2019I 2020 GLOBAL INVESTMENT COMPETITIVENESS REPORT Rebuilding Investor Confidence in Times of Uncertainty Global Investment Competitiveness Report 2019/2020 Global Investment Competitiveness Report 2019/2020 Rebuilding Investor Confidence in Times of Uncertainty © 2020 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 23 22 21 20 This work is a product of the staff of The World Bank with external contributions. The findings, 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. 2020. Global Investment Competitiveness Report 2019/2020: Rebuilding Investor Confidence in Times of Uncertainty. Washington, DC: World Bank Group. doi:10.1596/978-1-4648-1536-2. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third- ­ party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-1536-2 ISBN (electronic): 978-1-4648-1543-0 DOI: 10.1596/978-1-4648-1536-2 Cover design: Bill Pragluski, Critical Stages, LLC. Library of Congress Cataloguing-in-Publication data has been requested. Contents Foreword. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii About the Authors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii Main Messages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Overview. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Christine Zhenwei Qiang and Peter Kusek 1  Outlook and Priorities for Foreign Investors in Developing Countries: Findings from the 2019 Global Investment Competitiveness Survey in 10 Middle-Income Countries . . . . . 24 Peter Kusek, Abhishek Saurav, and Ryan Kuo 2  How Beneficial Are Foreign Acquisitions of Firms in Developing Countries? Evidence from Six Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 56 Alexandros Ragoussis 3 The Distributional Effects of FDI: Evidence from Ethiopia, Vietnam, and Turkey . . . . . . . . 84 Victor Steenbergen and Trang Thu Tran 4 Regulatory Risk and FDI. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128 Sarah Hebous, Priyanka Kher, and Trang Thu Tran 5 Increasing the Development Impact of Investment Promotion Agencies. . . . . . . . . . . . . . . . . 170 Armando Heilbron and Hania Kronfol Glossary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209 v v i    C o n t e n t s Boxes O.1 Key Success Factors in High-Performing IPAs in Developing Countries . . . . . . . . . . . . . 18 1.1 Key Findings of the 2019 GIC Survey Results, by Foreign Investor Type . . . . . . . . . . . . 43 2.1 Conceptual Overlaps between Mergers, Acquisitions, and Brownfield FDI. . . . . . . . . . . 59 2.2 The Evidence on Economic and Social Outcomes of Multinational Enterprises. . . . . . . 67 2.3 Systemic Benefits of Brownfield Foreign Direct Investment: Telecommunication Acquisitions in Africa. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73 3.1 FDI’s Potential to Improve Women’s Economic Opportunities . . . . . . . . . . . . . . . . . . . . 94 3.2 A Sector Typology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 3.3 The Limited Effectiveness of Traditional Active Labor Market Policies. . . . . . . . . . . . . 108 4.1 Data Sources for the Regulatory Risk Measure. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134 4.2 Examining How Regulatory Risk Affects Investor Location Decisions. . . . . . . . . . . . . . 141 4.3 Importance of Political Risk and a Stable Regulatory Framework for Investment Decisions: Confirmation from the 2019 GIC Survey. . . . . . . . . . . . . . . . . . . . . . . . . . . 144 4A.1 Scoring and Aggregating Rules for IIAs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 5.1 The Investment Life Cycle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172 5.2 From Sectors to Segments: Making IPA Targeting More Precise . . . . . . . . . . . . . . . . . . 182 5.3 The Role of Subnational IPAs in Attracting FDI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185 5.4 Costa Rica’s IPA (CINDE) Follows Several Good Practices, Particularly around Strategic Focus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 5.5 Learning from the Success of a 70-Year-Old IPA: IDA Ireland . . . . . . . . . . . . . . . . . . . 190 5.6 How Different Institutional Setups Respond to Multiple Mandates: The Cases of Ireland and Rwanda. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 194 5.7 Invest India: Building a High-Performing IPA from the Ground Up in Record Time. . . . 196 Figures O.1 FDI Inflows to Developing Countries over the Past Decade Have Been Mostly Flat and Have Declined as a Share of GDP. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 O.2 Investment Policies Regarding FDI Are Becoming More Restrictive. . . . . . . . . . . . . . . . . 5 O.3 Even before the COVID-19 Crisis, Investors Were Sensitive to Policy Uncertainty in Trade and Investment and Have Been Adversely Affected in the Past Year. . . . . . . . . . 6 O.4 The COVID-19 Pandemic Had Adversely Affected a Vast Share of MNEs by April 2020. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 O.5 FDI Affects Household Incomes through Several Broad Channels . . . . . . . . . . . . . . . . . . 9 O.6 Brownfield Investment Rose as a Share of Total FDI in Developing Countries, 2010–17. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 O.7 Greenfield and Brownfield FDI Firms Perform Better than Domestic Firms over the First Five Years of Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 O.8 Employment and Wages in Brownfield Firms Grow Faster than in Domestic Firms with Similar Characteristics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 O.9 FDI Has Had Varied Effects on National Income Distributions in Ethiopia, Vietnam, and Turkey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 O.10 Three Pillars of Addressing Regulatory Risk. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 O.11 FDI Inflows Are Higher in Countries with Lower Regulatory Risk. . . . . . . . . . . . . . . . . 15 O.12 The Legal and Regulatory Environment Was among the Top Three Factors for FDI in 2019. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 O.13 Core Elements for Increasing the Development Impact of Investment Promotion Agencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 C o n t e n t s   v ii 1.1 FDI Inflows to Middle-Income Countries Have Been Declining since the 2008–09 Global Financial Crisis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 1.2 Surveyed Countries Have High FDI Restrictiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 1.3 Respondents Are Evenly Split between Manufacturing and Services Firms and Represent Firms across Various Specific Sectors. . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 1.4 The Median MNE Affiliate Is Relatively Small, Well-Established, and Majority Foreign Owned. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 1.5 MNE Affiliates’ Trading Behavior Differs Widely across Host Countries. . . . . . . . . . . . 31 1.6 The Competitive Dynamics of FDI Firms in Host Markets Are Complex. . . . . . . . . . . . 32 1.7 Investors Are Sensitive to Policy Uncertainty in Trade and Investment and Have Been Adversely Affected in the Last Year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 1.8 Larger Firms and Those from Developing Countries Are More Likely to Expand Investment, on Average. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 1.9 Investors in China and Turkey Have Less Ambitious Investment Plans than Those in Other Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 1.10 Legal and Regulatory Environment Is among the Top Three Investment Decision Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 1.11 MNE Affiliates That Experience Legal and Regulatory Obstacles Are More Likely to Reduce or Withdraw Investments in the Future . . . . . . . . . . . . . . . . . . . . . . . . 40 1.12 Foreign-Owned Firms Perceive the Quality of Rules and Their Implementation as Obstacles in Government Conduct. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 2.1 Brownfield Investment Rose as a Share of Total FDI in Developing Countries, 2010–17. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58 2.2 Higher Shares of Brownfield Investments Occur in More Developed Economies . . . . . . 62 2.3 Brownfield FDI Is Likelier in Sectors That Rely on Land, Have Established Distribution Networks, or Are Highly Regulated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 2.4 Greenfield Affiliates Typically Have Larger Workforces than Brownfield Affiliates, Considering Age of the Firm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 2.5 Brownfield Affiliates and Greenfield Ventures Differ from the Average Domestic Firm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 2.6 Brownfield and Greenfield FDI Firms Perform Better than Domestic Firms over the First Five Years of Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70 2.7 Targeted Firms Improve Their Performance after Acquisition by Foreign Investors . . . . 71 2C.1 Evolution of Brownfield MNE Characteristics over Time . . . . . . . . . . . . . . . . . . . . . . . . 79 3.1 FDI Has Had Varied Effects on National Income Distributions in Ethiopia, Vietnam, and Turkey . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 3.2 FDI Affects Household Incomes through Several Broad Channels . . . . . . . . . . . . . . . . . 89 3.3 Conceptual Framework: FDI Has Varied Effects on Labor Market Outcomes. . . . . . . . 90 3.4 FDI Has Different Labor Market Outcomes by Sector. . . . . . . . . . . . . . . . . . . . . . . . . . . 93 3.5 FDI Has Increased Significantly in Ethiopia, Vietnam, and Turkey. . . . . . . . . . . . . . . . . 96 3.6 Ethiopia, Vietnam, and Turkey Have Attracted Different Types of FDI . . . . . . . . . . . . . 98 3.7 Ethiopia: FDI Effects on Poverty, Shared Prosperity, and Inequality. . . . . . . . . . . . . . . 103 3.8 Vietnam: FDI Effects on Poverty, Shared Prosperity, and Inequality. . . . . . . . . . . . . . . 104 3.9 Turkey: FDI Effects on Poverty, Shared Prosperity, and Inequality. . . . . . . . . . . . . . . . 105 4.1 Three Pillars of Regulatory Risk Frame the Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . 135 4.2 Regulatory Risk Levels Vary Significantly across Countries and Change Substantially for Some Countries over Time. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138 4.3 Higher Regulatory Risk Is Associated with Higher Expropriation Risk Insurance Premiums and Investment Risk Ratings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 4.4 Higher Regulatory Risk Is Associated with More Restrictive FDI Regulations. . . . . . . 139 v i i i    C o n t e n t s 4.5 FDI Inflows Are Higher in Countries with Lower Regulatory Risk. . . . . . . . . . . . . . . . 140 4.6 Higher Regulatory Risk Has a Negative Impact on Investors’ Likelihood of Entering or Expanding in a Host Country. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142 4.7 Different Dimensions of the Regulatory Framework (Transparency, Protection, and Recourse) Matter for Investment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 B4.3.1 Legal and Regulatory Environment Is the Third Most-Cited Investment Decision Factor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144 B4.3.2 Expropriation and Breach of Contract Are the Most Likely to Affect Investments Adversely. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145 B4.3.3 Investors Perceive the Quality of Rules and Their Implementation as Obstacles. . . . . . . 145 4.8 Kazakhstan’s Scores on the Three Pillars of the FDI Regulatory Risk Measure . . . . . . . 149 B4A.1.1 Model of a Country with Multiple Treaty Relationships. . . . . . . . . . . . . . . . . . . . . . . . 155 4A.1 The Regulatory Risk Measure Is Highly Correlated across Varying Data Sources and Aggregation Methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 4A.2 Regulatory Risk Varies across Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 4A.3 Countries That Perform Better in One Pillar Often Perform Better in Other Pillars of Regulatory Risk. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 4A.4 Higher Regulatory Risk Is Associated with Higher Expropriation Risk Insurance Premium. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 B5.1.1 The Investment Life Cycle from Investor and Host Country Perspectives. . . . . . . . . . . 172 5.1 The Best IPAs Tend to Be Dedicated Promoters. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174 5.2 Better-Quality IPA Services Attract More FDI Inflows . . . . . . . . . . . . . . . . . . . . . . . . . . 176 5.3 IPA Services across the Investment Life Cycle Are Important to Investors . . . . . . . . . . 177 5.4 Investors from Developing Countries Value IPA Services More than Those from High-Income Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177 5.5 IPAs Are Becoming More Proactive and Focusing More on Sector Targeting. . . . . . . . . 179 5.6 IPA Sector Targeting Maintains a Traditional Focus on the Primary Sector, Indicating a Misalignment with Evolving Global FDI Trends . . . . . . . . . . . . . . . . . . . . 180 5.7 IPA Inquiry Handling Deteriorated in 2012 from an Already Poor Level in 2006 . . . . 180 5.8 Eighty-Four Percent of IPAs Have Five or More “Priority” Sectors. . . . . . . . . . . . . . . . 182 5.9 Most IPAs Cover a Wider Range of Mandates than Just FDI Promotion, Especially in Developing Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 5.10 IPAs in Developing Countries That Have More Mandates Also Have Lower FDI Inflows. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184 5.11 IPAs in Developing Countries Indicate That Financial and Human Resources, as Well as the Economic Environment, Are Their Top Challenges. . . . . . . . . . . . . . . . . 185 5.12 Core Elements for Increasing the Development Impact of Investment Promotion Agencies. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 B5.4.1 Costa Rica’s FDI Inflows and GVC Participation Have Increased Tenfold since the 1990s. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189 5.13 Proposed Division of National and Subnational IPA Roles in Service Delivery to Investors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198 5A.1 Nearly All National IPAs Collect Data on the Amount of Investment Facilitated and Jobs Created as Indicators of IPA Performance. . . . . . . . . . . . . . . . . . . . . . . . . . . . 200 Tables O.1 The Effects of FDI on Labor Markets Vary by Sector and Workers’ Skill Levels. . . . . . . 13 1.1 Most Investors Come from High-Income Countries in Asia or Europe. . . . . . . . . . . . . . 28 C o n t e n t s   i x 1.2 Intraregional Investment Links Are Strong in East Asia and Pacific, and in Europe and Central Asia, While Interregional Investment Is Greater in Other Regions . . . . . . . 30 1.3 Investment Approval Processes Are the Top Most-Cited Obstacle across Surveyed Countries, Followed by Price, Technology, or Product Restrictions. . . . . . . . . . . . . . . . . 41 1A.1 2019 GIC Survey Sample, by Country and Sector. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 1A.2 Share of 2019 GIC Survey Respondents, by Subsector. . . . . . . . . . . . . . . . . . . . . . . . . . . 47 1A.3 Sampling Frame Sizes, by Country. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 1B.1 FDI Outlook Trends in Surveyed Countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 2A.1 Industrial Census Data Used for Cross-Country Analysis. . . . . . . . . . . . . . . . . . . . . . . . 78 2B.1 Variance Decomposition. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 3.1 The Effects of FDI on Labor Markets Vary by Sector and Workers’ Skill Levels. . . . . . . 88 3.2 The Relative Effects of FDI on Labor Markets (Formal Employment and Wages) in Ethiopia, Vietnam, and Turkey, by Broad Sector and Worker Skill Level. . . . . . . . . 101 3C.1 Broad Sectoral Classification and Their Subsectors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 3C.2 FDI Labor Market Effects: All Countries (Second-Stage IV Results). . . . . . . . . . . . . . . . 114 3C.3 FDI Labor Market Effects by Broad Sector and Education Level: Ethiopia (Second-Stage IV Results). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 3C.4 FDI Labor Market Effects by Broad Sector and Education Level: Vietnam (Second-Stage IV Results). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 3C.5 FDI Labor Market Effects by Broad Sector and Education Level: Turkey (Second-Stage IV Results). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 3C.6 FDI Labor Market Effects by Broad Sector and Education Level: All Countries (Population Weighted, Total Average Effects). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 3C.7 Vertical Spillovers Effect of FDI on Labor Market Outcomes: All Countries (Second-Stage IV Results). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120 4.1 Several Popular Cross-Country Risk Ratings Do Not Clearly Link Risk Levels to Specific Government Actions and Rely on Perception-Based Inputs . . . . . . . . . . . . . 132 4A.1 Underlying Variables of the Regulatory Risk Measure, by Dimension and Subindicators. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154 4A.2 Principal Component Analysis (PCA) Weights. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 4A.3 A Gravity Model of Bilateral FDI Inflows, 2014–16 . . . . . . . . . . . . . . . . . . . . . . . . . . . 160 4A.4 Regulatory Risk and Activities of Affiliates of US MNEs . . . . . . . . . . . . . . . . . . . . . . . 160 4A.5 A Discrete Choice Model of Global Investment Location, 2014–16 . . . . . . . . . . . . . . . . 161 5.1 IPAs Self-Report Providing Services Mostly at the Attraction Stage, with Service Delivery Rapidly Declining in Later Stages of the Investment Life Cycle. . . . . . . 181 5.2 Dos and Don’ts of Mandates for IPAs in Developing Countries. . . . . . . . . . . . . . . . . . . 192 5B.1 Regional Coverage of the 2017 World Bank Group Global IPA Survey. . . . . . . . . . . . . 201 5C.1 Selected IPA Characteristics with No Significant Changes between the 2005 and 2017/18 IPA Surveys. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202 Foreword The Global Investment Competitiveness to raise investor confidence in times of Report 2019/2020 comes at a critical time— uncertainty. a period of economic uncertainty marked by Although responsible fiscal and monetary the coronavirus (COVID-19) outbreak, a policies underpin macroeconomic stability, challenging global policy environment for governments can further reduce risk and build investment and trade, rising protectionism, confidence by implementing transparent and and shifting trade and investment prefer- predictable regulatory regimes. Reaffirming ences. These forces are changing the patterns commitments to market access and rules- of international production and corporate based international systems would decrease decision-making, creating both opportunities uncertainty related to protectionism and eco- and risks for foreign investment. nomic nationalism. Better regulation and The report was developed in the months implementation would reduce obstacles to that preceded the outbreak of COVID-19 investment. Finally, governments do not need and focuses on trade and investment policy to forgo growth to improve inclusion. uncertainty due to policy shifts, globally and Measures to improve labor market skills and nationally. It finds that rising policy uncer- local supplier linkages to multinational com- tainty is darkening the outlook for foreign panies can promote more equitable, broad- direct investment. Unfortunately, these nega- based economic growth. tive effects will only be exacerbated by the These recommendations reflect the economic challenges and policy uncertainty results of the 2019 Global Investment brought by the spread of the virus. Competitiveness Survey of more than 2,400 Considering the difficult global environment, business executives representing foreign direct this report focuses on what the governments investment (FDI) in 10 large developing coun- of developing countries can do to enhance tries: Brazil, China, India, Indonesia, investor confidence, maximize investments’ Malaysia, Mexico, Nigeria, Thailand, Turkey, contributions to inclusive growth, and foster and Vietnam. Two-thirds of investors said the investment competitiveness of their econ- that policy uncertainty due to protectionism omies. It delivers novel analytical insights, and economic nationalism in trade and invest- fresh empirical evidence, and actionable ment is “important” or “critically important” ­ r ecommendations for governments eager in their investment decisions. Over half of xi xii Global Investment Competitiveness Report 2019/2020 those who said policy uncertainty is a “criti- technology transfers, and stronger linkages to cally important” investment factor reported a global value chains. Still, governments need decrease in employment, firm productivity, or to remain vigilant about the potential for investments over the past year. A majority of worsening income inequality. A final chapter investors surveyed overall did not expect to articulates priorities for investment promo- expand investments over the next three years. tion agencies and other stakeholders seeking Interviews were conducted from June to to enhance their countries’ investment November 2019. competitiveness. Foreign investors said that supportive polit- The Global Investment Competitiveness ical environments, stable macroeconomic con- Report 2019/2020 draws insights from a vari- ditions, and conducive regulatory regimes are ety of sources, including the survey of busi- their top three investment decision ­ factors—​ ness executives, extensive analysis of data and even more important than low taxes, low evidence, and a thorough review of interna- labor and input costs, or access to natural tional best practices in investment policy resources. Moreover, the report’s new global design and implementation. Together, these database of regulatory risk shows that predict- sources underscore the significant threat to ability and transparency increase investor con- global economic growth from high levels of fidence and FDI flows. international and domestic policy uncertainty The report assesses the impact of FDI on in trade and investment. They also indicate poverty, inequality, employment, and firm that with timely action, policy makers can performance using empirical evidence from bolster foreign direct investment and reap its various countries. It shows that FDI in devel- benefits for short-term growth and long-term oping countries yields benefits to firms and economic transformation—both critical for citizens—including more and better-paid jobs, poverty reduction. Ceyla Pazarbasioglu Hans Peter Lankes Vice President Vice President Equitable Growth, Finance and Institutions Economics and Private Sector Development World Bank Group International Finance Corporation Acknowledgments This report was developed by the World Gaurav Nayyar, Richard Newfarmer, Habib Bank Group’s Global Investment Climate Rab, Rita Ramalho, Carlos Rodriguez Castelan, team. It was managed by Peter Kusek, under Jorge Luis Rodriguez Meza, Miguel Eduardo the guidance of Christine Zhenwei Qiang Sanchez Martin, Karl Sauvant, Pierre Sauve, (Practice Manager of the Investment Climate Bostjan Skalar, and Joshua Wimpey. Unit), Caroline Freund (Global Director for The report’s authors are grateful for the Trade, Investment, and Competitiveness), excellent research assistance and overall sup- and William Maloney (Chief Economist port of Sarur Chaudhary, Philippe Armand of the Equitable Growth, Finance, and De Bonneval, Maximilian Philip Eltgen, Institutions Vice Presidency). Jingyu Gao, Nisan Gorgulu, Carlos Griffin, The team would like to thank the follow- Nadia Hasham, Sarah Hebous, Benjamin ing donor partners of the World Bank Group Kett, Yan Liu, Alex Sanchiz Vicente, and for making this report possible through their Ramprakash Sethuramasubbu. generous financial contributions: the The team would also like to recognize Prosperity Fund of the United Kingdom’s numerous World Bank Group and other col- Foreign and Commonwealth Office and the leagues for their helpful guidance and assis- Government of Austria’s Federal Ministry of tance, without whom this report would not Finance. have been complete. These include Gokhan The team is grateful to the many internal and Akinci, Ivor Beazley, Tania Begazo Gomez, external technical reviewers and contributors David Bridgman, Pascaline della Faille, Wim who provided insight and guidance at various Douw, Thomas Farole, Tania Ghossein, stages of the report preparation, including Daniela Gomez-Altamirano, Justin-Damien Laura Alfaro, Robert Beschel, Richard Bolwijn, Guénette, Georges Vivien Houngbonon, Lee Paul Brenton, Jansen Calamita, Deepa Ann Jackson, Sergiy Kasyanenko, Anju Chakrapani, Ritam Chaurey, Jieun Choi, Lamichhane, Jan Loeprick, Syed Akhtar Andrea Ciani, Riccardo Crescenzi, Stephan Mahmood, Graciela Miralles Murciego, Ngan Dreyhaupt, Marcello Estevao, Neil Gregory, Thuy Nguyen, Ivan Nimac, Celia Ortega Mona Haddad, Mary Hallward-Driemeier, Sotes, Rodrigo Ortiz, Cecilia Sager, Damien Nathen Jensen, Mark Kantor, Ayhan Kose, Aart Shiels, Matthew Stephenson, Temel Taskin, Kraay, Anuj Mathew, Sebastien Miroudot, Robert Whyte, and Sandy Ye. xiii x i v    Global Investment Competitiveness Report 2019/2020 Nancy Morrison and Mary Anderson were contributions of the following Kantar the editors of the report. Mary Fisk and Jewel ­ c olleagues: Jamie Burnett, Stephan Paetz, McFadden provided editorial production and and Marco Pelucchi. logistics support. Aarushi Sinha provided The team acknowledges the many formal additional editorial and production assis- and informal contributions of individuals, tance. The communications efforts were led groups, and organizations that provided by Elizabeth Price and Melissa Knutson. meaningful comments and inputs to enable The 2019 Global Investment Competi­ the successful publication of this report. tiveness Survey analyzed in chapter 1 was Keeping in mind these contributions, the team undertaken on behalf of the World Bank apologizes if it has inadvertently omitted for- Group by Kantar Public, a global survey mally acknowledging any such valuable firm. We are particularly grateful for the contributions. About the Authors Sarah Hebous is a consultant at the World foreign direct investment–led development, Bank Group. She has extensive experience in investment attraction, institutional strength- data analysis and empirical economics ening, and trade and investment policy including conducting household surveys and reform. She holds a master’s degree in public analyzing job creation and sources of agrar- policy from Harvard University. ian distress. She holds a master’s degree in Ry a n K u o  i s a c o n s u l t a n t i n t h e mathematics from the University of Münster, Investment Climate Unit of the World Bank Germany. Group, where he conducts research and Armando Heilbron leads the World Bank advises governments on foreign direct Group’s investment promotion workstream. investment and competition policy. His His global experience spans international experience also includes work in the public business, investment promotion, private sec- and private sectors across East Africa and tor development, and special economic zones. Southeast Asia. He holds a master’s degree He is a Fulbright scholar with a master’s in public affairs from Princeton University. degree in international business from the Peter Kusek is a senior economist at the University of California, Berkeley. World Bank Group. He leads the Applied Priyanka Kher is a private sector specialist Research Program on Investment Climate in the Investment Climate Unit of the and is the managing editor of the Global World Bank Group. She leads research and Investment Competitiveness Report series. policy advisory on investment climate His research and advisory work focus on reforms, investment retention and expansion, international investment, competitiveness, and international trade and investment law. and private sector development. He holds a She holds a master’s degree in law from master’s degree in economic policy and inter- Harvard Law School. national development from Princeton University. Hania Kronfol is a private sector specialist in the Investment Climate Unit of the World Christine Zhenwei Qiang is the practice Bank Group, where she leads analytical and manager of the Investment Climate Unit of advisory projects on investment policy and the World Bank Group. Her team advises promotion.  Her areas of expertise include governments in more than 90 countries on xv x v i    Global Investment Competitiveness Report 2019/2020 catalyzing investment and business competi- advice to governments. He holds a doctorate tiveness through legal, policy, regulatory, and in public policy and administration from institutional reforms in the areas of business ­ George Washington University. regulation, competition policy, and Victor Steenbergen is an economist in the ­ investment policy and promotion. She holds a Investment Climate Unit of the World Bank doctorate in economics from Johns Hopkins Group. He has quantitative research expertise University. in investment, tax, trade, and public finance Alexandros Ragoussis is an economist policy. He has also provided longer-term tech- at the International Finance Corporation nical assistance to the governments of of the World Bank Group and an associate Malawi, Mexico, Nigeria, and Rwanda. He fellow at the German Development holds a master’s degree in public administra- Institute. He holds a doctorate in econom- tion and development economics from the ics from the University of Sydney and has London School of Economics. been working on firm dynamics, invest- Trang Thu Tran is a senior economist in ment, and their links with international the Firms, Entrepreneurship, and Innovation development. Unit of the World Bank Group. Her analytical Abhishek Saurav is an economist in the and advisory work have focused on entrepre- Investment Climate Unit of the World Bank neurship and innovation policy, international Group. He leads policy research on private investment, and business regulatory issues. investment, job creation, women’s eco- She is an applied microeconomist with a doc- nomic empowerment, and productivity. torate from the University of Maryland, He also ­p rovides evidence-based policy College Park. Abbreviations ABA Austrian Business Agency B2G business-to-government BIT bilateral investment treaty BPP Benchmarking Public Procurement (World Bank) CATI computer-assisted telephone interview CEO chief executive officer CINDE Costa Rican Investment Promotion Agency CISF comprehensive investor services framework CRM customer relationship management DDI domestic direct investment EIU Economist Intelligence Unit EPU Global Economic Policy Uncertainty Index EU European Union FDI foreign direct investment FET fair and equitable treatment FIAS Foreign Investment Advisory Service (World Bank) FTA free trade agreement GDP gross domestic product GIC Global Investment Competitiveness (survey) GIPB Global Investment Promotion Best Practice GIRG Global Indicators of Regulatory Governance (World Bank) GST goods and services tax GVC global value chain HICs high-income countries ICRG International Country Risk Guide xvii x v i i i    Global Investment Competitiveness Report 2019/2020 ICT information and communication technology IFC International Finance Corporation IIA international investment agreement IPA investment promotion agency IPI investment promotion intermediaries ISDS investor-state dispute settlement ISIC International Standard Industrial Classification of All Economic Activities ISIC2 ISIC two-digit classification IT information technology IV instrumental variables JV joint venture KPIs key performance indicators LICs low-income countries M&A mergers and acquisitions MFN most-favored nation MICs middle-income countries MNE multinational enterprise MREL monitoring, reporting, evaluation, and learning OECD Organisation for Economic Co-operation and Development OFDI outward foreign direct investment OSS one-stop-shop PCA principal component analysis PPP public-private partnership R&D research and development SBU Survey of Business Uncertainty SDG Sustainable Development Goal SDI Scottish Development International SEZ special economic zone SMEs small and medium enterprises UNCTAD United Nations Conference on Trade and Development WAIPA World Association of Investment Promotion Agencies WTO World Trade Organization Main Messages Overall 1. Even before the COVID-19 outbreak, global foreign direct investment (FDI) was in decline due to trade policy uncertainty, rising protectionism, falling rates of return on FDI, and changing forms of international production. 2. The COVID-19 crisis is presenting a new, unprecedented source of investor risk that is depressing business confidence to historic lows, resulting in a projected fall in global FDI by more than 40 percent in 2020. 3. More than two-thirds of multinational investors in developing countries are reporting disruptions in supply chains, declines in revenues, and falls in production as a result of COVID-19—and the impacts are projected to worsen in the coming months—based on a new World Bank survey on the impact of the pandemic. 4. FDI can alleviate the impact of the COVID-19 crisis and boost countries’ economic resilience by providing a critical source of external capital for financing public debt and continuing to create more and better-paid jobs, lift people out of poverty, and boost productivity. 5. Foreign acquisitions of local firms in developing countries have doubled as a share of FDI over the past decade, and they have made the acquired companies more export oriented, productive, and diversified in their product offering. 6. At the same time, the possible adverse effects of FDI on income inequality and on lower- skilled workers emphasize the critical mitigating role of labor market and education policies. 7. An extensive survey of more than 2,400 global business executives in 10 large middle- income countries conducted between June and November 2019 shows that government policies can influence FDI location decisions. 8. Government actions—such as reducing investor risk and increasing policy predictability—­ can rebuild investor confidence, based on the report’s new global database of regulatory risk. 9. Investment promotion agencies can boost their countries’ investment competitiveness by better aligning their FDI attraction and retention efforts with market signals and changing investor preferences. 10. Governments can leverage FDI for robust economic recovery from COVID-19 by avoiding protectionist policies, seizing new opportunities from changing FDI and supply chain trends, and fostering global cooperation. 1 Overview Christine Zhenwei Qiang and Peter Kusek Global FDI Flows Face an in decades relative to gross domestic product (GDP).2 Compared with the mid-2000s, when Unprecedented Decline FDI reached nearly 4 percent of GDP in devel- The COVID-19 pandemic is severely impact- oping countries, that share fell to under 2 per- ing multinational enterprises (MNEs) glob- cent in 2017 and 2018 (­ figure O.1, panel b). ally. The economic shock of the crisis to the This worrisome global trend in recent private sector is being transmitted through years has reflected a mix of (a) economic multiple channels, including falling demand, ­factors, including declining rates of return on reduced and disrupted input supply, tighten- FDI; (b) business factors, including adoption ing of credit conditions, a liquidity crunch, of digital technologies and increasingly asset- and rising uncertainty. The pre-COVID-19 light forms of international production; and global environment for foreign direct invest- (c) policy factors, including the erosion of ment (FDI) was already characterized by investor confidence due to policy uncertainty rapidly eroding investor confidence because and changes in US tax policy that drove repa- of trade and investment policy uncertainty, triation of capital back to the United States.3 lagging global growth, falling commodity More specifically, worsening business funda- prices, and rising protectionism. The mentals have driven much of the decline in COVID-19 crisis presents a new, unprece- FDI since 2015, when FDI flows reached their dented source of investor risk that is depress- postcrisis peak. The global average rate of ing investor confidence to new lows. return on FDI decreased from 8.0 percent in Even before the COVID-19 pandemic 2010 to 6.8 percent in 2018 (UNCTAD upended the global economy, global FDI was 2019). While the rates of return have dropped sliding to levels even below those last seen in in both developing and developed countries, the aftermath of the global financial crisis a the declines have been especially large in decade ago (figure O.1, panel a).1 The decline developing countries. was more concentrated in high-income coun- Furthermore, changing business models tries, where inflows of FDI fell by nearly resulting from technological advances have 60 percent in recent years. Although FDI to driven declines in FDI levels and returns. In developing countries did not decline as particular, increases in labor costs and the rise steeply, it nonetheless fell to its lowest levels of advanced manufacturing technologies have 3 4    Global Investment Competitiveness Report 2019/2020 FIGURE O.1  FDI Inflows to Developing Countries over the Past Decade Have Been Mostly Flat and Have Declined as a Share of GDP a. Net in ows, 2000–18 b. Net in ows as a share of GDP, 2000–18 3,500 7 Net FDI inflows as share 3,000 6 Net FDI inflows (US$, billions) 2,500 5 of GDP (%) 2,000 4 1,500 3 1,000 2 500 1 0 0 20 0 20 1 20 2 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 20 1 20 2 20 3 20 4 20 5 20 6 20 7 18 20 0 20 1 20 2 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 20 1 20 2 20 3 20 4 20 5 20 6 20 7 18 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 20 20 Developing countries High income World Source: World Development Indicators database. Note: All country income categories use World Bank-defined classifications; “developing” countries refers collectively to all low and middle-income countries. FDI = foreign direct investment; GDP = gross domestic product. eroded or decreased the significance of many than half the citizens in some of the world’s developing countries’ labor cost advantages. largest 27 countries believe that trade and At the same time, the increasing importance globalization help create jobs, and less of the digital economy and services is shifting than one-third find that they are good for businesses toward more asset-light models of wages, recent data from the Pew Research investment (UNCTAD 2019). In addition, Center indicate (Gramlich 2019).4 The anti-­ commodity price slumps have adversely globalization sentiment is also heightened by affected returns on FDI in more commodity- the ongoing shifts in economic and geopoliti- dependent markets (such as many economies cal power as well as concerns about national in Latin America and the Caribbean, the security. Such anxiety and discontent are Middle East and North Africa, and Sub- ­ fueling a rise in economic nationalism and Saharan Africa). protectionism. Recent events such as withdrawals from global trade agreements, tariff escalations, Uncertainty Has Been Rising and FDI and other trade tensions have contributed to a Rules Tightening new rise in trade and investment policy uncer- Even before the COVID-19 crisis, the num- tainty (Baker, Bloom, and Davis 2019). Free ber and magnitude of various global eco- trade, unhindered investment, and open mar- nomic, geopolitical, technological, and social kets are under threat. Although these fears are shifts have increased uncertainty for citizens, particularly pronounced in the industrialized businesses, and policy makers. These changes world, a growing number of developing coun- are reflected in the high values registered in try governments are also building their policy 2019 by various indicators such as the World agendas along similar themes. Uncertainty Index, the Global Economic The growing protectionist views have Policy Uncertainty Index, and the Trade gradually translated into more restrictive Policy Uncertainty Index (Baker, Bloom, rules on the entry of FDI. The United States and Davis 2019; Caldara et al. 2019). In and the European Union have enacted strict 2020, these indexes have reached unprece- screenings of foreign acquisitions in response dented levels. to perceived risks to national or economic Citizens are increasingly attributing grow- security. Cases of investment withdrawals— ing economic disparity and losses in local eco- investments that are either rejected or with- nomic opportunities to globalization. Less drawn over security concerns—tripled in O v e r v i e w   5 2018 alone, often receiving high publicity around the world show that 55 countries (UNCTAD 2019). undertook at least 112 policy measures Governments have also become increas- related to FDI in 2018 (UNCTAD 2019). Of ingly anxious about the potentially noncom- these, more than one-third restricted or regu- mercial objectives of foreign investment by lated FDI more tightly, whereas the share of state-owned enterprises or sovereign wealth measures that liberalized and promoted FDI funds. Of particular concern has been foreign fell to less than two-thirds (figure O.2). In ownership of core technologies, manufactur- contrast, it was only the previous year (2017) ing of health care products, sensitive business when around 80 percent of the measures pro- assets, and critical infrastructure. Various moted FDI. High-income countries have been governments blocked mergers and acquisi- the primary drivers of the trend toward more tions (M&A) deals worth more than US$150 restrictive rules on FDI. In 2018, more than billion in 2018 (more than 10 percent of total 70 percent of new FDI policy measures in global FDI) on the basis of national security developed countries were aimed at restricting concerns (UNCTAD 2019). Member coun- or regulating FDI (UNCTAD 2019). tries of the Organisation for Economic Although most developing countries have Co-operation and Development (OECD) on so far largely resisted increasing the restric- both sides of the Atlantic are tightening—or tiveness of their FDI regimes, there is a proposing to tighten—their rules governing ­ growing concern that the actions of the gov- the entry of FDI. ernments of developed countries will either In fact, a global cross-country analysis of set a precedent for the developing countries policy trends shows that the share of restric- to follow, or that developing countries will tive and regulatory measures against FDI do so as a retaliatory measure. For example, is the highest it has been in more than China and South Africa have recently intro- 20 years—and the trend may be worsening. duced new regulatory frameworks for FDI The United Nations Conference on Trade and screening for national security concerns Development’s data on FDI policy trends (UNCTAD 2019). FIGURE O.2  Investment Policies Regarding FDI Are Becoming More Restrictive 100 75 Share of FDI policies (%) 66 50 34 25 0 2004 2006 2008 2010 2012 2014 2016 2018 Restriction/regulation Liberalization/promotion Source: UNCTAD 2019. Note: Sample in 2018 comprised 55 countries that undertook at least 112 FDI-related policy measures. FDI = foreign direct investment. 6    Global Investment Competitiveness Report 2019/2020 Policy and COVID-19 Uncertainties that policy uncertainty due to protectionism Adversely Affect Jobs, Investment, and and economic nationalism in trade and invest- Productivity ment was “important” or “critically impor- Between June and November 2019, a Global tant” in their investment decisions in the past Investment Competitiveness (GIC) Survey of year (figure O.3, panel a). Furthermore, more than 2,400 global business executives in among those investors who considered such 10 large middle-income countries was con- policy uncertainty in trade and investment to ducted for this report (see chapter 1). Without be a “critically important” investment factor, taking into account the additional effects more than half have already experienced a of the COVID-19 pandemic, two-thirds of decrease in employment, firm productivity, or investors—particularly firms that import a ­ investments as a result (figure O.3, panel b). greater share of their inputs and larger firms These negative effects have been further (employing more than 250 people)—reported exacerbated by the economic challenges and FIGURE O.3  Even before the COVID-19 Crisis, Investors Were Sensitive to Policy Uncertainty in Trade and Investment and Have Been Adversely Affected in the Past Year a. Question: In the past financial year, how important was rising policy uncertainty due to protectionism and economic nationalism in trade and investment for your company’s investment decisions in this country? All MNEs 29 35 Critically important Important 0 20 40 60 80 100 Share of respondents (%) The 29 percent of MNE affiliates that considered policy uncertainty in trade and investment to be “critically important” were asked the following question: b. Question: In the last financial year, what impact has rising policy uncertainty due to protectionism and economic nationalism in trade and investment had on your company’s operations in this country? Jobs 31 Among these respondents, 51 percent experienced a Productivity 34 decline in at least one category–jobs, productivity, or investment—in Investment 35 the past year. 0 5 10 15 20 25 30 35 40 Share of firms reporting a decrease (%) Source: Computation based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. MNEs = multinational enterprises. The “past financial year” was a 12-month period between January 1, 2018, and September 30, 2019, depending on the country. O v e r v i e w   7 policy uncertainty brought by the spread of consumers, resulting in reduced consumer COVID-19. The closing of factories, disrup- spending and corporate orders. tions in transport, and unavailability of pro- On the supply side, three in four MNEs duction inputs are directly affecting how report declines in supply chain reliability, on companies operate across the globe. The average by 30 percent. Along with the liquid- shocks, having already spread from directly ity crunch (experienced by more than 60 per- hit sectors to others, are also spreading across cent of respondents) and a decline in worker regions through supply linkages. At the epi- productivity (reported by three-fourths of center of this turmoil are multinational cor- businesses), the aggregate effects of these porations that have shaped the geography of shocks include reductions of roughly one- global value chains (GVCs) over the past third in output and investment, reported by three decades. most businesses. The shock waves are also To assess the impact of the pandemic on reaching companies’ employees: two in five MNE affiliates in developing countries, the businesses report declines in jobs, on average World Bank conducted a foreign investor by 16 percent. “pulse” survey in March–April 2020.5 The Even more worrisome than these shocks results show that the COVID-19 pandemic over the first quarter of 2020 are companies’ has already adversely affected more than dire predictions that the impacts will likely three-fourths of investors through both intensify over the second quarter, with perfor- demand- and supply-side channels. Nearly mance deteriorating along every measured four in five MNEs report reductions in reve- dimension (­ figure O.4, panel b). More than nues and profits over the past three months, 85 percent of surveyed businesses expect that on average by 40 percent (figure O.4, panel a). their revenues and profits will decline in April Demand has fallen sharply because of high through June 2020, on average by more than uncertainty and precautionary behavior of 40 ­percent. Four in five businesses also expect FIGURE O.4  The COVID-19 Pandemic Had Adversely Affected a Vast Share of MNEs by April 2020 a. Recent impacts b. Expected impacts Question a: Over the last three months (Jan-Mar 2020), Question b: Over the next three months (Apr-Jun 2020), what has been the impact of COVID-19 on your company? what is the likely impact of COVID-19 on your company? Net income 40 80 Net income 46 86 Revenues 38 80 Investment 42 69 Worker productivity 34 74 Availability of finance 41 76 Output 32 67 Revenues 40 85 Reliability of supply chain 30 73 Reliability of supply chain 38 82 Investment 30 56 Output 35 80 Availability of finance 28 62 Worker productivity 33 78 Employment 16 39 Employment 25 62 0 50 100 0 50 100 Share of firms (%) Share of firms (%) Share of MNEs reporting decrease Average decrease across MNEs Source: World Bank, forthcoming. Note: Computation based on the World Bank’s Investor Confidence Global Pulse Survey, conducted March–April 2020. Sample represents 105 ­multinational enterprise (MNE) affiliates operating in 26 developing countries. The reference period of “last three months” ranges approximately from January to March 2020. 8    Global Investment Competitiveness Report 2019/2020 an average 35 percent reduction in output in The individual chapters of this report ana- the second quarter. The employment impacts lyze various facets of countries’ foundations are particularly likely to worsen: three in five for investment competitiveness. The GIC businesses expect to have to reduce employ- Survey analyzes the drivers of FDI and identi- ment in the second quarter, on average by fies priorities for countries to increase their 25 percent. In addition to the likely downsiz- FDI attractiveness. Several chapters provide ing of the workforce, the most precipitous new evidence on FDI’s contributions to job declines are anticipated in the availability of creation, poverty alleviation, and firm pro- finance (by 41 percent) and in investments ductivity. The report also explores how to (by 42 percent). boost investor confidence through specific The gloomy outlook reported by the sur- policy and regulatory actions that reduce reg- vey respondents is consistent with emerging ulatory risks. The report concludes with an evidence on declining investment activity. assessment of what governments—and espe- UNCTAD (2020) estimates that global FDI cially their investment promotion agencies could decline by up to 40 percent in 2020–21. (IPAs)—can do to help attract high-quality The world’s largest 5,000 MNEs, which FDI and transform their economies. If they account for a significant share of global FDI, succeed, FDI can continue to play a critical have revised their earnings estimates down- role in a robust economic recovery from the ward by an average of 30 percent. Because a COVID-19 pandemic. major share of FDI materializes through rein- vested earnings, FDI activity among existing investors is set to decline. Furthermore, in the Pillar 1: FDI Boosts Economic first quarter of 2020, M&A activity is Resilience—Easing the Impact of expected to drop by up to 70 percent. In Economic Crises by Creating Jobs, February 2020, new cross-border acquisitions Alleviating Poverty, and Boosting fell below US$10 billion, compared with the Productivity normal monthly average of US$40 billion— FDI has always been a key building block for US$50 billion before the crisis.6 the economic growth of developing coun- tries, often providing the largest source of external finance—surpassing remittances, FDI Can Help Countries Alleviate official development assistance, and portfo- the Impact of the Crisis, But lio investment flows. In the post-COVID recovery phase, FDI’s role is likely to further Governments Must Rebuild increase. Countries’ crisis-response policies, Investor Confidence such as financial and fiscal stimulus mea- sures, are generating debt. Domestic revenue With the expected massive global decline in sources will be insufficient to service that FDI, competition among developing countries debt. FDI is therefore likely to remain an to attract foreign investment has only intensi- essential source of capital. fied. What can developing countries do to Beyond capital, foreign investment also counter prevailing global headwinds and helps create jobs and reduce poverty. FDI can uncertainty and to rebuild investor confidence? affect welfare through three main channels How will the factors that affect countries’ (figure O.5): investment competitiveness change as a result of COVID-19? The report’s findings p ­ ertaining • Employment income: As FDI brings capi- to these questions are organized around two tal and new technologies to a sector, it core pillars focused on (1) FDI contributions often raises overall labor demand and to development and economic resilience, and productivity in the sector. This can raise (2) policy actions to rebuild investor confi- total employment and average wages, dence and boost investment. leading to higher household incomes. O v e r v i e w   9 FIGURE O.5  FDI Affects Household Incomes through Several Broad Channels Foreign direct investment Income to employees Prices for consumers Income to producers Change in real household income Poverty reduction Shared prosperity Equitable income distribution Source: World Bank and OECD 2017. Note: FDI = foreign direct investment. • Consumer prices: The entry of new foreign the wages they offer.7 To help fill this gap, this firms increases competition in markets. report focuses on acquired firms in develop- This may lower the prices of goods and ing countries—what they look like, how they services, thus raising household purchas- evolve, and whether conventional narratives ing power and consumption possibilities. do justice to their contributions to develop- • Producer income: As foreign firms com- ment goals. This is particularly pertinent as pete with, buy from, or sell to domestic brownfield investment has doubled as a share firms, they may influence the productiv- of FDI in developing countries over the past ity and profitability of these enterprises, 10 years (figure O.6, as further discussed in increasing or cutting into incomes of chapter 2). domestic producers. This report analyzes a unique set of indus- trial censuses from six developing countries— These FDI effects are seemingly more obvi- China, Côte d’Ivoire, Indonesia, Moldova, ous when it comes to greenfield FDI. Serbia, and Vietnam—to show that brownfield Greenfield investment adds new elements to FDI firms perform better than local firms on the economy: new facilities, new jobs, new some of the key dimensions that matter for production capacity. In contrast, brownfield development, such as export orientation, prod- FDI—acquisitions of domestic firms by f ­ oreign uct diversification, asset accumulation, labor investors—transforms existing production. productivity, and employment (figure O.7). Any positive effect of brownfield investment Results show that firms acquired by MNEs would therefore tend to materialize over lon- not only perform better than the average ger time frames and with varying intensity. domestic firm at the time of the acquisition, Most of the previous evidence on brown- but also improve their performance after field FDI has come from high-income coun- acquisition faster than local firms along some tries and has focused on macroeconomic of the key dimensions that matter for devel- growth, overlooking development outcomes opment. For example, over the first five years at the level of firms, the jobs they create, or of a firm’s operation, a brownfield affiliate is 1 0    Global Investment Competitiveness Report 2019/2020 FIGURE O.6  Brownfield Investment Rose as a Share of Total FDI in Developing Countries, 2010–17 60 50 Share of brownfield FDI in total FDI (%) 40 30 20 10 0 10 11 12 13 17 14 15 16 20 20 20 20 20 20 20 20 Low income Lower middle income Upper middle income High income Source: World Bank, based on the 2019 United Nations Conference on Trade and Development (UNCTAD) Mergers and Acquisitions (M&A) database, http://www.unctad.org/fdistatistics. Note: The trend is illustrated using a two-year moving average. All country income categories use 2017 World Bank-defined classifications; “developing countries” refers collectively to low- and middle-income countries. FDI = foreign direct investment. 70–100 percent more likely to export than a in three developing countries: Ethiopia, domestic firm (figure O.7, panel e). Wages in Vietnam, and Turkey (see chapter 3). Workers foreign takeovers at the end of the first five in sectors and regions with a higher presence years of operations are 40–50 percent higher of foreign firms are generally more likely to than in domestic firms (figure O.7, panel a). be formally employed and receive higher Furthermore, contrary to conventional wages. belief about the potential job-destroying In Vietnam, FDI allowed more than effects of foreign M&A, employment in 350,000 individuals to enter formal manufac- newly acquired firms tends to grow faster in turing employment between 2007 and 2016. most countries than employment in domestic In Turkey, FDI brought in at least 40,000 firms with similar characteristics. Specifically, additional formal manufacturing jobs two years after acquisition, the average between 2009 and 2016. FDI also raised employment in brownfield affiliates expands average manufacturing wages, which by approximately 4 percent, compared with increased by 32 percent in Ethiopia, 12 per- 1.5 percent in domestic firms with similar cent in Vietnam, and 8 percent in Turkey. characteristics (figure O.8, panel a). The Consequently, these wage increases brought firms’ asset value after the acquisition follows about by FDI helped reduce poverty in all a similar path. The experience of the six coun- three countries. Conservative estimates sug- tries analyzed in this study suggests that for- gest that FDI contributed to lifting at least eign acquisitions can be a helpful complement 35,000 individuals out of poverty in Ethiopia to greenfield FDI in all developing countries (2009–14), 24,000 in Vietnam (2007–16), seeking to leverage foreign investment for and 15,000 in Turkey (2009–16). advancing their development goals. Growth in formal jobs and wages due to Looking beyond formal enterprises, the FDI has also translated into increased report further finds that FDI has a significant shared prosperity: the FDI-induced wage effect on household employment and wages increases helped improve the income of O v e r v i e w   1 1 FIGURE O.7  Greenfield and Brownfield FDI Firms Perform Better than Domestic Firms over the First Five Years of Operation a. Wages b. Labor productivity Value added per employee (US$, millions, Wage growth (US$, thousands, constant 4 0.015 constant 2010), controlled average 2010), controlled average 3 0.010 2 0.005 1 0 0 1 2 3 4 5 0 1 2 3 4 5 Number of years after firm birth Number of years after firm birth Greenfield MNE Brownfield MNE Domestic c. Employment d. Multiple product offerings 3.0 0.20 Probability of multiple product offerings, Log (employees), controlled average 2.8 controlled average (%) 0.15 2.6 2.4 0.10 2.2 2.0 0.05 0 1 2 3 4 5 0 1 2 3 4 5 Number of years after firm birth Number of years after firm birth Greenfield MNE Brownfield MNE Domestic e. Exports f. Imports 0.5 0.7 Probability of imports, controlled Probability of exports, controlled 0.4 0.6 0.5 average (%) average (%) 0.3 0.4 0.2 0.3 0.1 0.2 0 1 2 3 4 5 0 1 2 3 4 5 Number of years after firm birth Number of years after firm birth Greenfield MNE Brownfield MNE Domestic Source: World Bank calculations, based on industrial censuses from six countries. Note: For this figure, industrial census data were analyzed from China, Côte d’Ivoire, Indonesia, Moldova, Serbia, and Vietnam (as further described in chapter 2, annex 2A). Vertical bars indicate the margin of error. “Domestic” refers to the firms that originated as domestic enterprises. Growth paths of firm outcomes can be captured in a simple framework using an interaction between indicators of firm group (greenfield, brownfield, and domestic) and years after entry in the following specification: y ics = groupi β ⋅ dt + α cs + δ dit0 + ε ics . The sample is restricted to cohorts whose entry is observed. To account for differences that might be driven by country characteristics, sector composition, and macroeconomic trends, the regressions also control for country-sector fixed effects and cohort fixed effects. FDI = foreign direct investment; MNE = multinational enterprise. 1 2    Global Investment Competitiveness Report 2019/2020 FIGURE O.8  Employment and Wages in Brownfield Firms Grow Faster than in Domestic Firms with Similar Characteristics a. Employment (aggregate sample) b. Wages (aggregate sample) 5.2 3.5 Log (employees), controlled average Wage growth (US$, thousands, constant 5.0 2010), controlled average 3.0 4.8 2.5 4.6 4.4 2.0 −2 −1 0 1 2 3 4 5 6 7 −2 −1 0 1 2 3 4 5 6 7 Years postacquisition Years postacquisition Domestic Foreign acquisitions Source: World Bank calculations, based on industrial censuses from six countries. Note: For this figure, industrial census data were analyzed from China, Côte d’Ivoire, Indonesia, Moldova, Serbia, and Vietnam (as further described in chapter 2, annex 2A). Vertical bars indicate the margin of error. “Domestic” refers to the firms that originated as domestic enterprises. Growth paths of firm outcomes can be captured in a simple framework using an interaction between indicators of firm group (greenfield, brownfield, and domestic) and years after entry in the following specification: y ics = groupi β ⋅ dt + α cs + δ dit0 + ε ics . The sample is restricted to cohorts whose entry is observed. To account for differences that might be driven by country characteristics, sector composition, and macroeconomic trends, the ­regressions also control for country-sector fixed effects and cohort fixed effects. Wage growth paths are calculated using constant deflated values in US dollars. FDI = foreign direct investment. the bottom 40 percent of the population see relative declines in formal employment in all studied countries. However, the distri- and wages. Overall, this skill premium is butional effects differ significantly across more pronounced for FDI in services than in the three countries (figure O.9). In Ethiopia, manufacturing. the benefits of FDI are more concentrated Given that FDI disproportionately benefits in the bottom 40 percent, while in Vietnam, better-educated and higher-skilled workers, the welfare gains are evenly distributed those labor force participants who lack these across the income distribution. Turkey had characteristics tend to be left behind. Such the greatest average wage benefits from FDI workers tend to be more concentrated in the but also experienced increases in income less economically advanced parts of their inequality in services.8 countries; as a result, FDI can exacerbate geo- These differences in the distributional graphic disparities within economies. In par- effects of FDI across the three countries are ticular, the analysis of Turkey presents a case likely driven by differences across sectors and of FDI-led skill premiums leading to wage dis- workers’ education levels (table O.1). In gen- persion, explaining why FDI in Turkey is eral, the average effects of FDI on formal associated with an increase in income inequal- employment and wages are positive for man- ity. This dynamic emphasizes the importance ufacturing and high-skilled services but neu- of a country’s labor market and education tral for extractive sectors and low-skilled policies. services. The effects of multinational firms’ pro- The analysis also finds significant evidence duction patterns on income and wage dis- of a skill premium for high-skilled versus low- parities are also explored in the recent World skilled workers in FDI affiliates. In regions Development Report on global value chains and sectors with higher MNE activity (rela- (World Bank 2020). The report finds that tive to those not receiving FDI), higher-skilled GVCs increase wage inequality in countries workers experience large benefits while low- at all income levels for at least three reasons: skilled workers may see no changes or even First, FDI and offshoring increase the O v e r v i e w   1 3 FIGURE O.9  FDI Has Had Varied Effects on National Income Distributions in Ethiopia, Vietnam, and Turkey 8 6 Wage effect from FDI (%) Turkey 4 Vietnam 2 Ethiopia 0 0–10 10–20 20–30 30–40 40–50 50–60 60–70 70–80 80–90 90–100 Income decile Ethiopia: bottom 40% Ethiopia: top 60% Vietnam: bottom 40% Vietnam: top 60% Turkey: bottom 40% Turkey: top 60% Source: World Bank calculations; individual labor market data from World Bank’s International Income Distribution Database (I2D2), supplemented with Labor Force Surveys; firm-level data from sources including Ethiopia’s Large and Medium Manufacturing Survey, Vietnam’s Enterprise Census, and Turkey’s Enterprise Information System data. Note: FDI = foreign direct investment. TABLE O.1  The Effects of FDI on Labor Markets Vary by Sector and Workers’ Skill Levels Broad sector Average effect Low-skilled workers High-skilled workers Extractives No effect No effect No effect Low-skilled manufacturing Positive Positive Positive High-skilled manufacturing Positive Neutral Positive Low-skilled services Neutral Negative Positive High-skilled services Positive Negative Positive Source: World Bank calculations; individual labor market data from World Bank’s International Income Distribution Database (I2D2), supplemented with Labor Force Surveys; firm-level data from sources including Ethiopia’s Large and Medium Manufacturing Survey, Vietnam’s Enterprise Census, and Turkey’s Enterprise Information System data. Note: In the table title, “effects of FDI on labor markets” refers to the effects on both wages and the probability of formal employment. The table summarizes an analysis of data from three countries: Ethiopia, Turkey, and Vietnam. (For complete results, see chapter 3, annex 3C.) “Low-skilled” workers are those with primary education or less, while “high-skilled” workers have completed at least secondary education. All results are relative to workers in sectors with less or no investment by multinational enterprises (MNEs). FDI = foreign direct investment. demand for skilled workers in low- and talent” in the developing countries and bid middle-income economies and put upward ­ up the wages of skilled workers. Third, firms pressure on wage inequality. Second, GVCs in GVCs tend to adopt more capital-­ intensive are often more skill-sensitive because they techniques than comparable domestic firms. tend to produce goods destined for quality- The deepening and upgrading of physical sensitive consumers in high-income coun- capital contribute to the increase in the rela- tries. This can in turn create “a war for tive demand for skilled workers. 1 4    Global Investment Competitiveness Report 2019/2020 Pillar 2: Government Actions Can their property rights and enforce those rights Rebuild Investor Confidence—Reducing in a predictable and efficient manner. Investor Risk, Fostering Investment Economic theory suggests that when investors Expansion, and Attracting New incur fixed and irreversible setup costs, uncer- FDI through Policy Predictability, tainty about the local conditions—especially Regulatory Certainty, and Targeted policy uncertainty—will have a dampening Investment Promotion effect that reduces investors’ response to new The COVID-19 pandemic has rapidly esca- investment opportunities (Bernanke 1983; lated business uncertainty, in turn magnify- Bloom 2009; Dixit 1989). Amid the COVID- ing investment risks and depressing foreign 19 outbreak, nationalization of essential sup- investor confidence. Multinational firms are ply chains, cancellation of government realizing that their historical push toward procurement contracts, and exchange control low-cost, low-inventory supply chains restrictions have come as sudden regulatory has opened them up to significant risk. changes. Investors identify these political risks In response, some of them are changing among their top concerns in the current crisis. their corporate strategies, reassessing their It is therefore vital for governments to approaches to sourcing production inputs, endeavor to reduce investor risk and help diversifying their suppliers, and making restore their confidence. greater use of digital technologies (Baldwin This report presents a new global database and Evenett 2020). and a novel quantitative measure of regula- They are also responding to changes tory risk (see chapter 4). This measure draws in the policy environments, which in some on, among others, data on the content of markets have seen introductions of more-­ domestic laws and international treaties to restrictive regulations, including during the assess countries’ regulatory frameworks for outbreak. For example, to protect sensitive investment in three dimensions (­ figure O.10): assets from foreign takeovers—notably in transparency, protection, and recourse. More sectors such as health, medical research, bio- specifically, it evaluates (a) transparency and technology, and infrastructure—some coun- predictability in both the content and process tries are adopting new foreign investment of making laws and regulations that apply to screening mechanisms. investors; (b) legal protection of investors Traditionally, investors rely on a country’s against arbitrary and nontransparent govern- legal and regulatory framework to recognize ment interference; and (c) investor access to FIGURE O.10  Three Pillars of Addressing Regulatory Risk Pillar 1. Transparency Pillar 2. Protection Pillar 3. Recourse Is there transparency regarding the What is the extent of legal protection Do investors have access to content as well as the process of provided to investors against effective mechanisms for recourse making laws and regulations that arbitrary, unpredictable, or in case of grievances or disputes? apply to investors? nontransparent government actions? • Systematic publication of and consultation on laws • Absolute treatment standards • Investor-state dispute settlement and prevention and regulations • Protection guarantees against direct and indirect • Land dispute resolution • Registries or ICT platforms and similar mechanisms expropriation, transfers of funds, fair and • Quality of judicial processes to allow investors to find information about relevant equitable treatment (FET) laws and regulations • Specificity and clarity of legal provisions (to reduce space for discretion) Source: World Bank GIC Report team. Note: ICT = information and communication technology. O v e r v i e w   1 5 effective mechanisms for recourse, including The effect of regulatory risk on FDI is siz- grievance management and dispute settlement. able and comparable in magnitude to the Evidence at both the country and investor investment-enhancing effects of trade open- levels suggests that regulatory risk—as mea- ness in the same regression models. In fact, in sured in this framework—matters for invest- some of the models, the effect of regulatory ment decisions. First, at the country level, risk on FDI exceeds that of trade openness, higher regulatory risk is correlated with showing that a 1 percentage point reduction higher risk premia measured by other indexes. in regulatory risk increases the likelihood of Second, higher regulatory risk is associated an investor entering or expanding in a host with lower FDI inflows (figure O.11). country by 0.5–2 percentage points. In con- Consistent with this result, investor data trast, a 1 percentage point increase in the host lend support at the microeconomic level to country’s trade-GDP ratio is associated with a the negative relationship found between 0.3–0.6 percentage point increase in an inves- regulatory risk and FDI. To test the rela- tor’s likelihood to enter or expand. tionship between a host country’s regula- The critical importance of the regulatory tory risk and foreign companies’ investment environment is further confirmed by results entry and expansion decisions, the report from the 2019 GIC Survey, in which investors uses a dataset of over 14,000 parent com- rank countries’ legal and regulatory environ- panies investing in nearly 28,000 FDI ments as one of the top three factors for greenfield and expansion projects across investment. In line with findings from 168 host countries between 2014 and the 2017 GIC Survey (World Bank 2018), 2016.9 Estimations from this investor loca- 84 percent of respondents list regulatory tion decision model suggest that regulatory ­environment as an “important” or “critically risk can deter MNEs from entering or important” factor in their investment deci- expanding operations in a country. sions (figure O.12). FIGURE O.11  FDI Inflows Are Higher in Countries with Lower Regulatory Risk a. 2014–17 panel score b. 2017 cross-section score 20 10 Net FDI inflows (log transformation) Net FDI inflows (log transformation) 10 5 0 0 –10 –20 –5 10 20 30 40 50 60 70 80 90 100 30 40 50 60 70 80 Nonweighted linear score, panel Nonweighted linear score, cross-section including laws Conditional on GDPPC, trade openness, and country FE Conditional on GDPPC and trade openness Non-tax-haven countries 95% CI Linear fit (excluding tax havens) Tax havens Source: World Bank calculations, from the World Development Indicators database. Note: The scatterplots show the correlation between net foreign direct investment (FDI) inflows and regulatory risk index. Panel a uses a 2014–17 panel score; panel b uses a 2017 cross-section score. CI = confidence interval; FE = fixed effects; GDPPC = GDP per capita. 1 6    Global Investment Competitiveness Report 2019/2020 FIGURE O.12  The Legal and Regulatory Environment Was among the Top Three Factors for FDI in 2019 How important were the following factors in your parent company’s decision to invest in this country? Political stability 49.4 34.9 Macroeconomic stability 49.0 35.7 Legal and regulatory environment 42.0 42.1 Talent/skills 41.5 43.5 Low taxes 41.3 35.5 Market size 39.3 42.6 Physical infrastructure 38.6 40.8 Ability to export 38.4 31.7 Intellectual property protections 35.4 35.4 Investor protections 34.5 37.5 Low labor and input costs 33.2 41.4 Supply chain coordination 30.1 44.2 Local input sourcing 26.6 42.2 Resource endowments 18.5 31.7 Local acquisition targets 14.4 35.6 0 20 40 60 80 100 Share of respondents (%) Critically important Important Source: Computation based on 2019 GIC Survey. Note: Affiliates of multinational enterprises (MNEs) were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. FDI = foreign direct investment. When it comes to investors who consider related to protectionism and economic nation- these factors “critically important,” the legal alism. To further advance this objective, a and regulatory conditions of the host growing group of countries is calling for a c ountries rank behind only political and ­ new multilateral framework on investment m acroeconomic stability, and ahead of ­ facilitation. Although the framework’s future considerations such as low taxes and low ­ is not yet clear, its emerging contours suggest input costs. The legal and regulatory environ- it could encompass a set of practical measures ment is especially important for larger firms. concerned with improving the transparency On average, large firms rank it as their top and predictability of investment frameworks; investment consideration.10 streamlining procedures related to foreign Overall, to reduce risks, governments need investors; and enhancing coordination and to remain committed to creating open and cooperation between stakeholders such as predictable environments for FDI. Given that host and home country governments, foreign the sources of policy uncertainty that erode investors, and domestic corporations as well investor confidence are both international as societal actors (Berger, Gsell, and Olekseyuk and domestic, solutions at both levels are 2019; WTO 2017). needed. In addition to pursuing global, multilateral, Internationally, reaffirming commitments or bilateral efforts toward providing clearer to market access and rules-based international policy directions and investment frameworks, systems would decrease policy uncertainty the 2019 GIC Survey results presented in this O v e r v i e w   1 7 report point to the central role of domestic IPAs can help increase FDI inflows, attract policy stability, with a particular emphasis on higher-quality FDI, and even transform their political and macroeconomic conditions. economies (Charlton and Davis 2007; Freund Indeed, evidence shows that the key elements and Moran 2017; Harding and Javorcik of stable political environments include strong 2012; Moran et al. 2018; Morisset and institutions, a level playing field, and predict- Andrews-Johnson 2004; Wells and Wint able policy making. Macroeconomic stability 2000). They can play a significant role in is also vital, centered on implementing macro- strengthening their countries’ investment prudential policies, ensuring central bank inde- competitiveness (see chapter 5). pendence, and optimizing fiscal policy. Yet although IPAs have proliferated over Governments can further help reduce risks the past two decades, success stories are still for investors by improving the legal, regula- scarce, especially in the developing world. tory, and institutional frameworks for FDI. Many IPAs are unfocused—with too many Business operations can be made more pre- mandates and target sectors—and are not dictable by improving transparency and providing the key services investors expect. At reducing room for bureaucratic discretion. the same time, many IPAs are not evolving Transparency can be strengthened by system- dynamically enough to align with both chal- atically consulting with the private sector and lenges and opportunities in the changing FDI other stakeholders, developing information landscape. The current literature, combined portals to make laws and regulations publicly with surveys of IPAs and operational experi- available, and articulating clear and specific ence by the World Bank Group, suggests that FDI-related legal provisions and administra- IPAs can have greater positive impact if they tive procedures. sharpen their strategic alignment and focus, Investment promotion agencies can play a adopt a coherent institutional framework, critical role in these efforts given their role as and strengthen their delivery of investor ser- governments’ key interlocutors with foreign vices (figure O.13). In contrast, IPAs should businesses. Empirical evidence shows that not overestimate the role of investment FIGURE O.13  Core Elements for Increasing the Development Impact of Investment Promotion Agencies Development Impact Higher Investor Satisfaction and Confidence Strategic Coherent Strong alignment institutional investor service and focus framework delivery National Development Plan, Investment Policies, FDI strategy Source: World Bank, based on Heilbron and Aranda-Larrey 2020. Note: FDI = foreign direct investment. 1 8    Global Investment Competitiveness Report 2019/2020 BOX O.1 Key Success Factors in High-Performing IPAs in Developing Countries World Bank research and operational experience have • A high degree of institutional and financial auton- identified the following key success factors common to omy (or semiautonomy), emulating private sector high-performing investment promotion agencies (IPAs) flexibility to act according to agreed-upon strategic in developing countries: plans and to hire staff using specified and transpar- ent job qualifications; avoiding political interfer- • High-level government support (from the presi- ence; and providing sustainability through political dent or prime minister), granting a high priority to cycles. investment (or foreign direct investment [FDI]) and • An independent and well-functioning board of direc- directly or indirectly championing the needed legal, tors or advisory board with strong and active private regulatory, and institutional reforms for investment. sector representation to better understand investors • Strong strategic alignment stemming from consulta- and provide direction in catering to their needs. tions with public and private sectors and cascading from a national development plan or FDI strat- • A strong investor-centric service orientation to egy to IPA corporate plans and industry-specific design and provide relevant and high-quality ser- strategies. vices to investors throughout their investment cycle. • A clear, uncontested mandate , ideally focused • Management and key promotion staff with strong on investment promotion, especially when start- private sector experience, as well as international ing or restructuring the IPA. Developing-country exposure and language skills, within the IPA’s IPAs with multiple mandates take much longer mix of employees with public and private sector to, or never do, deliver substantial FDI impact. experience. Regulatory functions (including one-stop shops) are • Sufficient and sustained financial resources over best performed by a separate public institution that three- to five-year periods to provide continuity of ensures proper delivery of this essential function strategic efforts over the long-cycle nature of invest- without compromising the equally essential invest- ment promotion and to avoid struggling over funds ment promotion mandate of an IPA. every year or having to charge fees. Source: Adapted from Heilbron and Whyte 2019. incentives in increasing a location’s overall of existing foreign investors as well as pres- investment competitiveness, although these ervation of supply chains connecting ­ foreign may be needed to help companies during the firms and their domestic suppliers. Through pandemic crisis.11 IPAs’ responses to market signals and When creating or strengthening their IPAs, MNEs’ needs, governments have an oppor- policy makers should consider critical success tunity to minimize the risk exposure of factors (box O.1). The right strategic and MNEs and their associated supply chain institutional frameworks vary, depending on linkages. Specific investment services to be the country’s political economy, the govern- prioritized by IPAs include (a) identifying ment’s existing institutional setup, available and directly contacting at-risk or systemi- legal formats, the civil service culture, and the cally strategic firms according to number of institutional collaboration culture (Heilbron employees, region, or sector; (b) expediting and Whyte 2019). foreign exchange approvals; and (c) advo- In the current COVID-19 crisis situation, cating for urgent government actions IPAs are in large part shifting their principal to solve the firms’ grievance issues more focus from FDI attraction to retention systematically. O v e r v i e w   1 9 Governments Can Leverage priorities and by fostering open and condu- cive trade and investment environments. FDI for Robust Recovery from COVID-19 Seize New Opportunities from Changing What can governments do, on the one hand, FDI and GVC Patterns to leverage FDI to strengthen the resilience of their economies and help absorb future Countries can seize new opportunities to shocks, and on the other hand, to turn increase their competitiveness for FDI as a the current COVID-19 crisis into new oppor- result of shifting trade and investment pat- tunities to increase their competitiveness terns and policies. In the face of higher tar- for FDI? iffs resulting from the 2019 trade war between China and the United States, importers have already sought new sourcing Avoid Protectionist Policies locations in the global marketplace Governments should avoid protectionist (Constantinescu et al. 2019). Trade diversion ­ p olicies, which would further exacerbate may in turn cause a shift in FDI as firms ­ disruptions to GVCs and amplify the already adjust global supply chains and centers of elevated uncertainty. Instead, to attract addi- production (Blanchard 2019). Developing tional investment, countries should counter economies with large export bases could the global protectionist trend by further eas- emerge as suitable FDI hosts. Specifically, ing FDI entry and operational restrictions. countries that already export similar prod- Being more open to FDI relative to peers ucts are likely to attract greater investment helps attract new investment. In fact, some (Cali 2018). countries are already using this crisis as an With the COVID-19 crisis, the push to opportunity to open new sectors of their diversify supply chains will likely be intensi- economies to foreign investment. fied. Yet no consensus has emerged on how Enhanced regional cooperation can also be the global FDI and GVC landscape will look a critical element in the removal of barriers to after COVID-19. Some economists hold the intraregional trade and investment. Regional view that no major changes will take place integration helps countries overcome divi- and that adjustments will concentrate in sions that impede the flow of goods, services, health-related industries, as the economic capital, people, and information. These rationale for GVCs holds the same (Baldwin divisions are a constraint to economic growth, ­ and Evenett 2020; Freund 2020; Miroudot especially in developing countries. While 2020). Others believe that COVID-19 has Europe, North America, and East Asia have become a wake-up call for a new balance historically led the way in regional integra- between risk and reward for GVCs, as pan- tion, the momentum has lately also increased demics, climate change, natural disasters, and in some of the less integrated regions—as other man-made crises may expose the world e videnced, for example, by the recently ­ to increased risks (Goldberg 2020; Javorcik c oncluded negotiations on the African ­ 2020). Regardless of which outcome prevails, Continental Free Trade Area (AfCFTA). as the main architects of GVCs, multinational Experience has shown that deepened regional firms will adjust production networks to integration allows countries to improve mar- improve their resilience and robustness in ket efficiency, accelerate reform processes in a response to COVID-19. coordinated and predictable manner, and Policy makers should reflect on these pos- foster multiregional cooperation.12 Bilateral ­ sible shifts in investment preferences and let and regional trade and investment agreements business realities guide their policy response. also help enhance policy certainty by commit- Countries should assess which sectors and ting national governments to specific policy value chains have proven resilient during the 2 0    Global Investment Competitiveness Report 2019/2020 COVID-19 crisis. This will involve evaluating Notes the risk exposure, value proposition, and 1. FDI data (here taken from the World Bank’s competitiveness factors of individual sectors World Development Indicators database) and value chains. In addition, governments should be interpreted with caution. Research should identify emerging competitive sectors shows that multinational corporations’ tax in their countries that may arise from the engineering and the role of investment hubs ­ p ossible reorganization of GVC and FDI distort traditional FDI statistics. A growing landscapes. body of evidence demonstrates that multi- Should new investment patterns emerge, national corporations are reallocating roy- they will require new priorities in investment alties and other intangible assets to low-tax policy and promotion reforms. These will locations to reduce their aggregate corporate entail realigning the investment incentive tax liability. Such “phantom investment” regimes to the new national development pri- into corporate shells may account for almost 40 percent of global FDI (Damgaard, Elkjaer, orities likely to emerge after COVID-19, such and Johannesen 2019). At the same time, the as job creation. Reforms may also be needed main results reported in figure O.1 still hold, to limit or phase out crisis-related investment even if the analysis excludes tax havens as screening and approval mechanisms. In FDI destinations. ­ a ddition, measures to address investor 2. “Developing” countries in this report refers to ­ protections and grievance issues might be low- and middle-income countries as defined appropriate in some countries to enable gov- by the World Bank. “Developed” countries ernments to resolve grievances before they are high-income countries. For the definitions become legal disputes. Finally, measures to of all income classifications and the coun- increase local firms’ resilience and to tries therein, see “World Bank Country and strengthen supplier development programs Lending Groups,” World Bank Knowledge Base: https://datahelpdesk.worldbank.org​ will be needed to enhance FDI linkages to the /knowledgebase/articles/906519-world-bank​ local economies. -country-and-lending-groups. 3. The 2017 Tax Cuts and Jobs Act (TCJA) essentially exempts US companies’ foreign Strengthen Global Cooperation earnings from taxation, albeit with a one-off Tackling the complex challenges presented tax on past profits to ease the transition to by the current global environment will the new system (Toder 2018). The implemen- tation of the TCJA led to a massive increase require global cooperation. The pandemic in the repatriation of foreign-earned profits has illustrated the shared public health and by the US multinationals back to the United economic vulnerabilities that countries face. States, resulting in negative FDI inflows from It has also highlighted the critical impor- the United States for the affected host coun- tance of exchanging data, sharing informa- tries (OECD 2019). Although reinvested FDI tion on good practices, and strengthening earnings returned to positive levels in the collaboration. first half of 2019—suggesting that many of The magnitude and scale of the crisis the negative FDI flows were from one-time require policy makers to employ their full repatriations of past profits—rates of rein- arsenal of policy tools to improve business vestment remain below averages observed in confidence and boost countries’ invest- the five years leading up to the implementa- tion of the TCJA (OECD 2019). This pattern ment competitiveness. During the global may signal a “new normal” for reinvestment financial crisis, an unprecedented synchro- levels as US companies now have fewer nized, coordinated policy response was crit- incentives to reinvest their foreign earnings ical to containing it. Once again, the times to avoid taxation (OECD 2019). are testing policy makers. They must rise to 4. The Pew Research Center’s Spring 2018 the o­ ccasion by acting quickly, decisively, Global Attitudes Survey included respond- and collaboratively. ents from 27 countries: Argentina, Australia, O v e r v i e w   2 1 Brazil, Canada, France, Germany, Greece, Uexkull 2018; UNIDO 2011; World Bank Hungary, India, Indonesia, Israel, Italy, 2018). Effective use of incentives requires Japan, Kenya, the Republic of Korea, greater regional and international coordi- Mexico, the Netherlands, Nigeria, the nation, political commitment, and common Philippines, Poland, the Russian Federation, reporting standards to enhance transparency South Africa, Spain, Sweden, Tunisia, the (IMF et al. 2015). United Kingdom, and the United States. 12. See “Overview,” Regional Integration topic 5. This short, English language, web-based sur- page, World Bank website: https://www​ vey was sent to known email addresses of .worldbank.org/en/topic/regional-integration​ MNEs, leveraging existing sampling frames /overview. for developing countries (World Bank, forth- coming). To extend reach, the survey was also circulated to known foreign investors through the countries’ investment promotion References agencies (IPAs). The period of data collec- Akame, Afuge Junior, Metougue Eric Ekwelle, tion was March 24 to April 24, 2020. Data and George Ndonghandou Njei. 2016. “The underlying the analysis comprise responses Impact of Business Climate on Foreign Direct from 105 MNE affiliates operating in 26 Investment in the CEMAC Region.” Journal developing countries. The results of the pulse of Economics and Sustainable Development survey are not generalizable to all develop- 7 (22): 66–74. ing countries but are an indicative estimate Alfaro, Laura, Sebnem Kalemli-Ozcan, and of impact of MNEs operating in developing Vadym Volosovych. 2008. “Why Doesn’t countries. Capital Flow from Rich to Poor Countries? An 6. Cross-border acquisition data are from the Empirical Investigation.” Review of Economics Thomson Reuters Refinitiv Mergers and and Statistics 90 (2): 347–68. Acquisitions Database (https://www.refinitiv​ Andersen, Maria R., Benjamin R. Kett, and .com/en/financial-data/company-data). Erik von Uexkull. 2018. “Corporate 7. At the same time, several notable exceptions Tax Incentives and FDI in Developing exist. These studies tend to focus on the Countries.” In 2017/2018 Global Investment employment and productivity of acquired Competitiveness Report: Foreign Investor firms in the context of a single developing Perspectives and Policy Implications, 73–100. country: Arnold and Javorcik (2009); Bircan Washington, DC: World Bank. (2019); Gong, Görg, and Maioli (2007); and Arnold, Jens Matthias, and Beata S. Javorcik. Lipsey, Sjöholm, and Sun (2013). 2009. “Gifted Kids or Pushy Parents? Foreign 8. Data analysis conducted for this chapter Direct Investment and Plant Productivity in finds that Turkey’s Gini coefficient currently Indonesia.” Journal of International Economics stands at 0.35, but it would have been 0.33 79 (1): 42–53. (indicating lower inequality) without FDI. Baker, Scott R., Nicholas Bloom, and Steven J. 9. These data are from fDi Markets, a Financial Davis. 2016. “Measuring Economic Policy Times dataset (https://www.fdimarkets.com). Uncertainty.” Quarterly Journal of Economics 10. These differences may be driven by the pres- 131 (4): 1593–636. ence of restrictions that are applicable only Baker, Scott R., Nicholas Bloom, and Steven J. Davis. to larger firms and the greater regulatory 2019. “The Extraordinary Rise in Trade Policy scrutiny that large companies tend to experi- Uncertainty.” VoxEU, September 17 (accessed ence. November 7, 2019), https://voxeu.org/article​ 11. This guidance suggests that developing coun- /extraordinary-rise-trade-policy-uncertainty. tries should be careful and conservative in Baldwin, Richard E., and Simon J. Evenett, eds. their use of tax incentives to stimulate their 2020. COVID-19 and Trade Policy: Why investment competitiveness. Other factors Turning Inward Won’t Work. London: CEPR such as good investment climates, politi- Press. cal stability, regulatory quality, and market Berger, Axel, Sebastian Gsell, and Zoryana opportunities are more critical to investors’ Olekseyuk. 2019. “Investment Facilitation initial location considerations than are tax for Development: A New Route to Global rates and incentives (Andersen, Kett, and von Investment Governance.” Briefing Paper 2 2    Global Investment Competitiveness Report 2019/2020 No. 5/2019, German Development Institute Freund, Caroline, and Theodore H. Moran. 2017. (DIE), Bonn. “Multinational Investors as Export Superstars: Bernanke, Benjamin S. 1983. “Irreversibility, How Emerging Market Governments Can Uncertainty, and Cyclical Investment.” Quarterly Reshape Comparative Advantage.” Working Journal of Economics 98 (1): 85–106. Paper No. 17-1, Peterson Institute for Bircan, Çağatay. 2019. “Ownership Structure and International Economics, Washington, DC. Productivity of Multinationals.” Journal of Goldberg, Pinelopi Koujianou. 2020. “A Silent International Economics 116: 125–43. Hero of the Coronavirus Crisis.” Project Blanchard, Emily J. 2019. “Trade Wars in the Syndicate, March 17. Global Value Chain Era.” In Trade War: The Gong, Yundan, Holger Görg, and Sara Maioli. Clash of Economic Systems Endangering 2007. “Employment Effects of Privatisation Global Prosperity , edited by Meredith A. and Foreign Acquisition of Chinese State- Crowley, 57–63. London: CEPR Press. Owned Enterprises.” International Journal of Bloom, Nicholas. 2009. “The Impact of Uncertainty the Economics of Business 14 (2): 197–214. Shocks.” Econometrica 77 (3): 623–85. Gramlich, John. 2019. “5 Trends in International Bloom, Nicholas. 2014. “Fluctuations in Public Opinion from Our Searchable Global Uncertainty.” Journal of Economic Perspectives Indicators Database.” Fact Tank (online 28 (2): 153–76. platform) report, April 25, Pew Research Caldara, Dario, Matteo Iacoviello, Patrick Center, Washington, DC. Molligo, Andrea Prestipino, and Andrea Harding, Torfinn, and Beata S. Javorcik. 2012. Raffo. 2019. “The Economic Effects of Trade “Investment Promotion and FDI Inflows: Policy Uncertainty.” International Finance Quality Matters.” Economics Series Working Discussion Paper 1256, Board of Governors of Papers 612, University of Oxford. the Federal Reserve System, Washington, DC. Heilbron, Armando, and Yago Aranda-Larrey. doi:10.17016/IFDP.2019.1256. 2020. “Strengthening Service Delivery Cali, Massimiliano. 2018. “The Impact of of Investment Promotion Agencies: The the US-China Trade War on East Asia.” Comprehensive Investor Services Framework.” VoxEU, October 16. https://voxeu.org/article​ Finance, Competitiveness and Innovation in /impact-us-china-trade-war-east-asia. Focus. World Bank, Washington, DC. Charlton, Andrew, and Nicholas Davis. 2007. Heilbron, Armando, and Robert Whyte. 2019. “Does Investment Promotion Work?” The B.E. “Institutions for Investment: Establishing a Journal of Economic Analysis & Policy 7 (1): High-­ P erforming Institutional Framework 1–21. for Foreign Direct Investment (FDI).” Invest- Constantinescu, Cristina, Aaditya Mattoo, Michele ment Climate In Focus Note, World Bank, Ruta, Maryla Maliszewska, and Israel Osorio- ­Washington, DC. Rodarte. 2019. “Global Trade Watch 2018: IMF, OECD, UN, and World Bank (International Trade Amid Tensions.” Report No. 137201, Monetary Fund, Organisation for Economic World Bank, Washington, DC. Co-operation and Development, United Damgaard, Jannick, Thomas Elkjaer, and Niels Nations, and World Bank). 2015. “Options Johannesen. 2019. “What Is Real and What for Low Income Countries’ Effective and Is Not in the Global FDI Network?” Working Efficient Use of Tax Incentives for Investment.” Paper No. 19/274, International Monetary Report to the G-20 Development Working Fund, Washington, DC. Group, issued as an IMF Policy Paper, Dixit, Avinash. 1989. “Entry and Exit Decisions Washington, DC. under Uncertainty.” Journal of Political Javorcik, Beata. 2020. “Global Supply Chains Economy 97 (3): 620–38. Will Not Be the Same in the Post-COVID-19 Freund, Caroline. 2020. “Governments Could World.” In COVID-19 and Trade Policy: Bring Supply Chains Home. It Would Defy Why Turning Inward Won’t Work , edited Economic Rationality.” Barron’s, May 1. by Richard E. Baldwin and Simon J. Evenett, Freund, Caroline, Michael Ferrantino, Maryla 111–16. London: CEPR Press. Maliszewska, and Michele Ruta. 2018. Knight, Frank H. 1921. Risk, Uncertainty, and “Impacts on Global Trade and Income of Profit. Boston: Hart, Schaffner, and Marx; Current Trade Disputes.” Macroeconomics, Houghton Mifflin. Trade & Investment (MTI) Practice Notes, Lipsey, Robert E., Fredrik Sjöholm, and Jing Sun. No. 2, World Bank, Washington, DC. 2013. “Foreign Ownership and Employment O v e r v i e w   2 3 Growth in a Developing Country.” Journal of UNIDO (United Nations Industrial Development Development Studies 49 (8): 1133–47. Organization). 2011. Africa Investor Report Miroudot, Sébastien. 2020. “Resilience versus 2011: Towards Evidence-based Investment Robustness in Global Value Chains: Some Promotion Strategies. Vienna: UNIDO. Policy Implications.” In COVID-19 and Trade Wells, Louis T. Jr., and Alvin G. Wint. 2000. Policy: Why Turning Inward Won’t Work , “Marketing a Country: Promotion as a Tool edited by Richard E. Baldwin and Simon J. for Attracting Foreign Investment.” Revised ed. Evenett, 117–30. London: CEPR Press. Occasional Paper No. 13, Foreign Investment Moran, Theodore H., Holder Görg, Adnan Advisory Service, World Bank, Washington, DC. Seri č, and Christiane Krieger-Boden. 2018. Wernick, David A., Jerry Haar, and Shane “Attracting FDI in Middle-Skilled Supply Singh. 2009. “Do Governing Institutions Chains.” Discussion Paper No. 2018-2, Affect Foreign Direct Investment Inflows? Kiel Institute for the World Economy, Kiel, New Evidence from Emerging Economies.” Germany. International Journal of Economics and Morisset, Jacques, and Kelly Andrews-Johnson. Business Research 1 (3): 317–22. 2004. The Effectiveness of Promotion Agencies World Bank. 2018. Global Investment Com- at Attracting Foreign Direct Investment . petitiveness Report 2017/2018: Foreign Occasional Paper No. 16, Foreign Investment Investor Perspectives and Policy Impli- Advisory Service, World Bank, Washington, DC. cations . Washington, DC: World Bank. OECD (Organisation for Economic Co-operation doi:10.1596/978-1-4648-1175-3. and Development). 2019. “FDI in Figures, World Bank. 2019. Global Economic Prospects, October 2019.” Semiannual data report, OECD, June 2019: Heightened Tensions, Subdued Paris. Investment. Washington, DC: World Bank. Rodrik, Dani. 1991. “Policy Uncertainty and Private World Bank. 2020.  World Development Investment in Developing Countries.” Journal of Report 2020: Trading for Development in the Development Economics 36 (2): 229–42. Age of Global Value Chains. Washington, DC: Toder, Eric. 2018. “Explaining the TCJA’s World Bank. International Reforms.” TaxVox (blog) , World Bank and OECD (Organisation for February 2. http://www.taxpolicycenter.org​ Economic Co-operation and Development). /taxvox/explaining-tcjas-international-reforms. 2017. A Step Ahead: Competition Policy for UNCTAD (United Nations Conference on Trade Shared Prosperity and Inclusive Growth . and Development). 2019. World Investment Washington, DC: World Bank and OECD. Report 2019: Special Economic Zones. WTO (World Trade Organization). 2017. New York: UNCTAD. “Joint Ministerial Statement on Investment UNCTAD. 2020. Investment Trends Monitor Facilitation for Development.” Adopted at the (Special Issue): Impact of the Coronavirus 11th Ministerial Conference, December 13, Outbreak on Global FDI. Geneva: UNCTAD. WTO, Geneva. 1 Outlook and Priorities for Foreign Investors in Developing Countries: Findings from the 2019 Global Investment Competitiveness Survey in 10 Middle-Income Countries Peter Kusek, Abhishek Saurav, and Ryan Kuo Note: Since the Global Investment Competitiveness survey was conducted between June and November 2019, the results do not capture the effects of the COVID-19 pandemic on foreign investors. Key Findings • An extensive survey of more than 2,400 foreign investors in 10 large middle-income countries, conducted between June and November 2019, shows that foreign-owned firms face significant trade and investment policy uncertainty that can negatively affect future investment decisions. Since the survey was conducted before the COVID-19 outbreak, the results do not capture the effects of the pandemic on foreign-owned firms. The 10 countries covered by the survey are Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. The surveyed companies cumulatively represent around US$400 billion in total investment (about 10 percent of FDI stock in the surveyed countries) and employ nearly 1 million workers, based on conservative estimates. • Two-thirds of investors report that policy uncertainty due to protectionism and economic nationalism in trade and investment is either “important” or “critically important” in their investment decisions—and among the latter group, more than half have already experienced a decrease in employment, firm produc- tivity, or investments in the last year. Investor confidence decreases when the direction of policy making is unclear or unpredictable. Large firms and importers have been particularly sensitive to the effects of policy uncertainty in trade and investment. • Even before the COVID-19 outbreak, many investors were holding off expansion plans—based on the survey, less than half of foreign businesses planned to expand investment over the next three years. However, results vary by country. Foreign businesses in China (17 percent of investors planning to expand ­ investments) and Turkey (35 percent) report being much less likely to expand in the future than those in other surveyed countries. In contrast, about four-fifths of foreign affiliates in Nigeria and two-thirds in India plan to expand their investment stocks over the next three years. The effect of policy uncertainty in trade and investment—combined with domestic factors, such as macroeconomic fundamentals, political developments, and the legal and regulatory environment—are likely to shape foreign investors’ invest- ment plans in the surveyed countries. • The top three factors influencing investment decisions are political stability, macroeconomic stability, and a country’s legal and regulatory environment; nearly 9 in 10 businesses consider them to be “­ important” or “critically important.” These factors rank ahead of considerations such as low tax rates, low labor and input costs, and access to resource endowments. Furthermore, large firms (those with more than 250 ­ employees) rank an enabling regulatory environment as their top investment consideration. Investors that encounter major legal and regulatory obstacles are more likely to reduce or withdraw investment. • The COVID-19 pandemic represents an unprecedented shock to the global economy and MNEs, under- scoring the need for policies to bolster investor confidence. Against the backdrop of heightened policy uncertainty in trade and investment, the pandemic is set to further escalate uncertainty, magnify investment risks, and depress foreign investor confidence. These extraordinary challenges warrant a crisis management approach to governments’ responses. In addition to short-term crisis response, governments should address international and domestic sources of policy uncertainty by reaffirming commitments to global and regional trade and investment systems, promoting political stability, enhancing macroeconomic stability, and improv- ing legal and regulatory frameworks for FDI. Creating a predictable, business-friendly regulatory environ- ment goes beyond the rules on the books and includes their full and consistent implementation in practice. Introduction This chapter presents the results of the population, one-quarter of global gross 2019 Global Investment Competitiveness domestic product (GDP), and one-fifth of Survey (GIC Survey), a survey of execu- global trade. From a foreign direct tives of the affiliates of multinational investment (FDI) perspective, they accounted enterprises (MNEs) in 10 developing for 37 percent of global inflows and countries. 1 The phone-based survey data 75 percent of inflows to developing countries cover more than 2,400 foreign investors in 2018. As with developing countries in with operations in 10 middle-income general, FDI as a share of GDP has declined countries (MICs): Brazil, China, India, in the selected countries since the global Indonesia, Malaysia, Mexico, Nigeria, ­ figure 1.1). From financial crisis in 2008–09 ( Thailand, Turkey, and Vietnam. Using a precrisis average of 3 percent of GDP per self-reported data from surveyed execu- year, FDI inflows have contracted to less tives, the chapter serves two analytical than 2 percent in recent years.2 objectives: First, it assesses the effect of Most of the surveyed countries have high rising trade and investment policy uncer- statutory restrictions on FDI relative to the tainty on investors’ confidence and future global average (figure 1.2). Furthermore, investment prospects. Second, it examines countries more exposed to global mega- the role of a country’s legal and regulatory trends such as rising protectionism, eco- environment in shaping investment deci- nomic nationalism, and trade policy sions and identifies specific market entry tensions are in turn more vulnerable to and operational constraints faced by for- investment risks and declines in investor eign investors. confidence. The countries covered by the survey In most of the selected countries, FDI account for more than half of the global growth rates have stalled or declined from FIGURE 1.1  FDI Inflows to Middle-Income Countries Have Been Declining since the 2008–09 Global Financial Crisis 4.0 3.5 3.0 Net FDI inflows (% of GDP) 2.5 2.0 1.5 1.0 0.5 0 0 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 0 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 10 surveyed countries All developing countries Source: World Development Indicators database. Note: FDI = foreign direct investment. Surveyed countries are Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. “Developing” countries refers to low- and middle-income countries as defined by the World Bank. 25 2 6    Global Investment Competitiveness Report 2019/2020 FIGURE 1.2  Surveyed Countries Have High FDI Restrictiveness 0.35 0.30 0.31 0.25 0.25 0.25 Index score 0.20 0.21 0.19 0.15 Global 0.13 average 0.094 0.10 0.09 0.05 0.06 0 ia sia m o y es ina il ia xic rke tna az lay Ind on Ch Me Br Tu Vie Ma Ind Source: Organisation for Economic Co-operation and Development (OECD) 2018 FDI Regulatory Restrictiveness Index, https://www.oecd.org/investment​ /fdiindex.htm. Note: FDI = foreign direct investment. FDI Regulatory Restrictiveness Index scores range from 0 (open) to 1 (closed). The index scores are not available for Nigeria and Thailand. their levels a decade ago, and growth has Survey Methodology and even been negative in some countries (such as Brazil and Nigeria) in recent years. Respondent Profile An acute slowdown in FDI can sap growth The data used in this study are from the momentum, lower participation in global 2019 GIC Survey, conducted June– value chains, and limit positive spillovers to November 2019 through 30-minute phone domestic firms. Sluggish growth exacer- interviews in the primary business bates the countries’ risk of being trapped in language(s) of the host economies. The sur- middle-income status, limiting their ability vey was administered to senior executives of to undertake “second generation” struc- foreign-owned firms. Information was col- tural reforms. lected on the companies’ general characteris- FDI has been the largest source of exter- tics, the importance and effect of global nal finance for many developing countries— megatrends on business operations, contri- greater than remittances, private debt and bution to the host economy, and the impor- portfolio equity, or official development tance of investment policy factors and ­a ssistance. 3 Higher FDI inflows can ease operational obstacles they face. ­ c apital constraints, contribute to output The 2019 GIC Survey was designed to and employment growth, and increase generate results that are representative at aggregate productivity through positive the country level and comparable across productivity spillovers and technology countries. It targeted a statistically repre- transfers. sentative sample of foreign-owned firms This chapter offers practical evidence to across the 10 surveyed MICs.4 The target strengthen investment competitiveness by was to reach 125 interviews per sector identifying policy levers that can relax FDI (manufacturing and services). Each country barriers, de-risk countries’ investment cli- sample comprises roughly 250 MNE affili- mates, and facilitate additional FDI ates with at least five employees. The only inflows. Through its systematic, data- exception is Nigeria, where because of sam- driven identification of investment climate pling frame limitations, the sample com- policy barriers, the chapter reflects the col- prises 164 respondents (55 manufacturing lective voice of foreign investors on the target and 109 services). Thus, across the 10 ­ design and prioritization of investment countries, more than 2,400 responses were policy reforms. collected. O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    27 To assess changes in investor experience • Size: About one-quarter of surveyed and perceptions, a second round of the survey MNE affiliates are large, with more than is planned in 2020–21. To the extent possible, 250 employees. The remainder are small the second round will target respondents and medium enterprises (SMEs) with from the first round. For more details on the 250 or fewer employees, roughly half survey methodology, including sample repre- of which have 100 or fewer employees sentation and survey administration, see (figure 1.4, panel a). annex 1A. • Investment stock: Roughly one-quarter The remainder of this section outlines of the MNE affiliates have invested more the  survey respondent profiles and than US$10 million in host countries. ­ additional methodological features, More than one-tenth have invested more as follows: than US$50 million (figure 1.4, panel b). • Age: On average, the surveyed MNE • Sector and subsector: Survey respondents affiliates are fairly established in their represent a range of sectors and source respective markets. Nearly two-thirds of countries. By design, about half of the them have been in the host country for MNE affiliates were in the manufacturing more than a decade, and one-third for sector, and about half were in services. more than 20 years (figure 1.4, panel c). Within each sector, the sample covers • Ow nership: Roughly t wo-thirds of many subsectors (figure 1.3 and annex respondents are fully owned by foreign 1A, table 1A.2). investors (that is, foreign MNEs hold a FIGURE 1.3  Respondents Are Evenly Split between Manufacturing and Services Firms and Represent Firms across Various Specific Sectors Share of 2019 GIC Survey respondents, by subsector (percent) Services Manufacturing Financial services Construction Chemicals and Computer and including Rubber and Business services 4.8 2.4 chemical products software services insurance plastic products 4.5 3.5 3.5 2.0 Hotels, Real estate restaurants Electricity, 0.7 and Agroprocessing, gas, and tourism 0.7 Metals and metal products 5.1 Electrical and Textiles, water 1.3 food products, Telecom- Health and beverages electronic apparel, and Logistics, transport, and Administrative and Arts and recreation 0.2 equipment, and leather munications 0.3 3.0 storage 4.2 support services Other 0.5 components 2.1 2.1 Education 2.4 professional, 0.2 Residential care and social work 0.1 scientific, and Media tech. 0.8 0.4 Scientific research and Wood products, development services 0.2 Automobiles, other motor Information technology paper, and Water supply and waste vehicles, and transport and telecommunications printing 1.9 management 0.1 equipment 4.8 2.9 Refined petroleum products, coke, and nuclear fuel. 0.3 Pharmaceuticals, biotechnology, and medical devices 0.7 Wholesale and retail trade 11.1 Services: Other or unclassified 14.7 Machinery and equipment 6.2 Manufacturing: Other or unclassified 12.0 Source: Computation based on the 2019 GIC Survey. Note: The relative size of the rectangles represents the relative share of respondents in each overall sector (“services” or “manufacturing”). Services subsectors comprising less than 1 percent include scientific research and development (R&D), arts and recreation, and others. For the number and shares of respondents by subsector, see annex 1A, table 1A.2. 2 8    Global Investment Competitiveness Report 2019/2020 FIGURE 1.4  The Median MNE Affiliate Is Relatively Small, Well-Established, and Majority Foreign Owned Share of 2019 GIC Survey respondents (percent) a. Question: At the end of the last financial year, how b. Question: How much has your company invested in many employees did your company have? this country in total to date? 52% 22% 19% 6% 0% 58% 6% 11% 11% 13% 0 20 40 60 80 100 0 20 40 60 80 100 Percent Percent <100 100–250 251–1,000 10,000 Don’t know US$10–50 million Don’t know c. Question: How long has your company been operating d. Question: What percentage of your company is owned by in this country? foreign individuals, companies, or organizations? 4% 33% 33% 19% 11% 3% 6% 10% 15% 66% 0 20 40 60 80 100 0 10 20 30 40 50 60 70 80 90 100 Percent Percent 0–10 years 11–20 21–30 >30 Don’t know <25 25–49.9 50–74.9 75–99.9 100 Source: Computation based on the 2019 GIC Survey. Note: MNE = multinational enterprise. Percentages may not total 100 because of rounding. TABLE 1.1  Most Investors Come from High-Income Countries in Asia or Europe Share of 2019 GIC Survey respondents (percent) Question: In which country is your parent company or your company’s largest foreign owner located? Region Developed Developing Total East Asia and Pacific 37 7 44 Europe and Central Asia 33 1 34 North America 13 0 13 Sub-Saharan Africa 0 3 3 Middle East and North Africa 1 1 2 Latin America and the Caribbean 1 1 2 South Asia 0 2 2 Total 85 15 100 Source: Computation based on the 2019 GIC Survey. Note: “Developing” countries are low- and middle-income countries, and “developed” countries are high-income countries, as defined by the World Bank. “North America” is defined here as Canada and the United States. O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    29 100 percent stake), and about 10 percent invested more than US$5 million in host are minority owned by foreign investors economies, and more than 10 percent have (that is, MNEs hold less than a 50 percent invested more than US$50 million stake) (figure 1.4, panel d). (figure 1.4, panel b). Not all of this capital • Origin: Eighty-five percent of the for- comes from abroad: MNE affiliates reinvest eign investors come from high-income about 60 percent of their profits back into countries, while the vast majority of host economies. This result is in line with the remainder come from middle-­ income earlier survey evidence and literature show- countries (table 1.1). Low-income coun- ing that reinvested earnings are an important tries account for less than 1 percent of the source of FDI (Kusek and Silva 2018; foreign investors surveyed. UNCTAD 2016). Parent-affiliate relationships observed in survey data reveal both North-South The Development Contributions and South-South FDI flows. As noted of FDI in Host Economies earlier, 85 percent of parent MNEs are head- quartered in high-income countries, the A large body of literature explores FDI’s remainder coming mostly from middle- contribution to host economies through capi- income countries. tal infusion and productivity spillovers. A breakdown of country-level invest- Researchers theorize that MNE affiliates may ments, by source region, shows that the benefit local economies by introducing more respondents in surveyed countries in two advanced technology and management prac- regions—East Asia and Pacific, and Europe tices (especially to their suppliers), increasing and Central Asia—exhibit strong intrare- demand for local products, providing gional links (table 1.2). Roughly three- improved inputs, driving exports, and intro- quarters of respondents from these regions ducing competition (Arnold and Javorcik have parent MNEs based within the same 2009; Djankov and Hoekman 2000; Javorcik region, confirming high regional economic 2004; Lin and Saggi 2005; UNCTAD 2013; integration. In contrast, MNE affiliates in World Bank 2020). countries such as Brazil, India, Mexico, and At the same time, FDI may be harmful if Nigeria are more likely to have parent local firms lose market share to foreign- companies from other regions. backed competitors (Aitken and Many surveyed MNE affiliates are active Harrison 1999). From an empirical stand- in sourcing from local suppliers. On average, point, the literature has generally found posi- the GIC Survey respondents source tive upstream spillovers (that is, benefits of 55 percent of their inputs locally However, FDI for firms that are suppliers to MNE affili- there is significant cross-country variation, ates) and insignificant and sometimes nega- likely driven by the availability and quality tive spillovers for competitors of MNEs of local ­suppliers. In China and India, which (Havránek and Iršová 2011; Iršová and have well-­ d eveloped domestic product Havránek 2013). markets, MNE affiliates source a higher To assess the development contributions of share of their inputs locally (67 percent MNE affiliates in host economies, the 2019 and 60 percent, respectively), while respon- GIC Survey asked respondents about their dents in Turkey and Vietnam source less investments and reinvestment behavior, input than half of their inputs locally. In addition, sourcing, export and import activity, and ­ s ervices sector MNE affiliates, minority competition dynamics in host countries. foreign-owned affiliates, large employers The survey data suggest that MNE (more than 250 employees), and domestic affiliates make strong contributions to local ­ market-oriented affiliates (those with less economies through capital infusions. than half of revenues from exports) source a Roughly one-third of respondents have greater share of their inputs locally. 3 0    Global Investment Competitiveness Report 2019/2020 TABLE 1.2  Intraregional Investment Links Are Strong in East Asia and Pacific, and in Europe and Central Asia, While Interregional Investment Is Greater in Other Regions Share of respondents (percent) Source region Latin America Sub- Middle East East Asia and Europe and and the Saharan North and North Host country Pacific Central Asia Caribbean South Asia Africa America Africa China 58 25 1 1 0 15 0 Indonesia 77 17 0 1 0 4 1 Malaysia 64 22 1 2 0 11 1 Vietnam 86 13 0 0 0 1 0 Thailand 80 11 0 1 1 7 1 Turkey 10 77 1 0 1 8 4 Brazil 15 60 6 1 0 17 1 Mexico 15 38 9 1 0 37 0 India 21 51 1 1 2 23 2 Nigeria 9 33 0 8 26 12 11 Source: Computation based on the 2019 GIC Survey. Note: The figure excludes respondents who reported they did not know where their foreign parent multinational enterprises (MNEs) were based (less than 1 percent of the sample). Shaded cells denote investments originating from the same region as the host country (i.e., intraregional investment). From a trade perspective, many surveyed compared with just 16 percent that lost mar- MNE affiliates are active in driving exports. ket share (figure 1.6, panel a). In addition, Survey respondents derive about one-third of roughly two-thirds report primarily compet- their revenues from exports on average, and ing with firms operating in the host economy: nearly one-quarter are majority exporters (at 29 percent compete primarily with other least 50 percent of revenues derived from MNE affiliates, and 35 ­ percent compete with exports). In general, MNE affiliates in the ser- local firms (figure 1.6, panel b). These results vices sector derive a smaller share of their rev- suggest that MNEs apply competitive pres- enues from exports than do affiliates in the sure on domestic competitors, although the manufacturing sector (figure 1.5, panel a). net effect on domestic competitors’ produc- Lower tradability of certain ­ s ervices and tivity is unclear. services potentially higher barriers to trade in ­ explain these results. This pattern holds across most surveyed countries: in Vietnam, Policy Uncertainty and Foreign MNE affiliates in the manufacturing sector derive 74 percent of their revenues from Investors’ Outlook exports on average, compared with a modest The slowdown in FDI flows has come amid 29 percent for ­ s ervices affiliates. Similar rising policy uncertainty in trade and invest- trends are observed in China, Indonesia, ment. Policy uncertainty increases when the Malaysia, and Mexico. direction of policy decision making is Finally, survey data show that MNE affili- unclear or erratic, limiting businesses’ abil- ates promote competition in host countries. ity to forecast the likelihood of future events Owing to technological, financial, and man- and outcomes (Bloom 2014; Knight 1921). agement advantages, foreign affiliates per- The increase in policy uncertainty is form well on average in the surveyed reflected in the high values registered in countries: about half report having gained 2019 by various indicators such as the market share in the past three years, Global Economic Policy Uncertainty (EPU) O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    31 FIGURE 1.5  MNE Affiliates’ Trading Behavior Differs Widely across Host Countries a. Question: For the last financial year, what percentage b. Question: For the last financial year, what percentage of your company’s sales were exports? of your company’s inputs were imported? 42 48 All MNEs All MNEs 26 43 74 64 Vietnam Vietnam 29 52 68 63 Malaysia Mexico 26 52 54 53 Mexico Turkey 20 47 48 53 China Malaysia 21 42 43 51 Indonesia India 19 27 38 45 Thailand Brazil 40 38 32 43 Turkey Indonesia 29 44 30 41 India Nigeria 40 46 15 37 Brazil Thailand 14 41 14 29 Nigeria China 17 37 0 20 40 60 80 100 0 50 100 Share of respondents (%) Share of respondents (%) Manufacturing Services Source: Computation based on the 2019 GIC Survey. Index, World Uncertainty Index, World FDI rules (such as restrictive screenings Trade Uncertainty Index, and Trade Policy and approval requirements); and economic Uncertainty Index (Baker, Bloom, and Davis nationalism. The pattern of economic 2019; Caldara et al. 2019).5 nationalism extends to investment policy For MNE affiliates, both international as well as trade: more than a third of and domestic sources of policy uncertainty national investment policies introduced in could escalate risk sentiment.6 Recent and 2018 were measures related to new FDI ongoing global events such as withdrawals restrictions or regulations (UNCTAD from global trade agreements; new trade 2019). barriers (such as bilateral tariff escala- Business survey data enhance our under- tions); geopolitical developments (such as standing of investor behavior amid trade and Brexit); and other trade tensions have con- investment policy uncertainty. A growing tributed to an unprecedented rise in trade body of literature has investigated the and investment policy uncertainty (Baker, behavior of foreign investors in times Bloom, and Davis 2019). In host econo- of policy uncertainty more generally mies, MNE affiliates can further be subject (Bonaime, Gulen, and Ion 2018; Cao, Li, to uncertainties related to domestic politi- and Liu 2017; Chen and Funke 2003; Chen, cal and electoral outcomes; unpredictable Nie, and Ge 2019; Julio and Yook 2016; 3 2    Global Investment Competitiveness Report 2019/2020 FIGURE 1.6  The Competitive Dynamics of FDI Firms in Host Markets Are Complex a. Question: Considering the last 3 financial years, did your b. Question: In the last financial year, what was the company’s market share of its main product in this main source of competitive pressure for your country increase, decrease, or stay the same? company? All MNEs 16 50 All MNEs 35 29 22 8 5 India 5 66 China 55 18 20 3 4 Nigeria 21 57 India 44 23 22 8 3 Vietnam 12 55 Brazil 40 37 8 12 3 Indonesia 20 54 Nigeria 38 34 12 14 1 Turkey 17 53 Malaysia 33 15 36 7 7 Brazil 13 53 Thailand 32 32 24 3 9 Mexico 12 49 Turkey 31 26 27 12 4 China 17 39 Mexico 30 42 20 7 2 Malaysia 15 37 Indonesia 25 29 24 15 6 Thailand 28 36 Vietnam 24 31 30 5 11 100 50 0 50 100 0 20 40 60 80 100 Share of respondents (%) Share of respondents (%) Decreased Increased Domestic firms Foreign firms operating Foreign firms in other locally markets Imports Don’t know Source: Computation based on the 2019 GIC Survey. Note: In panel b, percentages may not total 100 because of rounding. FDI = foreign direct investment; MNEs = multinational enterprises. Rodrik 1991), but little research directly • What are foreign investors’ predominant examines the business effects of trade and investment plans in relation to expand- investment policy uncertainty on foreign- ing, maintaining, and reducing invest- owned firms.7 ments in host countries over the next The 2019 GIC Survey data provide direct three years? evidence of the effects of trade and invest- ment policy uncertainty on business opera- tions of foreign-owned firms in the 10 MICs. Effects of Policy Uncertainty on By capturing current perceptions and expec- Foreign-Owned Businesses tations of foreign investors with investments Survey results show that foreign-owned in the surveyed MICs, the survey addresses firms are sensitive to recent increases in pol- the following questions related to policy icy uncertainty due to protectionism and uncertainty and investments: economic nationalism in trade and invest- ment. Nearly two-thirds of respondents • How important is the increasing policy report that such policy uncertainty was uncertainty due to protectionism and “important” or “critically important” to economic nationalism in trade and invest- their investment decisions in the last year ment in shaping foreign companies’ (figure 1.7, panel a).8 The survey data indi- investment decisions? cate that, on average, MNE affiliates that • How has policy uncertainty affected import a greater share of their inputs and foreign-owned businesses in terms of those that employ more than 250 workers jobs, productivity, investments, and (large firms) are more sensitive to policy changes in production locations? O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    33 FIGURE 1.7  Investors Are Sensitive to Policy Uncertainty in Trade and Investment and Have Been Adversely Affected in the Last Year a. Question: In the past financial year, how important was rising policy uncertainty due to protectionism and economic nationalism in trade and investment for your company’s investment decisions in this country? All MNEs 29 35 Critically important Important 0 20 40 60 80 100 Share of respondents (%) The 29 percent of MNE affiliates that considered policy uncertainty in trade and investment to be “critically important” were asked the following question: b. Question: In the last financial year, what impact has rising policy uncertainty due to protectionism and economic nationalism in trade and investment had on your company’s operations in this country?a Jobs 31 Among these respondents, 51 percent experienced a Productivity 34 decline in at least one category—jobs, productivity, or investment, in Investment 35 the past year. 0 5 10 15 20 25 30 35 40 Share of firms reporting a decrease (%) Source: Computation based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. MNEs = multinational enterprises. a. Questions about the impact of rising policy uncertainty on business operations were only asked of respondents who reported rising policy uncertainty as “critically important” in the past financial year (a 12-month period between January 1, 2018, and September 30, 2019, depending on the country). uncertainty (both differences being signifi- MNE affiliates reported being adversely cant at p < 0.01).9 affected by policy uncertainty (figure 1.7, To assess the direct impact of policy panel b). About a third of respondents to uncertainty on foreign-owned firms, MNE this question reported declines in the affiliates that consider policy uncertainty number of jobs (31 percent), productivity to be “critically important” were also (34 percent), and investment (35 percent) in asked whether policy uncertainty has the last year. Taken together, over half caused increases, decreases, or no impact in (51 percent) the respondents have terms of the number of jobs, productivity, experienced a decline in jobs, productivity, investments, and changes in production or investments owing to rising policy locations (such as global supply chain uncertainty in the past financial year. About adjustments). A substantial share of these a third (32 percent) reported positive 3 4    Global Investment Competitiveness Report 2019/2020 impacts along at least one outcome (and no Theoretical explanations posit that an negative impact on any outcome). The increase in uncertainty increases the option remaining 17 percent reported no effect or value of delaying investments when faced did not know. with adjustment costs, resulting in declines The increase in trade and investment in both investments and new hiring (Abel, policy uncertainty could disrupt existing Dixit, and Eberly 1996; Abel and Eberly global value chains, reflecting MNEs’ 1996; Bernanke 1983). These theoretical gradual shift toward reshoring or other predictions consistently find support in changes to the locations of production the empirical literature (Anand and Tulin (World Bank 2020). Thus, the survey 2014; Baker, Bloom, and Davis 2013; Bloom asked the MNE affiliates that considered et al. 2012; Cebreros, Heffner, and Salcedo policy uncertainty to be “critically 2019; Gulen and Ion 2016). In turn, stalled important” whether their decision making investment activity impedes productivity about production locations had changed growth because of limited reallocation across as a result of policy uncertainty in the companies (Bloom 2009). In the face of last financial year. Over 40 percent of trade and investment policy uncertainty, respondents in China, Indonesia, Mexico, MNEs have incentives to revisit global Nigeria, and Thailand have adjusted production and sourcing decisions to avert how they organize their supply chains as vulnerabilities from possible supply chain a result of policy uncertainty. In other dislocations (Blanchard 2019; IMF 2019). countries, the results are less pronounced: Firms’ cautious investment behavior can less than one-third of respondents in thus curtail global economic activity and Brazil, India, and Malaysia reported slow growth (Caldara et al. 2019). adjusting their supply chains because of To assess the investment outlook of policy uncertainty. MNEs operating in the surveyed economies, Policy uncertainty is most likely to have their affiliates were asked about their affected the configuration of supply predominant investment plans in relation to chains and adjustments of production expanding, maintaining, and reducing locations among the smaller MNE affili- investments in host countries over the next ates . Among those that are SMEs (with three years.11 Figure 1.8 presents aggregate fewer than 250 employees), 40 percent results for the 10-country pooled sample. report having adjusted how they organize Fewer than half (48 percent) are planning to their production locations, compared with expand their investments over the next three 30 percent of large firms. 10 This pattern years. A similar share of respondents may reflect the larger firms’ greater capacity (44 percent) across both manufacturing and to weather policy-related challenges as well services sectors are planning not to invest as the smaller affiliates’ greater organiza- further, keeping their investment stock at tional agility in adjusting production current levels in respective host economies. locations. A small share of existing respondents (4 percent) are planning to withdraw or reduce their investments. Predominant Investment Plans MNE affiliates that are more exposed to The nature of FDI (being partially sunk policy uncertainty exhibit more cautious after the investment is made) renders it par- investment outlooks. In line with the litera- ticularly vulnerable to trade and investment ture, survey data suggest that policy policy uncertainty. Adopting a “wait and uncertainty in trade and investment is poised see” approach, cautious firms delay or to shift investment patterns. Firms that cite cancel planned investments and technologi- policy uncertainty in trade and investment cal upgrades. as a “critically important” investment O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    35 FIGURE 1.8  Larger Firms and Those from Developing Countries Are More Likely to Expand Investment, on Average Question: Over the next three years, what is your company’s predominant investment plan in this country? 1 All MNEs 48 44 3 4 2 2 > US$50 million 57 36 3 1 1 Investment stock*** US$10–50 million 49 47 3 1 2 US$5–10 million 53 39 4 1 < US$5 million 46 46 3 3 1 High-income country 46 46 3 4 HQ country** 2 Developing country 63 27 5 3 1 > 250 employees 50 41 2 5 Workforce* 1 ≤ 250 employees 47 45 3 4 0 50 100 Share of respondents (%) Expand Retain Reduce Withdraw Don’t know Source: Computation based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises (MNEs) were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. “Developing countries” refers to low- and middle-income countries as defined by the World Bank. Differences for investment stock were tested by comparing the “ > US$10 million” group with the rest of the sample. Percentages may not total 100 because of rounding. The inter- views were conducted June–November 2019, so the implied three-year time horizon for this question is 2019–22. HQ = headquarters. *p < .10 **p < .05 ***p < .01 consideration are more than twice as likely as investment consideration expect to be less other affiliates (7 percent versus 3 percent) to expansionary on average.12 Relatedly, those reduce or withdraw their investments. Even respondents that experienced reductions in after controlling for host country and various either workforce, productivity, or investments firm characteristics, MNE affiliates that cite as a result of policy uncertainty are nearly policy uncertainty as a “critically important” four times more likely to expect to reduce or 3 6    Global Investment Competitiveness Report 2019/2020 withdraw investments in the next three years more likely to plan on expanding their (9.2 percent versus 2.4 percent).13 investments. Nearly two-thirds of affiliates Larger MNE affiliates (by size of workforce with global headquarters in other developing or investment) are more positive in their countries plan to expand their investments investment outlook. Foreign-owned firms over the next three years. The factors driving with more than US$10 million in investment this result could include a higher appetite for stock are nearly 10 percentage points more risk and policy uncertainty among investors likely to expand investments in the next three from developing countries as well as current years.14 Similarly, foreign-owned firms with trade tensions being concentrated between more than 250 employees are also more likely developed and developing countries (Beamish to expand and less likely to reduce investment and Banks 1987; Gonzalez, Qiang, and Kusek over the next three years. 15 Results are 2018; Wei, Liu, and Wang 2008). consistent with literature that suggests that MNE affiliates report heterogeneous future firms with greater financial resources are investment plans across the surveyed countries. better equipped to weather uncertainty if they Figure 1.9 presents disaggregated results by continue to believe that the long-term country for the question on predominant fundamentals of their investments are investment plans in relation to expanding, attractive (Ghosal and Loungani 2000). maintaining, and reducing investments in MNE affiliates with parents from other host countries over the next three years. In developing countries are also significantly terms of expansion, foreign affiliates in China FIGURE 1.9  Investors in China and Turkey Have Less Ambitious Investment Plans than Those in Other Countries Question: Over the next three years, what is your company’s predominant investment plan in this country? Nigeria 81 11 5 21 0 India 64 33 12 0 0 Indonesia 57 40 3 Brazil 49 41 4 2 4 Share of respondents (%) Vietnam 46 45 1 8 Mexico 45 52 121 1 Thailand 43 45 6 5 Malaysia 40 49 13 7 Turkey 35 50 8 2 6 0 China 17 73 5 4 0 10 20 30 40 50 60 70 80 90 100 Expand Retain Reduce Withdraw Don’t know Source: Computation based on the 2019 GIC Survey. Note: For sample composition, see annex 1A. Percentages may not total 100 because of rounding. The interviews were conducted June–November 2019, so the implied three-year time horizon for this question is 2019–22. O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    37 (17 percent) and Turkey (35 percent) are sensitivity to policy uncertainty in trade and much less likely to expand than those in other investment or average investment outlook. surveyed countries. In stark contrast, about This relative homogeneity suggests that four-fifths of foreign affiliates in Nigeria and foreign-owned firms’ relative pessimism in two-thirds in India plan to expand their Turkey is driven by domestic macroeconomic investment stocks over the next three years. uncertainty. The Turkish economy has The effect of policy uncertainty in trade and recently suffered from sharp financial out- investment—combined with domestic factors flows driven by concerns related to high such as macroeconomic fundamentals, current account deficits, high corporate political developments, and the legal and indebtedness, and the direction of domes- regulatory environment, among others—is tic economic policy (World Bank 2019b). likely to shape foreign companies’ investment Such issues are likely to affect all MNE affili- plans in the surveyed countries. ates in the country. In China, both efficiency-seeking and In contrast to China and Turkey, more market-seeking investors report similar than 80 percent of the surveyed MNE and relatively pessimistic investment out- affiliates in Nigeria plan to expand their looks for the next three years. However, investment stocks in the next three years. the future investment plans for these two This expansionary outlook holds regardless subgroups are likely shaped by different of the affiliates’ sector, export intensity, and factors. primary export destination, suggesting that Efficiency-seeking investors (majority economy-wide factors are driving investors’ exporters) that primarily export to the relative optimism. United States are about 15 percentage points Survey results suggest that Nigeria may more likely than those with other primary experience a strengthening in investor export destinations to consider policy confidence following a period of declining uncertainty to be “important” or “critically FDI inflows. The positive investment out- important” in their investment decisions.16 look observed in the survey is also supported Plausibly, the investment sentiment of by investment forecasts for the next three efficiency-seeking investors reflects the years. The projected investment growth rate detrimental effect of ongoing trade tensions is higher in Nigeria than in all other surveyed with the United States. Although investment countries (see annex 1B). In recent years, flows to China have not slowed to date FDI inflows to Nigeria have declined because (UNCTAD 2019), the relative pessimism of of factors such as falling commodity prices, affiliates in China suggests that changes in uncertainty regarding elections, new trade patterns may lead to investment diver- regulations establishing local content sion in the near future. requirements, restrictions on visas for expa- In contrast, the subdued future investment triate workers, and disputes between the plans for domestic market-seeking investors government and foreign investors related to are plausibly driven by domestic factors, repatriation of profits and taxation including high levels of corporate indebted- (UNCTAD 2019; World Bank 2019b). ness (a future growth risk) and an overall Foreign affiliates’ positive outlook for invest- slowdown in China’s economic growth ment over the next three years reflects (World Bank 2019b). developments including political stability In Turkey, MNE affiliates with different after completion of the 2019 election cycle, characteristics exhibit similar future invest- a marked improvement in the ease of doing ment plans: on average, only 35 percent plan business as a result of business environment to expand. There are no significant differences reforms, and a new policy to reduce public across major dimensions of interest ownership in joint-venture oil assets (manufacturing and services, efficiency seeking (UNCTAD 2019; World Bank 2019a). and market seeking, large and small affiliates, However, the longer-term outlook may older and newer affiliates) in terms of either depend on further strengthening of the 3 8    Global Investment Competitiveness Report 2019/2020 country’s economic governance framework important” for investments, a credible policy (World Bank 2019b). response should enhance predictability and In India, most of the MNE affiliates— investor confidence. about two-thirds—also plan to expand their investment stocks in the next three years. Key Role of Political, Macroeconomic, They display no significant differences in and Regulatory Environments either sensitivity to policy uncertainty in trade and investment or future investment The survey results show that nearly 9 in plans across most major respondent 10 respondents consider political stability, characteristics. macroeconomic stability, and a country’s The overarching trend observed across legal and regulatory environment to be major respondent categories (manufacturing “important” or “critically important” and services, efficiency seeking and market for investment decisions (figure 1.10). These seeking, large and small affiliates, older and factors rank ahead of considerations such as newer affiliates) suggests that foreign low tax rates, low labor and input costs, and investors are likely responding to a base access to resource endowments. set of supportive economic fundamentals The findings are consistent with the in India. A variety of positive factors can 2017/2018 GIC Survey and extant empirical plausibly explain their robust investment literature showing that higher instability outlook, including an accommodating related to the political and macroeconomic monetary policy that has supported credit environment17 imposes additional transac- growth and policy stability as a result of tion costs and risks for businesses. It thus political continuity. Investor confidence plays a critical role in shaping long-term further stands to gain from the country’s investment decisions. Empirical research strong performance on business regulatory shows that there is significant negative effect reforms and a streamlined nationwide goods of the resulting risk and uncertainty on FDI and services tax (GST) regime, among other inflows (Asiedu 2006; Busse and Hefeker factors (IMF 2019; Kazmin 2019; World 2007; Jun and Singh 1996; Krifa-Schneider Bank 2019a, 2019b). and Matei 2010; Schneider and Frey 1985; Sekkat and Veganzones-Varoudakis 2007; Walch and Wörz 2012). Importance of Predictability for Relatedly, a transparent and predictable regulatory environment is crucial for Foreign Investment attracting new investment as well as for Evidence presented in the preceding section retaining existing foreign investors. A large suggests that policy uncertainty has adversely body of research suggests that the quality of affected many investors. These impacts, a country’s legal and regulatory environment combined with the looming threat of a is positively associated with FDI (Akame, synchronized global economic slowdown, Ekwelle, and Njei 2016; Buchanan, Le, and mean that competition between countries for Rishi 2012; Daude and Stein 2007; Gani FDI is likely to further intensify. 2007; Globerman and Shapiro 2002; This raises important questions for host- Vogiatzoglou 2016; Wei 2000; Wernick, country policy makers. Crucially, what can Haar, and Singh 2009). Furthermore, developing countries do to inspire investor evidence from previous investor surveys rein- confidence, counter prevailing global forces the claim that a supportive business headwinds and policy uncertainty, and climate is among the top priorities for foreign leverage FDI for their development objectives? investors (A.T. Kearney 2019; Kusek and With nearly two-thirds of the 2019 GIC Silva 2018). Survey respondents considering policy uncer- Countries’ legal and regulatory environ- tainty to be “important” or “critically ments are especially important for larger O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    39 FIGURE 1.10  Legal and Regulatory Environment Is among the Top Three Investment Decision Factors Question: How important were the following factors in your parent company’s decision to invest in this country? Political stability 49.4 34.9 Macroeconomic stability 49.0 35.7 Legal and regulatory environment 42.0 42.1 Talent/skills 41.5 43.5 Low taxes 41.3 35.5 Market size 39.3 42.6 Physical infrastructure 38.6 40.8 Ability to export 38.4 31.7 Intellectual property protections 35.4 35.4 Investor protections 34.5 37.5 Low labor and input costs 33.2 41.4 Supply chain coordination 30.1 44.2 Local input sourcing 26.6 42.2 Resource endowments 18.5 31.7 Local acquisition targets 14.4 35.6 0 20 40 60 80 100 Share of respondents (%) Critically important Important Source: Computation based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises (MNEs) were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. firms. On average, large firms (those with Cumbersome investment approval processes more than 250 employees) rank the legal and and operational restrictions are the most- regulatory environment as their top invest- cited regulatory barriers for FDI in the sur- ment consideration, while smaller affiliates veyed MICs. consider it to be only the fourth most impor- Respondents most commonly identify tant consideration.18 These differences may investment approval processes as a key issue, be driven by investment restrictions that are with 56 percent listing them as “moderate” applicable only to larger firms and by the or “major” obstacles to operations. On greater regulatory scrutiny that large firms average, MNE affiliates need more than two tend to experience. months (64 days) to obtain such approvals, but times vary widely across countries and types of investment—and 10 percent of affil- Investment Policy and Regulatory iates report wait times of five months or Regimes that Enable FDI more. Restrictions on prices, technology, or The 2019 GIC Survey assessed which spe- products are another key barrier, with cific investment policy and regulatory obsta- 44 percent of respondents citing them as cles hinder an enabling regime for FDI. “moderate” or “major” obstacles. The 4 0    Global Investment Competitiveness Report 2019/2020 salience of these top two concerns holds investments to have experienced higher legal across most countries and sectors. In and regulatory obstacles in investment addition, these findings are consistent with approvals, local sourcing requirements, prior work that also finds that investment research and development (R&D) invest- approval processes and restrictions on prices, ment requirements, minimum investment technology, or products can be significant requirements, and expatriate staff restric- obstacles for foreign affiliates (Mistura and tions. For example, 35 percent of respon- Roulet 2019; UNCTAD 2019). dents planning to reduce or withdraw The survey data also show that firms plan- investments cite investment approvals as a ning to reduce or withdraw investments in “major” obstacle, compared with just the next three years are more likely than 26 percent of other respondents (figure 1.11). those planning to retain or expand These stark differences suggest that legal and FIGURE 1.11  MNE Affiliates That Experience Legal and Regulatory Obstacles Are More Likely to Reduce or Withdraw Investments in the Future Question: To what degree are the following factors an obstacle for your company to operate in this country? Planning to retain or expand investment 26 30 Investment approvals** Planning to reduce or withdraw investment 35 30 Planning to retain or expand investment 16 29 Price, technology, or product restrictions Planning to reduce or withdraw investment 22 23 Planning to retain or expand investment 15 22 Foreign investment limits Planning to reduce or withdraw investment 18 22 Planning to retain or expand investment 12 24 Expatriate restrictions Planning to reduce or withdraw investment 23 20 Planning to retain or expand investment 15 20 Joint venture requirements** Planning to reduce or withdraw investment 17 27 Planning to retain or expand investment 12 21 Local sourcing requirements Planning to reduce or withdraw investment 17 18 Planning to retain or expand investment 10 20 Research and development investment requirements Planning to reduce or withdraw investment 16 24 Planning to retain or expand investment 8 20 Minimum investment requirements Planning to reduce or withdraw investment 11 13 0 20 40 60 80 100 Share of respondents (%) Major obstacle Moderate obstacle Source: Computation based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises (MNEs) were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. The interviews were conducted June–November 2019, and the implied future time horizon for this question is a three-year period: 2019–22. **p < .05 O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    41 regulatory barriers play a key role in MNE MNE affiliates’ operations. In contrast, affiliates’ investment decisions. ­ limits on the amount of ­ foreign investment Notably, some factors that are only rarely are relatively bigger concerns for affiliates in considered major obstacles by most MNE China, India, Indonesia, Thailand, and affiliates (such as restrictions on expatriate Turkey. In Vietnam, local sourcing require- staff) rank among the top obstacles for affili- ments and restrictions on hiring expatriate ates looking to reduce or withdraw their staff routinely hold back affiliates’ investments. Some of these issues may matter operations. a great deal to a subset of affiliates, while Recent literature has shed light on others may be relatively rare but important how businesses navigate the regu- when they do arise. latory environment in developing Table 1.3 disaggregates the top three countries and the divergence between regu- legal and regulatory obstacles, by country. latory provisions and their implementation Cumbersome investment approvals to start (Freund, Hallward-Driemeier, and Rijkers and operate a business are the top-cited 2014; Hallward-Driemeier and Pritchett obstacle in most countries and rank in the 2015). To assess factors that contribute to top two in all surveyed MICs. Restrictions obstacles for foreign-owned firms in the sur- on setting prices, production technology, veyed countries, the survey asked respon- or the format of products also rank in the dents about specific aspects of government top three in all surveyed MICs except conduct related to the quality and implemen- Vietnam. tation of investment rules. MNE affiliates Other top-three obstacles vary by coun- cite both the quality of laws (rules and regu- try: for example, affiliates in Brazil and lations) and challenges in their implementa- Mexico cite those countries’ r ­ elatively strin- tion as contributing to their obstacles in the gent joint venture ­requirements as hindering surveyed countries (figure 1.12). TABLE 1.3  Investment Approval Processes Are the Top Most-Cited Obstacle across Surveyed Countries, Followed by Price, Technology, or Product Restrictions Country Most-cited obstacle Second most-cited obstacle Third most-cited obstacle Brazil Investment approvals Price, technology, or product JV requirements restrictions China Investment approvals Price, technology, or product Foreign investment limits restrictions India Investment approvals Foreign investment limits Price, technology, or product restrictions Indonesia Investment approvals Price, technology, or product Foreign investment limits restrictions Malaysia Expatriate restrictions Investment approvals Price, technology, or product restrictions Mexico Investment approvals Price, technology, or product JV requirements restrictions Nigeria Investment approvals Expatriate restrictions Price, technology, or product restrictions Thailand Investment approvals Price, technology, or product Foreign investment limits restrictions Turkey Investment approvals Price, technology, or product Foreign investment limits restrictions Vietnam Investment approvals Local sourcing requirements Expatriate restrictions Source: Computation based on the 2019 GIC Survey. Note: Rankings are based on frequency counts. JV = joint venture. 4 2    Global Investment Competitiveness Report 2019/2020 FIGURE 1.12   Foreign-Owned Firms Perceive the Quality of Rules and Their Implementation as Obstacles in Government Conduct Question: To what degree are the following factors an obstacle for your company to operate in this country? Administrative complexity 29 35 Bureaucratic discretion 30 33 Quality of laws 23 37 Interagency coordination 22 33 Public agency capacity 21 32 Accessibility of laws 17 29 0 10 20 30 40 50 60 70 80 90 100 Share of respondents (%) Major obstacle Moderate obstacle Source: Computation based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. Major drivers of legal and regulatory curtailed, such uncertainty can present a sig- challenges include the complexity of admin- nificant threat for the global economy. istrative procedures, discretion exercised by Survey data suggest that policy uncertainty the bureaucracy, and the quality of laws and has already adversely affected many inves- regulations. In other words, both the sub- tors and could drive future investment stantive content of laws and the way in slowdowns. Over the next three years, most which they are implemented contribute to investors in the surveyed MICs are not legal and regulatory obstacles in the surveyed planning to expand their investment stock. countries. Given the demonstrated contributions of Outside of the main findings of this report, FDI to host economies in terms of capital the 2019 GIC survey also revealed important infusion, technology transfer, and linkages differences between various categories of to global and local value chains, policy MNE affiliates (box 1.1). uncertainty poses a serious threat to both short-term growth and long-term structural transformation. Conclusions and Policy In this environment of uncertainty, govern- ments of developing countries can neverthe- Implications less take steps to bolster FDI by strengthening High levels of international and domestic their investment competitiveness. Current policy uncertainty in trade and investment insights from the 2019 GIC Survey highlight have emerged as an investment risk. If not several policy priorities: O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    43 BOX 1.1 Key Findings of the 2019 GIC Survey Results, by Foreign Investor Type The 2019 GIC survey revealed some important • Majority exporters use and value investment incen- differences between various categories of multina- tives more, on average. tional enterprise (MNE) affiliates.a Some of the results listed below concern topics addressed in the survey but Differences by Origin Country (Developed not covered in detail in this chapter (such as incen- versus Developing) tives and investment promotion agencies). These topics • Investors from developing countries are more likely are addressed in detail in subsequent chapters of this to plan on expanding investments over the next report. three years. Differences by Sector (Manufacturing versus • Investors from developing countries value invest- Services) ment incentives and investment promotion agency • Manufacturers are more sensitive than services (IPA) services more than investors from developed affiliates to the availability and costs of local countries do, on average. inputs. • Services affiliates consider joint venture require- Differences by Size (Firm Employment Level) ments and limits on foreign investment to be larger • Large employers (those with more than 250 employ- obstacles, on average, than do manufacturers. ees), on average, place greater importance on the • Manufacturers tend to use and value investment legal and regulatory environments of their host incentives more than services affiliates do. countries. Differences by Degree of Export Activity • Smaller employers, on average, encounter more legal and regulatory obstacles, particularly concerning Several findings showed differences between “major- investment approvals, joint venture requirements, ity exporters” (MNE affiliates whose revenues come research and development (R&D) investment require- mostly from exports) and affiliates with less than half ments, and minimum investment requirements. of revenues from exports: • Large employers receive more fiscal or financial • Majority exporters are less sensitive to local mar- incentives, on average, but do not necessarily value ket size and stability but are more sensitive to input them more than smaller firms do. cost, availability, and quality. • Majority exporters consider local sourcing require- a. Unlike other analyses in this chapter, differences presented in this box are from logistic and ordered logistic regressions without controls, because they are meant ments and restrictions on hiring expatriate staff to primarily for descriptive purposes. They should not be interpreted as indicative of be bigger obstacles, on average. causal impacts. 1. Counter international policy uncertainty a. Uphold the multil ate ral tra ding by reaffirming commitments to global and ­system. Honoring existing global and regional trade systems. Survey data consis- regional trade and investment agree- tently demonstrate how rising policy ments would improve the government’s uncertainty due to protectionism and credibility and commitment regard- economic nationalism in trade and ing the course of future policy. In the investment weighs on investor sentiment. absence of such commitments, foreign- To counter such uncertainty, policy mak- owned firms may be hesitant to invest ers should signal and follow through on if they think existing rules may not be their commitments to multilateral and honored in the future. regional trade and investment arrange- b. C o n t i n u e t ra d e a n d e c o n o m i c ments in several ways: liberalization . The continuation of 4 4    Global Investment Competitiveness Report 2019/2020 ongoing trade and investment liberal- factor among MNE affiliates. Underlying ization efforts through multilateral and inflationary pressures are still present in regional mechanisms would increase many low- and middle-income countries, investor confidence by setting expecta- as are risks of short-term capital outflows. tions of further reductions in trade and To counter those pressures, governments investment restrictions in the future. In should pursue the following measures: parallel, governments should continue to a. Implement macroprudential policies. reduce economic distortions (for exam- Policies such as countercyclical capital ple, from subsidies) through domestic buffers and limits on foreign currency reforms, which would both facilitate borrowing can help limit exposure to domestic growth and contribute to the future currency, interest rate, or debt resolution of ongoing trade disputes. rollover shocks. c. Clearly communicate policy directions. b. Ensure central bank independence. Clearly communicating changes in Insulating central banks from political trade and investment policy in advance interference would help establish cur- would reduce investors’ uncertainty on rency and interest rate credibility. the direction of future policy. c. Optimize fiscal policy. From a fiscal 2. Promote political stability, strong perspective, individual country situa- institutions, and good governance. On tions vary widely. In general, however, average, MNE affiliates cited political sta- countries would do well to preserve bility as their top investment decision- growth-enhancing spending and tax making factor. Indeed, research highlights reforms while ensuring fiscal space how potential reescalation of conflicts, through fiscal consolidation, broaden- electoral violence, and political turbulence ing of the tax base, and strengthening pose risks to many countries’ economic of tax administration. Such measures outlooks. Hence, policy makers should would help keep debt-to-GDP ratios focus on these overarching goals: manageable while limiting adverse a. Strengthen institutions. Political and effects on economic growth. economic institutions establish “rules 4. Improve the legal and regulatory frame- of the game” by promoting, among work for FDI. Foreign investors consis- other things, openness, transparency, tently identify the legal and regulatory and stability. Strengthening institu- environments for FDI in host countries as tions to ensure peaceful transitions of being critical considerations for their power and some degree of continuity investment decisions. Indeed, these issues in structures of governance and policy are even more important to the large firms making thus contributes to greater pre- that disproportionately contribute to dictability and investor confidence. employment growth in host countries. b. Ensure fair governance. Establishing Policy makers should thus remain commit- capacity rules and building institutional ­ ted to fair market access for foreign firms to ensure a level playing field for inves- while removing critical administrative tors and to eliminate political favorit- barriers to investment, as follows: ism toward specific businesses is also a. Remain committed to fair access. crucial for attracting investments. Policy makers should resist the temp- A level playing field ensures that the tation to engage in protectionism and most efficient MNEs have adequate economic nationalism in their own incentives to invest, helping to maxi- markets. This entails minimizing mize benefits from FDI. foreign investment limits, excessive 3. Optimize macroeconomic policy . and ­d iscretionary foreign investment Macroeconomic stability is the second screening, and discrimination against most-cited investment decision-making foreign firms. O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    45 b. Remove critical administrative barriers. vulnerabilities. The prospects for recovery Across all countries, survey respondents rest on the breadth and depth of policy sup- consistently cite cumbersome invest- port extended to MNEs in the face of the ment approval processes as well as extraordinary global shock. restrictions on pricing, technology, or Finally, the severity of the pandemic under- products as key obstacles to their oper- scores the need for timely policy insights. ations. ­G overnments of developing A responsive policy research agenda should countries should thus invest in mak- seek to fill knowledge gaps (for example, esti- ing approval processes more efficient mate the effects on markets, businesses, and and optimize operational regulations workers) and enable the design of policy mea- to minimize adverse impacts on busi- sures that increase the resilience of MNEs to ness operations without sacrificing the shocks and preserve their viability. regulations’ original policy objectives. c. Focus on improving both the imple- mentation and quality of laws. ­Survey Annex 1A. Survey and Data data suggest that both suboptimal Analysis Methodology design of regulations and the ways in The data used in this study are from the 2019 which they are enforced contribute to Global Investment Competitiveness (GIC) operational ­obstacles. Thus, although Survey, which captures the experiences and improving the regulations should perceptions of multinational enterprise remain a priority, governments should (MNE) affiliates on global megatrends and also work to streamline procedures investment climate factors in 10 middle- and clarify roles to limit administrative income countries (MICs). The survey involved complexity and bureaucratic discretion. interviewing senior executives of foreign- owned firms who possess a broad understand- The COVID-19 pandemic reinforces the ing of their companies’ business strategies, importance of these policy priorities, espe- policy barriers, operational obstacles, and cially measures to build investor confidence. investments in the host economy. Against the backdrop of heightened policy The survey complements other investor uncertainty in trade and investment, the pan- surveys by focusing on investment climate vari- ­ demic will further escalate uncertainty, mag- ables, such as administrative and legal barriers, nify investment risks, and depress foreign rather than on broader economywide factors. investor confidence. With higher dependence These specific investment climate variables are on imports of intermediate goods and actionable areas for policy makers. broader exposure to export markets, MNEs The survey comprised four sections: are particularly vulnerable to supply and demand shocks induced by the pandemic. 1. General information on the company, Although large-scale impacts are already including sector, number of employees, observable, the full extent and duration of the the total investment stock to date, and effects of the pandemic remain uncertain. predominant investment plan over the Beyond such medium- to long-term mea- next three years in the host country. sures, the extraordinary challenges associated 2. Importance and effect of global megatrends with the pandemic also warrant crisis on the company’s business operations— management measures by governments. The ­ including on jobs, productivity, investments, pandemic represents an unprecedented shock and changes in location of production—in to the global economy, and the economic fall- the last year. out for MNEs is expected to be very high 3. Foreign-owned firms’ contribution to the (IMF 2020; UNCTAD 2020). Support to host economy through reinvestments, local MNEs should be deployed rapidly, benefit a sourcing, and pro-competition effects, as broad cross-section, and respond to pressing well as foreign-owned firms’ integration in 4 6    Global Investment Competitiveness Report 2019/2020 global value chains through estimates of TABLE 1A.1  2019 GIC Survey Sample, by Country imports and exports. and Sector 4.  Importance of investment policy factors Manufacturing Services MNE and operational obstacles faced by the Country MNE affiliates affiliates Total foreign-owned firms’ affiliates, including Brazil 125 125 250 investment restrictions, services offered by China 125 125 250 investment promotion agencies (IPAs), tax India 125 125 250 and financial incentives, and investment Indonesia 133 125 258 protection guarantees. Malaysia 125 125 250 Mexico 124 125 249 The survey was designed to generate Nigeria 55 109 164 results that are representative at the country Thailand 125 125 250 level and comparable across countries. An Turkey 125 125 250 assessment of changes in affiliate experience Vietnam 128 125 253 and perceptions over time will be possible Total 1,190 1,234 2,424 with a second wave of data collection in Source: Computation based on the 2019 GIC Survey. 2020–21. To the extent possible, the second Note: MNE = multinational enterprise. round will target foreign-owned businesses from the first round. select affiliates were contacted before the target was reached. Nonresponse bias can occur when those Sample Representation who respond to the survey are systematically The survey represents experiences and different from nonrespondents in terms of perceptions of a representative sample of basic characteristics. The likelihood of non- foreign-owned firms in each of 10 MICs: response bias in this survey is minimal because Brazil, China, India, Indonesia, Malaysia, no systematic differences were found when Mexico, Nigeria, Thailand, Turkey, and respondents and nonrespondents were com- Vietnam. Each country sample comprises pared based on observed characteristics (such roughly 250 foreign-owned firms with at least as sectoral affiliation and country of origin). five employees.19 In each country, roughly To address any possibility of non­ response bias 125 respondent firms operate in the manu- due to target respondents’ varying willingness facturing sector, and roughly 125 respon- or ability to respond to the survey, data were dent firms operate in the services sector weighted for non­ r esponse. This did not (tables 1A.1 and 1A.2).20 The only exception change the results and findings derived from is Nigeria, where because of sampling frame the survey and presented in the chapter. limitations, the sample comprises 164 respon- To ensure representativeness, analyses con- dents (55 manufacturing and 109 services). tained in the chapter incorporate weights to Thus, across the 10 target countries, more account for different sample sizes across than 2,400 responses were collected. countries, different probabilities of sampling, Sampling frames comprising partially or and bias due to nonresponse. Design weights fully foreign-owned businesses in the 10 MICs have been included to ensure that the differ- were constructed using commercially available ent strata (country-sector intersections) are and proprietary sources (Dunn & Bradstreet, given equal weight. Sampling weights were Orbis/Bureau van Dijk, Sample Solutions, and included to account for different probabilities others). The sampling frame sizes by country of being sampled, weighting each observation are presented in table 1A.3. The frames were by the inverse probability of selection. Finally, de-duplicated and cleaned, and data quality nonresponse weights are applied to maintain was enhanced using standard sample framing consistency between the distribution of MNE and data manipulation techniques. In some affiliates in the sampling frame and results sampling frames, all affiliates were contacted from the sample along observable characteris- to reach the target sample size. In others, only tics To check the robustness of results in this O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    47 TABLE 1A.2  Share of 2019 GIC Survey Respondents, by Subsector Sector and subsector N Share of total sample (%) Manufacturing 1,190 49.1 Machinery and equipment 151 6.2 Metals and metal products 124 5.1 Automobiles, other motor vehicles, and transport equipment 116 4.8 Rubber and plastic products 108 4.5 Chemicals and chemical products 84 3.5 Agroprocessing, food products, and beverages 72 3.0 Information technology and telecommunications 71 2.9 Electrical and electronic equipment and components 52 2.1 Textiles, apparel, and leather 51 2.1 Wood products, paper, and printing 45 1.9 Pharmaceuticals, biotechnology, and medical devices 18 0.7 Refined petroleum products, coke, and nuclear fuel 8 0.3 Manufacturing: Other or unclassified 290 12.0 Services 1,234 50.9 Wholesale and retail trade 270 11.1 Business services 116 4.8 Logistics, transport, and storage 101 4.2 Computer and software services 85 3.5 Administrative and support services 59 2.4 Construction 58 2.4 Financial services, including insurance 49 2.0 Electricity, gas, and water 32 1.3 Other professional, scientific, and technical services 20 0.8 Real estate 18 0.7 Hotels, restaurants, and tourism 16 0.7 Telecommunications 13 0.5 Media 9 0.4 Health 7 0.3 Arts and recreation 6 0.2 Education 5 0.2 Scientific research and development services 5 0.2 Water supply and waste management 3 0.1 Residential care and social work 2 0.1 Public administration and defense services 1 0.0 Services: Other or unclassified 359 14.8 Source: Computation based on the 2019 GIC Survey. TABLE 1A.3  Sampling Frame Sizes, by Country chapter to different weighting approaches, all Country Total number of affiliates analyses in this report were also run without weights and with sampling weights only. The Brazil 5,007 results in this chapter were found to be robust China 15,668 to these changes in weighting approach. India 9,120 Indonesia 4,153 Malaysia 5,673 Survey Administration Mexico 7,992 Nigeria 7,089 The World Bank Group commissioned an Thailand 9,789 international survey firm, Kantar Public, to Turkey 4,248 conduct 30-minute phone interviews with tar- Vietnam 2,739 get respondents. The interviews were Source: Computation based on the 2019 GIC Survey. conducted by enumerators, and response data 4 8    Global Investment Competitiveness Report 2019/2020 were entered in a computer system, a setup questionnaire, in the country’s primary busi- commonly referred to as computer-assisted ness language, was used to ensure than it telephone interviews (CATIs). The interviews could be completed within a reasonable time were conducted in nine languages: Bahasa frame. The online read-only version of the Indonesia, Mandarin Chinese, Cantonese questionnaire was available to be consulted Chinese, English, Portuguese, Spanish, Thai, during the interview. The fieldwork manag- Turkish, and Vietnamese. In addition to ers and CATI enumerators were screened to the main survey questions, each inter- ensure experience in conducting business view included a screener phase to ensure and market research, and they underwent the eligibility of respondents. The inter- specific interviewer training to prepare for views were conducted between June and this survey. A questionnaire manual with November 2019. detailed explanations of the questionnaire The survey was piloted in all 10 countries also served as a reference source while the to test the survey instrument in various lan- survey was being administered. guages and to identify effective strategies to Survey administration arrangements prior- increase response rates. The lessons from the itized respondents and constraints on the time pilot phase were used to reduce administra- of senior executives. Sensitive to variability in tion time and enhance overall clarity of the typical business hours and local norms survey instrument. around time use, 23 CATI enumerators The overall response rate for the survey attempted to establish contact and schedule was 9.3 percent. This response rate is consis- interviews during conducive time periods. tent with the current expected range for Survey administration arrangements such as phone-based business surveys.21 The main timing of calls, language options, repeat fieldwork of the survey leveraged lessons follow-up attempts, and scheduled callbacks from empirical research in survey design and were ­implemented to maximize the likelihood administration to implement the strategies of obtaining responses from the contacted described below to ensure high response sample. In case of initial failure to reach the rates.22 intended respondent, 5–10 follow-up call Potential respondents were notified by attempts were made. email before the survey. Building on research As a token of appreciation, respondents evidence (Dillman 2000; Lynn, Turner, and were promised a set of nonmonetary incen- Smith 1997), a prenotification email with tives. A key constraint to survey participation World Bank Group and International Finance is the opportunity cost of time. Business Corporation (IFC) branding was sent to poten- surveys impose a net cost on respondents, tial respondents to signal that the survey would requiring executives to apportion productive contribute to important global policy research. time away from work. Research largely The prenotification emails also directed poten- supports the use of incentives as an effective tial respondents to an informational website means to increase response rates (Singer and (www.investorsurvey.net) to obtain additional Ye 2013). To encourage potential respondents information about the survey, including to “invest” time in the survey, interviewers a (view-only) copy of the survey questionnaire. (a) emphasized the important policy research These measures aimed to lower information that the survey will inform; (b) promised to barriers and enhance trust between respon- send respondents a copy of the final research dents and interviewers, thereby improving the report; (c) promised to send respondents a cer- likelihood of securing an interview with senior tificate of appreciation; and (d) noted that a executives. charitable donation would be made to the An easy-to-follow survey questionnaire United Nations Children’s Fund (UNICEF) was administered by well-trained profes- when the target number of surveys was sional CATI enumerators. The survey reached.24 O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    49 Data Analysis percent of revenues, sector-export interac- tions, import share of inputs as a continu- Throughout the chapter, tests of the signifi- ous variable, sector-import interactions, cance of differences are conducted using source country income group, a dummy ordered logistic (for ordinal variables such for employment over 250 employees, a as ratings for importance) or logistic (for dummy for investment stock over binary variables) regressions. Unless other- US$10 million, number of years in host significance wise noted, tests for statistical ­ country, percentage of foreign ownership, of differences control for sector, a dummy and country fixed effects. variable for exports constituting over 50 Annex 1B. Country-Level FDI Outlook Trends TABLE 1B.1  FDI Outlook Trends in Surveyed Countries Forecast data Historical data Survey data (share planning to (inward FDI expand investment over next value, CAGR (inward FDI value, (Greenfield FDI project Country three years)a 18–21F) CAGR 13–18) announcements, CAGR 13–18) China 17% −7% –7% –6% Turkey 35% 2% –1% 8% Malaysia 40% 8% –5% 0% Thailand 43% −12% –4% 2% Mexico 45% −3% –5% –1% Nigeria 81% 29% –19% –4% Vietnam 46% 6% 12% 11% Brazil 49% –5% 3% –4% Indonesia 57% 4% –3% –9% India 64% 8% 8% 7% Sources: 2019 GIC Survey; Economist Intelligence Unit forecast data; World Development Indicators database (historical inward FDI value); and historical greenfield FDI data from fDi Markets, a Financial Times dataset (https://www.fdimarkets.com/). Note: Green shading indicates more growth, while red indicates low or negative growth. CAGR = compound annual growth rate; FDI = foreign direct investment. a. The interviews were conducted June–November 2019, so the implied three-year time horizon for this question is 2019–22. Notes 2008 to 2018, FDI inflows to China fell sharply (from 3.7 percent to 1.5 percent) but increased 1. “Developing countries” in this report refers to in Brazil (from 3.0 percent to 4.7 percent). As low- and middle-income countries as defined discussed in the Overview, a mix of economic by the World Bank. For the definitions of all factors are plausibly shaping global FDI income categories and the countries therein, trends, including declining rates of return see “World Bank Country and Lending on FDI, changes in US tax policy, increasingly Groups,” World Bank Knowledge Base: https:// asset-light forms of international production datahelpdesk.worldbank.org​/knowledgebase​ on the backs of digital technologies, and rising /articles/906519-world​ - bank-​ c ountry-and​ policy uncertainty. -lending-groups. 3. Recent projections show remittances exceed- 2. Although FDI inflows as a share of GDP ing FDI for low- and middle-income coun- have declined in most of the surveyed MICs, tries in 2019, although they are not projected experiences have varied. For example, from to do so for the 10 surveyed MICs given their 5 0    Global Investment Competitiveness Report 2019/2020 relatively higher FDI compared to remittances noted, tests for statistical significance of differ- (Global Knowledge Partnership on Migration ences are done via ordered logistic regression of and Development [KNOMAD] database: the dependent variable, controlling for a wide https://www.knomad.org/data/remittances). variety of company characteristics and country Foreign bank lending is also a significant fixed effects. For more details on the analytical source of private external finance and repre- methodologies such as modeling techniques sents about half of all external liabilities of and control variables, see annex 1A. emerging-market countries (Bräuning and 10. The 10 percentage point difference between Ivashina 2019). large and small firms is significant at the 4. The 2019 GIC Survey covers foreign companies p < 0.10 level. that have invested in the 10 surveyed countries. 11. The interviews were conducted between June It does not represent the perceptions and expe- and November 2019, so the implied time riences of companies that have never invested horizon for this question is 2019–22. in foreign countries or that have invested only 12. The coefficient in the ordered logistic regres- in countries other than those surveyed. The sion is negative and significant at the p < .10 results of the survey are not generalizable to level after controlling for firm characteristics all developing countries but are highly relevant and host country fixed effects. because the surveyed countries account for a 13. In the logistic regression with full controls substantial share of FDI inflows to developing and country fixed effects, the coefficient on countries (75 percent in 2018). having been adversely affected by policy 5. For the various uncertainty indexes, see the uncertainty is negative and significant at the Economic Policy Uncertainty index website: p < .05 level. https://www.policyuncertainty.com/. 14. Differences are significant at the p < .01 level 6. For a discussion on the sources of policy in the ordered logistic regression with full uncertainty, see the Overview of this volume. controls. 7. Studying this relationship empirically is par- 15. Differences are significant at the p < .10 level ticularly challenging using modeling techniques in the ordered logistic regression with full because of the strong assumptions required to controls. measure trade and investment policy uncer- 16. These differences are statistically significant tainty across heterogeneous firms and the lack at the p < .05 level in the ordered logistic of firm-level data on relevant economic out- regression with full controls. comes. Several 2019 surveys have attempted 17. Political instability includes high incidence to address these limitations by directly asking of political turbulence and internal con- firms about their investment plans in response flicts. Macroeconomic instability includes to trade tensions and uncertainty. They volatility in inflation and in real exchange include the seminannual Survey of Business rates. Uncertainty (SBU) in the United States con- 18. Ranking of importance is based on the ducted by the Federal Reserve Bank of Atlanta, percentage of investors that rate a factor Stanford University, and the University of as “ ­ important” or “critically important.” Chicago Booth School of Business (July Differences in average importance are signif- 2019); the annual US-China Business Council icant at the p < .01 level in the ordered logis- (USCBC) Member Survey of US businesses in tic regression with full controls and country China (August 2019); the monthly Reuters fixed effects. Tankan survey of Japanese manufacturers 19. Because of sampling frame limitations, (September 2019); and the quarterly UBS the Nigeria sample is 164 respondents Investor Sentiment survey (March 2019). (55 manufacturing and 109 services firms). 8. The “last year” refers to the last financial 20. The sample size of 125 respondents per sector year preceding the 2019 GIC Survey—that is, per country is greater than the required sam- a year spanning a 12-month period between ple strength for estimates with 7.5 ­ percent January 1, 2018, and September 30, 2019, precision in 90 percent confidence intervals. depending on the country. 21. The response rate calculation follows the 9. Throughout the chapter, tests of the significance Response Rate 3 (RR3) methodology out- of differences are done through either ordered lined in the latest guidance on response logistic (for ordered ordinal variables such as rates from The American Association for ratings for importance) or logistic (for binary Public Opinion Research (AAPOR 2016). variables) regression analysis. Unless otherwise This approach estimates the proportion of O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    51 cases of unknown eligibility that is actually AAPOR (American Association for Public eligible. The response rate was calculated as ­ Opinion Research). 2017. “The Future of U.S. follows: General Population Telephone Research.” Report, AAPOR Task Force, Oakbrook • Response rate = Interview / (Interview Terrace, IL. + Eligible Non-Interview + e (Unknown Abel, Andrew B., Avinash K. Dixit, and Janice Eligibility Non-Interview)) C. Eberly. 1996. “Options, the Value of • Response rate = (2424) / (2424 + 7309 + Capital, and Investment.” Quarterly Journal of 0.38 (42783)) = 9.33 percent Economics 111 (3): 753–77. • e = the estimated proportion of cases of Abel, Andrew B., and Janice C. Eberly. 1996. unknown eligibility that are eligible. “Optimal Investment with Costly Reversibility.” • e = (Confirmed Eligible) / (Confirmed Review of Economic Studies (63) 4: 581–93. Eligible + Confirmed Not Eligible) Aitken, Brian J., and Ann E. Harrison. • e = (2424 + 7309) / (2424 + 7309 + 1999. “Do Domestic Firms Benefit from 15818) = 0.38. Direct Foreign Investment? Evidence from   The Pew Research Center reported that Venezuela.” American Economic Review 89 response rates in 2017 and 2018 telephone (3): 605–18. surveys fell to 7 percent and 6 percent, Akame, Afuge Junior, Metougue Eric Ekwelle, respectively, a decline from the prior norm and George Ndonghandou Njei. 2016. “The of 9 percent (Kennedy and Hartig 2019). Impact of Business Climate on Foreign Direct Gallup reported attaining a similar 7 per- Investment in the CEMAC Region.” Journal cent average response rate in the Gallup Poll of Economics and Sustainable Development Social Series in 2017 (Marken 2018). The 7 (22): 66–74. AAPOR reported that response rates from Anand, Rahul, and Volodymyr Tulin. 2014. leading survey research firms were about 9 “Disentangling India’s Investment Slowdown.” percent for landlines and 7 percent for cell Working Paper 14/47, International Monetary phones in 2015 (AAPOR 2017). Fund, Washington, DC. 22. For global data collection from formal busi- Arnold, Jens Matthias, and Beata S. Javorcik. nesses in developing countries, the use of 2009. “Gifted Kids or Pushy Parents? Foreign telephone-based surveys can be advanta­ Direct Investment and Plant Productivity in geous. Compared with face-to-face surveys, Indonesia.” Journal of International Economics telephone surveys take less time and are less 79 (1): 42–53. expensive, and the near-universal prevalence Asiedu, Elizabeth. 2006. “Foreign Direct of telephones supports the generation of rep- Investment in Africa: The Role of Natural resentative samples (von der Lippe, Schmich, Resources, Market Size, Government Policy, and Lange 2011). However, this administra- Institutions and Political Instability.” World tion mode has unique challenges, including the Economy 29 (1): 63–77. growing aversion to divulging business infor- A.T. Kearney. 2019. “Facing a Growing Paradox: mation by phone (de Leeuw and Hox 2004) The 2019 A.T. Kearney Foreign Direct and the proliferation of answering machines Investment Confidence Index.” Annual survey and caller ID (Callegaro, McCutcheon, and report, A.T. Kearney, Chicago. Ludwig 2010). Baker, Scott R., Nicholas Bloom, and Steven 23. Local norms include, in some countries, the J. Davis. 2013. “Measuring Economic Policy prevalence of an afternoon break from work Uncertainty.” Working Paper No. 21633, or breaks for prayers. National Bureau of Economic Research, 24. A small monetary incentive for respondents Cambridge, MA. was also introduced in China. Baker, Scott R., Nicholas Bloom, and Steven J. Davis. 2019. “The Extraordinary Rise in Trade Policy Uncertainty.” VoxEU, September 17 (accessed November 7, 2019), https://voxeu.org/article​ References /extraordinary-rise-trade-policy-uncertainty. AAPOR (American Association for Public Beamish, Paul W., and John C. Banks. 1987. Opinion Research). 2016. “Standard “Equity Joint Ventures and the Theory of Definitions: Final Dispositions of Case Codes the Multinational Enterprise.” Journal and ­ Outcome Rates for Surveys.” 9th edition. of International Business Studies 18 (2): AAPOR, Oakbrook, IL. 1–16. 5 2    Global Investment Competitiveness Report 2019/2020 Bernanke, Benjamin S. 1983. “Irreversibility, Mexico.” Working Paper, Bank of Mexico, Uncertainty, and Cyclical Investment.” Quarterly Mexico City. Journal of Economics 98 (1): 85–106. Chen, Kexin, He Nie, and Zhenyu Ge. 2019. “Policy Blanchard, Emily J. 2019. “Trade Wars in the Uncertainty and FDI: Evidence from National Global Value Chain Era.” In Trade War: The Elections.” Journal of International Trade and Clash of Economic Systems Endangering Global Economic Development 28 (4): 419–28. Prosperity, edited by Meredith A. Crowley, Chen, Yu-Fu, and Michael Funke. 2003. “Option 57–63. VoxEU.org e-book, CEPR Press. Value, Policy Uncertainty, and the Foreign Bloom, Nicholas. 2009. “The Impact of Uncertainty Direct Investment Decision.” Discussion Shocks.” Econometrica 77 (3): 623–85. Papers 234, Hamburg Institute of International Bloom, Nicholas. 2014. “Fluctuations in Economics, Hamburg, Germany. Uncertainty.” Journal of Economic Perspectives Daude, Christian, and Ernesto Stein. 2007. “The 28 (2): 153–76. Quality of Institutions and Foreign Direct Bloom, Nicholas, Max Floetotto, Nir Jaimovich, Investment.” Economics and Politics 19 (3): Itay Saporta-Eksten, and Stephen J. Terry. 317–44. 2012. “Really Uncertain Business Cycles.” de Leeuw, Edith D., and Joop J. Hox. 2004. Working Paper No. 18245, National Bureau of “I Am Not Selling Anything: 29 Experiments in Economic Research, Cambridge, MA. Telephone Introductions.” International Journal Bonaime, Alice, Huseyin Gulen, and Mihai Ion. of Public Opinion Research 16 (4): 464–73. 2018. “Does Policy Uncertainty Affect Mergers Dillman, Don A. 2000. Mail and Internet Surveys. and Acquisitions?” Journal of Financial 2nd ed. Hoboken, NJ: John Wiley & Sons. Economics 129 (3): 531–58. Djankov, Simeon, and Bernard Hoekman. 2000. Bräuning, Falk, and Victoria Ivashina. 2019. “U.S. “Foreign Investment and Productivity Growth Monetary Policy and Emerging Market Credit in Czech Enterprises.” World Bank Economic Cycles.”  Journal of Monetary Economics . Review 14 (1): 49–64. Published ahead of print, February 22, 2019. Freund, Caroline, Mary Hallward-Driemeier, doi:10.1016/j.jmoneco.2019.02.005. and Bob Rijkers. 2014. “Deals and Delays: Buchanan, Bonnie G., Quan V. Le, and Meenakshi Firm-Level Evidence on Corruption and Policy Rishi. 2012. “Foreign Direct Investment Implementation Times.” The World Bank and Institutional Quality: Some Empirical Economic Review 30 (2): 354–82. Evidence.” International Review of Financial Gani, Azmat. 2007. “Governance and Foreign Analysis 21: 81–89. Direct Investment Links: Evidence from Panel Busse, Matthias, and Carsten Hefeker. 2007. Data Estimations.” Applied Economics Letters “Political Risk, Institutions and Foreign Direct 14 (10): 753–56. Investment.”  European Journal of Political Ghosal, Vivek, and Prakash Loungani. 2000. Economy 23 (2): 397–415. “The Differential Impact of Uncertainty on Caldara, Dario, Matteo Iacoviello, Patrick Molligo, Investment in Small and Large Businesses.” Andrea Prestipino, and Andrea Raffo. 2019. Review of Economics and Statistics 82 (2): “The Economic Effects of Trade Policy 338–43. Uncertainty.” Journal of Monetary Economics. Globerman, Steven, and Daniel Shapiro. 2002. Published ahead of print, November 8, 2019. “Global Foreign Direct Investment Flows: The doi:10.1016/j.jmoneco.2019.11.002. Role of Governance Infrastructure.” World Callegaro, Mario, Allan L. McCutcheon, and Jack Development 30 (11): 1899–919. Ludwig. 2010. “Who’s Calling? The Impact Gonzalez, Anabel, Christine Zhenwei Qiang, of Caller ID on Telephone Survey Response.” and Peter Kusek. 2018. “Overview.” In Field Methods 22 (2): 175–91. 2017/2018 Global Investment Competitiveness Cao, Chunfang, Xiaoyang Li, and Guilin Liu. Report: Foreign Investor Perspectives and 2017. “Political Uncertainty and Cross-Border Policy Implications, 1–18. Washington, DC: Acquisitions.” Review of Finance 23 (2): World Bank. 439–70. Gulen, Huseyin, and Mihai Ion. 2016. “Policy Cebreros, Alfonso, Aldo Heffner, and Alejandrina Uncertainty and Corporate Investment.” Salcedo. 2019. “Trade Policy Uncertainty Review of Financial Studies 29 (3): 523–56. and Its Effect on Foreign Direct Investment Hallward-Driemeier, Mary, and Lant Pritchett. and Export Participation: Evidence from 2015. “How Business Is Done in the Developing O u t l o o k a n d Pri o ri t i e s f o r F o r e ig n I n v e s t o r s i n D e v e l o p i n g C o u n t ri e s    53 World: Deals versus Rules.”  Journal of Lin, Ping, and Kamal Saggi. 2005. “Multinational Economic Perspectives 29 (3): 121–40. Firms and Backward Linkages: A Critical Survey Havránek, Tomáš, and Zuzana Iršová. 2011. and a Simple Model.” In Does Foreign Direct “Estimating Vertical Spillovers from FDI: Investment Promote Development? edited by Why Results Vary and What the True Effect Theodore H. Moran, Edward Graham, and Is.” Journal of International Economics Magnus Blomstrom, 159–74. Washington, DC: 85 (2)​ : 234–44. Peterson Institute for International Economics IMF (International Monetary Fund). 2019. and Center for Global Development. “World Economic Outlook Update, July Lynn, P., R. Turner, and P. Smith. 1997. “The 2019: Still Sluggish Global Growth.” IMF, Effect of Complexity and Tone of an Advance Washington, DC. Letter on Response to an Interview Survey.” IMF (International Monetary Fund). 2020. World Survey Methods Centre Newsletter 17: Economic Outlook, April 2020: The Great 13–17. Lockdown. Washington, DC: IMF. Marken, Stephanie. 2018. “Still Listening: Iršová, Zuzana, and Tomáš Havránek. 2013. The State of Telephone Surveys.” Gallup “Determinants of Horizontal Spillovers from Methodology Blog, January 11. https://news​ FDI: Evidence from a Large Meta-Analysis.” .gallup.com/opinion/methodology/225143​ World Development 42: 1–15. /listening-state-telephone-surveys.aspx. Javorcik, Beata Smarzynska. 2004. “Does Foreign Mistura, Fernando, and Caroline Roulet. Direct Investment Increase the Productivity 2019. “The Determinants of Foreign Direct of Domestic Firms? In Search of Spillovers Investment: Do Statutory Restrictions Matter?” through Backward Linkages.” American OECD Working Papers on International Economic Review 94 (3): 605–27. Investment 2019/01, Organisation for Economic Julio, Brandon, and Youngsuk Yook. 2016. “Policy Co-operation and Development, Paris. Uncertainty, Irreversibility, and Cross-Border Rodrik, Dani. 1991. “Policy Uncertainty and Flows of Capital.” Journal of International Private Investment in Developing Countries.” Economics 103: 13–26. Journal of Development Economics 36 (2): Jun, Kwang W., and Harinder Singh. 1996. “The 229–42. Determinants of Foreign Direct Investment Schneider, Friedrich, and Bruno S. Frey. 1985. in Developing Countries.”  Transnational “Economic and Political Determinants Corporations 5 (2): 67–105. of Foreign Direct Investment.”  World Kazmin, Amy. 2019. “India Slashes Corporate Development 13 (2): 161–75. Taxes to Give a Boost to Economy.” Financial Sekkat, Khalid, and Marie-Ange Veganzones– Times, September 20. Varoudakis. 2007. “Openness, Investment Kennedy, Courtney, and Hannah Hartig. 2019. Climate, and FDI in Developing Countries.” “Response Rates in Telephone Surveys Have Review of Development Economics  11 (4): Resumed Their Decline.” Fact Tank (online 607–20. platform) report, February 27, Pew Research Singer, Eleanor, and Cong Ye. 2013. “The Use and Center, Washington, DC. Effects of Incentives in Surveys.” Annals of Knight, Frank H. 1921. Risk, Uncertainty, and the American Academy of Political and Social Profit. Boston: Hart, Schaffner, and Marx; Science 645 (1): 112–41. Houghton Mifflin. UNCTAD (United Nations Conference on Trade Krifa-Schneider, Hadjila, and Iuliana Matei. 2010. and Development). 2013. World Investment “Business Climate, Political Risk and FDI in Report 2013: Global Value Chains: Investment Developing Countries: Evidence from Panel and Trade for Development. Geneva: Data.” International Journal of Economics and UNCTAD. Finance 2 (5): 54–65. UNCTAD (United Nations Conference on Trade Kusek, Peter, and Andrea Silva. 2018. “What and Development). 2016. World Investment Matters to Investors in Developing Countries: Report 2016: Investor Nationality: Policy Findings from the Global Investment Challenges. Geneva: UNCTAD. Competitiveness Survey.” In 2017/2018 Global UNCTAD (United Nations Conference on Trade Investment Competitiveness Report: Foreign and Development). 2019. World Investment Investor Perspectives and Policy Implications, Report 2019: Special Economic Zones. 19–50. Washington, DC: World Bank. New York: UNCTAD. 5 4    Global Investment Competitiveness Report 2019/2020 UNCTAD (United Nations Conference on Trade Wei, Yingqi, Xiaming Liu, and Chengang Wang. and Development). 2020. Investment Policy 2008. “Mutual Productivity Spillovers between Monitor 23. April. UNCTAD. Foreign and Local Affiliates in China.” Vogiatzoglou, Klimis. 2016. “Ease of Doing Cambridge Journal of Economics 32 (4): Business and FDI Inflows in ASEAN.” 609–31. Journal of Southeast Asian Economies 33 (3): Wernick, David A., Jerry Haar, and Shane Singh. 343–63. 2009. “Do Governing Institutions Affect Foreign von der Lippe, Elena, Patrick Schmich, and Direct Investment Inflows? New Evidence from Cornelia Lange. 2011. “Advance Letters as a Emerging Economies.” International Journal of Way of Reducing Non-Response in a National Economics and Business Research 1 (3): 317–22. Health Telephone Survey: Differences between World Bank. 2019a. Doing Business 2020: Listed and Unlisted Numbers.” Survey Research Comparing Business Regulation in 190 Methods 5 (3): 103–16. Economies. Washington, DC: World Bank. Walch, Nadja, and Julia Wörz. 2012. “The Impact Wo r l d B a n k . 2 0 1 9 b . G l o b a l E c o n o m i c of Country Risk Ratings and of the Status Prospects, June 2019: Heightened Tensions, of EU Integration on FDI Inflows in CESEE Subdued Investment . Washington, DC: Countries.” Focus on European Economic World Bank. Integration 3: 8–26. World Bank. 2020. World Development Report Wei, Shang Jin. 2000. “How Taxing is Corruption 2020: Trading for Development in the Age on International Investors?” Review of of Global Value Chains. Washington, DC: Economics and Statistics 82 (1): 1–11. World Bank. 2 How Beneficial Are Foreign Acquisitions of Firms in Developing Countries? Evidence from Six Countries Alexandros Ragoussis Key Findings • Acquisitions of domestic firms by foreign investors have doubled as a share of FDI in developing countries over the past 10 years. The acquired firms in developing countries can be diverse. They can be in any sector but more commonly in activities that rely on land and established distribution net- works, as well as in sectors where entry is highly regulated. For developing countries, income level and market size are strong predictors of the intensity of “brownfield” investment. • This report explores differences in the performance and development impact of brownfield FDI relative to greenfield FDI and domestically owned firms by analyzing a unique set of industrial censuses from six developing countries: China, Côte d’Ivoire, Indonesia, Moldova, Serbia, and Viet- nam. Although FDI’s benefits to economic development have been well studied using aggregate sta- tistics and country case studies, this report brings new evidence on the contributions of brownfield FDI to developing countries’ competitiveness, productivity, and labor markets. • Results show that firms acquired by multinational enterprises not only perform better than the average domestic firm at the time of the acquisition but also improve their performance after acqui- sition faster than local firms along some of the key dimensions that matter for development. For example, over the first five years of a firm’s operation, a brownfield affiliate is 70–100 percent more likely to export than a domestic firm. Wages in foreign take-overs at the end of the first five years of operations are 40–50 percent higher than in domestic firms. • Furthermore, contrary to conventional belief about the potential job-destroying effects of for- eign mergers and acquisitions, employment in newly acquired firms tends to grow faster in most countries than employment in domestic firms with similar characteristics. Specifically, two years after acquisition, the average employment in brownfield affiliates expands by approximately 4 per- cent, compared with 1.5 percent in domestic firms with similar characteristics. The firms’ asset value after the acquisition follows a similar path. In addition, wages in brownfield affiliates tend to increase, compared with relatively ­ stagnant wages in the domestic firms. The experience of the six countries analyzed in this study suggests that foreign acquisitions could be adding more value in markets at the lower end of the development spectrum—that is, in countries where most FDI still takes place through greenfield investment. • A policy framework that is supportive of brownfield investment should emphasize (a) streamlin- ing investment screening mechanisms and approval processes, (b) increasing the effectiveness of competition policies to reduce the administrative burden of merger and acquisition controls, and (c) enhancing cooperation between competition and investment authorities. Although legal safe- guards to protect the public interest ought to be preserved, an alleviation of the administrative bur- den, unpredictability, cost, and time involved in brownfield investment would facilitate the process greatly. Differential treatment of brownfield multinationals with respect to investment incentives should also be avoided. Introduction Acquisitions of domestic firms by foreign modest effects on aggregate growth (Harms investors have doubled as a share of total and Meon 2018; Wang and Wong 2009). foreign direct investment (FDI) in developing Narratives likening cross-border mergers and countries over the past 10 years. In the past, acquisitions (M&A) to “bad cholesterol” or foreign investors in developing countries qualifying them as “useless” rely on this evi- would typically establish new facilities in dence to play down their contribution (Beattie unused “green fields” rather than investing 2014; Harms and Meon 2018). These critiques in established companies in potentially con- are not new. Following the 1990s surge of taminated “brown fields,” as the analogy acquisitions of state-owned enterprises, the goes. That is still how most FDI takes place United Nations Conference on Trade and outside the industrialized world. Development (UNCTAD) noted “concerns in The balance between the two modes of political discussions and the media that foreign entry, however, is shifting toward more brown- acquisitions as a mode of entry are less benefi- field investment. Upper- and lower-middle- cial for economic development, if not posi- income countries lead the way ­ (figure 2.1). The tively harmful” (UNCTAD 2000). trend is also discernible in outward investment For the development community, one by upper- and lower-middle-income countries complicating factor in discerning the effect to other developing countries and, notably, to of brownfield investment is that most of the high-income economies. evidence comes from high-income coun- The rise of foreign acquisitions brings ten- tries, where the impact of investment can sions in the investment landscape. The United be different in scope and depth than in States and the European Union have enacted developing countries. In addition, much of strict screenings of foreign acquisitions in what captures the public eye tends to focus response to perceived challenges to national on macroeconomic growth, overlooking security. Cases of investment withdrawals— the shift of attention to development out- either rejected or withdrawn over security comes at the level of firms, the jobs they concerns—tripled in 2018 alone, often receiv- create, or the wages they offer. Little is ing high publicity (UNCTAD 2019). China known about acquired firms in developing and South Africa have also changed their FDI countries1—what they look like, how they screening mechanisms in recent years for the evolve, and whether conventional narra- same reasons. Although tensions arise pri- tives do justice to their contribution in marily over assets in high-income and large development terms. emerging economies, they entail a risk of This study uses a unique set of industrial shaping narratives, policy precedents, and censuses from six developing countries responses beyond their own jurisdictions. (China, Côte d’Ivoire, Indonesia, Moldova, But how beneficial is brownfield FDI for Serbia, and Vietnam) to shed additional light developing countries? Opinions vary. on these three questions. The discussion that Greenfield FDI adds new elements to the follows shows that acquired firms in develop- economy: new facilities, new jobs, and addi- ing countries can be diverse. They can be in tional production capacity. Brownfield invest- any sector but are more common in activities ment, by contrast, transforms existing that rely on land and established distribution production. Any positive effect would there- networks, as well as in sectors where entry is fore tend to materialize over longer time highly regulated. frames and with varying intensity. Motivations matter: firm acquisitions are That brownfield investment represents rents strongly associated with market-seeking and to existing assets is the prevailing explanation asset-seeking FDI. For developing countries, offered by various empirical studies for the income and the size of the market are strong 57 5 8    Global Investment Competitiveness Report 2019/2020 FIGURE 2.1  Brownfield Investment Rose as a Share of Total FDI in brownfield investment takes place more fre- Developing Countries, 2010–17 quently, such as agriculture or services. 60 Governments have various means to foster the potential of brownfield ventures in these contexts. Improving the predictability of Share of brownfield FDI in total FDI (%) 50 screening mechanisms; strengthening competi- tion enforcement; ensuring equal applicability 40 of incentives such as tax credits, preferential rates, or subsidies; and facilitating the partici- 30 pation of foreign investors in development of firms can go a long way in that direction. 20 Although legal safeguards to protect the public interest ought to be preserved, an alleviation of 10 the administrative burden, unpredictability, cost, and time involved in the process would 0 facilitate brownfield investment greatly. 2010 2011 2012 2013 2014 2015 2016 2017 The rest of the chapter is organized as fol- Low income Lower middle income lows: The Choice between Brownfield and Upper middle income High income Greenfield FDI reviews existing knowledge Source: World Bank, based on 2019 United Nations Conference on Trade and Development on brownfield FDI as the starting point for (UNCTAD) Mergers and Acquisitions (M&A) database, http://www.unctad.org/fdistatistics. Note: The trend is illustrated using a two-year moving average. All country income categories use the investigation. Characteristics of Markets 2017 World Bank-defined classifications; “developing” countries refers collectively to low- and Affecting Brownfield FDI describes countries middle-income countries. FDI = foreign direct investment. and sectors where brownfield investment predictors of the intensity in which acquisi- grows more rapidly. Differences between tions take place. And while there is a sorting Brownfield Affiliates and Other Firms delves of average outcomes between the three into the firm-level outcomes, growth, and groups—greenfield affiliates doing better than transformation paths of firms taken over by brownfield affiliates, and the latter doing bet- foreign investors. Policy Considerations for ter than domestic firms—brownfield affiliates Brownfield FDI explores policy options are shown to develop significant advantages for countries to foster their development over domestic firms, and those advantages are potential. consolidated in the first five years of their operations. Steeper transformation paths of domestic firms taken over by foreign investors The Choice between Brownfield highlight important contributions of brown- and Greenfield FDI field FDI to some key outcomes that matter for development. There are two main ways for a foreign inves- Although more evidence would be neces- tor to enter a market: either set up a new sary to establish general conclusions, the firm or acquire existing facilities (box 2.1). experience of the six countries analyzed in The choice between the two depends natu- this study suggests that foreign acquisitions rally on which yields the greatest return. The could be adding more value in markets at the benefits and costs differ substantially by lower end of the development spectrum—that mode of entry. is, in countries where most FDI still takes The value of a firm’s assets is the prime place through greenfield investment. This driver of brownfield investment. Access to a study’s findings could therefore be more rele- successful firm’s technology, machinery, or vant for rapidly growing economies where the brand name represents a future stream of rev- share of foreign acquisitions is rising as well enues to the acquirer, especially when those as for investment into sectors in which assets yield lower returns domestically than H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    59 BOX 2.1 Conceptual Overlaps between Mergers, Acquisitions, and Brownfield FDI “Mergers and acquisitions” (M&A) is a general term as (a) ­horizontal: between firms that produce and used to describe the consolidation of companies or sell the same products—that is, between competing assets through various types of financial transactions. firms; (b) vertical: between firms operating at dif- The terms “merger” and “acquisition” are often used ferent stages of the value chain; or (c) conglomerate: interchangeably, although they have slightly different between firms in unrelated businesses. meanings. “Brownfield foreign direct investment (FDI)” is a In an acquisition , a company purchases another broader term for any purchase by a foreign entity of entity, partially or entirely, and establishes itself as assets that corresponds to more than 10 percent of the a new owner. From a legal point of view, the target total assets of a target company, which is the thresh- company does not cease to exist. In other words, old for a foreign investment to be considered direct acquisitions involve the purchase of an entity’s assets (FDI) according to the International Monetary Fund without a change in market structure. (IMF) and Organisation for Economic Co-operation In a merger, on the other hand, one or several enti- and Development (OECD). ties involved cease to exist, and a new entity may be Ownership of a 10 percent share does not necessar- created; thus there is a change in market structure. ily grant control over the firm. The investor’s ability In the case of public firms, the boards of directors of to make independent decisions would require a major- the two companies approve the combination of their ity share, although shareholders can significantly assets and seek shareholders’ approval. If the purchase influence the firm strategies and managerial decisions of assets takes place without the consent of the board at lower thresholds, generally over 30 percent. The or shareholders of the target company, the operation purchase can be friendly or unfriendly and result in is called a “hostile take-over.” A hostile take-over can various combinations of outcomes in terms of creat- also result in a merger, whereby companies’ stocks are ing a new legal entity, including a simple acquisition surrendered, and new company stock is issued in its or a merger. Joint ventures do not fall under the cat- place. egory of brownfield foreign investment because they Depending on the activities exercised by the refer to the establishment of new facilities—greenfield buyer and seller, M&A can further be classified investment—involving a local and a foreign entity. ­ in international markets. The foundation of target firm, the larger the valuation gains rela- the decision to acquire lies in the investor’s tive to operational synergies. ability to raise the value of the assets, which These sources of surplus can become a often involves a substantial but lower relatively more important motive for brown- ­ commitment of resources than setting up a field investment during financial crises—a subsidiary from scratch. Partly for this situation often referred to as “fire-sale FDI,” reason, brownfield FDI seems to attract less ­ coined by Paul Krugman to describe the productive investors than greenfield invest- ­ surge in foreign acquisitions of Asian firms ment, on average.2 during the 1997–98 financial crisis (Krugman Fundamentally, there can be two sources 2000). Indeed, a crisis is associated with a of surplus from foreign acquisitions: (a) the 30 percent increase in the probability of a efficiency gains due to operational synergies, foreign acquisition of a typical target relative and (b) the valuation gains associated with to the noncrisis average (Alquist, Mukherjee, the relaxation of the target firm’s liquidity and Tesar 2016). But more generally, valua- constraints. The more credit constrained the tion gains appear to explain an important 6 0    Global Investment Competitiveness Report 2019/2020 share of variation in cross-country mergers and consumers, and greater costs in setting up and acquisitions (M&A): firms in countries new facilities. Sectors such as real estate, finan- whose stock market has increased in value, cial services, or pharmaceuticals are examples whose currency has recently appreciated, (Davies, Desbordes, and Ray 2018). The high and that have a relatively high market-to- cost of setting up a local supply network in a book value tend to be purchasers, while vertically integrated market with strong back- firms from weaker-­ performing economies ward linkages also favors cross-border acqui- tend to be targets (Erel, Liao, and Weisbach sition relative to greenfield investment (Milliou 2011). and Pavlou 2014). In high-income countries, The value of the acquired assets is always the location, geography, and cultural barriers, assessed against the purpose of the invest- together with tariff rates applied on ment. If the main objective is to sell in inputs from the origin country, are other fac- the domestic market—a “market-seeking” tors affecting the attractiveness of acquisitions investment—then acquiring a company that ­ (Davies, Desbordes, and Ray 2018; di is already operating is a common way to Giovanni 2005; Roberto 2004). gain access. Market intelligence is part of that Overall, a variety of factors explain the advantage: existing firms know the demand decision to acquire rather than set up a new and know the risks, so the investor does not venture: the particularities of the sector and have to start at the bottom of the learning the location, the motivation, the level of con- curve. Complementarity between acquired trol over intellectual property, the macroeco- assets and foreign owners’ tangible and intan- nomic environment, restrictions to alternative gible assets is another factor that often tips modes such as costly procedures for construc- the balance in favor of brownfield FDI tion permits, and the length of the commit- (Balsvik and Haller 2010; Curran, Lv, and ment wished by the investor. Francesa Spigarelli 2017). In all cases, the average domestic firm in a host country Impacts of FDI on Acquired Firms and would rarely combine all these qualities, so Host Economies foreign investors “cherry pick” the more suc- cessful, productive, and profitable ones that What is perhaps more critical from a develop- suit their plans (Almeida 2007; Balsvik and ment standpoint, however, is the question of Haller 2010; Bertrand et al. 2012; Guadalupe, the impact of foreign take-overs on acquired Kuzmina, and Thomas 2012). firms. Employment tops the list of concerns. There are reasons why investors may pre- In developed economies, both job losses and fer to set up a new venture rather than acquire gains have been documented in acquired an existing firm. Control over precious intel- firms over time. In developing economies, lectual property, operations, and management there are fewer records of effects of acquisi- would justify that preference. To secure their tions on employment, but they are generally property, investors are often willing to bear positive.3 These findings are consistent with higher costs of construction and navigate a the findings that foreign investment in exist- host country’s regulatory system and tax ing firms improves productivity (Bircan 2019; structure, in what consists overall in a longer- Conyon et al. 2002; Hale and Xu 2016; term commitment to the market and host Lichtenberg and Siegel 2000; Maksimovic country. and Phillips 2001; Maksimovic, Phillips, and The activity of the firm itself matters in the Prabhala 2008; Schoar 2002), which tends to decision to build or buy. Acquisitions do not be associated with larger size. Skill and make equal sense in all sectors of economic knowledge transfers, as well as increased activity; they can be more beneficial for the labor efficiency are the main channels through investor in markets where there is higher con- which improvements happen. And productiv- tractual intensity, higher informational asym- ity improvements in turn lead gradually to metries among firms as well as between firms better wages. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    61 Positive valuation gains are also reported A distinct feature of brownfield investment for the acquiring firm when it buys a majority is also that it can save jobs and revenue rather stake in an enterprise in a developing country than create new ones, by restructuring com- (Chari, Ouimet, and Tesar 2010). The size of panies that would otherwise fail to sustain the stock price increase for the buyer is more operations (Grzegorz 2014). And although pronounced when the contracting environ- little has been written on the evolution of ment is weak and in industries with high asset domestic linkages of foreign take-overs in intangibility. broader areas like revenue and taxes, there is When it comes to the contribution of for- agreement that, unless they transfer their eign take-overs to the growth of the host entire profit to the parent company, brown- economy as whole, the prevailing view is field affiliates make a significant contribution that the growth impact of greenfield invest- to the host economy (Bandick and Karpaty ment is stronger than that of acquisitions 2007; Beattie 2014). (Calderón, Loayza, and Servén 2004). The Feedback loops whereby brownfield commitment involved in greenfield ventures investment induces more greenfield invest- also makes their impact more lasting ment by reducing informational asymme- (Bandick and Karpaty 2007; Harms and tries, and greenfield investment induces Meon 2018). Greenfield ventures not only more brownfield investment, have been raise the production capacity and create jobs studied less and have shown mixed results. but also intensify competition by increasing In high-income countries, the limited evi- the number of suppliers, which adds to their dence suggests that foreign acquisitions are appeal (Burger and Ianchovichina 2017; associated with more greenfield FDI over Claeys and Hainz 2007). time, while the reverse has been observed in Cross-border acquisitions, by contrast, developing countries: greenfield investment keep the number of market players is associated with more brownfield invest- unchanged. Their effects on market structure ment over time (Calderón, Loayza, and thus are often thought to be neutral, if not Servén 2004). negative, which is a recurring concern when foreign investors enter by this mode (OECD 2012). In developing countries in particular, Characteristics of Markets newly acquired firms can in principle capital- Attracting Brownfield FDI ize on advantages associated with foreign ownership and concentrate market power The decision to set up a firm or acquire more easily where competition enforcement is one is driven by multiple factors. How ineffective. Evidence on the general validity of does a country’s level of development that effect across countries and contexts influence that decision? Stronger institu- remains limited. tions associated with development enable In the same way that greenfield investment more effective protection of investors’ transforms markets, brownfield investment intellectual property and more reliable can also bring about job creation, innovation, contractual arrangements, both of which and competition, but it takes longer for that make brownfield investment more appeal- impact to materialize as the acquired firms ing. A m­ arket’s level of development, gradually improve their position in the host moreover, reflects the attractiveness of market. Recent evidence from Turkey, for acquisition targets—the presence of suc- example, indicates that foreign acquisitions cessful firms that require less time and increase physical productivity in acquired effort to bring returns—as well as struc- firms while lowering competitor prices tural shifts to services in which brown- (Bircan 2019). This finding suggests that their field affiliates are more common. But is procompetitive effects may have been under- the relationship between share of acquisi- estimated to date. tions and level of development linear, and 6 2    Global Investment Competitiveness Report 2019/2020 which sectors and firms are more likely to investment and its relative intensity. Of the receive foreign investment in developing US$313 billion of brownfield investment in countries? developing countries between 2014 and 2017, Macroeconomic investment data by three-quarters consisted of acquisitions of mode of entry (aggregated from commercial assets in upper-middle-income countries. The sources) and firm-level microdata from more developed among developing countries industrial censuses in six developing coun- also tend to receive higher shares of brown- tries—China, Côte d’Ivoire, Indonesia, field investment in total foreign investment Moldova, Serbia, and Vietnam—employed (figure 2.2). in this study yield a wealth of insights. A Regional concentration is also evident few caveats are in order, however. The data- because of, or perhaps in addition to, income. sets have imperfections: in all cases they Latin American economies such as Argentina, include some transactions that do not Chile, Colombia, Mexico, and Peru have strictly fall under the standard definition of disproportionally high shares of brownfield ­ FDI. They also have not been thoroughly FDI, while in Sub-Saharan Africa, the share benchmarked in terms of representativeness of foreign acquisitions in total FDI is very and coverage. (See annex 2A for more low. The size of the market matters as well. details on the sources, their contents, and The BRICS (Brazil, the Russian Federation, limitations.) India, China, and South Africa), Argentina, Brownfield investment occurs frequently in and Mexico—where larger shares of brown- developing countries, but more so in some field investment are recorded—are some of than in others. Income is a strong predictor of the largest economies in the developing both the absolute volume of brownfield world. FIGURE 2.2  Higher Shares of Brownfield Investments Occur in More Developed Economies 100 90 Share of M&A in total FDI inflows 2007–17 (%) 80 70 60 50 40 30 20 10 0 100 1,000 10,000 10,0000 GDP per capita (mean 2007–17) Source: World Bank calculations, based on data from the United Nations Conference on Trade and Development (UNCTAD) Mergers and Acquisitions (M&A) database (http://www.unctad.org/fdistatistics) and fDi Markets, a Financial Times dataset (https://www.fdimarkets.com/). Note: The figure shows investment destination countries with per capita GDP below US$15,000. Data are averages for 2007–17. FDI = foreign direct investment; GDP = gross domestic product; M&A = mergers and acquisitions. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    63 The sectoral composition of these large networks are hard to build from scratch, such economies might also lie behind cross-­ country as food and beverages, wholesale and retail differences in the intensity of acquisitions. trade, and health services; and (c) sectors that The aggregate data confirm that although are highly regulated, such as financial services brownfield investment can take place in any (figure 2.3). Manufacturing activities with sector, it occurs more intensely in (a) activities strong backward or forward linkages to these that rely on land (agriculture, mining, and real industries—such as food processing, tobacco, estate, for example) where access is restricted; or pharmaceuticals—also attract significantly (b) activities where distribution and client greater shares of brownfield investment. FIGURE 2.3  Brownfield FDI Is Likelier in Sectors That Rely on Land, Have Established Distribution Networks, or Are Highly Regulated a. High-income countries Public administration and defense; compulsory social security 100.0 Other service activities 100.0 Construction 100.0 Agriculture, forestry, and fishing 98.6 1.4 Administrative and support services 77.0 23.0 Mining and quarrying 63.6 36.4 Wholesale and retail trade 60.1 39.9 Human health and social work 51.0 49.0 Professional, scientific, and technical 48.5 51.5 Financial and insurance 47.9 52.1 Manufacturing 43.5 56.5 Water supply, sewerage, waste management, and remediation 43.0 57.0 Accommodation and food service 35.8 64.2 Arts, entertainment, and recreation 34.4 65.6 Transportation and storage 31.2 68.8 Education 25.3 74.7 Real estate 23.5 76.5 Electricity, gas, steam, and air conditioning supply 14.4 85.6 Information and communication 11.4 88.6 0 20 40 60 80 100 Share of total FDI (%) b. Upper-middle-income countries Public administration and defense; compulsory social security 100.0 Other service activities 100.0 Construction 100.0 Agriculture, forestry, and fishing 84.8 15.2 Human health and social work 57.2 42.8 Education 37.8 62.2 Financial and insurance 27.6 72.4 Administrative and support services 24.0 76.0 Professional, scientific, and technical 21.2 78.8 Arts, entertainment, and recreation 19.8 80.2 Transportation and storage 13.1 86.9 Real estate 10.9 89.1 Mining and quarrying 9.4 90.6 Electricity, gas, steam, and air conditioning supply 8.3 91.7 Water supply, sewerage, waste management, and remediation 8.0 92.0 Manufacturing 6.1 93.9 Information and communication 3.6 96.4 Accommodation and food service 3.1 96.9 0 20 40 60 80 100 Share of total FDI (%) M&A Greenfield Figure continues next page 6 4    Global Investment Competitiveness Report 2019/2020 FIGURE 2.3  Brownfield FDI Is Likelier in Sectors That Rely on Land, Have Established Distribution Networks, or Are Highly Regulated (continued) c. Lower-middle-income countries Construction 100.0 Agriculture, forestry, and fishing 78.6 21.4 Mining and quarrying 22.5 77.5 Human health and social work 21.1 78.9 Wholesale and retail trade 11.1 88.9 Financial and insurance 11.1 88.9 Education 9.8 90.2 Water supply, sewerage, waste management, and remediation 6.6 93.4 Administrative and support services 6.5 93.5 Information and communication 4.4 95.6 Manufacturing 3.8 96.2 Arts, entertainment, and recreation 3.0 97.0 Transportation and storage 2.7 97.3 Accommodation and food service 2.0 98.0 Real estate 1.1 98.9 Electricity, gas, steam, and air conditioning supply 1.0 99.0 Professional, scientific, and technical 0.9 99.1 0 20 40 60 80 100 Share of total FDI (%) d. Low-income countries Wholesale and retail trade 30.0 70.0 Mining and quarrying 9.1 90.9 Transportation and storage 5.3 94.7 Manufacturing 1.4 98.6 Information and communication 1.3 98.7 Financial and insurance 0.5 99.5 Real estate 0.2 99.8 Professional, scientific, and technical 0.0 100.0 Human health and social work 0.0 100.0 Electricity, gas, steam, and air conditioning supply 0.0 100.0 Education 0.0 100.0 Agriculture, forestry, and fishing 0.0 100.0 Administrative and support services 0.0 100.0 Accommodation and food service 0.0 100.0 0 20 40 60 80 100 Share of total FDI (%) M&A Greenfield Source: World Bank calculations, based on data from the United Nations Conference on Trade and Development (UNCTAD) Mergers and Acquisitions (M&A) database (http://www.unctad.org/fdistatistics) and fDi Markets, a Financial Times dataset (https://www.fdimarkets.com/). Note: Data are for 2010–17. Volumes are aggregated by country income group, following the World Bank classifications for fiscal year 2017 throughout, to maintain consistency. Only shares of foreign direct investment (FDI) over total industry values exceeding US$20 million are represented in the panels. FDI = foreign direct investment; M&A = mergers and acquisitions. In upper-middle-income countries, a sub- place in agriculture, mining, wholesale stantial share of brownfield investment trade, and construction. In other words, takes place in all these sectors in addition this mode of entry prevails in industries to important volumes in manufacturing. where foreign investors have few to no In lower-middle-income countries and low- alternatives. income countries, by contrast, the ­volumes More generally, country-specific factors remain lower and sectoral concentration is such as the market size, level of develop- high: brownfield investment tends to take ment, and quality of institutions explain a H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    65 much greater share of variation in modes of almost 54,000 observations of multinational entry than sectoral characteristics. Firm-level firms that have been acquired by foreign data from six developing countries lend investors at some point in their lifetime. Of ­ support to this conclusion (see annex 2B, these, about 16,700 observations are from table 2B.1). Although both country-specific firms whose time of origin is observed (see factors and sector-specific factors drive annex 2A, table 2A.1). the intensity by which foreign investors The countries themselves are quite choose one mode over another, the former diverse, spanning three continents and three explain twice as much of the variation in income levels, with stark differences in ­ f requency of foreign acquisition as the terms of market size (ranging from China to latter. Moldova) and varied industrial structures. The regulatory context of the six countries varies as well. With the exceptions of Serbia Differences between Brownfield and Vietnam, the rest fare rather poorly during the period studied in terms of the Affiliates and Other Firms World Bank’s Doing Business scores— Delving further into how brownfield affiliates particularly on the subcomponent dealing differ from the rest of firms in host econo- with construction permits, which would mies is not simple because of data con- incentivize modes of entry other than straints. Many industrial surveys do not ask greenfield. 4 The sample overall allows firms to report the origin of their capital, cross-country evidence over multiple whether domestic or foreign. This reduces dimensions that previous studies have not the number of surveys that can be used for been able to capture. this exercise. In surveys that do include ownership infor- Different Outcomes of Brownfield mation, greenfield affiliates can be identified Ventures Relative to Greenfield only if a firm’s activity is observed in the year Ventures or Domestic Firms it is established, which is rare. Many firms do not report activity figures until several years A closer examination of the six countries’ after entry, which adds a significant margin of industrial censuses reveals important differ- error in the estimates. To identify brownfield ences within the multinational segment by affiliates, by contrast, information on the time mode of entry, as expected from previous of establishment is not needed. All that is work. The average brownfield affiliate is required is a moment when foreign owner- larger than a greenfield affiliate, measured by ship turns positive. The sample of firms where workforce size. However, this statistic masks this happens is older and larger; it tells us a the fact that greenfield affiliates that can be lot about the characteristics of firms that for- identified in the sample are often younger eign investors acquire. But to reliably study than the rest of the firms. The bulk of acqui- differences between greenfield affiliates, sitions takes place within the first 10 years of brownfield affiliates, and domestic firms, the a firm’s lifetime, but that margin makes three samples need to be comparable to avoid acquired firms on average older than green- attributing age differences to ownership and field affiliates at the time of observation mode of entry. This requirement constrains figure 2.4, panel a). (­ this exercise to firms that are younger and When accounting for how long firms have smaller. been in operation, that is, at a given age, The six countries used for this study— greenfield affiliates are typically larger than China, Côte d’Ivoire, Indonesia, Moldova, brownfield affiliates in all sample countries. In Serbia, and Vietnam —represent a solid sam- turn, both types of multinational enterprises ple for the exercise. Covering various periods (MNEs) are significantly larger than their from the early 2000s to recently, they contain domestic counterparts ­ (figure 2.4, panel b). 6 6    Global Investment Competitiveness Report 2019/2020 FIGURE 2.4  Greenfield Affiliates Typically Have Larger Workforces than Brownfield Affiliates, Considering Age of the Firm a. Age distribution b. Employment premiums net of sector and agea 0.4 5 Density of kernel distribution 4 Controlled premium 0.3 3 0.2 2 0.1 1 0 0 0 10 20 30 40 China Côte Indonesia Moldova Serbia Vietnam d’Ivoire Age of acquisition (years) China Indonesia Vietnam Greenfield lower 95% CI / upper 95% CI Moldova Serbia Côte d’Ivoire Brownfield lower 95% CI / upper 95% CI Source: World Bank calculations, based on industrial censuses from six countries. Note: Industrial census data were analyzed from China, Côte d’Ivoire, Indonesia, Moldova, Serbia, and Vietnam (for data details, see annex 2A). CI = confi- dence interval. a. In panel b, the graph shows premia relative to the benchmark of a domestic firm, set at 0. “Domestic” refers to the firms that have not changed their own- ership from local to foreign at any point in the observed sample. Myriad dimensions and layers can be stud- Productivity, the value of the firm’s assets, ied under the umbrella of firm-level outcomes. exports, and diversification of production are In terms of development impact, more conventional measures of how competitive a emphasis is placed on impacts that reflect two firm is, while the extent to which these returns broad objectives: competitiveness (which cap- benefit local workers and suppliers can be tures productivity and returns to investment) measured by wages and imports, respectively. and inclusiveness (the extent to which these Starting from competitiveness, both types returns benefit a broad range of society). of multinationals are significantly more pro- These two objectives conceptually track two ductive than the average domestic firm in all pillars of sustainable development—economic six countries observed (figure 2.5). The value and social—to which typically a third pillar, of firms’ assets is a distinct feature of brown- environmental, is added. Firm-level data used field affiliates (figure 2.5, panel c); indeed, it for this exercise only shed light on the first is a major motivation for investment. two, although several hypotheses can be Valuable assets, however, are not reflected in made on the differential effect on the environ- significantly better labor productivity for ment of brownfield versus greenfield brownfield affiliates than greenfield affiliates investment. (figure 2.5, panel b): the averages across the The economic and social outcomes of two groups of firms are within the margin of foreign-owned firms differ substantially error, and in all countries are significantly from those of domestic firms (box 2.2). better than domestic firms. When it comes to These differences in turn help shift macro- internationalization, both brownfield and economic and social outcomes in host coun- greenfield affiliates have greater exposure tries, depending on the volume of investment than domestic firms, and comparable levels and the presence of complementary condi- in exports and imports (figure 2.5, panels d tions that facilitate the absorption of this and e), exhibiting closer supply linkages to foreign investment by firms, regions, and the domestic market in only one case: countries. Indonesia. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    67 BOX 2.2 The Evidence on Economic and Social Outcomes of Multinational Enterprises Although foreign direct investment (FDI) is gener- and North Africa, for example, multinationals dif- ally associated with positive effects on development, fer significantly in terms of export propensity and whether these effects will materialize is neither auto- geographical diversification because of the concen- matic nor monotonic; it varies considerably across tration of FDI in natural resource sectors. Multi- various types of enterprises, sectors, regions, and coun- nationals also contribute more to gender empower- tries, and it is highly dependent on mediating factors ment in this region than anywhere else by employing and local conditions. significantly more women in managerial positions A recent examination of the World Bank Enterprise than do domestic firms, potentially because of social Surveys highlighted systematic differences between and cultural differences between the home and host foreign multinational enterprises (MNEs) and domes- countries. tic firms across 63 countries over 10 dimensions that Differences in some key areas that drive competi- matter for development. Although there appears to tiveness (such as productivity, innovation, and skills be no striking trade-off between competitiveness and transfer) appear to increase with income, while pre- inclusiveness of foreign multinationals, their premi- miums in all other areas are greatest in lower-mid- ums over domestic firms differ substantially across dle-income or low-income markets, highlighting the regions and income groups. relevance of foreign multinationals for socioeconomic Relative to other regions, foreign MNEs estab- progress in these contexts. lished in eastern Europe and central Asia, for exam- Outcomes of foreign-owned firms, however, should ple, exhibit better outcomes than domestic firms not be confused with the aggregate development impact on most of the dimensions relating to competitive- of international business; they are one among a num- ness (such as productivity, outward orientation, and ber of such ­drivers that shift the macroeconomic and innovation) as well as inclusiveness (wages or provi- social outcomes of host countries. The actual aggre- sion of training). Foreign MNEs established in Latin gate impact of foreign multinationals on host countries America stand out in terms of productivity and remains dependent on the v ­ olume of investment and skills transfer, while in Sub-Saharan Africa, foreign the presence of complementary conditions that facili- MNEs stand out with respect to job expansion and tate their absorption by firms, regions, and countries. wages. These conditions include the policy environment, qual- The mix appears to be highly specific to the ity of local institutions and ­ fi nancial markets, sector type of multinationals each region attracts, includ- characteristics, and spatial colocation of domestic with ing the industry and investor motivations as well foreign firms. as the host economy conditions. In the Middle East Sources: Alfaro 2017; Lejárraga and Ragoussis 2018. Improvements in the competitiveness of firms, yet greenfield affiliates pay consider- firms acquired by foreign investors, how- ably more. ever, pass through only partially to workers. Why this discrepancy? Greenfield FDI The productivity premiums of both brown- adds production to the economy, so it boosts field and greenfield affiliates over domestic demand for labor, driving wages up for new firms should translate into broadly similar hires. Paying “efficiency” wages (those above wage premiums for workers. However, this market clearance) is also to be expected as is not the case: there appears to be a clear frontier firms enter new countries. On the sorting between greenfield, brownfield, and other hand, “sticky” wages (those that domestic firms in terms of wages (figure 2.5, respond slowly to changes in the perfor- panel a). Both brownfield and greenfield FDI mance of a company or the economy) would pay their workers better than domestic tend to keep wages of brownfield ventures at 6 8    Global Investment Competitiveness Report 2019/2020 FIGURE 2.5  Brownfield Affiliates and Greenfield Ventures Differ from the Average Domestic Firm a. Wages b. Labor productivity 8 0.08 0.06 6 Controlled premium Controlled premium 0.04 4 0.02 2 0 0 −0.02 China Côte Indonesia Moldova Serbia Vietnam China Côte Indonesia Moldova Serbia Vietnam d’Ivoire d’Ivoire c. Fixed assets d. Importers 2 0.4 Controlled premium Controlled premium 0.3 0 0.2 −2 0.1 0 −4 –0.1 China Côte Indonesia Moldova Serbia Vietnam Côte Indonesia Serbia Vietnam d’Ivoire d’Ivoire e. Exporters f. Multiple product offerings 0.3 1.0 0.8 Controlled premium Controlled premium 0.2 0.6 0.4 0.1 0.2 0 0 China Côte Indonesia Serbia Vietnam China Côte d’Ivoire Indonesia Vietnam d’Ivoire Greenfield Brownfield lower 95% CI / upper 95% CI lower 95% CI / upper 95% CI Source: World Bank calculations, based on industrial censuses from six countries. Note: For this figure, industrial census data were analyzed from China, Côte d’Ivoire, Indonesia, Moldova, Serbia, and Vietnam (further described in annex 2A). The graphs show premia relative to the benchmark of a domestic firm, set at 0. “Domestic” refers to the firms that have not changed their ­ownership from local to foreign at any point in the observed sample. Vertical bars indicate the confidence interval (CI). Industrial censuses vary in their ­coverage of different variables, and so some countries are missing from selected panels. lower levels that persist after the acquisition. firms evolve as well in directions that In the smaller countries in the sample, such depend on the country context and period as Côte d’Ivoire, Moldova, and Serbia, these of study (see annex 2C). In China, for wage premiums of multinational firms example, the average size of firms acquired are significantly greater than in bigger by foreign investors relative to the rest economies. increased over 1998–2007, while the oppo- As brownfield investment increases in a site happened in Serbia and Indonesia over host country, the characteristics of acquired 2006–13 and 2009–15, respectively. In both H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    69 China and Vietnam, the trend in average Average wages in the three categories of productivity over the periods studied has firms are clearly sorted, with brownfield affili- been toward acquisitions of firms that ates paying marginally less than greenfield resemble more the rest of firms in the econ- affiliates throughout the first years of their omy, while in Serbia and Indonesia, that has operations. Wages in foreign take-overs at the not been the case. General patterns in the end of the first five years of operations are evolution of acquired firm characteristics 40–50 percent higher than domestic firms, would require study of more country cases and the gap can reach 70 percent in greenfield over longer periods. affiliates (figure 2.6, panel a). The same pattern arises when it comes to importer and exporter status: greenfield affili- Differences in Growth Paths between ates are significantly more internationalized Brownfield Affiliates and the Average than brownfield affiliates, which in turn are Domestic Firm significantly more exposed to global mar- kets than domestic firms (figure 2.6, panels Foreign investors do not target the average e and f). Specifically, over the first five years of domestic firms for acquisition. So questions the firm’s operations, a brownfield affiliate is naturally arise as to whether brownfield 70–100 percent more likely to export than affiliates look different from the first day domestic firms, while greenfield affiliates are they are established and whether differentia- at least three times more likely to export tion is the result of (a) growth over their life- throughout the period. By the fifth year of times, or (b) direct influence of the foreign their operation, brownfield and greenfield investor. affiliates tend to significantly narrow their By comparing the growth and transforma- gap in terms of internationalization. tion paths of firms in the different categories Overall, much of the growth in the firms’ over several key outcomes such as employ- employment takes place within the first three ment, labor productivity, wages, and levels of years of their operations. After that time, internationalization, the evidence points to a firm sizes seem to stabilize. Both brownfield positive answer to both questions, to varying and greenfield affiliates begin with more extents. Track records of firms allow obser- ambitious undertakings, stabilizing at levels vations for only five to seven years after firms that are 15–25 percent larger than the aver- enter.5 A deeper analysis of transformation age domestic firm (figure 2.6, panel c). When following acquisition yields insights for only it comes to product offerings, while all types the same number of years. These first years of of firms diversify within the first five years of a firm’s life cycle capture important dynam- their operations, greenfield and brownfield ics: whether in developing countries or in affiliates accelerate their diversification more mature economies such as the United States, rapidly, and by the fifth year of their life- half of start-ups fail within that period. times end up with a profile that includes dif- Growth paths of brownfield affiliates and ferent activities (figure 2.6, panel d). greenfield ventures. Firms that get acquired Firm transformation after firm acquisition. by foreign investors look different at origin Growth paths show that acquisition targets from firms that do not. They start off larger start off with above-average potential, which and more productive from the first year of translates into better outcomes in the medium their operations, and they offer better wages to long term. But is it foreign ownership that for their workers than the average domestic improves firm performance, or rather advan- firm. These differences are not statistically tages at birth and the ability of firms to grow discernible at the very origin but become differently? apparent already within the first year of a This exercise examines changes in out- firm’s lifetime. comes for firms that transition into MNE 7 0    Global Investment Competitiveness Report 2019/2020 FIGURE 2.6  Brownfield and Greenfield FDI Firms Perform Better than Domestic Firms over the First Five Years of Operation a. Wage growth (US$, thousands, constant 2010), b. Value added per employee (US$, millions, constant 2010), Value added per employee (US$, millions, controlled average controlled average Wage growth (US$, thousands, constant constant 2010), controlled average 4 0.015 2010), controlled average 3 0.010 2 0.005 1 0 0 1 2 3 4 5 0 1 2 3 4 5 Number of years after rm birth Number of years after rm birth Green eld MNE Domestic Brown eld MNE c. Log (employees), controlled average d. Probability of multiple product o erings (%), controlled average Log (employees), controlled average Probability of multiple product o erings, 3.0 0.20 2.8 controlled average (%) 0.15 2.6 2.4 0.10 2.2 2.0 0.05 0 1 2 3 4 5 0 1 2 3 4 5 Number of years after rm birth Number of years after rm birth Green eld MNE Domestic Brown eld MNE e. Probability of exports (%), controlled average f. Probability of imports (%), controlled average Probability of imports, controlled Probability of exports, controlled 0.5 0.7 0.4 0.6 average (%) average (%) 0.5 0.3 0.4 0.2 0.3 0.1 0.2 0 1 2 3 4 5 0 1 2 3 4 5 Number of years after rm birth Number of years after rm birth Green eld MNE Domestic Brown eld MNE Source: World Bank calculations, based on industrial censuses from six countries. Note: For this figure, industrial census data were analyzed from China, Côte d’Ivoire, Indonesia, Moldova, Serbia, and Vietnam (as further described in annex 2A). “Domestic” refers to the firms that have not changed their ownership from local to foreign at any point in the observed sample. Growth paths of firm outcomes can be captured in a simple framework using an interaction between indicators of firm group (greenfield, brownfield, domestic) and years after entry in the following specification: y ics = groupi β ⋅ dt + α cs + δ dit0 + ε ics. The sample is restricted to cohorts whose entry is observed. To account for differences that might be driven by country characteristics, sector composition, and macroeconomic trends, the regressions also control for country-sector fixed effects (FE) and cohort fixed effects. Wage growth paths are calculated using constant deflated values in US dollars. Regression includes country FE*2 -digit sector FE as well as cohort start year dummies. Vertical bars indicate the margin of error. FDI = foreign direct investment; MNE = multinational enterprise. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    71 status relative to firms of similar profile that employment in brownfield affiliates expands by remain domestic. The procedure involves approximately 4 percent, compared with matching every firm acquired by foreign 1.5 percent in domestic firms with similar char- investors to a firm that remains domestic and acteristics. The value of firms’ assets after the has similar characteristics (in terms of acquisition follows a similar path. In addition, employment, age, and sector) during the wages in brownfield affiliates appear on average year before the acquisition takes place. 6 to increase, marginally widening the differences Average outcomes in the two groups of firms with domestic firms (figure 2.7, panel b). are then tracked over the years before and Transformation paths are highly depen- after the acquisitions in a framework identi- dent on the context. The options available to cal to the growth paths used in the previous domestic firms of similar characteristics differ section. across markets, and so do the limitations in Contrary to conventional belief about the foreign owners’ decision making—all of potential job-destroying effects of M&A, which affect the value added from foreign employment in newly acquired firms grows at capital. Large upper-middle-income countries similar or often faster rates than the control (notably, China) offer more growth opportu- group of domestic firms for the first few years nities to firms independently of the origin of after acquisition (figure 2.7, panel a). More spe- their capital. Indeed, transformation paths in cifically, two years after acquisition, the average the sample of five smaller or lower-income FIGURE 2.7  Targeted Firms Improve Their Performance after Acquisition by Foreign Investors a. Employment (aggregate sample) b. Wage (aggregate sample) 5.2 3.5 Log (employees), controlled average Average wage (US$, thousands, constant 5.0 3.0 2010), controlled average 4.8 2.5 4.6 4.4 2.0 −2 −1 0 1 2 3 4 5 6 7 −2 −1 0 1 2 3 4 5 6 7 Years post-acquisition Years post-acquisition c. Labor productivity (without China) d. Wage (without China) Value added per employee (US$, millions, 0.05 6 constant 2010), controlled average constant 2010), controlled average Average wage (US$, thousands, 0.04 5 0.03 4 0.02 3 0.01 −2 −1 0 1 2 3 4 5 −2 −1 0 1 2 3 4 5 Years post-acquisition Years post-acquisition Domestic Foreign acquisitions Figure continues next page 7 2    Global Investment Competitiveness Report 2019/2020 FIGURE 2.7  Targeted Firms Improve Their Performance after Acquisition by Foreign Investors (continued) e. Exporter (without China) f. Product diversi cation (without China) 0.40 0.8 offerings (%), controlled average Probability of multiple product Probability of exports (%), 0.35 0.7 controlled average 0.30 0.6 0.25 0.5 0.20 0.4 0.15 0.3 −2 −1 0 1 2 3 4 5 −2 −1 0 1 2 3 4 5 Years post-acquisition Years post-acquisition Domestic Foreign acquisitions Source: World Bank calculations, based on industrial censuses from six countries. Note: “Without China” refers to results for five countries: Côte d’Ivoire, Indonesia, Moldova, Serbia. and Vietnam. Industrial census data analyzed for this figure are further described in annex 2A, table 2A.1. “Domestic” refers to the firms that have not changed their ownership from local to foreign at any point in the observed sample. Growth paths of firm outcomes can be captured in a simple framework using an interaction between indicators of firm group (brownfield, domestic) and years after entry in the ­following specification: y ics = groupi β ⋅ dt + α cs + δ dit0 + ε ics. The sample is restricted to equinumerous groups of brownfield affiliates and domestic firms matched on the basis of employment, age, and sector, in the year before the firm was acquired. To account for differences that might be driven by country characteristics, sector composition, and ­macroeconomic trends, the regressions also control for country-sector fixed effects (FE) and year fixed effects. Regression includes country FE* 2-digit sector. Vertical bars ­indicate the margin of error. countries excluding China (Côte d’Ivoire, sectoral effects conceals important variations Indonesia, Moldova, Serbia, and Vietnam) of benefits that are specific to these levels. confirm greater value addition of foreign cap- Benefits from greater market access and ital in some dimensions (figure 2.7, panels diversification, for example, are expected to c–f). Acquired firms in these countries signifi- be more pronounced in manufacturing and cantly increase their ability to access markets primary commodities than in services. There through exports and product diversification, are also a number of dimensions—such as while wages adjust to a higher level than liabilities and the skill composition of work- expected. Specifically, by the fifth year follow- force, or relative measures such as export-to- ing acquisition, the average brownfield affili- sales ratio—where differences are less ate has increased wages by 10–30 percent discernible in the sample studied following relative to no adjustment in the wages of acquisition. domestic firms. The rise in labor remunera- Overall, and despite important data limi- tion reflects, at least partly, adjustment to a tations, the evidence supports a value addi- marginally higher level of productivity. These tion of foreign investors in some key premiums for new brownfield affiliates dimensions related to development, such as ­ persist for the observed period following employment and market access comple- acquisition. menting systemic benefits of acquisitions These findings come with caveats. That reported from other sources (box 2.3). The ownership shares are unobserved dilutes evidence also suggests that these advantages potentially stronger impact in cases in which are more pronounced in markets that are foreigners gain majority or full control of less developed, smaller, or both, which is a domestic firms.7 Moreover, that the reported general hypothesis warranting further estimates are conditional on country and investigation. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    73 BOX 2.3 Systemic Benefits of Brownfield Foreign Direct Investment: Telecommunication Acquisitions in Africa Starting in the mid-1990s, state-owned telecommu- them access to free electricity 24 hours a day. In addi- nications operators in Africa were privatized in large tion, it provided four incubators for entrepreneurs numbers, with the vast majority acquired by for- in four countries and six accelerators as well as eign brownfield investors. Acquirers included global programs and e-­ e ducation services for digital skills firms in the industry from Europe such as Orange, to schoolchildren, university students, and young Vodafone, and Portugal Telecom, but also developing- professionals in partnership with local ministries country multinational firms such as Maroc Telecom, of ­e ducation. Vodafone invested in Healthline South Africa’s MTN, and India’s Bharti Airtel. The in Ghana, the first medical call center in Africa. sector grew rapidly, with a subsequent phase of boom- The rapid growth has helped boost the economy ing greenfield investment through license acquisitions and employment across Africa. In 2015, mobile tech- and a wave of new brownfield investment in existing nologies and services generated 6.7 percent of Africa’s operators over the past decade. gross domestic product (GDP), or around US$150 Firms that were acquired by foreign investors often billion in economic value. Africa’s mobile ecosystem invested in network expansion and upgrade, espe- directly supported over 1 m ­ illion jobs. The expansion cially for mobile and fixed internet access. Follow- of the sector has also supported an additional 2.4 mil- ing the acquisition of Ghana Telecom, for example, lion jobs indirectly through production inputs, wages, Vodafone invested around US$1 billion in improving public funding, and profits spent in other sectors. the digital infrastructure in the country. There were This development has strengthened economic activ- 400 sites in 2008, increased to more than 2,000 sites ity in other industries through improved i ­nformation within six years. sharing and increased access to data and mobile Employment reductions during an initial restruc- broadband. turing of the acquired firms have not been rare but The outcomes of foreign acquisitions have been were often followed by fast growth and subsequent more spectacular in some countries than in others, job creation. The acquisition of Burkina Faso’s Airtel with the context making a difference. Gasmi et al. by Orange in 2016, for example, was associated with (2013) note poor outcomes of some acquisitions in a drop of the enterprise’s full-time headcount from resource-scarce, landlocked African countries as well 291 to 259, followed by a 10 percent average annual as in resource-rich African countries because of weak increase annually that increased the headcount to contractual design, inadequate policy enforcement in ­ 365 by 2019. Jobs offered to local populations by the infrastructure sector, and insufficient aggregate these multinational firms often served as a vehicle demand. The bundling of telecommunications with for skills upgrade: in 2016, for example, 80 percent banking services allowed some incumbents to compete of the 20,000 employees of the Middle East and successfully against brownfield ventures in some mar- Africa branch of Orange received training averaging kets. Sector-specific taxation imposes additional costs 26 hours. to the investor in others. G ­ enerally, in the absence Engagement of the operators in areas outside of strong state capacity, competition is a necessary the strict boundaries of their markets has also been complement to foster development outcomes of such ­ common. Orange funded the Africa Cup of Nations investments. football tournament in 2013 and supported more than Sources: Estrin and Pelletier 2018; Gasmi et al. 2013; GSMA Intelligence 2016; GSM 30 rural radio stations in 13 counties by providing Association (GSMA) financial statements; Orange 2017; Staff 2015. 7 4    Global Investment Competitiveness Report 2019/2020 Policy Considerations for Sensitivity of Brownfield FDI to Brownfield FDI Investment Restrictions Statutory restrictions on foreign investment A policy framework for foreign investment generally apply to all modes of entry. comprises incentives, rules, and restrictions Thresholds on foreign equity (for example, that firms need to comply with through their investment screenings), restrictions on life cycle. These policy elements are not sys- movement of people, or repatriation of prof- tematically designed to favor one mode of its do not discriminate between newly estab- foreign entry over another. Yet, typically, lished firms and brownfield ventures. They some are more relevant to investors entering apply regardless of whether equity was through a particular mode or to sectors acquired or created. By contrast, licensing where brownfield investment takes place requirements, or limitations on the number more intensively. of firms in a market segment, are less rele- vant to firms that are already operating and, Investment Incentives by Mode of Entry by extension, to brownfield investment tar- geting those firms. Investment incentives such as tax credits, Although most restrictions apply equally preferential rates, or subsidies are commonly to all modes of entry, brownfield investment made conditional on characteristics of the tends to be more sensitive than greenfield firm. Most developing countries that grant investment to the general severity of restric- tax holidays (77 percent) condition them on tions and to certain types of restrictions in location requirements within the country particular. Evidence in the literature, while (World Bank 2018). Less common is the limited, finds that a similar reduction in total requirement to export or to sell to exporting investment restrictions is associated with a firms and engage in research and develop- greater increase in cross-border acquisitions ment (R&D). The mode of entry can be than in greenfield foreign investment, inde- ­ s pecified—as was done in the Czech pendent of a country’s size or level of develop- Republic, Mozambique, or South Africa, ment (Lee 2016; Mistura and Roulet 2019). where preferential treatments and conces- This pattern could be partly explained by the sions are explicitly applicable only to green- greater frequency of acquisitions in services field investments or acquired firms that industries, which are also subject to higher plan on expanding production capacity— restrictions. although no widespread discrimination has Restrictions that are more frequently been reported in the literature over that encountered in sectors where brownfield dimension.8 investment takes place, such as services, Incentives are not often a deal breaker for could have greater impact on that mode of foreign investment. They might add an entry. Restrictions on movement of people important element to the equation when and on board nominations are examples that other, more fundamental components are deter investment by hindering participation present. Asset-seeking investment, such as of foreign investors in development of the FDI in natural resource sectors, tends to be firm. This type of restriction can be particu- less responsive to incentives (World Bank larly stringent when investment originates 2018), which suggests that any discrimina- from other developing countries (Borchert, tion by mode of entry might be less relevant. Gootiiz, and Mattoo 2012). But more generally, the signal that some for- The limited evidence to date confirms that eign investors are less welcome than others cross-border M&A is highly sensitive to can hurt the growth prospects of countries variable requirements that are left to the that would otherwise benefit from brownfield ­ discretion of authorities, such as investment FDI in the medium term. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    75 screenings or economic needs tests competition is loosely delimited in the litera- (Lee 2016; Mistura and Roulet 2019). These ture, although there is general agreement procedures end up being more relevant to that the feedback loops between the two are brownfield investment because of the pre- important. vailing ambiguity over the value this type of Competition is typically enforced by an investment brings to the host economy. It is independent authority that has the capacity to on the basis of this evidence that the rest of detect anticompetitive behavior, such as collu- the policy discussion emphasizes investment sion and abuses of dominance, as well as the screenings and variable requirements. power to penalize misbehavior. This authority Investment screenings vary in scope and typically requires notification of prospective depth across countries; they tend to include M&A, issues approvals on the basis of its a range of administrative burdens in terms reviews of likely effects, and proposes reme- of contracts and can be applied in non- dies to minimize the anticompetitive effects of transparent ways. Russia, for example, market consolidation. requires a national security review for for- In principle, the absence of proper compe- eign investments in more than 40 sectors tition enforcement can deter entry for both but provides no criteria for evaluating an greenfield and brownfield investors. Private application on these grounds. In Tunisia, impediments to acquisitions, such as cross- multiple sectors require preauthorization holding or tactical obstacles by incumbents, for foreign acquisitions of a majority share are common, all falling under the realm of of a company, yet again without specific competition barriers (Nolan 2019). These criteria for review. barriers can be addressed through a well- Many countries do not screen prospective functioning competition authority that limits FDI extensively, although notable recipients the ability of incumbents to deter entry. of brownfield investment do, including In practice, the rules and their enforce- China, India, Malaysia, Mexico, Russia, ment often pose additional burden to inves- Tunisia, and Vietnam.9 Investment screen- tors. Although more than 120 countries and ings, moreover, are significantly on the rise in regional blocs have relevant legal frame- higher-income economies (UNCTAD 2019). works, implementation is limited in countries By submitting more sectors or activities to whose competition frameworks or authori- review, lowering the triggering thresholds, or ties are more recent. In these cases, the broadening the definition of foreign invest- administrative burden of information ment, countries increasingly strengthen these requirements, their cost, and the time they mechanisms. The risk of these policies is that take can be an impediment to an acquisition they shape narratives and establish prece- of a firm. To minimize undue burdens on the dents that will likely influence the policy investor, many competition authorities con- stance of developing countries in the future, duct two-phase investigations or have a for- either through reciprocity or established mal simplified notification procedure for practice. certain transactions that allow for fast-track decisions (World Bank 2016). This procedure filters out investments that are less likely to Competition Frameworks: Another have an impact on competition. Layer of Frictions The independent application of competi- Several types of brownfield investment aim- tion reviews and investment screenings by ing at control or a vertical merger of acquired different authorities, under different frame- firms are also the subject to review from works, can be the source of numerous competition authorities because of their failures. Of 40 developing economies across ­ potential impact on market dynamics. The regions surveyed by the World Bank in 2016, nexus between foreign investment and two-thirds keep the two processes separate.10 7 6    Global Investment Competitiveness Report 2019/2020 The frameworks are often disconnected not • Increase the efficiency of competition only in terms of the criteria used to evaluate policies to reduce the administrative bur- transactions but also in their procedural den of controls, strengthen enforcement requirements. Generally, although competi- capacity, and reduce the scope for tactical tion reviews are often more focused on impediments to foreign acquisitions by ­ efficiency considerations and potential mar- incumbents. The objective here is to both ket effects, public interest considerations are safeguard competition and minimize the pervasive in investment assessment frame- burden of administrative procedures on works. Therefore, the degree of subjectivity business by using public resources more of investment reviews is higher, and so is the effectively, ultimately fostering the inter- uncertainty of the process. The lack of pre- national exposure of markets. Competi- dictability can be potentially discouraging for tion authorities are often ill equipped to brownfield investments that qualify for such deliver these mandates in lower-income review. countries (Berger, Gsell, and Olekseyuk An alternative proposed by some practi- 2019; World Bank 2016). tioners is to merge the two processes into one, • Enhance cooperation between competi- in a so-called single-review model (Bakhoum tion and investment authorities with a and Fox 2019). An example of full conver- view to reducing inconsistencies between gence on the substantive criteria to review different time lines for reviews, different mergers and investment is South Africa, thresholds and considerations triggering where the investment framework specifically review, or their mandatory nature. This delegates the review to the competition con- cooperation stands to facilitate both pro- trol framework.11 The adoption of that model cesses without necessarily compromising remains the exception rather than the rule in their content: independent assessments practice. of competition effects are essential to Overall, a policy framework that supports market well-being. A wider application brownfield investment should emphasize the of the single-review model, along with following: improvements in its design, could poten- tially prove beneficial in countries where • Streamline investment screening mecha- brownfield investment is rising fast. nisms and approvals, focusing on trans- • Avoid differential treatment of brown- parency, well-defined criteria, consistency field multinationals with respect to with competition frameworks, reliable investment incentives. Although more time frames for reviews and decisions, information is necessary to assess the and judicial redress for the investor . extent of discrimination in this area, and Clearly defined methodologies to assess its potential cost, eligibility for incentives “public interest” could expand the is an essential element of a supportive empirical basis underpinning the assess- framework for brownfield investment. ments and increase the efficiency of the It is also an element that ensures that the process. In addition, a multilateral right signals are being sent to investors investment facilitation framework would likely to enter in that mode. go a long way toward improving the • A d d re s s o p e ra t i o n a l b a r r i e r s t o transparency and predictability of a multinationals, such as to movement range of other administrative procedures of people and board nominations, to to deal with investments.12 International facilitate participation of foreign investors rules for investment facilitation are rare in development of the firm. Many of these (Polanco Lazo 2018). Only a few inter- restrictions are not specific to brownfield national investment agreements cover investment but are more frequently rules on the facilitation of business encountered in sectors where this mode of activity. entry is more intensive, such as services, H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    77 and can be ­ particularly stringent with therefore be more relevant for fast growing, respect to investment originating from lower-middle-income economies where the other developing countries. share of foreign acquisitions is rising. They could also be more relevant for attracting The reasons why brownfield investment is investment into sectors where brownfield more common in higher-income economies investment takes place more frequently, such extend beyond the specific considerations as agriculture or services. Governments in already discussed. By engaging actively in these contexts have the means to foster the the process of improving institutions and potential of foreign acquisitions by addressing laying strong market foundations, govern- administrative frictions, enhancing the pre- ments are supporting brownfield investment dictability of controls, and safeguarding without that necessarily being the explicit competition. objective. Stronger institutions ensure the Future research should expand the evi- protection of intellectual property, respect of dence base with analysis of outcomes of contractual arrangements, and shareholders’ acquisitions over longer periods of time and minority rights, all of which make this mode in different country contexts. Additional evi- of entry more appealing. The growth of via- dence would also be warranted on the effects ble stock markets and successful firms that of acquisitions of intangible as opposed to are attractive for foreign investment can tangible assets, as well as on the develop- have the same effect. Causality could go in ment effects on the acquiring firm, motivated both directions, as brownfield acquisitions by the booming outward investment from can potentially bring longer-term benefits major developing economies to the rest of like stronger corporate governance, which in the world. All these extensions will allow a turn can foster stronger institutions (Bris, more nuanced case to be made for brown- Brisley, and Cabolis 2008). field investment in the development process. Finally, assessing the extent of “masked effects” of this mode of entry on domestic firms that would otherwise exit the market, Concluding Remarks saving jobs and revenue, could shed light on This study documents the characteristics, the ways that brownfield investment con- growth paths, and outcomes of multina- tributes to sustaining economic activity that tional firms established through brownfield other modes of foreign entry cannot. investment relative to those established Coupled with a systematic mapping of com- through greenfield investment and domestic petition and investment screening frame- firms in six developing countries to shed works, this evidence could improve technical additional light on their contribution. A key assistance to developing countries and takeaway from this analysis is that brown- attract more investment that works for field affiliates add value to the development development. process in ways that do not differ in their essence from those established through greenfield investment. Annex 2A. Data Sources Although more evidence is necessary to establish general conclusions and macroeco- UNCTAD Aggregate Data nomic effects, the experience of the six coun- Cross-border mergers and acquisitions tries analyzed in this study suggests that (M&A) statistical sources from the United foreign acquisitions could be adding more N a t i o n s C o n f e r e n c e o n Tr a d e a n d value in markets at the lower end of the devel- Development (UNCTAD) are based on infor- opment spectrum—that is, in countries where mation reported by Thomson Reuters. Such most FDI still takes place through greenfield M&A conform to the standard definition of investment. The findings reported here could foreign direct investment (FDI) as far as the 7 8    Global Investment Competitiveness Report 2019/2020 equity share is concerned. However, the data or c­ ensus longitudinal data from the six also include purchases via domestic and inter- developing countries investigated in this national capital markets, which should not be study: China, Côte d’Ivoire, Indonesia, considered to be FDI flows. Cases of round- Moldova, Serbia, and Vietnam. For the tripping (also known as round-trip transac- growth path analysis, this study analyzed tions) are also considered on the basis of the all cohorts of firms whose entry as well as immediate acquiring country and immediate survival for at least five years was observed target country principles. (table 2A.1). Data on announced greenfield FDI proj- Cross-country, firm-level data have ects sourced from UNCTAD are based on important limitations. One of the most crit- the information provided by fDI Markets, a ical concerns is the issue of comparability of service of the Financial Times (https://www​ ­ employment and capital measures, which .fdimarkets.com/). fDI Markets tracks all can vary from one survey to another. new investment projects and expansion of Harmonization has been undertaken to existing investments but does not include address this issue. However, the analysis is information on the equity participation by necessarily constrained by the data avail- investors. This suggests that the data may able (or not available) in the raw surveys or include investments that are not qualified censuses. In addition, all surveys available ­ as FDI. Joint ventures are also included for this study record foreign ownership as a only where they lead to a new physical binary indicator (yes or no), bundling port- operation. folio investment of less than 10 percent with FDI (>10 percent of total assets) and not allowing a separate treatment of Industrial Censuses majority-owned affiliated of foreign firms, ­ The cross-country microeconomic evi- where effects of acquisitions could be more dence draws firm and establishment survey pronounced. TABLE 2A.1  Industrial Census Data Used for Cross-Country Analysis Full sample Sample of firms whose entry is observed GDP per Observations Observations capita Time Brownfield Brownfield Country (2017) coverage Restrictions Firms Total MNE MNE Firms Total MNE MNE China 8,827 1998–2007 Manufacturing firms 570,108 2,048,525 421,544 32,759 292,978 829,869 173,041 10,184 (legally independent subsidiaries) with sales ≥ RMB 5 million Côte d’Ivoire 1,662 2003–13 Firms in all sectors 40,424 75,326 11,127 7,937 29,380 45,541 6,063 2,290 Indonesia 3,847 2010–15 Manufacturing estab- 33,131 169,324 15,985 5,014 6,768 19,097 1,363 712 lishments with at least 20 employees (L ≥ 20) Moldova 2,290 2004–14 Firms in all sectors 31,591 122,423 12,895 1,566 31,295 120,699 12,874 1,566 Serbia 5,900 2006–16 Firms in all sectors, 35,402 159,487 14,029 827 10,698 31,798 3,880 171 with at least 6 employees (L ≥ 6) Vietnam 2,343 2007–12 Full census of large firms, 504,916 1,573,373 13,248 5,858 289,779 777,300 2,641 1,797 limited information on small firms Source: Data provided by country authorities to the World Bank Group. Note: GDP = gross domestic product; L = number of employees; MNE = multinational enterprise; RMB = renminbi. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    79 Annex 2B. Relative Importance of Country-Level versus Sectoral Factors in Favoring Greenfield or Brownfield Investment TABLE 2B.1  Variance Decomposition Contribution to Source Partial SS Df F Prob > F (%) model SS (%) Model 1,042.00 1,327 13.21 0.00 42.9 Country 21.38 5 71.96 0.00 2.1 Sector 8.49 86 1.66 0.01 0.8 Year 18.61 17 18.42 0.00 1.8 country*sector 36.16 263 2.31 0.00 3.5 country*year 237.41 31 128.87 0.00 22.8 sector*year 124.35 925 2.26 0.00 11.9 Source: World Bank calculations, based on industrial censuses. Note: Industrial census data came from six countries: China, Côte d’Ivoire, Indonesia, Moldova, Serbia, and Vietnam (further described in annex 2A). Variance components are estimated on the linear probability of a multinational firm transitioning to foreign ownership from domestic status at some point during the period it is observed; on country, sector, and year fixed effects; and on their two-way interactions. Df = degrees of freedom; F = F-value; SS = sums of squares. Annex 2C. Evolution of Brownfield MNE Characteristics over Time FIGURE 2C.1  Evolution of Brownfield MNE Characteristics over Time 4.0 2012 2011 2009 Size (employment ratio of brownfield MNE avg. to the rest) 2010 3.5 2010 2015 2009 3.0 2007 2006 2011 2014 2012 2009 2008 2.5 2011 2013 2010 2013 2012 2007 2.0 2006 2005 2004 2003 1.5 2002 2001 2000 1999 1998 1.0 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 2.3 2.5 Labor productivity (ratio of brownfield MNE avg. to the rest) China Serbia Vietnam Indonesia Source: World Bank calculations, based on industrial censuses from four countries. Note: Industrial census data analyzed for this figure are further described in annex 2A, table 2A.1. MNE = multinational enterprise. 8 0    Global Investment Competitiveness Report 2019/2020 Notes 7. In a seminal contribution 10 years ago, Arnold and Javorcik (2009) reported that 1. Notable exceptions of research focusing on foreign acquisitions of more than 20 ­percent employment and productivity of acquired equity had stronger effects in nearly 400 firms in the context of a single develop- establishments in Indonesia: an average ing country are Gong, Görg, and Maioli 24 percent increase in employment and (2007) (China); Arnold and Javorcik (2009) 40 percent increase in wages within the first (Indonesia); Lipsey, Sjöholm, and Sun (2013) two years following acquisition. (Indonesia); and Bircan (2019) (Turkey). 8. In South Africa, for example, the addi- 2. Nocke and Yeaple (2008) find this for multi- tional investment allowance for industrial national enterprises (MNEs) making brown- policy projects may not exceed R 900 field investments in the United States, while million for any greenfield project with a Raff, Ryan, and Stähler (2011) find a similar preferred status, R 550 million for any pattern in Japan. other greenfield project, R 550 million for 3. Geluebcke (2015) shows the negative impacts any brownfield project with a preferred of foreign take-overs on employment in status, or R 350 million for any other Germany. Bellak, Pfaffermayr, and Wild brownfield project. (2006), Martins and Esteves (2008), and 9. Data on FDI screening come from the World Mattes (2010) find no statistically signifi- Development Indicators database and the cant impact of foreign acquisitions on the OECD FDI Restrictiveness Index 2017 employment reduction in Austria, Brazil, and (https://www.oecd.org/investment/fdiindex​ Germany, respectively. Yet foreign ownership .htm). is found to contribute to formal employment 10. According to a review of country-­ specific in developing countries like Indonesia (Lipsey, merger control and foreign investment Sjöholm, and Sun 2013) and in Nigeria’s regimes in 40 developing economies in manufacturing sector (Inekwe 2013), as well multiple regions in 2016, 86 percent of as China (Gong, Görg, and Maioli 2007), countries have a predefined merger con- in addition to some high-income coun- trol framework, and only 70 percent have tries including New Zealand (Fabling and investment review mechanisms (World Bank Sanderson 2014), Portugal (Almeida 2007), 2016). Where investment control frame- and Sweden (Bandick and Görg 2010). works exist, the institutions in charge of the 4. Data are from the World Bank’s Doing Business review as well as the regulatory framework Indicators database (accessed December 2019), applied tend to be both economywide as https://www.doingbusiness.org/. well as sector specific. 5. The growth and evolution of firms continue 11. The exceptions are certain sectors that after the first five years of their operations. require government approval for foreign par- However, data constraints allow examina- ticipation, including energy, mining, bank- tion of cohorts of firms through the end of ing, insurance, and defense. the seventh year without major loss of sta- 12. In 2017, during the 11th World Trade tistical power. The paths in cohorts with that Organization (WTO) Ministerial Conference duration do not differ substantially in any of in Buenos Aires, 70 WTO members adopted the dimensions discussed. a “Joint Statement on Investment Facilitation 6. Propensity score matching for this exercise for Development,” announcing discussions implements the full Mahalanobis matching toward a multilateral framework on invest- based on employment, age, sector, and year to ment facilitation (WTO 2017). The discus- adjust for pretreatment observable differences sions aim primarily at achieving transparency between a group of firms targeted by foreign and predictability of investment measures; investors and firms that remained domestic streamlining and speeding up administrative throughout the observed period. Matching is procedures and requirements; and enhancing performed separately for each country, one- international cooperation, information shar- to-one with the nearest neighbor the year ing, the exchange of best practices, and rela- before acquisition takes place. Weights are tions with relevant stakeholders, including not used in the transformation paths. dispute prevention. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    81 References Entry?” Canadian Journal of Economics 45 (3): 1083–124. Alfaro, Laura. 2017. “Gains from Foreign Direct Bircan, Çağatay. 2019. “Ownership Structure and Investment: Macro and Micro Approaches.” Productivity of Multinationals.” Journal of T h e Wo r l d B a n k E c o n o m i c R e v i e w International Economics 116: 125–43. 30 (Supplement 1): 2–15. Borchert, Ingo, Batshur Gootiiz, and Aaditya Almeida, Rita. 2007. “The Labor Market Mattoo. 2012. “Policy Barriers to International Effects of Foreign-Owned Firms.” Journal of Trade in Services: Evidence from a New International Economics 72 (1): 75–96. Database.” Policy Research Working Paper Alquist, Ron, Rahul Mukherjee, and Linda 6109, World Bank, Washington, DC. Tesar. 2016. “Fire-Sale FDI or Business as Bris, Arturo, Neil Brisley, and Christos Cabolis. Usual?” Journal of International Economics 2008. “Adopting Better Corporate Governance: 98: 93–113. Evidence from Cross-Border Mergers.” Journal Arnold, Jens Matthias, and Beata S. Javorcik. of Corporate Finance 14 (3): 224–40. 2009. “Gifted Kids or Pushy Parents? Foreign Burger, Martijn J., and Elena I. Ianchovichina. Direct Investment and Plant Productivity in 2017. “Surges and Stops in Greenfield and Indonesia.” Journal of International Economics M&A FDI Flows to Developing Countries: 79 (1): 42–53. Analysis by Mode of Entry.” Review of World Bakhoum, Mor, and Eleanor M. Fox. 2019. Making Economics 153 (2): 411–32. Markets Work for Africa: Markets, Development, Calderón, Cesar, Norman Loayza, and Luis and Competition Law in Sub-Saharan Africa. Servén. 2004. “Greenfield Foreign Direct Oxford: Oxford University Press. Investment and Mergers and Acquisitions: Balsvik, Ragnhild, and Stefanie A. Haller. 2010. Feedback and Macroeconomic Effects.” World “Picking ‘Lemons’ or Picking ‘Cherries’? Bank Policy Research Paper 3192, World Bank, Domestic and Foreign Acquisitions in Washington, DC. Norwegian Manufacturing.” Scandinavian Chari, Anusha, Paige P. Ouimet, and Linda Tesar. Journal of Economics 112 (2): 361–87. 2010. “The Value of Control in Emerging Bandick, Roger, and Holger Görg. 2010. “Foreign Markets.” Review of Financial Studies 23 (4): Acquisition, Plant Survival, and Employment 1741–70. Growth.” Canadian Journal of Economics Claeys, Sophie, and Christa Hainz. 2007. 43 (2): 547–73. “Acquisition versus Greenfield: The Impact of Bandick, Roger, and Patrik Karpaty. 2007. the Mode of Foreign Bank Entry on Information “Foreign Acquisition and Employment Effects and Bank Lending Rates.” Working Paper in Swedish Manufacturing.” Research Paper Series No. 210, Sveriges Riksbank (Central No. 2007/35, University of Nottingham, U.K. Bank of Sweden), Stockholm. Beattie, Alan. 2014. “Foreign Direct Investment: Conyon, Martin J., Sourafel Girma, Steve It’s Not All Good.” Financial Times, October Thompson, and Peter W. Wright. 2002. “The 22 (accessed March 23, 2019), https://www​ Productivity and Wage Effects of Foreign .ft.com/content/6f71229d-d74d-34fa-a30d​ Acquisition in the United Kingdom.” Journal of -39e4ac07de8b. Industrial Economics 50 (1): 85–102. Bellak, Christian, Michael Pfaffermayr, and Curran, Louise, Ping Lv, and Francesa Spigarelli. Michael Wild. 2006. “Firm Performance after 2017. “Chinese Investment in the EU Ownership Change: A Matching Estimator Renewable Energy Sector: Motives, Synergies Approach.” Applied Economics Quarterly and Policy Implications.” Energy Policy 101: 52 (1): 29–54. 670–82. Berger, Axel, Sebastian Gsell, and Zoryana Davies, Ronald B., Rodolphe Desbordes, and Olekseyuk. 2019. “Investment Facilitation for Anna Ray. 2018. “Greenfield versus Merger Development: A New Route to Global Investment and Acquisition FDI: Same Wine, Different Governance.” Briefing Paper No. 5/2019, Bottles?” Canadian Journal of Economics German Development Institute (DIE), Bonn. 51 (4): 1151–90. Bertrand, Olivier, Katariina Nilsson Hakkala, di Giovanni, Julian. 2005. “What Drives Capital Pehr-Johan Norbäck, and Lars Persson. 2012. Flows? The Case of Cross-Border M&A “Should Countries Block Foreign Takeovers Activity and Financial Deepening.” Journal of of R&D Champions and Promote Greenfield International Economics 65 (1): 127–49. 8 2    Global Investment Competitiveness Report 2019/2020 Erel, Isil, Rose Liao, and Michael Weisbach. Sebastian Edwards, 43–58. Chicago: University 2011. “Determinants of Cross-Border Mergers of Chicago Press. and Acquisitions.” Journal of Finance 67 (3): Lee, Hyun-Hoon. 2016. “Policy Factors 1045–82. Influencing FDI Inflows to Developing Estrin, Saul, and Adeline Pelletier. 2018. Countries.” Background paper for Asian “Privatization in Developing Countries: Economic Integration Report 2016: What What Are the Lessons of Recent Experience?” Drives Foreign Direct Investment in Asia and The World Bank Research Observer 33 (1): the Pacific? Manila: Asian Development Bank. 65–102. Lejárraga, Iza, and Alexandros Ragoussis. 2018. Fabling, Richard, and Lynda Sanderson. 2014. “Beyond Capital: Monitoring Development “Foreign Acquisition and the Performance of Outcomes of Multinational Enterprises.” Policy New Zealand Firms.” New Zealand Economic Research Working Paper 8686, World Bank, Papers 48 (1): 1–20. Washington, DC. Gasmi, Farid, Alexis Maingard, Paul Noumba Lichtenberg, Frank R., and Donald Siegel. 2000. Um, and Laura Recuero Virto. 2013. “Productivity and Changes in Ownership of “The Privatization of the Fixed-Line Manufacturing Plants.” In Innovation, Evolution Telecommunications Operator in OECD, Latin of Industry and Economic Growth, edited by America, Asia, and Africa: One Size Does Not David B. Audretsch and Steven Klepper, 322–62. Fit All.” World Development 45: 189–208. Cheltenham, U.K.: Edward Elgar Publishing. Geluebcke, John P. 2015. “The Impact of Foreign Lipsey, Robert E., Fredrik Sjöholm, and Jing Sun. Takeovers: Comparative Evidence from Foreign 2013. “Foreign Ownership and Employment and Domestic Acquisitions in Germany.” Growth in a Developing Country.” Journal of Applied Economics 47 (7–9): 739–55. Development Studies 49 (8): 1133–47. Gong, Yundan, Holger Görg, and Sara Maioli. Maksimovic, Vojislav, and Gordon Phillips. 2001. 2007. “Employment Effects of Privatisation “The Market for Corporate Assets: Who Engages and Foreign Acquisition of Chinese State- in Mergers and Asset Sales and Are There Efficiency Owned Enterprises.” International Journal of Gains?” Journal of Finance 56 (6): 2019–65. the Economics of Business 14 (2): 197–214. Maksimovic, Vojislav, Gordon Phillips, and Grzegorz, Pac. 2014. “Foreign Acquisition and Nagpurnanand Prabhala. 2008. “Post-Merger Post-Privatization Exit of State-Owned Firms.” Restructuring and the Boundaries of the Firm.” Journal of International Trade & Economic Journal of Financial Economics 102 (2): 317–43. Development 23 (4): 540–77. Martins, Pedro S., and Luiz Alberto Esteves. 2008. GSMA Intelligence. 2016. “The Mobile Economy: “Foreign Ownership, Employment and Wages in Africa 2016.” Report, GSMA Intelligence, Brazil: Evidence from Acquisitions, Divestments GSM Association, London. and Job Movers.” Discussion Paper No. 3542, Guadalupe, Maria, Olga Kuzmina, and Catherine Institute of Labor Economics (IZA), Bonn. Thomas. 2012. “Innovation and Foreign Mattes, Anselm. 2010. “International M&A: Ownership.” American Economic Review Evidence on Effects of Foreign Takeovers.” 102 (7): 3594–627. Discussion Paper 60, Institute for Applied Hale, Galina, and Mingzhi Xu. 2016. “FDI Effects Economic Research (IAW), Tübingen, Germany. on the Labor Market of Host Countries.” Milliou, Chrysovalantou, and Apostolis Pavlou. Working Paper 2016-25, Federal Reserve Bank 2014. “Foreign Direct Investment Modes and of San Francisco. Local Backward Linkages.” Working Paper Harms, Philipp, and Pierre-Guillaume Meon. Series 4623, CESifo Group, Munich. 2018. “Good and Useless FDI: The Growth Mistura, Fernando, and Caroline Roulet. Effects of Greenfield Investment and Mergers 2019. “The Determinants of Foreign Direct and Acquisitions.” Review of International Investment: Do Statutory Restrictions Matter?” Economics 26 (1): 37–59. Working Papers on International Investment Inekwe, John Nkwoma. 2013. “FDI, Employment 2019/01, Organisation for Economic and Economic Growth in Nigeria.” African Co-operation and Development (OECD), Paris. Development Review 25 (4): 421–33. Nocke, Volker, and Stephen Yeaple. 2008. Krugman, Paul. 2000. “Fire-Sale FDI.” In Capital “An Assignment Theory of Foreign Direct Flows and the Emerging Economies: Theory, Investment.” Review of Economic Studies Evidence, and Controversies , edited by 75 (2): 529–57. H o w B e n e f ici a l Ar e F o r e ig n Ac q u i s i t i o n s o f Fir m s i n D e v e l o p i n g C o u n t ri e s ?    83 Nolan, Marcus. 2019. “Competition Policy and Staff, Matthew. 2015. “Ghana Undergoes a FDI: A Solution in Search of a Problem?” Vodafone Transformation.” Africa Outlook Working Paper No. 99-3, Peterson Institute for 23/24: 26–41. International Economics, Washington, DC. UNCTAD (United Nations Conference on Trade OECD (Organisation for Economic Co-operation and Development). 2000. World Investment and Development). 2012. “Cross-Border Report 2000: Cross-Border Mergers and Merger Control: Challenges for Developing Acquisitions and Development. New York and and Emerging Economies.” Competition Geneva: United Nations. Policy Roundtable proceedings, DAF/COMP​ UNCTAD. 2019. World Investment Report 2019: /­ GF(2011)13, OECD Publishing, Paris. Special Economic Zones. New York: United Orange. 2017. “Orange–Action Statement: Nations. Support for Innovative Digital Ecosystems in Wang, Miao, and Man C. Sunny Wong. 2009. Middle East and Africa.” Note for the Make-IT “What Drives Economic Growth? The Case Alliance, Orange, Paris. of Cross-Border M&A and Greenfield FDI Polanco Lazo, Rodrigo. 2018. “Facilitation 2.0: Activities.” Kyklos 62 (2): 316–30. Investment and Trade in the Digital Age.” World Bank. 2016. Breaking Down Barriers: Paper produced under the RTA Exchange Unlocking Africa’s Potential Through Vigorous of the International Centre for Trade and Competition Policy. Washington, DC: World Sustainable Development (ICTSD), Geneva; Bank. and Inter-American Development Bank (IDB), Wo r l d B a n k . 2 0 1 8 . G l o b a l I n v e s t m e n t Washington, DC. Competitiveness Report 2017/2018: Foreign Raff, Horst, Michael Ryan, and Frank Stähler. Investor Perspectives and Policy Implications. 2011. “Firm Productivity and the Foreign- Washington, DC: World Bank Market Entry Decision.” Working Paper, Kiel WTO (World Trade Organization). 2017. University, Kiel, Germany. “Joint Ministerial Statement on Investment Roberto, Basile. 2004. “Acquisition versus Facilitation for Development.” Adopted Greenfield Investment: The Location of Foreign at the 11th Ministerial Conference, Manufacturers in Italy.” Regional Science and December 13, WTO, Geneva. https:// Urban Economics 34 (1): 3–25. d o c s . w t o . o r g ​/ d o l 2 f e / P a g e s / F E _ S e a r c h​ Schoar, Antoinette. 2002. “Effects of Corporate / F E _ S _ S 0 0 9 - D P ​ . a s p x ? l a n g u a g e =​ Diversification on Productivity.” Journal of E&CatalogueIdList=240870. Finance 57 (6): 2379–403. 3 The Distributional Effects of FDI: Evidence from Ethiopia, Vietnam, and Turkey Victor Steenbergen and Trang Thu Tran Key Findings • Many countries aim to attract foreign investment to help create jobs and reduce poverty. Yet empirical evidence on the direct poverty-reducing effects of FDI is surprisingly scarce, especially in developing countries. Little is also known about the aggregate effects of FDI on income distributions. • Analysis of unique firm-level and household data from Ethiopia, Vietnam, and Turkey shows that FDI firms create new jobs and pay higher wages than domestic firms. Work- ers in sectors and regions with higher foreign-firm presence are generally more likely to be formally employed and receive higher wages. FDI allowed more than 350,000 individuals to enter formal manufacturing employment in Vietnam between 2007 and 2016, and at least 40,000 in Turkey between 2009 and 2016. FDI also raised average manufacturing wages by 32 percent in Ethiopia, 12 percent in Vietnam, and 8 percent in Turkey. • Consequently, FDI-induced wage increases helped reduce poverty in all three countries. Conservative estimates suggest that FDI contributed to lifting at least 35,000 individuals out of poverty in Ethiopia during 2009–14; 24,000 in Vietnam (2007–16); and 15,000 in Turkey (2009–16). Although the FDI-induced wage increases helped improve the incomes of the bottom 40 percent of the population in all three countries, the effects across the entire income distribution differed significantly across countries. In Ethiopia, the benefits of FDI were more concentrated in the bottom 40 percent, while in Vietnam, the welfare gains were evenly distributed across the income distribution. Turkey had the greatest aver- age wage benefits from FDI but also experienced increases in wage inequality. • However, FDI can also contribute to inequality by disproportionately benefiting better- educated and higher-skilled workers. When comparing regions and sectors with higher FDI activity with those with no FDI, higher-skilled workers experience large benefits, while low-skilled workers may see no changes or even experience relative short-term declines in formal employment and wages. • The possible adverse effects of FDI on income inequality and on lower-skilled workers emphasize the importance of a country’s labor market and education policies. Key policies include strengthening the absorptive capacity of d ­ omestic firms and workers (for example, through programs that foster FDI-­ supplier linkages and employment ­ supporting training); ­ vulnerable communities (such as lower-skilled workers, youths, and women) with active jobs information, provision, and skills certification; and establishing programs to stimu- late labor mobility within countries. Introduction Recent findings about trade liberalization Many countries around the world aim to indeed confirm that some evidence backs such attract foreign investment to help create popular sentiments. For higher-income coun- jobs and reduce poverty. Yet direct empirical tries, import liberalization may have increased evidence on the direct poverty-reducing competition in less-skilled, labor-intensive effects of foreign direct investment (FDI) industries while favoring demand for skilled is surprisingly scarce. Most of the earlier workers (Acemoglu and Autor 2011; literature focused on the ability of FDI to Acemoglu and Restrepo 2017; Goldberg and raise economic growth, which in turn is Pavcnik 2007; Maloney and Molina 2016; associated with reductions in poverty (Chen Michaels, Natraj, and Van Reenen 2014; and Ravallion 2007). However, it is notably Pavcnik 2017). Tariff reductions on Chinese difficult to estimate the growth effects of products contributed to substantial job losses FDI precisely (Lipsey 2003). FDI’s poverty- in US manufacturing in the 1990s and 2000s alleviating effects may also be greater or less (Autor, Dorn, and Hanson 2013; Pierce and than average because of its direct influence Schott 2016). Similarly, in Brazil, trade liber- on a country’s aggregate employment alization and import competition strongly numbers and average wages (Nunnenkamp, affected local labor markets, resulting in wage Schweickert, and Wiebelt 2007). increases for skilled workers but wage A second generation of literature then declines for unskilled workers. These effects argued that FDI helps raise household income persisted up to 20 years after import liberal- because formal firms pay premium wages.1 ization (Dix-Carneiro and Kovak 2015). While important, this literature focuses on It is important to better understand the firm-level effects. This can present a biased role that FDI plays in national income distri- picture because foreign-owned firms may butions. This could help counter nationalist be “cherry-picking” the most productive sentiments around FDI by providing oppos- workers, possibly leading to labor shifts ing evidence or by stressing the need for com- among firms with no real change in overall plementary interventions that proactively employment or household income. Using address FDI’s impact on income inequality. To firm-level data also means that the aggregate do so, it is critical to understand the potential effects on labor markets that most policy impact that FDI may have across different makers care about (such as creation countries, sectors, and workers with different of formal jobs and growth in average skill levels. wages) cannot be observed. To better This study aims to answer two vital ques- establish the relationship between FDI and tions around the contribution of FDI to development, it is therefore important to development: consider FDI’s effect at the household level. • How does the presence of FDI firms So far, robust economic analysis doing so influence labor market outcomes (formal has been limited. employment and wages)? Little is also known about the aggregate • What are the effects of FDI firms’ pres- effect of FDI on income distributions. This ence on poverty and income distributions? relationship has become particularly impor- tant in recent years, as backlashes against This analysis will help demonstrate the globalization have been attributed to growing effects of FDI on jobs and offer practical concerns around the effects of trade and insights into how investment promotion investment on rising levels of income can support inclusive growth. It will investi- inequality. This may have played a role in gate the effect of FDI presence on the World reducing investor confidence and FDI flows in Bank Group’s twin goals to end extreme recent years (see Overview). poverty and boost shared prosperity. 2 85 8 6    Global Investment Competitiveness Report 2019/2020 This report also ties in with the wider dis- Income equality is measured through two cussions around FDI and development and metrics: the Gini coefficient and the Palma relates to the renewed focus by the United ratio (the ratio of the richest 10 percent of the Nations (UN), Organisation for Economic population’s share of gross national income Co-operation and Development (OECD), divided by the poorest 40 percent’s share). and others on “Promoting Investment in The analysis finds that increases in MNE the Sustainable Development Goals” activities have a significant differential effect (UNCTAD 2018).3 on formal job creation and wages. Workers To examine the effects of FDI, this chapter in sectors and regions with a higher share studies three countries where FDI increased of MNE output are more likely to be for- greatly in the past 20 years: Ethiopia, mally employed and receive higher wages Vietnam, and Turkey (in order of lowest to (relative to workers in sectors with a lower highest income level). The analysis links firm- share of MNE output). In aggregate, the esti- level data with household survey data to mations suggest the following: examine the effect that economic activity from multinational enterprises (MNEs) has • In Vietnam, on average, FDI has contrib- on local labor market outcomes (jobs and uted to more than 350,000 individuals wages). switching from informal to formal manu- The analysis focuses on MNE presence facturing employment each year between rather than FDI inflows for two reasons: First, 2007 and 2016. FDI also resulted in FDI inflow data broken down by sector are 12 percent higher wages in manufacturing rarely available across countries. Second, FDI and 2 percent higher wages in services. inflows merely measure the amount of cross- • In Turkey, manufacturing FDI is associated border financing, which may or may not with around 40,000 additional formal materialize into meaningful economic activi- jobs from 2009 to 2016 annually and ties. By focusing on the activity of MNEs, this an 8 percent increase in average wages. study follows recent research that has increas- No effects were identified on formal ingly analyzed firm-level operational data to employment or on average wages in better understand the impact of FDI on host Turkey’s services sector. economies (Alfaro and Chauvin, forthcom- • In Ethiopia, aggregate formal employment ing; Antràs and Yeaple 2014). effects are insignificant, but FDI contrib- The exercise exploits variations in MNE uted to a large increase (32 percent) in total sales (as a share of a sector’s and manufacturing wages. No information region’s total output) to compare the formal was available for Ethiopia’s services. employment and average wages of workers FDI-induced growth in formal jobs and employed in sectors and regions with higher wages has translated into increased shared MNE activity with those with lower MNE prosperity and reduced poverty. The wage activity. To explore the distributional impact benefits from FDI were positive and sizable of FDI firms, the analysis focuses on the for all three countries studied. Yet, the distri- potential skill bias by comparing the effects butional effects differ across countries of MNE activities on labor market outcomes (figure 3.1): in higher- and lower-skilled sectors and for workers with different education levels. • In Ethiopia, the wage effects from low- It uses these estimated effects to estimate a skilled manufacturing FDI are highly simple back-of-the-envelope counterfactual positive but limited in magnitude because income (without MNE activities). From of the manufacturing sector’s small scale. this, it assesses the aggregate effect of FDI The income benefits are with the bottom on poverty reduction and shared prosperity 40 percent and linked to declining income (income gain of the bottom 40 percent). inequality. Wage increases from FDI are T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    87 FIGURE 3.1  FDI Has Had Varied Effects on National Income Distributions in Ethiopia, Vietnam, and Turkey 8 6 Wage effect from PDI (%) Turkey 4 Vietnam 2 Ethiopia 0 0–10 10–20 20–30 30–40 40–50 50–60 60–70 70–80 80–90 90–100 Income decile Ethiopia: bottom 40% Ethiopia: top 60% Vietnam: bottom 40% Vietnam: top 60% Turkey: bottom 40% Turkey: top 60% Source: World Bank. Note: FDI = foreign direct investment. found to have reduced poverty for at The aggregate FDI effects across the three least 35,000 individuals between 2009 countries mask significant variations by sec- and 2014. tors and workers’ education levels (summa- • In Vietnam, the wage benefits from rized in table 3.1). In general, the average FDI have also been positive and are effects of FDI on formal employment and the most widespread and evenly dis- wages are positive for manufacturing and tributed across incomes. Correspond- high-skilled services but neutral for extrac- ing to these effects, FDI has almost tives and low-skilled services. no effect on income inequality. These There is also evidence of FDI’s skill wage increases from FDI contributed to premium. In regions and sectors with higher l ifting at least 24,000 individuals out of ­ MNE activity, higher-skilled workers experi- poverty between 2007 and 2016. ence large increases, while low-skilled • In Turkey, FDI had the largest average workers may see no changes or relative wage benefits across the three countries. declines in formal employment and wages It increased average wage income for the (relative to the sectors not receiving FDI). The bottom 40 percent and helped reduce pov- skill premium is more pronounced in services erty for over 15,000 individuals between than in manufacturing. 2009 and 2016. However, FDI was asso- This analysis suggests that FDI, especially ciated with substantial benefits for high- in tradable sectors, can contribute meaning- skilled workers and evidence of displace- fully to development by stimulating formal ment and potential wage reductions for job creation, poverty reduction, and shared the lowest-skilled workers—thus contrib- prosperity. FDI brings about productivity uting to increased income inequality. improvements and structural transformation 8 8    Global Investment Competitiveness Report 2019/2020 TABLE 3.1  The Effects of FDI on Labor Markets Vary by Sector and Workers’ Skill Levels Broad sector Average effect Low-skilled workers High-skilled workers Extractives No effect No effect No effect Low-skilled manufacturing Positive Positive Positive High-skilled manufacturing Positive Neutral Positive Low-skilled services Neutral Negative Positive High-skilled services Positive Negative Positive Source: World Bank calculations, summarizing table 3.2. Note: Low-skilled workers are defined as those with primary education or less, while high-skilled workers have completed at least secondary e ­ ducation. All results are relative to workers in sectors with less or no investment by multinational enterprises (MNEs). For all sector- and country-specific data, see annex 3C, table 3C.6. FDI = foreign direct investment. that boost long-term economic growth. Yet, women) by providing active jobs informa- in some cases, skill-intensive FDI can be asso- tion and skills certification. ciated with a skill premium that increases • Establish programs to stimulate inter- wage inequality. To maximize FDI’s contribu- nal migration, which can further help tion to inclusive growth, countries should improve access to employment, with therefore complement investment policy and important household welfare benefits. investment promotion efforts with progres- sive labor market policies to counter FDI’s potential effect on any (temporary) declines in Conceptual Framework: The Link employment and increased income inequality. The chapter makes six recommendations (dis- between FDI and Welfare cussed in detail in the concluding “Policy FDI can affect welfare by both changing Implications” section): individuals’ incomes and altering the prices of goods and services available to consumers • Better align investment promotion through three channels (figure 3.2): strategies with a country’s labor skill base, related sectors, and relevant source • Employment income: As FDI brings cap- countries. ital and new technology to a sector, it • Strengthen the absorptive capacity of often raises the overall labor demand domestic firms and workers (for example, and productivity in the sector. This can through programs supporting FDI-supplier raise total employment and average linkages and employment training) to help wages, leading to higher household extend the labor market benefits from FDI. incomes. • Open up services sectors to foreign • Consumer prices: The entry of new (and investment to help (indirectly) create new possibly more productive) foreign firms jobs. Governments may wish to combine to markets also increases competition. the promotion of services FDI with This may lower the price of goods progressive labor market interventions and services, thus raising household to ensure that both high- and low-skilled purchasing power and consumption workers in the services sector benefit. possibilities. • Improve bargaining power and knowl- • Producer income: As foreign firms com- edge spillovers for workers by enforcing pete with, buy from, and sell to domestic sufficient labor standards and supporting ­ roductivity firms, they may influence the p labor representation. and profitability of these enterprises, • Support vulnerable communities (such increasing or cutting into income for as lower-skilled workers, youth, and domestic producers. T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    89 FIGURE 3.2  FDI Affects Household Incomes through Several Broad Channels Foreign direct investment Income to employees Prices for consumers Income to producers Change in real household income Poverty reduction Shared prosperity Equitable income distribution Source: World Bank, based on World Bank and OECD 2017. Note: FDI = foreign direct investment. Depending on how important such effects Finally, metrics of income inequality reflect are for individuals along the income distribu- on a country’s overall (wage) income distribu- tion, FDI will have different consequences on tion. Recent research suggests that, for many three welfare goals often considered by policy countries, inequality trends are mainly driven makers around the world: poverty reduction, at the top and the bottom (with limited shifts shared prosperity, and an equitable income in the middle of the income distribution). distribution. Poverty is often defined as those Such research thus promotes the use of the households that fall below a minimum basic Palma ratio, which considers the aggregate level of real household income (based on income share for the top 90 percent versus the income or consumption patterns). To the bottom 40 percent (Krozer 2015). extent that FDI raises income for these house- This analysis focuses on how FDI affects holds, it also helps reduce poverty. labor income, which has been shown to be Yet governments often face a parallel chal- the main channel through which individuals lenge to ensure that any general income gains escape poverty. Historically, poverty reduc- are equitably distributed within the country. tion has been shaped most by growth in This is captured by two different metrics. labor income rather than by changes in non- Shared prosperity has been defined as labor income or demographics (Ferreira “expanding the size of the pie continuously 2010). Among 21 developing countries, and sharing it in such a way that the welfare growth in labor income accounts for more of those at the lower end of the income distri- than half of the poverty reduction in bution rises as quickly as possible” (World 12 countries, while it accounts for more than Bank 2013). It is measured as income growth 40 percent of poverty reduction in another of the bottom 40 percent of the income distri- 6 countries, Inchauste et al. (2014) find. bution in the population. In some lower- They note that wage growth (rather than income countries, this goal will coincide with increased employment) contributed most to that of reducing poverty. poverty reduction. 9 0    Global Investment Competitiveness Report 2019/2020 The Effect of FDI on Labor Market MNEs (upstream) or that rely on MNE inputs in producing their goods or services Outcomes (downstream). The overall labor market impact of FDI is Effects on the demand for skilled versus theoretically and empirically ambiguous unskilled labor (skill-biased labor because of the opposing effects it can have demand). Alongside overall shifts in on labor demand for different types of work- employment demand, there may also be ers (figure 3.3). changes in the employment and remunera- Effects on the aggregate demand for labor. tion across different types of workers. Inflows of FDI affect a sector’s labor demand MNEs often bring in new technology, through scale, competition, and productivity. which requires higher-skilled workers to Scale effects may take place when the foreign operate. For that reason, high-skilled firm produces new or extra goods and ser- workers may be more likely to be employed vices (for example, for export-oriented firms). and could enjoy a wage premium. In con- This often increases overall labor demand. trast, demand for lower-skilled workers FDI can also take away market share from might fall. A limited supply of skilled labor domestic firms (for example, for domestic in the local economy further helps raise products). benefits for skilled work by affecting their The effect this has on employment depends wage bargaining power, further intensify- on the relative labor intensity of foreign ver- ing the skill bias for FDI. sus domestic firms. MNE productivity may be reflected in higher employment and workers’ Effects on Aggregate Labor Demand compensation. Alternatively, MNEs may adopt labor-saving technologies that could be Empirically, FDI has often been found to associated with a net negative employment have a positive impact on wages and employ- effect. In addition, there may be labor demand ment, particularly for lower-income coun- effects from other sectors that supply to tries. Much of the evidence points to FDI’s FIGURE 3.3  Conceptual Framework: FDI Has Varied Effects on Labor Market Outcomes Skill-biased FDI effects technological on labor demand change (scale, competition, (compositional productivity) employment effects) Aggregate Skill-biased labor labor demand demand effects effects from FDI from FDI Vertical effects Local scarcity of (spillovers) on labor skilled labor demand supply (upstream product (wage bargaining demand, downstream power) inputs prices) Source: World Bank. Note: FDI = foreign direct investment. T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    91 potential to raise wages, driven primarily by This can lead to lower production, lower new technology and increased labor produc- employment, and possibly lower wages tivity (Hale and Xu 2016). In many cases, the (Hale and Xu 2016). So far, the overall effect literature also finds a positive effect of FDI is unclear, given that few studies consider the on aggregate employment. For example, FDI vertical spillovers of FDI to labor market was found to increase the employment rate in outcomes. China, the Czech Republic, and Uruguay (Dinga and Münich 2010; Karlsson et al. 2009; Peluffo 2015). FDI also had a strong Effects on Skill-Biased Labor Demand positive employment effect on Mexico’s man- FDI often introduces new technologies that ufacturing FDI, with stronger effects in raise the demand for higher-skilled workers export-oriented industries (Waldkirch, and increase the wage gap between skilled Nunnenkamp, and Bremont 2009). and unskilled workers. There is considerable However, the evidence is more mixed for empirical evidence confirming that FDI con- higher-income countries. Although manufac- tributes to rising wage inequality in host turing FDI in Sweden was associated with an increase in employment (Bandick and countries. Karpaty 2011), FDI in central and eastern In developing countries, wage inequality Europe led to job losses through competitive increases with stocks of inward FDI, a cross- pressure and introduction of labor-saving country study shows (Figini and Görg 2011). technology (Jude and Silaghi 2015). For A rise in Japanese FDI in developing coun- higher-income countries, FDI may decrease tries is associated with an increase in nonpro- the number of jobs in the short term by duction wages (for more-skilled workers) introducing labor-saving technology, but it relative to production wages (for less-skilled will likely increase job growth in the long workers), according to Head and Ries (2002). term by enhancing labor productivity (Hale Similar effects of foreign investments have and Xu 2016). been found for firms in Indonesia and Mexico There is limited and inconclusive evidence (Feenstra and Hanson 1997; Lee and Wie 2015). on the effect of FDI’s vertical spillovers. The However, technological change is not nec- overall effect of FDI on upstream firms that essarily biased in favor of skilled workers supply inputs to foreign-invested firms is (Luo 2017). There is an important sector bias ambiguous. In some cases, product demand in the type of FDI attracted. FDI in some types rises, and positive technological spillovers of low-skilled sectors (such as textiles and may push up employment and the average food processing) could disproportionally ben- wage. On the other hand, target firms might efit unskilled workers (Cruz et al. 2018; switch from domestic to foreign suppliers of Leamer 1998). For this reason, FDI in labor- intermediate inputs, in which case produc- intensive manufacturing and infrastructure is tion, labor demand, and wages of upstream associated with declining inequality in industries will decline (Reyes 2017). Ethiopia, Ghana, and Mozambique (Cornia The effect of FDI on downstream firms’ 2016). wages is similarly unclear. Access to cheaper The effects of FDI can be local, at least in or higher-quality inputs (such as in business the short term. Overall employment in receiv- services) can improve domestic firms’ pro- ing industries tends to increase with FDI. Yet ductivity, increase output, and raise wages because of these industries’ greater reliance (Arnold et al. 2016; Arnold, Javorcik, and on technology that requires complementary Mattoo 2011; Duggan, Rahardja, and Varela skills, a larger presence of foreign firms or 2013). Yet FDI firms may shift production affiliates in the region and industry also focus from domestic to international mar- increases demand for skilled labor. Because kets, so that the cost of intermediate inputs the supply of skilled labor is highly inelastic in may increase for domestic downstream firms. 9 2    Global Investment Competitiveness Report 2019/2020 the short and medium term, this further Depending on a sector’s labor and skill con- pushes up the wages of skilled workers in the tent, FDI can have markedly different impacts regions and industries with higher FDI pres- on employment and wages. Much of the evi- ence (Hale and Xu 2016). Given that most dence considered so far has either been cross- developing countries have considerable sectoral or focused only on m ­ anufacturing. restrictions on worker mobility between However, policy makers often must decide regions, the effects tend to be rather concen- which sectors should receive priority in their trated in local labor markets (Dix-Carneiro investment promotion efforts (Javorcik 2004). and Kovak 2015; Pavcnik 2017). This also For that reason, it is important to consider the means that FDI can lead to another form of various impacts of FDI by sector. inequality—geographical inequality—as has been found in Bolivia and Vietnam (McLaren Labor Market Impact in the and Yoo 2016; Nunnenkamp, Schweickert, Services Sector and Wiebelt 2007). Overall, the literature suggests that FDI The distributional impact of FDI in the ser- has positive but unequal effects on host coun- vices sector is underresearched but deserves tries’ labor markets. FDI is associated with greater policy attention. FDI in services dif- higher aggregate employment and a rise in fers from FDI in manufacturing in three average wages. Many of these benefits accrue important ways. to higher-skilled workers, while lower-skilled First, the services sector tends to increase workers may experience adverse effects. Yet the demand for higher-skilled labor. Many the literature also suggests that FDI might service-oriented firms rely more than manu- change local norms about labor conditions facturers on intellectual capital and may (Hale and Xu 2016; Javorcik 2015). Although therefore exhibit larger skill premiums the increase in inequality resulting from dis- (Kianto, Hurmelinna-Laukkanen, and Ritala proportionate growth of demand for skilled 2010). For example, in the Philippines, ser- labor is a worrying dynamic, this may also vice liberalization in banking, distribution, induce the labor force, in the long term, to and telecommunications created employment seek additional education and training (Heath opportunities for higher-skilled workers and and Mobarak 2015). generated negative impacts on the employ- ment and wages of low-skilled workers (Amoranto, Brooks, and Chun 2010). The Sectoral Impact of FDI on Greenfield FDI in business support services Labor Market Outcomes (for example, professional services, informa- tion and communication technology [ICT], and research and development [R&D]) "FDI flows come in at least three—and across 17 higher-income countries benefited probably four—separate forms: FDI in high-skilled workers at the expense of extractive industries, FDI in infrastructure, medium-skilled workers (Davies and FDI in manufacturing, plus the Desbordes 2015). underresearched field of FDI in services. Second, FDI in the services sector is more Each form presents such distinctive policy likely to reduce domestic employment challenges for developing-country host because of labor-saving productivity authorities, and generates such diverse improvements. Although some services are impacts on the developing host economy, as export-oriented (such as ICT), many operate to undermine the usefulness of any research exclusively within the domestic market. This that does not disaggregate the FDI flows." means that FDI in services often lacks scale effects and instead captures market share —Theodore H. Moran, Foreign Direct from domestic firms, often resulting in little Investment and Development (2011) or no aggregate employment effects. T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    93 Evidence from China on services liberaliza- price channel in figure 3.2. However, those tion finds that although output increased in researchers also find evidence of a sizable almost all services industries, employment reduction in the monthly incomes of workers losses occurred because of labor-saving in traditional retail sectors as well as some improvements in productivity (Li, Wang, decline in the labor incomes of workers in and Zhai 2003). Although these job losses modern retail sectors. In total, the household could be offset by expansion of overall labor benefits are positive but regressive. demand in nonservice industries and by long-term growth of aggregate labor Labor Market Impact in the demand, these findings suggest that there are Manufacturing Sector important short-term labor adjustment costs from FDI in services sectors. A synthesis of the literature previously dis- Third, services sector FDI exerts competi- cussed suggests that FDI in different sectors tive pressure on smaller businesses. Services has different effects on overall and skill- sectors in developing countries are more often biased labor demand. FDI in low-skilled dominated by small, family-owned businesses manufacturing is expected to have the larg- (such as retail operations). Unlike manufac- est effects on labor demand, with limited turing, which is often dominated by a few skill premiums. Higher-skilled manufactur- large firms, competitive pressures on these ing is more skill-biased but with some aggre- smaller businesses might lead to a more gate benefits to labor demand. FDI in regressive impact of FDI. A notable example low- and high-skilled services has been found is the entry of foreign supermarkets in to create few jobs but to have notable effects Mexico, which helped lower the cost of living on average wages and skill premiums. As a and substantially benefited the average house- capital-intensive sector, FDI in extractives is hold (Atkin, Faber, and Gonzalez-Navarro expected to have little overall impact on 2018)—an effect represented by the consumer wage incomes (figure 3.4). FIGURE 3.4  FDI Has Different Labor Market Outcomes by Sector Low-skilled manufacturing High Aggregate labor demand effects from FDI High-skilled manufacturing Medium Low-skilled High-skilled Extractives services services Low Low Medium High Skilled-biased labor demand effects from FDI Source: World Bank. Note: The figure is a stylistic representation based on available literature. Each aggregated sector (manufacturing, services, and extractives) is subdivided based on workers’ average education: “high skilled” are those in which at least 50 percent of workers had secondary or postsecondary education; the remainder is considered “low skilled.” For more about the subclassification, see annex 3C, table 3C.1. FDI = foreign direct investment. 9 4    Global Investment Competitiveness Report 2019/2020 Finally, it is worth noting that while this 2002 and 2007. Despite the slowdown fol- chapter focuses on the potential skill bias of lowing the global financial crisis, cumulative FDI, there are other channels through which inflows in the 15 years between 2003 and FDI can have different consequences on 2017 are still almost 13 times higher than income distribution and inclusive growth. total inflows in the preceding 30 years.4 One such example comes from gender-specific In Vietnam, FDI growth has been more labor market effects. Although this chapter gradual but more persistent. FDI there first does not explicitly analyze this issue in the jumped in 2007–08 and has since steadily case studies, a discussion on past findings on increased, leading to inflows matching those FDI’s impact on women’s opportunities is of Turkey by 2017. presented in box 3.1. As the least developed country among the three countries, the FDI takeoff in Ethiopia Country Case Studies: Ethiopia, lags Turkey’s and Vietnam’s by almost a decade. Nevertheless, acceleration of FDI Vietnam, and Turkey since 2012 has made Ethiopia one of the larg- To examine the effects of FDI, this chapter est FDI recipients in Africa (UNCTAD 2019). studies three countries where FDI has taken These episodes of FDI growth took place off significantly in the past 20 years: when the three countries were at different Ethiopia, Vietnam, and Turkey (figure 3.5). stages of development. The countries have Turkey’s first episode of significant growth also had distinct economic structures and sup- was in the early 2000s, when FDI increased plies of skills.5 In 2018, Turkey’s income per twentyfold in the five-year period between capita was close to US$28,000 (in purchasing BOX 3.1 FDI’s Potential to Improve Women’s Economic Opportunities Through its impacts on labor market outcomes, ­ ignificant step up in pay for women relative to s foreign direct investment (FDI) can notably informal employment. affect women’s economic opportunities (for Empirical studies from economies in three example, by raising the female participation different parts of the world provide support- rate in the country or reducing the gender wage ive evidence that FDI can help raise women’s gap).a Unfortunately, little evidence currently labor demand. In Honduras, FDI inflows exists on the gender-specific effects of FDI. What were critical in establishing export-oriented evidence does exist indicates that FDI generally manufacturing maquiladoras . b Sur veys has a positive effect on gender equality, partly found that women who moved to this kind of by raising the overall demand for all labor employment for the first time earned about and partly through cultural norm transfers 50 percent more than in their previous jobs from source countries with more gender-equal (Ver Beek 2001). In China, the establish- cultures. Yet much of these benefits ultimately ment of foreign firms helped raise the female depend on the sectoral FDI type and skill level participation rate; female employment rates of women in the host economy. in such firms were 13 percent higher than FDI can increase gender equality by rais- for domestic-owned firms (Chen, Li, and ing labor demand. Women in many developing Shapiro 2012). In Madagascar, FDI gener- countries are relatively overrepresented in the ated jobs in export-processing zones (EPZs) informal sector, which is often poorly remu­ that provided women with high-wage jobs nerated. More formal employment oppor- relative to their skill level and with similar tunities brought by FDI can thus present a pay between men and women. Looking at Box continues next page T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    95 BOX 3.1 FDI’s Potential to Improve Women’s Economic Opportunities (continued) firm-level data from 1995 to 2002, the study and low-training tasks in foreign firms, with found that after three initial years, wage lower relative wages as a result (Chen, Li, growth for women even outstripped that of and Shapiro 2012). China provides some men: 35 percent versus 25 percent (Glick and evidence of this dynamic effect. Looking at Roubaud 2006). household data from 1995–2002, FDI was FDI may also affect cultural norm trans- found to have positive effects on both female fers. Women’s position in the labor force and male wages. At the beginning, women may be affected by gender-biased norms and experienced larger wage increases from FDI perceptions.c Recent evidence suggests mul- than men. At the end of the sample period, tinational firms may be less subject to such however, this trend reversed, and men expe- gender-biased norms and can help the global rienced larger wage increases (Braunstein diffusion of gender-equal norms. Cross- and Brenner 2007). country analysis for 94 developing countries Finally, the sectoral dimension of FDI finds that higher FDI inflows are associated also inf luences its gender impact. T he with increases in gender development (includ- overall skill requirements of labor differ ing female participation rates) and with significantly across sectors. FDI’s impact declines in gender inequality (including the o n wo m e n’s e m p l oy m e n t a n d w a g e s gender wage gap) (Ouedraogo and Marlet may therefore depend on FDI’s sectoral 2018). Similarly, in China, foreign affiliates differences and women’s skill level. Evidence from countries with a more gender-equal from this comes from a study in rural culture are found to employ proportion- Indonesia, which considered the effect of FDI ally more women and appoint more female on women’s employment in lower-skilled managers. They also generate cultural spill- plantation employment and higher-skilled overs, increasing domestic firms’ female labor hotel employment. The study found that low shares in the same industry or city (Tang and wages affected employment in plantations Zhang 2017). the most, while skills availability was the Women’s skill level affects the impact of main determinant for employment in hotels. FDI on women’s empowerment. In many On the whole, female workers in rural developing countries, women are relatively Indonesia were both low waged and low unskilled and face lower wages relative skilled. As a result, relatively more women to men. These features can be a pull fac- ended up being employed on plantations, tor for FDI in low-skilled, labor-intensive while relatively fewer women ended up sectors (such as textiles) and help increase working in the new hotels (Siegmann 2007). women’s access to employment. Yet as a. This box is based in part on Fang, Shamseldin, and Xu (2019) and countries move toward better technology on extensive inputs from Matthew Stephenson. and higher demand for skilled labor, wom- b. Maquiladoras are factories producing labor-intensive products with imported goods; a high share of their employment is female. en’s employment and wages may decline or c. For example, about two-thirds of firms surveyed in Pakistan’s even reverse (Braunstein 2006; Seguino and Enterprise Survey in 2013 reported gender-discriminatory attitudes as Grown 2006). Lower-skilled women work- reasons for not hiring women. These reasons include “women should focus on family responsibilities”; “women employees ‘disrupt’ the ing within these firms may be disproportion- workplace”; and “male colleagues/customers are hesitant to interact ately assigned to low-value-added, low-tech, with women” (Amir et al. 2018). 9 6    Global Investment Competitiveness Report 2019/2020 FIGURE 3.5  FDI Has Increased Significantly in Ethiopia, Vietnam, and Turkey 25 20 FDI inflows (US$, billions) 15 10 5 0 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 Ethiopia Vietnam Turkey Source: World Development Indicators database. Note: FDI = foreign direct investment. power parity terms), more than 3 times that sectors—these investments reflecting the of Vietnam and 14 times that of Ethiopia. The increased share of output and of employment Turkish economy is typical of an upper-mid- due to FDI (VFIA 2018). Turkey’s FDI inflows dle-income economy, with services contribut- have been the most broad based, covering ing 65 percent of gross domestic product substantial investments in both manufactur- (GDP). Vietnam still has a substantial agricul- ing and services. The financial sector has ture and industry base (contributing 15 per- attracted the highest amount of FDI, closely cent and 34 percent of GDP, respectively), followed by manufacturing, energy, and ICT while Ethiopia still depends heavily on agri- services (ISPAT 2018). culture (31 percent of GDP). Gaps in educa- The differences in economic structure, tional attainment also remain substantial. In labor supply composition, and types of 2017, lower-secondary completion rates were investments the countries have attracted pro- 95 percent and 87 percent for Turkey and vide a rich setting to study the distributional Vietnam, respectively, but were only 30 per- impact of FDI. As discussed earlier, the cent in Ethiopia. impact of FDI will depend on the scale of Ethiopia, Vietnam, and Turkey have investments as well as the interaction between attracted FDI in different types of sectors, the labor and skill content of FDI activities reflecting both their current economic struc- and the domestic skills supply. To organize ture and educational attainment. Ethiopia has the empirical analysis, the exercise explores mainly attracted FDI in agroprocessing and these interactions through a sector typology manufacturing, notably in the textile and that classifies FDI activities based on their food and beverages sectors (EIC 2017). labor content (extractives versus nonextrac- Vietnam’s FDI remains concentrated in manu- tives); tradability (manufacturing versus ser- facturing but with increasing diversification vices); and skill intensity (high skilled versus from less-skilled activities (such as textiles/ low skilled), as detailed in box 3.2 and the clothing and plastics/rubber) to more next section. Analyzing the labor market sophisticated activities, particularly in the impact of FDI based on these characteristics electronics sector. Investments in real estate will improve understanding of the channels and retail/wholesale have also increased fol- through which FDI affects aggregate income lowing the recent opening up of these distributions. T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    97 BOX 3.2 A Sector Typology To better understand the sectoral dynamics The subsectoral classification is broadly of foreign direct investment (FDI), this study based on Hallward-Driemeier and Nayyar uses a typology that clusters various sectors (2017). To provide additional detail at the as grouped by the International Standard ISIC2 level, the exercise divided up the low- Industrial Classification of All Economic and high-skilled sectors based on the average Activities (ISIC) at the two-digit level (ISIC2), education profile of that sector’s workers in based on their general classification (extractives, Vietnam’s and Turkey’s household surveys. If manufacturing, or services) combined with their at least 50 percent of workers had secondary average workers’ skill intensity (low skilled or or postsecondary education, the sector was high skilled). The analysis focuses only on these classified as “high skilled.” Otherwise, it was sectors (and excludes others, such as agricul- considered “low skilled.” The exact subclas- ture, utilities, and personal recreation) because sification is provided in annex 3C, table 3C.1. these are easier to analyze using firm- and a. Share of global FDI from World Bank calculations of 2010–15 global household-level data, and they jointly make up FDI, based on greenfield FDI announcements from fDi Markets, a around 87 percent of global FDI (with almost Financial Times dataset (https://www.fdimarkets.com/) and Thomson Reuters’s Mergers & Acquisitions database (https://www.refinitiv​ all other FDI going to the utilities sector).a .com/). A breakdown of foreign firms’ investments Empirical Strategy across the five sectors (extractives, high- skilled manufacturing, low-skilled manufac- The significant takeoff of FDI in different turing, high-skilled services, and low-skilled sectors experienced by Ethiopia, Vietnam, services) shows that the three countries differ and Turkey creates an ideal setting to study substantially in the type of FDI they have the impact of MNEs’ presence on labor mar- attracted (figure 3.6). Given the varying skill ket outcomes. The analysis links household content and potential tradability of sectors survey data with firm-level data to examine receiving FDI in the three countries, the labor the relationship between individual workers’ market outcomes will also likely differ exposure to multinational activities and their significantly. employment and wages in each country. Ethiopia’s FDI is heavily concentrated in The main data source for individual labor low-skilled manufacturing activities, with market outcomes is the World Bank’s very little FDI going to other sectors. Vietnam International Income Distribution Database has had sizable increases in both low-skilled (I2D2), supplemented with the countries’ and high-skilled manufacturing as well as Labor Force Surveys. Firm-level data come small but consistent increases in extractives from various sources, including the Large and both types of services. Turkey has had the and Medium Manufacturing Survey highest growth of investment in high-skilled (Ethiopia), Enterprise Census (Vietnam), manufacturing (driven by pharmaceuticals, and Enterprise Information System (Turkey). motor vehicles, and chemicals) and low- The period of analysis is limited by the skilled service activities (driven by employ- overlap of household and firm-level data. ­ ment services or activities). FDI in low-skilled Nevertheless, the analysis still captures peri- manufacturing is also significant (driven ods with significant FDI growth in all three mainly by tobacco). countries: Ethiopia (2009–14), Vietnam 9 8    Global Investment Competitiveness Report 2019/2020 FIGURE 3.6  Ethiopia, Vietnam, and Turkey Have Attracted Different Types of FDI a. Ethiopia, 2000–15a b. Vietnam, 2006–15a 2,500 10,000 2,000 FDI inflows (US$, millions) FDI inflows (US$, millions) 1,500 1,000 1,000 500 100 0 –500 10 00 06 09 03 12 15 06 08 10 12 14 20 20 20 20 20 20 20 20 20 20 20 Low-skilled manufacturing FDI High-skilled manufacturing FDI Extractives FDI Low-skilled manufacturing FDI Other FDI (not observed) High-skilled manufacturing FDI Low-skilled services FDI High-skilled services FDI Other FDI c. Turkey, 2007–16 100,000 FDI inflows (US$, millions) 10,000 1,000 2007 2009 2011 2013 2015 Extractives FDI Low-skilled manufacturing FDI High-skilled manufacturing FDI Low-skilled services FDI High-skilled services FDI Other FDI Source: World Bank calculations based on the World Development Indicators database, Ethiopia’s Large and Medium Manufacturing Survey, Turkey’s Enterprise Information System data, and Vietnam’s Enterprise Census. ­ anufacturing, or Note: Total foreign direct investment (FDI) inflow is calculated as the total investments in assets made by all foreign companies in the data. Sectors (extractives, m ­services) are aggregated based on International Standard Industrial Classification (ISIC) two-digit classification. Each is subdivided based on workers’ average education: “high skilled” are those in which at least 50 percent of workers had secondary or postsecondary education; the remainder is considered “low skilled.” For more about the s ­ ubclassification, see annex 3C, table 3C.1. a. Breakdown for Ethiopia includes only manufacturing FDI because very little of the country’s FDI goes to other sectors. (2007–16), and Turkey (2009–16). For fur- output of a sector and region within a coun- ther details on data, see annex 3A. try. This approach takes the sector and region To study the direct impact of FDI, the anal- as the relevant labor market, as in Dix- ysis looks at different cohorts of workers over Carneiro and Kovak (2015) and Cruz et al. time in each country and compares labor (2018). The hypothesis is that FDI inflows market outcomes for workers who are into a sector have two potential opposing employed in sectors and regions with higher effects on the local (regional) labor market: versus lower MNE activity. It focuses on two Increased MNE activities create higher labor main outcomes: formal employment and demand, which results in an increase in formal wages.6 To proxy for MNE activity, it uses the employment and wages. At the same time, share of foreign firms’ revenue in the total FDI can compete away market share from T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    99 domestic firms, reducing their labor demand. FDI interacts with workers’ skills. That is, the If, for example, foreign firms are more effi- impact of FDI on labor market outcomes is cient, the net effect on employment could be compared for high- versus low-skilled sectors negative. and for workers with different education lev- Increased FDI in upstream (selling) and els. If FDI concentrates more in skill-intensive downstream (buying) sectors can also affect sectors, then the more-educated workers will labor demand. 7 To examine this vertical likely benefit the most, potentially increasing impact of FDI, the exercise analyzes how inequality. If FDI concentrates in less-skilled workers’ outcomes vary with the total sectors, then it has the potential to improve amount of MNE activity in their upstream or employment and wage outcomes for those at downstream sectors as well as the strength of the lower end of the income distribution. linkages between them. More specifically, Finally, the estimated average impact of this variable is calculated as the sum of FDI FDI is used to calculate the aggregate firms’ output shares—the share of revenue impact on (wage) income in a simple back- (employment) by foreign firms in total output of-the-envelope counterfactual exercise. To (employment)—in all upstream or down- compare the actual (wage) income distribu- stream sectors, weighted by how much those tion to the hypothetical case without FDI sell to or buy from a workers’ own sector presence, the exercise assumes that there is a (see annex 3B). constant effect of FDI on all workers that is The main empirical challenge is to separate equal to the estimated average effect. The the impact of FDI from other unobserved counterfactual (wage) income is then equal changes in policies or market trends that can to the actual wage minus the estimated aver- affect the labor markets at the same time. For age income gain (loss) due to FDI. From example, infrastructure spending can attract these two income distributions, estimates are FDI as well as other domestic investments made of FDI’s aggregate contributions to that boost employment and wages. MNEs in poverty reduction, shared prosperity (other certain sectors can also choose to locate in income gains within the bottom 40 percent), low-wage regions because of cost consider- and income equality—the latter measured ations, in which case higher FDI activity through the Gini coefficient and the Palma might appear to be associated with lower ratio, which considers the aggregate income wages. As a result, a simple correlation share for the top 90 percent versus the bot- between FDI activity and labor market out- tom 40 percent (Krozer 2015).9 Annex 3B comes can either inflate or underestimate the discusses these different empirical specifica- true impact of FDI. To account for this poten- tions in more detail. tial bias, lagged global FDI growth was used as an instrument to capture supply-side changes that affect FDI inflows and eventu- Results ally MNE presence but are unlikely to be cor- This section discusses results on the average related with other domestic shocks. impact of FDI and its implications for Operationally, the instrument is the growth in ­ a ggregate poverty, shared prosperity, and global FDI (greenfield FDI and mergers and inequality. As discussed, results using ordi- acquisitions [M&A]) interacted with a nary least squares (OLS) can either underes- region’s original shares of FDI per ISIC2 timate or overestimate the true impact of sector.8 FDI; this exercise finds indication of both Beyond the average impact, how FDI types of biases across different countries and affects the distribution of income will depend sectors. For the sake of brevity, only the on what types of workers benefit most from results from the instrumental variables (IV) these investments. The exercise attempts to estimations are presented. A full comparison answer this question from one main angle: and discussion of results can be found in the how the skill content of the sector receiving online appendix.10 1 0 0    Global Investment Competitiveness Report 2019/2020 FDI in Manufacturing Has Been Most sectors and workers’ education levels. In Effective in Shifting Employment toward annex 3C, tables 3C.3, 3C.4, and 3C.5 pres- the Formal Sector and Increasing Wages ent the FDI labor markets effects across the five broad sectors and by workers’ education On average, increases in MNE activities are for Ethiopia, Vietnam, and Turkey, respec- associated with increased formal employ- tively (see annex 3B, equation [3B.2]). To get ment in manufacturing but not in services. In a better sense of the magnitude of FDI’s annex 3C, table 3C.2 presents the second- effects, annex 3C, table 3C.6 presents the stage IV results from the baseline specifica- overall population-weighted average effects tion (see annex 3B, equation [3B.1]), for each of the three countries, across the estimated separately for manufacturing and five broad sectors. These are further summa- services. The results suggest that a worker’s rized stylistically in table 3.2. relative probability of formal employment Overall, there is considerable evidence of (versus informal employment in that sector) FDI’s skill premium in Turkey and Vietnam. tends to increase with the output share of Although higher-skilled (more educated) MNEs in the worker’s sector and region.11 workers benefit—enjoying greater formal However, this effect is significant only for employment opportunities and higher average manufacturing FDI (with positive average wages—the effects of FDI on low-skilled effects in Turkey and Vietnam). The esti- (less-educated) workers are less positive. For mated average effect of MNE presence in them, there is either no effect or their proba- services is both smaller in magnitude and bility of employment and wages decline rela- statistically imprecise. tively in regions with higher services FDI. Based on the coefficient estimates, the total As expected, the skill premium seems to be impact of FDI activity was calculated as a stronger in services than in manufacturing population-weighted average effect.12 These (annex 3C, tables 3C.4 and 3C.5). These calculations suggest that in the sample period, results are broadly consistent with previous manufacturing FDI has created around findings, as summarized in figure 3.4. 350,000 additional formal jobs in Vietnam The exercise also finds considerable differ- and around 40,000 new formal jobs in Turkey ences in the total effect of FDI across the five each year.13 Aggregate employment effects are broad sectors. minimal for Ethiopia because both baseline Extractives. FDI in the extractives sector manufacturing employment and FDI activities has little impact on household income, as have been relatively small. expected of a capital-intensive sector. Few The effects on average wages also vary by households are employed in this sector, and sector and by country. In annex 3C, table no significant effects from FDI on wage 3C.2 shows that increases in MNE presence growth or formal employment were identified in Vietnam were associated with very large, in Turkey or Vietnam, except among workers significant average wage increases in both with primary education in Vietnam. manufacturing (12 percent nationwide) and Low-skilled manufacturing. Benefits on services (2 percent across the country). In employment and wages are positive. Low- Turkey, average wages increased in manufac- skilled manufacturing FDI has the highest turing (nearly 8 percent), but there was no average effect in Ethiopia (+5 percent in for- significant effect on services. In Ethiopia, mal jobs, +24 percent in average wages) but large increases in manufacturing wages also has positive effects on Vietnam (+5 per- (almost 32 percent) were identified; no infor- cent in formal jobs, +12 percent in average mation was available for services. wages) and Turkey (+0.8 percent formal jobs, +6 percent in average wages). FDI Resulted in a Skill Premium, Interestingly, the beneficiaries from these Especially in Services Sectors gains differ significantly across the three countries: In Ethiopia, FDI has primarily ben- The effects from FDI presence on labor mar- efited those with no education or only kets vary significantly across the broad T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    101 TABLE 3.2  The Relative Effects of FDI on Labor Markets (Formal Employment and Wages) in Ethiopia, Vietnam, and Turkey, by Broad Sector and Worker Skill Level High-skilled Country Broad sector Average effect Low-skilled workers workers Ethiopiaa Low-skilled manufacturing Positive Positive No effect High-skilled manufacturing Positive Positive No effect Vietnam Extractives No effect No effect No effect Low-skilled manufacturing Positive Positive Positive High-skilled manufacturing Positive Positive Positive Low-skilled services Neutral Negative Positive High-skilled services Positive Negative Positive Turkey Extractives No effect No effect No effect Low-skilled manufacturing Positive Positive Positive High-skilled manufacturing Positive Positive Positive Low-skilled services Neutral Negative Positive High-skilled services Positive Negative Positive Source: World Bank summary, based on annex 3C, table 3C.6. Note: Low-skilled workers are defined as those with primary education or less, while high-skilled workers have completed at least secondary education. All results are relative to workers in sectors with less or no multinational enterprise (MNE) involvement. Sectors (extractives, manufacturing, or services) are aggre- gated based on International Standard Industrial Classification (ISIC) two-digit classification. For more about the subclassification, see annex 3C, table 3C.1. a. No data were available on foreign direct investment (FDI) in extractives or services for Ethiopia. primary education (reflecting low-skilled fac- workers with postsecondary education tory work). In Turkey, those with primary or (+5.5 percent), while workers with no educa- secondary education benefited most. In tion or only primary education both experi- Vietnam, the total average effects are concen- ence a relative decline in wages (–6.6 percent trated and similar among those with primary and –4.1 percent, respectively). education. In Turkey, the effects differ even more High-skilled manufacturing. FDI in strongly by education. Workers with second- high-skilled manufacturing sectors results in ary and postsecondary education enjoy both the highest average benefits for Vietnam formal employment benefits (+2 percent for (+5 percent in formal employment, workers with secondary education, +3 per- +14 percent in average wages). For Turkey cent for postsecondary education) and sizable and Vietnam, the benefits from this sector are wage benefits (+12 percent and +18 percent, concentrated among those with primary or respectively). In contrast, low-skilled workers secondary education. Turkey also had notable experience sizable relative declines in their positive average effects in employment probability of formal employment and wages. (+1.3 percent in formal jobs) and wages This applies particularly for those with no (+10.8 percent in average wages). Although education (–10 percent probability of formal Ethiopia saw some benefits in high-skilled employment, –49 percent in average wages) manufacturing (+0.3 percent in formal jobs, as well as for those with primary education, +23 percent in average wages), few were but to a much lesser extent (–2 percent in for- affected given that the whole sector employed mal employment, no significant change in fewer than 25,000 individuals, on average. wages, respectively). Low-skilled services. FDI in low-skilled High-skilled services. FDI in high-skilled services results in significant wage dispersion. services also results in significant wage disper- In Vietnam, wages increase relatively for sion. FDI in this sector has the biggest average 1 0 2    Global Investment Competitiveness Report 2019/2020 effects in Turkey, with positive effects on for- effect on services. Increased FDI in down- mal employment (+1.2 percent) and wages stream (buying) sectors is associated with a (+12.9 percent), concentrated among workers reduction in both formal manufacturing with secondary and postsecondary education. employment (–5 percent) and wages Yet effects are negative for the small group of (–36 percent).14 One potential explanation is workers with no education working in these that MNEs are switching from domestic to sectors, who are most likely to be displaced foreign suppliers of intermediate inputs, (with very high reductions in wages as a prompting a decline in domestic production, result). In Vietnam, the result is neutral in labor demand, and wages of sectors with terms of employment but positive in terms of forward links to FDI. Such a finding would average wage effect (+4.9 percent), driven warrant additional analysis to better under- entirely by workers with postsecondary edu- stand potential constraints between MNEs cation (+11.1 percent). and domestic suppliers. The Vertical Spillover Effects of FDI FDI’s Aggregate Effects on Poverty, Are Mixed Shared Prosperity, and Inequality Vary Notably among the Sample Countries When controlling for FDI’s direct effects, the results in other sectors (vertical labor market The aggregate effects of FDI on poverty, spillovers) are less conclusive. Findings are shared prosperity, and income inequality mixed across the three countries. The results were estimated by comparing the observed from the second-stage IV (specifications in income distribution against a counterfactual annex 3B, equation [3B.2]) are presented in distribution with no FDI presence. This cal- annex 3C, table 3C.7. These present sector- culation combines the estimated direct region coefficients that interact FDI output effects from FDI activity in manufacturing in share with intensity of a vertical sector’s the case of Ethiopia and from FDI activity engagement with FDI. The magnitude can be across extractives, manufacturing, and hard to interpret. To aid interpretation, this services in the cases of Turkey and Vietnam. table also includes the population-weighted The counterfactual income distribution is average effects from each regression. assumed to be the predicted wage income Ethiopia. No significant effect is found on where foreign firms’ revenue share is set to either backward or forward linkages. This zero. could indicate that relatively few domestic Undoubtedly, deriving counterfactual pre- firms are currently supplying MNEs in dictions from estimated differential responses Ethiopia. Another possibility is that the over- across regions or sectors might lead to under- all manufacturing sector is too small for any estimation or overestimation of the true total statistically significant results to appear. effects (see Adão, Arkolakis, and Esposito Vietnam. The FDI backward link appears [2019] for a discussion). Nevertheless, these to be the most important channel. When FDI estimates provide for a rough approximation in upstream (selling) sectors increases, wages as a useful first step to gauge the potential go up in the services sector (+5 percent), and aggregate effect of FDI. formal employment in both manufacturing Ethiopia. In Ethiopia, the effects from FDI and services increases (+4.2 percent and +1.7 are highly positive, with noticeable effects on percent, respectively). Both types of increase poverty reduction and shared prosperity, but may be the result of productivity increases they are limited in magnitude. The FDI wage linked to labor market benefits from accessing benefits accruing to low-skilled workers have cheaper or higher-quality inputs (such as an important pro-poor effect, concentrating MNE producers of intermediate inputs, or in the FDI benefits in the bottom 40 percent business services). ­ (figure 3.7, panel b). Consequently, FDI is Turkey. The effect of FDI’s forward link- associated with a lower average Palma ratio ages is negative in manufacturing and has no and Gini coefficient—that is, a decline in T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    103 FIGURE 3.7  Ethiopia: FDI Effects on Poverty, Shared Prosperity, and Inequality a. Poverty reductiona b. Shared prosperity 0 0 6 –5,000 Poverty headcount –0.01 Poverty ratio (%) Wage effect (%) –10,000 4 –15,000 –0.02 –20,000 –25,000 2 –0.03 –30,000 –35,000 –0.04 0 09 10 11 12 14 13 10 0 0 0 0 0 0 0 0 00 –2 –3 –4 –5 –6 –7 –8 –9 20 20 20 20 20 20 0– –1 10 20 30 40 50 60 70 80 90 Poverty headcount effect (at US$1.90/day) Income decile Poverty ratio effect (at US$1.90/day) Wage effects from FDI for bottom 40% Wage effects from FDI for upper 60% c. Income inequality c1. Palma ratiob c2. Gini coef cientc 2.8 0.48 Gini coefficient Palma ratio 2.6 0.46 2.4 0.44 2009 2010 2011 2012 2013 2014 2009 2010 2011 2012 2013 2014 Palma ratio (income share top 90%/bottom 40%) without FDI Gini coefficient without FDI Gini coefficient with FDI Palma ratio (income share top 90%/bottom 40%) with FDI Source: World Bank calculations. Note: Estimates of the effect on poverty headcount, Palma ratio, and Gini coefficients for 2013 are not available because firm-level data are not available for that year; FDI = foreign direct investment. a. The poverty headcount effect is the effect of multinational enterprises (MNEs) on raising workers’ income above the poverty headcount (at US$1.90/day). b. The Palma ratio is the ratio of the richest 10 percent of the population’s share of gross national income (GNI) divided by the poorest 40 percent’s share. c. The Gini coefficient summarizes the dispersion of income across the entire income distribution. It ranges from 0 (indicating perfect equality, where everyone receives an equal share) to 1 (perfect inequality, where only one recipient or group of recipients receives all the income). income inequality (figure 3.7, panel c). prosperity, but they are minimal on income However, given that relatively few households inequality. The wage increases from FDI have are employed in the manufacturing sector, the helped lift almost 24,000 individuals out of effects apply to only a small share of the poverty each year between 2007 and 2016, population. on average (figure 3.8, panel a). The wage Wage increases from FDI are found to have income benefits from FDI are positive for all reduced poverty for only around 35,000 indi- workers along the income distribution. While viduals in 2010 (0.04 percent of the popula- the largest wage increases are in the middle of tion) (figure 3.7, panel a). This suggests that the distribution, increases in income are also although FDI can offer a powerful tool for significant for the bottom 40 percent poverty reduction, a greater focus on invest- ­ (figure 3.8, panel b). Given the fairly evenly ment promotion is needed to extend the ben- distributed wage benefits, the effects from efits to a wider share of the population. FDI on income inequality are minimal, with Vietnam. In Vietnam, effects from FDI are almost no changes in the Palma ratio and very positive for poverty reduction and shared small increases of the Gini coefficients over 1 0 4    Global Investment Competitiveness Report 2019/2020 time, with and without FDI (figure 3.8, (figure 3.9, panel a). In contrast, higher- panel c). skilled workers enjoyed strong rises in their Turkey. In Turkey, FDI has had the most income when FDI flowed in, thus presenting pronounced distributional effects. Although evidence of skill premiums leading to wage manufacturing FDI helped raise wages for dispersion. Overall, FDI has contributed to lower-skilled workers, there is evidence of shared prosperity (with positive effects for displacement for lower-skilled workers from both the bottom 40 percent and top 60 per- services FDI. In total, FDI provided minor cent) ­(figure 3.9, panel b), but it did struc- benefits to poverty reduction (affecting at turally contribute to income inequality (as most 15,000 individuals, or around identified by the Palma ratio and Gini coef- 0.02 percent of Turkey’s population) ficient) (figure 3.9, panel c). FIGURE 3.8  Vietnam: FDI Effects on Poverty, Shared Prosperity, and Inequality a. Poverty reductiona b. Shared prosperity 0 0 4 Wage effect from FDI (%) –0.02 Poverty headcount –20,000 3 Poverty ratio (%) –0.04 –40,000 2 –0.06 –60,000 1 –0.08 –80,000 –0.10 0 11 0 0 0 0 0 00 08 09 10 12 13 16 10 0 0 14 07 15 0 –2 –4 –5 –7 –9 –6 –8 –3 20 –1 20 20 20 20 20 20 0– 20 20 20 10 30 40 60 80 50 70 20 90 Poverty headcount effect: lower-middle-income countries (at US$3.20/day) Income decile Poverty ratio effect: lower-middle-income countries (at US$3.20/day) Wage effects from FDI for bottom 40% Wage effects from FDI for upper 60% c. Income inequality c1. Palma ratiob c2. Gini coefficientc 1.4 0.45 1.3 0.40 Gini coefficient Poverty ratio 1.2 0.35 1.1 0.30 1.0 0.9 0.25 11 8 09 10 11 12 13 16 08 2 13 16 10 09 4 07 14 15 7 15 0 1 1 0 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Palma ratio (income share top 90%/bottom 40%) without FDI Gini coefficient without FDI Gini coefficient with FDI Palma ratio (income share top 90%/bottom 40%) with FDI Source: World Bank calculations. Note: Estimates of the effect on poverty headcount, Palma ratio, and Gini coefficients for 2008, 2009, 2011, and 2012 are not available because firm-level data are not available for those years. FDI = foreign direct investment. a. The poverty headcount effect is the effect of multinational enterprises (MNEs) on raising workers’ income above the poverty headcount (at US$3.20/day). b. The Palma ratio is the ratio of the richest 10 percent of the population’s share of gross national income (GNI) divided by the poorest 40 percent’s share. c. The Gini coefficient summarizes the dispersion of income across the entire income distribution. It ranges from 0 (indicating perfect equality, where everyone receives an equal share) to 1 (perfect inequality, where only one recipient or group of recipients receives all the income). T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    105 FIGURE 3.9  Turkey: FDI Effects on Poverty, Shared Prosperity, and Inequality a. Poverty reductiona b. Shared prosperity 0 0 8 Wage effect from FDI (%) –5,000 Poverty headcount Poverty ratio (%) 6 –0.01 –10,000 4 –15,000 –0.02 –20,000 2 –25,000 –0.03 0 10 0 0 0 0 0 0 0 0 00 09 11 13 15 –2 –3 –4 –5 –6 –7 –8 –9 0– –1 20 20 20 20 10 20 30 40 50 60 70 80 90 Poverty headcount effect: upper-middle-income countries (at US$5.50/day) Income decile Poverty ratio effect: upper-middle-income countries (at US$5.50/day) Wage effects from FDI for bottom 40% Wage effects from FDI for upper 60% c. Income inequality c1. Palma ratiob c2. Gini coef cientc 1.4 0.45 1.3 0.40 Gini coefficient Poverty ratio 1.2 0.35 1.1 0.30 1.0 0.9 0.25 2009 2010 2011 2012 2013 2014 2015 2016 2009 2010 2011 2012 2013 2014 2015 2016 Palma ratio (income share top 90%/bottom 40%) without FDI Gini coefficient without FDI Palma ratio (income share top 90%/bottom 40%) with FDI Gini coefficient with FDI Source: World Bank calculations. Note: FDI = foreign direct investment. a. The poverty headcount effect is the effect of multinational enterprises (MNEs) on raising workers’ income above the poverty headcount (at US$5.50/day). b. The Palma ratio is the ratio of the richest 10 percent of the population’s share of gross national income (GNI) divided by the poorest 40 percent’s share. c. The Gini coefficient summarizes the dispersion of income across the entire income distribution. It ranges from 0 (indicating perfect equality, where everyone receives an equal share) to 1 (perfect inequality, where only one recipient or group of recipients receives all the income). Policy Implications all have experienced notably different effects on poverty reduction, shared prosperity, and The evidence in this chapter so far has shown income inequality. In general, FDI in lower- that FDI plays an important role in shaping skilled, tradable, labor-intensive sectors have labor markets, affecting both aggregate had the most significant pro-poor impact. labor demand and skill-biased employment FDI in higher-skilled, less tradable sectors and wage dynamics. The three countries tend to benefit the more-educated workers at examined in more detail (Ethiopia, Vietnam, the expense of those at the lower end of the and Turkey) further confirm the varied income distribution. impact that FDI can have across different This section extracts lessons from the types of sectors and by workers’ education experience of the analyzed countries and levels. Although all three countries have incorporates other empirical evidence to been relatively successful at attracting FDI, suggest which complementary policies can 1 0 6    Global Investment Competitiveness Report 2019/2020 enhance FDI’s contribution to poverty A country’s skill base may also have impli- reduction and inclusive growth. Specific cations for relevant FDI source countries. recommendations to help improve the Source economies to be targeted may be those developmental impact of FDI fall into two that have firms with capacity in these sectors, areas: (a) investment policy and promotion and especially at a level of technology that is efforts, and (b) labor market policies. complementary to the host economy. If the technology gap between foreign and domestic firms is too wide, it will be difficult for domes- tic firms to benefit from direct knowledge Investment Policy and Promotion for transfer or spillovers, limiting the develop- Development mental impact of the FDI on inclusive growth Recommendation 1: Align investment promo- (Perea and Stephenson 2018). tion with a country’s labor skill base, related Recommendation 2: Strengthen the sectors, and relevant source countries. absorptive capacity of domestic firms and Not all FDI will have the same impact workers (such as through programs fostering on an economy. Thus, investment promo- FDI-supplier linkages and employment train- tion officials may wish to target FDI that is ing) to help extend the labor market benefits most likely to bring the impact they seek from FDI. (Javorcik 2004). For the purposes of Host governments may wish to increase designing investment promotion strategies the level of absorptive capacity of their econ- and adopting investment promotion omy to increase the potential benefit of FDI. measures, officials may wish to consider The results on vertical spillovers illustrated targeting based on the country’s skill the potential impact that FDI-supplier link- base ,  related sectors, and relevant source ages can have on household income. In the countries. case of Ethiopia, the study found evidence The case studies showed that FDI has the that FDI in forward-linked areas has a posi- biggest effect on inclusive income growth tive effect on employment and wages (likely when it is aligned with a country’s skill base. capturing the broader demand effects from While FDI in low-skilled manufacturing manufacturing FDI on their suppliers). resulted in the highest average welfare bene- However, in the case of Turkey, the study fits for Ethiopia and Vietnam, Turkey bene- found a negative effect on forward-linked fited most from FDI in higher-skilled FDI, suggesting that MNEs are switching manufacturing. In addition, Turkey was the from domestic to foreign suppliers of interme- only country in the sample where benefits diate inputs, in which case production, labor from higher-skilled services increased signifi- demand, and wages of sectors forward linked cantly, on average. A likely reason is that to FDI can decline. Turkey’s population is higher-skilled on aver- These two cases suggest that absorptive age, making it easier to employ domestic capacity is relative to the type of FDI attracted. workers and absorb knowledge transfers Turkey’s domestic suppliers likely have higher from MNEs. absorptive capacity than Ethiopia’s suppliers, In practical terms, this suggests that to best yet Turkey’s MNEs are also involved in more- use FDI for household income growth, invest- complex production (placing additional ment promotion strategies should start with demands on Turkey’s suppliers). To better an assessment of the country’s labor skill base improve FDI-supplier linkages would there- (possibly through a jobs diagnostic).15 Sectors fore warrant additional analysis to better to be targeted may then be those in which the understand potential constraints between host economy has some skill endowment and MNEs and domestic suppliers in Turkey. in which it wishes to develop greater capacity Policy makers have a role to play in help- when considering its national development ing develop the hard infrastructure and soft strategy. skills needed for domestic firms to assimilate T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    107 knowledge and technology brought by for- can thus capture market share from domestic eign firms when MNEs invest in the economy firms, resulting in little or no aggregate (Amann and Virmani 2014). As seen in the employment effects. Negative employment empirical evidence, skill level can make the effects possibly indicate the use of labor- difference between being locked into a low- saving technology. income, low-skill cycle of FDI or moving up However, liberalization of services can also to a high-income, high-skill cycle of FDI (Te stimulate long-term economic development Velde and Xenogiani 2007). Measures to by raising an economy’s overall total factor consider include promoting firm linkages, productivity (the efficiency with which societ- boosting R&D expenditures, increasing ies combine labor, capital, and technology) R&D employment, providing training pro- (Van der Marel 2012). For 86 developing grams, building business networks, establish- countries from 1985 to 1999, those that ing institutional partnerships, creating opened up their financial and telecommunica- national infrastructure, helping to diffuse tions sectors grew, on average, 1.5 percentage information, and designing appropriate points faster than countries that did not open school curricula. The specific type of relevant up these sectors (Mattoo, Rathindran, and policy will depend in large part on a country’s Subramanian 2006). Similarly, for 20 transi- economy and stage of development. What is tion economies from 1990 to 2004, enabling most critical is an approach of continuous entry of FDI in select services (finance, power, learning and adaptation for domestic firms transport, and telecommunications) was asso- and the domestic economy to maximize the ciated with large new investments and pro- benefits of FDI. vided statistically significant explanatory Policy makers’ interventions to boost variables for their economic performance absorptive capacity should be considered at after 1990 (Eschenbach and Hoekman 2006). the national level. Absorptive capacity can be This suggests that services FDI may have measured at the levels of the firm and the important dynamic effects in stimulating economy. To provide the most transforma- household income growth. tional benefits for the country from FDI, it is Removing restrictions on foreign equity in most important to adopt the necessary hori- the services sector is also known to improve zontal measures that can boost the absorptive productivity in other sectors and can capacity of the economy as whole rather than therefore improve indirect job creation. try to boost the absorptive capacity of indi- Evidence from the Czech Republic, India, vidual firms (Perea and Stephenson 2018). and Indonesia shows that opening up Recommendation 3: Opening up services services to foreign investors improves the sectors to foreign investment can (indirectly) productivity of domestic firms in manufac- create new jobs. Governments may wish to turing, which may raise average wages in this combine the promotion of services FDI with sector (Arnold et al. 2016; Arnold, Javorcik, progressive labor market interventions to and Mattoo 2011; Duggan, Rahardja, and ensure that both high- and low-skilled work- Varela 2013). This chapter found evidence ers in the services sector benefit. that in Vietnam, FDI in backward-linked The three case studies suggest that FDI in sectors (such as professional business services has little direct effect on aggregate services) led to higher employment in manu- (formal) job creation. Although FDI in high- facturing and to higher wages in the service skilled services led to a minor increase in sector. This suggests that services liberaliza- formal employment for Turkey, FDI in tion can assist indirectly in creating jobs. lower-skilled services led to a decline To limit the potential increase in wage (Turkey) or no effect on employment inequality from services FDI in the short run, (Vietnam). A possible reason for this is that governments can prioritize promoting ser- most services firms operate exclusively vice sectors that are export oriented or that within the domestic market. FDI in services most affect (export-oriented) manufacturing. 1 0 8    Global Investment Competitiveness Report 2019/2020 The case studies showed that FDI in services and increases in income inequality. This chapter is associated with the largest increases in shows that FDI is associated with skill premi- wage inequality because lower-skilled work- ums that can increase wage inequality (as seen ers are displaced in favor of higher-skilled in Turkey). Yet evidence also suggests that FDI workers. As shown in the conceptual frame- can bring about productivity improvements work, FDI is most associated with wage and structural transformation that boost long- inequality in services sectors that are (a) term growth. Hence, rather than undermine more skill intensive (see the broad sectoral FDI flows, the best way to ensure inclusive ­ ox 3.2 and annex 3C, table classification in b growth is to complement investment policy 3C.1); (b) domestically oriented (such as with progressive labor market policies.16 construction and security services) rather Recent evidence suggests that some active than export oriented (such as tourism, ICT, labor market policies, including vocational and business process outsourcing); and (c) training and employment subsidies, are not dominated by small, family-owned cost-effective (box 3.3). However, that does businesses (such as retail). not mean that government policy cannot For that reason, it can make sense to priori- improve labor market outcomes. Three types tize promoting services in areas that are either of labor market policies have had the most export oriented or that strongly affect export- beneficial effect on labor markets: (a) oriented manufacturing (such as logistics, improving labor standards and labor transport, and wholesale trade) and adopt representation; (b) providing labor market robust social safety net policies to compensate information and skills certification to vulner- for potential losers, as discussed below. able communities; and (c) stimulating inter- nal labor mobility. These are discussed in Recommendations 4–6. Progressive Labor Market Policies Recommendation 4: Improve bargaining Progressive labor market policies can be impor- power and knowledge spillovers for workers tant complements to counter FDI’s effect on by enforcing sufficient labor standards and any (temporary) declines in ­ formal employment supporting labor representation. BOX 3.3 The Limited Effectiveness of Traditional Active Labor Market Policies A systematic survey of 24 randomized control E mploy ment subsid ie s a re a lso not trials of active labor market programs in 10 effective, especially when firms had to developing countries found that “these pro- formally register workers, the literature grams have at best modest impact in most cir- suggests. In the one study, initial posi- cumstances” (McKenzie 2017). tive results disappeared quickly once the Vocational training programs were found subsidy ended. No study identified any to have modest effects on employment. “For long-term impact on employment from every 100 people offered vocational training, subsidies (McKenzie 2017). Such evidence fewer than 3 will find a job they would not from developing countries is consistent with have otherwise found” (McKenzie 2017). The findings of ineffective active labor market study finds that the cost of employing one policies in developed countries (Crépon and additional person ranged from approximately van den Berg 2016). US$17,000 to US$60,000. T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    109 While FDI may be job creating, FDI bene- little opportunity to learn on the job or receive fits in terms of creating “good” jobs are not knowledge spillovers from MNEs. This leads automatic. The OECD defines job quality to low labor efficiency for the firm and little based on three dimensions: wages, labor mar- in productivity-linked wage benefits for the ket security, and the working environment workers.17 Government can play a role by (Cazes, Hijzen, and Saint-Martin 2015). enforcing labor standards throughout an Although this study finds that FDI has an industry, which is likely to reduce turnover, impact on wages, evidence regarding FDI and raise employee health benefits, and give firms labor market security and the quality of the a reason to invest in training their workers. working environment is mixed (Hijzen et al. Better labor representation can also reduce 2013). Some studies have found that foreign wage inequality, but more research is needed firms adapt to local labor practices rather on ways to balance the costs and benefits of than export the labor practices of their home labor laws (Blattman and Dercon 2018). economy (Almond and Ferner 2006). Others There is considerable evidence from devel- have argued that governments have been oped countries that unions can reduce wage tempted to lower labor standards to attract inequality (see, for example, Dinardo and FDI, resulting in an international “race to the Lemieux 1997; Lemieux 1993) and that dif- bottom” (Bernard and Jensen 2007; Bernard ferences in the rate of deunionization are cor- and Sjöholm 2003). Yet, in many cases, such related with differences in the growth of policy is misguided, given that efficiency- inequality (Card, Lemieux, and Riddell 2004, seeking firms care about low unit labor costs 2018; Gosling and Lemieux 2004). This evi- (the average cost of labor per unit of output) dence has led some experts to argue that labor rather than the lowest employment costs law reforms should be part of any policy overall. Government-enforced labor stan- response to rising inequality (see, for exam- dards may raise labor costs but can also ple, Stiglitz 2012). Yet regulation also risks reduce unit labor costs by reducing turnover. raising labor costs (which could scare off FDI One way to reduce labor turnover is to in the country) and risks benefiting insiders improve safety standards—in turn creating (union members) at the expense of outsiders better opportunities for knowledge spillovers (those without employment). to workers and increasing labor productivity More research is needed in this area. A key for firms. Many of the manufacturing jobs example comes from labor standards cam- associated with FDI in low-income countries paigns in Indonesia, which led to large real are associated with significant health risks. wage increases in targeted firms, with some Blattman and Dercon (2018) find that, in costs (falling profits) but no adverse employ- Ethiopia, chances of a chronic health issue ment effects (Harrison and Scorse 2010). In went up 1 percentage point for every month the absence of union representation (which someone works in an industrial firm (associ- takes time to establish), governments can also ated with chemical use and dirty air). foster workers’ councils to grant employees a Combined with extremely low pay, this situa- voice and a venue to air grievances (Barrett tion meant that worker turnover was very and Baumann-Pauly 2019). high, and 77 percent of workers quit within Recommendation 5: Support vulnerable their first year. Other estimates also find evi- communities (such as lower-skilled workers, dence of extreme annual turnover in Ethiopia, youth, and women) by providing jobs infor- ranging from 60 percent to 120 percent (with mation and skills certification. complete turnover of more than one cohort Governments can also help vulnerable occurring in some years) (Barrett and communities by lowering search costs in the Baumann-Pauly 2019). When workers are labor market by offering public information employed for such a short period, they have about jobs and organization of job fairs. 1 1 0    Global Investment Competitiveness Report 2019/2020 Providing job seekers with information about is costly, and informal insurance in the form the labor market can help improve employ- of relying on families and communities means ment. In rural India, informing young women that individuals rarely move (Munshi and about urban ICT jobs and helping them with Rosenzweig 2016). the application process meant that these Some of the most successful labor market women were 4.6 percentage points more interventions are those that helped workers likely than others to work in ICT jobs (Jensen access job opportunities in a different loca- 2012). In Ethiopia, providing job seekers with tion (Jensen 2012) or subsidized job searches bus fares to search for vacancies or attend job in different parts of the city (Abebe et al. interviews improved labor market outcomes 2016; Franklin 2015).18 More striking evi- (Abebe et al. 2017). Finally, in the Philippines, dence comes from Bryan, Chowdhury, and a job fair also allowed attendees to learn Mobarak (2014), who show that a small sub- about their labor market prospects, increasing sidy equal to the cost of a bus ticket spurred the probability of working in a formal job by new seasonal migration in Bangladesh, which 11 percentage points (compared with the con- improved employment opportunities and trol group) (Beam 2016). increased household consumption (analogous Recent experimental evidence also shows to income) by 30–35 percent. that programs that certify existing skills can Overall, this report calls for an integrated also help job seekers find better jobs. In approach that combines (a) proactive Ethiopia, workers who attended job applica- investment policy and promotion (including tion workshops that provided skill certificates targeting FDI, tackling absorptive capacity, as well as training on résumés, cover letters, and liberalizing services with the best and job interviews had 20 percent higher prospects for employment and wage earnings than comparable workers in the con- increases) with (b) progressive labor market trol group. These gains were concentrated policies (including support for workers’ labor among those with the least education and standards and labor representation, active experience (Abebe et al. 2018). In Uganda, provision of information to vulnerable job certificates of soft skills led employed workers seekers, and support programs to stimulate to earn 11 percent more in the two years after internal migration). Policies that can align the intervention (Bassi and Nansamba 2018). these different elements have the best chance Certificates work best when they focus on of ensuring that FDI will bring benefits to the general skills rather than apprenticeships, host economy, stimulate poverty reduction, which focus on firm-specific skills that are promote shared prosperity, and produce harder to certify and were valued less by other inclusive growth. firms in the market (Alfonsi et al. 2017). In sum, reducing information gaps can increase employment quality and earnings for job seekers in vulnerable communities (Caria and Annex 3A. Data Description Lessing 2019). The analysis links household survey data Recommendation 6: Establish programs to with firm-level data. The main source of stimulate internal migration. household data comes from the World Regulatory changes and support programs Bank’s International Income Distribution to stimulate internal migration can further Database (I2D2). The I2D2 is a harmonized help improve access to employment, with dataset covering more than 900 nationally important welfare benefits (Newman et al. representative household surveys from more 2016). FDI often creates more jobs close by, than 160 countries.19 The data include the which often means that large wage benefits industry of employment, which can be har- are restricted to urban areas. Stimulating monized in all countries to two-digit ISIC urbanization can help expand some of these (rev. 4) industries. In addition, it includes benefits to the wider population. Yet moving information on workers’ characteristics T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    111 (gender, age, education) and geographic indicator takes a value of 1 if a worker is location (region). The analysis uses this sec- either a paid employee or employer, and 0 if tor and location information to match with the individual is a nonpaid employee or self- the variable of interest concerning MNE employed. In Turkey, this indicator takes a activities. For Ethiopia, this resulted in five value of 1 if a worker contributes to social surveys between 2009 and 2014. For security (a common indicator of formal Vietnam, to increase overlap with the firm- employment), and 0 otherwise. The measure level data, the I2D2 was supplemented with of wages was normalized to monthly pay- Labor Force Surveys for 2013–16. For ments in constant 2010 terms. Turkey, household surveys recorded two- The main variable of interest is defined as digit industrial sectors only from 2009 the share of FDI firms’ revenue in a sector onward (previously, it was one-digit). and region. Because of differing data avail- Various firm-level datasets in each country ability, there are some variations in how this were used to match with the household-level variable is measured across the three sample data. For Vietnam and Turkey, the analysis countries. In all countries, a region is defined relies on information from the Enterprise as a level-2 administrative region (NUTS2 in Census and Enterprise Information System, the case of Turkey, and province in the case respectively. Both include information on of Ethiopia and Vietnam), which results in firms from all sectors in the economy. For 11 regions in Ethiopia, 26 in Turkey, and 64 Ethiopia, a census of all manufacturing estab- in Vietnam. Foreign firms are defined as lishments with 10 or more employees, the those with at least 10 percent foreign owner- Large and Medium Manufacturing Survey, ship in Ethiopia and Turkey. In Vietnam, was used. foreign firms are those identified as having From firm-level data, the share of reve- any positive foreign shares, given the lack of nue (employment) by foreign firms in total data on shareholders in some years. output (employment) was calculated as a Nevertheless, when this variable is available, proxy for FDI activity. This MNE output a high correlation is found between the two share is estimated annually, across each definitions. region and ISIC two-digit sector. In Vietnam and Turkey, the data include information on firms in all sectors. In Ethiopia, the Annex 3B. Empirical survey covers manufacturing activity only. Specifications As a result, the analysis in Ethiopia is restricted to FDI in the manufacturing The analysis is performed separately for each sector, which nevertheless captures 70–89 of the three country case studies. The base- percent of annual FDI inflows since 2007 line model estimates the following: (EIC 2017). The analysis covers the periods yisrt = β × FDI srt −1 + δ Xirst + γ ∗ tariff st + d s with significant FDI growth in Ethiopia (2009–14), Vietnam (2007–16), and Turkey + drt + ε irst , (3B.1) (2009–16).  Before conducting the analysis, all the where, y denotes formal employment or (log) datasets were restricted to only working-age wages, i is the specific individual, s is the two- individuals (ages 15–65). The focus is on digit sector, r is the region within a country, employed individuals. Two main outcomes and t is the year. FDIsrt–1 denotes lagged FDI were constructed: an indicator for whether a activity, calculated as the share of foreign worker is formally employed and the work- firms’ revenue in the total output of a sector er’s wage level. Because of differences in the and region within a country. b is the main household survey, the definition of formal coefficient of interest, which measures the employment varies slightly across the three change in the probability of formal countries. In Ethiopia and Vietnam, this 1 1 2    Global Investment Competitiveness Report 2019/2020 employment or the percentage change in yisrt = β fw × ∑ j FDI j→s,srt + β fw wages associated with a unit change in FDI activity. FDI j→k,srt + δ Xirstγ ∗ (3B.2) ×∑ k Implicitly, this specification assumes that tariff st + d s + drt + ε irst , the relevant labor market is within a sector and region. There is an adjustment cost to where ∑ j FDI j→s,srt and ∑ k FDI s→k,srt denote move between sectors and regions, so there are differences in individual employment and the weighted sum of FDI output shares in all wages due to differences in FDI activities. The sectors supplying to or buying from sector s, analysis controls for a set of individual char- where the weights are the technical acteristics in Xirst’ including age, gender, and coefficients in the input-output table for each education level to account for potential selec- country, to capture the degree of linkages tion of workers into regions and sectors with between sectors. higher FDI. The sectoral fixed effects, ds’ Intuitively, this specification examines the control for inherent differences in sectoral extent to which labor demand in a sector is labor demand that could be correlated with affected by backward and forward linkages FDI attractiveness. with other foreign-invested sectors. The Equation (3B.1) is estimated using an instrument is the weighted sum of the original instrumental variable, where FDIsrt is instru- instrument for FDI in each sector and region. mented for by growth in global FDI (green- To understand the distribution impact of field and M&A) in sector s in year t -4, FDI, the following specification is estimated: interacted with the original shares of FDI in region sector rs (that is, the shares at the begin- yisrt = β FDI rst + nFDI rst × Eduisrt + δ Xirst ning of the sample period). A quadratic term is + γ ∗ tariff st + d s + drt + ε irst . (3B.3) included of the instrument to capture potential  nonlinear effects between the instrument and Here, the interactions between FDI and the variable of interest. For example, agglom- workers’ education is included to capture eration effects might imply that the original how worker outcomes vary with their educa- shares of FDI would have an exponential tion levels. In aggregate, the total effect will effect on FDI activities in later years. also depend on the types of sectors receiving Including another excluded instrument FDI and the nature of FDI (low skill versus also allows the analysis to formally test for high skill, tradable versus nontradable, labor the exogeneity of the instruments and the intensive versus capital intensive). error terms. Global FDI captures supply The impact of FDI on the income distri- shocks that are unlikely to be correlated with bution is then aggregated using estimates other domestic changes. Nevertheless, this from equation (3B.3). For each individual, instrument is not exogenous if regional shocks the counterfactual wage without FDI is pre- occur that affect both the labor markets and dicted assuming a zero FDI value. The aggre- FDI shares. To account for this, the model gate poverty level, total income of the also controls for a set of region-year fixed bottom 40 percent, and the inequality effects. Finally, for all nonservices sectors, the indicators (Palma ratio and Gini coeffi- model also controls for average tariff in the cient) are estimated for the actual and this sector to separate out the potential impact of counterfactual (wage) income distribution. FDI from trade liberalization, given that FDI The contribution of FDI to poverty reduc- reforms are often accompanied by trade tion, shared prosperity, and inequality are liberalization. then calculated as the differences between The effect of vertical FDI is estimated using these numbers. the following specification: T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    113 Annex 3C. Additional Tables TABLE 3C.1  Broad Sectoral Classification and Their Subsectors Broad sector Subsectors Extractives Mining of coal and lignite Extraction of crude petroleum and natural gas Mining of metal ores Other mining and quarrying Mining support service activities Low-skilled manufacturing Food, beverages, and tobacco products Wood and wood products Other nonmetallic mineral products Fabricated metal Paper and paper products; printing and publishing Rubber and plastics products Basic metals Textiles, wearing apparel, and leather products Furniture; manufacturing n.e.c. (not specified) High-skilled manufacturing Coke and refined petroleum products Chemicals and chemical products Machinery and equipment n.e.c. (not specified) Transport equipment Electrical machinery and equipment Computer, electronics, and optical equipment Pharmaceutical products Low-skilled services Construction Wholesale and retail trade; repair of motor vehicles and motorcycles Transportation and storage (land, warehousing) Accommodation and food service activities Security, landscape, and employment activities High-skilled services Transportation and storage (water, air, postal) Information and communication Financial and insurance activities Professional, scientific, and technical activities Travel agencies and tour operators Office administration and other business support activities Source: World Bank, based on Hallward-Driemeier and Nayyar 2017. Note: Based on International Standard Industrial Classification (ISIC) rev. 4 classification. Extractives follows ISIC 05–09. Low-skilled manufacturing covers ISIC 10–18, 22–24, 31–32. High-skilled manufacturing covers ISIC 19–21, 25–30, 33. Low-skilled services covers ISIC 41–49, 52, 55–56, 68, 77–78, 80, 81. High-skilled services covers ISIC 50–51, 53, 58–66, 69–75, 79, 82. n.e.c. = not elsewhere classified. 1 1 4    Global Investment Competitiveness Report 2019/2020 TABLE 3C.2  FDI Labor Market Effects: All Countries (Second-Stage IV Results) Vietnam Turkey Ethiopia (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Outcome Formal Formal Formal Formal Formal Variables Wages (LN) Wages (LN) Employment Employment Wages (LN) Wages (LN) Employment Employment Wages (LN) Employment Household Manu- Manu- Manu- Manu- Manu- Manu- Sample facturing Services facturing Services facturing Services facturing Services facturing facturing MNE output share 0.340*** 0.577** 0.141** -0.0200 1.054*** 1.033 0.122*** -0.010 1.156* 0.124 (lagged Y1) (2.78) (2.20) (2.45) (-0.16) (0.356) (1.247) (0.045) (0.168) (0.654) (0.100) Additional controls Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Gender, Age Gender, Age Gender, Age Gender, Age Gender, Age Gender, Age Gender, Age Gender, Age Gender, Age Gender, Age Sectoral tariffs Yes No Yes No Yes No Yes No Yes Yes Region × year fixed Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes effects ISIC2 sector fixed Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes effects Observations 265,335 489,660 393,905 885,941 202,855 392,785 211,900 418,975 10,683 11,084 R-squared 0.405 0.342 0.260 0.296 0.050 0.049 0.053 0.049 0.120 0.082 Hansen J statistic 0.222 0.168 0.000 0.061 0.7107 0.6907 0.9815 0.3867 0.2904 0.2700 (p-value) Kleibergen-Paap 107.6 34.8 115.9 37.2 42.4 25.3 42.4 25.1 11.4 11.6 LM test (test) Kleibergen-Paap 4.21e-24 2.83e-08 6.91e-26 8.42e-09 0.0000 0.0000 0.0000 0.0000 0.0033 0.0030 LM test (p-value) Kleibergen-Paap 68.39 35.28 88.94 34.56 142.44 24.11 139.64 23.17 10.36 10.93 Wald rk F statistic Total average 12.0 2.0 5.2 0 7.7 0  0.9  0 31.5 0  effects (%) Source: World Bank. Note: FDI = foreign direct investment; IV = instrumental variables; ISIC2 = International Standard Industrial Classification of All Economic Activities two-digit code; LM = Lagrange Multiplier; LN = natural logarithm; MNE = multinational enterprise; Y1 = Year 1. Standard errors are clustered at the region*sector level: *** p < 0.01 ** p < 0.05 * p < 0.10 T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    TABLE 3C.3  FDI Labor Market Effects by Broad Sector and Education Level: Ethiopia (Second-Stage IV Results)   (1) (2) (3) (4) (5) (6) Outcome variables Wages (LN) Wages (LN) Wages (LN) Formal Emp. Formal Emp. Formal Emp. Low-skilled High-skilled Low-skilled High-skilled Household sample Manufacturing manufacturing manufacturing Manufacturing manufacturing manufacturing MNE output share (lagged Y1)*no education 1.810*** 1.793*** 17.20*** 0.239** 0.233** 2.295*** (0.637) (0.643) (3.780) (0.104) (0.107) (0.484) MNE output share (lagged Y1)*primary education 1.595** 1.913** 1.108 0.199* 0.235** 0.0983   (0.729) (0.781) (1.177) (0.109) (0.117) (0.145) MNE output share (lagged Y1)*secondary education 2.741 2.750 — 0.332 0.198 —   (2.672) (3.143) (.) (0.400) (0.467) (.) MNE output share (lagged Y1)*postsecondary education –0.772 –0.847 1.255 –0.246 –0.260 0.0120   (1.225) (1.609) (1.408) (0.171) (0.225) (0.196) Additional controls Education, Education, Education, Education, Education, Education, Gender, Age Gender, Age Gender, Age Gender, Age Gender, Age Gender, Age Sectoral tariffs Yes Yes Yes Yes Yes Yes Region-year fixed effects Yes Yes Yes Yes Yes Yes ISIC2 sector fixed effects Yes Yes Yes Yes Yes Yes Observations 10,683 10,024 659 11,084 10,411 673 R-squared 0.112 0.106 –0.082 0.071 0.069 –0.182 Source: World Bank. Note: See first-stage results in the supplementary appendix; FDI = foreign direct investment; ISIC2 = International Standard Industrial Classification of All Economic Activities two-digit code; IV = instrumental variables; LN = natural logarithm; MNE = multinational enterprise; Y1 = Year 1. Both the dash and ( . ) mean that there were insufficient observations for this, and so no results were identified. Standard errors are clustered at the region*sector level: *** p < 0.01 ** p < 0.05 * p < 0.10 115 1 1 6    Global Investment Competitiveness Report 2019/2020 TABLE 3C.4  FDI Labor Market Effects by Broad Sector and Education Level: Vietnam (Second-Stage IV Results) (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) Outcome Wages Wages Wages Wages Wages Wages Wages Formal Formal Formal Formal Formal Formal Formal variables (LN) (LN) (LN) (LN) (LN) (LN) (LN) Emp. Emp. Emp. Emp. Emp. Emp. Emp. Low- High- Low- High- Low- High- Low- High- Household Manu- skilled skilled skilled skilled Manu- skilled skilled skilled skilled sample Extractives facturing manuf. manuf. Services services services Extractives facturing manuf. manuf. Services services services MNE output 1.560*** 0.627 0.836 -0.217 -3.791** -3.995** 4.230 -0.514*** 0.372*** 0.443*** 0.223* -0.401 -0.457 1.242 share (lagged Y1)* (9.48) (1.48) (1.58) (-0.75) (-2.30) (-2.22) (1.40) (0.16) (2.70) (2.72) (1.77) (-0.79) (-0.79) (1.60) no education MNE output -1.418 0.356** 0.445*** -0.005 -1.57** -1.838** 0.425 -0.750*** 0.259*** 0.329*** 0.043 -0.422 -0.509 -0.048 share (lagged Y1)* (2.48) (2.42) (2.62) (-0.03) (-2.13) (-1.99) (0.49) (0.32) (3.83) (4.17) (0.69) (-1.38) (-1.28) (-0.31) primary education MNE output 0.783 0.319** 0.368** -0.108 0.478 0.853* 0.028 0.029 0.149** 0.226*** -0.035 0.185 0.352* 0.004 share (lagged Y1)* (0.51) (2.43) (2.40) (-0.83) (1.21) (1.72) (0.10) (0.14) (2.57) (3.25) (-0.64) (1.31) (1.89) (0.02) secondary education MNE output 0.370 0.348*** 0.969*** -0.166 0.837*** 1.115** 0.211 -0.096 -0.103* -0.052 -0.199*** -0.162 -0.120 0.072 share (lagged Y1)* (0.44) (2.64) (3.33) (-1.28) (2.95) (2.22) (0.97) (0.14) (-1.65) (-0.56) (-3.46) (-1.16) (-0.64) (0.59) postsecondary education Additional Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, controls Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Age Age Age Age Age Age Age Age Age Age Age Age Age Age Sectoral tariffs Yes Yes Yes Yes No No No Yes Yes Yes Yes No No No Region-year fixed Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes effects ISIC2 sector fixed Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes effects Observations 11,436 265,325 215,664 49,661 489,641 446,685 42,956 17,450 393,895 324,270 69,625 885,922 826,071 59,851 R-squared 0.0339 0.0221 0.0121 0.0329 0.0140 0.0122 0.0534 -0.0284 0.0667 0.0654 0.0202 0.0944 0.0923 0.0956 Source: World Bank. Note: See first-stage results in the supplementary appendix. FDI = foreign direct investment; ISIC2 = International Standard Industrial Classification two-digit code; IV = instrumental variables; LN = natural logarithm; MNE = multinational enterprise; Y1 = Year 1. Standard errors are clustered at the region*sector level: *** p < 0.01 ** p < 0.05 * p < 0.10 TABLE 3C.5  FDI Labor Market Effects by Broad Sector and Education Level: Turkey (Second-Stage IV Results)   (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) Wages Wages Wages Wages Wages Wages Wages Formal Formal Formal Formal Formal Formal Formal T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    Outcome variables (LN) (LN) (LN) (LN) (LN) (LN) (LN) Emp. Emp. Emp. Emp. Emp. Emp. Emp. Low- High- Low- High- Low- High- Low- High- Manu- skilled skilled skilled skilled Manu- skilled skilled skilled skilled Household sample Extractives facturing manuf. manuf. Services services services Extractives facturing manuf. manuf. Services services services MNE output share 0.119 0.189 0.155 0.546 -11.690*** -11.040*** -2.772 -0.510 0.260** 0.290* 0.075 -2.344*** -2.324*** -0.316 (lagged Y1)*no (1.070) (1.145) (1.372) (0.763) (2.377) (2.558) (1.725) (0.488) (0.122) (0.148) (0.092) (0.381) (0.404) (0.203) education MNE output share 0.184 1.424*** 1.902*** 0.571* -1.859 -0.762 -1.989 0.007 0.174*** 0.236*** 0.053 -0.467*** -0.353* -0.204 (lagged Y1)*primary (0.326) (0.414) (0.571) (0.312) (1.167) (1.433) (1.243) (0.013) (0.052) (0.073) (0.039) (0.158) (0.191) (0.153) education MNE output -0.512 0.813*** 1.099** 0.232 3.269*** 5.469*** -0.191 0.016 0.076** 0.087 0.021 0.363*** 0.610*** -0.075 share (lagged Y1)* (0.730) (0.307) (0.509) (0.245) (1.018) (1.346) (0.967) (0.015) (0.038) (0.063) (0.031) (0.133) (0.176) (0.109) secondary education MNE output 0.322 0.431 0.550 0.024 4.037*** 2.505 1.412 0.007 0.041 0.031 0.009 0.690*** 0.712*** 0.089 share (lagged Y1)* (0.489) (0.317) (0.522) (0.280) (1.245) (1.600) (1.256) (0.019) (0.039) (0.063) (0.034) (0.186) (0.240) (0.142) postsecondary education Additional controls Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Age Age Age Age Age Age Age Age Age Age Age Age Age Age Sectoral tariffs Yes Yes Yes Yes No No No Yes Yes Yes Yes No No No Region-year fixed Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes effects ISIC2 sector fixed Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes effects Observations 5,948 202,855 153,448 49,407 444,774 392,785 51,989 6,102 211,900 159,955 51,945 475,068 418,975 56,093 R-squared 0.021 0.050 0.061 0.025 0.042 0.048 0.012 -0.004 0.053 0.066 0.022 0.039 0.046 0.027 Source: World Bank. Note: See first-stage results in the supplementary appendix; FDI = foreign direct investment; ISIC2 = International Standard Industrial Classification two-digit code; IV = instrumental variables; LN = natural logarithm; MNE = multinational enterprise; Y1 = Year 1. Standard errors are clustered at the region*sector level: *** p < 0.01 ** p < 0.05 * p < 0.10 117 1 1 8    Global Investment Competitiveness Report 2019/2020 TABLE 3C.6  FDI Labor Market Effects by Broad Sector and Education Level: All Countries (Population Weighted, Total Average Effects) a. Ethiopia (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Outcome variables Wages (LN) Wages (LN) Wages (LN) Wages (LN) Wages (LN) Formal Emp. Formal Emp. Formal Emp. Formal Emp. Formal Emp. Low-skilled High-skilled Low-skilled High-skilled Low-skilled High-skilled Low-skilled High-skilled Household sample Extractives manufacturing manufacturing services services Extractives manufacturing manufacturing services services Total average effects (%)                     FDI effect: average — 37.1 23.5 — — — 4.7 0.3 — — FDI effect: no — 53.1 29.4 — — — 6.9 4.0 — — education FDI effect: primary — 51.4 0 — — — 6.3 0 — — education FDI effect: secondary — 0 0 — — — 0 0 — — education FDI effect: — 0 0 — — — 0 0 — — postsecondary education b. Vietnam (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Outcome variables Wages (LN) Wages (LN) Wages (LN) Wages (LN) Wages (LN) Formal Emp. Formal Emp. Formal Emp. Formal Emp. Formal Emp. Low-skilled High-skilled Low-skilled High-skilled Low-skilled High-skilled Low-skilled High-skilled Household sample Extractives manufacturing manufacturing services services Extractives manufacturing manufacturing services services Total average effects (%)                     FDI effect: average 0.2 11.6 14.4 -0.8 4.9 -0.8 5.1 5.5 0 0 FDI effect: no 11.2 0 0 -6.6 -27.3 9.8 13.6 0 0 0 education FDI effect: primary 0 12.6 13.0 -4.1 -18.2 -4.0 5.3 5.4 0 0 education FDI effect secondary 0 11.4 14.1 0 0 0 5.3 6.6 0 0 education FDI effect: 0 11.9 17.6 5.5 11.1 0 0 0 0 0 postsecondary education table continues next page T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    TABLE 3C.6  FDI Labor Market Effects by Broad Sector and Education Level: All Countries (Population Weighted, Total Average Effects) (continued) c. Turkey (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Outcome variables Wages (LN) Wages (LN) Wages (LN) Wages (LN) Wages (LN) Formal Emp. Formal Emp. Formal Emp. Formal Emp. Formal Emp. Low-skilled High-skilled Low-skilled High-skilled Low-skilled High-skilled Low-skilled High-skilled Household sample Extractives manufacturing manufacturing services services Extractives manufacturing manufacturing services services Total average effects (%)             FDI effect average 0 6.0 10.8 5.1 12.9 0 0.8 1.3 –0.7 1.2 FDI effect: no 0 0 0 –49.3 –116.0 0 1.1 2.0 –9.8 –23.0 education FDI effect: primary 0 7.5 14.7 0 0 0 0.9 1.8 –2.2 –4.2 education FDI effect: secondary 0 5.5 11.4 18.0 20.0 0 0.5 1.1 2.1 2.3 education FDI effect: 0 0 0 22.4 14.3 0 0 0 3.8 2.4 postsecondary education Source: World Bank. Note: Based on population-weighted, total average effects in the country, using regression coefficients from tables 3C.3 through 3C.5. FDI = foreign direct investment; LN = natural logarithm; — = not available. 119 1 2 0    Global Investment Competitiveness Report 2019/2020 TABLE 3C.7  Vertical Spillovers Effect of FDI on Labor Market Outcomes: All Countries (Second-Stage IV Results) Vietnam Turkey Ethiopia (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Wages Wages Formal Formal Wages Wages Formal Formal Wages Formal Outcome variables (LN) (LN) Emp. Emp. (LN) (LN) Emp. Emp. (LN) Emp. Household sample Manuf. Services Manuf. Services Manuf. Services Manuf. Services Manuf. Manuf. MNE output share 0.307*** 0.141*** 0.201*** –0.089 1.036*** 0.760 0.119** 0.032 1.237* 0.135 (lagged Y1) (0.116) (0.052) (0.030) (0.130) (0.378) (0.895) (0.048) (0.123) (0.679) (0.104) MNE output share 0.829 0.452** 0.415*** 0.169** 2.761 –0.194 0.409 –0.080 0.323 –0.015 (lagged Y1)*backward (0.559) (0.212) (0.066) (0.084) (2.251) (0.666) (0.309) (0.119) (3.648) (0.487) intensity (backward link) MNE output share –0.515 0.146 0.205*** –0.024 –6.793*** 0.056 –0.945*** 0.076 3.337 0.471 (lagged Y1)*forward (0.628) (0.266) (0.063) (0.087) (2.112) (0.571) (0.282) (0.088) (2.794) (0.411) intensity (forward link) Education, Education, Education, Education, Education, Education, Education, Education, Education, Education, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Gender, Additional controls Age Age Age Age Age Age Age Age Age Age Sectoral tariffs Yes No Yes No Yes No Yes No Yes Yes Region-year fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes ISIC2 sector fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Observations 265,335 393,905 489,660 885,941 202,855 444,774 211,900 475,068 10,683 11,084 R-squared 0.404 0.260 0.342 0.296 0.048 0.043 0.050 0.044 0.117 0.078 (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Total average effects (%) MNE output share (lagged Y1) 10.8 5.0 0.9 0.0 8.4 0 1.0 0 33.8 0 MNE output share (lagged Y1)*backward intensity (backward link) 0 5.0 4.2 1.7 0 0 0 0 0 0 MNE output share (lagged Y1)*forward intensity (forward link) 0 0 1.7 0.0 -35.7 0 -4.9 0 0 0 Source: World Bank. Note: See first-stage results in the supplementary appendix; FDI = foreign direct investment; ISIC2 = International Standard Industrial Classification two-digit code; IV = ­instrumental variable; LN = natural logarithm; MNE = multinational enterprise; Y1 = Year 1. Standard errors are clustered at the region*sector level: *** p < 0.01 ** p < 0.05 * p < 0.10 Notes 2. More specifically, the goals are to reduce extreme poverty in the world to less than 1. For an overview of the general literature 3 percent by 2030 and to foster income on FDI and firm dynamics, see Iršová and growth of the bottom 40 percent of the pop- Havránek (2013) and Javorcik (2015). For ulation in each country (World Bank 2015). examples from Africa, see El Badaoui, Strobl, 3. The UN estimates the gap in financing to and Walsh (2008) and Söderbom and Teal achieve the Sustainable Development Goals (2004). Other examples include Bernard, (SDGs) at US$2.5 trillion per year in devel- Robertson, and Schott (2010); Feenstra and oping countries alone (UNCTAD 2014). Hanson (1997); and Verhoogen (2008). Although governments and the public sector T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    121 will continue to play a key financing role, by MNEs’ output share) across regions and there is greater scope for the private sector to sectors experiencing a higher or lower share engage in financing many areas of the SDGs of MNE activities, not aggregate national to help close the gap (UNCTAD 2018). Out effects. Dix-Caneiro and Kovak (2015) raise of the 17 SDGs, this chapter focuses on the a similar point in their study on trade liberal- link between FDI and three of the SDGs in ization in Brazil. particular: (a) SDG 1: End poverty in all its 12. To identify aggregate effects on formal forms everywhere; (b) SDG 8: Promote inclu- employment, the regression coefficients from sive and sustainable economic growth and annex 3C, table 3C.1, are used and multi- decent work for all; and (c) SDG 10: Reduce plied by the total yearly number of workers in inequalities within and among countries. manufacturing in the country using sampling 4. “FDI in Turkey,” Investment Office, Presidency weights. The results are averaged over time. of the Republic of Turkey website: http:// 13. Similar findings on FDI for Turkey’s man- www.invest.gov.tr/en-US​ / investmentguide​ ufacturing employment were identified by /investorsguide/Pages/FDIinTurkey.aspx. Kalemli-Ozcan, Sánchez-Martín, and Thirion 5. Data for the following discussion come from (2016). World Bank’s various World Development 14. The aggregate regressions (annex 3C, table Indicator (WDI) statistical tables: http://wdi​ 3C.3) find that Turkey’s manufacturing FDI .worldbank.org/tables. is positively though insignificantly associated 6. Ideally, one would examine how FDI in a with manufacturing wages and employment. sector where an individual was previously This suggests that any negative vertical effect employed affects their subsequent labor is undone by the positive direct effects. market outcomes. However, a lack of panel 15. For more details on the World Bank’s Jobs household data precludes the analysis from Diagnostics, see Hallward-Driemeier (2015) following individuals over time. As a result, and Merotto, Weber, and Aterido (2018). the analysis cannot observe when a worker 16. Similar arguments could be made for social transitions between employment and unem- policy (including unemployment and disability ployment and through different jobs. benefits) to complement FDI and temporarily 7. An example of an upstream effect comes from accommodate labor market adjustment. Yet a car parts manufacturer increasing sales to such policies are often not within the fiscal an MNE auto manufacturer. An example of a space of developing countries, which is why downstream transaction may be professional this section focuses on labor market policies. services (such as lawyers and accountants) 17. More research is needed to better under- that are used for an MNE’s bookkeeping. stand why firms do not try to combat high 8. Data for sectoral greenfield FDI come turnover (for example, through efficiency from public announcements recorded by wages or self-enforced standards). It is the Financial Times’s fDi Markets dataset possible that firms are poorly managed or (https://www.fdimarkets.com/), while M&A constrained in unobserved ways. But it is data capture public announcements recorded also possible that high turnover at very low by Thomson Reuters. wages, where only those workers with the 9. The Gini coefficient uses the Lorenz curve to poorest outside options remain, is the firm’s define the income distribution, with a number profit-maximizing choice (Blattman and ranging from perfect equality (0) to perfect ine- Dercon 2018). Given the positive external- quality (1). This analysis includes both the Gini ities associated with workers’ knowledge coefficient and Palma ratio because the Gini transfer, this would call for government coefficient is oversensitive to the middle of the intervention. distribution, and undersensitive to the tails, 18. One notable exception comes from coun- and thus might underestimate inequality for tries that have seen high internal migration most countries today (Krozer 2015). patterns (such as in the Middle East, where 10. For the full results of the empirical analysis, some countries experienced a large inflow of see the online appendix accessible through refugees in a short period). Because it takes the report website: http://www.worldbank.org​ time for labor markets to adapt to such labor /gicreport supply shocks, in such cases, there may be 11. Note that the effects estimated in this chapter limited benefits (or possible harm) from fur- capture relative effects of FDI (as proxied for thering internal migration. 1 2 2    Global Investment Competitiveness Report 2019/2020 19. As the World Bank’s repository of household and Anna Vitali. 2017. “Tackling Youth surveys, I2D2 harmonizes nationally repre- Unemployment: Evidence from a Labour sentative household surveys—both welfare Market Experiment in Uganda.” Development and labor force surveys—from around the Economics Papers No. 64, Suntory and world, presenting data using the same varia- Toyota International Centres for Economics bles and coding in each country and survey. and Related Disciplines, London School of Economics. Almond, Phil, and Anthony Ferner, eds. 2006. American Multinationals in Europe. Oxford: References Oxford University Press. Abebe, Girum, Stefano Caria, Marcel Fafchamps, Amann, Edmund, and Swati Virmani. 2014. Paolo Falco, Simon Franklin, and Simon Quinn. “Foreign Direct Investment and Reverse 2016. “Anonymity or Distance? Experimental Technology Spillovers: The Effect on Total Evidence on Obstacles to Youth Employment Factor Productivity.” OECD Journal: Opportunities.” Unpublished research study, Economic Studies 2014 (1): 129–53. Stanford University, Stanford, CA. Amir, Saman, Aphichoke Kotikula, Rohini P. Abebe, Girum, Stefano Caria, Marcel Fafchamps, Pande, Laurent Loic Yves Bossavie, and Paolo Falco, Simon Franklin, and Simon Quinn. Upasana Khadka. 2018. “Female Labor 2018. “Anonymity or Distance? Job Search and Force Participation in Pakistan: What Do We Labour Market Exclusion in a Growing African Know?” Note for the Pakistan Gender and City.” Discussion Paper 13136, Centre for Social Inclusion Platform and Social Protection Economic Policy Research, London. and Jobs Teams, World Bank, Washington, DC. Abebe, Girum, Stefano Caria, Marcel Fafchamps, Amoranto, Glenita, Douglas H. Brooks, and Paolo Falco, Simon Franklin, Simon Quinn, Natalie Chun. 2010. “Services Liberalization and Forhad Shilpi. 2017. “Job Fairs: Matching and Wage Inequality in the Philippines.” Firms and Workers in a Field Experiment in Working Paper 239, Asian Development Bank, Ethiopia.” Policy Research Working Paper Manila. 8092, World Bank, Washington, DC. Antràs, Pol, and Stephen R. Yeaple. 2014. Acemoglu, Daron, and David Autor. 2011. “Skills, “Multinational Firms and the Structure Tasks and Technologies: Implications for of International Trade.” In Handbook of Employment and Earnings.” In Handbook of International Economics, Volume 4, edited by Labor Economics, Volume 4B, edited by Orley Gita Gopinath, Elhanan Helpman, and Kenneth Ashenfelter and David Card, 1043–171. San Rogoff, 55–130. Oxford and Amsterdam: Diego, CA, and Amsterdam: North Holland. North-Holland Acemoglu, Daron, and Pascual Restrepo. 2017. Arnold, Jens Matthias, Beata Javorcik, Molly “Robots and Jobs: Evidence from US Labor Lipscomb, and Aaditya Mattoo. 2016. “Services Markets.” Working Paper 23285, National Reform and Manufacturing Performance: Bureau of Economic Research, Cambridge, Evidence from India.” The Economic Journal MA. 126 (590): 1–39. Adão, Rodrigo, Costas Arkolakis, and Federico Arnold, Jens M., Beata S. Javorcik, and Aaditya Esposito. 2019. “Spatial Linkages, Global Mattoo. 2011. “Does Services Liberalization Shocks, and Local Labor Markets: Theory and Benefit Manufacturing Firms? Evidence from Evidence.” Working Paper No. 2019-10, Becker the Czech Republic.” Journal of International Friedman Institute for Economics, University of Economics 85 (1): 136–46. Chicago. Atkin, David, Benjamin Faber, and Marco Alfaro, Laura, and Jasmina Chauvin. Forthcoming. Gonzalez-Navarro. 2018. “Retail Globalization “Foreign Direct Investment, Finance, and and Household Welfare: Evidence from Economic Development.” In Encyclopedia of Mexico.” Journal of Political Economy 126 International Economics and Global Trade, (1): 1–73. Vol. 3: Foreign Direct Investment and the Autor, David H., David Dorn, and Gordon H. Multinational Enterprise, edited by Mariana Hanson. 2013. “The China Syndrome: Local Spatareanu, 1–32. Singapore: World Scientific. Labor Market Effects of Import Competition in Alfonsi, Livia, Oriana Bandiera, Vittorio Bassi, the United States.” American Economic Review Robin Burgess, Imran Rasul, Munshi Sulaiman, 103 (6): 2121–68. T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    123 Bandick, Roger, and Patrik Karpaty. 2011. The Roles of Gender, Skill and Public Sector “Employment Effects of Foreign Acquisition.” Employment.” Working Paper 25313, National International Review of Economics & Finance Bureau of Economic Research, Cambridge, 20 (2): 211–24. MA. Barrett, Paul M., and Dorothée Baumann-Pauly. Caria, Stefano, and Theres Lessing. 2019. 2019. “Made in Ethiopia: Challenges in the “Filling the Gap: How Information Can Help Garment Industry’s New Frontier.” Report, Jobseekers.” Growth Brief, International NYU Stern Center for Business and Human Growth Centre, London. Rights, New York. Cazes, Sandrine, Alexander Hijzen, and Anne Saint- Bassi, Vittorio, and Aisha Nansamba. 2018. Martin. 2015. “Measuring and Assessing Job “Screening and Signaling Non-Cognitive Quality: The OECD Job Quality Framework.” Skills: Experimental Evidence from Uganda.” Social, Employment and Migration Working Research Paper No. 19-08, University College Paper 174, Organisation of Economic London. Co-operation and Development, Paris. Beam, Emily A. 2016. “Do Job Fairs Matter? Chen, Shaohua, and Martin Ravallion. 2007. Experimental Evidence on the Impact of Job- “Absolute Poverty Measures for the Fair Attendance.” Journal of Development Developing World, 1981–2004.” Proceedings Economics 120: 32–40. of the National Academy of Sciences 104 (43): Bernard, Andrew B., and J. Bradford Jensen. 16757–62. 2007. “Firm Structure, Multinationals, and Chen, V. Zitian, Jing Li, and Daniel M. Shapiro. Manufacturing Plant Deaths.” Review of 2012. “International Reverse Spillover Effects Economics and Statistics 89 (2): 193–204. on Parent Firms: Evidences from Emerging- Bernard, Andrew B., Raymond Robertson, and Market MNES in Developed Markets.” Peter K. Schott. 2010. “Is Mexico a Lumpy European Management Journal 30 (3): 204–18. Country?” Review of International Economics Cornia, Giovanni Andrea. 2016. “An Econometric 18 (5): 937–50. Analysis of the Bifurcation of Within-Country Bernard, Andrew B., and Fredrik Sjöholm. 2003. Inequality Trends in Sub-Saharan Africa, 1990– “Foreign Owners and Plant Survival.” Working 2011.” Africa Report 267781, United Nations Paper 10039, National Bureau of Economic Development Programme, New York. Research, Cambridge, MA. Crépon, Bruno, and Gerard J. van den Berg. Blattman, Christopher, and Stefan Dercon. 2018. 2016. “Active Labor Market Policies.” Annual “The Impacts of Industrial and Entrepreneurial Review of Economics 8 (1): 521–46. Work on Income and Health: Experimental Cruz, Marcio, Gaurav Nayyar, Gerhard Toews, Evidence from Ethiopia.” American Economic and Pierre-Louis Vézina. 2018. “FDI and Journal: Applied Economics 10 (3): 1–38. the Skill Premium: Evidence from Emerging Braunstein, Elissa. 2006. Foreign Direct Economies.” Policy Research Working Paper Investment, Development and Gender Equity: 8613, World Bank, Washington, DC. A Review of Research and Policy. Geneva: Davies, Ronald B., and Rodolphe Desbordes. United Nations Research Institute for Social 2015. “Greenfield FDI and Skill Upgrading: Development. A Polarized Issue.” Canadian Journal of Braunstein, Elissa, and Mark Brenner. 2007. Economics 48 (1): 207–44. “Foreign Direct Investment and Gendered Dinardo, John, and Thomas Lemieux. 1997. Wages in Urban China.” Feminist Economics “Diverging Male Wage Inequality in the United 13 (3–4): 231–37. States and Canada, 1981–1988: Do Institutions Bryan, Gharad, Shyamal Chowdhury, and Ahmed Explain the Difference?” ILR Review 50 (4): Mushfiq Mobarak. 2014. “Underinvestment in 629–51. a Profitable Technology: The Case of Seasonal Dinga, Marián, and Daniel Münich. 2010. “The Migration in Bangladesh.” Econometrica Impact of Territorially Concentrated FDI on 82 (5): 1671–748. Local Labor Markets: Evidence from the Czech Card, David, Thomas Lemieux, and W. Craig Republic.” Labour Economics 17 (2): 354–67. Riddell. 2004. “Unions and Wage Inequality.” Dix-Carneiro, Rafael, and Brian K. Kovak. 2015. Journal of Labor Research 25 (4): 519–62. “Trade Liberalization and the Skill Premium: Card, David, Thomas Lemieux, and W. Craig A Local Labor Markets Approach.” American Riddell. 2018. “Unions and Wage Inequality: Economic Review 105 (5): 551–57. 1 2 4    Global Investment Competitiveness Report 2019/2020 Duggan, Victor, Sjamsu Rahardja, and Gonzalo United States.” In Seeking a Premier Economy: Varela. 2013. “Service Sector Reform and The Economic Effects of British Economic Manufacturing Productivity: Evidence from Reforms, 1980–2000, edited by David Card, Indonesia.” Policy Research Working Paper Richard Blundell, and Richard B. Freeman, 6349, World Bank, Washington, DC. 275–312. Chicago: University of Chicago Press. EIC (Ethiopia Investment Commission). 2017. Hale, Galina, and Mingzhi Xu. 2016. “FDI Effects “Ethiopian Investment Report 2017.” EIC, on the Labor Market of Host Countries.” Addis Ababa. http://www.investethiopia.gov.et​ Working Paper 2016-25, Federal Reserve Bank /images/pdf/EIC%202017%20Report%20.pdf. of San Francisco. El Badaoui, Eliane, Eric Strobl, and Frank Walsh. Hallward-Driemeier, Mary C. 2015. “Jobs 2008. “Is There an Informal Employment Diagnostics Guidance: Why, What, and How?” Wage Penalty? Evidence from South Africa.” Working Paper No. 96328, World Bank, Economic Development and Cultural Change Washington, DC. 56 (3): 683–710. Hallward-Driemeier, Mary, and Gaurav Nayyar. Eschenbach, Felix, and Bernard Hoekman. 2006. 2017. Trouble in the Making? The Future of “Services Policy Reform and Economic Growth Manufacturing-Led Development. Washington, in Transition Economies.” Review of World DC: World Bank. Economics 142 (4): 746–64. Harrison, Ann, and Jason Scorse. 2010. Fang, Sheng, Heba M. M. Shamseldin, and “Multinationals and Anti-Sweatshop Activism.” L. Colin Xu. 2019. “Foreign Direct Investment American Economic Review 100 (1): 247–73. and Female Entrepreneurship.” Policy Research Head, Keith, and John Ries. 2002. “Offshore Working Paper 9083, World Bank, Washington, Production and Skill Upgrading by Japanese DC. Manufacturing Firms.” Journal of International Feenstra, Robert C., and Gordon H. Hanson. Economics 58 (1): 81–105. 1997. “Foreign Direct Investment and Heath, Rachel, and A. Mushfiq Mobarak. 2015. Relative Wages: Evidence from Mexico’s “Manufacturing Growth and the Lives of Maquiladoras.” Journal of International Bangladeshi Women.” Journal of Development Economics 42 (3–4): 371–93. Economics 115: 1–15. Ferreira, Francisco. 2010. “Distributions in Hijzen, Alexander, Pedro S. Martins, Thorsten Motion: Economic Growth, Inequality, and Schank, and Richard Upward. 2013. Poverty Dynamics.” Policy Research Working “Foreign-Owned Firms around the World: Paper 5424, World Bank, Washington, DC. A Comparative Analysis of Wages and Figini, Paolo, and Holger Görg. 2011. “Does Employment at the Micro-Level.” European Foreign Direct Investment Affect Wage Economic Review 60: 170–88. Inequality? An Empirical Investigation.” The Inchauste, Gabriela, João Pedro Azevedo, World Economy 34 (9): 1455–75. B. Essama-Nssah, Sergio Olivieri, Trang Van Franklin, Simon. 2015. “Location, Search Costs Nguyen, Jaime Saavedra-Chanduvi, and and Youth Unemployment: A Randomized Trial Hernan Winkler. 2014. Understanding Changes of Transport Subsidies in Ethiopia.” Working in Poverty. Directions in Development Series. Paper WPS/2015-11, Centre for the Study of Washington, DC: World Bank. African Economies, University of Oxford. Iršová, Zuzana, and Tomáš Havránek. 2013. Glick, Peter, and François Roubaud. 2006. “Export “Determinants of Horizontal Spillovers from Processing Zone Expansion in Madagascar: FDI: Evidence from a Large Meta-Analysis.” What Are the Labour Market and Gender World Development 42: 1–15. Impacts?” Journal of African Economies ISPAT (Investment Office of the Presidency of the 15 (4): 722–56. Republic of Turkey). 2018. “FDI in Turkey.” Goldberg, Pinelopi Koujianou, and Nina Pavcnik. Online statistical report, ISPAT website: http:// 2007. “Distributional Effects of Globalization www.invest.gov.tr/en-US/investmentguide​ in Development Countries.” Working Paper /investorsguide/Pages/FDIinTurkey.aspx. 12885, National Bureau of Economic Research, Javorcik, Beata Smarzynska. 2004. “Does Foreign Cambridge, MA. Direct Investment Increase the Productivity Gosling, Amanda, and Thomas Lemieux. 2004. of Domestic Firms? In Search of Spillovers “Labor Market Reforms and Changes in Wage through Backward Linkages.” American Inequality in the United Kingdom and the Economic Review 94 (3): 605–27. T h e D i s t ri b u t i o n a l E f f e c t s o f F D I : E v id e n c e f r o m E t h i o p i a , Vi e t n a m , a n d T u r k e y    125 Javorcik, Beata S. 2015. “Does FDI Bring Good Lipsey, Robert E. 2003. “Foreign Direct Investment, Jobs to Host Countries?” World Bank Research Growth, and Competitiveness in Developing Observer 30 (1): 74–94. Countries.” In The Global Competitiveness Jensen, Robert. 2012. “Do Labor Market Report 2002/2003, edited by P. K. Cornelius. Opportunities Affect Young Women’s Work New York: Oxford University Press. and Family Decisions? Experimental Evidence Luo, Rui. 2017. “Skill Premium and Technological from India.” Quarterly Journal of Economics Change in the Very Long Run: 1300–1914.” 127 (2): 753–92. Discussion Papers in Economics 17/09, Jude, Cristina, and Monica Ioana Pop Silaghi. University of Leicester, U.K. 2015. “Employment Effects of Foreign Direct Maloney, William F., and Carlos A. Molina. Investment: New Evidence from Central and 2016. “Are Automation and Trade Polarizing Eastern European Countries.” International Developing Country Labor Markets, Too?” Economics 145: 32–49. Policy Research Working Paper 7922, World Kalemli-Ozcan, Sebnem, Miguel Eduardo Sánchez- Bank, Washington, DC. Martín, and Gilles Thirion. 2016. “A Difficult Mattoo, Aaditya, Randeep Rathindran, and Relationship: Declining (but Productive) FDI Arvind Subramanian. 2006. “Measuring Inflows in Turkey.” Policy Research Working Services Trade Liberalization and Its Impact on Paper 7919, World Bank, Washington, DC. Economic Growth: An Illustration.” Journal of Karlsson, Sune, Nannan Lundin, Fredrik Sjöholm, Economic Integration 21 (1): 64–98. and Ping He. 2009. “Foreign Firms and McKenzie, David J. 2017. “How Effective Are Chinese Employment.” The World Economy Active Labor Market Policies in Developing 32 (1): 178–201. Countries? A Critical Review of Recent Kianto, Aino, Pia Hurmelinna-Laukkanen, and Evidence.” Policy Research Working Paper Paavo Ritala. 2010. “Intellectual Capital in 8011, World Bank, Washington, DC. Service- and Product-Oriented Companies.” McLaren, John, and Myunghwan Yoo. 2016. “FDI Journal of Intellectual Capital 11 (3): 305–25. and Inequality in Vietnam: An Approach with Krozer, Alice. 2015. “The Inequality We Want: Census Data.” Working Paper 22930, National How Much Is Too Much?” Working Paper Bureau of Economic Research, Cambridge, MA. 2015/015, UNU-WIDER (United Nations Merotto, Dino, Michael Weber, and Reyes University World Institute for Development Aterido. 2018. “Pathways to Better Jobs in IDA Economics Research), Helsinki. Countries: Findings from Jobs Diagnostics.” Leamer, Edward E. 1998. “In Search of Stolper- Jobs Series Issue No. 14, World Bank, Samuelson Effects on U.S. Wages.” In Exports, Washington, DC. Imports and the American Worker, edited by Michaels, Guy, Ashwini Natraj, and John Van Susan M. Collins, 141–214. Washington, DC: Reenen. 2014. “Has ICT Polarized Skill Brookings Institution Press. Demand? Evidence from Eleven Countries over Lee, Jong-Wha, and Dainn Wie. 2015. Twenty-Five Years.” Review of Economics and “Technological Change, Skill Demand, and Statistics 96 (1): 60–77. Wage Inequality: Evidence from Indonesia.” Moran, Theodore H. 2011 . Foreign Direct World Development 67: 238–50. Investment and Development: Launching Lemieux, Thomas. 1993. “Unions and Wage a Second Generation of Policy Research: Inequality in Canada and the United States.” Avoiding the Mistakes of the First, Reevaluating In Small Differences that Matter: Labor Policies for Developed and Developing Markets and Income Maintenance in Canada Countries. Washington, DC: Peterson Institute and the United States, edited by David Card for International Economics. and Richard B. Freeman, 69–108. Chicago: Munshi, Kaivan, and Mark Rosenzweig. 2016. University of Chicago Press. “Networks and Misallocation: Insurance, Li, Shantong, Yan Wang, and Fan Zhai. 2003. Migration, and the Rural-Urban Wage Gap.” “Impact of Service Sector Liberalization on American Economic Review 106 (1): 46–98. Employment and Output: A CGE Analysis.” Newman, Carol, John Page, John Rand, Abebe Paper for the Joint Senior Policy Seminar, Shimeles, Måns Söderbom, and Finn Tarp. “China’s Service Sector in the Post-WTO Era,” 2016. Made in Africa: Learning to Compete National Development and Reform Commission in Industry. Washington, DC: Brookings and World Bank, Beijing, September 22–24. Institution Press. 1 2 6    Global Investment Competitiveness Report 2019/2020 Nunnenkamp, Peter, Rainer Schweickert, and Te Velde, Dirk Willem, and Theodora Xenogiani. Manfred Wiebelt. 2007. “Distributional 2007. “Foreign Direct Investment and Effects of FDI: How the Interaction of FDI International Skill Inequality.” Oxford and Economic Policy Affects Poor Households Development Studies 35 (1): 83–104. in Bolivia.” Development Policy Review UNCTAD (United Nations Conference on Trade 25 (4): 429–50. and Development). 2014. World Investment Ouedraogo, Rasmané, and Elodie Marlet. 2018. Report 2014. Investing in the SDGs: An “Foreign Direct Investment and Women Action Plan. New York and Geneva: United Empowerment: New Evidence on Developing Nations. Countries.” Working Paper No. 18/25, UNCTAD (United Nations Conference on International Monetary Fund, Washington, DC. Trade and Development). 2018. “Promoting Pavcnik, Nina. 2017. “The Impact of Trade on Investment in the Sustainable Development Inequality in Developing Countries.” Working Goals.” Investment Advisory Series, Series A, Paper 23878, National Bureau of Economic No. 8, United Nations, Geneva. Research, Cambridge, MA. UNCTAD (United Nations Conference on Trade Peluffo, Adriana. 2015. “Foreign Direct and Development). 2019. World Investment Investment, Productivity, Demand for Skilled Report 2019: Special Economic Zones . Labour and Wage Inequality: An Analysis of New York and Geneva: United Nations. Uruguay.” World Economy 38 (6): 962–83. doi:10.18356/8a8d05f9-en. Perea, Jose Ramon, and Matthew Stephenson. 2018. Van der Marel, Erik. 2012. “Trade in Services “Outward FDI from Developing Countries.” and TFP: The Role of Regulation.” The World In Global Investment Competitiveness Report Economy 35 (11): 1387–429. 2017/2018, 101–34. Washington, DC: World Ver Beek, Kurt Alan. 2001. “Maquiladoras: Bank Group. Exploitation or Emancipation? An Overview Pierce, Justin R., and Peter K. Schott. 2016. “The of the Situation of Maquiladora Workers Surprisingly Swift Decline of US Manufacturing in Honduras.” World Development 29 (9): Employment.” American Economic Review 1553–67. 106 (7): 1632–62. Verhoogen, Eric A. 2008. “Trade, Quality Reyes, Jose Daniel. 2017. “FDI Spillovers and Upgrading, and Wage Inequality in the Mexican High-Growth Firms in Developing Countries.” Manufacturing Sector.” Quarterly Journal of Policy Research Working Paper 8243, World Economics 123 (2): 489–530. Bank, Washington, DC. VFIA (Vietnam Foreign Investment Agency). Seguino, Stephanie, and Caren Grown. 2018. “Investments in Vietnam 2018.” [In 2006. “Gender Equity and Globalization: Vietnamese.] Online statistical report, VFIA Macroeconomic Policy for Developing website: http://fia.mpi.gov.vn/tinbai/6110/Tinh​ Cou n tr ie s .” J o u r n a l o f I n ter national -hinh-thu-hut-Dau-tu-nuoc-ngoai-nam-2018. Development 18 (8): 1081–104. Waldkirch, Andreas, Peter Nunnenkamp, and Siegmann, Karin Astrid. 2007. “Globalisation, Jose Eduardo Alatorre Bremont. 2009. Gender, and Equity: Effects of Foreign Direct “Employment Effects of FDI in Mexico’s Non- Investment on Labour Markets in Rural Maquiladora Manufacturing.” Journal of Indonesia.” Proceedings of the German Development Studies 45 (7): 1165–83. Development Economics Conference, Göttingen World Bank. 2013. “Shared Prosperity Links 2007, Verein für Socialpolitik, Research to Growth, Inequality, and Inequality of Committee Development Economics, No. 30. Opportunity.” Policy Research Working Paper Söderbom, Måns, and Francis Teal. 2004. “Size 6649, World Bank, Washington, DC. and Efficiency in African Manufacturing Firms: World Bank. 2015. A Measured Approach Evidence from Firm-Level Panel Data.” Journal to Ending Poverty and Boosting Shared of Development Economics 73 (1): 369–94. Prosperity: Concepts, Data, and the Twin Stiglitz, Joseph. 2012. The Price of Inequality. Goals. Washington, DC: World Bank. New York: W.W. Norton. World Bank Group and OECD (Organisation Tang, Heiwai, and Yifan Zhang. 2017. “Do for Economic Co-operation and Development). Multinationals Transfer Culture? Evidence on 2017. A Step Ahead: Competition Policy for Female Employment in China.” Working Paper Shared Prosperity and Inclusive Growth . No. 6295, CESifo Group, Munich. Washington, DC: World Bank and OECD. 4 Regulatory Risk and FDI Sarah Hebous, Priyanka Kher, and Trang Thu Tran Key Findings • This report presents a new global database on the content of legal instruments and a quan- titative measure of regulatory risk, focused on transparency, investment protection, and recourse. It evaluates (a) transparency and predictability in the content as well as in the process of making laws and regulations that apply to investors; (b) legal protection of investors against arbitrary and nontransparent government interference; and (c) investor access to effective mechanisms for recourse, including grievance management and dispute settlement. • Evidence from this database, which analyzes laws affecting investment, shows that inves- tor confidence and FDI flows increase with regulatory transparency, investment protec- tion, and effective recourse. The constructed measure of regulatory risk is predictive of investment risk premium. Lower regulatory risk is associated with higher investment, in regressions using a global dataset of over 14,000 parent companies investing in nearly 28,000 FDI greenfield and expansion projects across 168 host countries. • The effect of regulatory risk on FDI is sizable and comparable in magnitude to the invest- ment-enhancing effects of trade openness in the same regression models. In fact, in some of the models, the effect of regulatory risk on FDI exceeds that of trade openness, showing that a 1 point reduction in regulatory risk increases the likelihood of an investor entering or expanding in a host country by 0.5–2 percentage points. In contrast, a 1 percentage point increase in the host country’s trade-GDP ratio is associated with a 0.3–0.6 percent- age point increase in an investor’s likelihood to enter or expand. • Business survey results confirm the importance of transparent, predictable regulatory environments to investors. Investors rank countries’ legal and regulatory environments as one of the top three factors shaping investment decisions, along with political and macroeconomic stability. Exposure to regulatory risks in host countries triggers exist- ing investors to consider withdrawing investments or canceling planned investment. Moreover, results of the 2019 Global Investment Competitiveness (GIC) Survey show that complex administrative procedures are a further obstacle for nearly two-thirds of investors. • To enhance investor confidence and reduce regulatory risk, governments need to remain committed to creating open and predictable environments for FDI. Given that sources of policy uncertainty that erode investor confidence are both international and domestic, solutions at both levels are needed. Governments can reduce risks for investors by improv- ing transparency and predictability in policy making and implementation, reducing room for bureaucratic discretion, aligning domestic rules with international legal frameworks, and facilitating access to a wide range of dispute settlement mechanisms, including mecha- nisms to prevent disputes by early detection and resolution of investor grievances. Introduction Competitiveness (GIC) Surveys in 2017 (World Bank 2018) and 2019 (see chapter 1) Investors in a country rely on its legal and suggest that two-thirds of investors consider regulatory framework to recognize their policy uncertainty as “important” or “criti- property rights and enforce those rights in a cally important” to their investment deci- predictable and efficient manner. Economic sions. Further, they confirm that political theory suggests that when investors incur stability and a country’s legal and regulatory fixed and irreversible setup costs, uncertainty framework are the two most important fac- about the local conditions—especially policy tors for firms’ decision to invest in develop- uncertainty—will have a dampening effect ing countries. As FDI to developing countries that reduces investment response to new has been slowing, competition between investment opportunities (Bernanke 1983; developing countries for investments has Bloom 2009; Dixit 1989). only intensified. Among studies that look at developed countries, Baker, Bloom, and Davis (2016) A Regulatory Framework to Reduce Risk construct a news-based index of United and Boost Investor Confidence States economic policy uncertainty and ­ document that the most frequent references Even though the shift toward more protec- to perceived policy uncertainty are related to tionist policies has so far concentrated in macroeconomic and regulatory policy. Their large economies, it is vital for all govern- empirical analysis indicates that aggregate ments to improve the regulatory framework investment and output decline as uncertainty to reduce risk and help restore investor con- in the United States surges. Among the fidence for several reasons. First, country ­ studies that specifically look at developing risk is difficult to manage from the firm countries, Alfaro, Kalemli-Ozcan, and standpoint. In a survey of chief financial offi- Volosovych (2008) find that low institu- cers across the globe, only 15 percent of tional quality, a source of uncertainty, is a respondents state that they use political risk major deterrent for foreign capital flows into insurance,1 and nearly half avoid investing in low-income countries. A similar effect has a risky country altogether (Giambona, been documented for components of institu- Graham, and Harvey 2017). tional quality including corruption (Wei Second, the risk profile in each country 2000); government transparency (Gelos and will influence the types of firms that enter its Wei 2005); predictability of laws, regula- markets because multinational enterprises tions, and policies (Daude and Stein 2007); (MNEs) vary systematically in their vulnera- and property rights protection (Papaioannou bility to and ability to manage different risks. 2009). A joint survey by the Multilateral Insurance With the global rise in protectionism in Guarantee Agency and the Economist response to nationalist sentiments and eco- Intelligence Unit (MIGA 2013) finds that nomic security considerations, policy uncer- firms that outperformed their competitors tainty has once again become a key concern paid significantly more attention to assessing for investors. Cross-country evidence docu- and taking measures to manage political risk. ments the highest share of newly introduced Better-performing companies, with better restrictive measures against foreign direct capabilities to assess political risk, also expe- investment (FDI) in high-income countries in rienced fewer cases of expropriation, default the past two years (UNCTAD 2019d). of government payments, cancellation of Further uncertainty over the development of import/export licenses, or restrictions on cur- the i­nternational policy frameworks for rency transfer than other firms. More trade and investment is likely to erode inves- recently, using a new measure of firm-level tor ­c onfidence. The Global Investment political risk, Hassan et al. (2019) find that a 129 1 3 0    Global Investment Competitiveness Report 2019/2020 large share of the variation in political risk to arbitrary government conduct. Specifically, appears to play out at the firm level and is the new regulatory risk measure examines significantly associated with investment and (a) whether there is transparency in both the hiring decisions. Moreover, they find that dis- content and process of making laws and reg- persion in firm-level risk often increases with ulations that apply to investors; (b) the aggregate risk level. This result suggests that extent of legal protection provided to inves- reducing country-level risk can improve mac- tors against arbitrary, unpredictable, and roeconomic outcomes through an additional nontransparent government interference; channel—by lowering the distortion of and (c) whether investors have access to resource allocation across firms as they effective mechanisms for recourse. The regu- respond to varying levels of risk. latory risk measure developed in this study, Finally, a country’s attractiveness for therefore, serves as a tool to help countries FDI can suffer in the long term from a bad identify specific areas for further track record of government conduct. From a improvement. ­ signaling perspective, it is important to reduce This study draws on existing indicators not only actual risk but also perceived levels and collects new data, including on the con- of risk. tent of legal instruments, that cover several regulatory areas: investment laws and trea- ties, public procurement, property registra- A New Regulatory Risk Measure tion, and other cross-cutting regulatory An important response by countries to the governance measures. Importantly, and in increasing policy uncertainty and associated contrast to other risk ratings, the constructed risks is to create supportive, predictable measure of regulatory risk does not rely on legal and regulatory regimes to de-risk inputs based on perception. investments. The question is: How is the The underlying data are organized into role of government action reflected in coun- three dimensions: transparency, protection, try risk ratings? Relating government actions and recourse. For each of these dimensions, to the measurement of risks is key to inform- two aggregate scores are calculated from indi- ing policy makers. Attracting FDI requires vidual data points using a simple average and improving investors’ perception about a weighted average, where weights are derived uncertainty in the economy, which in turn from a principal component analysis (PCA). requires an effective signal and government Because of limited data overlap, two different actions to boost confidence in its overall pol- sets of the overall regulatory risk measure icy. Yet country risk ratings often include a are developed: (a) a panel version including wide range of measures, from quantitative data between 2014 and 2017, and (b) a cross-­ macroeconomic indicators to qualitative sectional version that includes richer data but expert perception of political and invest- is available only for 2017. ment risk. These qualitative measures are This study finds that these constructed often beyond the influence of or have no measures carry meaningful signals of risk for direct relationships with government investors. The measures show that countries actions. differ substantially in regulatory risk. While To fill in this gap, this study develops a many countries’ risk levels have stayed rela- quantitative measure of regulatory risk—a tively stable, some have experienced signifi- subset of political risk—that is linked directly cant changes in their risk levels over time. to specific legal and regulatory provisions. Countries with higher regulatory risk in this Regulatory risk, as defined here, is related to framework tend to have a higher expropria- select features of countries’ regulatory frame- tion risk premium (that is, higher prices to work that can reduce risks for investors and ensure against expropriation risk) and tend to limit the potential for unexpected losses due be considered as riskier in other ratings such R e g u l a t o r y R i s k a n d F D I    131 as the International Country Risk Guide’s significant changes in line with changing gov- (ICRG) investment risk profile indicators. ernment policy despite the limited number of Notably, higher regulatory risk appears to regulatory areas that can be covered because be associated with a more restrictive FDI reg- of the lack of cross-country data. Across ulatory framework. Although restrictiveness countries, performance on the three dimen- per se does not necessarily constitute risks for sions of regulatory risk measured—­ investors, it can increase uncertainty where transparency, protection, and recourse—is rules are imprecise or unclear, leaving room often correlated. Yet statistically, they all for discretion in implementation. This result appear to have some predictive power lends empirical support to the concern about concerning investor behavior, suggesting that ­ growing protectionism and further highlights it is important for countries to pay attention the need for countries to manage their regula- to all three aspects in their regulatory frame- tory frameworks to restore investor work. The constructed risk measure and the confidence. underlying data can provide a starting point to help guide further research, diagnostics, and more specific policy recommendations to How Regulatory Risk Affects FDI reduce risk for investors. Importantly, regulatory risk matters for investments. This study finds that lower risk is associated with higher FDI inflows. Consistent with this result, estimations from Analytical Framework a model of investor location choices sug- Measuring risk and uncertainty is inherently gests that regulatory risk can deter the deci- challenging. Research has often relied on sions of MNEs to enter or expand in a host some measures of volatility or dispersion as country. This effect is of meaningful magni- proxies of uncertainty, which might or might tude: if the median country improves its not be tightly linked to true underlying eco- performance to the level of a top 25th per- nomic uncertainty (Jurado, Ludvigson, and centile performer, investors will be 5.5–22 Ng 2015). A growing literature attempts to percentage points more likely to locate in quantify policy uncertainty and political risk the country. To put this result in further per- based on the frequency that “keywords” spective, in the same model, the effect of the related to “risk” or “uncertainty” appear in regulatory risk measure on investment deci- news publications or corporate disclosures sion making is comparable in magnitude to (Baker, Bloom, and Davis 2016; Hassan trade openness: a 1 standard deviation et al. 2019). These proxies are useful to track increase in trade openness is associated with the movements of risk and to study their a 28 percentage point increase in the likeli- consequences on market participants. hood of investor entry, on average. In com- However, they are not intended to provide a parison, a 1 standard deviation decrease in direct link to specific government actions regulatory risk is associated with a 9 per- that cause such movements in risk. In addi- centage point increase in likelihood of inves- tion, because of the nature of the data and tor entry. textual analysis required, these proxies are To summarize, these results suggest that not easily expanded or comparable across the legal provisions and other regulatory fea- countries. tures selected and scored in this study can In contrast, the risk measure in this study provide a meaningful framework for govern- aims to capture risk as implied by policy and ment actions and reforms to reduce regula- regulatory choices. It bears some similarities tory risk. Country case studies suggest that to other popular risk ratings, such as the the regulatory risk measure can capture ICRG’s political risk rating; the Economist 1 3 2    Global Investment Competitiveness Report 2019/2020 Intelligence Unit’s (EIU) legal and regulatory • The ICRG political risk rating is a com- policy risk rating; and the country risk clas- posite risk rating of 12 components, sification of the Organisation of Economic including government stability, invest- Co-operation and Development (OECD), ment profile, and conflict. 2 It ranges from which allows for a cross-country risk com- 0 (low confidence, high risk) to 100 (high parison (table 4.1). These risk ratings confidence, low risk). The scores are ­ typically include a mix of quantitative mac- determined by political risk experts and roeconomic indicators, business environ- editorial staff. ment indicators, and qualitative expert • The OECD country risk classification assessment based on political events. The forms the basis for minimum risk pre- model to arrive at the final rating is often mium categories for many official export proprietary. In summary, they assess risk as credit agencies. It is a measure of transfer follows: and convertibility risk and cases of force majeure, and it relies on both quantitative • The EIU legal and regulatory policy risk inputs and expert opinions.3 rating is a component of its operational risk model, ranging from 0 (low risk) to Other characteristics of these risk ratings 100 (high risk). It is the (rescaled) sim- are summarized in table 4.1. Among other ple average of various subindicators, factors, these commercially available risk which are scored on a 0–4 scale by the ratings often inform investors’ opinions on EIU’s analysts working in regional countries’ investment attractiveness. Yet teams using open and closed sources despite the importance of risk rating for (EIU 2017). investment decisions, the methodology often TABLE 4.1  Several Popular Cross-Country Risk Ratings Do Not Clearly Link Risk Levels to Specific Government Actions and Rely on Perception-Based Inputs EIU legal and regulatory Characteristic ICRG political risk rating policy risk rating OECD country risk classification Objective To assess the political stability of To assess the risk that the legal To provide country risk classification, encompassing the country system will fail to safeguard transfer and convertibility risk and cases of force investment majeure (such as war, expropriation, revolution, civil disturbance, floods, and earthquakes) Type of analysis Subjective (perception-based) Subjective (perception-based) Objective data and subjective (perception-based) analysis only analysis only analysis Underlying data Twelve components, including Ten questions, including “How Objective data: macroeconomic indicators on the government stability, vulnerable is the legal process to country’s financial and economic situation investment profile, corruption, interference or distortion to serve Subjective analysis: qualitative assessment to and external and internal particular interests?” and “What integrate information not fully taken into account by conflict is the risk that business financial quantitative data Scored on a scale from 4 (very statements are inconsistent or Answered by country risk experts from export credit low risk) to 1 (very high risk) misleading?” agencies by political risk experts and Answered on a scale from 0 (very editorial staff little risk) to 4 (very high risk) by expert opinion of analysts in the regions Aggregation Simple average of individual Simple average of individual Model-based: two-step procedure, including a methodology components components quantitative model with possible adjustments through qualitative assessment Source: World Bank. Note: EIU = Economist Intelligence Unit; ICRG = International Country Risk Guide; OECD = Organisation for Economic Co-operation and Development. R e g u l a t o r y R i s k a n d F D I    133 does not clearly link the level of risk to spe- • What is the extent of legal protection cific government actions. provided to investors against arbitrary This study’s measure of regulatory risk and unpredictable, or nontransparent, contrasts with other risk ratings by measuring government actions? regulatory risk without relying on inputs • Do investors have access to effective based on perception. It is also narrower in mechanisms for recourse in case of griev- scope and excludes other sources of risks, ances or disputes? such as macroeconomic uncertainty or politi- These questions allow for a systematic way cal violence. It captures sources of regulatory to think about dimensions of the regulatory risk solely through quantitative indicators framework that countries can influence to that are linked directly to specific, actionable improve perceived and actual levels of risk. legal and regulatory provisions. It is therefore They also serve as a guide to organize the intended to help identify concrete weaknesses, data sources used (box 4.1). give specific policy recommendations, and By improving transparency and reducing trace real changes in the regulatory room for discretionary behavior of regula- environment. tors, countries help reduce risks for inves- tors as entry and operating costs become Capturing Sources of Regulatory more predictable ex ante. De jure protection of rights (that is, protection based on legal Risk provisions) provides guarantees for inves- Regulatory risk, in this study’s definition, is tors against unexpected interferences. An related to selected features of the legal and effective recourse mechanism can help mini- regulatory framework that might affect the mize the ex post costs of disputes for inves- expected profitability of a business. In the tors and provide “teeth” for the protection same vein that political risk reflects the vari- guarantees by making it costlier for govern- ability in economic returns that stems from ments to violate them. In other words, it uncertainty about political events, the con- provides a way to sanction deviating behav- cept here is closely related to uncertainty ior and determine the credibility of legal about laws and regulations. 4 Given that promises. sources of regulatory risk can be extremely Accordingly, even though the analysis heterogeneous and new risks will continue to measures regulatory risk through these three emerge, the measure in this study does not dimensions, their effectiveness is dependent include specific regulatory changes. Rather, on one another. In fact, as will be shown the analysis aims to capture features of coun- later, countries that perform better on one tries’ regulatory frameworks that can limit dimension are often better performers in the potential for unexpected losses due to other dimensions of the regulatory risk arbitrary government conduct that generates ­ measure as well. uncertainty for investors. The analysis combines existing indicators and collects new data sources, including the content of legal instruments, to measure Three Questions to Assess Risk countries’ regulatory risk, guided by the three analytical questions. To ensure cross- The study uses three analytical questions to country availability, the data focus on a few assess how the legal and regulatory frame- regulatory areas that apply to investors, both work affects risk for investors: foreign and domestic, or in some cases, only • Is there transparency regarding the con- one of them (for example, international tent as well as the process of making investment agreements [IIAs]). The areas laws and regulations that apply to covered are investment (specifically, invest- investors? ment laws and treaties); public procurement; 1 3 4    Global Investment Competitiveness Report 2019/2020 BOX 4.1 Data Sources for the Regulatory Risk Measure This study’s measure of regulatory risk relies on both The “cross-sectional” version of the regulatory risk primary and secondary data that cover cross-cutting data maximizes the number of underlying variables regulatory governance as well as specific regulatory covering but only for a single year (2017 data only), ­ areas: investment, public procurement, and property 86 countries. In addition to the information used registration. In annex 4A, table 4A.1 lists the full in the panel version, its data sources include the set of seven data sources, the scale of the raw values following: for each variable, and which of the three pillars they • A new investment laws database, compiled for this belong to. study, codifies the content of publicly available B e c au s e t h e d at a s ou r c e s d i f f e r i n t h e i r investment laws of 102 countries.c A framework of geographical and time coverage, the constructed 162 questions was developed to assess the content of risk measure includes two different versions. The the laws and create a database that includes a series “panel version” of regulatory risk data maximizes of 0–1 indicators, depending on whether the laws comparability over time (using 2014 –17 data), include specific provisions. Only those investment covering up to 167 countries. Its data sources laws that are currently in force are included. include the following: Thus, investment law scores are included only in • A database of 2,103 international investment agree- the cross-sectional indicator for 2017 and not in the ments (IIAs) a includes those that were in force panel version. between 2014 and 2017, whose content is mapped • T h e Wo r l d B a n k ’s B e n c h m a r k i n g P u b l i c by the United Nations Conference on Trade and Procurement (BPP) database is used to create a Development (UNCTAD).b The IIA provisions are score for public procurement. scored using principles similar to those applied to • Subindicators from the World Bank’s Global the investment laws database (used in the “cross- Indicators of Regulatory Governance (GIRG) cover ­ sectional” version described below). The IIAs virtually all countries (more than 186). are publicly available from UNCTAD’s online Investment Policy Hub. a. The UNCTAD Investment Policy Hub database (https://investmentpolicy​ .unctad.org/) includes bilateral investment treaties (BITs) as well as other treaties • A panel database was compiled of members of the with investment provisions. Convention on the Recognition and Enforcement b. The measure does not include survival clauses that extend the application of a treaty to a certain number of years beyond its termination. Although BITs of Foreign Arbitral Awards (the New York between members of the European Union (EU) were included in the analysis, Convention) and Convention on the Settlement of they are expected to terminate over the coming months. On January 15, 2019, Investment Disputes between States and Nationals EU member states endorsed a political declaration to terminate their intra-EU BITs by December 6, 2019. This followed a recent decision by the European of Other States (the ICSID Convention) for the uni- Court of Justice (Achmea v. Slovak Republic), which ruled that investor-state verse of all countries. arbitration provisions in BITs between EU member states are not compatible • World Bank Doing Business subindicators are used with EU law. c. In total, the database includes investment laws for 102 countries (as of this that were available from 2014 onward and thus ­writing) that are publicly available on UNCTAD’s Investment Policy Hub or that define the time coverage for the panel index. were received directly from government counterparts. cutting property registration; and other cross-­ protection, and recourse). The legal provi- regulatory governance measures. sions included across all these data sources An important criterion for including data are “scored” based on how they increase sources is that they can be linked to specific transparency, protection, and access to regulatory provisions that lend themselves to recourse through a specific set of principles, government action. The measure covers as shown below. select aspects of these distinct regulatory The dimension of transparency includes areas that fit into one of the three pillars three elements: (a) systematic publication of (transparency and predictability, investment and consultation on laws and regulations; R e g u l a t o r y R i s k a n d F D I    135 (b) availability of portals and similar mecha- is rooted in the legal and international politi- nisms that enable investors to find information cal economy literature. It aims to capture about relevant laws and regulations; and (c) to characteristics of the overall regulatory envi- a more limited extent, the specificity and clar- ronment and instruments such as investment ity of legal provisions on the applicable admin- laws and treaties that governments often istrative procedures (to increase predictability adopt to limit the risk to outside investors. and reduce chances of abuse of discretion).5 Substantively, protection from unpredic­ The standards of protection are selected table government conduct is generally one of based on the importance of the protections the core purposes of IIAs and investment to investment operations, especially in the laws. They present the fundamental principles context of unpredictable or nontransparent of investment protection within a country’s government actions. These include provi- investment policy regime. Some evidence sug- sions on expropriation, fair and equitable gests that IIAs can work as either a commit- treatment (FET), and transfer of funds. 6 ment device that protects investors covered by Further, the focus is on standards that are the relevant IIAs or as a signal that countries “absolute” in nature.7 have laws and policies in place that protect all For the recourse pillar, investors’ access to foreign investment. (For a review of the evi- investor-state dispute settlement (ISDS) is dence, see Bonnitcha, Skovgaard Poulsen, and reviewed as well as proxies for overall quality Waibel 2017.) Importantly, IIAs appear to of domestic dispute resolution, including the matter more for investments that entail high quality of land dispute resolution and the sunk costs—such as infrastructure (Bauerle judicial processes. Danzman 2016) or fixed capital investments Figure 4.1 summarizes the three pillars of (Kerner and Lawrence 2014)—or that are the regulatory risk measure and the key prin- capital intensive (Colen, Persyn, and Guariso ciples for evaluating the legal provisions 2016). under each pillar. For further details on how Further, it has been argued that it is not the these principles translate into specific provi- ratification of IIAs per se but the treaty sions selected and the rationale for such selec- “strength” (including dispute provisions) that tion, see the first section of annex 4A. is important for FDI inflows. For example, This analytical framework, including the Dixon and Haslam (2016) find a positive three pillars and areas of regulations covered, association between the strength of protection FIGURE 4.1  Three Pillars of Regulatory Risk Frame the Analysis Pillar 1: Transparency Pillar 2: Protection Pillar 3: Recourse Is there transparency regarding the What is the extent of legal protection Do investors have access to content as well as the process of provided to investors against effective mechanisms for recourse, making laws and regulations that arbitrary, unpredictable, or in case of grievances or disputes? apply to investors? nontransparent government actions? • Systematic publication of and consultation on laws • Absolute treatment standards • Investor-state dispute settlement and prevention and regulations • Protection guarantees against direct and indirect • Land dispute resolution • Registries or ICT platforms, and similar mechanisms appropriation • Quality of judicial processes to allow investors to find information about relevant • Provisions on transfers of funds and fair and laws and regulations equitable treatment (FET) • Specificity and clarity of legal provisions (to reduce space for discretion) Source: World Bank. Note: ICT = information and communication technology. 1 3 6    Global Investment Competitiveness Report 2019/2020 clauses in bilateral investment treaties (BITs) used for this research are available at the and FDI inflows. Frenkel and Walter (2019) economywide level only. Relatedly, the find that stronger international dispute settle- focus is on a few select legal instruments ment provisions in BITs are associated with such as IIAs and investment laws—again, positive effects on FDI activity. driven by limitation of cross-country Because of various empirical challenges, comparable data.8 evidence about the impact of IIAs remains ­ Finally, an important caveat is that the mixed. One common finding, however, is that framework sidesteps the issue of the “right to IIAs act as complements rather than substitutes regulate,” given its varied and unsettled impli- for local property rights and that countries cations. There have been growing concerns must have the necessary domestic institu- around limitations on the sovereign right to tions in place to make these international com- regulate and increased vulnerability to mitments credible and valuable to investors investor-state disputes. As a result, states have ­ (Hallward-Driemeier 2003; Tobin and Rose- started including provisions in IIAs to clarify Ackerman 2011). As such, the framework in their right to regulate. This is being done in a this study aims to cover regulatory aspects that few ways, such as by defining terms like apply to foreign investors but also serve as “indirect expropriation” and “fair and equi- proxies for the overall domestic regulatory table treatment” to limit the scope of inter- environment. pretation of these terms. In other cases, carve-outs and exceptions are included—for Caveats in the Regulatory Risk Measure example, to limit the application of provi- sions on expropriation and transfer of cur- Although these pillars provide a useful rency for legitimate regulatory measures,9 or framework to guide government actions to adjustments are made to secure states’ right to reduce regulatory risk, the regulatory risk regulate through jurisdictional exclusion of measure carries some caveats: regulatory disputes. For example, the • First, it focuses heavily on de jure legal Comprehensive and Progressive Agreement provisions. Because of lack of data, the for Trans-Pacific Partnership (Article 29.5) study does not review important imple- allows states to revoke the benefit of ISDS mentation aspects that can affect regula- with respect to claims challenging a state’s tory risk, such as the quality of day-to-day measures to control tobacco. functioning of all regulatory bodies and Indeed, provisions clarifying the right to variations in the interpretation and appli- regulate give arbitral tribunals and courts the cation of laws and regulations. option to fully weigh them against investor • Second, although the measure considers protection guarantees, and carve-outs allow whether there is consultation with stake- for nonapplication of guarantees. Arguably, holders during the rulemaking process, such provisions enable greater regulatory existing data cannot capture how widely flexibility and consequently can also generate or how well such consultation occurs and uncertainty for investors. Yet they also pro- the extent to which comments have led to vide predictability regarding the situations changes in laws and regulations. where exceptions can apply. Given the varia- • Third, even though regulations often vary tion in how “right to regulate” provisions by sector, given the limited availability of are drafted, the high subjectivity in their cross-country information on sectoral interpretation, and their multidimensional regulatory frameworks, the regulatory impact on investors and states, these provi- risk measure does not cover this dimen- sions have not been scored for the purpose of sion. Both primary and secondary data this study. R e g u l a t o r y R i s k a n d F D I    137 Constructing a Composite first components in a PCA (annex 4A, table 4A.2). The second approach has the advan- Measure of Regulatory Risk tage of considering the correlation structure Based on the three pillars in the analytical of the individual variables, but it is less trans- framework, individual data sources are com- parent in how they contribute to the overall bined into the three component scores as index.10 well as a composite risk score. The process Given these criteria and trade-offs, the anal- to construct the composite regulatory risk ysis ends up with two different sets of the over- score requires choices regarding how to nor- all regulatory risk measure: the panel version malize original data—which often have dif- (including the 2014–17 data) and the cross- ferent measurement units—and how to sectional version (richer data but available for aggregate the rescaled data. This study’s 2017 only). In addition, within each version, choices are guided by previous literature on the robustness of results is tested with the composite indicator methodology (Filmer inclusion (exclusion) of certain variables and and Prichett 2001; Gelman and Stanig 2016; two different aggregation methods discussed. Kaufman, Kraay, and Mastruzzi 2011; One caveat of the methodology is that it does OECD 2008), taking into account the fol- not calculate a range of uncertainty, such as in lowing criteria: Kaufmann, Kraay, and Mastruzzi (2011). Uncertainty makes comparison of adjacent • Comparability over time and across coun- points less credible (Høyland, Moene, and tries: The index should allow for assessing Willumsen 2012). For this and other reasons— progress of countries over time and rela- particularly, limited available cross-country tive to one another. data—caution should be taken in comparing • Maximized information content: The countries that are closer to one another in the index should consider the correlation constructed risk score. For further methodol- structure among different variables. ogy details, see Data Normalization and • Transparency: The construction method Aggregation in annex 4A. The data for all the should be simple enough to allow for an subcomponents are available in the online sup- understanding of how individual varia- plementary appendix.11 bles contribute to the composite index. Because of the varying time and country coverage of the different data sources (as Characteristics of the described in box 4.1), the study faces a trade- Constructed Measure of off between maximizing the information con- tent (number of variables used) and Regulatory Risk maximizing the size of the cross-section and The results show high correlations of the time dimension. In addition, to allow for the constructed risk measure across different inherent trade-off between maximizing infor- aggregation methods and inclusions of mation content and transparency, the study ­ d ifferent data sources (see annex 4A, tests for two different aggregation choices to ­ f igure 4A.1). Therefore, for brevity allow for an understanding of how individual throughout the chapter, unless indicated variables contribute to the composite score: otherwise, results using the panel version First, both the component scores and the of the composite risk score are reported. composite score are calculated as a simple The constructed score suggests that coun- average of the underlying (normalized) vari- tries differ substantially in the types and ables. Second, the component scores are cal- extent of regulatory risk, and some have culated as a weighted average of individual made regulatory changes that signifi- data sources, where weights are given as the cantly affect the level of risk over time. 1 3 8    Global Investment Competitiveness Report 2019/2020 FIGURE 4.2  Regulatory Risk Levels Vary Significantly across Countries and Change Substantially for Some Countries over Time 100 Nonweighted linear score overall, panel 80 60 40 20 Countries Source: World Bank. Note: Each bar is a boxplot of the regulatory risk measure (panel version) for one country during the sample period (2014–17), covering 167 countries. The bars are ordered by the average of the countries’ regulatory risk score (ranging from 0 to 100) during the sample period. The farther a bar is situated to the left (right), the lower (higher) is the country’s average regulatory risk. The bigger (smaller) a bar’s height, the bigger (smaller) is the country’s change in regulatory risk over time. For the full list and descriptions of data sources used to calculate aggregate scores, see annex 4A. Figure 4.2 plots the distribution of risk by for countries with more expensive risk pre- country as well as the range of changes in miums, the median as well as the 25th and each country between 2014 and 2017. On 75th percentiles of regulatory risk also tend average, the risk scores vary between 20 for to be higher. Figure 4.3, panel b, shows the the lowest risk score and 95 for the highest correlations of the regulatory risk measures risk score (out of a 0–100 range). While with the ICRG investment risk rating.13 In many countries have fairly small changes general, the riskier countries in the frame- over time, some have experienced significant work are also rated as riskier in the ICRG changes, especially those with a higher aver- rating. age risk level. In addition, higher regulatory risk appears Importantly, the results show that regula- to be associated with a more restrictive FDI tory risk, as measured in the framework, can regulatory framework. Figure 4.4 shows this explain perception of investment risk. This is result for a subset of 69 countries where stat- reflected in the correlation of regulatory risk utory restrictions to FDI, such as screening with the risk premium and other risk ratings. or equity restrictions, are captured by the Figure 4.3, panel a, presents the variations in OECD FDI Restrictiveness Index. Although the constructed regulatory risk across coun- restrictiveness per se does not necessarily ­ expropriation tries, based on the category of “ constitute risks for investors, it can increase and government action” risk insurance pre- uncertainty through higher chances for abuse mium (that is, risk insurance prices) evalu- of discretion when regulations are imprecise ated by Credendo, a major risk insurance or unclear. Growing p ­ rotectionism over the group.12 It shows that although there is some past few years has exacerbated policy and overlap in regulatory risk for countries across regulatory u­ ncertainty. The results here lend different risk premium categories, in general, empirical support to this concern and further R e g u l a t o r y R i s k a n d F D I    139 FIGURE 4.3  Higher Regulatory Risk Is Associated with Higher Expropriation Risk Insurance Premiums and Investment Risk Ratings a. Correlation with risk insurance premiuma b. Correlation with ICRG investment risk ratingb 100 15 Nonweighted linear score overall, panel 80 ICRG investment risk rating 10 60 40 5 20 0 0 1 2 3 4 5 6 7 0 50 100 Credendo risk rating: expropriation and government action Nonweighted linear score overall, panel Sources: Panel a: World Bank and Credendo; panel b: World Bank and ICRG. Note: The “nonweighted linear score overall, panel” refers to the “panel version” of the regulatory risk score (0–100), calculated from 2014–17 data, covering 167 countries. For the full list and descriptions of data sources used to calculate aggregate scores, see annex 4A. See additional results on the correlation of the cross-sectional version in annex 4A, figure 4A.4. a. Figure is a boxplot of regulatory risk score (panel version) over the seven risk premium categories of “expropriation and government action risk.” The country risk rating categories data come from Credendo, a major credit insurance group. The higher (lower) the risk premium, the higher (lower) the risk. The correlation coefficient between Credendo’s risk premium rating and regulatory risk is 0.52 (significant at 1% level). b. Figure is a scatterplot of regulatory risk score (panel version) on the x-axis and International Country Risk Guide (ICRG) investment profile rating on the y-axis. ICRG investment risk is measured on a scale from 1 (low confidence, high risk) to 12 (high confidence, low risk). The correlation coefficient between ICRG investment risk and regulatory risk is -0.24 (significant at 1% level). FIGURE 4.4  Higher Regulatory Risk Is Associated with More Restrictive FDI Regulations 80 Nonweighted linear score overall, panel 60 40 20 0 0 0.1 0.2 0.3 0.4 0.5 FDI Restrictiveness Index 2014 2015 2016 2017 Sources: World Bank and Organisation for Economic Co-operation and Development (OECD) FDI Regulatory Restrictiveness Index, https://www.oecd​ .org/investment/fdiindex.htm. Note: Foreign direct investment (FDI) restrictiveness is measured by the OECD FDI Restrictiveness Index, which measures statutory restrictions on FDI in 22 economic sectors across 69 countries, including all OECD and Group of Twenty (G-20) countries. The correlation between the OECD FDI Restrictiveness Index and this study’s regulatory risk score is 0.33 (significant at 1% level). The “nonweighted linear score overall, panel” refers to the “panel version” of this study’s regulatory risk score (0–100), calculated from 2014–17 data. For the full list and descriptions of data sources used to calculate the panel aggregate scores, see annex 4A. 1 4 0    Global Investment Competitiveness Report 2019/2020 highlight the need for countries to manage openness, and country fixed effects (where the regulatory framework to restore investor appropriate). This relationship is robust to confidence. the choice of different index versions: panel a In summary, these results suggest that the shows the panel index, while panel b shows legal and regulatory provisions selected and the cross-section version. scored in the framework carry meaningful In addition, results using bilateral FDI signals of risk for investors. With these data (where the source countries can be results in mind, the next section examines identified) hold that FDI inflows decrease as a ­ whether the regulatory risk measure can host ­c ountry’s regulatory risk increases. predict FDI inflows and other investor Arguably, aggregate FDI inflows mask hetero- decisions. geneity in the individual composition of a host country’s FDI. By using a newly con- structed panel dataset of bilateral FDI inflows, Regulatory Risk and FDI the model is able to consider both home and At the Country Level, Lower Regulatory host countries’ characteristics.15 Risk Is Associated with Higher FDI In this model, regulatory risk also has a significant and negative impact on bilateral On average, total FDI inflows to a country FDI inflows. Moreover, conditional on the are negatively correlated with the level of destination country’s risk, there appears to regulatory risk as measured in the frame- be a negative correlation (though statisti- work. Figure 4.5 depicts this relationship, cally not significant) between FDI inflows using a log transformation of real net FDI and the difference in risk levels between inflows. 14 The model controls for gross the destination and source countries (see domestic product (GDP) per capita, trade annex 4A, table 4A.3). It suggests not only FIGURE 4.5  FDI Inflows Are Higher in Countries with Lower Regulatory Risk a. 2014–17 panel score b. 2017 cross-section score 20 10 Net FDI inflows (log transformation) Net FDI inflows (log transformation) 10 5 0 0 –10 –20 –5 10 20 30 40 50 60 70 80 90 100 30 40 50 60 70 80 Nonweighted linear score, panel Nonweighted linear score, cross-section including laws Conditional on GDPPC, trade openness, and country FE Conditional on GDPPC and trade openness Non-tax haven countries 95% CI Linear fit (excluding tax havens) Tax havens Source: World Bank calculations, from the World Development Indicators database. Note: The scatterplots show the correlation between net FDI inflows and regulatory risk index. Panel a uses a 2014–17 panel score; panel b uses a 2017 cross-section score. CI = confidence interval; FDI = foreign direct investment; FE = fixed effects; GDPPC = GDP per capita. R e g u l a t o r y R i s k a n d F D I    141 that the destination country’s risk matters Increased Likelihood that MNEs but also that investors from a lower- risk country might be affected more Will Invest in Locations with disproportionally. Lower Regulatory Risk To focus on the interaction between risk Investor-level data lend microfoundation and irreversibility of investments, the analy- support to the negative relationship found ­ sis examines activities of existing investors between risk and aggregate FDI. The analysis that are subject to some adjustment costs. uses a dataset of 14,335 parent companies These include the total amount of capital investing in 27,886 FDI greenfield or expan- invested that is illiquid or employment sion projects across 159 host countries expansion. Aggregate data from the US between 2014 and 2016. This dataset enables Bureau of Economic Analysis on activity of the exercise to test the relationship between a foreign affiliates are used, including expendi- host country’s regulatory risk and MNE-level ture on fixed capital, research and develop- investment size using an investor decision ment (R&D) expenditure, and number of model (box 4.2). employees. For established MNEs, the corre- The analysis confirms that regulatory lation between regulatory risk and invest- risk deters investor entry and expansion. ment and hiring activities is negative. This result is robust to the inclusion of host However, the correlation is not statistically country controls. Figure 4.6 depicts the significant, conditional on the host country’s estimated coefficients for the different mod- GDP per capita, trade openness, and country els tested, showing that higher regulatory fixed effects—possibly driven by a small risk has a negative impact on the likelihood sample size. (See the full regression results in ­ that foreign investors will enter or expand annex 4A, table 4A.4.) BOX 4.2 Examining How Regulatory Risk Affects Investor Location Decisions Models of determinants of aggregate FDI suffer from logit model is estimated to investigate the determi- many potential biases, driven by the difficulty in nants of multinational enterprises’ (MNEs) location accounting for unobservable country characteristics choices. A random utility model is assumed (McFad- and characteristics of investment types that can drive den 1974) whereby an investor chooses one location investment flows. An alternative approach is to study among other alternatives. It can be thought of as a individual investors’ location choices, especially in the profit-maximizing problem in which investors enter context where investors’ decisions are observed over the country with the highest expected profit, where time. In such a setting, the assumption that ­ location expected profits depend on regulatory risk and other choices are driven by location characteristics tends to be country characteristics, as follows: a less restrictive assumption when investor fixed effects Pr˚(investijt)= F(b 0 + b1riskjt + X'jtg + d i + εijt), (B4.2.1) (and hence investor heterogeneity) can be accounted for. To look for more rigorous evidence on the rela- where Pr°(investijt) represents the probability of firm tionship between regulatory risk and foreign direct i investing in country j; F() is the logistic cumulative investment (FDI), this chapter takes advantage of distribution function; riskjt denotes regulatory risk; data from fDi Markets, a Financial Times dataset x'jt is a vector of other host-country characteristics; of greenfield FDI transactions that allows identifica- d i denotes firm fixed effects; and the error term e ijt tion of the parent company and sector information captures residuals. not typically available in aggregate FDI inflows data.a Note: Previous studies of the location choices of multinationals using this approach The fDi Markets transaction data are transformed include Chen and Moore (2010) and Joyez (2015). into an investor-level dataset, and a conditional a. For more on the fDi Markets dataset, see https://www.fdimarkets.com/. 1 4 2    Global Investment Competitiveness Report 2019/2020 FIGURE 4.6  Higher Regulatory Risk Has a Negative Impact on Investors’ Likelihood of Entering or Expanding in a Host Country 0 Marginal effect of risk on probability of MNE investment (%) -0.5 –1.0 –1.5 –2.0 Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Model numbers Sources: World Bank, using data from fDi Markets, a Financial Times dataset (https://www.fdimarkets.com/). Note: The figure presents coefficient estimates on regulatory risk from a discrete location choice model using fDi Markets data. All models control for ln(population), ln(GDP per capita), GDP growth (annual %), and trade openness. In addition, models 2–9 include one of the following variables in this order: lower secondary completion rate; bank deposits (% GDP); top combined corporate income tax rate (%); World Governance Indicators (WGI) regulatory quality; Polity IV democracy measure; volatility of GDP per capita growth; exchange rate volatility; and Fitch Sovereign Rating (categorical variables). GDP = gross domestic product; MNE = multinational enterprise. For more detailed model explanations, see box 4.2. For all coefficients, by model, see annex 4A, table 4A.5. in a host country. All specifications control level of the top 25th percentile performer, for economic fundamentals. Furthermore, investors will be 5.5–22 percentage points different measures of a country’s institu- more likely to locate in the country. tional environment or other indicators of To put this result in further perspective, in macro or sovereign risks are included. All the same model, the explanatory power of the results are in the same direction and are regulatory risk score on FDI is comparable in statistically significant at the 1 percent magnitude to trade openness. When control- level. ling for both variables at the same time, a The effect of regulatory risk—or, 1 percent increase in the host country’s trade- conceived more positively, certainty—on GDP ratio is associated with a MNEs’ investment is of meaningful 0.3–0.6 ­ percentage point increase in an inves- magnitude. The results across different tor’s likelihood to enter or expand.16 That is, specifications suggest that, on average, a a 1 standard deviation increase in trade open- 1 percent reduction in regulatory risk ness is associated with a 28 percentage point increases the likelihood of an investor increase in likelihood of investor entry, on entering or expanding in a host country by average. In comparison, a 1 ­ standard devia- 0.5–2 percentage points. In other words, all tion decrease in regulatory risk is associated else equal, if the median country improves its with a 9 percentage point increase in likeli- performance (reduces regulatory risk) to the hood of investor entry. R e g u l a t o r y R i s k a n d F D I    143 How Elements of the Regulatory Finally, the econometric evidence that regulatory risk matters for investor decisions Framework Matter for is consistent with investor perception as doc- Investment umented in the 2019 GIC Survey: First, The three dimensions of the regulatory risk investors, especially large investors, consider measure—transparency, protection, and a host country’s legal and regulatory envi- recourse—are not independent of one ronment to be one of the most important another. Countries with better protection factors shaping their parent firm’s decision also tend to have higher measured quality of to invest in the country. Second, exposure to recourse mechanisms and regulatory trans- political risk in host countries triggers exist- parency (see annex 4A, figure 4A.3, panel a). ing investors to consider withdrawing invest- Nevertheless, most countries still have mark- ments or canceling planned investment. edly different performances across the three Third, for investors that experienced an pillars (see annex 4A, figure 4A.3, panel b). adverse political risk event, both the quality That is, they tend to perform substantially of rules and their implementation appear to better in one pillar than in others, implying be a major obstacle in such cases. (See the potential to improve the overall level of box 4.3 for more detailed survey results and risk by focusing on certain pillars. chapter 1 of the report for the survey’s meth- Consequently, the analysis also finds that odology and sample.) These findings are also each of the three risk pillars still has a posi- consistent with other surveys, which consis- tive effect on investor entry decision, condi- tently identify political risk and regulatory tional on other dimensions of the regulatory uncertainty as major concerns for foreign framework (figure 4.7). businesses.17 FIGURE 4.7  Different Dimensions of the Regulatory Framework (Transparency, Protection, and Recourse) Matter for Investment Transparency Dimension of regulatory risk framework Protection Recourse 01 3 02 06 04 5 0 0 .0 .0 .0 .0 .0 .0 −0 −0 −0 −0 −0 −0 Marginal effect of risk on probability of MNE investment (%) Sources: World Bank, using data from fDi Markets, a Financial Times dataset (https://www.fdimarkets.com/). Note: The figure presents results from a discrete location choice model using fDi Markets data, controlling for ln(population), ln(GDP per capita), GDP growth (annual %), trade openness, WGI (World Governance Index) regulatory quality, and volatility of GDP per capita growth. The plot shows coefficient estimates and 95% confidence intervals of the three components of regulatory risk. GDP = gross domestic product; MNE = multinational enterprise. 1 4 4    Global Investment Competitiveness Report 2019/2020 BOX 4.3 Importance of Political Risk and a Stable Regulatory Framework for  Investment Decisions: Confirmation from the 2019 GIC Survey Legal and Regulatory Environment importance is based on the percentage of investors Besides political and macroeconomic stability, inves- who rate a factor as “important” or “critically impor- tors consider the legal and regulatory environment tant.”) These differences may be driven by the pres- to be one of the most important factors shaping ence of restrictions that are applicable only to larger their investment entry decisions. In the 2019 Global firms and the greater regulatory scrutiny that large Investment Competitiveness (GIC) Survey, 42 percent companies tend to experience. of respondents consider it a “critically important” fac- Political Risk tor in their investment decisions (figure B4.3.1). Most (two-thirds) of existing investors would con- Further, countries’ legal and regulatory envi- sider withdrawing investments or canceling planned ronments are especially important for larger firms. investment in the face of political risk exposure in host On average, large firms (those with more than 250 countries (figure B4.3.2). This result is in line with the employees) rank the legal and regulatory environment finding that more than 70 percent of existing investors as their top investment consideration, while small and consider investment protection guarantees (against medium enterprises (SMEs) consider it to be only the political risk) to be “important” or “critically impor- fourth most important consideration. (Ranking of tant” for investment decisions (figure B4.3.1). FIGURE B4.3.1  Legal and Regulatory Environment Is the Third Most-Cited Investment Decision Factor Question: How important were the following factors in your parent company’s decision to invest in this country? Political stability 49.4 34.9 Macroeconomic stability 49.0 35.7 Legal and regulatory environment 42.0 42.1 Talent and skills 41.5 43.5 Low taxes 41.3 35.5 Market size 39.3 42.6 Physical infrastructure 38.6 40.8 Ability to export 38.4 31.7 Intellectual property protections 35.4 35.4 Investor protections 34.5 37.5 Low labor and input costs 33.2 41.4 Supply chain coordination 30.1 44.2 Local input sourcing 26.6 42.2 Resource endowments 18.5 31.7 Local acquisition targets 14.4 35.6 0 20 40 60 80 100 Share of respondents (%) Critically important Important Source: Computation based on 2019 GIC Survey. Note: Affiliates of multinational enterprises were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. “Political stability” is ranked ahead of “macroeconomic stability” because 49.4 percent of respondents cited it as “critically important,” versus 49.0 percent for macroeconomic stability. Box continues next page R e g u l a t o r y R i s k a n d F D I    145 BOX 4.3 Importance of Political Risk and a Stable Regulatory Framework for  Investment Decisions: Confirmation from the 2019 GIC Survey (continued ) FIGURE B4.3.2  Expropriation and Breach of Contract Are the Most Likely to Affect Investments Adversely Question: How would the risk of the following events affect your investments in this country? At least one of the below 39 27 16 11 4 2 Expropriation 24 23 13 23 12 5 Government breach of contract 17 22 16 28 13 6 Sudden, adverse change in laws 12 16 25 36 7 4 Currency restrictions 9 13 29 33 12 5 Delay in permits and approvals 6 9 38 34 9 4 0 20 40 60 80 100 Share of respondents (%) Withdraw existing investment Cancel planned investment Delay planned investment Consider delay or cancellation None Don't know Source: Calculations based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. FIGURE B4.3.3  Investors Perceive the Quality of Rules and Their Implementation as Obstacles Question: How much of an obstacle are these areas of government conduct for your company in this country? Administrative complexity 30 40 Quality of laws 28 39 Bureaucratic discretion 29 37 Interagency coordination 22 38 Public agency capacity 20 38 Accessibility of laws 20 36 0 20 40 60 80 Share of respondents (%) Major obstacle Moderate obstacle Source: Calculations based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. Box continues next page 1 4 6    Global Investment Competitiveness Report 2019/2020 BOX 4.3 Importance of Political Risk and a Stable Regulatory Framework for  Investment Decisions: Confirmation from the 2019 GIC Survey (continued ) Risks of expropriation and government breach Quality of Rules and Their Implementation of contract evoke particularly negative investment Finally, the quality of rules and their implementation reactions. Experiencing such events would cause about contributes to investors’ exposure to political risk. 50 percent and 40 percent of investors, respectively, Among respondents who reported having experi- to consider withdrawing existing investments or enced exposure to political risk, the top three obsta- canceling planned ones. Sudden legal changes, currency cles related to government conduct are de jure factors restrictions, and delays in obtaining permits and (the complexity of administrative procedures, quality approvals elicit less severe reactions. Such risks are more of laws) as well as discretion on the part of bureau- likely to cause investors to delay investments rather crats who apply these laws or procedures in practice than to cancel or withdraw investments completely. (figure B4.3.3). Country Case Studies its land registry and cadaster and thus also improved on the Doing Business The primary differentiating feature of this Transparency of Information Index (within study’s constructed regulatory risk measure the “Registering Property” topic). Finally, is the link to specific actionable policy Senegal’s bilateral investment treaty (BIT) and regulatory levers. As a result, low per- with Canada, with strong transparency pro- formance on the measure—under any of the visions, came into force in 2016. three pillars—can in most cases be influenced The improvements documented in the by taking specific policy actions. data for Senegal reflect part of a broader To illustrate the policy and operational package of reforms initiated under the coun- implications of the regulatory risk measure, try’s Plan for an Emerging Senegal, adopted the discussion that follows presents cases of in 2014 (Republic of Senegal 2014). The countries that (a) significantly improved plan targets making Senegal an emerging over the sample period (Senegal); (b) signifi- market by 2035, attaining GDP growth of cantly declined (Indonesia); and (c) main- 7–8 percent, creating 600,000 formal jobs, tained consistently high performance and reaching GDP per capita of US$1,500. It (Kazakhstan). is based on three pillars: • Structural transformation of the economy Substantial Improvement: Senegal by consolidating current engines of growth and developing new sectors with a Senegal’s score on the regulatory risk mea- strong capacity to export and attract sure improved from 2014 to 2017, driven investment to create wealth, jobs, and mainly by improvements on the transparency social inclusion pillar. On the World Bank’s Doing Business • Promoting human capital by improving Communication of Tariffs and Tariff the well-being of the population Changes Index (within the “Getting • Enabling good governance in order to Electricity” topic), Senegal substantially out- strengthen security, stability, protection performed other Sub-Saharan African of rights and liberties, and consolidation ­countries.18 It made registering property eas- of the rule of law to create better condi- ier in 2016 by increasing the transparency of tions for social peace. R e g u l a t o r y R i s k a n d F D I    147 This period of reforms coincides with an Around the time of the terminations, increase in investment inflows. FDI inflows to Indonesia also became a respondent to a con- Senegal increased from US$409 million in troversial, high-stakes investment arbitration 2015 to US$587 million in 2017 (UNCTAD case. Churchill Mining PLC (a British com- 2019b). Six countries steadily increased pany) and Planet Mining Pty Ltd. (an investment during the period: China, the Australian company) filed arbitration cases Republic of Korea, Luxembourg, Morocco, against Indonesia, claiming over US$1 billion Poland, and Turkey.19 in damages.22 In February 2014, the two cases were consolidated, and the tribunal found that it had jurisdiction to decide on the case, negating Indonesia’s arguments oppos- Consistent Decline: Indonesia ing jurisdiction.23 The case was ultimately Indonesia’s score on all the three pillars fell decided in favor of Indonesia, and the claim- substantially from 2014 to 2017, reflecting ants were ordered to pay costs and arbitration an increase in risk levels. One main driver fees of nearly US$9.5 million.24 Nonetheless, was the termination of its IIAs.20 The con- from the experience of the Churchill case, the tent of IIAs is covered under all three pillars government perhaps also realized the value of of the regulatory risk measure, and therefore having clear treaty language to safeguard their termination affects performance on all. against claims based on conduct that was The number of IIAs in force fell from 41 unlawful or contrary to international public (mapped) in 2014 to 21 in 2018. All but one policy.25 were unilaterally denounced. IIA scores for Indonesia’s performance on the regulatory Indonesia across all three pillars declined risk measure aligns with investment climate consistently, with the largest decline in assessments indicating that regulatory uncer- 2016—also the year when most of the uni- tainty and lack of transparency are key fac- lateral terminations took place. FDI declined tors that impede operations of investors (US from US$16.641 billion in 2015 to US$3.921 Department of State 2017, 2019b). Investors billion in 2016 (UNCTAD 2019a). report that draft laws and regulations are Indonesia’s decision to terminate its BITs selectively published for public comment,26 came at a time when other countries also regulations are often vague and leave much started expressing concerns about IIAs room for interpretation, and drafts can take (including BITs) and the ISDS regime. Indeed, years to become law. Indonesia’s significantly there are legitimate concerns around expan- decentralized framework on lawmaking cre- sive or inconsistent interpretations of treaty ates further uncertainties.27 provisions; the qualifications and indepen- Indeed, political risk and regulatory uncer- dence of arbitrators; treaty shopping; lack of tainty remain critical issues for investors. In transparency; and high costs of dispute settle- the 2019 GIC Survey, more than 90 percent ment. As the broader IIA regime undergoes of respondents in Indonesia consider invest- reform, the challenge for developing countries ment protection against political risk to be is in making adequate adjustments to address “important” or “critically important.” To shortcomings yet ensuring that IIAs remain an address the issue of regulatory uncertainty, effective risk mitigation tool for the country. Indonesia has made specific efforts since More specifically, most of Indonesia’s BITs 2017. For example, Presidential Instruction were signed in the 1990s, when the realities of No. 7/2017 was issued, requiring ministries the country were very different from today:21 to coordinate before issuing regulations, to it was not a Group of Twenty (G-20) member, conduct regulatory impact assessment, and to was relatively unstable, and was not a capital provide opportunity for public consultation. exporter. The global economic landscape and Further, Presidential Regulation No. 95/2018 political economy has changed substantially on e-government was issued, requiring that since then. all levels of government (central, provincial, 1 4 8    Global Investment Competitiveness Report 2019/2020 and municipal) implement online governance political leadership. Providing impetus tools to improve overall transparency. These toward this goal is Kazakhstan’s strategic initiatives happened after the period covered location along China’s Belt and Road in the regulatory risk measure. Initiative. And after joining the World Trade As Indonesia progresses with its IIA reform Organization (WTO) in 2015, Kazakhstan efforts, key aspects that it may consider are made several reforms in its regulatory frame- (a) clarifying the definitions of investment, work (such as eliminating local content FET, and indirect expropriation; (b) including requirements). reasonable and limited exceptions and carve- Despite institutional and legal improve- outs to ensure regulatory space for states; and ments, investment climate assessments indi- (c) refining the scope of ISDS.28 cate that challenges remain relating to Notably, countries are also exploring insti- continued corruption, inefficient bureaucracy, tutional mechanisms to prevent disputes by and arbitrary law enforcement, especially at ensuring better implementation of core invest- the regional and municipal levels (US ment protection obligations (similar to what Department of State 2019a). Other reported Vietnam is setting up, for example).29 Such areas of concerns are the government’s ten- mechanisms are part of the regulatory risk dency to challenge contractual rights, unan- framework, and thus Indonesia’s performance nounced tax audits, imposition of high and can be improved by putting them in place. In ad hoc fines, and other interventions in com- addition, this may also be an opportunity for panies’ operations. On paper, the government Indonesia to ensure greater harmonization has obligations to publish draft legislations. between all of its IIAs-BITs, and regional free However, investment climate assessments trade agreements (FTAs). Finally, it is impor- indicate that the legal and regulatory pro- tant that Indonesia not only adjust its IIAs cesses are largely opaque. Draft bills are avail- but also harmonize its domestic legal frame- able for public comment, but the process work to ensure consistency in its legal frame- occurs without notice, and some bills are work and its implementation. excluded altogether. From these investor perceptions reported in investment climate assessments, it appears Strong Performance: Kazakhstan that the real challenge in Kazakhstan is the Kazakhstan’s performance on the overall lack of enforcement of the legal framework. risk measure has been strong. Over the past There is also lack of trust in the court sys- few years, Kazakhstan has consistently tems—the main avenue for seeking improved on various indexes based on de enforcement. jure legal and regulatory provisions. For Lack of enforcement of the legal frame- example, in 2019, Kazakhstan came 28th work is also reflected in Kazakhstan’s out of 190 countries in the World Bank’s investor-state disputes. It has had 19 investor- ­ Doing Business rankings. It ranks 4th on the state disputes (based on publicly available “Enforcing Contracts” topic, 18th on the information), of which 5 were decided in ­ “Registering Property” topic, and 36th on favor of investors, 5 were decided in favor of the “Starting a Business” topic. Kazakhstan the state, and the remaining are either pend- also performs well on the OECD FDI settled.31 In all cases decided in favor of ing or ­ Regulatory Restrictiveness Index, with a the investors, the publicly available informa- score similar to Austria’s.30 All these indica- tion indicates, the tribunals found violation of tors suggest consistent improvement on de FET or expropriation provisions—both of jure measures. which are core legal guarantees provided in This is a result of Kazakhstan’s efforts in Kazakhstan’s legal framework. several areas: the importance of attracting In 2017, the government adopted the more FDI as a tool to advance productivity 2018–22 National Investment Strategy with the and growth is recognized by the country’s aim of increasing FDI by 25 percent by 2022 R e g u l a t o r y R i s k a n d F D I    149 (US Department of State 2019c). Priority areas More specifically, Kazakhstan does not identified in the strategy include investment include a negative list (which would clarify climate improvements, privatization, and the sectors and activities that are prohibited greater economic diversification. Mineral or restricted for FDI) either in its Investment extraction continues to dominate Kazakhstan’s Law of 2003 or any other instruments. economy, with 75 percent of its FDI stock in However, it continues to maintain restric- the extractives sectors.32 Diversification has tions on foreign ownership (such as a ceiling consistently been identified as one of the on foreign ownership of media). In addition, country’s priority areas. As Kazakhstan bolsters foreign investors report that new laws and its efforts to attract a more diverse range of decrees are passed that impose penalties for FDI, ensuring effective implementation of its periods before the laws or decrees came into laws and regulations and minimizing risks will force—without “grandfathering” existing be key—indeed, manufacturing and services investments (US Department of State sector investors tend to be more mobile. 2019c). Notably, Kazakhstan’s investment In addition to strengthening the level of law does not include a provision to address regulatory enforcement, Kazakhstan’s perfor- this aspect. mance on the measure also indicates certain Kazakhstan’s strong performance on the de jure areas that it can further improve. regulatory risk measure also raises the ques- Figure 4.8 benchmarks Kazakhstan’s perfor- tion of why higher-income countries are not mance on each risk pillar relative to its neigh- the highest performers on the regulatory risk boring countries in Central Asia and other measure. The measure is linked to specific comparators. It suggests that Kazakhstan legal and regulatory instruments, and some could focus on improving the transparency countries lack these instruments—in some pillar, in which it lags the furthest relative to cases because they might not be required, all comparator countries except Poland and given the other laws and regulations of the Turkey. country. For example, the analysis finds that FIGURE 4.8  Kazakhstan’s Scores on the Three Pillars of the FDI Regulatory Risk Measure a. Transparency b. Protection c. Recourse 100 100 100 Nonweighted linear score of transparency, panel Nonweighted linear score of transparency, panel Nonweighted linear score of transparency, panel 80 80 80 60 60 60 40 40 40 20 20 20 0 0 0 MES X L R POR L S X R PO S X L PO ME ME RU RU RU TU TU TU KAZ Comparators Other countries Source: World Bank. Note: “Other comparators” refers to Mexico, Poland, Turkey, and the Russian Federation. The “nonweighted linear score, panel” refers to the “panel version” of this study’s regulatory risk score (0–100), calculated from 2014–17 data. For the full list and descriptions of data sources used to calculate the panel scores, see annex 4A. FDI = foreign direct investment; KAZ = Kazakhstan. 1 5 0    Global Investment Competitiveness Report 2019/2020 fewer countries with high GDP per capita tend investor levels suggests that regulatory risk to have an investment law. Arguably, countries matters for investment attraction and reten- with a relatively higher GDP have invested tion. Further, the chapter has demonstrated more heavily in creating comprehensive legal that the overall risk measure and underlying frameworks and in building institutional data sources can be used as a starting point capacity to ensure implementation. Therefore, to detect broad areas of weakness in a coun- lesser reliance may be placed on instruments try and guide further research and like investment laws, which are often used as diagnostics. signaling devices to generate investor confi- The primary differentiating feature of the dence. The dataset on investment laws shows risk measure is the link to specific action- that countries that perform better on Doing able policy and regulatory levers. As a Business tend not to have an investment law. result, low performance on the measure, Nonetheless, investment laws continue to under any of the three pillars, can in most be a powerful instrument leveraged by coun- cases be influenced by taking specific policy tries for different purposes.33 They serve as an actions. important risk mitigation tool. This is also Performance on the transparency pillar. confirmed in the dataset, which indicates that Transparency can be strengthened through countries with higher political risk ratings actions that improve (a) systematic publica- tend to have investment laws. tion of and consultation on laws and regula- tions; (b) the availability of portals and other similar mechanisms enabling investors to find information about relevant laws and regula- Concluding Remarks tions; and (c) the specificity and clarity of spe- Growing protectionism has exacerbated pol- cific provisions. icy and regulatory uncertainty—with coun- Countries can adopt specific legal provi- tries adopting a variety of new measures to sions to mandate publication of laws, regula- protect national security or for other public tions, and regulatory plans as well as purposes. With the global decline in FDI consultations on proposed regulations. In over the past few years, competition between addition, they can clearly publish information developing countries to attract it has only on sectors in which there are investment intensified. Attracting FDI will require effec- restrictions. To increase the accessibility of tive government actions to reduce real and legal and regulatory information, countries perceived risk for investors. Existing risk can set up an online portal or other unified indicators often help inform investors’ deci- website. sions, yet those indicators often rely heavily To improve overall predictability, specific on perception and do not have a direct link actions can be taken on precise drafting of to what governments can do or have done to administrative procedures. For example, affect risk. Relating specific government where an investment approval is required actions to the measurement of risks is one under the investment law, countries can potential avenue to inform policy makers in improve their performance by specifying the their quest to reduce investment risks. criteria on which approval would be granted This chapter has presented a new frame- and periods within which such approval work to measure regulatory risk that is linked should be granted. In the area of procure- directly to specific legal and regulatory provi- ment, countries can improve their perfor- sions, drawing on existing indicators and mance on the transparency pillar by ensuring newly constructed data on the content of that tendering documents include criteria for selected legal instruments. It shows that regu- evaluation of bids as well as the main terms of latory risk, as captured in this framework, the contract and payment schedule. carries meaningful signals of risk for inves- Performance on the protection pillar. tors. The evidence at both the country and Protection guarantees for investors can be R e g u l a t o r y R i s k a n d F D I    151 strengthened largely by improving select legal improves (by regulatory areas, time period, provisions in a country’s investment law or and geographical coverage), additional IIAs, in accordance with relatively well-­ research can be undertaken to better under- established good practices. For example, legal stand the impact of regulatory risk on differ- provisions should protect against both direct ent types of investors as well as which and indirect expropriation and should components of risk matter most for ­ mandate timely and adequate compensation. investors. Legal provisions that guarantee the investors’ ability to transfer funds in convertible cur- rency in a timely manner can also improve Annex 4A. Construction and performance on this pillar. Of course, drafting of any legal provisions will entail not only Characteristics of the Composite consideration of well-established principles of Regulatory Risk Score investor protection but also the country’s Selection and Scoring of Legal and overall context, legal traditions, and political Regulatory Provisions economy realities (including flexibilities that need to be provided to reflect the right to The selection of the specific legal and regula- regulate). tory provisions to be included in the risk Performance on the recourse pillar. score, and how to evaluate their contribu- Recourse for investors can be improved by tion to risk, is an inherently subjective allowing access to a wide range of dispute e xercise. This section details the decision ­ settlement mechanisms, including state- rules used in the study, based on the analyti- state as well as investor-state arbitration. cal framework and data availability, within Membership in the New York Convention, three pillars: transparency, protection, and which can facilitate enforcement of recourse. awards, can also improve performance. Pillar 1: Is there transparency regarding In addition, countries can consider setting the content as well as the process of making up an institutional mechanism to laws and regulations that apply to investors? systematically prevent investor-state Here transparency includes three dimensions: ­disputes.34 Overall strengthening of judicial (a) systematic publication of and consultation processes—through availability of on laws and regulations; (b) availability of specialized commercial courts, stipulation portals and other similar mechanisms, to of time periods for judicial processes, and allow investors to find information about rel- implementation of case management evant laws and regulations; and (c) to a lim- systems—can also improve performance on ited extent, the specificity and clarity of legal this pillar. provisions to increase transparency on the This chapter sets the foundation for fur- applicable administrative procedures (to ther research on several related aspects. increase predictability and reduce chances of A major limitation of the current framework abuse of discretion). is the relatively small number of regulatory The measure covers areas that can be evaluated because of a lack • Whether states have an obligation to pub- of comparable data across countries. In using lish laws and regulations affecting invest- this measure, important caveats discussed ear- ment, and whether they do publish laws lier in this chapter—including the current and regulations either on a unified web- focus on de jure legal provisions and coverage site or in an official gazette;35 of a relatively limited set of regulatory areas— • Whether countries publish their negative should be kept in mind. To improve the pre- or positive lists either in their investment dictive power of the risk measure, additional law or elsewhere in the legal and policy regulatory areas may be incorporated, such as framework;36 trade regulations. As data availability 1 5 2    Global Investment Competitiveness Report 2019/2020 • Whether international investment agree- process (before the final law or regulation is ments (IIAs) contain provisions con- approved), the measure covers whether states cerning any mechanisms for technical have an obligation under investment laws or cooperation (including for provision of IIAs to publish and consult on proposed laws information to the private sector); and regulations, and whether there is a period • Whether procurement laws and regula- set by law during which the text of the pro- tions, notices of calls for tender, tender posed regulations should be made publicly documents, notices of award, and minutes available.42 It also covers whether regulatory of bids are made publicly available; and plans are published,43 public consultation is • Whether bids are opened electronically undertaken on proposed regulations (not yet and whether minutes of bid processes are passed), and reports are issued on the consul- published online. tation process.44 Pillar 2: What is the extent of legal It incorporates the World Bank’s Doing p rotection provided to investors against ­ Business indicators to measure arbitrary, unpredictable, or nontransparent • Whether information on land ownership, government actions? For the purpose of this documentation requirements for land reg- study, the standards of protection reviewed istration,37 fee schedules, and electricity were selected based on the centrality of the tariffs were made publicly available and protections to investment operations;45 their changes notified;38 and particular relevance in the context of arbi- • Whether requirements for obtaining a trary, unpredictable, and nontransparent building permit are clearly specified in government conduct; and whether they are the building regulations or on any acces- “absolute” in nature.46 The standard of pro- sible website, brochure, or pamphlet as tection available to investors was measured well as whether building laws and regula- on the basis of the quality of provisions on tions were publicly accessible.39 expropriation, transfer of funds, fair and equitable treatment (FET), and nonderoga- Indeed, the degree of specificity and clarity tion. Indeed, variations may exist in the of drafting of legal and regulatory provisions interpretation of various provisions and determines the room regulators have to exer- jurisprudence; however, as discussed earlier cise discretion—and thus affects regulatory in the chapter, because of lack of availability risk. The measure covers this aspect to a lim- of cross-country comparable data, such vari- ited extent, largely focusing on a few specific ations are not included in the measure. administrative processes affecting foreign The measure covers whether protection direct investment (FDI). For example, it covers is explicitly provided against both direct whether, in cases where foreign investors need and indirect expropriation47 and whether to obtain an investment approval to invest in a several key elements are included to ensure country, the criteria and time frames for grant- the legality of expropriation:48 that expro- ing such approvals are stipulated in the law. priation is done (a) only for public purpose; On procurement, the measure covers (b) in a nondiscriminatory manner; (c) fol- whether procedures for acceptance of com- lowing due process; and (d) against pay- pleted works and termination of contracts are ment of prompt, adequate, and effective specified in the law. 40 Further, it covers compensation. These specific conditions for whether tender notices and documents expropriation constitute a widely accepted include specific criteria for the evaluation of legal standard. The measure also covers bids, main terms of the contract, and pay- whether investors are guaranteed the ability ment schedule. It also considers whether to freely transfer funds in a timely manner grandfather clauses are included.41 and in a freely convertible or freely usable To measure the extent to which countries currency.49 ensure transparency in the rulemaking R e g u l a t o r y R i s k a n d F D I    153 Further, the measure covers whether a spe- • Whether countries are members of the cific FET provision is included. FET is a com- ICSID and the New York Convention. 55 posite or a bundle of rights available to ICSID membership allows investors to investors. Although the FET standard is gener- pursue arbitration proceedings against ally not precisely defined in IIAs, it has been the host state under the ICSID Conven- clarified through various decisions of arbitral tion, 56 which requires automatic rec- tribunals. These interpretations indicate that ognition and enforcement of the pecu- FET is an obligation on states to act in a trans- niary aspects of awards by all member parent, consistent, reasonable, and propor- states. 57 tional manner and to respect legitimate The World Bank’s Doing Business indica- expectations of investors generated from writ- tors were incorporated to measure ten commitments. Investors have often used the FET standard to seek regulatory stability.50 • Whether countries have adopted good The FET provision may either be “qualified” practices in their court system in four (with reference to international law or to a list areas58—court structure and proceedings, of underlying obligations) or “unqualified.”51 case management, court automation, and Finally, the protection measure covers alternative dispute resolution—including whether, if the legal instruments conflict with aspects such as law regulating the number other legal norms (other laws, regulations, of adjournments allowed, availability of a and IIAs), the more favorable rules apply to case management system and electronic investors. filings, and availability of commercial Pillar 3: Do investors have access to effec- courts; and tive recourse mechanisms in case of griev- • Whether countries have adopted good ances or disputes? The measure covers practices in ensuring accessibility to land dispute resolution mechanisms, 59 includ- • Whether investors have recourse to ing availability of out-of-court compensa- investor-state dispute settlement (ISDS)52 ­ tion mechanisms and databases to verify and the full scope of such a right;53 accuracy of government-issued identity • Whether investors could submit an invest- documents. ment dispute under the ICSID Convention and United Nations Commission on Inter- Table 4A.1 lists the underlying variables national Trade Law (UNCITRAL) rules;54 of the three index components, indicating • Whether investors had recourse to other the original scale of the raw variables. types of alternate dispute resolution mechanisms, such as mediation or concili- Data Normalization and Aggregation ation, either voluntarily or as a mandatory procedure before any adjudicatory proce- The process to construct the composite regu- dures (such as arbitration) can begin; latory risk measure is described below. • Whether investors had access to domestic courts either as an option alongside other Normalization ISDS forums or as a mandatory step To preserve comparability of the constructed before submitting a claim to investor- min-max scores over time and cross-country, a ­ state arbitration; aggregation approach was chosen, whereby • Whether state-state dispute settlement is all individual variables will be rescaled as available; (Max–Min)/Range of (possible) ­ values. Each • Whether domestic investment laws pro- underlying variable is transformed to a scale vide access to any alternate institutional from 0–100, where 0 indicates the best possi- mechanisms to address investor issues ble outcome and 100 the worst—reflecting before they escalate into legal disputes; that the constructed regulatory risk measure and is an index of risk. Some data sources, such as 1 5 4    Global Investment Competitiveness Report 2019/2020 TABLE 4A.1  Underlying Variables of the Regulatory Risk Measure, by Dimension and Subindicators Pillar 1: Transparency Pillar 2: Protection Pillar 3: Recourse UNCTAD IIA mapping: Provisions UNCTAD IIA mapping: Provisions on fair UNCTAD IIA mapping: Dispute on transparency and technical and equitable treatment, expropriation, settlement provisions (0–100) cooperation (0–100) transfers (0–100) Doing Business (World Bank): Doing Business (World Bank): • “Registering Property”: “Enforcing Contracts”: Quality of •  Transparency of Information Index Judicial Processes Index (0–18) (0–6) “Registering Property”: Land Dispute •  • “Getting Electricity”: Resolution Index (0–8) Communication of Tariffs and Tariff Changes Index (0–1) • “Dealing with Construction Permits”: Quality of Building Regulations Index (0–2) Global Indicators of Regulatory Global Indicators of Regulatory Governance (World Bank): Governance (World Bank): Laws are publicly available; Challenging regulations (0–1) regulatory plans are published; public consultation is conducted on proposed regulations; results of consultation process are reported (0–4) Investment laws: Provisions on sector Investment laws: Provisions on Investment laws: restrictions, screening/approval/ expropriation, transfers, and fair and Provisions on dispute settlement and notifications, access to laws, equitable treatment (0–100) dispute prevention (0–100) transparency, and grandfathering (0–100) Benchmarking Public Procurement (World Bank): Transparency, clarity, access to information (0–100) Membership in ICSID: (0-1) Membership in the New York Convention: (0–1) Source: World Bank. Note: IIA = international investor agreement; ICSID = International Centre for Settlement of Investment Disputes (World Bank Group); UNCTAD = United Nations Conference on Trade and Development. The “New York Convention” refers to the United Nations Convention on the Recognition and Enforcement of Foreign Arbitral Awards. the World Bank’s Doing Business and Global each of the three pillars, provisions of interest Indicators of Regulatory Governance (GIRG) are identified. The legal texts are broken subindicators as well as convention member- down into a number of simple, mostly yes/no, ship dummy variables, which are numerical questions. Responses to questions relevant to values at the country level, only require this provisions of interest are chosen and con- straightforward transformation to a common verted into scores. Scores for the questions/ scale and direction. answers were then aggregated and normal- Box 4A.1 describes how legal texts (IIAs, ized to a [0,1] score for each provision. All investment laws, and public procurement reg- relevant provisions for a given pillar are then ulations) are transformed into scores. For aggregated and normalized to the pillar level. R e g u l a t o r y R i s k a n d F D I    155 BOX 4A.1 Scoring and Aggregating Rules for IIAs Step 1: Scoring the transparency, protection, and • The weighted average of country pair scores is taken, recourse pillars where weights are the partner country’s share in world gross domestic product (GDP). Underlying • Each provision is scored using a set of more detailed assumptions are that the pool of potential investors 1/0 (yes/no) questions. increases with the size of the international investment • The scores for each question are then added up and agreement (IIA) network and the partner’s economy. normalized to a scale [0–1] for each provision. • The scores are aggregated as in the following Step 2: Aggregating provisions to treaty level notation: • For each law or treaty, the total score of all provi- sions is taken as a simple average of all provisions. pi = ∑ j ∈ IIAij w o MFN { } w j × max pijt + t ∑ j ∈ IIAij with MFN wj × max { pik }, k A higher score indicates lower risk. • Scores are meant to be on an ordinal scale. (B4A.1.1) Step 3: Aggregating scores to the country level where pi presents the score for each pillar (transpar- • The “relevant” treaty for a country pair is deter- ency, protection, and recourse) at the country level; wj mined, considering the network of available treaties presents the weights; and and pijt presents the score of (figure B4A.1.1): the IIA for the country pair. FIGURE B4A.1.1  Model of a Country with Multiple Treaty Relationships Partner A A-X treaty: No MFN B-X treaty: MFN Partner Country B X D-X treaty 1 D-X treaty 2 Partner Partner D C Note: MFN = most-favored nation. For example, a relevant provision for the are used to construct the overall transparency transparency score of investment laws is score for investment laws. Other provisions “access to laws.” This is scored based on one are more complicated to score and require var- question: “Does the act guarantee accessibility ious questions (and legal know-how). of laws, regulations, and other legal instru- However, in the case of IIAs, this leaves a ments to investors?” If the answer is yes, a score at the treaty level rather than the coun- score of 1 is assigned; otherwise, a score of 0 is try level. To obtain a country score for each of assigned. This provision as well as four others the pillars, the relevant treaty is first identified 1 5 6    Global Investment Competitiveness Report 2019/2020 at the country pair level. If a pair of countries World Bank’s Benchmarking Public has more than one treaty at the same time, the Procurement [BPP] and GIRG databases); and better score is considered the relevant one. (b) a cross-section version with more variables Furthermore, if a treaty between a pair of (from the aforementioned sources) for a single countries has a most-favored nation (MFN) data year (2017). (See box 4.1 for a more provision, then the best score out of all host detailed description of the two versions.) country scores is assigned.60 In addition, the two versions were tested In a second step, these country-pair scores including different data sources. The overall are collapsed to the host country level by cal- score is robust to the inclusion (or exclusion) culating the partner country gross domestic of the investment law database, as shown by product (GDP) weighted average. This high correlations of the scores constructed weighting reflects the following assumptions: using different sets of data (figure 4A.1, all else equal, the level of protection increases panel a). with (a) the number of IIA partners; and The overall index of regulatory risk is the (b) the size of the partners’ economies. simple average of the three component scores. There are opposing views in the literature Each component score (transparency, about the marginal effects of additional IIAs protection, and recourse) is a composite score in attracting FDI. On the one hand, if invest- of its underlying variables. Two methods were ment treaties are pure signaling devices about tested to aggregate individual variables into the a host country’s commitment to protect inves- component scores: a simple average and a tors, then additional treaties have decreasing weighted average where weights are given by returns (Bubb and Rose-Ackermann 2007). the first component from a principal On the other hand, as Montt (2009) argues, component analysis (PCA). Table 4A.2 shows IIAs can have increasing returns because the weights for the cross-section version of the investors could expect a more predictable and risk measure, derived using PCA for each of efficient jurisprudence to evolve with the size the three pillars—transparency, protection, of the treaty network. Given these possible and recourse. It suggests that none of the opposing effects, a simple rule was followed variables included in the framework has an in which the level of protection increases lin- outsize influence on the overall risk early with the number of partner countries. components. The two aggregation methods yield very Aggregation of the Composite Score high correlations (figure 4A.1, panel b). In combining different data sources, a trade-off This chapter refers to the simple average ver- arises between maximizing the number of sion when referring to the index. (informative) variables used to construct the index and maximizing the size of the cross- section and time dimension. Thus, two Characteristics of the Regulatory Risk v ersions of the index were constructed: ­ Measure: Additional Results (a)  a panel version of comparable data for 2014–17, which excludes some data sources Additional results are presented in figures not available for the full period (an investment 4A.2, 4A.3, and 4A.4, in addition to tables law database constructed for this study and the 4A.3, 4A.4, and 4A.5. R e g u l a t o r y R i s k a n d F D I    157 FIGURE 4A.1  The Regulatory Risk Measure Is Highly Correlated across Varying Data Sources and Aggregation Methods a. Correlations across different data sourcesa Nonweighted linear score, overall, panel 80 60 Nonweighted linear score overall, 40 cross-section 20 80 Nonweighted 60 linear score overall, cross-section, including 40 laws 20 0 50 100 20 40 60 80 b. Correlation between aggregation methods (simple average and PCA-weighted average) 100 Nonweighted linear score overall, panel 80 60 40 20 20 40 60 80 100 (PCA) weighted linear score overall, panel Source: World Bank. Note: “Nonweighted [or weighted] linear score overall, panel” refers to the “panel version” of this study’s regulatory risk score (0–100), calculated from 2014–17 data. “Nonweighted linear score overall, cross-section” refers to the “cross-section version” of the risk score (0–100), calculated from 2017 data. For the full list and descriptions of data sources used to calculate the panel aggregate scores, see annex 4A. PCA = principal component analysis. a. The correlation matrix plots between different versions of the risk measure—panel version and cross-section version—with additional underlying variables. 1 5 8    Global Investment Competitiveness Report 2019/2020 TABLE 4A.2  Principal Component Analysis (PCA) Weights Pillar Variable PCA weights Transparency IIAs: Provisions on transparency and technical cooperation 0.183 Investment laws: Provisions on sector restrictions, screening/approval/notifications, access to 0.026 laws, transparency, and grandfathering GIRG: Laws publicly available, regulatory plans published, public consulted on proposed regulations, 0.207 consultation results reported BPP: Transparency, clarity, and access to information 0.088 DB: “Registering Property”: Transparency of Information Index 0.216 DB: “Getting Electricity”: Communication of Tariffs and Tariff Changes Index 0.142 DB: “Dealing with Construction Permits”: Quality of Building Regulations Index 0.139 Protection IIAs: Provisions on FET, expropriation, transfers 0.500 Investment laws: Provisions on FET, expropriation, transfers 0.500 Recourse IIAs: Dispute settlement provisions 0.221 Investment laws: Dispute settlement provisions 0.041 GIRG: Challenging regulations 0.086 ICSID membership 0.084 New York Convention membership 0.142 DB: “Enforcing Contracts”: Quality of Judicial Processes Index 0.221 DB: “Registering Property”: Land Dispute Resolution Index 0.204 Source: World Bank. Note: Weights have been normalized to sum up to 1. “New York Convention” refers to the Convention on the Recognition and Enforcement of Foreign Arbitral Awards. BPP = Benchmarking Public Procurement (World Bank); DB = Doing Business (World Bank); FET = fair and equitable treatment; GIRG = Global Indicators of Regulatory Governance (World Bank); ICSID = International Centre for Settlement of Investment Disputes; IIAs = international investment agreements. FIGURE 4A.2  Regulatory Risk Varies across Countries a. Distribution of the panel version b. Distribution of the cross-sectional version (including investment laws data) 0.04 0.08 0.03 0.06 Density Density 0.02 0.04 0.01 0.02 0 0 20 40 60 80 100 30 40 50 60 70 80 Source: World Bank. Note: Histograms show distribution of the panel version (panel a), and the cross-sectional version of the regulatory risk measure, including investment laws data (panel b). The “panel version” of the regulatory risk score (0–100) is calculated from 2014–17 data. The “cross-section version” is calculated from 2017 data. For the full list and descriptions of data sources used to calculate the panel aggregate scores, see annex 4A. R e g u l a t o r y R i s k a n d F D I    159 FIGURE 4A.3  Countries That Perform Better in One Pillar Often Perform Better in Other Pillars of Regulatory Risk a. Correlations between three pillars’ scoresa b. Variations across risk pillarsb Nonweighted linear score of 100 transparency, panel 100 Nonweighted linear score overall 80 50 Nonweighted linear score 60 0 of protection, panel 100 40 Nonweighted 50 linear score of 20 recourse, panel 0 0 0 50 100 0 50 100 Countries Source: World Bank. a. “Nonweighted linear score . . . , panel” refers to the “panel version” of the regulatory risk score (0–100), calculated from 2014–17 data. For the full list and descriptions of data sources used to calculate the panel aggregate scores, see annex 4A. b. In panel b, the blue dots present the overall regulatory risk scores for 2017 (average of the three risk pillars), covering 166 countries. The bars present the variations by the three pillars (that is, they denote the range determined by the two pillars with the lowest and the highest scores). FIGURE 4A.4  Higher Regulatory Risk Is Associated with Higher Expropriation Risk Insurance Premium 80 Nonweighted linear score overall, cross-section 70 60 50 40 30 1 2 3 4 5 6 7 Credendo risk premium level, expropriation and government action Source: World Bank and Credendo. Note: Figure is a boxplot of the regulatory risk score (cross-sectional version) over the seven categories of Credendo’s “expropriation and government action risk premium.” The higher (lower) the risk premium, the higher (lower) the risk. The “Nonweighted linear score overall, cross-section” refers to the cross-sectional version of the regulatory risk score (0–100), calculated from 2017 data. For the full list and descriptions of data sources used to calculate aggregate scores, see annex 4A. 1 6 0    Global Investment Competitiveness Report 2019/2020 TABLE 4A.3  A Gravity Model of Bilateral FDI Inflows, 2014–16   (1) (2) (3) (4) Regulatory risk (panel index, not −0.0501** −0.0308 including dispute data) (0.0225) (0.0354) Regulatory risk (panel index, including −0.0705*** −0.0560* dispute data) (0.0255) (0.0335) Difference between destination and −0.0199 −0.0155 origin countries’ regulatory risk (0.0298) (0.0239) Market size (GDP) 0.240 0.255* 0.241 0.256* (0.147) (0.148) (0.147) (0.148) Difference in income per capita −0.141** −0.153** −0.142** −0.154** (0.0598) (0.0607) (0.0599) (0.0608) Ln(distant) −0.503*** −0.529*** −0.502*** −0.529*** (0.0933) (0.0950) (0.0932) (0.0950) Contiguity −0.0321 −0.0541 −0.0302 −0.0518 (0.386) (0.383) (0.385) (0.382) A language is spoken by at least 9% of 0.429* 0.415* 0.428* 0.413* the population in both countries (0.245) (0.246) (0.245) (0.246) Ever in colonial relationship −0.0349 −0.0138 −0.0346 −0.0134 (0.234) (0.233) (0.233) (0.232) Common colonizer post-1945 −0.00605 −0.0128 −0.00519 −0.0119 (0.414) (0.414) (0.414) (0.414) Constant 5.416 4.289 6.319 5.406 (4.653) (4.587) (4.524) (4.477) Observations 68,086 61,396 68,086 61,396 Source: World Bank. Note: FDI = foreign direct investment; GDP = gross domestic product. Robust standard errors in parentheses: *** p < 0.01 ** p < 0.05 * p < 0.10 TABLE 4A.4  Regulatory Risk and Activities of Affiliates of US MNEs   Total employment CAPEX R&D expenditures Regulatory risk (panel index, not including dispute data) −0.101 −104.449 −7.594 (1.599) (113.909) (17.580) GDP per capita (constant 2010 US$) 0.000 −0.163** 0.048*** (0.000) (0.067) (0.006) Trade openness (trade as % of GDP) 0.025 6.411 −3.536 (0.116) (14.091) (2.536) Constant 241.723*** 11,666.918** 71.836 −77.31 −5,481.32 −732.201 Country fixed effects Yes Yes Yes Number of observations 105 100 101 Adjusted R2 −0.027 0.049 0.120 Source: World Bank. Note: Data on employment, CAPEX (capital expenditure), and R&D (research and development) expenditures are from the US Bureau of Economic Analysis (BEA) on activity of foreign affiliates. GDP = gross domestic product; MNE = multinational enterprise. Robust standard errors in parentheses: * p < .10 ** p < .05 *** p < .01 R e g u l a t o r y R i s k a n d F D I    161 TABLE 4A.5  A Discrete Choice Model of Global Investment Location, 2014–16   Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8 Model 9 Regulatory risk (panel −0.00887*** −0.0149*** −0.00645*** −0.00989*** −0.00793*** −0.00751*** −0.00767*** −0.00511*** −0.0200*** index, not including (0.000595) (0.000828) (0.000629) (0.000619) (0.000624) (0.000606) (0.000607) (0.000716) (0.000852) dispute data) ln(GDP per capita) 0.756*** 0.678*** 0.803*** 0.738*** 0.305*** 0.743*** 0.745*** 0.748*** 0.424*** (0.00571) (0.00783) (0.00647) (0.00593) (0.0111) (0.00617) (0.00574) (0.00697) (0.0144) ln(population) 0.808*** 0.836*** 0.932*** 0.792*** 0.824*** 0.827*** 0.782*** 0.824*** 0.798*** (0.00461) (0.00665) (0.00598) (0.00487) (0.00471) (0.00499) (0.00480) (0.00529) (0.00676) GDP growth 0.0373*** 0.0447*** 0.0290*** 0.0404*** 0.0313*** 0.0319*** 0.0428*** 0.0238*** 0.0476*** (annual %) (0.00205) (0.00244) (0.00210) (0.00210) (0.00226) (0.00210) (0.00221) (0.00241) (0.00287) Trade openness 0.00499*** 0.00578*** 0.00861*** 0.00511*** 0.00344*** 0.00610*** 0.00493*** 0.00457*** 0.00388*** (trade as % of GDP) (9.73e-05) (0.000123) (0.000130) (9.95e-05) (0.000100) (0.000126) (9.68e-05) (0.000105) (0.000111) Lower secondary completion rate, −0.0124*** total (% of relevant age group) (0.000524) Bank deposits (% GDP) −0.00536*** (0.000165) Top Combined CIT 0.00822*** Rate (%) (0.000904) WGI Regulatory 0.733*** Quality (0.0162) Polity IV: 0.00391*** Institutionalized (0.00138) Democracy Volatility of GDP per −0.110*** capita growth (0.00663) Exchange rate −6.239*** volatility (0.290) Fitch Sovereign Rating YES dummies Observations 2,825,609 1,349,349 2,308,090 2,564,859 2,825,609 2,563,529 2,825,609 1,641,111 1,805,949 Source: World Bank. Note: Results from discrete choice model as described in box 4.2, where investors choose to invest a location based on its characteristics relative to other locations. Investors are identified from parent company information from transaction data of fDI Markets, a Financial Times dataset (https://www.fdimarkets.com). CIT = corporate income tax; IV = instrumental variable. Standard errors in parentheses. * p < .10 ** p < .05 *** p < .01 Notes 1. In any case, political risk insurance generally https://www.prsgroup.com/explore-our​ does not cover the entire spectrum of a state’s -products/international-country-risk-guide/. conduct that can generate regulatory risks 3. For more information, see “Country Risk for investors. Classification,” OECD website: http:// 2. The ICRG model, created in 1980 and pro- www​ . oecd.org/trade/topics/export-credits​ duced since 1992 by investment risk com- /arrangement-and-sector-understandings​ pany PRS Group, issues ratings comprising /financing​-terms-and-conditions​/country-risk​ 22 variables in three subcategories of risk: -classification/. political, financial, and economic. For more 4. For a discussion of uncertainty about polit- information, see “The International Country ical events and political risk, see Kobrin Risk Guide (ICRG),” PRS Group website: (1979). 1 6 2    Global Investment Competitiveness Report 2019/2020 5. This principle concerns, for example, where individual variables are assumed to whether in cases where foreign investors need be noisy measures of a “true” underlying to obtain an investment approval to invest in (latent) governance indicator, as estimated by a country, the criterion and time frames for an unobserved components model. granting such approvals are stipulated in a 11. To access the online appendix, see www​ legal instrument. .worldbank.org/gicreport. 6. Expropriation, inability to transfer funds 12. Credendo is a European credit i nsurance ­ outside the host country, and instability in group that covers risk worldwide. In financial policy and regulatory environment have year 2017, the value of transactions insured consistently been identified as critical factors by Credendo amounted to €85 ­billion. affecting investor decisions to stay and poten- 13. The correlation patterns suggest that the reg- tially expand operations in a country. (See ulatory risk measure appears to distinguish World Bank [2018] as well as the 2019 GIC well between countries with very high or low Survey, which is covered in chapter 1 of this level of risks, but not the countries with very report and also discussed later in this study.) low risk. This is likely because the measure Breach of these standards has led to most largely includes de jure measures. For coun- international investor-state disputes. Also see tries that already have “good” rules on the the UNCTAD online Investment Policy Hub: book, additional information on implemen- http://investmentpolicyhub​.­unctad.org. tation would be needed to distinguish high 7. In contrast, the “national treatment” stand- and low performers. ard is a relative standard of treatment under 14. A common log transformation is used to which treatment of foreign investors and preserve negative and zero values of net FDI investments is assessed relative to the treat- inflows: the natural logarithm of 1 plus the ment accorded to domestic investors and absolute value of FDI, multiplied by (–1) if investment. the original FDI variable is negative. 8. Nonetheless, IIAs and investment laws are relatively standardized legal instruments 15. The bilateral FDI data are constructed where comparable (text) data are available from various sources, including the OECD across all countries. ­ bilateral FDI database and the International 9. Transfer of funds can be restricted tempo- Monetary Fund (IMF) balance-of-payments rarily in a nondiscriminatory manner and in International Investment Position (IIP) data. good faith in cases of a balance of payments (A forthcoming publication on the bilateral crisis or on legitimate application of certain FDI database will provide further details.) national laws—specifically, those relating A basic gravity model is estimated where host to bankruptcy, insolvency, or the protection and source country fixed effects are used to of the rights of creditors; issuing or trading control for the multilateral resistance term. in securities and other stock market instru- The model is estimated using the Poisson ments; criminal offenses; compliance with pseudo maximum likelihood and has the lim- orders or judgments in judicial or adminis- itation that observations with negative FDI trative proceedings; and compliance with inflows are omitted. labor or tax obligations. 16. Using the same model, the regulatory risk 10. Intuitively, the first principal component of measure has less predictive power than other a set of variables is the linear index of all risk ratings, such as the ICRG’s political risk the variables, which captures the largest rating and the EIU’s legal and regulatory pol- amount of information that is common to icy risk rating. This suggests that investor all the variables (Filmer and Pritchett 2001). perception plays an important role, which While different in purpose, results from a the measure presented in this chapter is not PCA in practice often closely approximate well equipped to capture fully. factor analysis, which is often used to 17. For example, political risk ranks second estimate an underlying structure of (a) latent among nine categories of possible impedi- variable(s) (Jolliffe 2002). As such, this ments to FDI, according to the MIGA-EIU aggregation methodology resembles results Political Risk Survey 2013 (MIGA 2013); from the noise extraction approach used in political risk and uncertainty is ranked 15th Kaufmann, Kraay, and Mastruzzi (2011), among 69 organizational risks, according to R e g u l a t o r y R i s k a n d F D I    163 the Aon Global Risk Management Survey sufficient due diligence in carrying out their 2019 (Aon 2019); and political risks and investment. regulatory uncertainty is ranked 4th among 24. In March 2017, the claimants applied for 12 risks, according to the Association for annulment at ICSID. On March 18, 2019, Financial Professionals and Risk Survey the ICSID Annulment Committee dismissed 2019 (AFP 2019). the claimants’ application to annul the 18. For all Doing Business data, see https://www​ award. .doingbusiness.org/en/data. 25. Notably, the tribunal observed that, although 19. This list is generated using the database generally BITs do not contain provisions on of bilateral FDI; see discussion in the the consequences of unlawful conduct by “Regulatory Risk and FDI” section. investors, arbitral decisions have clarified that general principles exist independent of 20. Indonesia’s decision does not immediately specific treaty language. affect all foreign investors. Several of the 26. Law No. 12/2011 on the Development of treaties terminated or being renegotiated Laws and Regulations and its implementing have sunset clauses allowing for continued government regulation 87/204 allow public application of the treaty for a specific period. comment on draft laws and regulations. Indonesia continues to be a party to other mul- 27. See Law No. 23 of 2014 on Regional tilateral treaties—in particular, the treaties of Government. In June 2016, the central gov- the Association of Southeast Asian Nations ernment, exercising its authority under Law (ASEAN)—and foreign investors can avail No. 23 of 2014, repealed more than 3,000 these to seek protection. For the purpose of regional bylaws that were overlapping with this study’s constructed regulatory risk meas- other laws and regulations. However, the ure, not all ASEAN treaties in force could Constitutional Court in its Decision No. be included because they are not mapped in 56/PUU-XIV/2016 limited the central gov- the UNCTAD database. Further, Indonesia’s ernment’s authority to repeal these local national Law Concerning Investment, 2007, ­ regulations and allowed local governments also provides protection guarantees but lim- to appeal the decision. See Butt (2017). ited recourse. 28. Refining the scope includes reform of the 21. For a comprehensive historical account of ISDS process. The EU’s free trade agree- foreign investment and property rights in ments (such as with Canada, Singapore, and Indonesia, see Wells and Rafik (2007). Vietnam) include a standing investment court 22. ICSID Case No. ARB/12/14 and ICSID Case system, which includes an appellate tribunal. No. ARB/12/40. Other FTAs have either excluded ISDS provi- 23. The companies, in collaboration with local sions or diluted its scope, such as in Brazil’s Indonesian companies (Ridlatama com- Cooperation and Facilitation Investment panies), invested in the East Kutai Coal Agreements and the United States–Mexico– Project (EKCP) in the Regent of East Kutai. Canada Agreement. In 2010, the Regent of East Kutai revoked 29. The World Bank Group has been helping the licenses (for activities such as survey and client countries to set up institutional mech- exploration) related to the EKCP, alleging anisms to enable them to better detect and that they were forged. The claimants first resolve investor problems or grievances, filed domestic legal proceedings against the which can potentially escalate in investor-​ revocation, alleging that they had obtained state legal disputes (Echandi and Kher 2014; the licenses lawfully through their partner- World Bank 2019). ship with the local companies. The tribunal 30. For more on the OECD FDI Regulatory ultimately decided in favor of Indonesia, Restrictiveness Index, see https://www.oecd​ stating that the claims were “based on doc- .org/investment/fdiindex.htm. uments forged to implement a fraud aimed 31. Investor-state dispute data from the at obtaining mining rights” and thus were UNCTAD Investment Policy Hub, https://​ inadmissible. The tribunal indicated that investmentpolicy.unctad.org/investment​ the local business partner of the claimants -dispute-settlement. was likely the source of the fraudulent con- 32. Data from “Direct Investment Statistic duct but that the claimants failed to exercise according to the Directional Principle,” 1 6 4    Global Investment Competitiveness Report 2019/2020 External Sector Statistics, National Bank of the amount of performance guarantee and Kazakhstan: https://nationalbank.kz/?docid​ whether it also specifies that the procuring =469&switch=english. entity cannot request more than a certain 33. First, as a single instrument capturing all percentage of the contract value as a bid the most important guarantees for foreign security amount. investors, an investment law may have an 41. A grandfather clause would exempt appli- important signaling effect on the country’s cation of a new law due to conditions that openness to investment and reform. Second, were in place before the new law was imple- it can substantively complement the stand- mented. In a sense, it provides continuity ards of treatment already available under the and predictability for existing investments. country’s existing legal framework. Third, 42. “Proposed regulation” means any draft rule it can serve as an opportunity to reflect, in affecting business activities proposed by a a country’s domestic legislation, its core government’s executive authority, minis- international commitments under its IIAs. try, or regulatory agency that, if fi ­ nalized, Fourth, it can be an opportunity to level the is intended to bind any individuals or playing field between all investors ensur- companies covered by it. This includes ing that all are equally protected. Finally, it subordinated legislation, administrative for- ­ can also allow for unifying a country’s legal malities, decrees, circulars, and directives. and regulatory framework, consolidating a The term also includes rules proposed by the diverse set of legal instruments currently in government that require final approval by the ­ force. parliament, other legislative body, or head of 34. The World Bank Group has been supporting state. See “Transparency of Rulemaking,” client countries to set up institutional mech- Global Indicators of Regulatory Governance, anisms to enable them to better detect and World Bank: https://rulemaking.worldbank​ resolve investor problems or grievances that .org/en/data/comparedata/transparency. can potentially escalate into investor-state 43. This refers to forward regulatory plans— legal disputes (Echandi and Kher 2014; that is, a public list of anticipated regulatory World Bank 2019). changes or proposals intended to be adopted 35. “Accessing Laws and Regulations,” Global or implemented within a specified time frame. Indicators of Regulatory Governance, World 44. See “Transparency in Rulemaking,” Global Bank: https://rulemaking.worldbank.org/en​ Indicators of Regulatory Governance, World /data/comparedata/accessibility. Bank: https://rulemaking.worldbank.org/en​ 36. Countries list sectors and activities that are /­data/comparedata/transparency. prohibited or restricted for FDI in their nega- 45. Expropriation, inability to transfer funds tive list. Sectors and activities not listed on the outside the host country, and instability in negative list are open to FDI. Alternatively, policy and regulatory environment have countries may choose to follow a positive list consistently been identified as critical fac- approach, wherein they list the sectors and tors affecting investors’ decisions to stay and activities that are open to FDI. potentially expand operations in a country 37. “Registering Property”: Transparency of (World Bank 2018). Breach of these stand- Information Index, World Bank Doing Business ards has led to most international inves- indicators: https://www.doingbusiness.org/en​ tor-state disputes (see UNCTAD’s online /data/exploretopics/registering-property. Investment Policy Hub: http://investmentpol- 38. “Getting Electricity”: Communication of Tariffs icyhub.unctad.org/ISDSn). and Tariff Changes Index, World Bank Doing 46. Absolute standards of treatment—such as Business indicators: https://www​ .doingbusiness​ protection from expropriation, and fair and .org/en/data/exploretopics/getting-electricity. equitable treatment (FET)—are to be guar- 39. “Dealing with Construction Permits”: Quality anteed to all investors, irrespective of their of Building Regulations Index, World Bank nationality or other characteristics. On the Doing Business indicators: https​ ://www​ other hand, the “national treatment” stand- .doingbusiness.org/en/data/exploretopics​ ard is a relative standard under which treat- /dealing-with-construction-permits. ment of foreign investors or investments is 40. Other areas that were reviewed were whether assessed relative to the treatment of domestic the law specifies the method to determine investors or investment. R e g u l a t o r y R i s k a n d F D I    165 47. “Direct” expropriation refers to the direct 52. Disputes emerging from commercial trans- seizure or taking of property. “Indirect” actions between enterprises are considered expropriation refers to cases where actions commercial disputes, and those arising from ­ (such as regulatory measures) of the govern- intergovernmental relations are considered ment may be tantamount to or have an effect state-state disputes. Investor-state disputes are equivalent to taking of the property. disputes between foreign investors and host 48. The aspect of the legality of expropriation is states. Such disputes are a relatively unique covered only in investment laws because of feature of international investment law. the lack of availability of comparable data 53. This criterion concerns whether, for exam- based on other legal instruments. ple, investors are allowed access to ISDS 49. “Freely usable” currency means a currency for (a) any disputes relating to invest- determined by the International Monetary ment; (b) only those disputes involving spe- Fund (IMF) under the IMF Articles of cific bases for claims other than the treaty Agreement [Article XXX(f)] to be a currency such as investment contracts and investment that is, in fact, widely used to make payments authorizations; or (c) only those disputes for international transactions and widely involving alleged breach of the treaty. The traded in the principal exchange markets. first case allows investors to submit a very The US dollar, Japanese yen, British pound, broad range of disputes to ISDS, while the euro, and Chinese renminbi are currently latter two cases progressively limit the types determined to be freely usable currencies. of disputes that can be submitted to ISDS. 50. In some cases, investors have argued that 54. This question focuses on the ease of access to the FET standard encompasses the obliga- various recourse mechanisms for ­ enforcement tion to maintain a stable and predictable of investment protection guarantees in a rel- legal framework (Bayindir v. Pakistan, ICSID atively cost-effective and neutral manner. Case No. ARB/03/29; CMS v Argentina, Therefore, although the extensively docu- ICSID Case No. ARB/01/8), while in oth- mented shortcomings of ISDS are well recog- ers, they have argued that the stability of a nized and noted—in particular, on transpar- legal framework is essential to meet inves- ency of the process—this study has not delved tors’ legitimate expectations (Occidental v. into this issue in detail. For further information Ecuador I, LCIA Case No UN3467). A recent on ISDS, see “ICSID Rules and Regulations review of arbitral decisions on this topic indi- Amendment Process” (https://icsid.worldbank​ cates that tribunals have recognized either a .org/en/­amendments); UNCTAD 2019c; and strict or soft regulatory stability obligation “Working Group III: Investor-State Dispute of states under the FET standard. In the first Settlement Reform,” UNCITRAL (https:// case, a mere change in the regulatory frame- uncitral.un.org/en/working_groups/3/investor​ work applicable to investment can trigger a -state). FET violation, while in the other, procedural 55. The 1958 Convention for Recognition and fairness and substantive reasonableness need Enforcement of Foreign Arbitral Awards to be considered to determine whether a FET (referred to as the New York Convention) violation has occurred. Although the obliga- requires the courts of a member state to tion to provide a stable legal and regulatory recognize and enforce an award rendered framework is fairly settled, tribunals have in another member state. It also limits the had mixed views on the scope of the obliga- grounds on which courts of member states tion (Ortino 2018). may refuse recognition and enforcement of 51. In terms of the IIA’s text per se, an “unqual- foreign arbitral awards. Under Article V, the ified” FET provision provides wider pro- following are some of the grounds: incapac- tection because its interpretation is not ity of the parties to the arbitration agree- confined—for example, to specifically enu- ­ ment; invalidity of the arbitration agreement; merated rights or other principles (depending natural justice grounds; arbitral authority or on text of the IIA). The general rules of inter- procedure was not in accordance with the pretation (under Article 31 of the Vienna agreement of the parties; the subject matter Convention on the Law of Treaties or Article of the arbitration cannot be referred to arbi- 38 of the Statute of International Court of tration under the national law of the enforc- Justice) continue to apply. ing country; and contrary to public policy of 1 6 6    Global Investment Competitiveness Report 2019/2020 the enforcing country. These exceptions are considers inclusion of an MFN provision not easy to establish. Therefore, countries and not any specific exceptions regarding can rarely use them, making the New York regional integration agreements, ISDS proce- Convention a fairly effective means of ensur- dural provisions, or phase of application that ing enforcement of awards. On the other may be included in treaties and may change hand, enforcement of foreign court judg- the applicable treatment on a case-by-case ments is available when states have passed a basis. specific law allowing reciprocal enforcement of foreign judgments. 56. Nonmember states can also pursue arbitral References proceedings against host states under ICSID’s AFP (Association for Financial Professionals). Additional Facility Rules, although without 2019. “2019 AFP Risk Survey Report: The the benefit of automatic recognition and Evolving Treasury Ecosystem.” Survey report, enforcement of the arbitral awards. However, AFP, Bethesda, MD. Article 19 of the Additional Facility Rules Alfaro, Laura, Sebnem Kalemli-Ozcan, and Vadym requires that arbitration proceedings con- Volosovych. 2008. “Why Doesn’t Capital Flow ducted under the rules be held only in states from Rich to Poor Countries? An Empirical that are parties to the New York Convention. Investigation.” Review of Economics and Therefore, in these cases, the regime under Statistics 90 (2): 347–68. the New York Convention will apply. Aon. 2019. “Global Risk Management Survey 57. This implies that ICSID awards are generally 2019.” Survey report, Aon, London. not subject to any review process by local Baker, Scott R., Nicholas Bloom, and Steven J. courts in host states and are automatically Davis. 2016. “Measuring Economic Policy enforced. Under Article 53(1) of the ICSID Uncertainty.” Quarterly Journal of Economics Convention, an arbitral award of the tribunal 131 (4): 1593–636. is binding on all parties to the proceeding. In Bauerle Danzman, Sarah. 2016. “Contracting with case of a failure by a party to comply with Whom? The Differential Effects of Investment an award, then under Article 54(1), the other Treaties on FDI.” International Interactions party may seek to have the pecuniary obliga- 42 (3): 452–78. tions recognized and enforced in the courts of Bernanke, Benjamin S. 1983. “Irreversibility, any ICSID member state as though it were a Uncertainty, and Cyclical Investment.” Quarterly final judgment of that state’s courts. Typically, Journal of Economics 98 (1): 85–106. if a party informs the ICSID Secretariat Bloom, Nicholas. 2009. “The Impact of about nonenforcement by another party, the Uncertainty Shocks.” Econometrica 77 (3): Secretariat contacts the noncomplying party 623–85. to request information on the steps taken to Bonnitcha, Jonathan, Lauge N. Skovgaard Poulsen, ensure compliance. See “ICSID Convention and Michael Waibel. 2017. The Political Arbitration Rules,” ICSID Documents, ICSID Economy of the Investment Treaty Regime. website: https​://icsid.worldbank.org/en/Pages​ Oxford: Oxford University Press. /icsiddocs/ICSID​ - Convention-Arbitration​ Bubb, Ryan J., and Susan Rose-Ackerman. 2007. -Rules​.aspx. “BITs and Bargains: Strategic Aspects of 58. “Enforcing Contracts”: Quality of Judicial Bilateral and Multilateral Regulation of Foreign Processes Index, World Bank Doing Business Investment.” International Review of Law and indicators: https://www.doingbusiness.org/en​ Economics 27 (3): 291–311. /data/exploretopics/enforcing-contracts. Butt, Simon. 2017. “Constitutional Court 59. “Registering Property”: Land Dispute Lets Local Governments Off the Leash.” Resolution Index, World Bank Doing Business Indonesia at Melbourne (blog), July 4. indicators: https://www.doingbusiness.org/en​ https​://​indonesiaatmelbourne.unimelb.edu.au​ /data/exploretopics/registering-property. /constitutional-court-lets-local-governments​ 60. Multinationals can also make use of differ- -off-the-leash/. ent treaties through investing from a third Chen, Maggie Xiaoyang, and Michael O. Moore. country. We sidestep this issue because not all 2010. “Location Decision of Heterogeneous investors can take advantage of restructuring Multinational Firms.” Journal of International to the same extent. Further, the scoring only Economics 80 (2): 188–99. R e g u l a t o r y R i s k a n d F D I    167 Colen, Liesbeth, Damiaan Persyn, and Andrea “Firm-Level Political Risk: Measurement and Guariso. 2016. “Bilateral Investment Treaties Effects.” Quarterly Journal of Economics and FDI: Does the Sector Matter?” World 134 (4): 2135–202. Development 83: 193–206. Høyland, Bjørn, Karl Moene, and Fredrik Daude, Christian, and Ernesto Stein. 2007. “The Wi l l u m s e n . 2 0 1 2 . “ T h e Ty r a n n y o f Quality of Institutions and Foreign Direct International Index Rankings.”  Journal of Investment.” Economics and Politics 19 (3): Development Economics 97 (1): 1–14. 317–44. Jolliffe, I. T. 2002. Principal Component Analysis. Dixit, Avinash. 1989. “Entry and Exit Decisions 2nd ed. New York: Springer Verlag. under Uncertainty.” Journal of Political Joyez, Charlie. 2015. “Location Choices and Economy 97 (3): 620–38. Foreign Direct Investment Motives of Dixon, Jay, and Paul Alexander Haslam. 2016. Heterogeneous Firms.” Unpublished paper. “Does the Quality of Investment Protection https://www.etsg​.org/ETSG2015/Papers/144​ Affect FDI Flows to Developing Countries? .pdf. Evidence from Latin America.” The World Jurado, Kyle, Sydney C. Ludvigson, and Serena Economy 39 (8): 1080–108. Ng. 2015. “Measuring Uncertainty.” American Echandi, Roberto, and Priyanka Kher. 2014. Economic Review 105 (3): 1177–216. “Can International Investor-State Disputes Be Kaufmann, Daniel, Aart Kraay, and Massimo Prevented? Empirical Evidence from Settlements Mastruzzi. 2011. “The Worldwide Governance in ICSID Arbitration.” ICSID Review Foreign Indicators: Methodology and Analytical Investment Law Journal 29 (1): 41–65 Issues.”  Hague Journal on the Rule of Law EIU. 2017. “World Risk: Alert—Guide to 3 (2): 220–46. Risk Briefing Methodology.” Risk briefing, Kerner, Andrew, and Jane Lawrence. 2014. ViewsWire online analytical service, EIU, “What’s the Risk? Bilateral Investment August 18. Treaties, Political Risk and Fixed Capital Filmer, Deon, and Lant H. Pritchett. 2001. Accumulation.” British Journal of Political “ E s t i m a t i n g We a l t h E f f e c t s w i t h o u t Science 44 (1): 107–21. Expenditure Data—or Tears: An Application Kobrin, Stephen J. 1979. “Political Risk: A Review to Educational Enrollments in States of and Reconsideration.” Journal of International India.” Demography 38 (1): 115–32. Business Studies 10 (1): 67–80. Frenkel, Michael, and Benedikt Walter. 2019. “Do McFadden, Daniel. 1974. “Conditional Logit Bilateral Investment Treaties Attract Foreign Analysis of Qualitative Choice Behavior.” Direct Investment? The Role of International In Frontiers in Econometrics, edited by Paul Dispute Settlement Provisions.” The World Zarembka. New York: Academic Press. Economy 42 (5): 1316–42. MIGA (Multilateral Investment Guarantee Gelman, Andrew, and Piero Stanig. 2016. Agency). 2013. World Investment and Political “Corruption Perceptions Index 2016: Risk 2013. Washington, DC: World Bank. Technical Methodology Note.” Transparency Montt, Santiago. 2009. State Liability in Investment International, Berlin. Treaty Arbitration: Global Constitutional and Gelos, R. Gaston, and Shang-Jin Wei. 2005. Administrative Law in the BIT Generation. “Transparency and International Portfolio Portland, OR: Hart Publishing. Holdings.” Journal of Finance 60 (6): 2987–3020. OECD (Organisation for Economic Co-operation Giambona, Erasmo, John R. Graham, and and Development). 2008. Handbook Campbell R. Harvey. 2017. “The Management on Constructing Composite Indicators: of Political Risk.” Journal of International Methodology and User Guide. Paris: OECD. Business Studies 48 (4): 523–33. Ortino, Federico. 2018. “The Obligation of Hallward-Driemeier, Mary. 2003. “Do Bilateral Regulatory Stability in the Fair and Equitable Investment Treaties Attract Foreign Direct Treatment Standard.” Journal of International Investment? Only a Bit … and They Could Economic Law 21 (4): 845–65. Bite.” Policy Research Working Paper 3121, Papaioannou, Elias. 2009. “What Drives World Bank, Washington, DC. International Financial Flows? Politics, Hassan, Tarek A., Stephan Hollander, Laurence Institutions and Other Determinants.” Journal van Lent, and Ahmed Tahoun. 2019. of Development Economics 88 (2): 269–81. 1 6 8    Global Investment Competitiveness Report 2019/2020 Republic of Senegal. 2014. “Plan Sénégal US Department of State. 2019a. “2019 Emergent.” Strategy document, Republic of Investment Climate Statements.” Annual Senegal, Dakar. business climate report, US Department Tobin, Jennifer L., and Susan Rose-Ackerman. of State, Washington, DC. https://www​ 2011. “When BITs Have Some Bite: The .state.gov/reports/2019-investment-climate​ Political-Economic Environment for Bilateral -statements/. Investment Treaties.” Review of International US Department of State. 2019b. “2019 Organizations 6 (1): 1–32. Investment Climate Statements: Indonesia.” UNCTAD (United Nations Conference on Trade Annual country statement, US Department and Development). 2019a. “Country Fact of State, Washington, DC. https://www​ Sheet: Indonesia.” Country FDI data sheet, .state.gov​ / reports/2019-investment-climate​ compiled from World Investment Report 2019: -statements/indonesia/ Special Economic Zones. New York and Geneva: US Department of State. 2019c. “2019 Investment United Nations. https://unctad.org/sections​ Climate Statements: Kazakhstan.” Annual /­dite_dir/docs/wir2019/wir19_fs_id_en.pdf. country statement, US Department of State, UNCTAD. 2019b. “Country Fact Sheet: Washington, DC. https://www.state.gov​ Senegal.” Country FDI data sheet, compiled /reports/2019-investment-climate-statements​ from World Investment Report 2019: /kazakhstan/ Special Economic Zones . New York and Wei, Shang Jin. 2000. “How Taxing Is Corruption Geneva: United Nations. https://unctad​ on International Investors?” Review of .org/sections/dite_dir/docs/wir2019​/wir19_fs​ Economics and Statistics 82 (1): 1–11. _sn_en.pdf. Wells, Louis T., and Rafik Ahmed. 2007. Making UNCTAD. 2019c. “Reforming Investment Dispute Foreign Investment Safe: Property Rights Settlement: A Stocktaking.” IIA Issues Note, and National Sovereignty. Oxford: Oxford Issue 1, UNCTAD, Geneva. University Press. UNCTAD. 2019d. World Investment Report Wo r l d B a n k . 2 0 1 8 . G l o b a l I n v e s t m e n t 2019: Special Economic Zones. New York and Competitiveness Report 2017/2018: Foreign Geneva: United Nations. Investor Perspectives and Policy Implications. US Department of State. 2017. “2017 Washington, DC: World Bank. Investment Climate Statements: Indonesia.” World Bank. 2019. “Retention and Expansion Annual country statement, US Department of Foreign Direct Investment: Political of State, Washington, DC. https://www​ Risk and Policy Responses.” Working .state.gov/reports/2017-investment-climate​ paper, Report No. 144312, World Bank, -statements/indonesia/ Washington, DC. 5 Increasing the Development Impact of Investment Promotion Agencies Armando Heilbron and Hania Kronfol Key Findings • Investment promotion agencies (IPAs) can help increase FDI inflows, attract higher quality FDI, and transform the economies of their home countries. Estimates of the magnitude of these effects vary in the literature, including a preliminary cost-benefit analysis indicating that US$1 spent on investment promotion yields US$189 in FDI inflows and that spending a relatively modest US$78 in investment promotion creates one additional job in the pro- moted sectors. • The number of IPAs has proliferated over the past two decades—including at national and subnational levels—and evidence shows they can play a ­ significant role in attracting, retaining, and growing investment. The contributions of IPAs are more pronounced in developing countries, where investors may know less about the location, struggle to obtain reliable information, find the regulatory environments more challenging, and encounter further obstacles stemming from institutional and cultural differences between the inves- tors’ home and host markets. • Many IPAs are struggling to reach their full potential: they are not nimble enough to respond to new market realities; they lack strategic focus; and they do not adequately provide services most valued by investors, such as advocating for improvements in busi- ness climate. Foreign investors appreciate IPA services offered across their investment life cycle—not just during the investment attraction and entry/establishment stages. For example, about two-thirds of surveyed investors consider IPA assistance with business operational issues to be “important” or “critically important.” • IPAs can increase their impact by s ­ harpening their strategic focus, building a coherent insti- tutional framework, and strengthening their delivery of investor services. Specifically, IPAs should (a) focus on a limited number of mandates and target segments; (b) adopt institu- tional features common to private companies; and (c) offer relevant and high-quality inves- tor services across the investment life cycle. These elements should help IPAs to rapidly adapt to sudden changes in the FDI landscape, such as those presented by the COVID-19 pan- demic, and to respond with relevant services to investors. • To maximize IPA’s impact, policy makers should (a) provide high-level government sup- port (from the president or prime minister); (b) foster strong strategic focus and align- ment; (c) grant a clear, uncontested mandate for investment promotion and a high degree of autonomy; (d) facilitate collaboration with ­ governments’ other investment institutions; and (e) provide sufficient and sustained financial resources. Introduction This chapter aims to advance research in the field of investment promotion by Investment promotion agencies (IPAs) are c onsolidating evidence and World Bank ­ national or subnational institutions man- Group operational experience, presenting dated to attract and grow investment—­ recent global data on IPA characteristics and usually foreign direct investment (FDI).1 on investor perceptions about their services, IPAs (also known as investment promotion assessing the evolution of IPAs over more intermediaries, IPIs) can play a significant ­ than a decade, and recommending new frame- role in fostering economic development in works for these agencies to improve their their countries. They can generate larger development impact. FDI inflows, attract quality FDI, deepen Bringing together different data sources ­ integration into global value chains (GVCs), and country experiences, the chapter and even transform the economies of their addresses the role that IPAs can play in home countries. f ostering FDI inflows and development ­ However, although IPAs have proliferated impact, examines the challenges they face, over the past two decades, their success s­ tories and presents recommendations to improve are not as widespread and are especially their effectiveness. To do so, the sections are scarce in developing countries.2 Many IPAs organized as follows: “Do IPAs Foster are stretched across many mandates and Development Impact?” draws on a literature target more sectors than they can properly ­ review to discuss the potential contribution handle, while not providing the key services that IPAs can make to economic development that investors expect. At the same time, IPAs and to identify agency characteristics that do not seem to be evolving as dynamically as contribute best to achieving this impact. “Are needed to align with changes in the FDI IPAs Delivering Good Value to Investors? l andscape as well as more sophisticated ­ What Challenges Do They Face?” analyzes investor requirements. The current literature, ­ results from the 2019 Global Investment combined with additional research and more Competitiveness (GIC) Survey of multina- than 30 years of World Bank Group opera- tional enterprises (MNEs) to discuss the tional experience, suggests that IPAs can have investment landscape and what investors much greater impact if they refine their strate- value from IPAs. The ­ section then compares gic focus, adopt institutional features that pri- IPA characteristic s ­ urveys over time—from oritize the investor, and improve ­ investor 2005 to 2017/18—to assess the extent to service delivery throughout the investment life which IPAs have evolved in line with FDI cycle (box 5.1). trends and investor needs. “What Can Given that IPAs are mostly financed by Governments Do to Improve the Development public funds, a reasonable question to ask is Impact of Their IPAs?” c ­ onsolidates key find- whether they provide a good return on such ings, insights from World Bank Group opera- funds. Do they work for development? And if tional experience, and examples of country so, what can policy makers do to maximize experiences to ­ provide a framework and pol- their impact? Currently, research on IPAs and icy guidelines for governments to improve the their impact—primarily considered in terms impact of their IPAs, focusing on three core of FDI inflows generated and direct jobs strategic alignment and focus, coherent areas: ­ ­ created (see annex 5A)—is highly dispersed institutional framework, and strong delivery ­ across different academic fields and has of investor services. “Conclusions and Future yielded few clear findings on the role of IPAs. Research” summarizes high-level policies gov- It remains limited in addressing global trends ernments may consider to support their IPAs and themes, particularly in the context of and increase their impact, and suggests areas developing countries. for further research. 171 1 7 2    Global Investment Competitiveness Report 2019/2020 BOX 5.1 The Investment Life Cycle The investment life cycle refers primarily to the investor’s the World Bank Group proposes that a host location journey from project p ­ lanning to site exploration and its investment promotion agency (IPA) mirror the and selection; i ­nvestment entry and establishment; cycle with a coherent investment policy and promotion operation; e ­ ­ xpansion, diversification, and linkages; offering that can be summarized across four key stages: and finally to transition or exit. To successfully attract attraction; entry and establishment; retention and foreign direct investment (FDI) and foster its growth, expansion; and linkages and spillovers (figure B5.1.1). FIGURE B5.1.1  The Investment Life Cycle from Investor and Host Country Perspectives PLAN TRANSITION EXPLORE Linkages and Attraction Spillovers EXPAND VALIDATE DIVERSIFY IPAs SELECTED LINK LOCATION Retention and Entry and Expansion Establishment OPERATE ESTABLISH Investor’s project cycle Host country cycle Source: World Bank. Note: IPAs = investment promotion agencies. Do IPAs Foster Development taking a stronger interest in the field of investment promotion around 20 years ago, Impact? when the World Bank Group’s Foreign Economists, researchers, economic develop- ­ ublished Investment Advisory Service (FIAS) p ment specialists, and policy makers started “Marketing a Country: Promotion as a Tool I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    173 for Attracting Foreign Investment” (Wells 2009; Spar 1998).8 The ­ attraction of Intel gen- and Wint 2000).3 IPAs have proliferated since erated a strong ­ signaling effect that helped then (Harding and Javorcik 2011). Between boost FDI inflows, diversify exports from 2002 and 2018, the number of national and mostly fruit c ­ ommodities to advanced manu- subnational IPAs registered in the World facturing, encourage deeper integration into Association of Investment Promotion GVCs, and upgrade to higher-value activities. Agencies (WAIPA) grew from 112 to 170.4 IPAs can also foster economic transforma- The 2000 FIAS publication and other tion (Freund and Moran 2017) and help link empirical evidence have shown that IPAs can economies to GVCs through FDI (World help generate larger FDI inflows, attract Bank 2020). Developing country examples quality FDI, deepen connections to GVCs, and ­ include Costa Rica, Malaysia, and Morocco, transform economies.5 Several studies indicate where policies supporting m ­ acroeconomic that IPAs increase FDI inflows to their home stability, skills development, and strong IPAs economies (Cho 2003; Crescenzi, Di Cataldo, contributed to attracting a few large, effi- and Giua 2019a; Morisset and Andrews- ciency-seeking MNEs, 9 which in turn Johnson 2004; Pietersen and Bezuidenhout boosted the countries’ revealed ­ comparative 2015). 6 Some estimate the magnitude of advantage (Freund and Moran 2017) and this FDI increase to be 29.7–45.3 percent their integration into GVCs. (Morisset and A ­ ndrews-Johnson 2004; Wells and Wint 2000). Another study finds 155 Strategic Focus Matters ­ percent higher FDI inflows and 68 percent greater employment in targeted sectors versus IPAs are more likely to succeed when they ­ nontargeted sectors (Harding and Javorcik focus strategically on promoting specific 2011). It also reveals a preliminary cost-­ sectors or business activities (Crescenzi, Di benefit analysis: US$1 spent on investment Cataldo, and Giua 2019a; ECORYS 2013; promotion yields US$189 in FDI inflows and Loewendahl 2001; Miškinis and Byrka 2014). a relatively modest US$78 spent on invest- (See box 5.2.) One study focusing on OECD ment promotion creates one additional job in countries finds that IPAs targeting industries the promoted sectors. increased FDI inflows into those targeted The contribution of IPAs is more pro- industries by 41 percent (Charlton and Davis nounced in developing countries, where 2007). Good examples of countries engaging (a) investors may know less about the l ­ocation in targeted promotion are Brazil, Chile, and and struggle to obtain good data (because of Costa Rica, which developed well-targeted, information asymmetries); (b) regulatory envi- responsive, and sustained strategies that ronments are more challenging (as reflected by attracted nontraditional FDI (Nelson 2005). poor ratings based on the World Bank Group’s All IPAs that belong to member countries of Doing Business indicators); and (c) cultural the Organization for Economic Co-operation distance from the United States is greater and Development (OECD), with the e ­ xception (Harding and Javorcik 2011, 2012).7 of France, prioritize sectors (OECD 2018). IPAs can also bolster the quality of FDI that Mandate clustering—adding or merging comes into their economies (Moran et al. other policy or economic development 2018), including knowledge-intensive FDI ­ f unctions to the IPA—is a controversial (Crescenzi, Di Cataldo, and Giua 2019b; topic in the investment promotion literature. Monge-González and Tacsir 2014). For exam- Some suggest that IPAs focus exclusively on ple, research points to the essential role of the investment promotion, while others recom- Costa Rican Investment Promotion Agency mend that economic development agencies (CINDE) in attracting Intel in 1996 and pursue multiple mandates, including invest- this firm’s overwhelmingly positive impact on ment promotion. For example, merging the country’s economic d ­ evelopment investment and trade promotion may hinder (MIGA 2006a; Nelson 1999, 2000, 2005, promotion results if done only as a 1 7 4    Global Investment Competitiveness Report 2019/2020 budget-cutting measure. Some countries have and services skills (Nelson 2009). Without achieved synergies by merging mandates the right staff, IPAs underperform because under a GVC-focused strategy. Other areas attracting of the highly competitive nature of ­ that could be merged are administration, FDI—which differs significantly from the research, image ­ building, and foreign offices role of typical government agencies. (UNCTAD 2013). Recent research, however, Certain institutional characteristics seem finds a negative association between FDI to be linked to better IPA performance, results and combining mandates of investment especially in the developing world: ­ ­ promotion with trade or with outward invest- ment (Lim 2018). Adding regulatory and • Strong support from the topmost levels of other responsibilities to an IPA can p ­ revent the government, sometimes linked to a high agency from focusing on the already demand- institutional status, hierarchy, or attach- ing role of catering to i­nvestors, as shown in ment to upper ministry levels (Lim 2018; figure 5.1 (Whyte, Ortega, and Griffin 2011). Morisset and Andrews-Johnson 2004; Volpe Martincus and Sztajerowska 2019) • Autonomy and operational ­ independence IPAs with Private Sector-Like (ECORYS 2013; Lim 2018; Loewendahl Institutional Characteristics are 2001; Nelson 2009; UNCTAD 1997; More Successful Wells and Wint 2000), which allow IPAs Most national IPAs are purely public bodies to receive consistent support even d ­ uring (around 80 percent, according to the 2017 periods of political transition, attain World Bank Group Global IPA Survey). 10 b etter understanding of investor needs, ­ While this may help with country branding and work more effectively alongside and advocacy, public bodies may need to private sector actors (Bauerle Danzman ­ operate under civil service rules. Traditional and Gertz, forthcoming)11 civil service recruitment and pay policies • Sufficient and sustained financial resources typically hamper an IPA’s potential to recruit ­ over periods of three years or longer, given qualified, specialized staff with the required the long cycle of i ­nvestment promotion private sector background, international (Morisset and Andrews-­ Johnson 2004; exposure, language skills, and marketing Volpe Martincus and Sztajerowska 2019) FIGURE 5.1  The Best IPAs Tend to Be Dedicated Promoters Best practice IPAs 100 Share of surveyed IPAs (%) Good IPAs 94 6 Average IPAs 57 43 Weak IPAs 32 68 Very weak IPAs 50 50 0 20 40 60 80 100 Dedicated promoters Promoter-regulators Source: Whyte, Ortega, and Griffin 2011. Note: The level of investment promotion agency (IPA) performance was proxied by the score in the World Bank Group’s “Global Investment Promotion Benchmarking 2009” report. A parallel internal World Bank Group survey of IPAs, deployed in 2009, provided the information as to whether an IPA was a dedicated promoter or promoter-regulator. I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    175 • Management and staff with ­ private sector staffed by government officials who tend to experience to develop i ­nvestor-​m inded, focus more on either carrying out investment service-oriented, and c ­ onsultancy-like events or providing a regulatory service, some- organizations (ECORYS 2013; ­ N elson times in an OSS. IPAs need staff with private 2009; UNCTAD 1997; Wells and Wint sector experience, service skills, and deep busi- ­ ector culture (Ortega 2000) with a private s ness knowledge (including understanding of and Griffin 2009) and a t ­ransnational investor needs, motivations, challenges, and learning capacity that better understands concerns but also sector terminology and the private sector and ­anticipates investor trends) as well as internal systems such as needs (Nelson 2009)12 ­ customer relationship management (CRM) • Staff with international exposure and systems to improve service and FDI perfor- an egalitarian and democratic organi- mance (Ortega and Griffin 2009). zational culture to foster transnational For a long time, IPAs have been consid- learning capacity (Nelson 2009) ered intermediaries between investors and • Private sector board representatives policy makers (Crescenzi 2018), possessing (ECORYS 2013; Miškinis and Byrka firsthand investor feedback that can be 2014) extremely ­useful to influence policy making. • S t rong in stit ution al c oll abo ratio n While 76 percent of IPAs include policy (Miškinis and Byrka 2014) advocacy as part of their mandates, • Longer IPA experience, more staff, and 35 ­percent recognize that they fail to engage greater overseas presence (Anderson in that activity (WAIPA 2019). Moreover, and Sutherland 2015; Lim 2018; Volpe according to the World Bank’s 2017 GIC Martincus and Sztajerowska 2019) ­ Survey,13 IPA services were most appreciated • Use of digitalization and emerging tech- by investors during the establishment, reten- nologies to reach target investors more tion, and expansion stages, including hands- efficiently (DCI 2017; WAIPA 2019). on assistance with issues during registration, setting up the business, and operation as Quality Services Are Linked to FDI well as advocacy to improve the business Performance environment (World Bank 2018). The quality of IPA information delivery is linked to FDI performance. As shown in Are IPAs Delivering Good Value ­ figure 5.2, an IPA’s score in the World Bank Group’s Global Investment Promotion Best to Investors? What Challenges Practices (GIPB) 2012 report on informa- Do They Face? tion ­services (such as website information and inquiry handling) is positively correlated Investors Value IPA Services across the with FDI inflows, and a one-unit increase in Investment Life Cycle the GIPB score is shown to be associated Insight on which investment promotion activ- with a 1.5 percent increase in FDI inflows ities matter most, and to whom, can be drawn (Harding and Javorcik 2012). from a review of investor responses from the IPA activities have been traditionally 2019 GIC Survey.14 The results reveal that ­ clustered under promotion, marketing, or tar- MNEs generally find IPA services valuable. geting; facilitation; servicing; one-stop-shop More than 60 percent of surveyed investors (OSS); and aftercare programs (FIAS 2011; consider at least one IPA service to be Loewendahl 2001; MIGA 2001; Ortega and “­critically important” to their operations, and Griffin 2009; UNCTAD 2007). However, nearly 90 percent consider at least one service many IPAs lack a service orientation or inves- to be “important” or “critically important.” tor service continuity throughout the invest- When looking more closely at the types ment life cycle, given that most IPAs are of services valued, foreign investors 1 7 6    Global Investment Competitiveness Report 2019/2020 FIGURE 5.2  Better-Quality IPA Services Attract More FDI Inflows 30 USA LUX GBR 25 CHN FRA DEU NLD ESP HKG CAN RUS MEX CHE BRA AUS SWE IND ITA Average FDI inflow, 2000–10 (log) SAU IRL HUN AUT POL SGP CHL JPN CZE TUR KAZ UKR THA MYS ISR DNK VNM EGY ROM COL PRT FIN LBN NGA ARG SVK KOR IDN ZAF BGR SRB GRC DZA PAK TUN PHL PER CYP SDN MAR HRV JOR EST ISL NZL LBY PAN DOM OMN MNE TTO SVN CRI AGO IRQ BLR AZE GEO VENGHAGTM BWABHR URY LTU COG SYR JAM LVA 20 MOZ ZMB TZA BHS ALBECU BIH MKD BOL YEM KHM BGD LKA HND SLV ARM UGA KSV NIC CMR GIN LBR ETH ATG BRN MDA TJK MNG KEN BRB FJI MUS PRY GABKGZ NAM MLI LAO PNG KNA SYC SEN LCA GRD BEN LSO BFA MDV MDG BLZ NER TGO DJI RWA WBG GUY VCT KWT IRN ERI BTN HTI SLE SLB SWZ DMA GMB VUT MWI TON WSM GNB NPL 15 BDI 10 20 30 40 50 60 70 80 90 100 IPA quality (score) Source: Harding and Javorcik 2012. Note: The dependent variable is the average FDI inflow received by each country during the 2000–10 period as reported in the IMF’s International Financial Statistics and expressed in the log form. Investment promotion agency (IPA) service quality is measured based on ratings on a scale from 0 to 100 from “Global Investment Promotion Best Practices 2012” (World Bank 2012) which measured the quality of each IPA’s inquiry handling and website. Country and economy labels are International Organization for Standardization (ISO) codes. FDI = foreign direct investment. indicate that IPA services offered across the important,” the highest percent of such investment life cycle—not just during the responses among IPA services. investment attraction and establishment Certain groups of investors value IPA stages—​ a re important to their business ­ services more than others (figure 5.4). IPA decisions ­(figure 5.3). About two-thirds or ­ s ervices are considered more important more of s ­ urveyed investors consider each to investors from developing countries. IPA service (including postinvestment ser- For example, about 33 percent of investors vices such as assistance with operational from developing countries report that prein- issues) to be “important” or “critically vestment assistance (such as site visits and brief- important.” The results also highlight the ings) is “critically important,” compared with importance of IPAs’ role in advocacy: 24 percent of investors from developed coun- 35 percent of the surveyed investors con- tries. Relative to investors from developed sider IPAs’ efforts to improve countries’ countries, investors from low- and middle-­ business environments to be “­ c ritically income countries may lack the financial I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    177 FIGURE 5.3  IPA Services across the Investment Life Cycle Are Important to Investors Question: How important are the following services offered by investment promotion agencies to your company in this country? Efforts to improve the business environment in the country 35 40 Preinvestment information 30 41 Assistance in setting up 31 39 Assistance with operational issues and grievances 29 38 Location marketing 29 37 Preinvestment assistance 25 40 Personalized contact or response to your company 24 40 0 20 40 60 80 100 Share of respondents (%) Critically important Important Source: Computation based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. IPA = investment promotion agency. FIGURE 5.4  Investors from Developing Countries Value IPA Services More than Those from High-Income Countries Question: How important are the following services offered by investment promotion agencies (IPAs) to your company in this country? Efforts to improve the business environment in the country*: HIC 33 41 MIC/LIC 46 36 Preinvestment information: HIC 29 41 MIC/LIC 39 36 Assistance in setting up: HIC 29 39 MIC/LIC 39 37 Assistance with operational issues and grievances: HIC 28 38 MIC/LIC 37 37 Location marketing: HIC 27 38 MIC/LIC 42 33 Preinvestment assistance*: HIC 24 39 MIC/LIC 33 43 Personalized contact or response to your company: HIC 24 40 MIC/LIC 28 42 0 20 40 60 80 100 Share of respondents (%) Critically important Important Source: Computation based on the 2019 GIC Survey. Note: Affiliates of multinational enterprises were surveyed in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. All country income categories use World Bank-defined classifications; “developing” countries refers collectively to all low- and middle- income countries. HIC = high-income country; LIC = low-income country; MIC = middle-income country. Statistical significance denoted by asterisk: *p < .10 in an ordered logistic regression incorporating controls for sector, a dummy variable for exports ­comprising over 50 percent of revenues, sector-export interactions, import share of inputs, sector-import interactions, source country income group, a dummy for employment over 250 employees, a dummy for investment stock over US$10 million, number of years in country, percentage foreign ownership, and country fixed effects. 1 7 8    Global Investment Competitiveness Report 2019/2020 capacity or connections to hire external IPAs are not advancing on several practices consultants to help scope and select potential ­ associated with better performance: investment location sites. They are also less • Only 55 percent of IPAs in 2017/18 likely to have access to binational chambers of compared with 53 percent in 2004) have (­ commerce in host countries. Thus, they are an investor tracking system more likely to rely on IPAs for preinvestment • Only 64 percent of IPAs are targeting assistance. speci f ic cou nt ries (compa red w it h In addition, IPA services that focus on early percent in 2004) 61 ­ stages of the investment life cycle are more • Few are reporting to higher levels of important to new investors with less in-­ government (12 percent in 2005 versus ­ country experience. Of investors with 0–10 16 percent in 2017/18) years of experience in surveyed middle-income • Despite significant growth in the num- countries, 27 percent consider p ­ reinvestment ber of overall staff, the growth of staff assistance such as site visits and briefings to be engaged in investment promotion has not “critically important,” versus 23 percent of increased as notably (suggesting that staff investors with more than 30 years of experi- are engaging in other mandates). ence in a market. Similarly, 35 percent of investors with less than 10 years of experience At the same time, IPAs are expanding, consider business setup ­ assistance (such as ­ ignaling their increasing prominence as ­ s public help with entry permits) to be “critically agencies. IPAs’ FDI and investment promotion important,” compared with 30 percent of budgets have grown since 2005. In 2005, investors with more than 30 years ­ of experi- more than 50 percent of agencies had budgets ence. In contrast, services like assistance with up to US$500,000, whereas in 2018, most of grievances and operational issues are valued them had budgets up to US$1 ­ m illion. 15 by newer and older investors alike. Alongside the growth in budgets, the average number of agency staff increased from 208 in 2005 to 337 in 2017/18. The average number IPAs May Not Be Evolving as of offices abroad also increased from 11 to 18. Dynamically as Needed Simultaneously, the role of subnational IPAs Despite important changes in the FDI has also been growing, with larger budgets landscape, IPAs have evolved surprisingly ­ and expanded mandates. little over the past 12–13 years. Comparing Encouragingly, national IPAs are ­engaging IPA characteristics surveys from 2005 to in more proactive, research-based sector 2017/18 (see annex 5B for summary of data ­ targeting (figure 5.5). When asked whether sources and methodology), a few shifts are the agency engages in proactive investor observed, but many IPA features remain the ­ t argeting, all IPAs in 2017/18 responded same (see annex 5B, table 5B.1). This sug- ­ affirmatively, compared with 77 percent in gests that IPAs may not be as dynamic or 2005. Similarly, all IPAs in 2017/18 reported responsive as needed, especially in the con- undertaking sectoral or market research, text of a more challenging FDI climate and compared with 74 percent in 2005. Nearly broader ­ economic shifts such as a heavier all IPAs (93 percent) reported targeting reliance on automation, the growth in s pecific sectors in 2017/18, whereas only ­ GVCs, and the rise of the services sector. 77 percent did so in 2005. Such limitations are particularly concerning However, IPAs are still targeting more-­ in light of the serious and unprecedented traditional economic sectors, mostly in the challenges posed by the novel 2019 corona- ­ p rimary sector (raw materials) without virus disease (COVID-19) pandemic; IPAs reflecting the global FDI emergence in the need to be able to adapt and provide the tertiary ­ s ector (services). Comparing the necessary support to investors and other ­ global percentage growth in 2009–18 FDI government agencies during crises. projects to percentage point increases in sector I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    179 FIGURE 5.5  IPAs Are Becoming More Proactive and Focusing More on Sector Targeting 100 Does the IPA undertake proactive investor targeting? (n = 39) *** 77 93 Does the IPA target specific sectors? (n = 43) ** 77 100 Does the IPA undertake sectoral or market research? (n = 38) *** 74 0 20 40 60 80 100 Share of surveyed IPAs (%) 2017/18 2005 Sources: 2005 World Bank Group IPA Census and 2017 World Bank Group Global IPA Survey; 2017 and 2018 World Association of Investment Promotion Agencies (WAIPA) annual surveys; OECD 2018. Note: For details about the data sources and methodology, see annex 5B. IPA = investment promotion agency; n = sample size (number of IPAs). Because the data are sourced from samples, the statistical significance of the increase over time is presented: *p < .10, **p < .05, ***p < .01 ­ argeting by IPAs,16 a m t ­ isalignment appears Group’s Global Investment Promotion Best between the sectors that IPAs are prioritizing Practice (GIPB) reports from 2006 to and the global trends in the FDI landscape 2012—IPAs have responded poorly to inqui- (figure 5.6). ries, a basic information service to inves- This trend suggests that IPAs may not be tors.17 Service delivery ratings for inquiry engaging in an evidence-based approach to handling were quite low in 2006, did not their proactive promotion efforts—focusing improve in 2009, and surprisingly declined on historical priorities rather than targeting in 2012, the year of the last GIPB report those segments with higher traction—or very (figure 5.7). Another basic service IPAs need likely, they could be under pressure from their to provide is information available through governments and local interest groups to pro- their websites. Between 2009 and 2012, the mote the primary sector regardless. The average website score remained relatively growth in IPA sector targeting in the primary low and stagnant, only reaching 61 percent sector from 2005 to 2017/18 also suggests a in 2012. ­ continued prioritization of natural-resource- In addition, as of 2012, only 3 percent of seeking FDI, which means that IPAs may be all national IPAs provided good-practice missing opportunities to support their home inquiry handling.18 Of 181 IPAs included in countries in leveraging efficiency-seeking FDI the GIPB report, 165 had a website, but only for ­development impact, especially given its 102 provided an email address, and only 53 potential role in GVC integration and eco- replied to investor inquiries in a reasonable, nomic upgrading. business-like time frame. Only 24 provided Nor have IPAs evolved much in their an adequate response, and only 6 followed targeting of source countries for FDI despite ­ up afterward to help convert interest to the global shift toward increased investment investment. flows from developing countries to other developing countries. Although IPAs should Common Challenges Prevent IPAs be consistent in their targeting efforts, they from Performing Better also need to adapt and revisit their strategies every year. Uneven Investor Service Coverage Looking at selected performance indica- As part of their goal to harness more and tors over time—drawing on the World Bank locations, IPAs better investment for their ­ 1 8 0    Global Investment Competitiveness Report 2019/2020 FIGURE 5.6  IPA Sector Targeting Maintains a Traditional Focus on the Primary Sector, Indicating a Misalignment with Evolving Global FDI Trends 30 26 60 Increase in IPA sector targeting, 2005–18 (ppt) Global FDI project growth, 2009–18 (%) 20 15 40 27 18 10 20 9 0 0 –10 –20 –20 –40 –30 –60 –31 –40 –80 Primary sector*** Secondary sector Tertiary sector* Global FDI project growth (%) IPA sector targeting increase (ppt) Sources: FDI project data from fDi Markets, a Financial Times dataset (https://www.fdimarkets.com/); sector targeting data from the 2005 World Bank Group IPA Census and a combination of 2017/18 IPA surveys from the World Bank Group, World Association of Investment Promotion Agencies (WAIPA), and Organisation for Economic Co-operation and Development (OECD). (For more information, see annex 5B). Note: Because of open responses to sector targeting in later investment promotion agency (IPA) surveys, the sector groups have been defined by the authors to allow concordance between fDi Markets data (comparing total global projects by sector over time) and the IPA surveys. The comparison date range uses the longest reliable information available in each of the data sources, so foreign direct investment (FDI) projects start in 2009 and sector targeting in 2005. For IPA sector targeting, the sample size of national IPAs varies by sector depending on available comparison data over time (51 IPAs for the primary sector; 45 for the secondary sector; and 44 for the tertiary sector). The number of FDI projects is a global figure. “Primary sector” refers to raw materials, “secondary sector” to manufacturing, and “tertiary sector” to services. ppt = percentage point(s). ­ urveys are samples, changes in targeting are presented with statistical significance (*p < .10, **p < .05, and ***p < .01) Significance levels: Given that the IPA s and represent percentage point changes in the fraction of IPAs targeting that specific sector. (The authors believe this is more informative than a percent- age change, given that a 1–2 percent change would otherwise represent a 100 percent increase in targeting.) FIGURE 5.7  IPA Inquiry Handling Deteriorated in 2012 from an Already Poor Level in 2006 70 60 61 58 50 GIPB score, average (%) 48 40 30 28 28 22 20 10 0 2006 2009 2012 Website Inquiry handling Sources: World Bank, using data from MIGA 2006b and the Global Investment Promotion Best Practices/Benchmark (GIPB) studies (World Bank 2009, 2012). Note: Average scores for website and inquiry handling were compiled by using the three rounds of the World Bank Group’s GIPB reports. Highest possible score = 100 percent. IPA = investment promotion agency. I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    181 are supposed to provide quality services to ­ pecificity when designating varying levels of s investors throughout the investment life their priority sectors—with some identifying cycle (box 5.1.). However, the 2017 World broad ­ e conomic categories (for example, Bank Group Global IPA Survey finds that manufacturing) and others, mostly in more ­ most IPAs are providing services pre­ advanced economies, selecting specific dominantly at the attraction stage, with segments. (See box 5.2 on the need for more ­ decreasing coverage starting at investment precise targeting.) World Bank Group entry and establishment, followed by the experience in developing countries suggests ­ retention and expansion stage. Services to that IPAs working with more than five foster linkages between MNEs and domestic priority sectors or segments may be diluting ­ firms are rarely covered by IPAs (table 5.1). their scarce resources, making it more d­ ifficult This distribution of effort reflects the focus to provide high-quality services demanded by of IPAs on the attraction stage but leaves investors, thus achieving weaker results. important gaps in terms of investor needs. ­ The World Bank Group is piloting detailed Wide Range of Mandates investor service satisfaction surveys in client Many IPAs are taking on too many mandates countries. Preliminary findings show inves- that cover a wide range of functions, requir- tors attach a high relevance to investment ing very different skill sets (figure 5.9). The assistance, advocacy, and information ser- 2017 World Bank Group Global IPA Survey vices provided by IPAs, but at the same time, finds that the average number of mandates investors largely do not give good satisfaction is 7.8 globally, with developing-country ratings in those three service categories. IPAs pursuing more (8.1, on average) than Poorly s ­erviced investors may decide to developed-country IPAs (7.0, on average). ­ discontinue investing or divest. In contrast, ­ Developing countries are more likely to be well-serviced investors may decide to ­ reinvest, mandated to ­ support domestic direct invest- either to expand the current activity or to ment (DDI), negotiate investment agreements, diversify into new activities. issue licenses, promote exports, negotiate public concessions, and administer public- Inadequate Sector Prioritization and Focus private partnerships (PPPs), in addition to Based on the 2017 World Bank Group Global providing core investment promotion IPA Survey, the vast majority of IPAs (84 per- functions. cent) list five or more “priority” sectors for Pairing data from the 2017 World Bank investment promotion; 44 percent have more Group Global IPA Survey with FDI figures than eight (figure 5.8).19 Moreover, IPAs show reveals a strong negative association between TABLE 5.1  IPAs Self-Report Providing Services Mostly at the Attraction Stage, with Service Delivery Rapidly Declining in Later Stages of the Investment Life Cycle Share of respondents (percent) Entry and Retention and Linkages and Service type Attraction establishment expansion spillovers Marketing 92 52 39 20 Information 94 61 49 23 Assistance 78 78 63 29 Advocacy 93 53 36 22 Source: Adaptation of Heilbron and Aranda-Larrey 2020. Note: The figures presented in this table are computations based on the 2017 World Bank Group Global IPA survey. The survey received responses from 83 national IPAs globally. (For more information, see annex 5B.) Stages are drawn from the World Bank Group’s investment policy and promotion life cycle (attraction, entry and establishment, retention and expansion, and linkages and spillovers). IPAs were asked which specific services they provided to ­ romotion agencies. ­investors. Their responses were grouped across the four investment life cycle stages and four service categories. IPAs = investment p 1 8 2    Global Investment Competitiveness Report 2019/2020 FIGURE 5.8  Eighty-Four Percent of IPAs Have Five or More “Priority” Sectors Share of IPAs claiming to have the indicated number of strategic sectors (percent) 12% 16% 32% 40% 1–4 sectors 5–7 sectors 8–10 sectors >10 sectors Source: World Bank Group 2017 Global IPA Survey. Note: The survey received responses from 83 national IPAs globally. (For more information, see annex 5B.) IPAs = investment promotion agencies. BOX 5.2 From Sectors to Segments: Making IPA Targeting More Precise Most sophisticated investment promotion agencies For exa mple , t he C ost a R ica n I nvest ment (IPAs) have evolved from ­ promotion during the 1980s Promotion Agency (CINDE) has been p ­ romoting of broad economic sectors—such as the primary s ­ ector segments such as the assembly and ­ s terilization of (natural resources) or the s­ econdary sector (manufac- therapeutic devices. IDA Ireland focuses its promotion turing)—to more specific industries within sectors. efforts by using a matrix of s ­ ectors and activities—for This evolution mirrored the advent in the 1990s of example, research and development (R&D) within the ­ better, more detailed data such as the Standard Industry ­ ­ ciences sector. life s Classifications and commercial databases. Many policy makers and practitioners in the field In the 2000s, the focus has sharpened toward often still refer generically to sectors, especially in the subindustries and specific business activities within developing world. Refinements toward more precise industries or global value chains (GVCs), such as identification of target segments follow good-practice “assembly of electronic components,” that could be techniques for marketing segmentation and can effec- clustered under an “advanced manufacturing” sector. tively improve IPA impact. the number of IPA mandates and FDI inflows development. For instance, the Rwanda in developing countries (figure 5.10). This Development Board, which has multiple suggests that the more mandates IPAs in mandates, has achieved important results developing countries cover, the more chal- but has taken many years to ramp up FDI lenges they may face in attracting FDI. inflows.20 The same association for devel- Findings from World Bank Group opera- oped-country IPAs does not seem to exhibit tional experience reflect this trend, espe- a significant statistical correlation, suggest- cially at the early stages of an IPA’s ing that the relationship between the I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    183 FIGURE 5.9  Most IPAs Cover a Wider Range of Mandates than Just FDI Promotion, Especially in Developing Countries Question: Which of the following functions are included in your IPA’s mandate? Foreign investment promotion Advocacy Domestic firms support* Other Regulatory Export promotion 0 20 40 60 80 100 Share of respondents (%) Developed countries Developing countries Source: Computation based on 2017 World Bank Group Global IPA Survey. Note: The survey received responses from 83 national IPAs globally. (For more information, see annex 5B) “Advocacy” includes mandates related to policy advocacy as well as investment climate reform. “Domestic firms support” includes mandates related to domestic investment promotion and development of small and medium enterprises and local suppliers, as well as matchmaking between foreign investors and local suppliers. “Regulatory” includes mandates related to incentive administration and investment project screening or approval, as well as issuing other licenses or permits. “Other” mandates include outward investment support, administration of public-private partnerships, special economic zones, industrial parks, or other state assets, as well as negotiation or administration of investment agreements or public concessions. All country income categories use World Bank-defined classifications; “developing countries” refers collectively to all low- and middle-income countries and “developed countries” to high-income countries. FDI = foreign direct investment; IPAs = investment promotion agencies. Significance levels:*p < .10, **p < .05, ***p < .01. number of mandates and FDI inflows is not pronounced for IPAs in developing countries, clear. IPAs may be able to adopt more man- which also carry wider mandates, in turn dates without curtailing FDI as they mature. putting more pressure on scarcer resources. Resource Constraints Inadequate Institutional Coordination Despite growing budgets and staffing over Institutional coordination and partnerships time, the 2017 World Bank Group Global are critical for IPAs to effectively service IPA Survey reveals that the biggest chal- investors. Based on the 2017 World Bank lenges facing IPAs in developing countries Group Global IPA Survey, 77 p ­ ercent of are related to financial resources (52 percent respondent national IPAs maintain close or of respondents c ­ iting it among their top regular contact with their subnational agen- three challenges), g ­ overnment support and cies (box 5.3). public policies (49 percent), and human Nearly all IPAs use coordination mecha- resources capacity (44 percent), as shown in nisms with other entities such as memoran- ­figure  5.11. dums of understanding (MoUs) or regular Developed-country IPAs indicate that they inter-agency meetings. Nevertheless, IPAs are most concerned with the economic envi- still face significant obstacles in their insti- ronment (57 percent), f ­ollowed by processes tutional coordination efforts: 65 percent and bureaucracy (43 percent), and then by cite having unresponsive partner entities, financial resources (33 percent). The chal- and 64 percent cite the absence of mandate lenge pertaining to limited resources is more or power to ensure effective cooperation. 1 8 4    Global Investment Competitiveness Report 2019/2020 FIGURE 5.10  IPAs in Developing Countries That Have More Mandates Also Have Lower FDI Inflows CHE 4 IPL BRA OYP TUR ARG AZE 2 VNM RUS CRY Relative log net FDI inflow ROU ESP URY SRB GEO MMR PHL ETH DMI WAM ISL COD DNK ALB JRN NIC GRC ZAF UGA MNE LUT TUN JAM 0 MPX 0.00 KTP PRT TIK SLV BIH ZWE XLX NOR MLI NGA MDA ITA IND USA CIV TTO KNA ARM HRV CPV ATG LKA LCA AFG BRB MUS –0.18** –2 MNG BLZ BDI TLS SLB PAK BHR RWA FJI BTN TON –4 –5 0 5 10 Relative number of mandates Developing-country trend Developed-country trend Sources: Computation based on 2017 World Bank Group Global IPA Survey; World Development Indicators database. Note: The figure is a scatterplot of the regression of log net FDI inflow and number of mandates, controlling for total investment promotion staff, country gross domestic product (GDP), and country population. “Relative number of mandates” refers to the number of mandates after partialing out the effects of the predictor variables used. The survey received responses from 83 national IPAs globally. (For more information, see annex 5B.) All country income categories use World Bank-defined classifications; “developing countries” refers collectively to all low- and middle-income countries and “developed countries” to high-income countries. Country labels are International Organization for Standardization (ISO) codes. FDI = foreign direct investment; IPAs = investment promotion agencies. Significance levels: *p < .10, **p < .05, ***p < .01 What Can Governments Do 1. Strategic alignment and focus 2. Coherent institutional framework to Improve the Development 3. Strong investor service delivery. Impact of Their IPAs? The three pillars should have a strong The World Bank Group proposes a new foundational base—that is, a national framework to help countries establish or development plan or vision, coupled with ­ strengthen their IPAs’ ability to achieve corresponding investment policies or FDI development objectives, especially through strategy. The pillars are also interrelated: the ­ increased and higher-quality FDI inflows strategies should inform the institutional (Heilbron, and Aranda-Larrey 2020). This framework, not the other way around, and framework is based on the literature as well both would determine service delivery as on the World Bank Group’s research and strength. The framework is proposed to serve operational experience.21 It consists of three as an assessment tool to help determine how thematic pillars that IPAs should consider to an IPA is performing on each of the specific increase investor satisfaction and confidence indicators under each pillar. It is also as a means toward higher development intended to function as a road map for IPA impact (figure 5.12): improvement. I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    185 FIGURE 5.11  IPAs in Developing Countries Indicate That Financial and Human Resources, as Well as the Economic Environment, Are Their Top Challenges Question: What are the three biggest challenges hindering your IPA’s performance? Lack of financial resources 52 33 Government support and public policies 49 29 Capacity issues of HR 44 29 Economic environment 41 57 Processes and bureaucracy 19 43 IT issues and digitalization of activities 14 0 11 Investment facilities and services 24 Lack of nonfinancial resources 10 10 0 10 20 30 40 50 60 Share of IPAs indicating item among three biggest challenges (%) Developing countries Developed countries Source: Computation based on 2017 World Bank Group Global IPA Survey. Note: The survey received responses from 83 national IPAs globally. (For more information, see annex 5B.) All country income categories use World Bank- defined classifications; “developing countries” refers collectively to all low- and middle-income countries and “developed countries” to high-income countries. HR = human resources; IPAs = investment promotion agencies; IT = information technology. BOX 5.3 The Role of Subnational IPAs in Attracting FDI The field of investment promotion is increasingly policy makers about a “race to the bottom” within ­ taking on a local focus, with subnational investment a country, whereby subnational locations compete on promotion agencies (IPAs) growing in prominence the basis of incentives for new investments coming alongside their national counterparts. In parallel, into the country or even displace investment from one investors are becoming ­ increasingly sophisticated in jurisdiction to another. evaluating the benefits of particular cities, provinces, Despite the apparent importance of these local or regions within a ­ c ountry, particularly in larger entities, the literature on subnational IPAs is almost countries. entirely absent. A recent study by the MASSIVE Subnational IPAs play unique roles and have (Multinationals, Institutions and Innovation in distinct characteristics compared with national IPAs. Europe) project, funded by the European Research Given their smaller jurisdictions, these agencies Council at the London School of Economics, has often have deeper knowledge of the local business taken a more thorough look (Crescenzi, Di Cataldo, environment and its value proposition to investors, and Giua 2019a). It leverages a survey on national and as well as stronger ties to local agencies more subnational IPAs in Europe to systematically evaluate heavily involved in the day-to-day operational needs the impact of investment promotion efforts to attract and issues facing investors. Despite these benefits, foreign direct investment (FDI) toward areas and subnational IPAs may raise particular concerns for sectors that would otherwise not be targeted. Box continues next page 1 8 6    Global Investment Competitiveness Report 2019/2020 BOX 5.3 The Role of Subnational IPAs in Attracting FDI (continued) The study finds that subnational IPAs help generate capacity for attracting FDI in knowledge-intensive sec- FDI inflows at the local level, in terms of both the tors than in other sectors. probability of receiving FDI and the total amount of Although further analysis and global coverage FDI received. This impact is more pronounced in less- is needed to address this underresearched topic of developed jurisdictions, likely pointing to the role of subnational investment promotion, these findings subnational IPAs in helping investors address infor- suggest that policy makers ought to carefully consider mation gaps, inadequate transparency, and weaker and potentially highlight the unique role that institutional conditions in these areas. Moreover, sub- subnational IPAs can play in fostering FDI. national IPAs in the study have demonstrated a better Source: Crescenzi, Di Cataldo, and Giua 2019a. FIGURE 5.12  Core Elements for Increasing the Development Impact of Investment Promotion Agencies Development Impact Higher Investor Satisfaction and Confidence Strategic Coherent Strong alignment institutional investor service and focus framework delivery National Development Plan, Investment Policies, FDI strategy Source: World Bank. See Heilbron and Aranda-Larrey 2020 for more information. Note: FDI = foreign direct investment. Pillar 1: Sharpen Strategic Alignment A strategic planning framework should and Focus foster stakeholder dialogue among the public and private sectors, organized labor, aca- Sharpening a country’s strategic alignment demia, and civil society on national priorities and focus is a key pillar to strengthen the and the role of each of them in development. development impact of IPAs. Doing so However, this process is difficult. Many coun- requires developing a shared vision while tries lack key pieces of this strategic frame- making difficult strategic decisions. work, leaving stakeholders to deal with I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    187 investment in an uncoordinated fashion, It lays out actions to establish the IPA (if new) sometimes duplicating functions or leaving or reinforce it (if existing), giving it a stronger important gaps in the investment ecosystem strategic focus, institutional capacity, and and investor services. commensurate resources. IPAs need to monitor global trends— Adopt a Cascading Strategic Framework including changing sources and modalities of for Investment FDI, dynamic segment and GVC trends, Several levels of strategic planning should technological advances, geopolitical tensions, ­ cascade down from an overarching national and other global opportunities or threats to development plan that would likely aim at inform the strategic development ­ process. achieving the Sustainable Development Notably, for example, in the context of the Goals (SDGs) of the United Nations’ 2030 COVID-19 pandemic, strong strategy, cou- Agenda for Sustainable Development. The pled with business intelligence, can help IPAs planning process should help clarify expecta- identify changes in GVC activities and advo- tions and roles for each stakeholder based on cate for needed reforms to support struggling a broad consultation.22 businesses, while proactively promoting new This step provides the opportunity for ­ opportunities. This type of data-driven, pro- countries to discuss sentiment concerning active approach is needed for IPAs to foster FDI, evaluate pros and cons, and decide how their influence to reposition their locations, to best leverage FDI for development. The respond to investor needs, and capture higher resulting national vision, strategy, or develop- levels of FDI. ment plan should clearly indicate the roles of IPAs should become active providers of the private sector and FDI while providing feedback on policy making and identifica- direction for more coherent policies and insti- tion of strategic segments, banking on the tutional frameworks for investment. That has valuable insights the IPAs gain from their been the case in Ireland since the 1960s, in daily contact with investors and by connect- Malaysia since the 1990s, and more recently ing investors and policy makers. Highly desir- in Colombia, Myanmar, and Rwanda.23 able segments that remain unattractive to The strategic process should continue cas- investors in the short-term require the IPA to cading down to develop the country’s advocate for related reforms or improvements investment and industrial policy as well as ­ in the investment e ­ cosystem. When economic its FDI strategy. The latter should identify, conditions turn unfavorable, including times through a data-intensive process, the coun- of global crisis like the COVID-19 pandemic, try’s most competitive segments to promote, IPAs should emphasize measures that support the reforms or measures to help improve the business continuity—by bolstering investor country’s competitiveness for desired seg- services around investor communications, ments, and the target markets and investors. retention, and advocacy. These high-level strategies should inform the more specific institutional strategies, such Define Target Segments, Source Markets, as the IPA corporate plan and investment and Investment Types ­ p romotion strategies for individual tar- IPAs need a strong focus on a few segments, geted segments. Scotland’s IPA, Scottish markets, and even investment types, which is Development International (SDI), illustrates achieved as part of the strategic planning this strategic development process well, with process and based on strong analytic capac- the added layer of alignment at the UK level. ity. Fortunately, GVCs have broken down An IPA’s corporate plan (also called an IPA products and services into business activities. business plan or strategy) is the road map for ­ncreasingly avail- At the same time, data are i the IPA to achieve meaningful developmental able, allowing countries and their respective goals, with clear milestones for each key per- IPAs to identify more specific “segments” formance indicator (KPI) along the way. and develop investment promotion strategies 1 8 8    Global Investment Competitiveness Report 2019/2020 for focused business or GVC activities (such considering the type of investment they are as manufacturing, ­ marketing and sales, cus- seeking to attract. Good alignment with tomer support c ­ enter, headquarters, and SDGs will result in higher-quality FDI, with R&D), as discussed in box 5.2. These activi- investment projects that go beyond e­ conomic ties determine motivation and location selec- considerations to integrate social and envi- ­ tion factors (Crescenzi, Pietrobelli, and ronmental dimensions. To date, IPAs have Rabellotti 2014). mostly focused on the e ­ conomic develop- Although countries may foster i ­nvestment ment and environmental dimensions in many segments, IPAs should focus on just a (Filippov and Guimon 2012; VCC and few (three to five) competitive segments for WAIPA 2010). proactive promotion efforts and high-level To maximize their impact, IPAs should service offerings. Such a focus does not mean not only target MNEs (for example, PVH in that IPAs would ignore or reject FDI in other Ethiopia or Volkswagen in Rwanda) but also segments but that a much higher level of consider their global suppliers to more fully resources would be dedicated to seeking and develop the respective cluster. IPAs could supporting projects in those identified strate- offer a linkages service by mapping MNE gic segments. Evidence suggests that the demand and introducing domestic suppliers number of targeted segments an IPA can man- to MNEs. However, building suppliers’ age ­ v aries with its level of maturity and capacity should not be part of an IPA’s resources (see figure 5.10). Developing or mandate. fragile and conflict-affected contexts may In certain countries, IPAs could consider require further focus (Whyte and Griffin brownfield forms of investment, beyond 2014). attracting and retaining greenfield FDI, their Identifying dynamic and competitive traditional domain (see chapter 2). Govern- ­ segments to proactively promote requires ments have ­various development motives and an evidence-based approach. The United means to foster the potential of brownfield Kingdom recently launched an approach to FDI ­ventures. IPAs could provide services, for prioritize segments on the basis of the value example, of connecting foreign investors to that FDI will have on the economy, as potential mergers and acquisitions (M&A) or ­ e stimated by an econometric model that joint venture candidates. includes several impact indicators (DIT 2018). Robust research and consultation Develop a Strong Investment Promotion with investors help IPAs understand trends, Mandate ­ investor factors, challenges, and needs as well Developing-country IPAs and most dis­ as how a country’s location compares with advantaged regions in developed countries competing locations for these segments so should strengthen their strategic focus on the that the IPA can build compelling value investment promotion mandate (Crescenzi, propositions and effectively attract FDI into ­ Di Cataldo, and Giua 2019a). This clarity these segments. CINDE (box 5.4) and IDA and focus of mandate are especially important Ireland (box 5.5) are examples of IPAs that for young IPAs or those that need to be have successfully operated this way for years. strengthened. Policy makers need to carefully World Bank Group experience indicates consider the pros and cons when adding that a top-down imposition of target s ­ egments either an investment promotion mandate to on an IPA by higher authorities often does not other entities or nonpromotion mandates to yield the expected results, likely because these the IPA, especially in developing countries. segments have not been properly validated on ­ Examples of good-practice IPAs exclu- competitiveness. As mentioned, IPAs can still ­ sively focused on investment promotion provide services reactively to investors. include the Austrian Business Agency (ABA- IPAs should also take into account the Invest in Austria); CINDE; Invest in SDGs when designing their strategies and Bogotá (Colombia’s subnational IPA); I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    189 BOX 5.4 Costa Rica’s IPA (CINDE) Follows Several Good Practices, Particularly around Strategic Focus Although the Costa Rican Investment Promotion It has also refined its sectoral focus from broad Agency (CINDE) is a private sector-led IPA with sig- light manufacturing in the early 1990s to strategic nificant autonomy, it has benefited from high levels of and specific global value chain (GVC) links by the government support and strong partnerships since the late 1990s, including the assembly and testing of late 1990s. These factors helped the IPA land not only electronic circuits, assembly and sterilization of technology giant Intel in 1996—which subsequently therapeutic devices, and legal and financial global shaped the country’s economic landscape (MIGA shared services. 2006a; Nelson 2000, 2005, 2009; Spar 1998)—but At the same time, CINDE improved its service also Abbott Laboratories (now Hospira), P&G, and offerings beyond the initial stage of attracting other anchor investors in the country’s most dynamic investment. It now proactively accompanies strategic segments (health sciences and information technology investors throughout their investment journeys. [IT]-enabled services). Advocating on behalf of investors and proactively CINDE has continuously sharpened its strategic connecting investors and government, CI N DE focus, evolving from an all-purpose development has helped catalyze key reforms, unlock strategic agency when founded in the mid-1980s to a fully investments and increase the country’s participation focused IPA attracting and expanding foreign direct i n G VC s t e n fold i n t he p a s t t h re e d e c ad e s investment (FDI) projects by the turn of the century. ­(figure  B5.4.1). FIGURE B5.4.1  Costa Rica’s FDI Inflows and GVC Participation Have Increased Tenfold since the 1990s 5,000 4,500 4,000 3,500 3,000 US$, millions 2,500 2,000 1,500 1,000 500 0 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Net FDI inflows (BoP) GVC participation (FVA + DVX) Sources: World Development Indicators Database; United Nations Conference on Trade and Development (UNCTAD)-Eora GVC database (https://worldmrio.com/unctadgvc/). Note: Global value chain (GVC) participation is measured by adding foreign value added (FVA) and domestic indirect value added (DVX). BoP = balance of payments; FDI = foreign direct investment. 1 9 0    Global Investment Competitiveness Report 2019/2020 BOX 5.5 Learning from the Success of a 70-Year-Old IPA: IDA Ireland Founded in 1949 and incorporated as an autonomous sectors that the IPA was targeting. An expert group state-sponsored body in 1969, Ireland’s investment on future skill needs was formed in 1997 to guide promotion agency (IPA), known as IDAa Ireland, has these decisions. Today, one-third of college graduates a decades-long track record of consistent achievement. specialize in sciences and engineering, and one-fourth When IDA Ireland was restructured in 1969, are in business. Ireland was not regarded as an attractive investment The government also created a combination of destination (because of economic stagnation, limited well-funded state agencies and advisory councils with natural resources, one of the lowest incomes per capita specialized functions, such as (a) IDA Ireland, which in Europe, and a total population of just 2.9 million). focused on FDI attraction; (b) Forfás, which focused IDA Ireland managed to change global perceptions on strategic planning for enterprise, trade, science, and helped transform Ireland into a foreign direct technology, and innovation—absorbed by Enterprise investment (FDI) powerhouse and a US$383 billion Ireland in 2014; (c) Enterprise Ireland, which supports economy.b indigenous industry and export development; and (d) Ireland’s cumulative FDI stock of US$909 billion Science Foundation Ireland, which fosters innovation. (UNCTAD 2019) is 237 percent of GDP and 2.6 In addition to their own synergies, these agencies times the European Union (EU) average. IDA Ireland have good working relationships with key regulatory reports that the 1,444 FDI companies it helped attract agencies at the national and local levels as well as with have generated 229,000 jobs (up to 2018), spent about private sector organizations. All employ professional US$20 billion in the Irish economy (in 2017), and and permanent staff who do not change when the represent 67 percent of the country’s total exports.c government changes. Several key elements contributed to this success. St rong mon itor i ng a nd eva luat ion (M&E) Clear mandate and sector strategy. With the capabilities. Government support for FDI attraction focused mandate of attracting FDI, IDA Ireland has was greatly helped by IDA Ireland’s development of developed a deliberate strategy to promote industrial a simple cost-benefit model, which demonstrated to development by targeting three sectors and three the government and taxpayers the economic benefits business activities (resulting in strategic segments) in and the inherent self-funding nature of investment which Ireland could achieve a competitive advantage. promotion. In 2018, IDA Ireland reported having created 124,000 The model uses a simple economic table to jobs in international and financial services; 64,000 in calculate the costs and benefits over a period of seven life sciences; and 22,000 in computers, electronics, years (to allow sufficient time for investment projects and optical equipment. It has managed to build a to build up to full production capacity) for every critical mass of firms in each of these sectors, which individual project supported by the government. IDA has a self-reinforcing clustering effect (IDA Ireland Ireland has established a target cost-benefit ratio of 2018). four to one—that is, the value of the future benefits Commitment toward improving the investment over seven years must be at least four times greater climate and forging effective partnerships. IDA Ireland’s than the (shared) costs to the state of running IDA success would not have been possible without a national Ireland and the cost of all financial incentives to the commitment to free trade from the 1960s onward and specific investment over that period. IDA Ireland a social partnership agreement whereby government, publishes the results of its aggregate cost-benefit employers, labor, farmers, and nonprofit organizations analyses every year to show that economic benefits collaborated closely to reach a consensus on development exceed costs over time. This model is now a standard priorities, moderate wage increases and cut taxes, and tool used by IDA Ireland in its requests to the share efforts to achieve national goals. government for funding. In addition, having identified low labor skills Adequate institutional and financial autonomy. as a main constraint, the government invested IDA Ireland has a separate legal mandate that grants significantly on education (amounting to 13.5 percent it a substantial degree of institutional and financial of public spending in 2016).d The Irish government autonomy and a sufficient and sustained budget, year focused education and training on the key technology in and year out. Box continues next page I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    191 BOX 5.5 Learning from the Success of a 70-Year-Old IPA: IDA Ireland (continued) Its board of directors counts on private sector offered merit bonuses tied to their performance against representation, but board members are clearly these targets. The strong support received from the appointed to represent public interests instead of government has enabled long-term, strategic, and private ones. Staff are paid at market rates and have consistent policies to survive outside political or both public and private experience. The stability of electoral time frames. senior management has avoided frequent changes at Sources: IDA Ireland 2018; IDA Ireland website: http://www.idaireland.com; the top level—for example, IDA Ireland has had only interviews with former IDA Ireland staff; UNESCO Institute for Statistics data: six chief executive officers (CEOs) since its creation. http://uis.unesco.org; UNCTAD 2019; World Bank’s World Development Indicators database. IDA Ireland has sector-based staff both in its a. The agency was founded in 1949 as the Industrial Development Authority. headquarters and in 20 offices in 13 countries. Its b. Ireland economic data from the World Bank World Development Indicators overseas offices are staffed by about 40 professionals. Database. c. FDI impact data from the IDA Ireland website: http://www.idaireland.com. Targets are set annually by sector, by country, by d. Education spending data from United Nations Educational, Scientific and Cultural office, by Irish region, and by staff member. Staff are Organization (UNESCO) Institute for Statistics 2016 data: http://uis.unesco.org. InvestHK (Hong Kong SAR, China); Invest • Inform the national vision for FDI, and India; and Scottish Development International develop and implement an investment (SDI). Standard components of an investment promotion strategy, in collaboration with promotion mandate include the following: relevant public and private sector part- ners in the country. • Attract FDI inflows by influencing inves- • Monitor, research, and gather intelligence tor location decisions with marketing, about investment. information, and assistance services (such as an outreach program) There could be a case for placing trade or • Support investments from ­ announcement export promotion and investment promotion to start-up with information and assis- under the same agency when there are ­ strategic tance services (such as an establishment looking to synergies, as in the case of a country ­ program) attract and expand investment geared toward • Retain and foster expansion of existing exports. Such functions should not merge investors with marketing, information, solely as a budget-cutting measure (Heilbron assistance, and advocacy services (such as and Whyte 2019; UNCTAD 2013). Each divi- full-service aftercare programs) sion must have sufficient resources to deliver • Encourage and facilitate business link- on its specific mandates. Cost savings may ages between foreign firms and domestic accrue from combining common back-office ones as well as other mechanisms for the functions—administration, finance, IT, human spillover of skills, technology, know-how, resources, legal—and some international and international market networks (such offices (when the market is a target for both as a linkages program) functions). However, this is rarely successful • Advocate for improvements to the loca- when institutions are ­ starting up or have low tion’s competitiveness, general invest- capacity. Policy ­makers need to recognize the ment climate, and sectoral ecosystems differences between the two mandates (differ- before the government, private sector, ent goals, company targets, company decision and any relevant stakeholders (such as an levels, markets, sales cycles, promotion instru- advocacy program) ments, job requirements, and budgets). 1 9 2    Global Investment Competitiveness Report 2019/2020 Support for outward FDI (OFDI) is a natu- the promoter is under the same roof as the ral extension of export promotion, as domes- regulator or incentives approver or when tic firms move from exporting to establishing resources favor domestic investment to the det- a presence in that foreign market. IPAs, espe- riment of FDI promotion (Heilbron and Whyte cially in developing countries, should not 2019). Table 5.2 summarizes this guidance for include or add a mandate to promote OFDI IPA mandates in the developing world. because the needs of such domestic firms are The World Bank Group’s operational quite different from those of foreign investors experience has shown that separation of coming into the home country (Heilbron and investment promotion function from other Whyte 2019). government units—including those dealing IPAs, particularly in the developing world, with regulatory and incentive approvals, should also ideally not have mandates cover- SME development, SEZ supervision, SEZ ing regulation, investment incentives, small development, and PPP administration or and medium enterprise (SME) development, concessions—produces better results for s pecial economic zones (SEZs), or PPPs. ­ both the investment and the specialized man- Governments often struggle when delivering dates in developing countries. At the same support to private sector investors. At times, time, strong intergovernmental cooperation mostly to cut costs, they combine several func- between the IPA and the specialized units is tions within the same institution. However, essential to improve effectiveness. especially in developing countries, this may If the IPA is assigned mandates beyond create issues—such as conflict of interest when promotion, sequencing the mandates over TABLE 5.2  Dos and Don’ts of Mandates for IPAs in Developing Countries Category Institutional function Investment promotion Marketing services Information Assistance Advocacy Other promotion types Foreign investor and local supplier matchmaking Export promotion Support for outward investment SME development Administration or regulation Administration of incentives Screening or approval of investment projects Issuance of noninvestment licenses or permits Administration or negotiation of government concessions (such as in infrastructure or extractive industries) Administration of public-private partnerships Management of state land or assets ■  Must do  ■  Okay to do, but with carea  ■  Do not do Source: World Bank. Note: This table captures general guidelines based on more than 30 years of World Bank Group operational experience in developing countries. Country- specific characteristics may warrant different approaches and considerations. FDI = foreign direct investment; IPA = investment promotion agency; SME = small and medium enterprise. a. These functions can be included with certain strict organizational prerequisites to avoid impeding investment promotion. I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    193 time may produce the best results. Many of In determining the right institutional the IPAs that have added mandates beyond f ­ramework for investment promotion, policy investment promotion have struggled to makers should address the following key ele- adapt effectively to their expanded portfolios ments and consider the recommendations without losing promotional effectiveness. described in more detail below: Among the few with decades-long success in managing additional mandates is IDA • A common vision and strategic alignment Ireland, which includes incentives and • Corresponding institutional network KPIs property management. More recently, the • Clearly identified institutional partners Rwanda Development Board has been with focused mandates, division of roles, forging a solid international reputation in focal points, joint promotion activi- this regard (See box 5.6 for more information ties, and some shared assets (such as an on the mandates of IDA Ireland and the i nformation library, online portals, and ­ Rwanda Development Board). These cases tracking tools) appear to have at least the following • Institutional coordination/collaboration characteristics in common: guidelines and protocols • Capacity building in the area of invest- • Their mandate expansions were ­ deliberate ment promotion. and motivated by strategic objectives—as opposed, for example, to cost-saving or Provide High-Level Government Support to political motives. FDI and the IPA • Different mandates (promotional and If the government is serious about leveraging nonpromotional ones like ­ r egulatory FDI for development, it must build an functions) have been sequestered, allow- institutional framework with a strong IPA ­ ing promotional staff to operate some- at its core. This IPA should integrate key insti- what autonomously according to their tutional characteristics that are associated own strategies and resources, and with a with stronger performance, starting with the private sector mindset. highest level of government support (from the • Promotional staff have been held accoun­ president or prime m ­ inister) directly or table for performance against their own ­ i ndirectly championing the needed legal, impact indicators. ­ regulatory, and institutional reforms. • There are no mandate overlaps with As part of its high-level support, the gov- other agencies, especially for investment ernment should send clear signals that promotion. instill confidence in investors and back the IPA with the appropriate legal status and h ierarchy within the government for the ­ Pillar 2: Build a Coherent Institutional d ifficult task of mobilizing stakeholders ­ Framework typically required to bring about significant ­ Building a coherent institutional frame- or transformational investment. Being a work for investment is crucial to FDI effec- unit of a department or ministry may, in tiveness, even if it takes significant time and most developing countries, undermine the effort. This is especially important when IPA’s influence to mobilize stakeholders to setting up an IPA or restructuring the exist- reform and obtain investment. The right ing one. The right institutional setup varies, attachment—​ b ut more importantly, the depending on the country’s political econ- right support from the highest levels of the omy, the government’s existing institutional ­ government—gives the IPA a certain stature framework, available legal institutional and visibility in the country. It also sends f ormats, the civil service culture, and the ­ strong signals to investors and government institutional collaboration culture alike about the priority the government (Heilbron and Whyte 2019). places on investment and FDI. 1 9 4    Global Investment Competitiveness Report 2019/2020 BOX 5.6 How Different Institutional Setups Respond to Multiple Mandates: The Cases of Ireland and Rwanda IDA Ireland Rwanda Development Board IDA Ireland a has undergone several major reforms Rwanda went the other way, by expanding mandates. to its mandate in the 70 years since its creation, as Since the Rwanda Development Board (RDB) was government priorities have shifted and as good founded around 2008, it has accumulated functions, practices in investment promotion have emerged. IDA building on those of its predecessor body, the Ireland was established in 1949 as a subministerial Rwanda Investment and Export Promotion Agency unit, during a period of protectionism, to stimulate (RIEPA). Today, the RDB includes a one-stop shop the development of exporting enterprises. Nine for investment-related procedures; sector development years later, its focus was changed to foreign direct of two of the government’s highest-priority sectors, investment (FDI) and export promotion. Recognizing information and communication technology (ICT) and the need for a more private-sector-like approach tourism; and the administration of special economic and greater operational flexibility, the government zones (SEZs), public-private partnerships (PPPs), and changed IDA Ireland’s legal form in 1969 from a special projects with international donors and partners. subministerial unit to an autonomous state-sponsored It reports directly to the president of Rwanda. body. Twenty-five years after that, in 1994, enterprise Given the RDB’s strong track record of economic development and export promotion were moved out reform and growth, the decision to add the one-stop of IDA Ireland so that it could be dedicated to the shop was seen as a way to extend its influence and good promotion and development of high-quality FDI. Two performance to areas that were impeding investment of its prior mandates, seen as assets in landing FDI, (such as procedures for establishment). This is a were retained: the regulatory function of incentive common motive for the assignment of one-stop shops administration and the management of industrial to IPAs. At the same time, a shift in organizational estates. culture away from promotion in favor of regulation In short, IDA Ireland came to its current set of is also a major reason for IPAs not performing on mandates not by addition but by subtraction. Although their promotion mandate. The RDB tries to mitigate it does more than other leading IPAs today, its this risk by dividing its investment division into (a) a mandates have actually decreased over time—the result promotion department with 11 sector-focused units, of many years of policy experimentation and lessons which list proactive outreach to potential investors as learned. In its current form, IDA Ireland organizes its their first activity; and (b) its “one-stop center,” which has limited its focus to business registration, duty promotional work through 11 units, mostly focused on exemptions, work permits, and environmental impact sectors, reporting to the chief executive officer (CEO). assessments. SEZs, PPPs, and sector development Its nonpromotional work is conducted through nine have their own divisions. Close attention and high units, mostly functional units, under a single executive expectations from the president are also seen as having director in charge of all noncore functions such as much to do with the RDB’s continued success. human resources, legal affairs, finance, and corporate Sources: Interviews with RDB staff and former IDA Ireland staff; RDB website strategy and planning as well as incentives and real (https://rdb.rw/); IDA Ireland website (http://www.idaireland.com). estate management. a. The agency was founded in 1949 as the Industrial Development Authority. Ethiopia’s recent tenfold growth in FDI Transformation Plans I and II (covering the stems from a high-level engagement by the five-year periods of 2010/11–2014/15 and former prime minister and his economic 2015/16–2019/20, respectively), included adviser in opening the economy to foreign ­ elevating the Ethiopian Investment investors and restructuring the institutional Commission to report to a newly created framework for investment. This restructur- Investment Board, chaired by the prime ing, guided by the government’s Growth and minister.24 I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    195 Turkey offers another good example of a and composition similar to a board of direc- national IPA with a high position in tors, even if it does not have legal authority government. The country’s Investment Office ­ over the IPA’s operations. is s­ ituated within the Presidency of the Republic of Turkey and reports directly to Ensure the IPA Is Staffed with the Right the ­ p resident. In India, Prime Minister Expertise Narendra Modi’s support of Invest India IPAs need a good mix of IPA management helped transform the ­ country’s IPA from a and staff with private sector experience. small, uninfluential 10-person unit to a Many IPAs are already incorporating more vibrant operation with more than 100 staff private sector expertise as both empirical that has contributed to higher levels of FDI research and operational experience highlight and is in the process of upgrading the capac- the need for IPAs to develop transnational ity of Indian subnational IPAs (box 5.7). learning capacity and adopt an investor- Some countries achieve high-level visibility minded, service-oriented, and ­ consultancy-like by attaching the IPA to the president’s or democratic organizational culture (ECORYS prime minister’s office. However, policy mak- 2013; Nelson 2009; Ortega and Griffin ers need to be aware of the risk that the IPA 2009). The IPA’s CEO, promotion director, may get little attention from an already busy and key promotion staff should have private p resident or prime minister and become ­ sector experience, international exposure, and unstable in a political transition. Accordingly, fluency in relevant languages, as well as an empowered autonomous unit reporting to strong interpersonal abilities. The CEO needs a formal board of directors, investment to enjoy a high level of credibility with both board, or advisory board chaired by the head the private and public sectors. For this to hap- of state or by a strong ministry may be a good pen, the IPA requires independence from the combination of high stature with a more civil service restrictions for human resources. ­sustainable approach. Invest India, for example, rapidly ramped up from a weaker promotional unit with little Grant the IPA a Sufficiently High Level of impact to an award-winning IPA that has Autonomy helped raise FDI inflows to new levels after IPAs tend to perform better when they have hiring high-caliber management and staff a high degree of financial and operational (90 ­percent of whom come from the private autonomy or ­ independence. IPAs should be sector) and adopting an operating model that allowed to operate following an approved mirrors consulting firms, with young talented strategic plan with minimal ­ political interfer- staff providing data-driven support to inves- ence; hire staff with private sector experience tors (box 5.7). independently from the civil ­ ­ service (as do CINDE, Invest Bogotá, and Invest India); be Provide Sufficient and Sustained Financial accoun­ table and report results to a board of Support to the IPA directors; and maintain continuity through- To perform effectively, IPAs require reliable out political cycles. funding over a three- to five-year period, In addition to top government officials, the given the long-cycle nature of investment pro- IPA board needs to have active and strong motion. In the 2017 World Bank Group private sector representation (ECORYS 2013; Global IPA Survey, IPAs indicated that their Miškinis and Byrka 2014) from key chambers most important challenge concerns financial of commerce or business associations, foreign resources and that financing comes mostly private sectors, and professional advisers. from public funds. Charging fees to investors Their expertise helps the IPA better is not a good practice because investment ­ understand investors and deliver relevant ser- promotion should be seen as a public good. vices to them. Nonautonomous IPAs should Fees can send a wrong signal to investors that at least have an advisory board with functions FDI is not prioritized or encouraged. 1 9 6    Global Investment Competitiveness Report 2019/2020 BOX 5.7 Invest India: Building a High-Performing IPA from the Ground Up in Record Time Invest India illustrates how a new investment pro- recruitment procedures. It adopted a consultancy-­ like motion agency (IPA) can be built up over just a few operating model designed to offer quality services years by following a few key ­ principles while avoid- to investors. The agency has now developed into a ing many typical mistakes. Invest India was estab- dynamic, service-oriented organization with highly lished in 2009 as a joint venture of the Department qualified staff. As of mid-2019, 51 percent of the of Industrial Policy and Promotion (DIPP) of the staff were women, 90 percent of its 138 dedicated Ministry of Commerce and Industry, the Federation professional staff had private sector experience, and of Indian Chambers of Commerce and Industry 60 percent had graduate degrees. (FICCI), and state governments of India. The agency Invest India also set about developing strong lay dormant for many years and, even as recently as relationships with state-level IPAs across India. It 2015, had no more than a handful of staff reactively took a proactive approach to better understand answering investor queries, with little or no traction the strengths and the needs of each state in terms with potential foreign investors. of competitiveness to attract new investment and T hen , i n S eptemb er 2014, Pri me M i n ister capacity to support incoming investors. It provided N arendra Modi launched “Make in I ndia,” a ­ direct support to subnational IPAs and invited all ­ g overnment initiative to persuade and encourage state IPAs to participate in a World Bank Group IPA companies globally to manufacture their ­ p roducts assessment designed to further strengthen capacity. in India (http://www.makeinindia.com). As part of Confidential reports were delivered in March 2018 this initiative, the government decided in 2015 to to 21 state IPAs offering tailored advice on areas for reinvigorate “Invest India,” recognizing the need for ­ improvement. a government agency to proactively tackle the attrac- Since the agency’s rejuvenation in 2015, Invest tion of foreign firms. Invest India was mandated to India has been transformed into an award-winning ramp up its ­ i nvestment promotion efforts and capi- IPA, receiving the United Nations Conference on talize on the potential of India’s economy. Trade and Development’s (UNCTAD) global award Following international good practices, Invest for best-practice IPA in 2016 and for sustainable I ndia received the full suppor t of senior-level development investments in 2019. It was also named government officials. A new chief executive officer ­ best IPA in South Asia, East Asia, and Oceania at (CEO) was appointed and given direct access to the Dubai’s Annual Investment Meeting (AIM) in 2016, line minister and the prime minister’s office. The 2017, 2018, and 2019, and was elected as co-vice IPA’s goals were linked to the country’s broader president of WAIPA for 2019–20. ­ development goals, and it was officially mandated By mid-2019, Invest India had responded to more to lead the ­ c ountry’s national investment promo- than 193,000 business requests from 126 countries tion as the single point of contact for foreign inves- and 41 sectors, 92 percent of which were answered tors. Following global good practices, it was not within 72 hours. Working with some 760 companies, assigned any regulatory functions. It was given it had generated a project pipeline of US$138 billion, sound financial support from the Department for of which an estimated US$22.7 billion had been exe- Promotion of Industry and Internal Trade (DPIIT, cuted, with 135,000 direct jobs in the process of being formerly DIPP) and a functioning board with 51 created and contributing to making India the world’s percent private sector representation. top-five greenfield destination in 2018.a Invest India was permitted to adopt a much fl atter organizational structure than typical Indian ­ Sources: Invest India website: https://www.investindia.gov.in/; interviews with Invest India management. civil ­ s ervice entities and to recruit high-caliber a. Data on Invest India’s outcomes from fDi Markets, a Financial Times dataset management and staff outside the normal civil ­ ­ service (https://www.fdimarkets.com/). I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    197 Develop Strong Partnerships within the country, such as subnational IPAs IPAs need to develop strong partnerships with competing for the same investors on the basis both the public and private sectors at the of incentives or concessions. Protocols of national and subnational spheres to harness engagement and coordination mechanisms FDI. They are interacting on average with 25 should help in this regard. India provides a different organizations (OECD 2018). good example (box 5.7). Strong partnerships develop through con- sultation and working together on common issues. For instance, jointly developing an Pillar 3: Strengthen Delivery of Investor investment vision or FDI strategy for the Services country helps strengthen these partnerships The 2019 GIC Survey reveals that investors and improves effectiveness, stemming from value IPA services: 90 percent of responding aligning stakeholders behind a few priorities investors value at least one IPA service, and for economic (and social) development. two-thirds or more appreciate IPA services Strong intergovernmental cooperation across different stages of the investment life between the IPA and specialized units dealing cycle, not just at attraction or entry. World with permits, incentives, and other noncore Bank Group experience and empirical IPA mandates is essential in effectively deliv- research demonstrate that IPAs are mostly ering services to investors. Other national and engaged in investment promotion events subnational stakeholders playing a role in while underdelivering services beyond the investment—such as sectoral ministries, envi- attraction stage. ronmental protection agencies, and utility IPAs should aim at improving their service providers—should also be integrated into the delivery across the investment cycle to increase institutional framework. investor satisfaction, especially for the services Strong national IPAs in larger countries that strategic investors consider most relevant. tend to have closer and more systematic The World Bank Group offers a new, working relationships with subnational comprehensive investor services framework IPAs—or at least more regular cooperation linking four service ­ categories (marketing, and contact with them. Institutional coordi- information, a ­ ssistance, and advocacy) across nation mechanisms are essential to avoid four stages of the investment life cycle investor confusion and frustration stemming (attraction, entry and establishment, retention from duplication or gaps in service delivery. and ­ expansion, and linkages and spillovers). Well-coordinated national and subnational The framework should be applied strategically IPAs complement each other and avoid such to each of the targeted segments. (Heilbron issues—taking on roles and delivering inves- and Aranda-Larrey 2020). The aim is to tor services based on their position within the provide a well-balanced mix of services typical investment cycle. For instance, it is depending on the ­ development level of the most effective and efficient for the national segments, with (a) p ­ roactive ­outreach and IPA to devote resources to marketing, espe- relevant i ­nformation in c ­ arefully identified cially when dealing with investors abroad, segments; (b) hands-on s ­upport to help as in the case of Apex-Brasil, Austrade investors e ­ stablish and expand (aftercare (Australia), and Germany Trade & Invest programs); and (c) a ­ dvocacy to continuously (GTAI). In contrast, the day-to-day problem improve the ­ investment ecosystem through solving for an established investor is generally fundamental reforms. For instance, quality IPA better dealt with at the subnational level (for information would increase transparency, and ­ example, by the SEZ or municipality). the IPA’s provision of effective assistance across Importantly, in creating a “national team” all stages would build investor predictability. spirit for promotion, protocols should be in Both improve investor confidence, which can place to avoid “race to the bottom” behavior support FDI growth (see chapter 4). 1 9 8    Global Investment Competitiveness Report 2019/2020 Well-respected IPAs such as ABA–Invest in Delivering top-quality, on-time services to Austria, CINDE, IDA Ireland, InvestChile, investors across all stages of the ­ investment InvestHK, and SDI all seem to apply the new life cycle requires systems, such as M&E on framework’s ­ principles and most of the ser- key KPIs; standard operating procedures; vices listed under it.25 The most notable dif- templates for repetitive tasks; a virtual ferences of approach among these leading library of the most frequently used docu- IPAs relate to the retention and expansion ments; a website where investors can meet stage, as some of these national IPAs seem to their initial information needs and download be working in partnership with subnational key documents and data; an investor IPAs taking care of such services. ­ relationship ­ management system with CRM As noted earlier, IPAs need to collaborate software at its core to track these interac- to provide ­ services to investors seamlessly tions over long periods; and investor sur- throughout their investment journeys. The veys. The United Kingdom has developed a national IPA is best placed to interact with monitoring, reporting, evaluation, and learn- investors ­ d uring the attraction stage but ing (MREL) framework to assess the effec- could decrease its role once investors get tiveness of all IPAs in the union, including its ­ established in a particular site if qualified national Department for International Trade subnational IPAs can take over. At that (DIT) and s ­ everal subnational IPAs in achiev- stage, the subnational IPAs can take ing impact relative to their respective regions’ ownership of the relationship and service characteristics (DIT 2018, 2019). Spain p rovision, and consequently provide a ­ has a portal to share investment ­ inquiries higher level of attention. This suggested broadly with all qualifying s ­ ubnational IPAs. framework with a diminishing role for the During the extremely uncertain times of national IPA and an increasing role for the COVID-19 pandemic, selected IPAs from the subnational IPA in different stages around the globe were showing resilience by of the investment life cycle is illustrated by operating remotely and delivering key ser- the triangles of figure 5.13. vices to investors. A WBG rapid survey, con- IPAs are (or could be) active connectors ducted in April 2020, showed that some between investors and policy makers, chan- IPAs were responding in the following ways: neling business-to-government (B2G) feed- back and highlighting the role IPAs can play • Strengthening communication via web- in delivering advocacy services. The 2019 sites, newsletters, and social media to GIC Survey revealed that investors consider update investors daily on developments advocacy to be the most critically important related to the virus and government service IPAs provide. responses FIGURE 5.13  Proposed Division of National and Subnational IPA Roles in Service Delivery to Investors Investment life cycle stage Service category Entry and Retention and Linkages and Attraction establishment expansion spillovers Marketing Information Subnational IPA Assistance National IPA Advocacy Source: World Bank. Note: IPA = investment promotion agency. I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    199 • Bolstering direct assistance (aftercare) • Grant the IPA a clear, uncontested man- services to established investors to solve date focused on investment ­ promotion, their individual issues especially when starting or restructuring • Boosting advocacy services to system- the IPA to empower the IPA and avoid atically solve pressing issues facing unnecessary duplication. investors. • Grant the IPA a high degree of institu- tional and financial autonomy (or semi- IPAs have a pivotal role to play in help- autonomy), emulating the ­ private sector’s ing governments respond to the COVID- flexibility to act according to ­ strategic plans 19 pandemic. This role extends beyond and hire staff according to specified and the immediate term to include the post- ­ transparent job ­ qualifications. This should outbreak recovery phase—providing rele- avoid political ­ interference and disruptions vant services to help investors return to ­ during p ­ olitical transitions. Having the IPA operations as quickly as possible, while report to an independent and functioning encouraging new investment and expan- board of directors or advisory board with sions into emerging strategic segments. strong and active private sector representa- tion would help it better understand inves- Conclusions and Future Research tors and provide direction in catering to The principal role of IPAs continues to be their needs. too often misunderstood, even by policy • Foster a strong, investor-centric service makers, as an exercise in simple marketing. orientation at the IPA and with partners. At the same time, most IPAs are struggling This includes allowing the IPA to hire to reach their full development impact management and key promotion staff because they lack strong support to evolve with strong private sector experience, quickly with FDI market trends and more international exposure, and language s ophisticated investor needs, they lack ­ skills as well as building capacity for strategic focus, and they do not adequately ­ the IPA and its partners to design and cover the services that investors value. provide relevant, high-quality services to For IPAs to improve their contribution to those strategic investors throughout the development, several policy r ­ ecommendations investment life cycle. are key: • Provide sufficient, sustained financial resources to the IPA over three- to five- • Provide the IPA with high-level govern- year periods to ensure continuity of stra- ment support (from the president or prime tegic efforts over the long-cycle nature of ­nvestment minister), giving high priority to i investment promotion and to avoid strug- (or FDI) and directly or ­indirectly champi- gling over funds every year or having to oning the needed legal, regulatory, and charge fees. institutional reforms for investment. • Foster strong strategic alignment, stem- Regarding future research, emphasis ming from consultations with the public should be placed on more rigorously assess- and private sectors and cascading from ing IPAs’ performance, especially to identify a national plan, vision, FDI strategy, or which IPA characteristics (autonomy, staff- industrial strategy that clearly states the ing, budgets, and so on) are most effective at role of the ­private sector. These higher- maximizing their development impact. More level national strategies also need to con- can be done to analyze the links between IPA sider IPA feedback and ­ permeate the IPA services (type and quality) and investor satis- corporate plans and sectoral strategies, ­ onfidence) faction (and potentially investor c which focus proactive investment promo- and their effect on FDI. tion efforts on a few properly identified To date, such research has been sparse, competitive segments. not only in the context of developing 2 0 0    Global Investment Competitiveness Report 2019/2020 countries but also in picking up on nuances Other areas that warrant deeper consider- different stages of economic develop- across ­ ation include further analyses on the role of ment. In this vein, econometric analysis at subnational IPAs as they become more preva- the global level should be pursued and ­ echanisms lent in developing countries; the m anchored in better collection of data across for better national-subnational institutional a wide international sample of IPAs, with collaboration and avoidance of “race to the time-series information on different IPA bottom” behavior; the role of IT and digitali- attributes and services, and critically, on zation on investment promotion; and the con- performance indicators (ideally with firm- tribution of IPAs to countries’ resilience in level data, or alternatively, with information times of crisis (notably, for example, the on IPA ­ segment targeting to link with data COVID-19 pandemic), integration into on FDI inflows). emerging GVCs, and in achieving the SDGs. Annex 5A. Key Measures of IPAs’ Impact Investment promotion agencies’ (IPAs’) top ­eatured in IPA strategies), particularly f performance indicators, in terms of their employment creation (VCC and WAIPA direct contribution to development, are 2010). However, most IPAs only collect including investment facilitated or generated (­ information on announced investment and foreign direct investment [FDI] inflows) and jobs, with few tracking FDI and employment jobs created. Worldwide, 95 percent of IPAs levels that are achieved (based on established responding to the 2017 World Bank Group investments). Global IPA Survey reported collecting data A few sophisticated IPAs such as IDA contribution to FDI, and 81 percent on their ­ Ireland and the United Kingdom’s Department collected data on job creation (figure 5A.1). ­ for International Trade (DIT) have been mea- IPAs are especially concerned about eco- suring return on investment of public funds, nomic development (the dimension most using a basket of indicators that include FDI FIGURE 5A.1  Nearly All National IPAs Collect Data on the Amount of Investment Facilitated and Jobs Created as Indicators of IPA Performance Question: For what performance or impact indicators did your IPA collect concrete data last year? Investments facilitated (number, US$ for investment inflows) 95 95% Jobs created (number) 81 81% Business climate reforms enacted (number) 62 62% Growth in a priority sector (% contribution to GDP) 41 41% Exports by companies facilitated (US$) 29 29% Jobs retained (number) 29 29% Other (please specify) 19 19% Increase in domestic sales to foreign affiliates (US$, percentage) 16 16% 0 10 20 30 40 50 60 70 80 90 100 Share of respondents (%) Source: 2017 World Bank Group Global IPA Survey. Note: The survey received responses from 83 national IPAs globally. (For more information, see annex 5B.) GDP = gross domestic product; IPA = investment promotion agency. I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    201 inflows, gross value added, wages, and measure the IPA’s performance as perceived by employment levels (DIT 2018; IDA Ireland survey respondents on the Likert scale, which 2018). Only a few developing countries mea- measures intensity of feeling: (a) FDI inflows, sure the degree of embeddedness of the invest- (b) target investment amount, (c) investment ments and their local impact in terms of promotion efficiency, and (d) survival rate of additional growth and jobs generated directly new invested ventures (Lim 2018).26 and indirectly. The United Kingdom is mov- Soft and hard factors (perceptions, FDI ing to net value added and employment. amounts, job numbers, and wages) need to be Most studies use the value of FDI inflows considered when evaluating IPAs (UNCTAD (Harding and Javorcik 2011, 2012; UNCTAD 2008). That said, data in general, but espe- 2001; Wells and Wint 2000) or the number of cially for indirect impact and qualitative FDI projects as the IPA’s measure of perfor- aspects, are not easily available. Such data mance, reflecting data availability. A more issues are even more pronounced in the devel- recent study proposes using four variables to oping world. Annex 5B. Analysis of IPA Surveys: Overview and Approach 2017 World Bank Group Global Comparison of Global IPA Surveys IPA Survey over Time From October 2016 to March 2017, the To compare IPA characteristics over time, data World Bank Group contacted 147 invest- from five different surveys27 are leveraged: ment promotion agencies (IPAs) to gather information on their characteristics and • 2005 World Bank Group IPA Census activities through a web-based survey. • 2017 World Bank Group Global IPA Eighty-three national IPAs responded Survey (a response rate of 56 percent). A regional • 2017 World Association of Investment breakdown of the respondents is presented Promotion Agencies (WAIPA) Annual in table 5B.1. Survey TABLE 5B.1  Regional Coverage of the 2017 World Bank Group Global IPA Survey Distribution of IPAs Number of respondents Region contacted (%) Response distribution (%) (number of IPAs) East Asia and Pacific 14 13 11 Europe and Central Asia 33 35 29 Latin America and Caribbean 18 18 15 Middle East and North Africa 10 7 6 North America 1 1 1 South Asia 4 6 5 Sub-Saharan Africa 21 19 16 Total 100 100 83 Source: 2017 World Bank Group Global IPA Survey. Note: The survey received responses from 83 national IPAs globally. “North America” includes Canada and the United States. IPA = investment promotion agency. 2 0 2    Global Investment Competitiveness Report 2019/2020 • 2018 WAIPA Annual Survey combines the remaining four surveys into sin- • 2018 OECD report, “Mapping of Invest- gle observations for the period 2017–18. The ment Promotion Agencies in OECD trend analysis is restricted to the subset Countries” (OECD 2018). of IPAs observed in both time periods, so ­ conclusions are not swayed by changes in Changes are evaluated across the following sample composition. Changes over time are key areas: governance structure, activities and calculated using IPA fixed effects to identify services, sector and country targeting, budget, within-IPA changes. staff, systems and tools, and foreign presence. A few caveats are in order. Sample size Because country and question coverage of the varies by question (based on presence and different surveys vary dramatically, precise format of questions in the survey). More comparison across time posed challenges. The broadly, caution should be exercised when 2005 World Bank Group IPA Census covered considering generalizability of results because the widest range of IPAs (106 in total) and the sample is nonrandom, and changes may provides a baseline for comparison over time. be associated with other unobserved To maximize sample size, the analysis characteristics. Annex 5C. Additional Data on Evolution of IPA Institutional Characteristics, 2005 to 2017/18 TABLE 5C.1  Selected IPA Characteristics with No Significant Changes between the 2005 and 2017/18 IPA Surveys Sample size Topic Question 2005 2017/18 (number of IPAs) % of IPAs that are government run 93 89 52 % of IPAs reporting to more senior levels of accountability (for example, prime minister’s or president’s office) 12 16 58 Governance % of IPAs with a board 71 75 56 structure % of IPAs with private sector members on their boards 52 55 21 % of IPAs that advertise in the media 78 90 40 % of IPAs engaging in analysis or policy advocacy to improve the investment climate 98 95 41 % of IPA’s budget devoted to image building 26 25 24 % of IPA’s budget devoted to investment generation 36 33 25 Activities % of IPA’s budget devoted to investor servicing 25 32 18 % of IPAs that target specific countries 66 63 35 Country targeting Average number of countries targeted by IPAs 6 4 32 Staff Average number of staff focusing on investment promotion 31 42 52 % of IPAs with private sector database 96 100 25 Systems and tools % of IPAs using an investor tracking system 53 55 38 Foreign presence % of IPAs with overseas representation 74 78 27 Sources: 2005 World Bank Group IPA Census and 2017 World Bank Group Global IPA Survey; 2017 and 2018 World Association of Investment Promotion Agencies (WAIPA) annual surveys; OECD 2018. Note: For details about the data sources and methodology, see annex 5B. Changes over time are calculated using IPA fixed effects to identify within-IPA changes. Significance of changes across time is calculated using IPA samples for which the question is commonly available in both 2005 and 2017/2018 to avoid changes being driven by changes in the sample composition. This table summarizes selected features where no significant changes, up to the 10 percent level, are observed. Nonsignificant changes do not necessarily indicate that no meaningful changes have occurred but may indicate that the sample size is too small to draw generalizable conclusions. Discrepancies in the phrasing of questions across surveys means that the authors have matched questions across surveys based on their judgment of sufficient comparability. I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    203 Notes 1. Throughout the chapter, investment promo- relevant. This may not be the case for FDI tion agencies (IPAs) refer to institutions that sourced from other countries, such as China. include an investment promotion function 8. CINDE identified the right competitive seg- or mandate (covering both dedicated agen- ment to target, engaged in one-on-one s ­ trategic cies as well as units that do so within larger outreach, and provided top-notch individual- institutions, such as economic development ized services to investors. boards). 9. “Efficiency-seeking” FDI leverages cost sav- 2. As throughout this report, “developing ings and competitive features of a l ­ocation to countries” refers to low- and middle-­ ­ income serve as an export base. countries, and “developed countries” to 10. For more information about the 2017 World high-income countries, based on World Bank Bank Group Global IPA Survey, see annex 5B. Group classifications: https://datahelpdesk​ 11. At the same time, Bauerle Danzman and .worldbank.org/knowledgebase/articles​ Gertz (forthcoming) also indicate that /906519-world-bank-country-and-lending​ autonomous IPAs are less likely to align their ­ -groups. activities with other government priorities. 3. FIAS—now called the Facility for Investment 12. Transnational learning capacity, in this con- Climate Advisory Services (still abbre- text, refers to the IPA’s capacity to understand viated as FIAS)—is a joint service of the multinational enterprises and learn from ­ World Bank and the International Finance them, usually by leveraging the expertise of Corporation (World Bank Group) that staff who have had experience working with, focuses on ­ helping countries attract ­ foreign or for, such enterprises. direct ­ investment, supporting reforms in 13. The 2017 GIC Survey data were collected more than 100 c ­ ountries over three decades. through telephone interviews with 754 For more information, see the World Bank’s business executives involved with o ­ ­ perations FIAS web page: https://www.worldbank​ .org​ in developing countries (picked from among /­e n​/ topic/competitiveness/brief​/­f acility-for​ 8,000 eligible companies in the Dun & -investment-climate-advisory-services-fias. Bradstreet database). The survey ­ captures 4. WAIPA, the World Association of Investment perceptions of international business execu- Promotion Agencies, is an international non- tives on the role that investment climate fac- governmental organization, established in tors play in their FDI decisions. 1995 by the United Nations Conference on 14. The 2019 GIC Survey data were collected Trade and Development (UNCTAD), that through telephone interviews with over acts as a forum, provides networking, and 2,400 business executives of MNE affiliates promotes best practices for IPAs. For more in 10 developing countries: Brazil, China, information, see the WAIPA website: https:// India, Indonesia, Malaysia, Mexico, Nigeria, waipa.org/. Thailand, Turkey, and Vietnam. The survey 5. For this section, the authors reviewed a wide captures perceptions of international b ­ usiness range of resources to provide a broad over- executives on the role that investment climate view of the literature on investment promo- factors play in their FDI decisions. tion and then classified them by topic but not 15. IPA data are from the 2005 World Bank by empirical strength. Group IPA Census and the 2017 World Bank 6. Crescenzi, Di Cataldo, and Giua (2019a) Group Global IPA Survey (see annex 5B). find a positive impact of subnational IPAs Regarding the comparison in IPA budget size, in attracting FDI and a mixed impact of the 2005 World Bank Group IPA Census asks national IPAs, based on empirical research in about the “FDI promotion” budget, while Europe. the 2017 World Bank Group Global IPA 7. “Cultural distance” is defined as differences Survey asks about the “investment promo- between one country and the United States tion” budget. Values are presented in 2018 in terms of language and business conduct. US dollars, adjusted for currency convertibil- The 2011 study by Harding and Javorcik ity and inflation. uses US FDI outflows, for which cultural 16. FDI project growth from 2009 to 2018 was cal- distance from the United States might be culated from the Financial Times’ fDi Markets 2 0 4    Global Investment Competitiveness Report 2019/2020 dataset (https://www.fdimarkets.com/), repre- within its 20-year plan (2000–20), it is now senting that survey’s coverage of more reliable preparing a 30-year plan (2020–50). data (through systematic collection of informa- 24. These reforms, which continue under the tion on FDI project announcements). Data to current government, have helped the c ­ ountry measure changes in IPA sector targeting were move toward needed labor-intensive indus- sourced from the 2005 World Bank IPA Census achievement in Sub- trialization (a significant ­ and a combination of 2017/18 IPA surveys from Saharan Africa) and contributed to boosting the World Bank Group, World Association of FDI inflows from US$279 ­ million in 2012 to Investment Promotion Agencies (WAIPA), and US$3.6 billion in 2017 (UNCTAD 2018). Organisation for Economic Co-operation and 25. This finding is based on two-hour, one-on- Development (OECD). (For more information, one detailed interviews carried out with the see annex 5B.) mentioned IPAs in 2019. 17. GIPB reports are based on the review of each 26. A Likert scale is a psychometric scale widely IPA’s website and its responses to investor used to gauge responses in survey research. requests for information. These inquiries are Specifically, it measures how people feel made using a “mystery shopper” approach, about something by asking respondents to whereby a global site selection firm submits choose from five to seven balanced responses. GIPB inquiries as the inquiries of a supposed 27. The analysis drew on the full datasets (only anonymous investor. Each IPA receives a available internally) of the 2005 World Bank confidential report with results. Group IPA Census; the 2017 World Bank 18. The GIPB was discontinued after 2012. There Group Global IPA Survey; and the 2017 have been no similar evaluation data since then. and 2018 World Association of Investment 19. The question in the 2017 World Bank Group Promotion Agencies (WAIPA) Annual Global IPA Survey was phrased as an open- Surveys. More information on the 2005 WBG ended question to IPAs to write in which IPA Census, and the WAIPA 2018 Annual sectors they are prioritizing. Consequently, ­ Survey can be found in Harding and Javorcik there were variations in how sectors were (2011), and WAIPA (2019), respectively. Any described and classified. But the numbers data leveraged from the 2018 OECD report, still indicate a comparable m ­ agnitude of “Mapping of Investment Promotion Agencies the ­ number of s ­ectors considered to be a in OECD Countries” were drawn from the “­priority” for investment promotion efforts. figures and charts included in the publicly 20. Rwanda’s FDI inflows have increased over available publication (OECD 2018) as the time, from US$14 million in 2005 to US$398 authors did not have access to the full dataset. million in 2018, according to UNCTAD World Investment Report data. 21. The World Bank Group has over 30 years of operational experience in the field of invest- References ment promotion, providing assistance to coun- Anderson, John, and Dylan Sutherland. 2015. tries across the world. “Developed Economy Investment Promotion 22. Countries are increasingly linking their Agencies and Emerging Market Foreign national plans with supranational visions Direct Investment: The Case of Chinese FDI and development plans. For instance, the in Canada.” Journal of World Business 50 (4): Association of Southeast Asian Nations 815–25. (ASEAN) community is working toward its Bauerle Danzman, Sarah, and Geoffrey Gertz. common Vision 2025. Rwanda has supra- Forthcoming. “Facilitating Sustainable national commitments through the ­ African Investment: The Role and Limits of Investment Union Agenda 2063 and East African Promotion Agencies.” In World Trade Forum Community Vision 2050. ­ 2018: Trade, Development, and Sustainability. 23. Malaysia developed “Wawasan 2020” Cambridge: Cambridge University Press. (Vision 2020), a 30-year national develop- Charlton, Andrew, and Nicholas Davis. 2007. ment plan spanning 1991–2020. Colombia’s “Does Investment Promotion Work?” The B.E. National Development Plan 2018–2022 Journal of Economic Analysis & Policy 7 (1): and Myanmar’s Sustainable Development 1–21. Plan 2018–2030 take steps toward the SDG Cho, Joong-Wan. 2003. “Foreign Direct targets for 2030. Although Rwanda is still Investment: Determinants, Trends in Flows and I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    205 Promotion Policies.” Investment Promotion Filippov, Sergey, and José Guimon. 2012. “From and Enterprise Development Bulletin for Asia Quantity to Quality: Challenges for Investment and the Pacific 1: 99–112. Promotion Agencies.” Institutions and Econo- Crescenzi, Riccardo. 2018. “Why Foreign mies 4 (2): 25–44. Investment Clicks in Some Cities and Regions, Freund, Caroline, and Theodore H. Moran. 2017. while Others Are Left Behind.” Global “Multinational Investors as Export Superstars: Investments and Local Development (blog), How Emerging-Market Governments Can London School of Economics, November 30. Reshape Comparative Advantage.” Working https://blogs.lse.ac.uk/gild/2018/11/30/why​ Paper 17-1, Peterson Institute for International -foreign-investment-clicks-in-some-cities-and​ Economics, Washington, DC. -regions-while-others-are-left-behind/. Harding, Torfinn, and Beata S. Javorcik. 2011. Crescenzi, Riccardo, Marco Di Cataldo, and Mara “Roll Out the Red Carpet and They Will Come: Giua. 2019a. ”FDI Inflows in Europe: Does Investment Promotion and FDI Inflows.” The Investment Promotion Work?” Working Paper Economic Journal 121 (557): 1445–76. 10/2019, Institute of Global Affairs, London Harding, Torfinn, and Beata S. Javorcik. 2012. School of Economics and Political Science. “Investment Promotion and FDI Inflows: Crescenzi, Riccardo, Marco Di Cataldo, Quality Matters.” CESifo Economic Studies and Mara Giua. 2019b. “Governments 59 (2): 337–59. Strive to Attract Investment, but So Heilbron, Armando, and Yago Aranda-Larrey. F a r T h e r e ’s N o E v i d e n c e o f W h a t 2020. “Strengthening Service Delivery Works.” LSE Business Review (blog), of Investment Promotion Agencies: The February 22. https://blogs.lse​ . a c . u k​ Comprehensive Investor Services Framework.” /businessreview/2019/02/22/governments​ Investment Promotion for Impact Series. -strive-to-attract-investment-but-so-far​ Investment Climate In Focus Note, World -theres​-no-evidence-of-what-works/. Bank, Washington, DC. Crescenzi Riccardo, Carlo Pietrobelli, and Roberta Heilbron, Armando, and Robert Whyte. 2019. Rabellotti. 2014. “Innovation Drivers, Value “Institutions for Investment: Establishing a Chains and the Geography of Multinational High-Performing Institutional Framework for Corporations in Europe.” Journal of Economic Foreign Direct Investment (FDI),” Investment Geography 14 (6): 1053–86. doi:10.1093/jeg​ Climate In Focus Note, World Bank, /lbt018. Washington, DC. DCI (Development Counsellors International). IDA Ireland. n.d. http://www.idaireland.com. 2017. “A View from Corporate America: IDA Ireland. 2018. “IDA Ireland Annual Report Winning Strategies in Economic Development & Accounts 2018.” IDA Ireland, Dublin. Marketing.” Survey report, DCI, New York. Lim, Sung-Hoon. 2018. “Determinants of DIT (United Kingdom, Department for International the Performance of Investment Promotion Trade). 2018. “Estimating the Economic Agencies: Evidence form a Mix of Emerging Impact of FDI to Support the Department Economies.” Emerging Markets Finance for International Trade’s Promotion Strategy: and Trade 54 (8): 1907–23. doi: 10.1080​ Analytical Report.” DIT, London. /1540496X.2017.1334144. DIT. 2019. “Department for International Loewendahl, Henry. 2001. “A Framework for Trade Inward Investment Results 2018–19.” FDI Promotion” Transnational Corporations Statistical report, DIT, London. https://www​ 10 (1): 1–42. .gov.uk/government/collections/dit-inward​ MIGA (Multilateral Investment Guarantee -investment-results. Agency). 2001. Investment Promotion Toolkit: ECORYS. 2013. “Exchange of Good Practice in A Comprehensive Guide to FDI Promotion. Foreign Direct Investment Promotion: A Study Washington, DC: World Bank. Carried Out under the Framework Contract MIGA. 2006a. “The Impact of Intel in Costa ENTR/2009/033.” Final report, 2nd rev., Rica: Nine Years after the Decision to Invest.” ECORYS, Rotterdam, the Netherlands. Working Paper No. 37402, Investing in FIAS (Foreign Investment Advisory Service). 2011. Development Series, World Bank Group, “Investment Generation Toolkit.” Online Washington, DC. resource (unavailable after 2017), World Bank, MIGA. 2006b. “Investment Promotion Agency Washington, DC. Performance Review 2006: Providing 2 0 6    Global Investment Competitiveness Report 2019/2020 Information to Investors.” Working Paper No. Pietersen, Pontévechio Hawarden, and Henri 37831, IPA Performance Review Series, World Bezuidenhout. 2015. “South African IPAs Bank Group, Washington, DC. Attracting FDI: Investment Promotion Miškinis, Algirdas, and Mariya Byrka. 2014. Strategies.” Journal of Applied Business “The Role of Investment Promotion Agencies Research 31 (3): 1057–72. in Attracting Foreign Direct Investment.” Spar, Debora. 1998. “Attracting High Technology Ekonomika 4 (93): 41–57. Investment: Intel’s Costa Rican Plant.” Monge-González, Ricardo, and Ezequiel Tacsir. Occasional Paper No. 11, Foreign Investment 2014. “Policy Coordination: From FDI to a Advisory Service, World Bank Group, Broader Framework to Promote Innovation— Washington, DC. The Case of Costa Rica.” In Science, Technology UNCTAD (United Nations Conference on Trade and Innovation Policies for Development: The and Development). 1997. “Survey of Best Latin American Experience, edited by Gustavo Practices in Investment Promotion.” Survey Crespi and Gabriela Dutrénit, 203–24. Basel, report, United Nations, New York and Geneva. Switzerland: Springer. UNCTAD. 2001. “The World of Investment Moran, Theodore H., Holder Görg, Adnan Promotion at a Glance: A Survey of Investment Serič, and Christiane Krieger-Boden. 2018. Promotion Practices.” Study, Advisory Services “Attracting FDI in Middle-Skilled Supply on Investment and Training (ASIT), United Chains.” Economics Discussion Papers No. Nations, New York and Geneva. 2018-2. Kiel Institute for the World Economy, UNCTAD. 2007. “Aftercare: A Core Function in Kiel, Germany.  Investment Promotion.” Investment Advisory Morisset, Jacques, and Kelly Andrews-Johnson. Series, Series A, No. 1, United Nations, Geneva 2004. “The Effectiveness of Promotion and New York. Agencies at Attracting Foreign Direct UNCTAD. 2008. “Evaluating Investment Investment.” Occasional Paper No. 16, Foreign Promotion Agencies.” Investment Advisory Investment Advisory Service, World Bank Series. Series A, No. 3, United Nations, New Group, Washington, DC. York and Geneva. Nelson, Roy. 1999. “Intel’s Site Selection Decision UNCTAD. 2013. “Optimizing Government in Latin America.” Global Business Case No. Services: A Case for Joint Investment and Trade A03-99-0016, Thunderbird School of Global Promotion?” IPA Observer No. 1, UNCTAD, Management, Arizona State University, Tempe. Geneva. Nelson, Roy. 2000. “Case Study: Intel’s Site UNCTAD. 2018. World Investment Report 2018: Selection Decision in Latin America.” Investment and New Industrial Policies. New Thunderbird International Business Review York and Geneva: United Nations. 42 (2): 227–49. UNCTAD. 2019. World Investment Report 2019: Nelson, Roy. 2005. “Competing for Foreign Direct Special Economic Zones . New York and Investment: Efforts to Promote Nontraditional Geneva: United Nations. FDI in Costa Rica, Brazil, and Chile.” Studies UNESCO (United Nations Educational, Scientific in Comparative International Development and Cultural Organization). n.d. Institution for 40 (3): 3–28. Statistics. https://uis.unesco.org. Nelson, Roy C. 2009. Harnessing Globalization: VCC and WAIPA (Vale Columbia Center and World The Promotion of Nontraditional Foreign Direct Association of Investment Promotion Centers). Investment in Latin America. University Park, 2010. “Investment Promotion Agencies and PA: The Pennsylvania State University Press. Sustainable FDI: Moving toward the Fourth OECD (Organisation for Economic Co-operation Generation of Investment Promotion.” Report and Development). 2018. “Mapping of of the findings of the VCC-WAIPA Survey on Investment Promotion Agencies in OECD Foreign Direct Investment and Sustainable Countries.” Inventory and survey report, OECD, Development, New York and Geneva. Paris. Volpe Martincus, Christian, and Monika Ortega, Celia, and Carlos Griffin. 2009. Sztajerowska. 2019. “How to Solve the “Investment Promotion Essentials: What Sets Investment Promotion Puzzle: A Mapping the World’s Best Investment Facilitators Apart of Investment Promotion Agencies in Latin from the Rest?” Investment Climate In Practice American and the Caribbean and OECD Note No. 6, World Bank, Washington, DC. Countries.” Monograph, Organisation for I n cr e a s i n g t h e D e v e l o p m e n t I m p a c t o f I n v e s t m e n t Pr o m o t i o n Ag e n ci e s    207 Economic Co-operation and Development Whyte, Robert, Celia Ortega, and Carlos and Inter-American Development Bank, Griffin. 2011. “Investment Regulation and Washington, DC. http://dx.doi.org/10.18235​ Promotion: Can They Coexist in One Body?” /0001767. In Practice Note No. 16, World Bank, WAIPA (World Association of Investment Washington, DC. Promotion Agencies). 2019. “Overview of World Bank. 2009. Global Investment Promotion Investment Promotion: Report of the Findings Benchmarking 2009 (Vol. 2): Summary Report. from the WAIPA Annual Survey of 2018.” Washington, DC: World Bank. WAIPA, Geneva. World Bank. 2012. Global Investment Promotion Best Wells, Louis T. Jr., and Alvin G. Wint. 2000. Practices 2012. Washington, DC: World Bank. “Marketing a Country: Promotion as a Tool World Bank. 2018. Global Investment Competitive- for Attracting Foreign Investment.” Revised ness Report 2017/2018: Foreign Investor Perspec- ed. Occasional Paper No. 13, Foreign tives and Policy Implications. Washington, DC: Investment Advisory Service, World Bank World Bank. doi:10.1596/978-1-4648-1175-3. Group, Washington, DC. World Bank. 2020. World Development Report Whyte, Robert, and Carlos Griffin. 2014. 2020: Trading for Development in the Age of “Promoting Foreign Investment in Fragile Global Value Chains. Washington, DC. and Conflict-Affected Situations.” Investment World Bank World Development Indicators Climate In Practice Note No. 22, World Bank, database. http://datatopics.worldbank.org​ Washington, DC. /world-development-indicators/. Glossary Bilateral investment treaty (BIT). A bilateral investment treaty is an agreement between two countries establishing the terms and conditions for private investment by an entity of one country in another country. Brownfield FDI. Brownfield FDI refers to any purchase of more than 10 percent of a target company’s assets by a foreign entity. Ten percent is the threshold for a foreign investment to be considered direct (FDI), according to the International Monetary Fund (IMF) and the Organisation for Economic Co-operation and Development (OECD). The purchase can be friendly or unfriendly and result in various combinations of outcomes in terms of creating a new legal entity, including a simple acquisition or a merger. Joint ventures do not fall under the category of brownfield foreign investment because they refer to the establishment of new facilities—greenfield investment—involving a local and a foreign entity. Developed countries. Developed countries include high-income countries as classified within the World Bank’s country and lending groups. Developing countries. Developing countries include low- and middle-income countries as classified within the World Bank’s country and lending groups. Doing Business. This World Bank project provides objective measures of business regulations and their enforcement across 190 economies and selected cities at the subnational and regional levels. Launched in 2002, the project looks at domestic small and medium-size companies and measures the regulations applicable to them throughout their life cycles. Efficiency-seeking FDI. Efficiency-seeking FDI occurs when investors seek to increase the cost efficiency of production by taking advantage of location-specific factors. These investors are also known as “cost-competitive investors,” and their main investment motivations include lowering production costs and establishing a new base for exports. FDI inflows. FDI inflows comprise 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. 209 2 1 0    G l o s s a r y FDI outflows. FDI outflows comprise 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 stock. According to the OECD, FDI stock measures total direct investment at a given point, usually at the end of a quarter or year. It represents the value of the resident investors’ equity in and net loans to enterprises resident in the reporting economy. Foreign affiliates. “Foreign affiliates” refers generically to various types of entities that a foreign investment might take. These affiliates may be subsidiaries, branches, or any other enterprise resident in a host country that is controlled by a nonresident institutional unit. Foreign direct investment (FDI). According to the IMF, FDI is a category of international investment made by a resident entity in one economy to establish a lasting interest in an enterprise resident in an economy other than the investor’s. A “lasting interest” refers to a long-term relationship between the direct investor and the enterprise as well as a significant degree of influence by the direct investor on the management of the direct investment enterprise. Components of FDI include equity, intracompany debt, and reinvested earnings. Global Investment Competitiveness (GIC) Survey. The GIC Survey is a World Bank survey of executives of the affiliates of multinational enterprises (MNEs) in developing countries. For the purposes of this report, the 2019 GIC Survey data cover more than 2,400 foreign investors with operations in 10 middle-income countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. Global value chain (GVC). A GVC refers to the series of stages required to produce a good or service that is sold to consumers, with each stage adding value and with at least two stages conducted in different countries. Government effectiveness. Part of the World Bank’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. Greenfield FDI. In greenfield FDI, the investor builds its business operations from the ground up. In this report, greenfield refers to a mode of FDI entry whereby a foreign investor builds its operations in a host economy. GVC activities (or stages). GVC activities, or stages, are those required to produce a good or service in the context of a global value chain. Spread across several locations, these activities span the conception of the good or service to its end use and include research, design, production, marketing, and distribution. GVC participation (or integration). GVC participation, or integration, refers to the engagement of a country, sector, or firm in at least one stage of a global value chain. Overall participation may take the form of two broad types: backward or forward participation. High-income countries. For the World Bank fiscal year 2020, high-income economies are defined as those with a gross national income (GNI) per capita (calculated using the World Bank Atlas method) of US$12,376 or more in 2018. Home economy. The home economy is the country of origin of the foreign investment. G l o s s a r y   2 11 Host economy. The host economy is the country that receives the foreign investment. International investment agreement (IIA). An IIA is a type of treaty between states that addresses issues on cross-border investments. IIAs exist on three levels: bilateral (such as BITs); 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. Investment incentives are measurable economic advantages that governments offer to specific enterprises or groups of enterprises to steer investments into preferred sectors or locations. These benefits can be either fiscal (for example, tax concessions) or nonfiscal (for example, loans or rebates). Investment linkages. Investment linkages are transmissions of foreign knowledge and practices that may improve the production capabilities of domestic suppliers, as a result of contractual arrangements between local suppliers and multinational corporations. Investment promotion agency (IPA). An IPA is a government agency or nonprofit organization whose mandate is to attract investment to the host economy. Investment protection guarantees. An investment protection guarantee is a guarantee or insurance provided by law, government, multilateral agency, or any party for an investment made. Lead firm. A lead firm is the hierarchically dominant actor within a GVC. Lower-middle-income countries. For the World Bank fiscal year 2020, lower-middle-income economies are defined as those with a GNI per capita (calculated using the World Bank Atlas method) between US$1,026 and US$3,995 in 2018. Low-income countries. For the World Bank fiscal year 2020, low-income economies are defined as those with a GNI per capita (calculated using the World Bank Atlas method) of US$1,025 or less in 2018. Mergers and acquisitions (M&A). M&A refers to transactions that result in the consolidation of companies or assets. Multinational enterprise (MNE). An MNE is an enterprise that has operations in more than one country and usually has a centralized head office that coordinates global management. Parent company. The parent company is the institutional unit that owns enough interest in another firm to manage or operate the firm. Regulatory risk. As defined for the purposes of this report, regulatory risk is a subset of political risk related to select features of countries’ regulatory frameworks that can reduce risks for investors and limit the potential for unexpected losses due to arbitrary government conduct. Specifically, the new regulatory risk measure introduced in this report examines (a) the level of transparency in both the content and process of making laws and regulations that apply to investors; (b) the extent of legal protection provided to investors against arbitrary, unpredictable, and nontransparent government interference; and (c) the existence of effective recourse mechanisms for investors. Reinvested earnings. Reinvested earnings are net earnings not paid out as dividends but retained by the firm for reinvestment in its business operations in the host country. 2 1 2    G l o s s a r y Trade diversion. Trade diversion is the process of diverting trade from a more efficient exporter to a less efficient one by means of a free trade agreement or a customs union. For example, when two countries sign a trade agreement, they could reduce their imports from the rest of the world and source their imports from each other. To the extent that this strategy of import reallocation has been triggered by the trade agreement, it can be considered a trade diversion. Upper-middle-income countries. For the World Bank fiscal year 2020, upper-middle-income economies are defined as those with a GNI per capita (calculated using the World Bank Atlas method) between US$3,996 and US$12,375 in 2018. 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. 2019I 2020 GLOBAL INVESTMENT COMPETITIVENESS REPORT The Global Investment Competitiveness Report 2019/2020 provides novel analytical insights, empirical evidence, and actionable recommendations for governments seeking to rebuild investor confidence in times of uncertainty. It focuses on the role of foreign direct investment (FDI) in alleviating the impact of the COVID-19 crisis and boosting countries’ economic resilience. It highlights FDI’s contributions to providing a critical source of external finance, creating jobs, lifting people out of poverty, and raising productivity. The report presents the results of a survey of more than 2,400 business executives representing multinational corporations in 10 large developing countries: Brazil, China, India, Indonesia, Malaysia, Mexico, Nigeria, Thailand, Turkey, and Vietnam. Results of the survey, as well as the report’s new global database of regulatory risk, highlight the critical role of government actions in reducing investor risk and increasing policy predictability for rebuilding investor confidence. The report also assesses the impact of FDI on poverty, inequality, employment, and business performance, using firm- and household-level evidence from various countries. It shows that FDI in developing countries yields benefits to firms and workers—including more and better-paid jobs—but governments need to remain vigilant about possible adverse consequences for income distribution. Lastly, the report articulates priorities for investment promotion agencies and other stakeholders seeking to strengthen their countries’ investment competitiveness and leverage FDI for a robust economic recovery. www.worldbank.org/gicreport UK Government ISBN 978-1-4648-1536-2 90000 9 781464 815362 SKU 211536