investment to flow. Domestic firms cannot latory, or skills deficiencies. At a critical mass attract equity investment and foreign inves- of companies, the zones can also foster tors cannot enter the market by partnering knowledge and skills transfer along local with domestic firms without formalization. 152 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 Investment climate reforms that help high- targeted investment promotion and attrac- growth economic activity shift to the formal tion. The political economy of diaspora sector are key to this process. Depending on engagement varies from country to country, the degree of fragility, the demand for reform and tailored strategies that take the reality of can be as basic as setting up a well-functioning conflict into account are critical. company registration process and introducing appropriate company laws. In other contexts, Taking a Regional Approach other incentives to formalization may be Many FCS are characterized by small domes- needed. tic markets and weak institutional capacity, which limits their ability to attract investment Linking Domestic Firms to Foreign Direct and mitigate risks for investors. For this rea- Investors son, a regional approach to investment cli- Another cost of conflict and fragility is the mate reform can enhance the market-creation fragmentation of the market and the loss of potential of the intervention. Interventions firm clusters and cross-sectoral linkages. The can benefit from a regional dimension in sev- underdevelopment of business clusters poses eral ways: a severe constraint specifically for fragile countries. This, combined with the typically 1. The investment opportunity for some small size of the local market, underscores FCS may lie in a large neighboring the importance of focusing on investment market. Investment opportunity derived climate reforms that target the development from market size is measured not just of local suppliers and link them to foreign by the size of the domestic market, or by investors operating in the country. access to the global market, but also by Since investment in fragile states concen- the size of the regional market bordering trates reforms in a small number of sectors— the fragile country. A small fragile state, such as extractives, construction, and such as Bosnia and Herzegovina or telecommunication, with variations across Kosovo, secures significant investment countries—to support the development of opportunities through its proximity to linkages, they should focus on sectors that affluent regional markets. Investment attract investment in the specific countries. climate reforms that aim to develop the domestic private sector and attract foreign Targeted Investment Promotion Efforts investment must be designed with this In addition to conflict and fragility, one of potential in mind. the key inhibitors of investment flows to FCS 2. One of the key reasons for fragile and con- is the lack of reliable and accessible country- flict-affected situations’ low growth and level information important for investor weak trade is a lack of investor confidence decision making. Better access to informa- and a high perception of risk. Commitment tion may help offset the adverse impact of mechanisms are needed for these countries poor country image and reputation that to signal commitment to change that result from media reporting of conflict and assures investors and raises their confi- fragility. For this reason, reforms must build dence (World Bank Group 2011, 283–84). the capacity of the country’s institutions to Regional integration agreements with carry out targeted investment promotion. market access commitments and legal har- The country must also be able to map its monization initiatives offer fragile states investment opportunities and identify sectors an opportunity to signal commitment by with potential for investment attraction. participating in such agreements and in Finally, as noted earlier, highly skilled labor their mutual monitoring mechanisms. and large domestic investors tend to flee the 3. Cooperation among regional actors to country during conflict. This potential pool of pool technical and administrative diaspora investors also demands a strategy of resources can compensate for lack of FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 153 institutional capacity in FCS (World Investment in FCS is thus well below Bank Group 2011, 283–84). Such potential. It is also concentrated in a limited approaches may be considered a part of number of capital-intensive sectors and cre- investment climate reform. For example, ates fewer jobs than it would in less fragile neighboring countries can set up national environments. Investors are naturally cau- quality infrastructure and standards nec- tious. If a firm decides to invest, the rewards essary for implementing them as shared must outweigh the risks. But the high risks in regional institutions. Good and well- fragile states render many investment oppor- enforced product quality standards are a tunities unviable. prerequisite for market access and com- Firms operating in FCS have several petitiveness in foreign markets. options for responding to the obstacles they face and to minimizing costs, risks, and Neighboring countries and countries challenges. Strategic choices of multination- within the same regional block have an incen- als in terms of scale, staffing, and location tive to support fragile states in transitioning often aim to address multiple challenges and out of fragility. Conflict dynamics do not stay risks at once. For example, hiring local staff within borders and both reputational and provides access to local intelligence that conflict risk tend to spill over to neighboring helps mitigate security risks, as well as countries. Regional organizations thus have a engage the local community. Some of the growing role to play in reducing fragility strategies documented by interviews with within their regions. investors (IFC 2017) include integrated In summary, investment climate reforms management and due diligence systems; are necessary for markets to move from con- strategically locating warehouses and pro- flict to peace, and from fragility to resilience. duction sites; tiered investments; interna- Deep changes to the rules of the game are tional standards; flexibility in scale, supply, essential. The limited capacity of governments and business plans; and supporting govern- in FCS, combined with the urgent need for ment functions. quick and positive returns on reform efforts, Investors with more knowledge of the local require a balanced strategy that substantially context, such as regional investors and dias- enhances the investment climate in the coun- pora investors, have still other mechanisms try. De-risking and retaining investment, tar- for coping in fragile states. Such investors geting investment promotion toward realistic familiar with the FCS environment tend to be investment opportunities, and optimally for- able to cope better, take more risk, and accept malizing the economy to promote linkages lower returns. Data show that intraregional between foreign and domestic investment are investment flows to fragile states are growing. key elements of such a strategy. These trends underscore the importance of regional sources of investment and regional approaches in transitioning out of fragility. As development assistance becomes more Conclusion constrained relative to the demands of recon- Investors in fragile countries face a wide struction and development in FCS, the role range of adverse market conditions, although of private investment in moving countries some are similar to what they face in other out of fragility will continue to grow. This developing markets—such as shortages of underscores the need for active strategies for skilled labor, capital scarcity, and infrastruc- attracting investment and for developing the ture shortcomings. The severity of these con- private sector in FCS. ditions in FCS, combined with security risks The central message of this chapter is that, and lack of institutional capacity and legiti- considering the centrality of the economic macy, create a seriously deficient investment underpinnings of conflict, graduating from climate in FCS. fragile status requires serious modifications to 154 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 conventional methods of economic develop- nondisclosure of CPIA ratings; IBRD coun- ment. Market-creating investment climate tries included here qualify only by the reforms that reduce the risk to investors and presence of a peacekeeping, political, or maximize the opportunity are crucial to peace-building mission—and their CPIA rat- success. ings are thus not quoted here. Firm-specific strategies are clearly limited The 2017 list of FCS includes Afghanistan, in what they can achieve. And, although they Burundi, the Central African Republic, Chad, can keep a company out of harm’s way, they the Comoros, the Democratic Republic of cannot address the risks associated with fra- Congo, Côte d’Ivoire, Djibouti, Eritrea, The gility holistically or permanently. Given that Gambia, Guinea-Bissau, Haiti, Kiribati, many risks are inherently part of the defini- Kosovo, Liberia, Madagascar, Mali, the tion of fragile states, even a large company’s Marshall Islands, the Federated States of strategies can go only so far in addressing Micronesia, Myanmar, Papua New Guinea, them. Investment climate reform tailored to Sierra Leone, the Solomon Islands, Somalia, the FCS context can go a long way toward South Sudan, Sudan, Togo, Tuvalu, and the reducing investors’ risks and creating markets Republic of Yemen. Territories: West Bank for investment. and Gaza. Blend: Zimbabwe; IBRD only: Iraq, Lebanon, Libya, and Syria. Countries that have appeared in the World Bank list since the first compilation include Annex 5A. Definitions of FCS Angola, Bosnia and Herzegovina, Cambodia, The World Bank Group’s (WBG) Fragile, Cameroon, the Republic of Congo, Georgia, Conflict and Violence Group (formally Guinea, the Lao People’s Democratic the Center on Conflict, Security and Republic, Lebanon, Malawi, Mauritania, Development, CCSD) annually releases the Nepal, Nigeria, São Tomé and Príncipe, Harmonized List of Fragile Situations. The Tajikistan, Timor-Leste, Tonga, Uzbekistan, first such list was compiled in fiscal year and Vanuatu. 2006 and has had a series of classification changes from the Low-Income Countries Under Stress List (LICUS) (2006–09), to the Fragile States List (2010), to the current Annex 5B. Investment Harmonized List of Fragile Situations (2011–15). Expectations Based on Economic The concept and the list have evolved as the WBG’s understanding of the development Fundamentals challenges in countries affected by violence Predicted values of FDI and deviations to and instability has matured (see World Bank actual investment are used in the litera- Group 2016). ture to form expectations based on specific “Fragile Situations” have either a harmo- questions of academic or policy interest nized average Country Policy and (For examples, see Bellak, Leibrecht, and Institutional Assessment (CPIA) country rat- Stehrer 2008; Brenton and Di Mauro ing of 3.2 or less or the presence of a UN or 1999; Demekas and others 2007). Non- regional peacekeeping or peace-building mis- econometric estimations in the form of com- sion during the past three years. This list posite indexes have also been published for includes only IDA-eligible countries and non- the same purposes (see Maza and Villaverde members or inactive territories/countries 2015; Rodríguez, Gómez, and Ferreiro without CPIA data. Countries in the 2009; UN 2012), although they do not map International Bank for Reconstruction and directly onto flows of foreign investment. Development (IBRD) with CPIA ratings These exercises have clear limitations: they below 3.2 do not qualify for this list owing to depend on assumptions about the drivers of FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 155 foreign investment, on past records rather are not suitable for forward-looking country- than forecasts, and they are designed specific policy advice, nor for country to answer questions that vary from one study rankings. to another. In addition, estimations are In this exercise, predicted values of FDI constrained by data availability for specific flows are calculated to determine how much countries. Data constraints are particularly FDI inflows would be expected, based on acute for FCS, most of which lack a complete recorded economic fundamentals nett of an and up-to-date set of drivers. As such, the estimated effect of fragility. Separating the estimates serve only to illustrate the cost of negative impact of fragility from the pre- fragility at some specific point in time, and dicted inflow (that is, fitted value) of this TABLE 5B.1 Regression Coefficients OLS Estimation Variables I II III IV V GDP growth (percent) 0.023*** 0.026 0.026 0.052*** (0.008) (0.017) (0.017) (0.012) GDP (log) 1.136*** 0.876*** 0.909*** 0.919*** 0.938*** (0.035) (0.053) (0.022) (0.021) (0.029) –0.207*** 0.042 Population (log) (0.034) (0.061) Trade openness ((X+M)/GDP) 0.011*** 0.011*** 0.011*** 0.011*** 0.010*** (0.001) (0.001) (0.001) (0.001) (0.001) Natural resources (percentage of GDP) 0.003 0.008*** 0.008*** 0.008*** 0.014*** (0.002) (0.003) (0.002) (0.002) (0.004) Landlocked (=1) –0.113 –0.261*** –0.253*** –0.203** –0.339*** (0.071) (0.092) (0.095) (0.087) (0.116) Proximity to world markets 0.111 0.115 0.098 0.091 0.275*** ∑ (Foreign GDP/distance) (0.079) (0.093) (0.094) (0.093) (0.094) Fragile States Index –0.016*** –0.014*** –0.013*** –0.019*** (0.003) (0.003) (0.003) (0.003) –0.014*** Savings (percent of GDP) (0.004) Country fixed effects No No No No No Time (year) fixed effects Yes Yes Yes Yes Yes N = Number of observations 2074 882 882 884 738 R2 0.753 0.766 0.765 0.761 0.771 Source: Computation based on data sources in the note. Note: Standard errors are provided in parentheses under the estimated coefficients. ***p < 0.01; **p < 0.05; *p < 0.1. Data sources for the table are as follows: Variable Source FDI inflows (log) Investment Map Database, International Trade Centre Fragile States Index The Fund for Peace (2014) GDP (log) World Development Indicators (2016), World Bank GDP growth World Development Indicators (2016), World Bank Population World Development Indicators (2016), World Bank Savings World Development Indicators (2016), World Bank Trade openness World Development Indicators (2016), World Bank Distance to world markets CEPII (2012) Landlocked CEPII (2012) Natural resources CEPII (2012); World Development Indicators (2016), World Bank 156 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 FIGURE 5B.1 Expected Inward Investment countries, to avoid the absorption of the effect Varies across FCS of fragility by the idiosyncratic effect. Fragility and predicted FDI flows, 2008–14 The prediction is decomposed into two 14 vectors: (a) a vector of covariates unrelated to fragility, and (b) the estimated impact of fra- FDI potential (post-2008 average) 12 gility and the error. The Structural Prediction corresponds to the first vector for country i at 10 year t only: 8  ( t + f γ  +d Iit = a + y it β ) it  25%  = structural prediction + fit γ 6 Granular FDI flows by origin or sector of 4 activity are available for some of the FCS, but 20 40 60 80 100 120 information based on investment potential at Fragile States Index (2014) that level reduced the sample enormously and Source: Computations based on Investment Map Database, International was not preferred. Trade Centre. Note: The green line crossing the panel indicates the upper threshold of the bottom quartile. The scatterplot shows predicted FDI and levels of fragility with FCS indicated in red and other developing countries in green. FCS = fragile and conflict-affected situations; FDI = foreign Notes direct investment. 1. See annex 5A for definition of FCS and list of economies. 2. Capital flees uncertainty and conflict (Knight, regression would yield an estimate of that Loayza, and Villanueva 1996; Fielding 2004); expectation. and foreign firms are frequently targeted dur- Investment is modeled using the following ing insurgencies (Czinkota and others 2010; log-linear equation: Lutz and Lutz 2014). 3. The Liberian Diaspora Fund is an exam- Iit = a + yita + fitg + dt + eit, Iit ∈ +, ple, with remittances from Liberians abroad pooled and matched to investments in a vari- ety of sectors. where the dependent variable Iit corresponds 4. National Accounts data do not include the to the logarithm of investment inflows of informal economy, which can be substantial in country i at year t; yit is a vector of country fragile and conflict-affected situations. Latest characteristics (table 5B.1); fit corresponds to estimates of the size of the informal economy the FFP’s Fragile States Index; dt are fixed- for the period 2005–2010 reveal very high effects for year t; and εit is the error term. numbers: 69 percent in Chad, 77 percent in The sample comprises all but the high- DRC, 86 percent in Nepal, and 87 percent in income countries of the world and standard Mozambique (see Charmes 2012). The num- errors are heteroscedasticity-consistent bers on the formal economy also likely suffer (robust estimation). In addition, the sample from errors due to the resource constraints is not bilateral, hence the presence of few of statistical agencies (inadequately trained zeros, not warranting regressions based on staff; absence of resources for surveys; obso- lete monitoring systems) also preventing reg- special distributions. ular updates. Conflict, lastly, brings about Among five variants, the preferred specifi- shocks to demographic and economic struc- cation used for the investment presented cor- tures that statistical agencies are only able responds to the last column of table 5B.1, to capture years after violence has ceased. In without population as a measure of size, but Eritrea, Libya, and Syria, for example, GDP includes savings as share of GDP. Fixed effects figures have not been updated for the last are only includes for years but not for six years. FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 157 5. Household enterprises can include various the prolonged conflict. Despite the recon- service activities (for example, hairdressing, struction in the wake of the conflict, ports repairs, selling of goods), as well as indus- and other essential infrastructure could not trial activities (for example, making of satisfy local demands. charcoal, bricks, iron work, grain process- 12. The Doing Business project by the World Bank ing), and artisanal activities (for example, Group provides objective measures of business woodworking, dressmaking, construction). regulations for local firms in 190 economies Household enterprises in manufacturing tend and selected cities at the subnational level. See to be replaced over time by factories, so they http://www.doingbusiness.org/. disappear faster over time than household 13. Such surveys are often conducted as part of enterprises in services (see Filmer and Fox the diagnostics necessary for advising gov- 2014). ernments on the best way to improve the 6. Post-conflict countries, for the purposes investment climate. They are not published of this chart, include the subgroup of FCS and, thus, specifying the country where such where conflict has occurred since 1990, survey was conducted is not possible. in addition to 11 outside the official list: Algeria, Colombia, Ethiopia, Guatemala, Mozambique, Nicaragua, Peru, Rwanda, Sri Lanka, Uganda, and Ukraine. Bibliography 7. Afghanistan, the Central African Republic, AfDB (African Development Bank). 2015. Special the Republic of Congo, Sudan, and Economic Zones in Fragile Situations—A Zimbabwe are where the growth of services Useful Policy Tool? Abidjan: African as a result of the associated shift of labor Development Bank. from agriculture to other sectors has been Arvanitis, Y. 2014. “Providing Efficient Banking most pronounced. Services in a Fragile Environment—Structure, 8. Growth in manufacturing is only experi- Performance and Perspectives of the Banking enced in small countries such as Lebanon, the Sector in Guinea-Bissau.” West Africa Policy Federated States of Micronesia, Timor-Leste, Note No. 1, Abidjan: African Development or Tuvalu for reasons that are probably Bank. unrelated to transition along the postconflict Barbieri, Katherine, and Rafael Reuveny. 2005. continuum. “Economic Globalization and Civil War.” 9. An Enterprise Survey is a firm-level survey Journal of Politics 67 (4): 1228–47. of a representative sample of an economy’s Bray, J. 2005. “International Companies and private sector. The surveys cover a broad Post-Conflict Reconstruction—Cross-Sectoral range of business environment topics includ- Comparisons.” Social Development Papers, ing access to finance, corruption, infrastruc- Conflict Prevention and Reconstruction Series ture, crime, competition, and performance no. CPR 22, World Bank, Washington, DC. measures. Since 2002, the World Bank has Bellak, C., M. Leibrecht, and R. Stehrer. 2008. collected this data from face-to-face inter- “Policies to Attract Foreign Direct Investment: views with top managers and business own- An Industry-Level Analysis.” Presentation at ers in over 155,000 companies in 148 econo- the OECD Global Forum on International mies. See www.enterprisesurveys.org. Investment, Paris, France, March 2008. 10. The World Economic Forum’s (WEF) Brenton, Paul, and Francesca Di Mauro. 1999. Executive Opinion Survey is one of the most “The Potential Magnitude and Impact of comprehensive datasets that provides detailed FDI Flows to CEECs.” Journal of Economic insights into the challenges that firms face Integration 14: 59–74. across countries. In its most recent version Campbell, A. 2002. “The Private Sector and for 2016, 14,723 business executives from Conflict Prevention Mainstreaming: Risk 141 countries assessed their domestic markets Analysis and Conflict Impact Assessment Tools and countries on more than 80 variables. for Multinational Corporations.” Mimeo. 11. In Liberia, for example, government’s con- Carleton University. sistent underinvestment in infrastructure Charmes, Jacques. 2012. “The Informal Economy resulted in a poorly maintained public road Worldwide: Trends and Characteristics.” network and energy infrastructure, much of Margin: The Journal of Applied Economic which was subsequently destroyed during Research 6 (2): 103–32. 158 GLOBAL INVESTMENT COMPETITIVENESS REPORT 2017/2018 Collier, Paul. 2009. “Post-Conflict Recovery: How Knight, M., N. Loayza, and D. Villanueva. 1996. Should Strategies Be Distinctive?” Journal of “The Peace Dividend: Military Spending African Economies 18 (1): 99–131. Cuts and Economic Growth.” Policy Czinkota, M. R., G. Knight, P.W. Liesch, and Research Working Paper 1577, World Bank, J. Steen. 2010. “Terrorism and International Washington, DC. Business: A Research Agenda . ” Journal of Liu, C., and E. Harwit. 2016. “The Effectiveness International Business Studies 41 (5): 826–43. of Private Sector Development in Fragile and Demekas, D. G., H. Balász, E. Ribakova, and Conflict Affected Situations: Evidence from Y. Wu. 2007. “Foreign Direct Investment in Evaluations.” IFC, Washington, DC. European Transition Economies—The Role of Lutz, B. J., and J. M. Lutz. 2014. “Terrorism and Policies.” Journal of Comparative Economics Its Impact on Foreign Economic Activity in 35 (2):369–86. Sub-Saharan Africa.” Journal of Business and Echandi, R. 2013. “Complementing Investor- Economics 5 (4): 249–58. State Resolution: A Conceptual Framework Maza, A., and J. Villaverde. 2015. “A New FDI for Investor-State Conflict Management.” In Potential Index: Design and Application to Prospects in International Investment Law the EU Regions.” European Planning Studies and Policy, edited by R. Echandi and P. Sauvé, 23 (12): 2535–65. 270–306. Cambridge: Cambridge University Mills, R., and Q. Fan. 2006. “The Investment Press. Climate in Post-Conflict Situations.” Policy Fielding, D. 2004. “How Does Violent Conflict Research Working Paper 4055. World Bank, Affect Investment Location Decisions: Evidence Washington, DC. from Israel During the Intifada.” Journal of Mueller, H. F., L. Piemontese, and A. Tapsoba. Peace Research 41 (4). 2017. “Recovery from Conflict: Lessons of Filmer, D., and L. Fox. 2014. Youth Employment Success.” Policy Research Working Paper 7970, in Sub-Saharan Africa. Africa Development World Bank, Washington, DC. Series. Washington, DC: World Bank. Polachek, S. W., and D. Sevastianova. 2012. Fund for Peace. 2016. “Fragile States Index.” “Does Conflict Disrupt Growth? Evidence of http://fundforpeace.org/fsi/. the Relationship between Political Instability Gissinger, R., and N. P. Gleditsch. 1999. and National Economic Performance.” The “Globalization and Conflict. Welfare, Journal of International Trade and Economic Distribution, and Political Unrest.” Journal of Development 21 (3): 361–88. World-Systems Research 5 (2): 274–300. Porter Peschka, M. 2011. “The Role of the Hallward-Driemeier, M., and L. Pritchett. 2015. Private Sector in Fragile and Conflict-Affected “How Business Is Done in the Developing States.” Background Paper for the World World: Deals versus Rules.” Journal of Development Report, The World Bank, Economic Perspectives 29 (3): 121–40. Washington, DC. Hoeffler, A. 2012. “Exporting from Fragile Ramachandran, V., and J. Walz. 2012. “Haiti: States: Challenges and Opportunities.” OECD Where Has All the Money Gone?” CGD Policy Working Paper 4, OECD, Paris. Paper 004. Center for Global Development, IFC (International Finance Corporation). 2017. Washington, DC. Private Enterprise in Fragile and Conflict Rodríguez, C., C. Gómez, and J. Ferreiro. 2009. Situations. Washington, DC. “A Proposal to Improve the UNCTAD’s IEG (Independent Evaluation Group). 2015. Inward FDI Potential Index.” Transnational Investment Climate Reforms: An Independent Corporations 18 (3): 85–114. Evaluation of World Bank Group Support to Selaya, P., and E. R. Sunesen. 2012. “Does Foreign Reforms of Business Regulations. Washington, Aid Increase Foreign Direct Investment?” DC: World Bank. World Development 40 (11): 2155–76. International Dialogue for Peace-Building Speakman, J., and A. Rysova. 2015. The Small and State-Building. 2016. International Entrepreneur in Fragile and Conflict-Affected Standards for Responsible Business in Situations. Washington, DC: World Bank. Conflict-Affected and Fragile Environment. UN (United Nations). 2012. World Investment Paris: OECD. Report, 2012. Geneva: UN. FDI IN FRAGILE AND CONFLIC T-AFFEC TED SITUATIONS 159 USAID (United States Agency for International ———. 2011. World Development Report 2011: Development). 2009. “Diaspora Direct Conflict, Security and Development: 2011. Investment (DDI): The Untapped Resource for Washington, DC: World Bank. Development.” Washington, DC. ———. 2014. “Promoting Foreign Investment ———. 2016. Local Private Sector Partnerships: in Fragile and Conflict-Affected Situations.” Assessing the Sate of Practice. Washington, DC: Investment Climate in Practice No. 22. World USAID. Bank, Washington, DC. WEF (World Economic Forum). 2016. The Global ———. 2016. “Information Note: The World Competitiveness Report 2016. Geneva: WEF Bank Group’s Harmonized List of Fragile Publishing. Situations.” http://pubdocs.worldbank.org World Bank Group. 2009. The Costs of Violence. / en/586581437416356109/FCS-List-FY16 Washington, DC: World Bank. -Information-Note.pdf. Glossary Behavioral incentives. Investment incentives intended to encourage certain investor behaviors, such as hiring local staff, investing in innovation, or using local suppliers to establish linkages. Bilateral investment treaty. An agreement between two countries establishing the terms and conditions for private investment by an entity of one country in another country. Competition effects. Competition between foreign firms and domestic firms that can lead to firms improving efficiency and upgrading production processes. Demonstration effects. A type of spillover from FDI to the host economy in which domestic firms increase productivity by replicating foreign technologies or managerial practices either through observation or by hiring workers trained by foreign firms. Developed countries. Developed countries refer to high-income countries as defined in this text. Developing countries. Developing countries include low-, lower-middle-, and upper-middle- income countries as defined in the text. For the chapter on outward foreign direct investment (OFDI), these economies are classified according to the income category for 1995 and remain in the developing category even if they eventually surpass the high-income threshold in later years. They include Argentina, Chile, Croatia, Czech Republic, Estonia, Equatorial Guinea, Greece, Hungary, Latvia, Lithuania, Mauritius, Oman, Poland, the Russian Federation, Saudi Arabia, Slovak Republic, Slovenia, Trinidad and Tobago, Uruguay, and República Bolivariana de Venezuela. Doing Business. This WBG project provides objective measures of business regulations and their enforcement across 190 economies and selected cities at the subnational and regional level. Launched in 2002, the project looks at domestic small and medium-sized companies and measures the regulations applicable to them throughout their life cycle. East Asia and Pacific (EAP). The World Bank Group (WBG) region that includes the econo- mies of American Samoa, Australia, Brunei Darussalam, Cambodia, China, Fiji, French 161 162 GLOSSARY Polynesia, Guam, Hong Kong SAR China, Indonesia, Japan, Kiribati, the Democratic People’s Republic of Korea, the Republic of Korea, Lao People’s Democratic Republic, Macao SAR China, Malaysia, the Marshall Islands, the Federated States of Micronesia, Mongolia, Myanmar, Nauru, New Caledonia, New Zealand, Northern Mariana Islands, Palau, the Philippines, Samoa, Singapore, the Solomon Islands, Taiwan China, Thailand, Timor- Leste, Papua New Guinea, Tonga, Tuvalu, Vanuatu, and Vietnam. For the purposes of this report, the countries surveyed for the region may be a smaller subset of the actual regional grouping. Efficiency-seeking FDI. One of the four motivations for FDI, efficiency-seeking FDI is when investors seek to increase cost efficiency of production by taking advantage of location-specific factors. These investors are also known as “cost-competitive investors.” In this report and the Global Investment Competitiveness (GIC) survey, they are respondents who identified “lower production costs” or “establish a new base for exports” as a motivation to invest. Enterprise Survey. A firm-level survey conducted by the WBG of a representative sample of an economy’s private sector. The survey covers a broad range of business environment topics including access to finance, corruption, infrastructure, crime, competition, and performance measures. Since 2002, the WBG has collected this data via face-to-face interviews with top managers and business owners in more than 155,000 companies in 148 economies. Europe and Central Asia (ECA). WBG region that includes the economies of Albania, Andorra, Armenia, Austria, Azerbaijan, Belarus, Belgium, Bosnia and Herzegovina, Bulgaria, Channel Islands, Croatia, Cyprus, Czech Republic, Denmark, Estonia, Faroe Islands, Finland, France, Georgia, Germany, Gibraltar, Greece, Greenland, Hungary, Iceland, Ireland, Isle of Man, Italy, Kazakhstan, Kosovo, Kyrgyz Republic, Latvia, Liechtenstein, Lithuania, Luxembourg, the former Yugoslav Republic of Macedonia, Moldova, Monaco, Montenegro, the Netherlands, Norway, Poland, Portugal, Romania, Russian Federation, San Marino, Serbia, Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Tajikistan, Turkey, Turkmenistan, Ukraine, the United Kingdom, and Uzbekistan. For the purposes of this report, the countries surveyed for the region may be a smaller subset of the actual regional grouping. Export share by sector. Calculated as non-host country sales divided by total sales based on the U.S. Bureau of Economic Analysis (BEA). FDI inflow. All liabilities and assets transferred between resident direct investment enterprises and their direct investors into the reporting economy for the reporting period, usually for one year. FDI outflow. All liabilities and assets transferred outward between resident direct investors and their direct investment enterprises away from the reporting economy for the reporting period, usually for one year. FDI spillover. The impact of foreign firms’ presence on domestic firms’ economic performance. Positive FDI spillovers indicate that domestic firms acquire foreign technology and frontier knowledge through direct and indirect interactions with MNCs. FDI stock. According to the Organisation for Economic Co-operation and Development (OECD), FDI stock measures total direct investment at a given point in time, usually at the end of a quarter or of a year. It represents the value of the resident investors’ equity in and net loans to enterprises resident in the reporting economy. Foreign affiliates. Generic term to describe various types of entities that a foreign investment might take. Affiliates may be subsidiaries, branches, or any other enterprise resident in a host country that is controlled by a nonresident institutional unit. GLOSSARY 163 Foreign direct investment (FDI). According to the International Monetary Fund (IMF), FDI is a category of international investment made by a resident entity in one economy with the goal of establishing a lasting interest in an enterprise, resident in an economy other than the investor’s. A lasting interest refers to the existence of a long-term relationship between the direct investor and the enterprise, and a significant degree of influence by the direct investor on the manage- ment of the direct investment enterprise. Components of FDI include equity, intra-company debt, and reinvested earnings. Fragile and conflict-affected situations (FCS). Group of economies that have either a har- monized average Country Policy and Institutional Assessment (CPIA) country rating of 3.2 or less; or the presence of a United Nations or regional peacekeeping or peace-building mis- sion during the past three years. The group of countries includes IDA-eligible countries and nonmember or inactive territories or countries without CPIA data. For fiscal year 2017, FCS include the following states and territories: Afghanistan, Burundi, Central African Republic, Chad, Comoros, the Democratic Republic of Congo, Côte d’Ivoire, Djibouti, Eritrea, The Gambia, Guinea-Bissau, Haiti, Iraq, Kiribati, Kosovo, Lebanon, Liberia, Libya, Madagascar, Mali, the Marshall Islands, the Federated States of Micronesia, Myanmar, Papua New Guinea, Sierra Leone, the Solomon Islands, Somalia, South Sudan, Sudan, the Syrian Arab Republic, Togo, Tuvalu, West Bank and Gaza, the Republic of Yemen, and Zimbabwe. Government effectiveness. Part of the WBG’s Worldwide Governance Indicators, government effectiveness is an aggregate indicator that reflects perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies. Global value chains (GVCs). International fragmentation of production where a single finished product results from manufacturing and assembly in multiple countries, with each step in the process adding value to the end product. Gravity model. Economic model used to estimate bilateral effects between two geographic points, based usually on economic sizes and distance between the two locations. Greenfield. Investment in which the investor builds its business operations from the ground up. In this report, greenfield refers to a mode of entry for FDI, where a foreign investor builds its operations in a host economy. Herfindahl–Hirschmann Index (HHI). A measure of market concentration. In this report, the HHI for geographic concentration is defined as the sum of the squares of all countries’ shares in the total number of FDI projects for a given sector. It would hence take the value of 1 in a hypothetical case where all FDI projects in a given sector went to one country. As the scale approaches 0, FDI projects are more dispersed among countries and the sector less geo- graphically concentrated. High-income countries. For fiscal year 2017, high-income economies are defined as those with a GNI per capita of $12,476 or more in 2015. For the chapter on OFDI, in 1995, these countries are defined as those with a gross national income (GNI) per capita of $9,386 or more. High-growth firms. Firms that have a disproportionately large role in job creation in the economy. Home economy. Country of origin of the foreign investment. Horizontal FDI. Investment abroad by a company in the same industry in which the company operates in in the home economy. 164 GLOSSARY Host economy. Country that receives the foreign investment. International investment agreement (IIA). A type of treaty between states that addresses issues on cross-border investments. IIAs exist in three levels: bilateral (such as bilateral investment treaties), regional or preferential (such as regional customs unions and free trade areas or preferential trade agreements), and multilateral (such as applicable rules in World Trade Organization agreements and other international investment conventions). Investment incentives. Measurable economic advantages that governments offer to specific enterprises or groups of enterprises with the goal of steering investments into preferred sectors or locations. These benefits can be fiscal (for example, tax concessions) or nonfiscal (for exam- ple, loans or rebates). Investment protection guarantees. Guarantee or insurance provided for by law, government, multilateral agency, or any party for an investment made. Investment promotion agency (IPA). Government agency or nonprofit organization whose job is to attract investment to the host economy. Knowledge effects. Acquisition of knowledge, through FDI, either directly by the investor or investee firm, or indirectly through spillovers to other firms. Knowledge can take the form of technology, production techniques, or management skills. Knowledge-seeking FDI. Type of FDI that aims to augment firm-specific advantage owned by the investor to improve its competitiveness by acquiring new knowledge. All knowledge- seeking FDI is strategic asset–seeking but not all strategic asset–seeking is knowledge-seeking. Latin America and the Caribbean (LAC). WBG region that includes the economies of Antigua and Barbuda, Argentina, Aruba, The Bahamas, Barbados, Belize, Bolivia, Brazil, British Virgin Islands, Cayman Islands, Chile, Colombia, Costa Rica, Cuba, Curaçao, Dominica, the Dominican Republic, Ecuador, El Salvador, Grenada, Guatemala, Guyana, Haiti, Honduras, Jamaica, Mexico, Nicaragua, Panama, Paraguay, Peru, Puerto Rico, Sint Maarten (Dutch part), St. Kitts and Nevis, St. Lucia, St. Martin (French part), St. Vincent and the Grenadines, Suriname, Trinidad and Tobago, Turks and Caicos Islands, Uruguay, República Bolivariana de Venezuela , and Virgin Islands (U.S.). For the purposes of this report, the countries sur- veyed for the region may be a smaller subset of the actual regional grouping. Linkages. The transmission of foreign knowledge and practices that may improve the produc- tion capabilities of domestic suppliers, as a result of contractual arrangements between local suppliers and multinational corporations. Locational incentives. Investment incentives that are intended to influence the location deci- sion of the investors. Low-income countries. For fiscal year 2017, low-income economies are defined as those with a GNI per capita of $1,025 or less in 2015. For the chapter on OFDI, in 1995, these econo- mies are defined as those with a GNI per capita of $765 or less. Lower-middle-income countries. For fiscal year 2017, lower-middle-income economies are defined as those with a GNI per capita between $1,026 and $4,035 in 2015. For the chapter on OFDI, in 1995, these economies are defined as those with a GNI per capita between $766 and $3,035. Manufacturing. Economic sector that produces goods. GLOSSARY 165 Market-seeking FDI. A motivation for FDI in which the investor seeks to access domestic markets by supplying goods and services to the host economy. Mergers and acquisitions (M&A). Transactions that result in the consolidation of companies or assets. In this report, M&A are FDI by nature, where the purchasing entity is a foreign investor that acquires the assets of a local firm. Middle East and North Africa (MENA). WBG region that includes the economies of Algeria, Bahrain, Djibouti, the Arab Republic of Egypt, Islamic Republic of Iran, Iraq, Israel, Jordan, Kuwait, Lebanon, Libya, Malta, Morocco, Oman, Qatar, Saudi Arabia, the Syrian Arab Republic, Tunisia, the United Arab Emirates, West Bank and Gaza,and the Republic of Yemen. For the purposes of this report, the countries surveyed for the region may be a smaller subset of the actual regional grouping. Multinational corporation (MNC). A corporation that has operations in more than one coun- try and usually has a centralized head office which coordinates global management. Natural resource–seeking FDI. A motivation for FDI in which investors seek to access natural resources—such as oil and gas, mining and minerals, water or solar power—in the host economy. North America. WBG region that includes the economies of Bermuda, Canada, and the United States. Outward FDI (OFDI). FDI from the perspective of the home economy. This is in contrast to FDI, which is from the perspective of the host economy. See entry for FDI. Parent company. Institutional unit that owns enough interest in another firm to manage or operate the firm. Postconflict countries. For this report, postconflict countries include the subgroup of FCS where conflict has occurred since 1990. In addition to the official list, 11 other countries include Algeria, Colombia, Ethiopia, Guatemala, Mozambique, Nicaragua, Peru, Rwanda, Sri Lanka, Uganda, and Ukraine. Preferential margin. The difference between the standard corporate income tax rate and the preferential rate granted as an incentive. Preferential trade agreement. A trading bloc that gives special treatment to participating entities. Primary. Economic sector that uses natural resources including farming, mining, and fishing. Reinvested earnings. Net earnings not paid out as dividends but retained by the firm for rein- vestment in its business operations in the host country. Scale effects. Average cost per unit decreases when production increases. Services. Economic sector that produces nongoods, including financial services and retail services. South Asia. WBG region that includes the economies of Afghanistan, Bangladesh, Bhutan, India, Maldives, Nepal, Pakistan, and Sri Lanka. For the purposes of this report, the countries surveyed for the region may be a smaller subset of the actual regional grouping. Strategic asset–seeking FDI. A motivation for FDI in which investors seek to control firm or country-specific asset including brand, distribution network, or supply chain. 166 GLOSSARY Sub-Saharan Africa (SSA). WBG region that includes the economies of Angola, Benin, Botswana, Burkina Faso, Burundi, Cabo Verde, Cameroon, Central African Republic, Chad, Comoros, the Democratic Republic of Congo, the Republic of Congo, Côte d’Ivoire, Equatorial Guinea, Eritrea, Ethiopia, Gabon, The Gambia, Ghana, Guinea, Guinea-Bissau, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Niger, Nigeria, Rwanda, São Tomé and Príncipe, Senegal, the Seychelles, Sierra Leone, Somalia, South Africa, South Sudan, Sudan, Swaziland, Tanzania, Togo, Uganda, Zambia, and Zimbabwe. For the purposes of this report, the countries surveyed for the region may be a smaller subset of the actual regional grouping. Tax holiday. Temporary complete removal of a tax granted to a specific firm or group of firms by a government. The World Economic Forum’s (WEF) Executive Opinion Survey. Conducted by the WEF, this survey captures information on a broad range of socioeconomic topics from executives across the world. In 2016, more than 13,000 responses in more than 130 countries were collected. Upper-middle-income countries. For fiscal year 2017, upper-middle-income economies are defined as those with a GNI per capita between $4,036 and $12,475 in 2015. For the chapter on OFDI, in 1995, these economies are defined as those with a GNI per capita between $3,036 and $9,385. Vertical FDI. Investment in an industry that produces inputs for the firms’ operations, and is often used to offshore immediate production steps to locations with lower costs. World Bank Group (WBG). Institutions that constitute the WBG include International Bank for Reconstruction and Development (IBRD), International Development Association (IDA), International Centre for Settlement of Investment Disputes (ICSID), International Finance Corporation (IFC), and Multilateral Investment Guarantee Agency (MIGA). 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 world- wide. 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.