92767 Jobs or Privileges MENA DEVELOPMENT REPORT Jobs or Privileges Unleashing the Employment Potential of the Middle East and North Africa Marc Schiffbauer Abdoulaye Sy Sahar Hussain Hania Sahnoun Philip Keefer © 2015 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 17 16 15 14 This work is a product of the staff of The World Bank with external contributions. The findings, interpre- tations, 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: Schiffbauer, Marc, Abdoulaye Sy, Sahar Hussain, Hania Sahnoun, and Philip Keefer. 2015. Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa. Washington, DC: World Bank. doi: 10.1596/978-1-4648-0405-2. 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 indi- vidual 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 the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. ISBN: 978-1-4648-0405-2 eISBN: 978-1-4648-0406-9 DOI: 10.1596/978-1-4648-0405-2 Cover art: © Aladdin El-Gendy. Used with permission. Further permission required for reuse. Library of Congress Cataloging-in-Publication Data has been requested Contents Acknowledgments xi About the Authors and Contributors xiii Abbreviations xvii Overview xix Introduction 1 Notes 7 1 Too Little Too Late: Private Sector Growth and Labor Demand 9 Economic Growth Has Been Moderate and Job Growth Weak 10 Drivers of Job Growth: Young Firms and Productive Firms Create More Jobs 24 MENA Needs a Larger Pool of Young Firms and Productive Firms 33 Notes 42 References 45 2 Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 49 Attracting FDI in Services Sparked Job Growth in Domestic Firms in Jordan 51 Business Regulations Limit Employment Growth among Young Firms in Morocco 61 Energy Subsidies in the Arab Republic of Egypt Discourage Growth in Labor-Intensive Industries 64 Discriminatory Policy Implementation Deters a Level Playing Field in MENA 68 Notes 77 References 79 v vi Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa 3 Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 81 Industrial Policy in MENA Has Had Limited Success and Many Instances of Policy Capture 84 What Did Successful Countries Do? The Case of the Republic of Korea 92 Lessons from East Asia Are More Difficult to Implement than Is Commonly Understood 96 Notes 98 References 99 4 Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 101 Privileges to Politically Connected Firms Undermine Competition and Job Creation: Evidence from the Arab Republic of Egypt and Tunisia 103 Available Qualitative Evidence Points to Similar Mechanisms of Policy Privileges in Other MENA Countries 131 What Explains the Different Outcomes in MENA and East Asia? 139 Notes 142 References 146 Implications for Policy 149 Appendix A Economic Growth and Structural Transformation 153 Appendix B Firm Censuses and Surveys: Countries, Time, and Sector Coverage 165 Appendix C Share of Employment in Large Firms among State-Owned Enterprises and Foreign Firms 169 Appendix D Employment Growth over Firms’ Life Cycles: Manufacturing Sector 171 Appendix E FDI Inflow and Employment in Jordan: Regression Analysis 173 Appendix F Quality of Business Environment and Jobs in Morocco: Data, Methods, and Main Findings 175 Appendix G Political Connections and Private Sector Growth in the Arab Republic of Egypt 179 Contents vii Boxes 1.1 Is Structural Change in Morocco Gender-Biased? 16 1.2 Who Create More Jobs? 25 1.3 Firm Dynamics and Productivity Growth in Morocco 41 2.1 FDI into Services Sectors Is Often Restricted in MENA Countries 54 2.2 Mobility Restrictions Reduces Competition and Job Growth in the West Bank 58 2.3 Misallocation of Capital in the Arab Republic of Egypt 67 3.1 Market Failure and Industrial Policy 82 3.2 Are GCC Countries an Exception? 90 4.1 Did Ben Ali Firms Dictate Amendments to the Investment Law in the 2000s? 119 4.2 Political Connections and Patronage in the Republic of Yemen 129 4.3 The Islamic Republic of Iran: Privatizations without the Private Sector 136 Figures 1.1 Decomposition of GDP Per Capita Growth in MENA and Other Developing Regions 11 1.2 Structural Change across Regions and among MENA Countries, 2000–05 14 1.3 Correlation between Entry and Exit Rates across Two-Digit Sectors 15 B1.1.1 Reallocation of Labor across Sectors, by Gender, 2000–11 17 1.4 Demographic Change and Composition of Working-Age Population 18 1.5 Employment Share, by Firm Size 19 1.6 Distribution of Employment across Nonagriculture Sectors 19 1.7 Percentage Change in Share of Employment in Medium and Large Establishments 20 1.8 Employment Transition, by Firm Size 22 1.9 Incidence of Gazelles in All Sectors and Manufacturing 23 1.10 Share of Jobs Created by Gazelles and Nongazelles in All Sectors and in the Manufacturing Sector 23 1.11 Net Job Creation, by Firm Size and Age 26 1.12 Net Job Creation, by Firm Size before and after Controlling for Firm Age 28 1.13 Net Job Creation, by Firm Age after Controlling for Firm Size 29 viii Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa 1.14 Employment Growth Is Strongest in First 4–5 Years after Firm Entry 30 1.15 Employment Growth over a Firm’s Life Cycle for All Nonfarm Sectors 31 1.16 Characteristics of Gazelles in the Arab Republic of Egypt, Lebanon, and Morocco 32 1.17 Entry Density of Formal Sector Limited-Liability Firms across Regions and Countries, 2004–12 34 1.18 Firm Turnover across Countries 35 1.19 Survival Rates Five Years after Entry 36 1.20 Employment Share of Young Medium or Large Establishments 36 1.21 Distribution of Employment, by Firm Size and Age across All Nonfarm Establishments: Arab Republic of Egypt and Turkey, 2006 37 1.22 Labor Productivity Growth over the Life Cycle of Manufacturing Establishments 38 1.23 Gap between Weighted and Unweighted Labor Productivity 40 1.24 Productivity, by Firm Size in Tunisia, Lebanon, and Turkey 40 B1.3.1 Decomposition of Firm Productivity Growth in Morocco’s Manufacturing Sector, 1996–2006 42 2.1 Share of FDI Inflows, by Sector, Selected MENA Countries, 2003–10 53 B2.1.1 Service Trade Restriction Index, by Sector and Region 54 B2.1.2 Service Trade Restrictions in Transportation Services in MENA, 2008 55 B2.2.1 Mobility Restrictions Reduce Net Entry, Employment Growth, and Local Output Growth 60 2.2 Distribution of Employment, by Energy Intensity and Size and Age 65 2.3 Employment Share, by Sector Factor Intensity in the Arab Republic of Egypt, 2006, and Turkey, 2010 66 B2.3.1 Productivity, in Manufacturing and Mining, by Size 67 2.4 Regulatory Policy Implementation Uncertainty in MENA 69 2.5 Variability in Days to Accomplish Various Regulatory Tasks across Firms, Selected MENA Countries 72 3.1 The Evolution of Average (Weighted) Tariffs and NTMs on Imports, 1995–2010 87 4.1 The Evolution of Net Profit Differentials between Connected and Other Firms, 2003–11 110 Contents ix 4.2 Authorization Requirements and FDI Restrictions Protect Politically Connected Firms in Tunisia 112 4.3 Share of Politically Connected Firms in High and Low Energy-Intensive Sectors in the Arab Republic of Egypt 115 4.4 Large Firms in Politically Connected Industries Are More Likely to Be Located in an Industrial Zone 117 4.5 Nontariff Barriers Are Frequently Imposed in MENA 132 4.6 Transparency International: Defence Anti-Corruption Index 133 4.7 Perceptions of Corruption in Government and Business, Middle East and North Africa, 2011 138 4.8 Worldwide Governance Indicators 139 A.1 Real GDP Per Capita Growth Decomposition 154 A.2 Long-Term Structural Change in Four MENA Countries 155 A.3 Structural Change, by Sector, 2000–05 156 A.4 The Product Space 161 A.5 Product Space in Selected Regions, 1976–78 and 2007–09 162 C.1 Number of Firms and Jobs in Foreign, Domestic Private, or Public Establishments 169 D.1 Manufacturing: Employment Growth over the Life Cycle 171 Map B2.2.1 Mobility Restrictions in the West Bank, 2006 and 2011 59 Tables 1.1 Employment Transition Matrix 21 1.2 Sectors with the Highest Rate of Job Growth across Countries 27 1.3 More Productive Firms Create More Jobs 32 2.1 Averages and Dispersion of Firms’ Waiting Days for Regulatory Services 70 2.2 Share of Firms That Disagree with the Statement That Implementation of Rules Is “Consistent and Predictable” 73 2.3 The Extent to Which Firms Take Action to Influence Policy Implementation across Types of Firms in MENA 74 2.4 Higher Policy Implementation Uncertainty Induces Senior Managers to Spend More Time with Government Officials 75 x Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa 2.5 Policy Implementation Uncertainty Reduces Innovation and Firm Growth in Jordan and the Arab Republic of Egypt 76 3.1 The Cost of Industrial Policy in Morocco, 2010 89 4.1 Number of Politically Connected Firms, by Economic Sectors 107 4.2 Within-Sector Differences, Politically Connected and Other Firms 109 4.3 Politically Connected Firms and All Firms Protected by Nontariff Trade Barriers in the Arab Republic of Egypt 113 4.4 Government Relations and Competition in Sectors with Politically Connected Firms versus Nonconnected Sectors in the Arab Republic of Egypt 116 B4.1.1 Correlation between New Barriers to Entry and the Presence of Ben Ali Firms 120 4.5 Firm Dynamics in Sectors with Politically Connected Firms versus Nonconnected Sectors in the Arab Republic of Egypt 125 4.6 Financial Corruption Risk Subindex: Asset Disposal and Links to Business, MENA Countries 135 A.1 Estimates of Labor’s Share and Marginal Productivities Using Harmonized Household Survey Data from the World Bank (I2D2) 158 A.2 Manufacturing Labor Productivity Growth Rates 159 E.1 Employment Spillovers from FDI, by Firm Characteristics 173 F.1 List of Regulatory Policy Variables 177 F.2 Job Growth Regression with Coefficients of the Policy and Environment Variables 177 G.1 Employment Growth Declines after Politically Connected Firms Enter Initially Unconnected Sectors 180 G.2 Entry of Connected Firms into Initially Unconnected Sectors, 1997–2006 180 Acknowledgments This report was prepared by a team composed of Marc Schiffbauer (task team leader), Abdoulaye Sy, Sahar Hussain, Hania Sahnoun, and Philip Keefer. In addition, the following persons contributed to the individual chapters of the report: Ishac Diwan (Chapter 4), Doerte Doemeland (Chapter 1), Bob Rijkers (Chapters 1 and 4), Dalia Al Kadi (Chapter 1), Izak Atiyas (Chapter 1), Ozan Bakis (Chapter 1), Michael Lamla (Chapter 2), and Michael Gasiorek (Chapter 2). Further inputs were provided by Jamal Haider, Hassen Arouri, Huy Nguyen, Karim Badr, Anna Raggl, Yeon Soo Kim, and Caroline Duclos. Clifton Wiens edited the report. Aladdin El-Gendy produced the cover picture of the report. Muna Abeid Salim, Seraphine Nsabimana, and Faythe Agnes Calandra provided administrative support. The report was prepared under the direction of Bernard Funck. The report benefited from the overall guidance of Shantayanan Devarajan, Chief Economist of the World Bank Middle East and North Africa region, as well as from Caroline Freund and Manuela Ferro. The team thanks Najy Benhassine (Practice Manager, Trade and Competitiveness, World Bank), Hafez Ghanem (Senior Fellow, Brookings Institution), and Adeel Malik (Research Fellow in Economics, University of Oxford) for their valuable comments. The report also ben- efited from comments and guidance from Kevin Carey, Ahmed Kouchouk, Tara Vishwanath, Celestin Monga, Daniel Lederman, Mary Hallward- Driemeier, Jorge Araujo, Nikola Spatafora, Aaditya Mattoo, Ana Fernandes, Peter Mousley, Simon Bell, and Randa Akeel. The team is grateful to the various statistics and research institutions across the region that facilitated access to data and collaborated with the team, including the Institut National de la Statistique in Tunisia; the Ministry of Planning and International Coordination and the Department of Statistics in Jordan; the Central Agency for Public Mobilization and Statistics in the Arab Republic of Egypt; and the Economic Research xi xii Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Forum team, in particular Ahmed Galal (Managing Director) and Hoda Selim (Economist). The team also thanks Antonio Nucifora, Eric Le Borgne, Orhan Niksic, Sibel Kulaksiz, Umar Serajuddin, Nour Jalal Nasser Eddin, Nada Choueiri, Amir Mokhtar Althibah, and the entire Poverty Reduction and Economic Management Department for supporting the country dialogue and for their collaboration during the preparation of various country analytical works undertaken in parallel to this regional report. About the Authors and Contributors Marc Schiffbauer is a senior economist and part of the team preparing the World Development Report 2016 on Internet for Development. He joined the World Bank in September 2009, working in the Poverty Reduction and Economic Management unit in the Eastern Europe and Central Asia region as well as in the Middle East and North Africa region. Before that, he worked for the Economic and Social Research Institute in Dublin, Ireland, and as a consultant for the European Central Bank and the International Monetary Fund on issues related to economic growth, firm productivity, and competition. Marc has a PhD in economics from the University of Bonn in Germany and was a one-year visiting scholar at Universidad Pompeu Fabra, Barcelona, and at the University of British Columbia, Vancouver. Abdoulaye Sy is an economist in the Macro and Fiscal Management Global Practice at the World Bank and currently works in the Middle East and North Africa region where he is the country economist for the Islamic Republic of Iran and Djibouti. Abdoulaye joined the World Bank in September 2011 as a Young Professional, first as an economist in the Sustainable Development Department in the Latin America and Caribbean region, later joining the Poverty Reduction and Economic Management unit in the Middle East and North Africa region. Abdoulaye has a PhD in agricultural and resource economics from the University of California–Berkeley and holds master’s degrees in science and economics from the Ecole Polytechnique, the Paris School of Economics (PSE), and the Ecole Nationale de la Statistique et de l’Administration Economique (ENSAE). Sahar Hussain joined the World Bank as an economist in February 2013 in MENA’s Poverty Reduction and Economic Management unit. Prior to that, she worked for the Egyptian Centre for Economic Studies in Cairo as an economist on issues related to the economics of transitions, xiii xiv Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa competition policy, and energy subsidies. She was also an economic consultant for the Planning Commission of Pakistan. Sahar has a master’s degree in development economics and policy analysis from the University of Nottingham and a bachelor’s degree from the London School of Economics. Hania Sahnoun is an economist and consultant. She joined the team preparing the World Development Report 2016 on Internet for Development. She has worked with the World Bank as a consultant since 2004. Hania holds a Diploma of Advanced Studies in economics from the University of Pantheon–Paris I in France. Philip Keefer is a principal advisor of the Institutions for Development Department of the Inter-American Development Bank. He was formerly a lead research economist in the Development Research Group of the World Bank. The focus of his work, based on experience in countries such as Bangladesh, Benin, Brazil, the Dominican Republic, Indonesia, Mexico, Peru, and Pakistan, is the determinants of political incentives to pursue economic development. His research, on issues such as the impact of insecure property rights on growth; the effects of political credibility on policy; and the sources of political credibility in democracies and au- tocracies; and the influence of political parties on conflict, political bud- get cycles, and public sector reform, has appeared in journals such as the Quarterly Journal of Economics and the American Political Science Review. Ishac Diwan is currently a research fellow at Paris Sciences et Lettres. He taught previously at the Harvard University Kennedy School of Government and at New York University. He has held several positions at the World Bank—in the Research Complex, the Middle East Department, and the World Bank Institute—and in Addis Ababa then in Accra as the country director for countries in East Africa and then in West Africa. His current research interests include growth strategies, the politi- cal economy of private sector development, and the analysis of public opinions, with a special interest in Africa and the Middle East. He directs the Economic and Political Transformation program of the Economic Research Forum. Doerte Doemeland is a senior economist in the Macroeconomics and Fiscal Management Global Practice. Before that, Doerte was a senior economist in the Development Research Group of the World Bank. The focus of her work, based on experience in Albania, Bulgaria, Malawi, Mexico, Nigeria, Senegal, Tunisia, Uganda, and Uruguay, is on the de- terminants of economic growth and poverty reduction resulting from trade and competitiveness, structural change, or productivity growth. About the Authors and Contributors xv Dalia Al Kadi is an economist in the Macroeconomics and Fiscal Management Global Practice at the World Bank. She joined the World Bank’s Poverty Reduction and Economic Management unit in the Mid- dle East and North Africa Department in February 2013. Before joining the World Bank, Dalia was a project manager at the Executive Council in Abu Dhabi, where she advised the government on economic policy and strategy. She was also a management consultant for McKinsey & Com- pany and advised clients in the banking, telecom, and public sectors in the Middle East and Pakistan. Dalia has an MPA in international develop- ment from Harvard University. Bob Rijkers is an economist in the Trade and International Integration Unit of the Development Research Group at the World Bank. He is interested in political economy, trade and labor market issues. He holds a BA in science and social sciences from University College Utrecht, Utrecht University, and an M.Phil. and D.Phil. in economics from the University of Oxford. Abbreviations AISS Annual Industry and Service Statistics CAPMAS Central Agency for Public Mobilization and Statistics CPI Corruption Perceptions Index EAP East Asia and Pacific ECA Europe and Central Asia EIDS Egypt Industrial Development Strategy ES Enterprise Survey FDI foreign direct investment GCC Gulf Cooperation Council GDP gross domestic product ICA Investment Climate Assessment ICT information and communications technology LAC Latin America and Caribbean LMIC lower middle-income country MENA Middle East and North Africa MFTI Ministry for Trade and Industry MNE multinational enterprise NTM nontariff measure OECD Organisation for Economic Co-operation and Development PPP purchasing power parity SOE state-owned enterprise SSO Social Security Organisation TFP total factor productivity TI Transparency International WBES World Bank Enterprise Survey WDI World Development Indicators WITS World Integrated Trade Solution xvii Overview Middle East and North Africa (MENA) countries face a critical choice as they strive to generate greater private sector growth and more jobs: pro- mote competition, provide equal opportunities for all entrepreneurs, and dismantle the current system of privileges for connected firms or risk per- petuating the current equilibrium of low job creation. This report shows that policies that stifle competition and create an uneven playing field abound in MENA and are a major constraint on private sector growth and job creation. These policies take different forms across countries and sec- tors but share several common features: they limit free entry in the domes- tic market, exclude certain firms from government programs, increase the regulatory burden and uncertainty for firms without connections, insulate certain firms and sectors from foreign competition, and create incentives that discourage domestic firms from competing in international markets. The report shows that such policies are often captured by a few privileged firms with deep political connections and that these policies persist despite their apparent cost to society. The millions of workers and consumers and the majority of entrepreneurs who bear the brunt of that cost are often unaware of the adverse impact of these policies on the jobs and economic opportunities to which they aspire. This limits the scope for critical internal debate on the economic future of MENA countries and curtails the policy dialogue necessary for reform. Labor markets in MENA have underperformed for a long time. This has left large segments of the population on the sidelines of the economy and created a sense of exclusion. MENA has a large reservoir of untapped human resources; it has some of the world’s highest unemployment rates among college graduates and youth and the lowest participation of women in the labor force. Strategies focused on increasing employment in the public sector have proved to be unsustainable, and private sector job cre- ation has been too weak to absorb the growing labor force. Desirable private sector jobs—those with high wages, a formal contract, and social xix xx Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa The Impact of Privileges on Policies, Competition, and Jobs The findings of this report highlight some • An additional firm with a politically of the economic impacts of the privileges connected chief executive officer (CEO) granted to politically connected firms: reduces the average waiting time for a construction permit in an industry by • In the Arab Republic of Egypt, 71 percent 51 days. of connected firms, but only 4 percent of all firms, sell products that are protected • Firms in industries with at least one by at least three technical import politically connected CEO are inspected barriers. by tax officials 4.6 times a year relative to 5.7 times a year for firms in sectors • In Tunisia, 64 percent of politically con- without a connected CEO. In addition, nected firms operate in sectors subject to the frequency of inspections by munici- restrictions on foreign direct investment palities is about 20 percent higher for (FDI), relative to only 36 percent of non- firms in nonconnected industries. connected firms. • The dispersion of reported inspections • In Egypt, 45 percent of all connected across firms is significantly higher within firms operate in energy-intensive indus- connected sectors. This suggests that tries such as cement or steel, compared politically connected firms receive very with only 8 percent of all firms. few inspections while nonconnected firms are inspected frequently. • In Tunisia, 64 percent of politically con- nected firms are in sectors requiring an • The entry of new firms into politically exclusive license to operate relative to connected sectors is about 28 percent only 45 percent of nonconnected firms. lower than into nonconnected firms. • Firms in politically connected industries • Aggregate employment growth declines (that is, with at least one connected firm) by about 1.4 percentage points annually are 11–14 percent more likely to have when connected firms enter new, previ- acquired land from the government. ously unconnected sectors in Egypt. security benefits—are few, pushing a growing number of workers to seek employment in unproductive subsistence activities, often in the informal economy. This situation has contributed to the widespread frustration with the lack of opportunities, of which Arab Spring uprisings were a powerful expression. Previous World Bank reports have linked MENA’s employment per- formance to supply-side factors, labor market policies, and qualitative evidence of weak competition as a result of privileges for connected firms. Two past regional World Bank reports provide the starting point Overview xxi to this report. First, the World Bank report Jobs for Shared Prosperity (2013a) analyzed how supply-side factors such as education and training, and labor market policies affect employment outcomes in MENA. The report concluded, however, that supply-side factors only partially explain employment outcomes in MENA and highlighted the importance of analyzing demand-side factors to explain the weak private sector job cre- ation record. Second, the World Bank report From Privilege to Competition: Unlocking the Private-Led Growth in the Middle East and North Africa (2009) provides rich qualitative evidence that policy capture in MENA countries leads to privileges for a few politically connected firms, which ultimately limits competition and private sector development. The report argued that privileges to politically connected firms in MENA resulted in policies—such as subsidized land acquisitions and directed bank lending—that limited competition and tilted the playing field. The authors of this pre–Arab Spring report used all information available at the time, but did not have access to the full array of data necessary to investigate the possible link between MENA’s weak aggregate job cre- ation, the lack of a level playing field, and the absence of competition as a result of prevailing privileges and policy capture in many countries and sectors across the region. This report fills this gap by analyzing the demand-side factors that constrain faster job creation in MENA countries and how they relate to weak competition and privileges for specific firms. This report aims to answer the following questions: What types of firms create more jobs in MENA? Are they different from job-creating firms in other regions? What policies in MENA prevent the private sector from creating more jobs? How do these policies affect competition? To what extent are these policies associated with privileges to politically connected firms? This report aims to address these questions by drawing on new data sources that became available after the Arab Spring. First, the report assembles firm census databases for several MENA countries that contain a wide range of firm characteristics and performance measures. This rich source of information is necessary to determine the fundamental drivers of aggregate job creation through the lens of firms. Second, the report combines this information with additional data sources to analyze how certain policies affect competition and the fundamentals of job creation. Third, the report merges these data with new detailed information on state-business relations that surfaced after the Arab Spring. Specifically, the report builds on two novel data sets that identify the first-tier politi- cally connected firms in the Mubarak and Ben Ali regimes in the Arab Republic of Egypt and Tunisia, respectively. These unique data are used to analyze the methods and extent of policy capture by politically con- nected firms. The report then provides for the first time quantitative xxii Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa evidence that these privileges limit competition, firm dynamics, and job creation. Five main findings of the report stand out. First, gross domestic product (GDP) per capita growth in MENA over the last two decades was moderate and driven by demographic change, while productivity growth was low. Real GDP per capita growth hovered around 2 percent in the past two decades; about 2–3 percent lower than in South and East Asia, respectively, but comparable to per capita growth rates in other developing regions. Demographic change, leading to an increase in the share of the working age population, accounted for about 50 percent of aggregate real GDP per capita growth over the past 20 years, substantially higher than in any other region. In contrast, aggregate pro- ductivity growth was low in MENA compared with other developing regions. Most countries in the region experienced structural change because of a decline of the labor share in agriculture. Aggregate productiv- ity growth was, however, mostly driven by productivity growth within sectors, which still lagged behind all other developing regions. The eco- nomic benefits from the ongoing demographic trend could have been much higher had MENA countries been able to absorb their fast-growing labor force into the higher-productivity activities. Instead, job creation was weak and informality, unemployment, and inactivity reached very high levels during this timeframe. Consistent with this trend, analysis of firm census data shows that most workers in MENA are employed in small-scale and low-productivity activities. Second, the report examines whether the fundamentals of job creation— the types of firms that create more jobs—differ in MENA countries than in fast-growing emerging or even high-income countries in other regions. It shows that they do not: younger firms and more productive firms grow faster and create more jobs in MENA as elsewhere. For example, firm census data show that micro-startups—firms less than five years old and with less than five employees—accounted for 92 percent of net job cre- ation in Tunisia between 1996 and 2010 and 177 percent in Lebanon between 2005 and 2010. In addition, young firms across all size categories contributed positively to net job creation in both countries while employ- ment in older firms tended to contract. However, MENA countries’ private sectors have been characterized by low firm turnover (firm entry and exit) and slow productivity growth, which ultimately reduces the pool of young firms and more productive firms. For example, for every 10,000 working-age persons, on average only six limited liability companies were created annually in MENA countries between 2009 and 2012; in contrast, the average across all 91 developing countries with available data was 20 per 10,000 working-age persons, and as high as 40 and 80 in Chile and Bulgaria, respectively. Moreover, we find that productivity growth in MENA is held back by a combination of slow within-firm productivity Overview xxiii growth and misallocation of labor and capital across firms. For instance, after 35 years in operation, firms in Tunisia and Egypt barely increase their productivity while firms in India, Mexico, and Turkey increase their productivity about two- or threefold over the same life cycle. Third, various policies across MENA countries limit competition and undermine the fundamentals of job creation by constraining firm startup and productivity growth. The report presents four case studies that dem- onstrate how different policies across MENA countries limit competition and result in lower firm turnover, productivity growth, and job creation. The first case study shows how foreign direct investment (FDI) inflow in Jordan led to a partial crowding-out of old and small domestic firms oper- ating in the same sector, but had positive employment spillovers among domestic service providers and young firms. Domestic manufacturing firms (suppliers) did not benefit from FDI spillovers, possibly reflecting a combination of weak competition in the sector and the absence of well- designed and effective technical supplier support programs. Overall, the findings suggest that removing restrictions on FDI into service sectors in Jordan can be expected to generate employment growth among domestic firms. In the second case study, we explore how several dimensions of Morocco’s business environment impact employment growth and dis- proportionately affect young firms. The findings suggest that more competition, equal and predictable treatment by tax administrations, less corruption and fewer obstacles in the judicial system, and lower cost of finance would raise employment growth among young firms. The third case study examines how large energy subsidies targeted to heavy industry in Egypt (equivalent to 2.9 percent of GDP or US$7.4 billion in 2010) affect competition and job creation. A government license is required to legally operate in energy-intensive industries (such as steel and cement), thereby limiting the entry of new firms, equal access for all entrepreneurs, and competition. Moreover, energy subsidies benefit energy-intensive industries thereby discouraging more labor-intensive activities and pre- venting the economy from fully exploiting its comparative advantage. In the last case study, we show that many firms in MENA identify “policy uncertainty” as a “severe” or “major” obstacle to growth, and that this reflects firms’ perception of “policy implementation uncertainty” result- ing from discriminatory practices. The large variation in policy imple- mentation leads to reduced competition and innovation in a number of MENA countries. The findings reveal a negative impact of discrimina- tory policy implementation on productivity growth and private sector dynamism (specifically the entry of new firms) in MENA. Fourth, past industrial policies in MENA did not reward firms based on performance and did not safeguard or promote competition. Efforts to stimulate private sector growth and jobs in MENA have xxiv Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa often taken the form of active industrial policies. But there is limited evidence of success while there are several instances of policy capture by a few insider firms. The report reviews the impact of these policies over the past two decades and compares them with the experience of East Asian countries. This comparison highlights several critical dif- ferences in policy design and implementation that underpin the suc- cess of industrial policies in East Asian countries when compared with MENA countries. First, East Asian countries seemingly reached broader consensus on a common strategic vision and objectives at the country level, and had a greater focus on new economic activities in sectors where market failures were more likely to constrain industrial development. Second, industrial policy in East Asia was performance oriented, and evaluation systems to assess the performance of policies and public officials were put in place. Third, by linking government support to measurable and verifiable performance, industrial policies in East Asia guaranteed equal access for all firms, while in MENA it often resulted in privileges for a limited number of firms. Fourth, industrial policy in East Asia promoted and safeguarded competi- tion in the domestic market and provided incentives for firms to com- pete in international markets. East Asian countries invested heavily in human capital and complementary infrastructures improvements, and undertook far-reaching public sector reforms that created a qualified and merit-based public administration. Fifth, the report provides direct evidence that policies in MENA have often been captured by a few politically connected firms. This has led to a policy environment that created privileges rather than to a level playing field, and has undermined private sector growth and job creation. We show that these privileges insulated firms from domestic and interna- tional competition and subsidized their operations via preferential and sometimes exclusive access to cheap inputs (electricity, land, and so forth). Using the theoretical framework proposed by Aghion et al. (2001),1 we discuss how such policies are likely to reduce competition, undermine equal opportunity for all entrepreneurs, and result in lower efficiency, innovation, and job creation. The report documents how this was the case during the Mubarak and Ben Ali regimes in Egypt and Tunisia, respec- tively, and provides qualitative evidence for the existence of similar mechanisms in other MENA countries. For example, we find that only a handful of politically connected firms received the bulk of the generous energy subsides to industry in Egypt. Moreover, barriers to entry and trade in Egypt and Tunisia insulated politically connected firms from competition and tilted their incentives towards producing for the domes- tic market. These policies are typically still in place in both countries; they include exclusive operating licenses that create monopolies Overview xxv in profitable services sectors, unequal access to land, or inconsistent implementation of rules and regulations across firms in the same sector. Furthermore, the report argues that the concentration of connected firms in (nontradable) backbone service sectors in MENA—which lowers the performance of these sectors and increases the relative price of nontrad- able to tradable goods and services—contributes to the overvaluation of the exchange rate through the phenomenon of weak links.2 The report provides direct quantitative evidence that the preferential treatment of politically connected firms lowers aggregate job growth in Egypt. The available qualitative evidence points to similar mechanisms of policy privileges in other MENA countries. In particular, governance and cor- ruption indicators are higher in MENA than in other regions, especially in corruption in defense as a result of military involvement in business. The findings of this report have several implications for policy. This report suggests that MENA countries’ quest for more jobs should not only include supply-side policies—education, wages, job training—but should also encompass significant policy reforms to stimulate labor demand. The report’s findings point to a roadmap for more jobs in MENA in four broad policy areas. Depending on the country context, additional and more specific policy areas also need to be considered: • First, governments in MENA should reform policies that unduly con- strain competition and equality of opportunity for all entrepreneurs. These policies include energy subsidies to industry, exclusive licenses requirements to operate in specific sectors, legal barriers to FDI, and trade barriers—including nontariff measures, administrative barriers to entry and firm growth, and barriers to access the judiciary, land, or industrial zones. Reforms to other policies not analyzed in this report, but potentially equally important in maintaining a level playing field, should also be considered when dealing with specific country cases. These include barriers to firm entry and exit resulting from restrictive hiring and firing laws, cumbersome bankruptcy laws, and so forth. In addition, the findings of this report suggests that if MENA governments want to pursue private sector development programs targeting specific types of firms, they would be well advised to focus on firm age or innova- tion and not on firm size as a primary targeting criterion, given that startups and more productive firms create most jobs in the region. • Second, policy makers should reduce the space for discretionary policy implementation and ensure that laws and regulations are enforced equally across firms. This involves ensuring that laws and regulations are clear, that the complexity of policy implementation is reduced, and that policy is designed and implemented by a strong, capable, and accountable administration. The latter can be supported by linking xxvi Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa entry into and promotions within the administration to merit, judged on the basis of potential or actual contributions to the legitimate goals of public policy. • Third, if MENA governments want to pursue state-led development policies, they would be wise to avoid the mistakes of the past and ensure that these new industrial policies—and the administrative structure that implements them—minimize the scope for capture, pro- mote competition, and tightly link support to measurable and verifi- able performance. • One critical aspect of this reform agenda is to create institutions that promote and safeguard competition and equal opportunities for all entrepreneurs. Such institutions include, but are not limited to, a strong, well-organized, and highly competent public administration necessary to implement critical policy changes, such as an effective competition law; an independent competition authority; appropriate procurement laws and implementation; and an independent judiciary. • Another component, just as important, is to ensure policy making is transparent and open, with a mechanism that facilitates and encour- ages citizen participation. Citizens should have access to information on proposed and ratified laws and regulations; be able to provide input into policy design and evaluation; be aware of politicians’ stakes in firms that benefit from government policies; and have full knowledge of who benefits from subsidies, procurement tenders, public land transactions, privatizations, and so forth. • This report provides a decision-making guide that summarizes the foregoing, which governments can use as a framework when designing and implementing policies. The decision-making guide is aimed to maximize the likelihood of success given inherent uncertainties and to maximize the positive impact of policies on growth and jobs by mini- mizing the risk for capture. Notes 1. The model proposed by Aghion et al. (2001) demonstrates that fair private sector neck-and-neck competition drives economic growth. In this model, competition increases firms’ incentives to adopt new technologies in order to reduce costs and escape competition (at least temporarily). However, if a few (colluding) firms have sizeable exogenous cost advantages, which are unbridgeable by competitors in the same sector, then all firms in the sector have fewer incentives to adopt new technologies and sector growth is lower. Overview xxvii In this case, the firms with the cost advantage have little incentive to invest in innovation since they do not face competitive pressures to reduce their costs further; the laggard firms are too far away from the frontier to bridge the cost gap and instead use vintage production technologies, focusing on local market niches to survive. Aggregate growth increases in the number of sectors that are characterized by neck-and-neck competition market structures. 2. See Jones (2011) for a presentation of the concept of weak links. Introduction This report argues that countries in the Middle East and North Africa (MENA) face a critical choice in their quest for higher private sector growth and more jobs: promote competition, equal opportunities for all entrepreneurs and dismantle existing privileges to specific firms or risk perpetuating the current equilibrium of low job creation. The report shows that policies which lower competition in MENA also constrain private sector development and job creation. These policies take different forms across countries and sectors, but share several common features: they limit free-entry in the domestic market, effectively exclude certain firms from government programs, increase the regulatory burden and uncertainty on nonprivileged firms, insulate certain firms and sectors from foreign competition, and create incentives that discourage domestic firms from competing in international markets. Such policies are often captured by a few privileged firms with deep political connections and persist despite their apparent cost to society. Furthermore, the millions of workers and consumers who bear the brunt of these policies are often unaware of their adverse impacts on the jobs to which they aspire, thereby limiting the scope of debate necessary for internal policy dialogue and reform. Labor markets in the MENA have been in low equilibrium for a long time. MENA has a large reservoir of untapped human resources with the world’s highest youth unemployment rate and the lowest participation of women in the labor force. Desirable private sector jobs, those that are high-paying and attached to a formal contract or social security, are very few. Strategies that consist in increasing employment in the public sector have proven to be unsustainable. On the other hand private sector job creation was too weak to absorb the growing labor force pushing a grow- ing number of workers to seek employment in unproductive, subsistence activities, often in the informal economy. The Arab Spring uprisings were 1 2 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa a powerful expression of the lack of employment opportunities and wide- spread sense exclusion. A recent Regional Flagship highlights how labor supply factors and labor laws affect employment in MENA. World Bank (2013a) shows that labor force participation is low, especially among women. Unemployment is persistently higher than in other regions and overwhelmingly affects youth. In some countries, such as the Arab Republic of Egypt and Tunisia, the highly educated are more likely to be unemployed (see also World Bank, 2014a). Too often, access to the few desirable jobs depends more on circumstances beyond individual control than on merit. Among those who are employed, the majority are engaged in low-quality jobs—those characterized by low pay and productivity, and without formal labor con- tracts or social security. The lack of high-quality private sector jobs is also reflected in the small size of formal firms. This report argued that a key challenge for the MENA region to improve the labor market and create more jobs is to “change the rules of the game to create a dynamic private sector that capitalizes on the full range of theregion’s human capital.” The report highlighted the central role of promoting competition to stimulate pri- vate sector growth. However, there is little evidence on the political econ- omy factors that perpetuate and/or accentuate the lack of competition in the region, nor on the type of policy distortions that weaken competition and how those distortions ultimately affect job creation. This report aims to fill these gaps. It tackles the following questions: What types of firms create more jobs in MENA? Are they different from other regions? What policies in MENA prevent the private sector from creating more jobs? How do these policies affect competition and job creation? To what extent are these policies associated with privileges to politically connected firms? Previous work has also linked MENA’s employment performance to qualitative evidence of privileges to specific firms and weak competition. World Bank (2009) develops the argument that policy privileges captured by a few connected firms limit competition and thus growth in MENA. World Bank (2009) provides rich qualitative empirical evidence describ- ing the lack of competition because of privileges in MENA economies. It used all data available at the time—before the Arab Spring—to character- ize the cost of privileges and weak competition. It outlined several rele- vant policy mechanisms, demonstrating how privileges suppressed competition, ranging from access to credit and land to industrial policy. Building on post-Arab Spring data, this present report broadens the findings of World Bank (2009) by exploring the possible links between privileges and policies limiting competition. Moreover, it provides for the first time direct quantitative evidence that privileges limit competition, firm dynamics, and job creation. Introduction 3 The literature identifies several other prominent explanations for weak private sector job creation in MENA. These explanations often focus on the idea of a natural resource curse. World Bank (2012a) exam- ines how, and the extent to which, overvalued real exchange rates lead to Dutch disease effects that limit diversification and growth in MENA. Overvalued real exchange rates can either originate from natural resource revenues or from weak links whereby low performance in nontradable domestic services increase the relative prices of tradable goods (and ser- vices). World Bank (2012a) also shows how volatility of commodity prices produces fiscal and real exchange rate volatility in the MENA region. Even oil-poor countries are sensitive to changes in the oil price because a large part of their economies depends on work remittances, aid, tourism revenues, and/or (real estate) investments from oil-rich countries (Dahi and Demir 2008). Previous studies also argue that the adoption of a pegged or fixed exchange rate regime to shelter oil-rich economies from oil price volatility led to a real exchange rate overvaluation, and thus losses in competitiveness in the region (Nabli and Veganzones-Varoudakis 2002; World Bank 2012a). Other explanations include low private invest- ment rates, investments in capital- rather than labor-intensive industries because of energy subsidy distortions (Lin and Monga 2010), cumber- some business and trade regulations (Malik 2013), or the lack of access to finance (World Bank 2011b). This report provides evidence that privileges granted to politically connected firms are associated with many of the policy distortions that the literature identifies to weaken private sector growth and job creation. The report follows the argumentation of World Bank (2009) and Malik (2013), and provides new data supporting this thesis. We provide direct quantitative evidence that the generous energy subsides to industry in Egypt profited a handful of politically connected firms. Moreover, we provide evidence that cumbersome business regulations, including barri- ers to entry and trade, protect a few politically connected firms in Egypt and Tunisia. Similarly, this report puts forth data that connects the unequal access to credit and land, and inconsistent implementation of business rules and regulations, to the presence of politically connected firms in Egypt. Furthermore, we argue that overvalued real exchange rates caused by weak links, at least in part, originate from the concentra- tion of politically connected firms in (nontradable) backbone service sec- tors in MENA. This report also highlights the dynamic interactions between shortcomings in the design of well-intentioned industrial policy and policy capture in MENA. We employ a rich set of techniques and newly available data to exam- ine the fundamentals of job creation in MENA, and how privileges may impede job growth. This report assembles the most comprehensive firm 4 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa census database ever put together for the MENA region. This allows us to measure accurate characteristics of and trends in firms’ demand for labor, and provides reliable representative estimates of both aggregate private sector job creation and productivity growth determinants.1 Recent academic contributions, based on comparable firm census data from other regions, provide the basis for benchmarking and give important new insights on specific firm characteristics and dynamics driving job creation (see, among others, Haltiwanger et al. 2011; Hsieh and Klenow 2012; Bartelsman et al. 2014). These latest state-of-the-art techniques are applied to provide novel empirical stylized facts on the fundamentals of job creation in MENA, based on the newly available firm census data. Moreover, following the Arab Spring and the regime changes in Egypt and Tunisia, two novel data sets were constructed that identify firms politically connected to the Mubarak and Ben Ali regimes or their close collaborators.2 These two unique data sets enable us to quantify how political connections lead to policy privileges that distort both competi- tion and the firm dynamics associated with job creation. Several findings stand out: • First, GDP growth in MENA over the past two decades was moderate and mostly driven by demographic and structural change, while within sectors productivity growth was slow, lagging behind all other devel- oping regions (chapter 1). While demographic change (an increase in the working-age population) contributed positively to growth, MENA did not fully harness the growth benefits of this trend, as unemploy- ment and inactivity remained high. In addition, most workers are employed in small-scale and low productivity activities. • Second, job growth in MENA is weak because there are not enough startups and productive firms (chapter 1). We test whether the funda- mentals of job creation in MENA countries are different from fast- growing emerging and high-income countries in other regions. They are not: young firms and more productive firms are the engines of private sector job creation in MENA as elsewhere. However, in MENA the pool of young firms is too low and productivity growth too slow to accelerate job creation. The low number of young firms is due to weak entry and growth of new firms. Productivity growth is held back by slow within-firm productivity growth and by misallocation of labor and capital across firms within sectors. • Third, many existing policies lower competition, create an uneven playing field and reduce the emergence and development of young firms and productive firms (chapter 2). We illustrate how different policies—ranging from energy subsidies and barriers to foreign direct Introduction 5 investment (FDI) into services sectors, to the regulatory environment and its implementation—shape and in most cases reduce private sector growth and job creation. The chapter shows how these policies sys- tematically led to reduced competition, created unequal opportunities between entrepreneurs, resulted in low entry or growth of new firms, and weakened productivity growth. • Fourth, shortcomings in the design and implementation of past indus- trial policies in MENA resulted in firm-specific policies and did not promote performance and competition (chapter 3). The chapter reviews industrial policies in a number of MENA countries over the past two decades and compares them with the experiences of East Asian countries, analyzing differences in policy design and implemen- tation. The chapter shows how the success of industrial policies in East Asian countries is related to several factors. The most critical include: a focus on new economic activities in sectors where market failures are more likely to have a binding influence on industrial development; the implementation of more performance-oriented policies; an evaluation system in which both the performance of policies and officials can be assessed; the promotion and safeguarding of competition; and equal access to all firms based on their performance. • Last, the report shows how many of these policies have benefitted a handful of politically connected firms, while reducing competition, tilting the level playing field, and impeding aggregate job growth (chapter 4). Novel data sets on first-tier politically connected firms in Egypt and Tunisia that became available after the Arab Spring allow us, for the first time, to provide quantitative evidence on how firm privileges distort regulations, competition, and job growth in the region. Together these findings shed light on the entire micro- economic transmission channels from political privileges to dis- torted competition and firm dynamics that slow aggregate job growth. Moreover, the report provides direct quantitative evidence that the presence of politically connected firms reduces aggregate job growth by lowering the growth opportunities of the majority of nonconnected firms in Egypt. The available qualitative evidence points to similar mechanisms of policy privileges in other MENA countries. The empirical findings and mechanisms of this report are rooted in the theoretical framework of Aghion et al. (2001), who demonstrate that fair private sector neck-and-neck competition drives economic growth. The authors show that competition increases firms’ incentives to invest in the adoption of new technologies to reduce their costs and escape 6 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa competition (at least temporarily).3 However, if a few (colluding) market leaders have sizeable exogenous cost advantages, which are unbridgeable by competitors in the same sector, then all firms in the sector have reduced incentives to adopt new technologies and sector growth is lower. The market leaders have little incentive to invest in innovation since they do not face competitive pressures to reduce their costs; the laggard firms are too far away from the frontier to bridge the cost gap and instead use vintage production technologies, focusing on local market niches to survive. In contrast, aggregate growth increases in the number of sectors that are characterized by neck-and-neck competition market structures. Put together our empirical findings demonstrate that more market com- petition and a level playing field are required to boost employment growth in MENA. This report is closely related to a parallel report providing firm level evidence for export and import dynamics and performances in MENA (World Bank, 2014b). Taken together, both reports provide a compre- hensive picture of macroeconomic trends and their underlying empirical microeconomic causes in MENA countries ranging from exports and imports to productivity growth and job creation. This report focuses on oil-importing developing countries in MENA. Most of the analysis excludes Gulf countries and other major oil and gas exporting countries in the region. This is not to say that the issues tackled in this report are not relevant in these countries. However, it is important to keep in mind when reading this report that the pat- terns of firm dynamics and job creation, and the specific policies that distort these dynamics in oil-exporting MENA countries could poten- tially be different. Nevertheless, privileges and capture of policies by firms connected to political leaders or monarchs are a major concern as they lead to policies distorting a level playing field in all countries in the region. As such, the main findings and policy implications of this report are relevant for the region as a whole. The report is organized in four chapters and proceeds as follows: • Chapter 1 analyzes the dynamics and determinants of job creation and tests whether the fundamentals of job creation in MENA are similar to those in fast-growing developing and high-income countries in other regions. • Chapter 2 shows how different policies in MENA countries shaped private sector competition and thus the firm dynamics associated with job creation identified in chapter 1. The report analyzes the effect of various policy distortions on firm dynamics and job creation across the MENA region, ranging from energy subsidies to industry in Egypt; FDI restriction and FDI domestic in Jordan; (mobility) restrictions to Introduction 7 market access in the West Bank; the relation between red tape and job growth in Morocco; and the impact of inconsistent implementation of regulations across the region. • Chapter 3 documents past industrial policies in MENA and compares the experiences with the experiences of East Asian countries, high- lighting how the differences are linked to policy objective, design, and implementation. • Chapter 4 analyzes how privileges to politically connected firms result in these policy distortions that undermine competition and constrain private sector growth and jobs in MENA. It uses novel data sets that identify first-tier politically connected firms in Egypt and Tunisia to quantify for the first time not only their impact on regulatory and policy distortions, but also how they impact on job growth. The chapter also presents and discusses qualitative evidence on state- business relations from other countries. • The report concludes by laying out the implications for policy of the various findings and highlights the specific areas for policy reform to create a roadmap for more private sector growth and jobs in MENA. Notes 1. Most of the analysis on job creation in MENA is based on survey data with small samples that cover only part of the economy; often merely including selected larger formal sector establishments. For example, the World Bank Enterprise Survey (WBES) data only survey a few formal firms. Apart from sampling issues, which are particularly severe for MENA countries, the larg- est survey for the Arab Republic of Egypt, for example, includes only about 1,100 out of 2.4 million establishments in 2006. Moreover, all of the sur- veyed 1,100 establishments have at least five employees while 91 percent of all establishments in Egypt in 2006 had less than five employees. 2. To the best of our knowledge, comparably rich data on politically connected firms across various economic sectors have only been employed for Indone- sia in the academic literature so far. 3. The framework is closely related to Parente and Prescott (2002). Its validity has been tested empirically by estimating the impact of increased product market competition on growth (Aghion et al., 2006, 2009) as well as entry deregulation in India (Aghion et al., 2008). CHAPTER 1 Too Little Too Late: Private Sector Growth and Labor Demand Over the past two decades, moderate GDP per capita growth in MENA was driven by demographic change rather than labor productivity and did not create enough formal private sector jobs. The economic benefits from the ongoing demo- graphic trend could have been higher if MENA countries were able to absorb the fast increasing labor force into their formal economy. Instead, weak formal private sector job creation has resulted in a large portion of the labor force being inactive. Why has private sector job creation been so weak? We first examine whether the fundamentals of job creation in MENA countries are different from fast-growing emerging or high-income countries in other regions. They are not: young firms and more productive firms are the engines of private sector job cre- ation in MENA as elsewhere. However, MENA countries’ private sector has been characterized by low firm turnover—firm entry and exit—and slow pro- ductivity growth, limiting the pool of both young and productive firms. This chapter examines the nature of labor demand in MENA countries’ private sectors, and discusses possible determinants of private sector growth and job creation. First, the chapter briefly examines MENA’s performance in aggregate growth and the drivers of economic growth. Growth in MENA over the last two decades appears to have been moder- ate and driven by demographic change, while productivity growth was low compared with other developing countries. Job creation was too weak over this period to absorb the growing working-age population. This resulted in high unemployment, inactivity, and informal jobs. The chap- ter then turns to the reasons why private sector job creation in MENA over the last two decades was weak. The analysis shows that the determi- nants of job growth in MENA countries do not differ from those in high-income or fast-growing emerging economies in other regions: in MENA, as elsewhere, it is young firms and more productive firms that create more jobs. The chapter contends that low firm turnover and slow 9 10 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa productivity growth limit the pool of young firms and productive firms, and undermine faster job creation. The analysis is based on newly available firm census data from MENA, which are crucial to identify the fundamentals of job creation. The deter- minants of job creation are analyzed through the lens of the firm, using unique firm census data collected in six MENA countries (Egypt, Tunisia, Morocco, Jordan, Lebanon, West Bank and Gaza), and Turkey, which is used as the benchmark country.1 To the best of our knowledge, this is the first time that these census data, apart from Morocco, are being used for research purposes.2 There are, however, important differences in the type of census, coverage of variables, and years across countries. For example, the census data in Egypt cover over 2,000,000 establishments across all sectors in 1996 and 2006, and a smaller annual manufacturing panel that includes all establishments with at least 10 employees between 2007 and 2011. The Turkish census comprises more than 2,400,000 establishments between 2005 and 2010. In Tunisia, Jordan, Lebanon, and West Bank and Gaza, the census data are also in panel format and cover all sectors, including between 100,000 and 600,000 economic establishments, depending on the country and year. In Morocco, the data is a panel of manufacturing firms covering all firms with at least 10 employees and some smaller firms between 1996 and 2006. These differences in data coverage across countries are carefully taken into account, and are high- lighted when presenting the analysis. Moreover, the same methodologies and definitions are used in each country to compute firm entry and exit, firm productivity, and so forth. A detailed summary of the available cen- sus data is provided in appendix B. The chapter is organized as follows. The first section examines the growth and job performance of MENA countries. The second section provides evidence that the fundamental mechanisms of job creation are the same in MENA as in other regions: young and more productive firms create jobs. The third section shows that low firm turnover and slow pro- ductivity growth limit the pool of young firms and productive firms and thus impede job growth in MENA.3 Economic Growth Has Been Moderate and Job Growth Weak MENA grew moderately during the last two decades. Growth was driven by demographic change (increased working-age population), while aggre- gate productivity growth was low. Real GDP per capita growth hovered around 2 percent in the last two decades; about 2–3 percent lower than in East and South Asia, but Too Little Too Late: Private Sector Growth and Labor Demand 11 comparable to per capita growth rates in the other developing regions. After prolonged economic stagnation during the 1980s, growth in MENA recovered in the 1990s as governments shifted away from state-led economic models towards more private sector-led growth and trade integration. Between 1991 and 2012, real GDP growth per capita aver- aged 2.2 percent in constant terms (figure 1.1). Thus, it was about 2 to 3 percent lower than real GDP per capita growth in South and East Asia. Nevertheless, it was comparable or even slightly exceeded per capita growth in Latin America and Caribbean, Eastern Europe and Central Asia, and Sub-Saharan Africa. This decent growth performance was not driven solely by MENA’s oil-exporting high-income countries. Real GDP per capita growth was comparable among MENA’s developing FIGURE 1.1 Decomposition of GDP Per Capita Growth in MENA and Other Developing Regions a. All countries, 1991–2012 b. All countries, 2000–12 5.0 6.0 4.0 5.0 3.0 4.0 2.0 3.0 Percent Percent 1.0 2.0 0 1.0 0 –1.0 –1.0 –2.0 MENA EAP LAC ECA SSA South MENA EAP LAC ECA SSA South Asia Asia c. Developing countries, 1991–2009 d. Developing countries, 2000–09 8.0 10.0 6.0 8.0 4.0 6.0 Percent Percent 2.0 4.0 0 2.0 –2.0 0 –4.0 –2.0 MENA EAP LAC ECA SSA South MENA* EAP LAC ECA SSA South Asia Asia Demographic change Change in employment Change in labor productivity GDP per capita growth Source: Calculations based on World Development Indicators. Note: Developing MENA effectively excludes the GCC (data missing for Qatar). 12 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa countries, averaging 2.1 percent from 1991–2009 and accelerating to 2.6 percent from 2000–2009. Demographic change accounted for about 50 percent of aggregate real GDP per capita growth over the past 20 years, substantially higher than in any other region. Demographic change is measured by the change in working-age population as a share of total population. The MENA region has the second highest population growth rate in the world. Its population growth rate between 1990 and 2012 averaged 2 percent and was only surpassed by population growth in Sub-Saharan Africa (2.7 percent). High fertility rates combined with rapidly declining mortality contributed to a sharp increase in MENA’s working-age population as a share of total population (figure 1.1, left), rapidly increas- ing MENA’s potential labor supply. Though its demographic profile is often blamed for MENA’s high youth unemployment, the relative size of the labor force is a key determinant of the region’s recent economic growth performance. Aggregate productivity growth was low in MENA compared with other developing regions. Figure 1.1 demonstrates that the change in labor productivity explained about 50 percent of GDP growth among MENA’s developing countries over the past two decades, generating 1 percent real GDP per capita growth annually in that period. Productivity growth was significantly lower than in other developing regions: it gen- erated about 4.5 percent real GDP per capita growth annually in East Asia, 4 percent in South Asia, and about 2 percent in Europe and Central Asia and Sub-Saharan Africa. For the MENA region as a whole, per capita growth increased between 1995–2000 and 2000–2005 when demo- graphic change accelerated. Among GCC countries, labor productivity did not contribute to economic growth over the past fifteen years. Among MENA’s developing countries, however, productivity growth averaged 1.3 percent over the past decade, primarily based on growth in non–oil-exporting countries. MENA experienced significant productivity growth through reallocation  across sectors, but within-sector productivity growth was the lowest among all regions The reallocation of workers from sectors with lower (marginal) productiv- ity to sectors with higher productivity can be an important driver of aggre- gate productivity growth. One key insight of development economics is that growth is driven by a structural shift from agriculture to manufactur- ing and services. This sectoral shift tends to be mirrored in the pattern of employment, so that over time the labor force in the nonagricultural sec- tor increases while employment in the agricultural sector declines (Kuznets 1996). As labor moves to the usually higher-productivity Too Little Too Late: Private Sector Growth and Labor Demand 13 industrial sector, overall productivity rises and incomes expand (Duarte and Restuccia 2010; Herrendorf, Rogerson, and Valentinyi 2013).4 As incomes rise, the demand for services increases. In many countries the share of the service sector in GDP rises almost linearly with the income level. Moreover, Eichengreen and Gupta (2011) reveal that in OECD countries service sector labor productivity as a share of average labor productivity tends first to rise at lower-income levels, then decline over an intermediate range, before increasing again. The second surge is most likely caused by the rise of modern services (business services, tele- communication, finance, and so forth). In many fast-growing developing countries, especially in Asia, the reallocation of workers from low pro- ductivity to high productivity sectors has contributed positively to growth during the last twenty years (Duarte and Restuccia 2010; Rodrik and McMillan 2012). All MENA countries in the sample, with the exception of Saudi Arabia, experienced aggregate productivity gains because of labor real- locations between sectors from 2000 to 2005. Labor productivity growth expressed as the change in output per worker can be decom- posed into within-sector change and reallocations “across” sectors or structural change (figure A.3 in appendix A). We note that the follow- ing results are based on measurements of average, not marginal labor productivity.5 However, as a robustness check, we also approximate marginal sector productivities based on wage data from harmonized household surveys for Egypt and Tunisia (World Bank, I2D2 data- base). The results show that the gaps in marginal productivities mea- sured by average wages across sectors are smaller than gaps measured by value added per worker, but sectoral differences remain significant (see table A.1 in appendix A). The contribution of labor reallocations (that is, structural change) to aggregate productivity growth was stron- gest in the Syrian Arab Republic and Egypt (figure 1.2a). In Syria, the country with the fastest structural change in the MENA region, real- location of labor contributed about 1.8 percentage points to aggregate productivity growth (which was 2.7 percent). In Egypt, it contributed one percentage point to aggregate productivity growth, which was negative (−2.2 percent) because of low within-sector productivity growth. The negative contribution in Saudi Arabia is a result of the influx of non-Saudi workers, many of whom were hired for low value added service activities. In Tunisia, the contribution of labor realloca- tion to growth (i.e., structural change) slowed after a wave of privatiza- tion came to an end in 2005. However, within-sector productivity growth was the lowest among all regions. Figure 1.2a illustrates that aggregate productivity growth among the seven MENA countries was the lowest among developing regions, 14 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 1.2 Structural Change across Regions and among MENA Countries, 2000–05 a. Regions b. MENA countries LAC Egypt, Arab Rep. Saudi Arabia High-income countries West Bank and Gaza MENA Tunisia Africa Jordan Syrian Arab Republic Asia Morocco –0.02 0 0.02 0.04 –0.04 –0.02 0 0.02 0.04 0.06 Within Structural change Source: World Bank calculations; data source see appendix A. because of low within-sector productivity growth. Figure 1.2b demon- strates that the regional average hides substantial variations across the seven MENA countries. Within-sector productivity growth has been negative in Egypt since 1982, primarily driven by declining labor produc- tivity in mining, manufacturing, and wholesale and retail trade. Within- sector productivity growth was also negative in Saudi Arabia and West Bank and Gaza between 2000 and 2005. In West Bank and Gaza, labor productivity fell steeply in agriculture, wholesale and retail trade, and transport and communication. Within-sector productivity growth was high in Jordan and Morocco. In Jordan, it was driven by manufacturing (through labor shedding), transport and communication, and finance (by attracting new workers); in Morocco, by agriculture, mining, and com- munity, social, personal, and government services. In Tunisia, human capital accounted for a significant share of labor productivity, but the analysis also reveals important misallocation of human capital. The lack of data prevented accounting for human capi- tal in the growth decomposition for other countries in the region. Several countries in the MENA region have undergone a steep increase in educational attainment during the last two decades. To understand better how recent increases in Tunisia’s educational attainment have affected the reallocation of human capital across sectors, we replicate the structural change analysis for the years 2005–10, using data on output per unit of human capital.6 Accounting for improvement in education of the labor force nuances some of the previous findings. For example, while both agriculture and the public sector employed Too Little Too Late: Private Sector Growth and Labor Demand 15 18 percent of the total working population in 2005, the share of imputed human capital was 12 percent for agriculture but 27 percent for the public sector. Human capital productivity growth within the agricultural sector was even negative, implying that growth of human capital exceeded overall employment growth. Overall, human capital exceeded employment growth by about 50 percent, accounting for a significant share of the labor productivity increase. Moreover, the adjusted productivity measure also reveals significant misallocation of human capital. In 2009, 75 percent of Tunisia’s human capital aug- mented labor was employed in sectors with below-average productiv- ity, 24 percent in public administration alone, with 12 percent in the public works program. Consistently, firm census data suggest that firm turnover in MENA is driven by structural change rather than creative destruction. Bartelsman, Haltiwanger, and Scarpetta (2004) suggest a way to assess if firm churn- ing is driven by structural change (resource reallocations between sectors) or creative destruction (resource reallocations among firms within a sector). In the former case, the correlation between entry and exit rates across sectors should be negative; in the latter, positive FIGURE 1.3 Correlation between Entry and Exit Rates across Two-Digit Sectors United States Hungary Turkey manufacturing (B) Turkey all sectors (b) Estonia Romania Mexico West Bank and Gaza (all sectors) Latvia Morocco manufacturing (a) Portugal Tunisia all sectors –0.4 –0.2 0 0.2 0.4 0.6 0.8 1.0 Source: Calculations based on census data. Note: The entry/exit rates are weighted by employment; correlation coefficients are significantly different from 0 at the 10% level in Tunisia, Estonia, Turkey, Hungary, and the United States; a) entry in /exit out 10+ employees; b) Entry in /exit out 20+ employees. Correlations are measured between 2005 and 2010 in Turkey, 1996–2006 in Morocco, 2004–2012 in West Bank and Gaza, and the 1990s for all other countries. 16 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa (i.e., new firms enter and old firms exit the same two-digit sector). Figure 1.3 reveals that countries at a later stage of development have higher positive within-sector correlations, indicating that the sectoral structures in these countries have converged so that the main force behind firm turnover is creative destruction. In contrast, in less devel- oped countries the correlations tend to be lower. Morocco, Tunisia, and West Bank and Gaza are among the lowest, suggesting that sectoral adjustment resulting from structural change is still ongoing. BOX 1.1 Is Structural Change in Morocco Gender-Biased? In the following, we disaggregate the rela- estate sectors increased their employment tive changes in sectors’ employment shares shares for both women and men, but the by gender to examine if structural change in numbers of new jobs in these sectors are Morocco increased the probability of female very small in proportion. These benefitted and/or male employment in higher produc- mostly educated women and men in cities. tivity sectors. The analysis is based on The overall number of jobs provided in World Bank (2014d). Figure B1.1.1 plots these two sectors is small, so relatively few the relative labor productivity of different employees benefitted from this trend. In sectors on changes in the employment share contrast, employment trends are very in these sectors. The sizes of the circles rep- different for the majority of uneducated resent the size of the sector. Sectors above women living in rural areas. About the (horizontal) dashed line have above- 60 percent of women in the labor force average labor productivity, while sectors to work in agriculture; more than 77 percent the left of the (vertical) dashed line increased of them worked as family helpers, and their employment share. The left panel 44 percent work part-time. The share shows the changes in the labor share among even slightly increased from 59 percent in women (on the x axis), while the right panel 2000 to 61 percent in 2011 (figure B1.1.1, illustrates the changes in the labor share left). Conversely, it declined for men among men (on the x axis). (figure B1.1.1, right). Given that the agri- The results show that structural change culture sector is by far the largest employer did not benefit women and men equally. in Morocco (39 percent of the total labor Figure B1.1.1 compares the reallocation force in 2011), this employment trend out- (changes in labor shares) of women and men weighs any other. Note that the aggregate across the different sectors. There are some labor share in agriculture still declined since important similarities. The high productiv- the overall labor force participation of men ity communications and finance and real is 2.8 times that of women. (continued on next page) Too Little Too Late: Private Sector Growth and Labor Demand 17 BOX 1.1 Continued FIGURE B1.1.1 Reallocation of Labor across Sectors, by Gender, 2000–11 a. Female, 2000–11 b. Male, 2000–11 Log (sectoral productivity/average productivity) Log (sectoral productivity/average productivity) 12 12 pu pu 10 comm 10 comm 8 8 fire fire 6 min 6 min 4 4 2 g mf e 2 mf g e mc mc mm mm mo tn mt motn os f c tch a a hmftd c 0 0 os –2 –2 –6 –4 –2 0 2 –10 –5 0 5 Change in employment state Change in employment state beta = 0.399; t-test = 0.70 beta = 0.065; t-test = 0.19 Note: From World Bank (2014d). Sectors: a (agriculture), f (fisheries), min (mining), mf (manufacturing, food), mt (manufacturing, textiles), mc (manufacturing, chemicals), mm (manufacturing, mechanical & electrical), mo (manufacturing, other), pu (public utilities), c (construc- tion), td (trade), h (hotels & restaurants), tn (transportation), comm (communications), fire (finance, insurance, real estate & business services), g (government), e (education & health), os (other services). The analysis showed that aggregate productivity growth was low in MENA compared with other developing region in the past 20 years. How did labor markets and in particular labor demand evolve during this period? MENA has had weak job performance. Most workers are employed in small-scale low productivity activities; this employment structure persisted and increased somewhat over the past decade MENA’s labor market failed to absorb the fast growing labor force. Formal sector workers as a share of working-age population in MENA is much lower than in other middle-income regions such as Latin America and Caribbean (LAC), or Eastern Europe and Central Asia (ECA); figure 1.4. While the measured share of informal labor is lower than in LAC, the share of the working-age population dropping out of the labor force is much higher, especially among women. Less than a quarter of all working-age women in the MENA region participated in the labor force in 2012 (see also World Bank 2014a). Small-scale activities provide the majority of jobs in MENA, albeit with some noteworthy differences across countries. Figure 1.5 illustrates the distribution of employment across firm size categories in the different MENA countries. The share of employment in micro establishments with less than five employees dominates the private sector in Egypt and 18 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 1.4 Demographic Change and Composition of Working-Age Population a. Demographic change b. Working-age population 75 60 Percentage of total population 50 (in % of total population) 70 Working-age population 40 65 30 60 20 55 10 50 0 Formal Informal Unemployed Inactive 90 92 94 96 98 00 02 04 06 08 10 12 19 19 19 19 19 20 20 20 20 20 20 20 workers workers individuals individuals ECA EAP LAC MENA LAC ECA SAS MENA SSA Source: a. Calculations based on World Development Indicators. b. World Bank 2013, based on ILO-KILM database. West Bank and Gaza, reaching almost 60 percent. It is significantly lower in Jordan (40 percent), Tunisia (37 percent), and lowest in Turkey (34 percent). In contrast, Tunisia (36 percent)7 and Jordan (33 percent) have the highest concentration of workers in large establishments,8 while Turkey has the highest share of workers in medium-size establish- ments (29 percent), also exceeding its share of workers in large ones (26 percent). The share of jobs in firms with at least 1,000 employees is less than 10 percent in all five countries, which starkly contrasts with the employment situation in high-income countries. For example, in the U.S., 48 percent of all employees work in firms with more than 10,000 employees. Overall, figure 1.5 highlights that small scale activities in micro enterprises are an important source of employment in MENA countries. The high share of jobs in micro establishments is alarming given that businesses with fewer than 10 employees are much more likely to be informal in MENA (World Bank 2011a). Moreover, informality in MENA is associated with a lower level of productivity relative to other regions at comparable stages of development. The concentration of jobs in micro establishments is also a reflection of MENA’s sectoral structure; low productivity services provide the most jobs. Most of these services are likely to represent subsistence activities rather than a vibrant informal sector. Figure 1.6 illustrates the distribu- tion of employment by sectors; economic sectors are approximately sorted by their share of formal sector employees. The majority of domestic pri- vate sector jobs are small scale; they are often low-productivity service sector activities. In Egypt and West Bank and Gaza, around 40 percent of Too Little Too Late: Private Sector Growth and Labor Demand 19 FIGURE 1.5 Employment Share, by Firm Size 60 40 Percent 20 Turkey 0 Tunisia Jordan ocr West Bank and Gaza Mi all Sm ium Egypt, Arab Rep. d Me rge La Source: Calculations based on census data. Note: The graph shows the share of employment by firm size according to the following classification: micro firms have less than 5 employees, small firms have between 5 and 9 employees, medium firms have between 10 and 99 employees, large firms have 100 employees or more. Periods covered by country: Turkey (2006), Tunisia (1996–2010), Jordan (2006), Egypt (2006), Palestine (2004, 2007, 2012). The employment shares in Tunisia are based on firm data while it is establishment data for the other countries. Lebanon is not included as the nature of the census data is different; its coverage limited to firms with a tax ID. FIGURE 1.6 Distribution of Employment across Nonagriculture Sectors Egypt, Arab Rep. West Bank and Gaza Jordan Turkey Lebanon Tunisia 0 20 40 60 80 100 Percent Mining Manufacturing Construction and utilities Other services Vehicle trade and repair Hotels and restaurants Personal services Retail trade Source: Calculations based on census data. 20 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa all jobs are in these sectors. All three sectors hold primarily what are often informal one- or two-person firms in MENA (see World Bank 2011a). For example, the average establishment size in retail trade is less than 2 percent in all of the countries, varying from 1.2 in Tunisia to 1.9 percent in Egypt (including one wage worker and the owner). Moreover, labor force survey data in Egypt (the Egypt Labor Market Panel Survey) indi- cate that 70 percent of employment in retail trade is informal: jobs with- out a formal contract or social insurance (World Bank 2014a). Retail trade, personal services, and hotels and restaurants still account for 28 percent in Jordan. In Lebanon and Tunisia, the highest share of jobs is in business services (which are included in other services). Business service firms have, on average, only slightly larger firm sizes than retail service firms in all MENA countries. The concentration of employment in small and micro-firms decreased slightly in recent years, but it is still higher than in the late 1990s in certain countries. The share of jobs in medium and large establishments increased somewhat in the oil-importing middle-income MENA countries (apart from Egypt) between 2005 and 2012, albeit at a much slower pace than in Turkey. Figure 1.7 shows that the share increased by almost 10 percent in Turkey at the end of the 2000s, compared with less than 5 percent in FIGURE 1.7 Percentage Change in Share of Employment in Medium and Large Establishments Turkey, 2005–10 Lebanon, 2005–10 West Bank and Gaza, 2004–12 Jordan, 2006–11 Tunisia, 1996–2010 Egypt, Arab Rep., 1996–2006 –10 –5 0 5 10 Percent Source: Calculation based on census data. Note: Medium and large establishments are defined as having more than 10 employees. Too Little Too Late: Private Sector Growth and Labor Demand 21 MENA countries. In Egypt and Tunisia, the share of employment in large establishments declined over the longer time horizon, reflecting stagnation in formal sector job growth preceding the recent crisis. Figure 1.7 high- lights that employment declined in Egypt by 7 percentage points (from 23 to 16 percent) between 1996 and 2006. In contrast, the dominance of small- scale activities in micro establishments with fewer than 10 employees increased over time (from 62 percent in 1996 to 72 percent in 2006).9 Most firms in MENA had weak employment growth; a few fast-growing firms account for a large share of job creation Small firms did not grow. Micro firms with fewer than 10 employees almost never enter larger size categories. This finding is illustrated in the case of Tunisia in table 1.1 which summarizes the probabilities that firms transitioned among different size categories (or exited them) in 2007–11. For example, of all one-person firms in Tunisia in 2007, 22 percent exited by 2011, 76 percent remained one-person firms, and only 2 percent hired at least one more worker. Overall, table 1.1 highlights that micro firms with fewer than 10 employees almost never grow beyond 10 employees over time. In particular, table 1.1 reveals that the probability of all nonfarm micro firms to grow beyond 10 employees four to five years later is 2 percent in the West Bank and Gaza (6 percent in the West Bank alone), 3 percent in Tunisia, and 12 percent in Lebanon. The very low probability that micro firms will transition to larger size categories is striking. This finding is consistent with those of the World Bank (2014a), which showed that most micro firms are informal, and that informal firms have a very low chance to formalize in MENA. The probability that medium-size manufacturing establishments grow to become large establishments four years later is low across TABLE 1.1 Employment Transition Matrix Percent Tunisia Transitions 2007–11 Status in 2011 Status in 2007 Exited 1-Person Micro SME Large 1-person 22 76 2 0 0 Micro 9 21 67 3 0 SME 6 11 16 64 4 Large 6 11 3 15 65 Source: Calculation based on census data. Note: Micro: 2–9 employees, SME: 10–99 employees, Large: ≥100 employees. Bold signifies stagnation (no growth in jobs), red signifies shrinkage (jobs or exit), and orange signifies expansion (growth in jobs). 22 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 1.8 Employment Transition, by Firm Size a. Firms with 2–9 employees (all sectors) b. Firms with 20–49 employees (manufacturing sector) 100 14 80 13 12 60 Percent Percent 11 40 10 20 9 0 8 Lebanon , Tunisia, West Bank Turkey, Egypt, Morocco, West Bank Jordan, 2005–10 2007–11 and Gaza, 2006–10 Arab. Rep.,1996-2006 and Gaza, 2006–11 2007–11 2007–11 2007–12 Grow to >10 Remain micro (2–9) Shrink or exit (<2) Probability of growth to >50 employees Source: Calculation based on firm census data. Note: Because of data constraints, the transition probability for Jordan is for incumbents only; excluding entrants and exiters. MENA countries. Figure 1.8 (right panel) shows that this probability for firms with 20–49 employees is 13.5 percent in Turkey, 11.9 percent in Egypt and Morocco, 10.7 percent in West Bank and Gaza, and 9.8 percent in Jordan. A few fast-growing firms (the gazelles) account for a high share of job creation in MENA. Gazelles are defined as firms that double their employment over a four-year period.10 The analysis is restricted to firms with more than 10 employees in the base year.11 Figure 1.9 shows the incidence of gazelles across MENA countries, the U.S., and Turkey. Lebanon has the highest share of gazelles (5.6 percent) out of the MENA countries. The shares are only slightly lower in Tunisia and Turkey. Jordan has the lowest incidence of gazelles (1.4 percent). However, gazelles accounted for a high share of employment growth in MENA. Figure 1.10 shows the share of jobs created by gazelles and non- gazelles. Gazelles accounted respectively for about 64 and 42 percent of total net job creation in Jordan and Tunisia. In contrast, Turkey’s job creation was broader-based across all firms, as gazelles only contributed 15 percent to job growth. Gazelles accounted for all net job creation in the manufacturing sector in Morocco, offsetting job destruction by all other formal manufacturing firms. In Egypt, manufacturing net job cre- ation was negative between 2007 and 2011, driven by substantial job destruction among nongazelle establishments, while in Jordan manufac- turing net job creation was positive, whereby nongazelles created more jobs (60 percent) than gazelles (40 percent). Too Little Too Late: Private Sector Growth and Labor Demand 23 FIGURE 1.9 Incidence of Gazelles in All Sectors and Manufacturing a. All sectors b. Manufacturing sector 7 7 6 6 5 5 Percent Percent 4 4 3 3 2 2 1 1 0 0 –112 1 0 11 0 08 –1 1 11 06 06 0 7– 005–1 006– 005–1 004– 06 007– 03– , 20 ,2 0 ,2 2 sia, anon, tates, 2 2 , 20 ., 2 co, 20 dan d Gaza urkey uni dan Rep roc Jor n T T e b d S Jor rab Mo a L te ,A ank Uni ypt e st B Eg W Source: Calculation based on census data. Note: Gazelles are defined as firms with at least 10 employees in the base year that double employment over any four-year period. Data for Turkey includes only firms with at least 20 employees. FIGURE 1.10 Share of Jobs Created by Gazelles and Nongazelles in All Sectors and in the Manufacturing Sector a. All sectors b. Manufacturing sector 200 150 100 150 50 100 0 Percent Percent 50 –50 –100 0 –150 –50 –200 –100 –250 11 11 10 08 06 ep. , 11 00 6– 006– 005– 004– 00 3– b R –11 0 06– a, 2 a isi da n, 2 ke y, 2 te s, 2 co, 2 r t, A 2007 dan , 2 Tun Jor Tur Sta roc gyp Jor d Mo E Unite Gazelles Nongazelles Source: Calculation based on census data. Note: Gazelles are defined as firms with at least 10 employees in the base year that double employment over any four-year period. Data for Turkey includes only firms with at least 20 employees. 24 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Drivers of Job Growth: Young Firms and Productive Firms Create More Jobs The previous section showed that MENA has had a weak jobs as well as aggregate productivity performance in the last 20 years. Moreover, most workers are employed in small-scale low productivity activities. It is important to understand the factors behind this weak performance at the microeconomic level. This section aims to shed light on this issue by answering the following questions: what types of firms drive job growth in MENA countries? Are these micro fundamentals of job creation differ- ent from those found in (fast-growing) countries outside of the region? Analysis of firm census data shows that it is younger firms and more productive firms that create more jobs in MENA, as in fast-growing and high-income countries Evidence from other regions suggests that younger and more productive firms create more jobs. Age, size, and productivity are fundamental deter- minants of firm employment growth. Understanding their relative importance in explaining job creation is critical to determine the policy mix for stimulating private sector growth. There is growing literature analyzing these questions (Box 1.2). For example, Haltiwanger, Jarmin, and Miranda (2010) find that in the Unites States, net employment growth is associated with firm age and not firm size. The literature also identifies productivity as an important determinant of firm growth in developing countries (e.g., Berman and Machin 2004; Vivarelli 2012). Another strand of the literature highlights the importance of firm growth over their life cycle; Hsieh and Klenow (2012) show that U.S. firms increase their size (number of employees) and productivity by a factor of 8 over their life cycle (within the first 35 years). In contrast, Mexican firms double and Indian firms do not increase their employees over the same period (both approximately double their productivity). Are the firm characteristics associated with job growth different in MENA countries? Job creation in MENA is dominated by young firms. Micro-startups create most jobs. These findings are illustrated in figure 1.11, which shows net job creation by firm size and firm age in Tunisia and Lebanon. Almost all net job creation in Lebanon and Tunisia was generated by young firms at their start-up period; i.e., in the first four years after they were estab- lished. In both countries, it was primarily micro-startups with between one and four employees that created most jobs. For example, micro-startups generated about 66,000 jobs in Lebanon between 2005 and 2010, account- ing for 177 percent of net job creation. The second largest number of jobs (12,000) was created by young large firms with 200–999 employees. Too Little Too Late: Private Sector Growth and Labor Demand 25 BOX 1.2 Who Create More Jobs? Young firms are an engine of job creation. of firm age as a determinant of firms’ poten- There is a large and growing literature link- tial to create jobs. ing employment growth to firm dynamics. Among the pool of young firms, a small Studies typically find that younger and number of fast-growing firms appear to cre- smaller firms have higher employment ate most new aggregate jobs in high-income growth rates than older and larger firms (e.g., countries. A recent stream of the literature Mansfield 1962; Hall 1987; Hart and Oulton linking employment growth to firm dynam- 1996; Ayyagari, Demirguc-Kunt, and ics suggests that a small group of fast- Maksimovic 2011). Likewise, Davidsson and growing firms, often referred to as gazelles, Delmar (2006) show that most of the growth are the main drivers of aggregate job cre- of younger and smaller firms is organic, ation (e.g., Bottazzi and Secchi 2007). In while for larger and older firms, job growth other words, a handful of firms experience a primarily comes through acquisitions. Hsieh period of accelerated employment growth and Olken (2014) contribute to the debate while most other firms hardly grow at all. on firm size and job creation, showing that Empirical studies for various developed large firms have higher average products of countries find that 5–10 percent of the firms capital and labor, which suggests that large deliver 50–80 percent of aggregate employ- (not small) firms are growth constrained. ment creation (e.g., Acs, Parsons, and Tracy Haltiwanger, Jarmin, and Miranda (2010) 2008; Coad and Hoelzl 2010). These fast- nuance these findings, showing that net growing firms can be found in all industries employment growth is associated with firm and are usually young firms that are more age and not firm size in the United States, innovative and take more risks (Bars et al. implying that young firms, especially start- 2006; Goedhuys and Sleuwaegen 2009; ups, are the drivers of job creation. However, Henrekson and Johansson 2010). as young firms tend to be small, there is also Most microeconomic studies find a posi- a positive bivariate correlation between firm tive relationship between productivity and size and net job growth in the data. employment creation (van Reenen 1997; Furthermore, Hsieh and Klenow (2012) cor- Blanchflower and Burgess 1998; Piva and roborate the importance of firm age growth. Vivarelli 2004; Coad and Hoelzl 2010; The authors show that U.S. firms increase Vivarelli 2012). In this regard, it is useful to their number of employees and productivity distinguish between product and process by a factor of 8 over their life cycle (within innovation. Product innovation is generally the first 35 years). In contrast, Mexican firms found to increase labor demand and hence double and Indian firms do not increase their firm-level employment growth. Process employees over the same period (both innovation is associated with productivity approximately double their productivity).a growth which might, however, compensate Again, the study highlights the importance labor. Indeed, the findings for process (continued on next page) 26 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa BOX 1.2 Continued innovation are less clear-cut and also indicate change” in the literature (e.g., Berman and job destruction in some cases, especially in the Machin 2004). Conte and Vivarelli (2010), short run (e.g., Hall, Lotti, and Mairesse Hanson and Harrison (1999), and Fuentes 2008; Harrison et al. 2008). and Gilchrist (2005) find that imported skill- Among developing countries, studies biased technological change is an important suggest that the adoption of foreign tech- determinant of the recent increase in the nologies increases firms‘ demand for labor, relative demand for skilled labor in develop- especially for skilled labor. Product and pro- ing countries. cess innovation in developing countries take Thus, these findings highlight a positive the form of diversification into new products relation between productivity and employ- and the adoption of foreign technologies (or ment in developing countries. Innovation, organizational structures), respectively. Both which takes the form of diversification into processes have been found to increase the new products and the adoption of foreign demand for labor in developing countries. technologies in developing countries, is Foreign technology adoption has been found found to increase the demand for labor, lead- to increase the demand for skilled labor, ing to a positive relation between productivity referred to as “skill-biased technological and job growth in developing countries. Note: a. The fact that older plants in India and Mexico are small may not have a large effect on aggregate outcomes if there are fewer surviving old plants. The authors show, however, that exit rates in India and Mexico are generally not higher than in the United States. FIGURE 1.11 Net Job Creation, by Firm Size and Age a. Tunisia, 1996–2010 b. Lebanon, 2005–10 500,000 60,000 400,000 50,000 40,000 Net job creation Net job creation 300,000 30,000 200,000 20,000 100,000 10,000 0 0 20 20– –100,000 10– –29 –100,000 10– 29 0– 14 0– 14 1–4 1–4 5–9 5–9 9 9 9 4 4 9 9 10–1 9 10–1 20–4 20–4 199 199 50–9 999 50–9 e e 999 00 00 Ag Ag 100– ≥1,0 100– ≥1,0 200– 200– Size Size Created jobs Shed jobs Source: Calculation based on census data. Too Little Too Late: Private Sector Growth and Labor Demand 27 In Tunisia, micro-startups created 580,000 jobs between 1996 and 2010, accounting for 92 percent of all net job creation. However, the aggregate performance masks important differences in the sectoral patterns of job creation across countries. In all MENA coun- tries with available data, job creation was driven by retail trade, business services, or personal and community services. World Bank (2011a) shows that job creation in micro-firms in these sectors is often part of the informal economy, which is less productive in MENA than in other developing regions. In particular, table 1.2 reports that many new jobs in micro estab- lishments are in retail trade and personal services, which are dominated by informal firms. In these sectors, the average firm size is below one worker. In Egypt, these two sectors generated more than 700,000 and 400,000 new jobs from 1996 to 2006, respectively, accounting for over 80 percent of total net job creation. Labor force survey data from the Egypt Labor Market Panel Survey show that this trend continued between 2006 and 2012. Certain higher productivity activities such as real estate and finance, tourism, ICT, and manufacturing also contributed to job creation. In Jordan, potentially higher productivity real estate and finance, chemicals and pharmaceuticals, and the food sector accounted for 28 percent of total net job creation between 2006 and 2011, counterbalancing somewhat the trend towards jobs in the informal sector (table 1.2). In Tunisia, 46 percent of total net job creation between 2006 and 2012 was concentrated in real estate and transport services, manufacturing of machinery and electrical equipment (mostly electric cables and switches), food products, and trans- port vehicles. The sectoral pattern of job growth in Turkey is different; there 77 percent of job growth between 2005 and 2010 was in real estate business services and construction, and other manufacturing. TABLE 1.2 Sectors with the Highest Rate of Job Growth across Countries Egypt, Arab Rep., West Bank and Gaza, Jordan, Tunisia, Turkey, 1996–2006 2004–12 2006–11 1996–12 2005–10 Δ Jobs Δ Jobs Δ Jobs Δ Jobs Δ Jobs Sector (%) Sector (%) Sector (%) Sector (%) Sector (%) Retail trade 39 Retail trade 26 Retail trade 18 Real estate, business 16 Real estate, business 37 service service Business service 17 Personal service 9 Hotels and 18 Machinery, electrical 12 Construction 25 restaurants equipment Other manufacturing 12 Hotels and 7 Health, social 17 Retail trade 12 Other manufacturing 15 restaurants Hotels and 9 Business service 7 Education 13 Transports 10 Hotels and restaurants 10 restaurants Personal service 6 Finance 6 Business service 12 Textiles 9 Food and beverages 6 Source: Calculation based on census data. 28 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa The ICT sector is an example of an emerging and dynamic sector that has experienced the entry and growth of new firms in several MENA countries. Consider the story of Eskadenia (World Bank 2009) founded by a Jordanian couple who worked for Ericsson in China, Dubai, Lebanon, and Sweden and decided to return to Jordan in 2000 to launch what grew into one of the largest and fastest-growing software firms in the Middle East. Their network of worldwide industry contacts from 30 years abroad helped them penetrate foreign markets quickly. Unable to tap startup capital from banks demanding high collateral, the partners self-financed the startup investment. By 2008 Eskadenia employed about 100 engineers and exports 80 percent of its products to countries in the Middle East, Eastern Europe, and North Africa. Even after controlling for sectoral heterogeneity, young firms are still the engine of job creation in MENA countries. We follow the methodology of Haltiwanger, Jarmin, and Miranda (2010) for the United States to test whether, after controlling for sector effects, young firms create more jobs regardless of their size. Figure 1.12 illustrates the rate of aggregate net job creation by firm size categories. The figure shows the coefficient estimate from a regression of firm employment growth on the various firm size cat- egories (controlling for sector and year dummies). The dashed-blue lines FIGURE 1.12 Net Job Creation, by Firm Size before and after Controlling for Firm Age 0.10 0.05 0 –0.05 –0.10 –0.15 1 2 4 9 9 9 9 9 9 3– 5– –1 –4 –9 19 99 0– 0– 10 20 50 10 20 Firm size Turkey size Lebanon size Tunisia size Turkey size + age Lebanon size + age Tunisia size + age Note: Calculation based on census data. The figure shows the results of a weighted regression of net job cre- ation, measured by the Davis-Haltiwanger-Schuh growth rate, on firm size dummies, controlling for sector and year effects. The figure plots the coefficients on the dummy variables representing the different firm size catego- ries before (dashed line) and after (solid line) controlling for firm age. The omitted category is firms with at least 1,000 employees. Too Little Too Late: Private Sector Growth and Labor Demand 29 show the impact of the different firm size categories on job growth when neglecting the joint distribution of firm size and age.12 It suggests that smaller firms create the majority of jobs in Lebanon and Tunisia.13 However, once the joint distribution of firm size and age is accounted for, the results change dramatically: smaller firms create fewer jobs than large firms. This indicates that the association between firm size and employment growth depends critically on firm age. A similar pattern can be observed among firms in Turkey (it is only possible to identify the same firms over time when they have at least 20 employees). Considering this finding, is it the case that young firms systematically create jobs regardless of their size? Figure 1.13 plots the relation between aggregate job creation and firm age (when accounting for the joint distribution of firm size and age). The findings shows that independent of firm size, young firms grow faster and create more jobs, particularly during their first four years of activity. The analysis is extended to additional MENA countries figure 1.14 by plotting the employment growth of entry cohorts in the first 10 years after they started operating. It confirms that employment growth is strongest in the first four to five years after firm entry and tends to level off thereafter. In Jordan, establishments from all nonagriculture eco- nomic sectors double their size in the first five years after entry, while manufacturing firms in Morocco are 1.7 times larger. The effects are comparable to growth rates of entrants in manufacturing and all other sectors in the first four years of operation. FIGURE 1.13 Net Job Creation, by Firm Age after Controlling for Firm Size 1.2 1.0 0.8 0.6 0.4 0.2 0 –0.2 10 8 9 2 3 1 4 9 9 4 7 5 6 –1 –1 –2 10 15 20 Firm size Turkey Lebanon Tunisia Source: Calculation based on census data. Note: The figure shows the results of a weighted regression of net job creation, measured by the Davis- Haltiwanger-Schuh growth rate, on firm size and age dummies, controlling for sector and year. The omitted category is firms older than 30 years. 30 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 1.14 Employment Growth Is Strongest in First 4–5 Years after Firm Entry 2.5 Employment, index = 1 for firms age 0 2.0 1.5 1.0 0 1 2 3 4 5 6 7 8 9 Years after entry Morocco (manufacturing), 2000 Turkey, 2006 Turkey, 2006 Jordan, 2006 Source: Calculation based on census data. Note: Employment is normalized to one for the entry year (age equal to zero). a) For Morocco (manufacturing), firm size exceeds 10 employees; for Turkey (manufacturing) and Tunisia, firm size exceeds 20 employees. If MENA governments want to pursue private sector development pro- grams targeting specific types of firms, they would be well advised to include firm age as a targeting criterion. SMEs have often been considered as the main source of employment growth, which explains the large vol- umes of access-to-finance support programs focused on small firms in developing countries in the past. The evidence for MENA countries, how- ever, highlights the critical role of firm age rather than size; i.e., young firms are in fact the engine of job creation. Thus, there is room for improv- ing existing SME support programs in MENA countries by targeting such programs to young firms, including startups and potential entrepreneurs. Average employment growth over firm’s life cycle in MENA is rela- tively weak. The analysis follows Hsieh and Klenow (2012), which shows the relationship between employment and establishment age among sur- viving firms based on cross-section census data (figure 1.15).14 The average weighted number of employees for the youngest age cohort (0–4 years after entry) is normalized to one. In contrast to Hsieh and Klenow (2012), the data allows the illustration of this relation among private establishments based on all economic (nonfarm) sectors, instead of the manufacturing sec- tor only. Figure 1.15 shows that after 25 years in operation, surviving firms approximately doubled their number of employees in the Arab Republic of Egypt, Jordan, Tunisia, and Turkey, with typically higher growth for younger age cohorts. Thereafter, employment for older age cohorts (founded before 1980) declined in Egypt, but increased in the other Too Little Too Late: Private Sector Growth and Labor Demand 31 FIGURE 1.15 Employment Growth over a Firm’s Life Cycle for All Nonfarm Sectors 12 Employment, index = 1 for firms age 0–4 10 8 6 4 2 0 0–4 5–9 10–14 15–19 20–24 25–29 30–34 35–39 ≥40 Age group Tunisia Jordan Turkey Egypt, Arab Rep. Source: Calculation based on census data. Note: The figure shows the average number of employees for different age-cohorts across establishments in all private nonfarm sectors (weighted by employment share of four-digit sectors following Hsieh and Klenow 2012). The average number of employees in each age cohort has been normalized to 1 for the youngest age category (age 0–4). The analysis for Turkey and the Arab Republic of Egypt is based on census data in 2006, for Tunisia in 2012, and for Jordan in 2011. Results for Jordan and Tunisia are similar for other years (e.g., 2006, 2010, or 2012). countries, most strongly in Turkey. For all MENA countries, the relation between employment and age is strongest in manufacturing, which also has the highest share of formal firms (figure D.1 in appendix D). More productive firms create more jobs. Apart from firm age, firm productivity is identified as an important determinant of job growth in fast-growing middle-income and high-income countries. We show this is also the case in MENA countries. Table 1.3 summarizes the results from regressions of job creation rates on base period (log) productivity levels, after controlling for firm size, age, and two-digit sector dummies.15 Using (log) value added per worker as a measure of productivity, we find that firms with higher labor productivity experience higher subsequent job growth.16 The result also provides some partial evidence of creative destruction in MENA economies, in the sense that establishments with higher productivity levels create more jobs. Consistent with the previous analysis, gazelles (fast-growing firms) are more productive and younger than nongazelles in MENA. The left panel of figure 1.16 shows the results of regressions of (log) labor productivity and age on a dummy variable equal to one for gazelle firms. For the two coun- tries with available data, Lebanon and Egypt, gazelles are significantly more productive than nongazelles. Moreover, gazelles are found to be about 32 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa TABLE 1.3 More Productive Firms Create More Jobs Lebanon, Tunisia, West Bank and Gaza, Egypt, Arab Rep., Egypt, Arab Rep., 2005–10 1997–2012 Turkey, 2005–10 2004–12 2007–11 2007–11 (all sectors) (all sectors) (all sectors, 20+) (all sectors) (manufacturing 10+) (manufacturing 10+) Labor productivity 0.039*** 0.029*** 0.007*** 0.022 0.007 Total factor productivity 0.019*** Controlling for firm size and age Yes Yes Yes Yes Yes Yes No. of observations 141,061 129,516 176,665 3,075 7,925 7,988 R2 0.40 0.34 0.03 0.41 0.10 0.09 Source: Calculation based on census data. Note: The dependent variable is the Davis-Haltiwanger-Schuh growth rate. Regressions are weighted by the average size of firms over the growth period. Job growth is measured annually, and productivity is measured at the beginning of the period. In Egypt, data include manufacturing and mining establishments with at least 10 employees; in Turkey, firms with at least 20 employees in all sectors are considered. Labor productivity in Egypt is significant at the 1 percent level when the job creation rate measured over the four-year period (2007–11) is regressed on initial labor productivity in 2007. Significance level: *** = 1%. FIGURE 1.16 Characteristics of Gazelles in the Arab Republic of Egypt, Lebanon, and Morocco a. Labor Productivity and Age in the Arab Republic of Egypt and Lebanon Country Labor productivity (log) Age Egypt, Arab Rep., 2007–11 0.880*** –5.619** Lebanon, 2005–10 0.261*** –4.723*** **p < 0.05. ***p < 0.01. b. Firms that are gazelles in Morocco, 2003–06 12 10 Percentage of firms 8 6 4 2 0 0–1 2–4 5–9 10–14 15–19 20–29 >29 Firm age (years) Source: Calculation based on census data. Note: Gazelles are defined as firms with at least 10 employees in the base year that double employment over any four-year period. Data for Turkey includes only firms with at least 20 employees. Too Little Too Late: Private Sector Growth and Labor Demand 33 4.7 and 5.6 years younger than other firms in Lebanon and Egypt, respec- tively. The right panel figure 1.16 shows that young manufacturing firms are more likely to be gazelles in Morocco than older firms; about 34 percent of all gazelles are at most four years old, and about 55 percent are less than 10 years old. Moreover, we find that gazelles emerge across all sectors of the economy. For example, the largest numbers of gazelles in Tunisia are in textiles, construction, and real estate. In Jordan, the highest incidence is in the construction sector. Nevertheless, gazelles also emerge in most other manufacturing or service sectors in both countries. MENA Needs a Larger Pool of Young Firms and Productive Firms The previous section showed that job creation in MENA countries is weak, but that the fundamentals of job creation in the region are similar to the fundamentals in fast-growing and high-income countries: younger firms and more productive firms create more jobs. Given the fundamen- tals of private sector job creation are the same as elsewhere, why has job creation been so weak? Low firm turnover (firm entry and exit) and weak productivity growth in MENA countries reduce the pool of young firms and productive firms Non-GCC MENA countries have the lowest formal sector entry rates, reducing the pool of young firms that grow and create jobs. MENA coun- tries have some of the lowest entry densities across all regions (figure 1.17, left panel). Entry density is defined as the number of newly registered lim- ited liability firms per 1,000 working-age people, and thus captures entry (of specific) formal sector firms. Formal sector entry in GCC countries is higher than in non-GCC MENA countries, but still relatively low by international comparison. Moreover, it declined somewhat between 2004 and 2012. Firm entry densities are particularly low in Algeria, Iraq, Egypt, and Syria, with less than 0.5 newly registered limited liability firms per 1,000 working-age people. Among MENA countries, Oman had the highest rate of limited liability firm creation per capita (figure 1.17, right panel) between 2009 and 2012. The entry density in Oman was, however, still lower than the average across all 91 (nonfinancial offshore) developing countries with available data. Among non-GCC countries, Tunisia and Morocco had the highest formal sector entry rates per capita; Algeria, Iraq, Egypt, and Syria had the lowest. The entry densities in many fast-growing developing countries such as Serbia, Brazil, Croatia, Chile, and Bulgaria are between two and eight times higher than in Morocco and Tunisia (the two non-GCC MENA countries with the highest entry densities). 34 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 1.17 Entry Density of Formal Sector Limited-Liability Firms across Regions and Countries, 2004–12 a. Regions 6 Number of newly registered limited liability firms per 1,000 working-age people 5 4 3 2 1 0 OECD ECA LAC GCC SSA EAP SAR non-GCC MENA 2004–06 2007–09 2010–12 b. Countries 4.0 3.5 Number of newly registered limited-liability firms per 1,000 working-age people 3.0 Average of all 123 countries 2.5 2.0 Average of all 91 non-OECD countries 1.5 1.0 0.5 0 ria Cro e a zil a Co an Tu a ia Mo y co Me a o Th an Alg d Sri ria a ian Ara q b R ep. lic rke il ati rbi bi an nk xic an Syr ypt, Ira nis ub roc Bra Ch Om d lga e lom Ara b R La Gh Jor ail Se Tu ep Bu Eg Countries of study MENA countries Source: Calculation and Klapper and Love (2010). Note: The average of 123 (91 non-OECD) countries represents the average entry density in all (nonfinancial offshore) countries with available data. Firm turnover rates (entry and exit rates) among MENA countries are low by international standards. High firm entry rates spur experimenta- tion, but also increase the likelihood of the marginal firm’s failure. Thus, one should expect a positive association between firm entry and exit rates in the data. Figure 1.18 plots the entry and exit rates in manufacturing and service sectors across MENA countries and developing countries from other regions. Overall, gross entry and exit rates in MENA countries are remarkably low by international standards. For example, entry and exit rates in manufacturing in Colombia are about 11 and 12 percent, Too Little Too Late: Private Sector Growth and Labor Demand 35 FIGURE 1.18 Firm Turnover across Countries 14 13 12 Turkey (serv) Hungary (all) 11 Columbia (man) Entry rate (%) Russia (man) 10 Tunisia (serv) Turkey (man) 9 Lebanon (serv) 8 Tunisia (man) 7 Morocco (man) Lebanon (man) 6 2 4 6 8 10 12 14 16 Exit rate (%) Non-MENA country MENA-manufacturing MENA-services Source: Calculation based on census data. Note: Entry (exit) in Turkey implies that firm size exceeds (falls below) 20 employees; in Morocco and Colombia it means exceeding (or falling below) 10 employees. man = manufacturing; serv = services. respectively, almost twice as high as in Morocco. Moreover, firm turnover in the services sector is higher than in the manufacturing sector; this reflects the higher dynamism of services, and also the smaller size and lower productivity firms found in the sector. Even after controlling for cohort effects, firm exit rates among MENA countries are low compared with a benchmark country like Turkey. Firm cohorts that entered in the manufacturing sector in Tunisia and Morocco in the early 2000s have high survival, and hence low exit rates, in the first five years after entry. Figure 1.19 shows the survival rates across MENA countries and Turkey. Apart from the different periods for entry cohorts across countries, it is important to note that firm exit definitions in Morocco and Turkey are somewhat different. Figure 1.19 reveals sub- stantially higher survival rates in Tunisia than in West Bank and Gaza and Jordan. In other words, fewer entrants are forced to exit after the first five years in operation, indicating low firm turnover in Tunisia. In contrast, about 60 percent of firms that exceeded 20 employees in Turkey in 2006 are projected to have fallen to fewer than 20 employees again by 2011. The low share of jobs in younger medium or large establishments highlights MENA’s challenge of missing young firms. Figure 1.20 shows the employment distribution by establishment size and age in Egypt, Jordan, Tunisia, and Turkey. It reveals that the share of employment in younger medium or large establishments (i.e., firms with at least 10 employees and created less than 15 years ago) is highest in Turkey, 36 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 1.19 Survival Rates Five Years after Entry 100 90 80 Percent 70 60 50 40 0 1 2 3 4 5 Morocco (manufacturing), 2000 Tunisia, 2002 West Bank and Gaza, 2007 Jordan, 2006 Turkey, 2006 Turkey (manufacturing), 2006 Source: Calculation based on census data. Note: For Morocco-manu, exit implies firm size falls below 10 employees; for Tunisia and Turkey-manu, exit implies firm size falls below 20 employees. The survival rates for Jordan and West Bank and Gaza were only avail- able for the fifth year after entry. The rates for years one through four were estimated assuming that the same fraction of firms exited every year. FIGURE 1.20 Employment Share of Young Medium or Large Establishments 60 40 Percent 30 20 10 Egypt, Arab Rep., 2006 Tunisia, 1996–10 Jordan, 2006 Turkey, 2005 Source: Calculation based on census data. Note: Medium and large: at least 10 employees; young: created less than 15 years ago. Too Little Too Late: Private Sector Growth and Labor Demand 37 significantly lower in Jordan and Tunisia, and particularly low in Egypt. These findings reflect a combination of low firm entry and overall weak employment growth among most young firms and point to severe constraints on business creation and startup growth in MENA. The shortage of medium- and large-size young establishments in Egypt is particularly noteworthy. Figure 1.21 illustrates the distribution of the total number of employees by detailed establishment size and age categories in Egypt and Turkey in 2006. It reveals that employment in Egypt is concentrated in micro establishments independent of their age, FIGURE 1.21 Distribution of Employment, by Firm Size and Age across All Nonfarm Establishments: Arab Republic of Egypt and Turkey, 2006 a. Egypt, Arab Rep. 450 Distribution of employment (in thousands) 400 350 300 250 200 150 20,29 100 9 50 6 3 Age 0 0 Firms do not grow as they 1 9 9 9 9 9 9 0 2 4 5– –1 –4 –9 19 99 ,00 3– get older 0– 0– 10 20 50 ≥1 10 20 Size Lack of jobs in young medium and large firms b. Turkey 450 400 Distribution of employment (in thousands) 350 300 250 200 20–29 150 9 100 6 50 3 Age 0 0 Firms grow as 1 2 4 9 9 9 9 9 9 0 they get older 3– 5– –1 –4 –9 19 99 ,00 0– 0– 10 20 50 ≥1 10 20 Size Source: Calculation based on census data. 38 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa and in the few very old and very large establishments; the latter accounted for less than 300,000 jobs out of more than 7 million in Egyptian eco- nomic establishments in 2006. The concentration of jobs in small, old establishments suggests that, in contrast to Turkey, small Egyptian estab- lishments do not grow over time. Furthermore, the high share of jobs in old establishments in Egypt is cause for concern. Either they remain small on purpose (to stay below the radar of scrutiny by public officials and large competitors), or they are unproductive and might be forced to exit in a more competitive environment (up-or-out dynamics). Firm productivity growth in MENA countries has been low Productivity growth over firms’ life cycle is weak in MENA countries and relatively stronger for the youngest cohorts.17 Figure 1.22 plots the evolu- tion of firm productivity over establishments’ life cycle. The productivity of the youngest cohort is normalized to one so that figure 1.14 effectively depicts life cycle productivity. It illustrates that average productivity of establishments in the U.S., and to a lesser extent also in Turkey, increases with age. After 35 years in operation, U.S. establishments increase their productivity eight-fold on average, while those in Mexico, India, and Turkey increase their productivity about two-or three-fold. In contrast, in Tunisia and Egypt establishments barely increase their productivity over their life cycle on average. Notably, firms are more productive at the beginning of FIGURE 1.22 Labor Productivity Growth over the Life Cycle of Manufacturing Establishments 6 Labor productivity, index = 1 for firms age 0–4 5 United states 4 3 Turkey 2 Mexico India 1 Tunisia 0 <5 5–9 10–14 15–19 20–24 25–29 30–34 35–39 ≥40 Age group United States, 1992–97 Mexico, 1998–2003 India, 1990–95 Egypt, Arab Rep., 2007–11 Turkey, 2005–10 Tunisia, 2009 Source: Calculation based on census data; India, Mexico, and United States from Hsieh and Klenow 2012. Note: The figure shows the average labor productivity over different age cohorts across establishments in manufacturing (weighted by the employment share of four-digit sectors, following Hsieh and Klenow 2012). Too Little Too Late: Private Sector Growth and Labor Demand 39 their life cycle in both countries, but initial productivity gains disappear for older cohorts.18 For example, establishment productivity in Egypt increases two-fold, peaking at the age of 10 years; in contrast, productivity of the sur- viving cohort 40 years after entry is, on average, only 1.4 times higher than the productivity of the youngest cohort. Similarly, Tunisian firms do not increase their productivity beyond 1.1 times the size of the youngest cohort. Productivity growth can ensue from within-firm growth or from the reallocation of resources across firms. We calculated the contribution of both sources of labor productivity growth in MENA countries with the latest available data. Olley and Pakes (1996), among others, show that the way resources are allocated in an economy has implications for productiv- ity growth. In the following, the analysis shows how the divergence in establishment dynamics between MENA and more competitive econo- mies is suggestive of a misallocation of resources. Low efficiency in resource allocation has limited productivity and employment growth. In the previous sections, we have highlighted that there is some evidence for creative destruction in that establishments with higher productivity create more jobs. This finding points to the existence of dynamics involving resource allocation to more productive firms. We quantify the resource misallocation across firms in MENA countries fol- lowing the productivity decomposition approach of Olley and Pakes (1996); these results are then compared with emerging economies from other regions. Figure 1.23 shows the Olley-Pakes covariance term, calculated as the difference between the weighted and un-weighted labor productivity across manufacturing firms.19 The term is a summary measure of the within-industry cross sectional covariance between size and productivity and indicates to what extent more productive firms within industries hire more employees.20 Figure 1.23 shows that the allocative efficiency is lower in Morocco and Egypt than in Chile, Colombia, or Indonesia.21 The results indicate higher resource misallocation (weaker creative destruction) across firms in MENA countries than in other developing regions. In contrast to fast-growing developing countries, large firms in the MENA region are not necessarily more productive. This low allocative efficiency is also reflected in the finding that large firms do not necessarily have higher labor productivity. If large firms are growth constrained (face higher marginal costs of labor and capital), we would expect that they would have higher average levels of value added per worker (and capital), to the extent that average and marginal products of labor (or capital) move together.22 In turn, small firms would be expected to have higher average labor productivity if they are more growth constrained relative to large firms (for given levels of value added per capital).23 The left panel in figure 1.24 shows the average log labor productivity for different firm size categories in Lebanon and Tunisia. Labor productivity hardly varies 40 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa among firms size categories in Lebanon. In contrast, firm productivity is lower for larger size categories in Tunisia, suggesting that small firms are more growth constrained (for given values of capital).24 These findings starkly contrast with Turkey (figure 1.24b) where large firms are much more productive (in terms of labor productivity and TFP). They also contrast with the findings of Hsieh and Olken (2014), who argue that large firms are (more) growth constrained in India, Indonesia, and Mexico on the basis that the average value added per labor and capital is higher among large firms in these countries. FIGURE 1.23 Gap between Weighted and Unweighted Labor Productivity 1.00 Weighted/unweighted labor productivity gap 0.75 0.50 0.25 0 Morocco Egypt, Arab Chile Colombia Indonesia Rep. Source: Calculation based on census data. Data points from Chile, Colombia, and Indonesia are from Bartelsman, Haltiwanger, and Scarpetta 2004. Note: Labor productivity gaps are the weighted (by employment shares) average of two-digit industries. Data cover firms with more than 10 employees in all countries. FIGURE 1.24 Productivity, by Firm Size in Tunisia, Lebanon, and Turkey a. Tunisia and Lebanon b. Turkey 1.2 1.2 3 Total factor productivity index = 1 for age 1–19 index = 1 for age 1–2 Labor productivity, Labor productivity, 1.1 1.1 2 1.0 1.0 1 0.9 0.9 0 1–2 3–5 6–9 10–49 50–99 ≥100 1–19 20–49 50–99 100–249 250–499 500+ Firm size Firm size Lebanon log LP Tunisia log LP Turkey log LP Turkey log TFP Source: Calculation based on census data. Note: Labor productivity is the average value added per employee and average TFP is weighted by value added. Too Little Too Late: Private Sector Growth and Labor Demand 41 BOX 1.3 Firm Dynamics and Productivity Growth in Morocco Productivity decomposition shows that net firm entry and improvements in allocative efficiency contributed largely to aggregate productivity growth in the Morocco’s manu- facturing sector between 1996 and 2006. However, the contribution of surviving firms (incumbents) to aggregate productivity growth was close to zero. The methodology proposed in Foster et al. (2001) was used to decompose productivity growth according to the following equationa: Δpst = ∑θ i ∈C is, t −1 Δpis + ∑ Δθ i ∈C is, t (pi, t −1 − ps, t −1 ) + ∑ Δθ i ∈C is, t Δpit + ∑θ i ∈C is, t (p i, t − ps, t −1 ) − ∑θ i ∈S is, t −1 (pi, t −1 − ps, t −1 ) where p refers to productivity; θ refers to a firm’s share of total sector output (thought of in terms of revenues); and the subscripts t, s, i, C, N, and S refer to time, sector, firm, con- tinuing (surviving) firms, new entrants and exiting firms, respectively. The first term on the right-hand side of equation (1) refers to the within effect. It represents internal restructur- ing effects stemming from changes in productivity of surviving firms. The second term shows the between effect for surviving firms. This is positive when the market shares increase for those survivors with above-average productivity in the previous period (t−1). The third term is an additional covariance term that is positive when market share increases (falls) for establishments with growing (falling) productivity. The BHC decomposition combines these two terms together by calculating the between effect as the sum of changes in market share weighted by ending period productivity (period t). The final two terms represent the contributions of firm entry and exit, respectively. These will be positive when there is entry (exit) of above (below) average productivity firms. The results are summarized in figure B1.3.1.b The within effect is quite unstable, with large oscillation around a mean of zero suggesting that surviving firms do not make a sys- tematic contribution to aggregate productive growth. Moreover, the lack of upward trends in the within effect point to the fact that surviving firms did not systematically improve their technical efficiency (through the adoption of better technologies, management prac- tices, worker training, and so forth) between 1996 and 2006. The between effect is negative over the entire sample period, but increased in later years suggesting that the allocative efficiency has improved in the Moroccan manufacturing sector while that scope for improvements remains. Between 1998 and 2004, the average productivity growth resulting from net entry was .03, or about 43 percent of average growth in the same period. Moreover, the contribution of net-entry to aggregate productivity growth seems to have accelerated (continued on next page) 42 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa BOX 1.3 Continued between 2000 and 2002. The contribution of net entry to productivity growth was largest in the electrical machinery sector, where the entry rate of large startups was highest in the sample period. FIGURE B1.3.1 Decomposition of Firm Productivity Growth in Morocco’s Manufacturing Sector, 1996–2006 0.10 Contribution to productivity growth 0.05 0 –0.05 –0.10 –0.15 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Covar Within Net entry Between Source: Calculation based on Morocco manufacturing census. Note: The methodology is explained in detail in the appendix and in Sy (2014). Notes: a. The decomposition is done using a window of three years to the contribution of entry to aggregate productivity growth. See Sy (2014) for details. b. The productivity decomposition cannot be conducted for Egypt, Jordan, West Bank and Gaza as data on firm exit or output are missing. See the data section in Appendix B for more details. Notes 1. To the best of our knowledge firm census data, including informal firms (below 5 employees), from all nonagriculture economic sectors has only been applied in research for very few other developing countries from other regions including India, Indonesia, and Mexico. 2. The data were collected over the course of more than a year. In the Arab Republic of Egypt, Tunisia, and Turkey the entire data sets were only acces- sible in the offices of the corresponding statistical departments in Cairo, Tunis, and Istanbul, respectively. 3. The different methodologies used and additional country specific analysis are described in detail in the corresponding companion papers of this report, including Sy (2014) for Morocco; Rijkers, et al. (2013) for Tunisia; Al Kadi (2014) for Jordan; Hussain and Schiffbauer (2014) for Egypt; and Atiyas and Bakis (2014) for Turkey. Too Little Too Late: Private Sector Growth and Labor Demand 43 4. Under perfect competition in input and output markets labor should move to the sector with the highest marginal productivity (i.e., wage) equalizing mar- ginal rates across sectors over time. In the presence of market failures, distor- tions, and rigidities (e.g., because of product or labor market regulations) wages and labor flows do not fully adjust driving a wedge between marginal productivities across sectors. While the impact of these distortions is difficult to measure, it is likely that they are more severe in developing countries. For example, Herrendorf and Valentinyi (2012) find large sectoral TFP differ- ences relative to the United States in agriculture, manufacturing, and services. Moreover, the sectoral TFP gaps relative to the United States are larger in agriculture and services than in manufacturing. 5. In fact, under a Cobb-Douglas production function specification, the mar- ginal productivity of labor is the average productivity multiplied by the share of labor in GDP. Thus, large differences in labor shares, i.e., in capital intensities across sectors, drive a wedge between marginal and average labor productivity levels. For example, among the aforementioned sectors, public utilities and mining are likely to have higher capital intensities potentially overstating their measured marginal productivities when approximated with averages. McMillan and Rodrik (2011) argue, however, that in the case of the other sectors, which employ most labor, it is not clear that there is a significant bias. Thus, we assume in the following that large gaps in average productivity across sectors within a country are positively correlated with the underlying unobservable gaps in marginal productivi- ties across sectors. See also Hsieh and Olken (2014) for a detailed discussion under which conditions the average and the marginal products of capital and labor move together. 6. For the years 2005–10, we have data on the amount of employees by sector with a primary, secondary, or post-secondary degrees. We assign 0, 6, 12 and 16 years of education to employees with no degree, primary degree, second- ary degree, and post-secondary degrees, respectively. Using a standard Mincerian technique and assuming a 10% return to each year of schooling, we assign each employee a human capital equal to e(1*years). 7. For more details, see World Bank (2014d). In contrast to all other countries, the employment distribution in Tunisia is based on firms instead of establish- ments; hence the share of jobs in large establishments is potentially slightly overstated. We note, however, that this bias is expected to be small since, for example, in Egypt only 1 percent of establishments were not firms, but part of larger entities in 2006. 8. Jordan’s, and to a lesser extent Tunisia’s, relatively high concentration of employment in large firms is in part explained by higher inflows of foreign direct investment (FDI). That is, 19 percent of all large firms in Jordan and Tunisia are foreign owned. These firms account for 30 and 19 percent of employment generated by large establishment in each countries, respectively (figure C.1 in appendix C). 9. These trends are consistent with survey data from the Egypt Labor Market Panel Survey showing an increase in the share of Egyptians working in the informal economy between 1998 and 2006 as well as between 2006 and 2012, respectively: jobs that provide neither social insurance nor a formal labor contract increased from 53 percent in 1998 to 61 percent in 2012. See World Bank (2014a). The report also shows that the trend to more informal work 44 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa materialized in all sectors. In addition, there has been a shift towards irregular work in the second half of the 2000s. 10. US gazelles are based on a somewhat stricter definition: firms whose sales and employment have at least doubled over the same four-year period (Spencer 2011). 11. This definition avoids considering micro businesses as gazelles that increased employment, for example, from two to four over a four-year period by hiring two more family members. 12. The results are based on regressions of the (Davis-Haltiwanger-Schuh) job growth rate following the methodology of Haltiwanger, Jarmin, and Miranda (2010) for the United States. All regressions control for two-digit sector and year dummies. The census data include all firms and economic sectors apart from agriculture (as in the US data). Thus, the results measure the aggregate job creation rate. The graph plots the coefficient estimates of the firm size dummies of two regressions. First, job growth is regressed on firm size dum- mies and controls only (blue-dashed line). Second, job growth is regressed on firm size dummies and controls as well as firm age dummies (red solid line). 13. The census data from the other MENA countries are not appropriate to apply the Haltiwanger methodology. Firm age in West Bank and Gaza is not included, while in Egypt the census data are not in panel format. In Jordan and Morocco the census data are only in a panel format for a subset of firms (e.g., manufacturing sectors). 14. Note that the analysis shows the relationship between average plant employ- ment and age based on cross-section census data, which conflates size differ- ences between cohorts at birth and employment growth of a cohort over its life cycle. Thus, when interpreting the results as reflecting dynamics over time, it is implicitly assumed that the relative size differences between differ- ent age cohorts are time-invariant. 15. The order of magnitudes of the coefficients are not directly comparable in Turkey and Egypt. The analysis tracks the same firms over time if they have at least 20 employees or 10 employees, respectively. 16. The corresponding coefficients are all statistically significant (at the 1 percent level) apart from Egypt. However, in Egypt, capital stocks of establishments are also taken into account, allowing for calculation of the preferred measure of total factor productivity (TFP) following the method of Caves, Christensen, and Diwert (1982). 17. Again, the analysis is based on cross section census data so that we have to assume that cohorts’ life cycle characteristics are time invariant. 18. Results are similar for manufacturing firms in Morocco, which increase their average productivity almost three-fold five years after entry, while average productivity is lower in the following five years. The results for Morocco are not reported here as the Moroccan (cross section) data only include firms above a certain size threshold (e.g., more than 10 employees). Note that in Turkey only firms exceeding 20 employees in panel format are included in yearly cross sections. 19. See also Bartelsman, Haltiwanger, and Scarpetta 2004 and 2013 or Hsieh and Klenow (2014). 20. Labor productivity gaps are the weighted average of two-digit industries (weighted by employment shares). 21. The data covers firms with more than 10 employees in all countries. Too Little Too Late: Private Sector Growth and Labor Demand 45 22. For example, Hsieh and Olken (2014) analyze differences in average labor productivity by firm size across countries and discuss the conditions under which the average and marginal products of labor move together. 23. In an efficient economy, competitive forces lead to a reallocation of resources to more productive firms equating (marginal) productivities across different categories of firms over time. In developing countries, firms are more likely to be growth constrained because of high growth opportunities (from adopt- ing new foreign technologies) paired with market failures (for example, access to finance, markets) preventing firms from harnessing these investment opportunities. 24. The same analysis is performed for manufacturing firms in Morocco and Egypt. However, there is no reliable data on firms with fewer than 10 employees. The findings suggest that larger firms in Morocco are more productive while in Egypt labor productivity is higher and TFP is lower for larger size categories. References Acs, Z., J., W. Parsons, and S. Tracy. 2008. “High-Impact Firms: Gazelles Revisited.” Working Paper 328, SBA Office of Advocacy, Washington, DC. Aghion, P., R. Blundell, R. Griffith, P. Howitt, and S. Prantl. 2009. “The Effects of Entry on Incumbent Innovation and Productivity.” The Review of Economics and Statistics 91 (1): 20–32. Aghion, P., R. Burgess, S. Redding, and F. Zilibotti. 2008. “The Unequal Effects of Liberalization: Evidence from Dismantling the License Raj in India.” American Economic Review 98(4): 1397–412. Aghion, P., R. Burgess, S. Redding, and F. Zilibotti. 2006. “Entry Liberalization and Inequality in Industrial Performance.” Journal of the European Economic Association 3 (2–3): 291–302. Aghion, P., C. Harris, P. Howitt, and J. Vickers. 2001. “Competition, Imitation and Growth with Step-by-Step Innovation.” The Review of Economic Studies 68 (3): 467–92. Al-Kadi, Dalia. 2014. “Firm dynamics in Jordan.” Mimeo, World Bank, Washington, DC. Atiyas, Izak and Ozan Bakis. 2014. “Firm Dynamics and Job Creation in Turkey.” Mimeo, Sabanci University, Istanbul, Turkey. Ayyagari, M., A. Demirguc-Kunt, and V. Maksimovic. 2011. “Small vs. Young Firms across the World: Contribution to Employment, Job Creation, and Growth.” Policy Research Working Paper 5631, World Bank, Washington, DC. Bars, F., S. Boiteux, M.-F. Clerc-Girard, and S. Janczak. 2006. “Entrepreneurship and the High Growth Companies: The Evolution of the Gazelles and Their Ties to the Territory.” Business School Working Paper 2006–02, ICN, Nancy, France. Berman, E., and S. Machin. 2004. “Skill-Biased Technology Transfer around the World.” Oxford Review of Economic Policy 16: 12–22. Bartelsman, Eric, J. Haltiwanger, and S. Scarpetta. 2004. “Microeconomic Evidence on Creative Destruction in Industrial and Developing Countries.” World Bank Policy Research Paper 3464, World Bank, Washington, DC. 46 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Bartelsman, Eric, John Haltiwanger, and Stefano Scarpetta. 2013. “Cross- Country Differences in Productivity: The Role of Allocation and Selection.” American Economic Review 103 (1): 305–34. Blanchflower, D., and S. M. Burgess. 1998. “New Technology and Jobs: Comparative Evidence from a Two-Country Study.” Economics of Innovation and New Technology 5: 109–38. Bottazzi, G., E. Cefis, G. Dosi, and A. Secchi. 2007. “Invariances and Diversities in the Patterns of Industrial Evolution: Some Evidence from Italian Manufacturing Industries.” Small Business Economics 29: 137–59. Caves, D.W., Christensen, L.R., and Diwert, W.E. 1982. “The Economic Theory of Index Numbers and the Measurement of Input, Output, and Productivity.” Econometrica 50: 1393–414. Coad, A., and W. Hoelzl. 2010. “Firm Growth: Empirical Analysis.” Paper on Economics and Evolution 1002, Max Planck Institute of Economics, Jena, Germany. Conte, A., and M. Vivarelli. 2010. “Imported Skill Biased Technological Change in Developing Countries.” Developing Economies 49: 36–65. Dahi, Omar and Firat Demir. 2008. “The Middle East and North Africa.” In International Handbook of Development Economics, edited by A.K. Dutt and J. Ros, Vol. 2, 522–535. London: Edward Elgar Publishing. Davidsson, P. and F. Delmar. 2006. “High-Growth Firms and Their Contribution to Employment: The Case of Sweden 1987–96.” In Entrepreneurship and the Growth of the Firm, edited by P. Davidsson, F. Delmar, and J. Wiklund. Cheltenham, U.K.: Edward Elgar. Davis, Steven J., John C. Haltiwanger, and Scott Schuh. 1996. Job Creation and Destruction. MIT Press. Doemeland, Doerte and Marc Schiffbauer. 2014. “Structural Change in the Middle East and North Africa.” Mimeo, World Bank, Washington, DC. Duarte, Margarida and Diego Restuccia. 2010. “The Role of the Structural Transformation in Aggregate Productivity.” Quarterly Journal of Economics 125: 129–73. Eichengreen, Barry, and Poonam Gupta. 2011. “The Service Sector as India’s Road to Economic Growth.” Working Paper 16757, NBER, Cambridge, MA. Foster, L., J. Haltiwanger, and C. J. Krizan. “Aggregate Productivity Growth: Lessons from Microeconomic Evidence.” In New Developments in Productivity Analysis, edited by C. R. Hulten, E. R. Dean, and M. J. Harper. Chicago: University of Chicago Press, 2001, 303–63. Fuentes, O., and S. Gilchrist. 2005. “Trade Orientation and Labor Market Evolution: Evidence from Chilean Plant-Level Data.” In Labor Markets and Institutions, edited by J. Restrepo and A. Tokman. Santiago: Central Bank of Chile. Goedhuys, M., and L. Sleuwaegen. 2009. “High-Growth Entrepreneurial Firms in Africa: A Quantile Regression Approach.” Small Business Economics 34(1): 31–51. Hall, B. H. 1987. “The Relationship between Firm Size and Firm Growth in the U.S. Manufacturing Sector.” Journal of Industrial Economics 35: 583–600. Hall, B. H., F. Lotti, and J. Mairesse. 2008. “Employment, Innovation, and Productivity: Evidence from Italian Microdata.” Industrial and Corporate Change 17: 813–39. Too Little Too Late: Private Sector Growth and Labor Demand 47 Haltiwanger, John, Ron S. Jarmin, and Javier Miranda. 2010. “Who Creates Jobs? Small versus Large versus Young.” Review of Economics and Statistics 95 (2): 347–61. Hanson, G., and A. Harrison. 1999. “Trade and Wage Inequality in Mexico.” Industrial and Labor Relations Review 52: 271–88. Harrison, R., J. Jaumandreu, J. Mairesse, and B. Peters. 2008. “Does Innovation Stimulate Employment? A Firm-Level Analysis Using Comparable Microdata from Four European Countries.” Working Paper 14216, NBER, Cambridge, MA. Hart, P. E., and N. Oulton. 1996. “The Growth and Size of Firms.” Economic Journal 106 (3): 1242–52. Henrekson, M., and D. Johansson. 2010. “Gazelles as Job Creators: A Survey and Interpretation of the Evidence.” Small Business Economics 35 (2): 227–44. Herrendorf, Berthold, Richard Rogerson, and Akos Valentinyi. 2013b. “Growth and Structural Transformation.” In Handbook of Economic Growth, edited by Philippe Aghion and Steven N. Durlauf. Philadelphia, PA: Elsevier. Herrendorf, Berthold, and Akos Valentinyi. 2012. “Which Sectors Make Poor Countries so Unproductive?” Journal of the European Economic Association 10: 323–41. Hidalgo, C. A., B. Klinger, A. L. Barabasi, and R. Hausman. 2007. “The Product Space Conditions and the Development of Nations.” Science 317: 482. Hsieh, Chang-Tai, and Peter J. Klenow. 2012. “The Life Cycle of Plants in India and Mexico.” Working Paper 18133, NBER, Cambridge, MA. Hsieh, Chang-Tai, and Benjamin Olken. 2014. “The Missing ‘Missing Middle’.” Working Paper 19966, NBER, Cambridge, MA. Hussain, Sahar, and Marc Schiffbauer. 2014. “Struggling for Growth: Labor Demand and Job Creation in Egypt.” Mimeo, World Bank, Washington, DC. Klapper, L., and I. Love. 2010. “The Impact of the Financial Crisis on New Firm Registration.” Policy Research Working Paper 5444, World Bank, Washington, DC. Kuznets, Simon. 1996. Modern Economic Growth. New Haven: Yale University Press. Mansfield, E. 1962. “Entry, Gibrat’s Law, Innovation, and the Growth of Firms.” American Economic Review 52 (5): 1023–51. McMillan, M., and D. Rodrik. 2011. “Globalization, Structural Change and Productivity Growth.” Working Paper 17143, NBER, Cambridge, MA. Nabli, Mustafa, and M. A. Veganzones-Varoudakis. 2002. “Exchange Rate Regime and Competiveness of Manufactured Exports. The Case of MENA Countries.” MENA Working Paper 27, World Bank, Washington, DC. Olley, G., and A. Pakes. 1996. “The Dynamics of Productivity in the Telecommunications Equipment Industry.” Econometrica 64: 1263–79. Parente, S., and E. Prescott. 1999. “Monopoly Rights: A Barrier to Riches.” American Economic Review 89: 1216–33. Piva, M., and M. Vivarelli. 2004. “The Skill Bias in Italy: A First Report.” Economics Bulletin 15: 1–8. Rijkers, Bob, Hassan Arrouri, Caroline Freund, and Antonio Nucifora. 2013. “Which Firms Create the Most Jobs in Developing Countries.” Working Paper, World Bank, Washington, DC. 48 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Rodrik, Dani. 2013. “Unconditional Convergence in Manufacturing.” The Quarterly Journal of Economics 128 (1): 165–204. Sahnoun, Hania, and Marc Schiffbauer. 2014. “Mapping MENA’s Manufacturing—The Export Performance and Prospects of MENA Countries since the First Structural Reforms.” Mimeo, World Bank, Washington, DC. Sy, Abdoulaye. 2014. “Firm Dynamics, Employment and Productivity Growth in Morocco.” Working Paper, World Bank, Washington, DC. Tracy Jr., Spencer L. 2011. “Accelerating Job Creation in America: The Promise of High-Impact Companies.” Corporate Research Board, LLC, Washington, DC. Van Reenen, J. 1997. “Employment and Technological Innovation: Evidence from U.K. Manufacturing Firms.” Journal of Labor Economics 15: 255–84. Vivarelli, M. 2012. “Innovation, Employment and Skills in Advanced and Developing Countries: A Survey of the Literature.” IZA Discussion Paper Series 6291, Institute for the Study of Labour (IZA), Bonn. World Bank. 2009. From Privilege to Competition: Unlocking the Private-Led Growth in the Middle East and North Africa. MENA Development Report. Washington, DC: World Bank. ———. 2011a. “Informality in the Middle East and North Africa.” Working Paper, World Bank, Washington, DC. ———. 2011b. “Financial Access and Stability: A Road Map for the Middle East and North Africa.” Working Paper, World Bank, Washington, DC. ———. 2012. “Natural Resource Abundance, Growth, and Diversification.” Mimeo, World Bank, Washington, DC. ———. 2013. “Jobs for Shared Prosperity: Time for Action in the Middle East and North Africa.” Mimeo, World Bank, Washington, DC. ———. 2014a. “More Jobs, Better Jobs: A Priority for Egypt.” Mimeo, World Bank, Washington, DC. ———. 2014b. “Why Doesn’t MENA Export More? A Firm-Level Perspective.” Mimeo, World Bank, Washington, DC. ———. 2014c. “Structural Change and Gender in Morocco.” Mimeo, World Bank, Washington, DC. ———. 2014d. “Tunisia Development Policy Review.” Mimeo, World Bank, Washington, DC. CHAPTER 2 Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth The previous chapter established that, in MENA, employment growth is limited by the small pool of younger firms and more productive firms. This chapter presents several case studies that show how various policies across MENA countries tend to lower competition and create unequal opportunities between entrepreneurs, thereby limiting the number of young firms and productivity growth. The case studies cover several policies ranging from energy subsidies to industry, cumbersome business regulations, uneven implementation of these regulations, to barriers to foreign direct investment. The Schumpeterian growth model predicts that fast-growing economies are characterized by specific firm dynamics echoing neck-and-neck competition market structures. Aghion et al. (2001) predict that the majority of sectors in fast-growing economies will exhibit high firm turnover, higher within-firm productivity growth, and low resource misallocations. In contrast, the firm dynamics in MENA identified in chapter 1 resemble market structures in which a few leading firms have large (exogenous) cost advantages, while potentially large numbers of infor- mal micro-firms use unproductive vintage technologies to serve local market niches. The Schumpeterian growth framework predicts that sec- tors in which leading firms enjoy large, exogenous cost advantages because of policy distortions should display a number of traits that dis- tinguish them from sectors in which leading firms do not enjoy such privileges. That is, Aghion et al. (2001) predict that sectors dominated by firms with large and exclusive cost advantages should face less com- petition and exhibit less entry and exit. Likewise, sectors dominated by these firms should have a more skewed firm distribution, characterized by a large privileged market leader, and a potentially large number of small and/or informal micro-firms using vintage technologies to serve local market niches. 49 50 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Low firm turnover, productivity growth, and resource misallocation, which hold back job growth in MENA, point to lack of competition. The lack of both entering and growing young firms also reduces the pool of firms that can put competitive pressure on incumbent firms. Thus, incumbents face less pressure to become more cost-effective over time or exit. Moreover, in the process of creative destruction, resources are real- located to more productive firms, either through the higher growth of more productive firms, or through firm churning, whereby the least pro- ductive firms are forced to exit. Chapter 1 contends that this process is undermined by various policies in MENA. Competition is a catalyst in the process of creative destruction, which has been identified in chapter 1 as the main driver of long-term employ- ment growth in MENA. Bartelsman, Haltiwanger, and Scarpetta (2004) demonstrate that for several Eastern European countries the threat of entry serves as a disciplining device, forcing incumbents to innovate more rapidly. For the MENA countries with available data, we also find a positive correlation between net entry and incumbents’ productivity growth in four-digit industrial sectors. In other words, sectors that are more contestable—that have more competition from entering firms— tend to exhibit rapider productivity improvements among existing firms. This chapter offers several case studies that demonstrate how policies in MENA shape (distort) private sector competition and thus firm dynamics associated with higher job growth. Thus, while chapter 1 doc- umented that firm dynamics in MENA are consistent with weak neck- and-neck competition in the sense of Aghion et al. (2001), this section highlights specific policies in MENA countries that lower competition by providing large exogenous cost advantages, in the form of policy priv- ileges to a few leading firms. Increasing the pool of younger firms and more productive firms—the engines of job creation—requires more competition and equal opportuni- ties for all entrepreneurs; in other words, it requires the removal of poli- cies that undermine competition by tilting the level playing field. Increasing private sector competition requires a comprehensive approach to compe- tition policy since a level playing field for all firms can be distorted in many different ways. For instance, the literature identifies several potential dis- tortions to fair competition in MENA countries including energy subsi- dies, access to finance, and access to land (World Bank 2009, 2011). The following sections summarize the main results from case studies evaluating: (a) the employment spillovers from FDI in Jordan (“Attracting FDI in Services Sparked Job Growth in Domestic Firms in Jordan” section); (b) the impact of mobility restrictions on firm dynamics in the West Bank; (c) the link between job growth and the quality of the business environment in Morocco (“Business Regulations Limit Employment Growth among Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 51 Young Firms in Morocco” section); (d) the impact of energy subsidies on employment and resource misallocation in the Arab Republic Egypt (“Energy Subsidies in the Arab Republic of Discourage Growth in Labor- Intensive Industries” section); (e) and how discretionary policy implemen- tation by public officials affect competition and innovation (“Discriminatory Policy Implementation Deters a Level Playing Field in MENA” section). Attracting FDI in Services Sparked Job Growth in Domestic Firms in Jordan We show that FDI inflow in Jordan led to a partial crowding-out of domestic firms in the same sector, but had positive spillovers on firms in supplying or using sectors. The analysis shows that FDI spillovers depend on specific characteristics of the domestic suppliers—domestic suppliers only benefit if they provide services, not goods, or if they are young. In turn, the employment contraction among firms in the same industry is concentrated in old and small firms. The results show that FDI benefits primarily the type of domestic firms that have been identified in chapter 1 to drive job growth. Moreover, domestic manufacturing firms (suppliers) did not benefit from FDI spillovers, possibly reflecting a combination of weak competition in the sector and the absence of larger scale technical supplier support programs. Overall, the findings suggest that removing the remaining restrictions to FDI into service sectors in Jordan is expected to generate employ- ment growth among domestic firms. Technology transfers through FDI to domestic suppliers, downstream sectors, or competitors are considered to have played a major role in the process of technology adoption, structural change and job creation of many East Asian economies including China, India, and Malaysia (Rodrik 2004, 2008; Sutton 2005). Policymakers in many developing economies provide incentives to attract FDI in the expectation that FDI inflows bring capital, new technologies, marketing techniques, and management skills. In fact, FDI is considered as one of the major channels for fostering technology transfers to developing countries (Keller 2004). Technology spillovers may take place when local firms copy technologies either through observation or by hiring workers trained by foreign affiliates. Moreover, entries of foreign firms change the market structure in the domestic economy typically increasing competition. In particular, it has been shown that FDI in backbone service sectors can increase the quality of services benefitting using firms (Arnold et al. 2012). This section aims to quantify the effects of FDI inflow on jobs in Jordan by accounting both on direct as well as spillover effects. Following the methodology of Javorcik (2004),1 the information on foreign 52 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa ownership used is based on the establishment census data in 2006 and 2011 covering both manufacturing and service sectors. The census includes panel information (and sample weights) for a subset of 15,500 establishment covering 53 percent of total employment in the economy (relative to labor force survey data). Firms with a share of foreign owner- ship of more than 10 percent account for 19 percent of all large firms in 2006 as well as 30 percent of total employment among large firms (see figure C.1 in appendix C). Data from the establishment census are com- bined with detailed data on input-output tables for about 80 two-digit sectors in 2006. This allows for approximating the linkages between for- eign firms and the domestic suppliers and users of foreign intermediates and services. Lamla and Schiffbauer (2014) provide more details on data and methodology, and additional results and robustness tests.2 The approach allows us to distinguish between horizontal spillovers to firms in the same sector and vertical spillovers to domestic suppliers (backward linkages) and downstream users (forward linkages). The dis- tinction is important, as vertical spillovers are more likely: while foreign firms have an incentive to prevent technology leakages to local competi- tors in the same industry, they benefit from technology diffusion to suppliers through improved input quality. In Lithuania and Romania, Javorcik (2004) and Javorcik and Spatareanu (2011) find positive spillovers from manufacturing FDI only for domestic suppliers in manufacturing (backward linkages). Jordanian firms appear to be relatively well placed to benefit from FDI spillovers in the form of foreign technology transfers that increase pro- ductivity and ultimately job growth. Jordan has some of the highest shares of foreign investment in its total investments: almost half of total invest- ment in Jordan is of foreign origin, according to the WDI in 2009. Likewise, FDI in Jordan accounted on average for about 11 percent of GDP from 2000 to 2009, which is among the highest shares in emerging economies. Figure 2.1 provides the breakdown of FDI inflows into Jordan from 2003 to 2010 by sector. More than half of all FDI is in real estate; FDI in manufacturing accounts for another 30 percent; foreign invest- ments in all other sectors are negligible, at only around 10 percent of total FDI combined. This pattern is comparable with other MENA countries (apart from FDI in the oil sector), but contrasts with the high shares of FDI inflows into manufacturing and ICT services in India, Indonesia, China, and Brazil. Foreign firms crowd out both small and old domestic firms in the same industry. Job creation declines among domestic firms producing the same product or service as foreign firms which operate in the same four-digit industry. These domestic firms are directly competing with foreign firms, which are often more productive at introducing superior technologies. Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 53 FIGURE 2.1 Share of FDI Inflows, by Sector, Selected MENA Countries, 2003–10 Syrian Arab Republic Egypt, Arab Rep. Iran, Islamic Rep. Iraq Jordan Lebanon Turkey China Brazil Indonesia India 0 10 20 30 40 50 60 70 Percentage of total FDI Mining Real estate Hotels and tourism Information technology and communications Financial and other services Chemicals and machinery Other manufacturing Source: Calculations based on FDI markets database. 54 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa BOX 2.1 FDI into Services Sectors Is Often Restricted in MENA Countries Restrictions on foreign firms entering ser- all other sectors, hence weak performance vice sectors in MENA countries are among in these services might lead to weak links in the highest in the world. These restrictions the economy dragging down productivity are generally larger in GCC countries rela- in using sectors (Jones 2011; Kremer 1993). tive to non-GCC MENA countries.a They In this case, foreign entry into these services are particularly high for professional ser- can improve performance and growth in vices (such as accounting, consulting, judi- using sectors by removing weak links.b ciary), transport, and finance; some service Jordan imposed some major restrictions trade restrictions also exist in telecommuni- on foreign entry in several backbone ser- cations and retail trade (figure B2.1.1). The vice sectors. Figure B2.1.2 summarizes partial protection from foreign competition restrictions on foreign firms to entry into in domestic service sectors has potentially different service sectors in MENA coun- led to lower productivity growth of services. tries, and (unweighted regional) averages Backbone services (banking, telecommuni- from other regions. Countries are ranked cation, transport) are important inputs for by their restrictiveness to foreign entry FIGURE B2.1.1 Service Trade Restriction Index, by Sector and Region 70 60 50 STRI index, 0 to 100 40 30 20 10 0 Financial Professional Retail Telecom Transport GCC SAR MENA EAP AFR LAC ECA Source: Calculation based on World Bank Service Trade Restriction Database (Borchert, Gootiiz, and Mattoo 2012). Note: STRI reflects simple country averages. The higher the index, the more restrictions are imposed on foreign firm entry: zero implies no restrictions on foreign owners, 100 implies foreigners are not allowed to operate in the sector at all. The Service Trade Restriction Index ranges from 0 (least restrictive) to 100 (most restrictive). (continued on next page) Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 55 BOX 2.1 Continued FIGURE B2.1.2 Service Trade Restrictions in Transportation Services in MENA, 2008 100 90 Service trade restrictions index 80 (100 = most restrictive) 70 60 50 40 30 20 10 0 . ECA LAC roc co ep. eria EAP dan ano n isia Rep GCC ep. Mo en, R Alg Jor Leb Tun Arab ic R am Yem pt, , Isl Egy Iran Transport Professional Financial Retail Telecom Source: Calculation based on World Bank Service Trade Restriction Database (Mattoo et al. 2012). Note: The higher the index, the more restrictions are imposed on foreign firm entry: zero implies no restrictions on foreign owners, 100 implies foreigners are not allowed to operate in the sector at all. across all service sectors (from lowest to transport services. The index reveals that highest). In 2008, Jordan imposed higher some transport sectors are virtually closed restrictions than the average country in to foreign competition in Jordan. For Latin America and Caribbean (LAC), example, in contrast to the majority of the Eastern Europe and Central Asia (ECA), or 81 coastal countries in the sample, Jordan East Asia and Pacific (EAP). Professional restricts foreign investors’ access to all and transport services were the most auxiliary port services (cargo handling, restricted in Jordan. The transport sector storage, maritime agency services, and comprises air, land, maritime, and auxiliary freight forwarding). Notes: a. A new (2008) World Bank database allows comparison of service trade restrictions in five key service sectors across 103 countries, includ- ing 13 MENA countries. The database on service trade restrictions provides comparable information across countries for the following five service sectors: telecommunications, finance, transportation, retail, and professional services. The indicators focus on policies and regula- tions discriminating against foreign service providers. Information on the de facto implementation of policies is captured in some cases, such as the extent to which the process of granting licenses is transparent and accountable. See Borchert, Gootiiz, and Mattoo (2012) for a detailed description of the data and sampling. b. Marotta, Ugarte, and Baghdadi (2014) analyze the extent to which weak links reduce productivity in the Tunisian economy. They show that weak links are consistently associated with lower levels of productivity per worker. Moreover, the authors identify an important spatial dimension in that the probability of facing weak links in intermediate inputs is higher in inland regions. In addition, economic sectors exposed to more international trade are less likely to be affected by the weak links. 56 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Thus, the results suggest at least a partial crowding-out effect of jobs in domestic firms to jobs in foreign firms.3 This crowding-out effect, however, is limited to small (less than 30 employees) or old (created before 1990) domestic establishments. Supposing that small and old establishments are less productive, the finding is consistent with a com- petition effect: employment is only crowded out by FDI in the least pro- ductive domestic firms, which either shrink (lose market shares) or exit. However, the entry of foreign firms leads to growth of domestic sup- pliers which are young and/or operate in service sectors. On aggregate, the analysis provides no evidence to suggest that the presence of foreign firms in 2006 led to employment growth over the subsequent five years (between 2006 and 2011) among domestic suppliers (backward linkages). Domestic suppliers do not have a stronger growth pattern if goods or services are supplied to sectors with a high initial concentration of foreign firms (that is, FDI). The analysis does show, however, that the existence of backward linkages from FDI spillovers depends on specific character- istics of the domestic suppliers. Domestic suppliers only grow if they pro- vide services, not goods, or if they are young—created after 1990. The results are summarized in table E.1 in appendix E. Thus, those domestic establishments supplying services to sectors with a high initial share of foreign firms experience higher subsequent employment growth. Thus, job creation among domestic service suppliers is strong in the medium term after the entry of foreign firms. Moreover, the findings suggest that young establishments, which started operations after 1990, created more jobs from 2006 to 2011 when they supplied their good or services to sec- tors with a larger presence of foreign firms in 2006. While the age of supplying firms matters—i.e., only young firms create more jobs—the size of domestic suppliers does not. Removing the remaining restrictions to FDI into service sectors in Jordan is expected to generate employment growth among domestic firms. The type of FDI also matters for jobs spillovers. FDI into services creates jobs among domestic firms in other service sectors. Domestic firms providing services to as well as using services from foreign firms experience significantly higher subsequent medium term growth (columns 8–10 in table E.1 in appendix E). In contrast, FDI into manu- facturing does not lead to growth among domestic firms in upstream or downstream sectors. The positive growth effect of services FDI on down- stream firms in Jordan using these services is consistent with the theory of weak links. In fact, figure B2.1.2 shows that FDI in some service sectors such as transport or professional service is restricted in Jordan. Our find- ings suggest that removing these restrictions would increase growth among domestic firms using these services or providing themselves services to these multinational companies. Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 57 The positive growth spillovers from FDI into service sectors to domestic service providers and young suppliers are permanent, lasting even after foreign firms exit. In contrast, the crowding-out effect of domestic firms operating in the same sector is only temporary, as domes- tic firm growth picks up again after the exit of the foreign competitor. Note that the time periods provided in the data allow for a clear empirical identification to test for asymmetric effects of foreign entry (FDI) versus foreign exit (sudden stops). That is, FDI to developing countries declined substantially when many foreign firms exited in 2009 and 2010, when multinationals adjusted their portfolios to reduce exposure to high-risk investments after the global financial crisis. The data show that the aver- age weighted share of foreign establishments, relative to all establish- ments, declined from 2.3 percent in 2006 to 1 percent in 2011. (The number of foreign-owned establishments declined from 338 to 142.) If initial employment spillovers from FDI before 2006 are truly technology spillovers, the growth effect for domestic suppliers is expected to endure. In contrast, if it is due to a temporary demand effect, job growth among domestic suppliers should disappear after the exit of the foreign firm. The results show the positive backward spillovers from foreign firms to domestic suppliers endure even after the exit of foreign firms. In contrast, the initial decline in employment among domestic competitors in the same sectors after foreign entry is reversed after the exit of the foreign firm (crowding in). The findings suggest that job creation among domes- tic suppliers is due to permanent technology spillovers and not tempo- rary demand effects. Furthermore, after the domestic supplier is able to supply goods or services to the foreign firms, the firm is well positioned to supply its services also to other firms afterwards, in Jordan or abroad. Attracting FDI can be a powerful tool to enhance private sector compe- tition and growth. The results show that FDI benefits primarily the type of domestic firms that have been identified in chapter 1 to drive job growth. In Jordan, FDI led to permanent growth spillovers to young firms supply- ing to foreign-owned firms. In line with previous contributions, these spill- overs emerge from vertical rather than horizontal foreign presence. While FDI spurs employment growth among young and service firms, it tempo- rarily crowds out employment growth in small or old domestic competi- tors in the same sector. The absence of positive spillovers to domestic suppliers in Jordan’s manufacturing sector, however, raises questions. The absence of linkages with domestic manufacturing suppliers ratio- nalizes an evaluation of targeted policy interventions in other developing countries. The results for Jordan contrast with evidence from other developing countries, where findings typically identify spillovers to domestic manufacturing suppliers as the main growth channel of FDI ( Javorcik 2004; Javorcik and Spatareanu 2011; Rodrik 2008; Sutton 2005). 58 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa The lack of spillovers to domestic manufacturing suppliers in Jordan also corroborates the findings of industry case studies. For example, the phar- maceutical sector hosts several large foreign multinationals and large domestic producers. Still, the sector appears to be only weakly linked to domestic suppliers: 90 percent of all chemicals used as inputs in the sector are imported. Only HIKMA Pharmaceuticals, the largest domestic pro- ducer, has a small spin-off supplying chemicals. The main reasons are said to be the high requested quality standards; the small economies of scale relative to East Asian suppliers such as India; and the relatively low trans- portation costs for chemicals. Similarly, other less-sophisticated inputs such as glass containers or packaging material are often imported rather than being supplied domestically. Government policies in Turkey, Malaysia, India, and China actively supported linkages between foreign multinationals and domestic sup- pliers by subsidizing technical training programs. For example, the gov- ernment in Turkey supported producers of domestic car components by promoting joint ventures and providing training programs to bridge the initial technology gap and enabling them to supply to foreign multina- tional automobile enterprises (MNEs) in the country. Once domestic producers satisfied MNEs’ quality standards in Turkey, they also auto- matically obtained the quality accreditation to export to factories of the MNEs in other countries. As a result, Turkey developed a domestic car parts industry supplying intermediate goods ranging from tires to motor parts to foreign MNEs such as Ford, Mercedes, Peugeot, and Isuzu. BOX 2.2 Mobility Restrictions Reduces Competition and Job Growth in the West Bank Competition among firms is distorted by between firms are not physical, but economic restrictions in their access to markets. distances. In the West Bank, economic dis- Market fragmentation increases the local tances can be large because of restrictions market power of firms, shielding them from on the mobility of firms. They distort firms’ potential competitors. Higher transport market access and hence the level playing costs increase market fragmentation and field and firm dynamics associated with job thus the degree of competition between growth. firms operating in these markets. Transport Political constraints on firms’ market costs are often determined by geographical access, in the form of mobility restrictions distances between markets. However, what imposed by Israel in the West Bank, dis- matters for the degree of competition tort the firm dynamics associated with (continued on next page) Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 59 BOX 2.2 Continued job growth. Figure C.1 highlights that the roadblocks, checkpoints, earth mounds, contribution of large domestic private sector trenches, and a separation barrier wall. firms to total employment in the West Bank In this section, we evaluate the extent to & Gaza is marginal (about 5 percent) even which these restrictions in access to markets by regional standards. Establishments in the in the West Bank shaped firms’ playing field West Bank also have low survival probabili- and thus their dynamics. The analysis is ties and low growth: the probability that based on an index of mobility restrictions micro establishments in 2007 grow beyond measuring the effective physical constraints 10 employees in 2012 is only 6 percent. faced by firms in accessing customers, sup- These stagnant firm dynamics are deter- pliers, and so forth. The index compares the mined by firms’ playing field, which is not population that can be reached within a spe- only shaped by domestic policies, but also by cific amount of time in a world with and mobility restrictions on firms’ access to cus- without the mobility restriction. Map B2.2.1 tomers, suppliers, and so forth. Mobility reveals that restrictions in market access for restrictions were installed in the West Bank firms in the West Bank declined between as part of the broader ‘closure’ regime, ini- 2006 and 2012. This section in based on the tially instituted by Israel in response to the analysis of Blankespoor, van der Weide, and first Palestinian uprising. They include Rijkers (2014) in World Bank (2014b). MAP B2.2.1 Mobility Restrictions in the West Bank, 2006 and 2011 a. 2006 b. 2011 mi200601 mi201201 0–0.200 0–0.200 0.201–0.400 0.201–0.400 0.401–0.600 0.401–0.600 0.601–0.800 0.601–0.800 0.801–1.000 0.801–1.000 Mobility index calculated with the negative potential accessibility index with a = 40 and b =2 Mobility index calculated with the negative potential accessibility index with a = 40 and b =2 (continued on next page) 60 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa BOX 2.2 Continued Mobility restrictions reduce net firm also associated with lower firm productivity entry. The relatively high gross entry and growth of incumbents. exit rates in West Bank and Gaza might be Mobility restrictions tilt the level play- related to changes in restrictions to market ing field reducing employment growth in access over time, leading to a more frequent the affected local economic centers. reshuffling of economic activity; for exam- Figure B2.2.1 illustrates the impact of ple, the closure and reopening of establish- higher mobility restriction on job growth in ments in different locations. Figure B2.2.1 the affected locations within the West plots the variations in entry and exit rates Bank. It shows that job growth declines between 2007 and 2012 among different with an intensification in mobility restric- subregions within the West Bank against tions providing evidence that distortions in the average mobility restriction index for firms’ exposure to competition (i.e., mar- these locations over the same time period. It kets access) reduces job growth. shows that gross entry and exit rates tend to The weaker firm dynamics resulting be higher in locations that suffer from from distortions to market access also trans- greater constraints to market access. late into lower output growth. Local eco- Figure B2.2.1 also reveals that the net effect nomic activity is measured by night time of these constraints on firm entry is nega- lights (Henderson et al. 2012) for the West tive. Lower net entry rates, in turn, translate Bank in total as well as among four major into lower competition from entry, hence economic centers affected by the restric- reducing incumbents firms’ incentives to tions. The strong decline in mobility restric- increase their efficiency. The lower net tions around 2009 coincides with an increase entry resulting from mobility restrictions is in local output. FIGURE B2.2.1 Mobility Restrictions Reduce Net Entry, Employment Growth, and Local Output Growth a. Entry and exit b. Employment growth vs. mobility restrictions, 2007–12 Change in log locality employment 0.6 0.4 0.4 0.35 0.3 Rate 0.2 0.25 0 0.2 –0.2 0.15 0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1 Average MRI Average MRI Entry Exit Net entry 95% CI Fitted Source: Blankespoor, van der Weide, and Rijkers (2014) in World Bank (2014b). Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 61 In Malaysia, the government subsidized training programs of foreign MNEs to domestic suppliers, while China and India used domestic con- tent requirements for foreign MNEs in the electronics and automobile sectors (Rodrik 2004, 2008; Sutton 2005). Technical support programs targeting potential domestic suppliers to foreign firms have shown some success in Jordan. Jordan implemented a technical support program operated by JEDCO targeting potential domestic suppliers to foreign firms in the mid-2000s. The program gen- erated some success stories despite small-scale funding, but was later abandoned. In one example, the program provided technical support for a local packaging firm so it could supply Kentucky Fried Chicken (KFC) after KFC entered the Jordan market. A few years later KFC began using this firm as its main supplier of packaging material for all stores in the Middle East region. Business Regulations Limit Employment Growth among Young Firms in Morocco We show how various dimensions of the business environment in Morocco impact employment growth and disproportionately affect young firms. The findings indicate that more competition, equivalent treatment by tax author- ities, less corruption and fewer obstacles in the judicial system, and lower cost of finance would raise employment growth among young firms. This section evaluates the extent to which red tape in the regulatory environment distorts manufacturing job growth in Morocco. Since the early 1990s Morocco has undertaken a range of macroeconomic, regula- tory, and social reforms to improve the functioning of the market econ- omy. Yet, GDP growth over the last decade was accompanied by stagnation in job creation. Figures 1.17, 1.18, 1.19, or 1.23 asserted that the firm dynamics driving job growth are limited in the Moroccan man- ufacturing sector. In this section, we relate these dynamics to cumbersome business regulations that distort private sector competition in Morocco. We empirically investigate how certain firm characteristics interact with constraints in the regulatory environment, finance, and competition, thus inhibiting job creation among Morocco’s manufacturing firms. Detailed data from the analysis presented here are found in the compan- ion paper by Gasiorek, Bottini, and Lai Tong (2014).4 Our approach allows for testing this hypothesis: do young firms with high growth potential suffer more than other firms in a less competitive business environment? We combine manufacturing census data with firm-level information from the World Bank Enterprise Surveys in 62 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Morocco.5 A unique feature of this version of the WBES is that it con- tains the same unique firm identifiers as the census. This allows us to use much more detailed firm-level information on job growth and reg- ulatory policy variables (for competition and finance) by combining the census and the WBES data at the firm level. Cleaning the data set results in an unbalanced panel containing 35,534 observations covering 6,119 firms over nine years (1997–2006). The list of regulatory policy, finance, and competition variables is reported in detail in table F.1. Startups and small firms create more jobs if they face more domestic competition. We measure three different components of competition: number of domestic competitors, unfair informal sector competition, and the extent of foreign competition. The informal sector accounts for a significant share of manufacturing firms. Since firms operating in the black market are not subject to government control or taxation, they could create a degree of “unfair competition” for other firms, which could negatively impact firm dynamics. We find that higher domestic competition (more competitors) is positively correlated with employ- ment growth for startups (those less than four years old), and small firms (those with less than 15 employees). The correlation between employ- ment growth and domestic competition is statistically zero for all other types of firms (larger and older ones). Similarly, startups and small firms create more jobs if they report higher domestic competition from the informal sector; medium-age and large firms tend to create less jobs when in competition with informal firms. Firms have lower employ- ment growth when they report higher foreign competition. This effect is particularly strong among old and large domestic firms. Startups grow faster when they face more transparent and predictable tax authorities. “Equivalent Fiscal Treatment” measures the percentage of firms stating their view as to whether all firms in their sector face equivalent treatment by authorities. Hence, it indicates that the firm perceives a more transparent and predictable fiscal regime in its sector and subregion. After their startup phase, younger firms create fewer jobs when they report more corruption in their industries, or face greater obstacles in their district’s judiciary. The judiciary indicators are dummy variables reflecting firms’ responses to whether the judicial system and dispute resolution dynamics constitute an obstacle to growth, respectively. These variables are aggregated to the sector level so that they reflect the share of firms considering the judiciary as a barrier. We find that large firms and startups have higher employment growth when they operate in sectors and locations with stronger constraints from the judiciary. This result could be reflecting the privileged position of some large firms resulting from their superior access to legal services; the positive correla- tion with startup employment growth might be due to self-selection Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 63 because politically connected firms that can circumvent judiciary con- straints through personal contacts enter districts and sectors where the judiciary is a constraint. In contrast, after their startup period, young firms (older than 4 years of age but younger than 10 years) have signifi- cantly lower employment growth when they face judiciary obstacles in courts or dispute resolution processes in their district and/or sector. In addition, we find that after their startup period, small and young firms have lower growth when they report more corruption in their industries. In contrast, large firms grow faster in these sectors, potentially pointing to the privileged positions of some large firms. Younger firms that operate in sectors or locations with a higher administrative burden have lower employment growth. We consider the following variables, which all reflect red tape in procedures of starting and conducting business: (a) number of days needed to obtain a construction permit (wait permit); (b) number of permits needed each year to continue to operate (administrative constraints); and (c) total number of permits required to create a new firm. We find that longer waiting periods for construction permits have a negative impact on employment growth for all types of firms. Young firms, after their startup period, and old firms both have lower employment growth when their sector (and subregion) has more barriers to entry—when a larger number of permits are required to start a business. Startup firms have higher job growth when they oper- ate in sectors with higher entry barriers, a finding which suggests that only the most promising potential entrepreneurs enter these sector or locations. In contrast, startups have lower job growth when they face higher administrative burdens in conducting their business. High administrative burden is frequently raised as a major constraint to firm growth by entrepreneurs across the region. The experience of a hotel manager in the capital city of the region exemplifies these con- straints. The hotel, a small operation of 40 rooms, with excellent quality service, had no restaurant. Here is the owner’s account of trying to set one up (World Bank 2009): To attract more clients, especially foreign visitors, I really needed a restaurant. The problem is that according to our laws, I needed a separate license for the restaurant. The hotel one was not enough. I eventually got it. I invested $200,000 in furniture and equipment. When I was ready to start, the whole venture collapsed: a representative of one of the four government agencies regulating the tourism industry visited the hotel, claimed that the license for the restaurant was not enough, and requested a large bribe for another license. I refused and decided to go out of the hotel business. I am now leasing my property on a long-term contract—a line of business that is regulated by only one government agency. 64 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Apart from startups, all types of firms that report higher costs of finance create fewer jobs. The results show that a high cost of external bor- rowing reduces the growth of firms of all sizes. Only startups grow when cost of external borrowing is higher, suggesting that they rely on other sources of finance (self-financing or informal sources). Moreover, older and large firms create fewer jobs when they report that access to bank finance is a growth constraint. These findings are consistent with Augier et al. (2012), who show that limited access to external finance reduces productivity growth among larger and older manufacturing firms in Morocco. Overall, the results suggest that cumbersome business regulations in Morocco constrain primarily the type of firms that have been identified in chapter 1 to drive job growth; i.e., startups and young firms. The anal- ysis indicates that greater administrative burdens, less transparent and predictable tax authorities, more obstacles in the judicial system, and higher corruption levels and less domestic competition reduce the growth opportunities for younger and, to a lesser extent, smaller manufacturing firms in Morocco. Morocco’s period of jobless growth over the past decade appears related to the growth constraints faced by young firms, which we identified as having a higher growth potential in chapter 1. Energy Subsidies in the Arab Republic of Egypt Discourage Growth in Labor-Intensive Industries Energy subsidies targeted to heavy industry in Egypt are large; in 2010, subsidies to energy-intensive sectors accounted for 2.9 percent of GDP, or US$7.4 billion (equal to nearly half of total public investments in 2010). A government license is required to legally operate in energy-intensive industries, such as steel and cement, thereby limiting the prospect for free-entry and competition. Moreover, energy subsidies affect the price of labor relative to capital, thereby dis-incentiv- izing more labor-intensive activities, and drifting the economy away from its core areas of comparative advantage. Implementation of the subsidies reduced the prospect for free-entry and competition. A few large and old firms disproportionally benefit- ted from the energy subsidies. Entry into energy-intensive industries typically requires large upfront fixed investments, which in turn demand access to land and credit. In addition, a government license is required to legally operate in energy-intensive heavy industries, such as steel and cement, thereby limiting free entry and competition. This license previously was issued by either the Ministry of Industry Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 65 and Trade or the Ministry of Investment and had to be renewed annu- ally, which meant that some firms could be excluded from the energy subsidies. Figure 2.2 illustrates the distribution of employment classi- fied by firm size or age and the intensity of industries’ consumption of energy.6 Note that this sample covers all establishments in the 2006 census. Large establishments accounted for half of the employment in high energy-intensive industries. In contrast, large establishments accounted for only about 24 and 23 percent of employment in moder- ate and low energy-intensive industries, respectively. In contrast, employment in these industries is concentrated in small establishments which employ 57 and 63 percent of all workers in moderate and low energy-intensive industries, respectively. The difference in the employment distribution across energy intensive industries is even more striking when we distinguish establishments by their age. That is, old establishments accounted for 73 percent of the employment in high energy-intensive industries while young establishments accounted for only 27 percent. The implied higher cost of labor—relative to capital—also helps to explain why old and large establishments failed to contribute significantly to job creation. FIGURE 2.2 Distribution of Employment, by Energy Intensity and Size and Age a. Energy intensity and size b. Age 70 80 60 70 60 50 50 40 Percent Percent 40 30 30 20 20 10 10 0 0 High-energy Moderate-energy Low-energy High-energy Moderate-energy Low-energy intensive intensive intensive intensity intensity intensity Large firms Medium firms Small firms Old firms Young firms Source: Calculations based on establishment census. Note: Large: at least 200 employees, medium: at least 10 but less than 200, small: less than 10. Young establishments are less than 10 years in operation and old establishments at least 10 years. 66 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa These distortions come at a significant cost to labor; the industrial sector in Egypt generates 1.4 million fewer jobs than in Turkey. Turkey serves as a good benchmark, as both countries have a comparable population (74 million in Turkey relative to 81 million in Egypt in 2012) while total GDP (in US$) is about three times lower in Egypt. Moreover, Turkey’s manufacturing sector grew strongly in the past 20 years, benefitting from integration into European value chains. This performance difference between manufacturing sectors in the two countries is reflected in the total number of jobs: the industrial sector in Turkey employed 4.8 million workers in 2012, compared with 3.4 million in Egypt.7 Despite Egypt’s relative comparative advantage in labor, the share of jobs in labor-intensive manufacturing sectors among industrial establishments is lower than in Turkey. Figure 2.3 plots the number of jobs by factor intensity based on the 2006 establishment census for Egypt. The figure shows that approximately 562,000 people work in labor-intensive manufacturing establishments in Egypt relative to about 886,000 in Turkey.8 The lower share in Egypt is striking given that Egypt’s lower stage of development (GDP per capita is about 3.2 times lower than in Turkey) entails a relative comparative advan- tage in labor-intensive sectors such as manufacture of textiles, garments, leather products, footwear, paper products, and publishing and printing. FIGURE 2.3 Employment Share, by Sector Factor Intensity in the Arab Republic of Egypt, 2006, and Turkey, 2010 1,200,000 1,000,000 800,000 Number employed 600,000 400,000 200,000 0 Egypt, Arab Rep., 2006 Turkey, 2010 Capital intensive Labor intensive Natural resources Source: Calculation based on establishment census data; Hussain and Schiffbauer (2014). Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 67 BOX 2.3 Misallocation of Capital in the Arab Republic of Egypt Larger industrial establishments in Egypt extent that average and marginal products are more capital-intensive but less produc- of labor and capital move together (Hsieh tive. Small firms in MENA have less access and Olken 2014). Figure B2.3.1 shows that to credit (World Bank 2011). Thus, there is this is the case in Egypt. Larger establish- good reason to expect that small firms are ments in manufacturing and mining more growth constrained than large firms have higher labor productivity and higher because they cannot finance all profitable capital intensities, but lower value added investment projects. Similarly, small firms per capital. Larger establishments also have might have less access to land, industrial lower total factor productivity (TFP) which zones, or subsidies, also suggesting that is the preferred productivity measure, as it they face higher marginal costs of capital controls for differences in capital intensi- than large firms. As discussed in Chapter 1, ties across establishments.a Thus, the sig- if small firms face higher growth constraints nificantly higher capital-to-labor ratios’ (higher marginal costs of labor or capital), of large firms over-compensate for their they should have higher average levels of lower TFP and helps push their labor value added per worker and capital, to the productivity. FIGURE B2.3.1 Productivity, in Manufacturing and Mining, by Size 11 3.0 11 2.5 2.0 10 1.5 10 1.0 9 0.5 9 0 8 –0.5 10–19 20–49 50–99 100–199 200–999 ≥1,000 Log TFP (right axis) Log value added/worker (right axis) Log capital per worker (left axis) Log value added/capital (left axis) Source: Calculation based on industrial census. (continued on next page) 68 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa BOX 2.3 Continued The results suggest that smaller firms in of resource misallocation across firm size is Egypt are capital constrained; in other striking since large establishments are typi- words, capital in the industrial sector is mis- cally found to be more productive in other allocated towards a few large old firms. In an countries, potentially reflecting past conver- efficient economy, competitive forces lead to gence because more productive firms grow a reallocation of resources to more produc- before marginal productivities equate. For tive firms, equating marginal productivities example, Hsieh and Olken (2014) argue that across different categories of firms over large, rather than small firms, are potentially time. Thus, reallocating capital from large to growth constrained, based on manufactur- smaller industrial establishments would raise ing census data in India, Indonesia, and aggregate productivity in Egypt. This type Mexico. Note: a. Higher labor productivity accompanied by lower TFP implies higher capital intensity, at least for conventional production functions. For example, in the case of a Cobb-Douglas production function, log labor productivity is the weighted sum of log TFP and log capital ( ) y intensity; i.e.,: log L L ) , where Y is output, L labor, K capital, and α the share of labor in output. = log(TFP) + (1− α )log( K Discriminatory Policy Implementation Deters a Level Playing Field in MENA Firms in MENA identify policy uncertainty as a “severe” or “major” obstacle to growth. We show that firms’ complaints about “policy uncertainty” reflect largely a perception of “policy implementation uncertainty” resulting from discrimina- tory policy implementation. The variations in policy implementation observed in the data are substantial, and firms spend a significant amount of time and effort to influence policy implementation. Moreover, the analysis indicates that policy implementation uncertainty reduces competition and innovation in a number of MENA countries, suggesting its potential negative impact on productivity growth and private sector dynamism, especially the entry and growth of new firms. Policy distortions in MENA are not limited to laws, but can also material- ize in the uneven implementation of rules and regulations across firms.9 Despite wide gaps in some countries and areas, macroeconomic and trade policy indices for most MENA countries are approximately on par with other fast-growing countries in East Asia and Eastern Europe. World Bank (2009) shows that gaps in macroeconomic and trade policy indicators are too small to explain the differences in performance between MENA and fast-growing countries in other regions. It concludes that, apart from a few exceptions, the region’s rank is as “average” as that of China, Malaysia, Poland, Thailand, and Turkey. Moreover, apart from Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 69 a few very restrictive countries (Iraq, the Islamic Republic of Iran, Djibouti, and to a lesser degree, the Syrian Arab Republic and West Bank and Gaza), the Doing Business indicators10 suggest that the legal business environment in most MENA countries is comparable to those in fast-growing dynamic emerging economies in other regions, especially if one abstracts from the restrictions in access to finance and judiciary contract enforcement.11 In MENA, an overwhelming majority of firms surveyed identify pol- icy uncertainty as a “severe” or “major” obstacle to firm growth. Over 50 percent of surveyed firms regard economic and regulatory policy uncertainty as an obstacle to their firms’ growth, and almost 35 percent regard it as a “severe” or “major” obstacle (figure 2.4). Though there is some variation across countries, regulatory policy uncertainty remains one of the biggest obstacles to growth in MENA, along with competition from the informal sector, access to finance, and macroeconomic uncer- tainty. For example, the biggest obstacle to growth in Egypt was competi- tion from the informal sector (over 19 percent of firms surveyed), with macroeconomic uncertainty and regulatory policy uncertainty close sec- onds (13.5 percent and 12.5 percent, respectively). While most other FIGURE 2.4 Regulatory Policy Implementation Uncertainty in MENA Yemen, 2010 Syrian Arab Republic, 2009 Libya, 2009 Egypt, Arab Rep., 2008 Egypt, Arab Rep., 2007 Morocco, 2007 Tunisia, 2013 Lebanon, 2006 Lebanon, 2009 Jordan, 2011 Jordan, 2006 Algeria, 2007 Iraq, 2011 0 10 20 30 40 50 60 70 80 90 100 Firms (%) Major/severe obstacle Moderate Minor Source: Calculations using Enterprise Surveys in 2006–13. 70 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa obstacles, such as infrastructure, macroeconomic uncertainty, or access to finance, are linked to the literature on economic growth, the interpreta- tion of what “regulatory uncertainty” is and how it impacts firm growth, is less obvious. In the following, we present evidence that MENA firms’ aversion to regulatory uncertainty is predominantly about discriminatory policy implementation that benefits selected firms with specific charac- teristics and thus hinders a level playing field among all firms. The variation in the outcomes of policy implementation across firms in MENA countries is considerable. One approach to understanding regulatory uncertainty is to look at firm-level variability in MENA. We examine the evidence provided by surveys of 8,120 firms in eleven coun- tries in MENA.12 The Enterprise surveys carried out by the World Bank cover qualitative and quantitative assessments by firms concerning numerous obstacles to their growth. The analysis looks at evidence of variation in policy implementation outcomes reported by firms, such as the time they had to wait to obtain an operating license, get a construction permit, or clear goods through customs. That is, table 2.1 summarizes the TABLE 2.1 Averages and Dispersion of Firms’ Waiting Days for Regulatory Services Average number of days Coefficient of variation Clear Clear Clear Clear Operating customs customs Import Construction Operating customs customs Import Construction Country name license imports exports license permit license imports exports license permit Jordan — — 2 — — — — 1.43 — — Egypt, Arab Rep. — 9 7 33 200 1.21 1.55 2.26 2.37 Egypt, Arab Rep. 282 9 6 59 346 2.89 1.22 1.05 1.47 1.93 Yemen, Rep. 13 — 8 24 48 2.84 — 1.43 2.51 1.54 Jordan 10 9 4 5 37 2.59 1.50 0.94 2.14 1.75 West Bank and Gaza 30 22 6 24 50 2.46 1.40 1.25 1.15 0.72 Tunisia 19 9 5 19 158 2.20 1.46 1.36 1.55 2.65 Morocco 4 4 2 2 61 1.87 1.46 1.32 0.61 1.72 Algeria 19 17 14 33 112 1.39 0.91 0.91 1.04 1.46 Lebanon 151 10 7 109 150 1.28 1.40 1.39 1.33 0.92 Syrian Arab Republic 184 10 5 39 245 1.26 1.13 1.13 1.81 1.08 Libya 50 13 6 — 90 1.22 0.87 0.22 — 1.24 Lebanon 81 7 7 30 218 0.87 1.27 1.32 0.52 1.53 Iraq 30 21 11 21 36 0.72 0.89 0.41 0.98 0.56 Turkey 37 10 — 21 42 2.88 1.34 — 1.67 1.65 Chile 84 17 — 17 143 2.62 1.39 — 1.59 1.94 Croatia 26 2 — 12 182 1.69 1.25 — 1.27 1.25 Bulgaria 62 3 — 21 94 1.59 1.1 — 1.17 1.04 Indonesia 21 3 — 11 32 1.43 1.09 — 0.94 1.93 India 29 14 — 15 28 1.4 1.02 — 1.82 1.33 Brazil 83 15 — 43 139 1.14 1.1 — 1.25 1.31 Source: World Bank Enterprise Surveys, various years between 2006 and 2011. Note: — = cells where not enough data are available because very few firms responded to the question or the information was not collected. Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 71 averages and dispersion of the number of days that firms in MENA coun- tries had to wait for different regulatory services in various years between 2006 and 2013. Although the results suggest that there are some differ- ences across countries in MENA, variations reported within-countries are larger than variations across countries. These large within-country variations are linked to actions undertaken by firms to control policy out- comes, such as spending time with government officials or paying bribes. Qualitative evidence supports the argument that policy implementation uncertainty is a severe constraint to firm growth. In 1998 a large and diver- sified family conglomerate in a country of the region opened the first supermarket of a new chain; in 2005 it had 18 stores across the country. When asked about their business constraints, an executive from this chain replied (World Bank 2009): We would have opened more than 50 stores by now to meet the growing demand if opening branches was not so cumbersome. This is by far our biggest challenge. We have to deal with 11 different authorities at the local level to get approval. Typically, you get only a temporary approval that allows you to start operating, but final approval may be delayed for months and sometimes for years. Their temporary licenses must be renewed every six months. Many of the laws we are subject to date back to more than five decades, when there were no supermarkets, so the actual application is almost entirely discretionary. Firm survey results reveal large variations in government officials’ implementation of regulations across firms in MENA, relative to most other emerging countries. The average waiting time to obtain an operat- ing license, a construction permit, and an import license is the longest in Egypt, Syria, and Lebanon. The firm survey results are consistent with the Doing Business indicators confirming that legal business regulations are, on average, relatively restrictive in Syria and Egypt, and more com- petitive in Tunisia or Jordan.13 However, table 2.1 shows that there are large variations in the implementation of regulations across firms in all MENA countries: the coefficient of variation in waiting times for differ- ent regulatory services is typically higher in MENA countries, especially in Egypt, Jordan, Tunisia, the Republic of Yemen, and Morocco, than in emerging economies from other regions.14 The large variations in policy implementation persist among firms operating in the same sector. Figure 2.5 shows the 90th, 50th, and 10th percentiles of the distribution of the number of days to get an operating license, a construction permit, or to clear customs from the most recent Enterprise Survey data for each MENA country. Again, the findings con- firm that the variation across firms within a country is, in many instances, 72 Number of days to obtain a construction permit Ye Number of days to obtain an operating license FIGURE 2.5 Ye me Number of days to clear customs Ye me n, me n, We n, 0 10 20 30 40 50 We 0 200 400 600 800 0 100 200 300 400 st B Re We Re Re p., st B p., st B p., an 20 an 20 an 20 ka 10 ka 10 ka 10 nd nd nd MENA Countries G aza Ga Ga ,2 za, za, 00 20 20 Sy Tu 6 Syr Tu 06 Tu 06 ria nis ian nis nis nA ia ,2 Ara ia, ia ,2 rab 01 20 01 Re pu 3 bR ep 13 3 bli ub c, 2 lic, 00 20 Mo 9 Mo 09 Mo roc roc roc co, co, co, 20 20 20 07 07 07 Lib Lib Lib 10th percentile ya ya, ya, ,2 20 20 00 Le ba 9 Leb 09 Leb 09 no an an n, on on 20 ,2 ,2 06 00 00 Le Leb 6 Leb 6 Source: Calculations using various World Bank Enterprise Surveys 2006–13. ba an an no on on n, ,2 ,2 50th percentile 20 00 00 09 9 9 Jor Jor Jor c. Clearing customs a. Operating license da da da b. Construction permit n, n, n, 20 20 20 11 11 11 Jor Jor Jor da da da n, n, n, 20 20 20 06 06 06 90th percentile Eg Ira Eg Ira Eg Ira yp q, yp q, yp q, t, A 20 t, A 20 t, A 20 rab 11 rab 11 rab 11 Eg R e Re Re yp p . , Eg yp p., p., t, A 20 t, A 20 20 rab 08 rab 08 08 Re Re p., p., 20 20 07 07 Alg Alg Alg eri eri eri a, 20 a, 2 a ,2 07 00 00 7 7 Variability in Days to Accomplish Various Regulatory Tasks across Firms, Selected Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 73 larger than the differences across countries. Among the eleven MENA countries, the country with the lowest median time to obtain an operating license is Jordan—one day. Some others report low median times—the median time in Tunisia and the Republic of Yemen is seven days. While sector specific characteristics might explain part of the variations, the large variations in policy implementation persist among firms operating in the same sector. For example, in the textile and garment sector in Jordan, 10 percent of firms waited only five days to obtain a construction permit, while the next 80 percent waited between 5 and 120 days. In other manufacturing, the median firm waited seven days for imports to clear customs, while the next 40 percent of firms waited between 7 and 21 days. At least part of the variations in policy implementation across firms within the same sector appears to reflect firm-level variations in deals. That is, the variations reflect differences in the way firms are treated based on their characteristics such as ownership (e.g., politically versus non–politically connected firms—see chapter 4). The outcomes of such deals are condi- tional on firms’ activities to influence public officials implementing the policies so that some of the firm-level variations could reflect firms’ uncer- tainty about the influence function; i.e., some obtained a good deal and some a bad deal (Hallward-Driemeier, Khun-Jush, and Pritchett 2010). The perceptions of firms that the implementation of policies is consis- tent and predictable vary by firm size and location. Table 2.2 shows the share of firms that disagree with the statement that implementation is consistent and predictable. In Egypt and Jordan there is a large difference in policy implementation perceptions between firms located in the capital TABLE 2.2 Share of Firms That Disagree with the Statement That Implementation of Rules Is “Consistent and Predictable” Egypt, Syrian Arab West Bank Arab Rep., Lebanon, Jordan, Morocco, Tunisia, Algeria, Yemen, Republic, Iraq, and Gaza, 2008 2006 2006 2007 2013 2007 Rep., 2010 2009 2011 2006 Small (5 to 19 employees) 47 42 34 53 23 66 65 45 70 68 Medium (20 to 99 employees) 50 49 50 57 28 52 55 42 62 66 Large (100 or more employees) 45 39 43 67 27 51 67 44 54 55 Food sector 53 49 40 56 25 59 45 44 62 56 Textiles and garments 48 56 51 68 26 62 59 46 86 75 Chemicals 41 44 67 63 18 48 88 35 54 67 Other manufacturing 47 39 39 60 32 53 59 46 70 68 Services na 43 33 51 25 na 67 43 67 68 Construction and transport na na 44 59 na 68 50 27 69 64 Capital city (or major city) 51 41 62 54 21 60 58 59 34 60 Outside capital city 46 50 24 73 30 57 63 63 67 74 Source: Calculations using various World Bank Enterprise Surveys, 2006–13. 74 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa city areas of Cairo and Amman and firms in the periphery. In the greater Amman area 62 percent of firms report inconsistent and unpredictable policy implementations, as compared with only 24 percent in peripheral areas. In the greater Cairo area 51 percent of firms report inconsistent and unpredictable policy implementations, compared with 46 percent in peripheral regions. These regional differences in policy implementation perceptions are not observable in the other MENA countries, where on average policy implementation perceptions show opposite perceptions, except in the case of Algeria. Geographical differences partly reflect dif- ferent attitudes or access to the government (municipal administrations). In many cases, most of the relevant business regulatory administrations are located within the capital city, reflecting an ease of access for firms in that area. A possible explanation for Jordan is that firms outside of Amman face lower competition, implying that higher costs resulting from varia- tions in regulatory services are less important. Firms outside of Amman are on average smaller, less likely to export, and operate more often in the service sector. The same may be inferred for Egypt. SMEs are more likely to complain about unpredictability of policy actions in Algeria, Egypt, Iraq, Lebanon, and West Bank and Gaza. Firms in MENA take costly actions to influence the outcomes of policy implementation by government officials. The degree to which firms take actions to influence policy implementation varies across firms within countries. Table 2.3 shows the time firms’ senior manage- ment spent, on average, with government officials. It reveals that firms TABLE 2.3 The Extent to Which Firms Take Action to Influence Policy Implementation across Types of Firms in MENA Percentage of management time spent dealing with officials (average) Egypt, Syrian Arab West Bank Jordan, Morocco, Lebanon, Arab Rep., Tunisia, Algeria, Yemen, Republic, Iraq, and Gaza, Country 2006 2007 2006 2008 2013 2007 Rep., 2010 2009 2011 2006 Total 8.4 13.2 10.5 11.0 24.8 25.1 17.4 13.2 6.2 7.1 Firms located in capital city 12.7 7.7 13.4 8.4 36.0 21.2 21.5 13.2 5.0 6.4 Firms not located in capital city 4.3 15.6 12.1 11.7 17.7 27.5 16.0 13.2 6.5 7.9 Small (5 to 19 employees) 7.0 10.6 11.6 8.5 27.2 22.9 14.7 14.6 4.2 6.6 Medium (20 to 99 employees) 10.6 14.0 13.3 13.1 24.8 27.7 19.8 14.1 10.2 7.7 Large (100 or more employees) 6.8 13.3 13.3 10.9 22.1 27.9 24.3 9.7 8.3 8.5 Food sector 10.1 11.9 13.3 11.6 21.0 26.9 19.1 15.0 6.8 7.8 Textiles and garments 7.1 9.0 12.8 12.7 22.1 24.7 14.2 8.8 1.6 6.2 Chemicals 7.9 12.2 11.1 11.0 20.5 23.8 37.5 11.8 4.4 8.4 Other manufacturing 9.9 14.6 17.5 10.2 25.9 28.4 15.1 12.0 4.1 6.6 Services 5.5 16.2 12.0 na 29.1 21.2 22.4 16.8 7.7 7.4 Construction and Transport 12.2 21.3 na na na 26.3 20.4 16.5 10.3 8.3 Source: Calculations using World Bank Enterprise Surveys in 2006–13. Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 75 are more likely to spend time with government officials where regula- tory implementation is uncertain. For example, in the Republic of Yemen senior managers spent on average 35 percent of their time influencing policy; the effort was especially high among firms in the chemicals sector. Moreover, large firms tend to spend more time influ- encing policy. In Tunisia, firms in the services sector, which includes tourism and hotels, are more likely to spend time dealing with govern- ment officials. In Morocco and Jordan, senior management in the con- struction and transport sectors spent more time dealing with government officials. Firms are more likely to take costly actions when faced with higher policy implementation uncertainty. The variations across firms’ access to regulatory services may stem either from a subset of firms that have privi- leged access, such as large politically connected firms, or from differences in the performance of officials implementing policy. Regression analysis shows, however, that there is a systematic correlation between policy implementation uncertainty and firms’ actions to influence the imple- mentation suggesting that differences in the enforcement by government officials is not random but discriminatory. In particular, we group firms by their location, sector, and size. The coefficient of variation is com- puted for the perceived consistency of policy implementation across firms in each group. We refer to this explanatory variable as policy implemen- tation uncertainty. The average management time firms spend interact- ing with, for example, lobbying with government officials, is also computed for each group (dependent variable); it is used as a proxy for quantifying firms’ actions to influence the outcome of policy implemen- tation. The results shown in table 2.4 indicate that more management time is spent dealing with officials when firm groups face greater policy implementation uncertainty. TABLE 2.4 Higher Policy Implementation Uncertainty Induces Senior Managers to Spend More Time with Government Officials Average management time Dependent variable: (in %) spent dealing with officials Coefficient of variation of firm reporting implementation as 0.234** consistent and predictable (2.13) R2 0.331 Number of location-sector-size firm groups 55 Source: World Bank (2012). Note: Each entry reflects the results of a regression including sector dummies and heteroscedasticity robust standard errors; t-values are presented in parenthesis. The correlation coefficient between management time and reported bribes is 0.305 (some countries are excluded). Significance level: * = 10% and ** = 5%. 76 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Policy implementation uncertainty is associated with lower competi- tion, innovation and firm growth in Jordan and Egypt. The WBES for Egypt and Jordan demonstrate that variations in implementing legislation among firms, rather than the legislation itself, distinguishes these coun- tries from fast-growing emerging economies in other regions. Moreover, we contend that discriminatory policy implementation reduces economic dynamism—competition and innovation—as well as firm growth. Table 2.5 reports empirical findings for Jordan and Egypt. It suggests that policy implementation uncertainty reduces perceived pressure from domestic competition.15 In contrast, it is not correlated with pressure from TABLE 2.5 Policy Implementation Uncertainty Reduces Innovation and Firm Growth in Jordan and the Arab Republic of Egypt Jordan Pressure from Pressure from domestic competition foreign competition Employment growth, to reduce cost to reduce cost 2003–06 Probability to innovate Dependent variables: (1) (2) (3) (4) (5) (6) (7) Coefficient of variation of firms reporting −1.01* 0.822 −.647** −.807 implementation as consistent and (−1.76) (0.75) (−2.24) (−.97) predictable Difference 75–25 percentile firms reporting −.155** −.072** −.206* implementation as consistent and (−2.04) (−1.95) (−1.85) predictable R-squared 0.120 0.122 0.336 0.199 0.197 0.238 0.241 Number of firms 467 467 419 436 436 487 487 Egypt, Arab Rep. Pressure from Pressure from domestic competition foreign competition Employment growth, to reduce cost to reduce cost 2006–07 Probability to innovate Dependent variables: (1) (2) (3) (4) (5) (6) (7) Coefficient of variation of firms reporting 0.025 0.060 −.111** −.787** implementation as consistent and (0.36) (1.00) (−2.57) (−2.96) predictable Difference 75–25 percentile firms reporting 0.013 −.002 −.090 implementation as consistent and (0.53) (−0.09) (−1.04) predictable R-squared 0.031 0.031 0.043 0.420 0.418 0.109 0.106 Number of firms 902 902 899 878 878 905 905 Source: World Bank (2012) and World Bank (2013). Note: Results are from World Bank 2012 for Jordan; calculations for the Arab Republic of Egypt. The results for Jordan are based on ES data for 2006, for Egypt on ES data for manufacturing firms in 2007. All regressions include sector dummies (apart from specification (1), heteroscedasticity robust standard errors that are clustered at the group level; t-values are presented in parenthesis. The average and standard deviation are computed over grouped firms in each location-sector-size group. Innovation is a binary variable equal to 1 if the firm introduced a new product or new process, or licensed a foreign technology in the last 3 years and 0 otherwise (roughly half of the firms in sample innovated). Significance level: * = 10%, ** = 5%. Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 77 foreign competition (specification 3). Taken together, this suggests an indirect test against spurious correlation, since variations, or uncertainty, in policy implementation are expected to reduce domestic competition, but should not affect competition from imports. Furthermore, empirical findings for Jordan and Egypt reveal that the greater the disagreement with the statement that government implementation is “consistent and predictable” within a location-sector-size firm group, the lower is employ- ment growth and the probability to innovate for firms in these groups. The results suggest that de facto discriminatory implementation of policies, rather than laws themselves, deter competition, innovation, and employment growth, by granting privileges to selected firms. In other words, we expect to find that firms with certain characteristics, or that undertake certain actions, benefit from streamlined regulatory services and procedures. This can lead to the lack of a level playing field, and under- mine the competitiveness of firms in the region. The data suggest that: a. firms’ characteristics (size, age, or ownership) and actions (bribes and lobbying) systematically influence policy implementation, and b. the resulting uncertainty reduces competition, innovation, and employment creation. Notes 1. The impact of FDI is measured on employment instead of productivity spill- overs as in Javorcik (2004), since no reliable output data for establishments was available. Focus was given to the long-term employment growth effects of the presence of foreign firms in 2006 and subsequent employment growth until 2011. Thus, we assume that over a five-year period learning effects (technology spillovers) of domestic suppliers materialize into job growth. Moreover, in contrast to Javorcik (2004), our data allow measuring spillovers to manufacturing and services firms. 2. See Marotta et al. (2014) for related work on Tunisia. 3. It is important to note that the net welfare effect might still be positive even in the case of complete crowding-out if foreign firms pay higher wages. 4. More details and additional analyses are described in the companion paper by Gasiorek et al. (2014). See also Appendix F for more details on data sources, methodology, and a summary table with the main empirical results. 5. Table F.2 in appendix F summarizes the results for regressions of the aggre- gate net job creation rate on business environment variables classified into (a) regulatory environment, (b) competition, and (c) access to finance. The first column shows the coefficients for these variables without any interac- tion, while the subsequent columns represent the policy-interacted coeffi- cients for different ‘types’ of firms. Access to finance and competition 78 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa variables are observed at the firm level (combining the census and WBES data at the firm level), while the regulatory variables are aggregated to the sector level. 6. The classification of industries in high, medium, and low energy-intensities is based on the UNIDO (2010), “Compilation of Energy Statistics for Economic Analysis,” Development Policy and Strategic Research Branch Working Paper 01/2010. High energy-intensive industries account for 22 percent of all mining and manufacturing four-digit industries, medium energy-intensive industries for 37 percent, and low energy-intensive for 42 percent. 7. The total number of employees working in the industrial sector in the Arab Republic of Egypt is based on the Egypt Labor Market Panel Survey, and in Turkey on the yearly labor force survey from Turkstat. 8. The numbers here differ from the Egypt Labor Market Panel Survey number of industrial employment in 2012 in the previous paragraph as the census was conducted in a different year (2006) and does not cover all informal or part- time workers. 9. This section follows the methodology in Hallward-Driemeier et al. (2010). 10. The Doing Business indicators measure the time and costs of official legal procedures for a representative domestic firm based in the capital or the larg- est business center of the country. The measured policy dimensions are the cost of starting a business, dealing with construction permits, registering property, getting credit, protecting investors, paying taxes, trading across borders, enforcing contracts, closing a business, and getting electricity. 11. MENA countries underperform systematically in two dimensions: access to bank finance and enforcing contracts. “Access to finance” primarily measures laws regarding credit information, collateral, and bankruptcy. “Enforcing con- tracts” measures the number of official procedures, time, and costs to enforce a sale of goods dispute from the moment of filing until actual payment. Hence, it indicates a problem of implementation rather than legislation. 12. The countries are Algeria, Egypt, Iraq, Jordan, Lebanon, Libya, Morocco, the Syrian Arab Republic, Tunisia, the Republic of Yemen, and West Bank and Gaza. 13. Comparing average waiting times across countries should be viewed with caution for some regulatory services. Receiving or renewing an operating license may be associated with mandatory complementary registrations or inspections (safety or health inspections) in many countries, which would bias the mean upwards. However, comparing the dispersion of waiting times across countries does not suffer from this bias, since the coefficient of varia- tion corrects for such level differences across each country. 14. The coefficient of variation, which is defined as the ratio of the standard deviation to the mean, is a normalized measure of dispersion of a probability distribution. The coefficient of variation should be used only for measures which take nonnegative values. It is independent of the unit in which the measure has been taken (in contrast to the standard deviation which can only be understood in the context of the mean of the data). Thus, one should use the coefficient of variation instead of the standard deviation for comparison between data with widely different means. 15. The variable approximating “policy implementation uncertainty” is con- structed as follows. Firms are grouped by their location, sector, and size, for Distorted Dynamics: The Impact of Policies on Firm Dynamics and Job Growth 79 30 groups of firms (each containing at least five firms). For each group, the coefficient of variation of the perceived consistency of policy implementa- tion across firms is computed. The spread between the 75th and 25th per- centiles of the perceived policy implementation consistency is computed as an alternative measure of policy implementation uncertainty. In addition, control variables measuring the initial size, location, age, exporting status, and the initial level of employment of firms are included in the regressions. The results for the control variables are consistent with findings in the lit- erature on firm growth (not shown in the table). The probability of innovat- ing is estimated with a probit regression, whereby the dependent variable is a binary variable equal to one if a firm either introduced a new product or a new process, or licensed a new technology within the last three years, and equal to zero otherwise. About 50 and 42 percent of firms in the sample in Jordan and Egypt, respectively, were innovators. References Aghion, P., C. Harris, P. Howitt, and J. Vickers. 2001. “Competition, Imitation and Growth with Step-by-Step Innovation.” The Review of Economic Studies 68 (3): 467–92. Arnold, Jens, B. Javorcik, M. Lipscomb, and A. Mattoo. 2012. “Services Reform and Manufacturing Performance—Evidence From India.” Policy Research Working Paper 5948, World Bank, Washington, DC. Augier, Patricia, Dovis Marion, and Gasiorek, Michael. 2012. “The Business Environment and Moroccan Firm Productivity.” Economics of Transition Volume 20(2): 369–99. Bartelsman, Eric, J. Haltiwanger, and S. Scarpetta. 2004. “Microeconomic Evidence on Creative Destruction in Industrial and Developing Countries.” Working Paper, World Bank, Washington, DC. Blankespoor, B., R. van der Weide, and B. Rijkers. 2014. “How Valuable is Market Access? Evidence from the West Bank.” Mimeo, World Bank, Washington, DC. Borchert, Ingo, Batshur Gootiiz, and Aaditya Mattoo. 2012. “Guide to the Services Trade Restrictions Database,” World Bank Policy Research Working Paper WPS6108, Washington, DC. Caves, D. W., Christensen, L. R., and Diwert, W. E. 1982. “The Economic Theory of Index Numbers and the Measurement of Input, Output, and Productivity.” Econometrica 50: 1393–414. Davis, Steven J., John C. Haltiwanger, and Scott Schuh. 1996. Job Creation and Destruction. MIT Press. Gasiorek, Michael, Novella Bottini, and Charles Lai Tong. 2014. “Employment Dynamics in Morocco: The Role of the Business Environment and Asymmetrical Policy Treatment.” Mimeo, World Bank, Washington, DC. Hallward-Driemeier, M., G. Khun-Jush, and L. Pritchett. 2010. “Deals versus Rules: Policy Implementation Uncertainty and Why Firms Hate It.” Working Paper 16001, NBER, Cambridge, MA. Henderson, Vernon, Adam Storeygard, and David N. Weil. 2012. “Measuring Economic Growth from Outer Space.” American Economic Review 102(2): 994–1028. 80 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Hsieh, Chang-Tai, and Benjamin Olken. 2014. “The Missing ‘Missing Middle’.” Working Paper 19966, NBER, Cambridge, MA. Hussain, Sahar, and Marc Schiffbauer. 2014. “Struggling for Growth: Labor Demand and Job Creation in Egypt.” Mimeo, World Bank, Washington, DC. Javorcik, Beata. 2004. “Does Foreign Direct Investment Increase the Productivity of Domestic Firms? In Search of Spillovers through Backward Linkages.” American Economic Review 94 (3): 605–27. Javorcik, Beata, and Mariana Spatareanu. 2011. “Does It Matter Where You Come From? Vertical Spillovers from Foreign Direct Investment and the Origin of Investors.” Journal of Development Economics 96 (1): 126–38. Jones, C. 2011: “Intermediate Goods and Weak Links in the Theory of Economic Development.” American Economic Journal: Macroeconomics 3 (2): 1–28. Keller, W. 2004. “International Technology Diffusion.” Journal of Economic Literature 42 (3): 752–82. Kremer, Michael. 1993. “The O-Ring Theory of Economic Development.” The Quarterly Journal of Economics 108 (3): 551–75. Lamla, Michael, and Marc Schiffbauer. 2014. “Employment Spillovers from FDI in Jordan.” Mimeo, World Bank, Washington, DC. Marotta, Daniela, Christian Ugarte, and Leila Baghdadi. 2014. “Weak Links in Tunisia.” Mimeo, World Bank Washington, DC. Olley, G., and A. Pakes. 1996. “The Dynamics of Productivity in the Telecommunications Equipment Industry.” Econometrica 64: 1263–79. Rodrik, Dani. 2004. “Industrial Policy for the Twenty-First Century.” Discussion Paper 4767, Centre For Economic Policy Research, London. ———. 2008. “Normalizing Industrial Policy.” Commission on Growth and Development, Working Paper 3, World Bank, Washington, DC. Sahnoun, Hania. 2014. “Discriminatory policy implementation in the Middle East and North Africa.” Mimeo, World Bank, Washington, DC. Sutton, John. 2005. “The Auto-Component Supply Chain in China and India: A Benchmark Study.” In Annual Bank Conference on Development Economics 2005: Lessons of Experience, edited by Francois Bourguignon and Boris Pleskovic. Washington DC: World Bank. World Bank. 2009. From Privilege to Competition: Unlocking the Private-Led Growth in the Middle East and North Africa. MENA Development Report. Washington, DC: World Bank. ———. 2011. “Financial Access and Stability: A Road Map for the Middle East and North Africa.” Working Paper, World Bank, Washington, DC. ———. 2012. “Jordan Development Policy Review.” Mimeo, World Bank, Washington, DC. ———. 2013. “Jobs for Shared Prosperity: Time for Action in the Middle East and North Africa.” Mimeo, World Bank, Washington, DC. ———. 2014b. “Why Doesn’t MENA Export More? A Firm-Level Perspective.” Mimeo, World Bank, Washington, DC. CHAPTER 3 Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia Efforts to stimulate private sector growth and jobs in MENA have often taken the form of active industrial policies with, however, limited evidence of success and many instances where policies have been captured by only a few firms. This chap- ter reviews these policies over the past decades and compares them with the experi- ences of East Asian countries. Several critical differences in policy design and implementation that underpin the success of industrial policies in East Asian countries compared with MENA countries are highlighted. These differences point to a list of key ingredients for an effective industrial policy: (a) there is consensus on a common strategic vision and objectives at the country level, and a focus on new economic activities where market failures are more likely to have a binding influence on industrial development; (b) policies are connected to perfor- mance and evaluation systems in which both the effectiveness of policies and offi- cials can be assessed; (c) policies promote and safeguard competition and equality of opportunity for all entrepreneurs in the domestic market and provide incen- tives for firms to compete in international markets. Many countries in MENA have taken the route of an active industrial policy in an effort to address the deficiencies in their business environ- ment and stimulate private sector growth, job creation, and structural transformation. The previous chapters have shown that a host of policies across MENA countries undermine the underlying firm-level fundamen- tals of job creation by limiting competition and tilting the playing field. While confronting these constraints directly would have been more effective, many countries in MENA have adopted industrial policies in an effort to encourage private sector growth, job creation, and structural transformation. In MENA, as in many other countries around the world, industrial policy has often included subsidies and tax breaks, which must be large to compensate for the deficiencies in the business environment and spur investment, growth, and job creation. Both MENA countries and many East Asian countries have used this alternative strategy 81 82 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa extensively. In the following, we discuss the extent to which these indus- trial policies have been successful and the factors that explain the differ- ent experiences of the two regions. The MENA region has many decades of experience with industrial policy, but there is limited evidence of success. Few observers argue that the experience has been successful: despite aggressive actions to drive industrial development, structural transformation, and job creation, results have been scarce and low. What should the region’s policy makers con- clude from this experience? Should they, in the future, rely more on mar- ket forces, and less on government direction? Or should they improve on the quality of government interventions? The second option appears par- ticularly attractive to policy makers and analysts who observe the remark- able success of East Asian economies, where the government role has been large and ongoing (Box 3.1).1 The analytical and data challenges in assessing whether a particular constellation of industrial policies triggered growth that otherwise would not have occurred are considerable. Did a sector emerge and prosper because of industrial policy? Despite it? More important, did industrial BOX 3.1 Market Failure and Industrial Policy Government intervention has a role to play into account when deciding whether to enter in structural transformation when market a sector. For example, firms may not know forces are disrupted. A long-standing argu- how costly it is to produce in a new sector, or ment for industrial policy is coordination how profitable an export market is. Their failure: firms may not invest when the profit- investments in discovering these costs yield ability of their own potential investments benefits for all firms that they individually depends on whether other firms make com- do not take into account (Hausmann and plementary investments.a This argument is Rodrik 2003). Thus, market forces are dis- more difficult to sustain when there is an rupted by information asymmetries related active world market in the complementary to the economic returns to investment; products, attenuating the need for within- coordination difficulties among entrepre- country coordination of investments. neurs in complementary industries; and the Harrison and Rodriguez-Clare (2010) sum- absence of markets. marize more recent versions of this argu- Three concerns about industrial policy ment. Learning externalities or knowledge preoccupy observers and analysts. First, can spillovers yield large productivity benefits market failure be reliably identified? A major for all firms, but no single firm takes them difficulty in identifying market failures (continued on next page) Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 83 BOX 3.1 Continued is that they may be difficult to firms are resolved.c Third, do governments disentangle from government policy fail- really want to fix it? Governments may place ures. Arguably, East Asian industrial policies a higher priority on alternative uses of funds have not targeted market failures directly, or pursue other goals that are potentially but appear to have particularly succeeded at incompatible with growth; these include, offsetting, for selected industries, govern- but are not limited to, incentives to extend ment-related obstacles to growth, such as open-ended benefits to supporters. those rooted in governance challenges, red While research is convincing that indus- tape, and political risk.b Second, can indus- trial policy aimed at attenuating the effects trial policy work given the significant techni- of market failure is necessarily selective, a cal and informational demands in crafting crucial point is that it must be selective at and implementing industrial policy? Rodrik the level of industries and sectors, not at the (2008) suggests an active public-private dia- level of firms.d Market failures do not, again logue to overcome information asymme- by definition, afflict some firms in a sector, tries, citing the positive experiences with but not others. Moreover, recent research deliberation councils or private-public ven- by Aghion et al. (2012) indicates that indus- ture funds in East Asia. However, these dia- trial policy can promote productivity growth logues are likely to succeed only to the extent when it favors competition—and that indus- that the obstacles to collective action among trial policy in China has done precisely this. Note: a. See also Murphy, Vishny, and Schleifer (1989). b. Where political risk is high—and in many East Asian countries it was very high—governments cannot easily attract private investment. To increase investment, they can either rely on state-owned enterprises—for which political risk is irrelevant—or offer large subsidies to private entrepreneurs to raise their risk-adjusted rates of return. As the subsequent discussion makes clear, during a period in which private investors confronted substantial political risk, the Republic of Korea embraced both strategies. c. When the problem is the identification of new markets, however, neither side of the dialogue is likely to be especially well-informed; it is precisely because they do not know about potential opportunities that the need for industrial policy might exist. Lin and Monga (2010) point out that a few private entrepreneurs might have already entered new profitable industries. They conclude that these local success stories are themselves informative. Public-private dialogues could bring such examples to light. d. The essentially selective characteristic of industrial policy prompts critics to describe it as “picking winners.” However, market failures are typically related to particular sectors or types of economic activity. Hence, industrial policies intended to correct them are necessarily selective. policy initiatives correct market failures, or did they simply offset, for some firms, policy distortions in other areas, such as cumbersome regu- lations, public infrastructure, financial markets, or the rule of law? Given this problem of missing data, this section follows the alternative strategy to directly compare elements of industrial policy design in MENA and East Asian countries, and in particular the Republic of Korea. It revisits in some detail the Arab Republic of Egypt and Morocco’s industrial pol- icy framework as well as some aspects of industrial policy in Jordan, the Syrian Arab Republic, and Tunisia (“Industrial Policy in MENA Has 84 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Had Limited Success and Many Instances of Policy Capture”) and com- pares it to the experience of East Asia (“What Did Successful Countries Do? The Case of the Republic of Korea” and “Lessons from East Asia Are More Difficult to Implement than Is Commonly Understood”). Industrial Policy in MENA Has Had Limited Success and Many Instances of Policy Capture Industrial Policy in Egypt After independence, the state invested in heavy industry and used its reg- ulatory powers to direct private sector investment into favored sectors. Among MENA countries, we know the most about the industrial policies of Egypt, the largest economy in the region. It has pursued policies meant to encourage particular economic sectors since its independence in 1952. From 1956 to 1970 the state invested in heavy industry, authorized favor- able tax treatment for some private investments, and heavily regulated private sector industrial activity. Confronted with the failure of state-led industrialization, but reluctant to abandon state-owned enterprises, from 1970 to 1981 Egypt focused even more intently on using its regulatory powers to direct private sector investment into favored sectors and to discourage it in others (Loewe 2013). Between 1981 and 1990, the most important adjustment in state-led development was a dramatic expansion in the business interests of the army. Price regulations, customs, and financial sector policies continued to favor state-owned enterprises from 1981 to 1991. However, the deval- uation of the Egyptian pound, incremental deregulation of domestic mar- kets, and some tax breaks for manufacturing brought benefits to the private sector as well. Private investment rose from approximately 16 percent of total investment over the period 1960–82, to 41 percent over the period 1983–90 (Loayza and Honorati 2007). During this latter time, the business interests of the army expanded dramatically into tour- ism, construction, white goods, vehicles, fertilizer, mineral water, olives, and bread, with much of it financed by the sale of government land in Cairo and on the seaside (Loewe 2013). The fiscal crisis forced a change in industrial policies in the 1990s. Egypt shifted somewhat to more favorable conditions for private invest- ment, though not to the point that the government embraced a more economic approach to industrial policy (identifying market failures and carefully constructing policies to correct them). While maintaining important privileges for favored sectors and enterprises, more favorable Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 85 conditions for all private investors included tax holidays and steps to lib- eralize the financial sector, deregulate commodity prices, and reduce bar- riers to trade and international capital movements (Loewe 2013). Private investment reached 51 percent of total investment during 2001–06, although due in part to declining public sector investment (Loayza and Honorati 2007). Despite numerous efforts, structural transformation did not fully materialize in Egypt. Galal and El-Megharbel (2005) indicate that indus- trial policies through 1999 did not achieve the goals of structural transformation. They consider two markers of structural transformation: whether product variety increased and total factor productivity improved. From 1980 to 1999, product concentration actually increased (variety fell), total factor productivity scarcely improved, and those industrial sectors that received the greatest assistance exhibited the lowest rates of productivity improvement.2 They argue that this is not surprising: policy over this period did not particularly target new activities; did not condi- tion assistance to firms on concrete goals, such as export success; left open the possibility that support to firms could continue indefinitely; and supported sectors rather than activities. The period from 2004 to 2011 is typically seen as representing a sharp turn towards a private sector-driven structural transformation, export growth, and job creation. In 2004–05, the government priva- tized 87 state-owned enterprises and reduced income taxes, before moving on to simplify customs procedures and business start-up regula- tions, while continuing to liberalize the financial sector. Policies seemed to focus on new markets (subsidies to exports), and new production technologies (subsidies for modernization), and they were more sub- stantial. However, vast areas of the economy remained closed to for- eigners, including aviation and engineering services, and heavy industry (energy production, steel and aluminum production, construction, insurance, and fertilizer). Moreover, in this period, more individual business people benefitted from first-tier personal connections with the government. Prior to 2000, approximately 8 percent of ruling party deputies were business people; from 2004 to 2011, these numbers increased to 17 percent of ruling party deputies and five ministers. The business people represented a miniscule fraction of firms in Egypt and were not politically accountable to them. Hence, they had stronger incentives to use their political positions to improve the investment climate for their own enterprises instead of the private sector in general. Because these business people typically repre- sented large enterprises, their closer ties with the government could have triggered observable improvements in Egyptian growth over the period. 86 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa A centerpiece of the 2004–11 period was the Egypt Industrial Development Strategy (EIDS), drafted by the Ministry for Trade and Industry (MFTI). The Strategy tracked closely the ideal prescriptions for successful industrial policy (Loewe 2013). It was designed to address: coordination failures in human capital by training workers and entrepreneurs (the Industrial Training Council); quality assurance (through the National Quality Council); financial markets (the Industrial Modernization Center); innovation and technology transfer (Technology and Innovation Centers); imperfect information about market opportunities (Export Council and Export Development Bank); and coordination problems in infrastructure and plant location (Industrial Development Agency). Loewe (2013) judges the EIDS to be an improvement over past industrial policies; he argues that FDI and exports surged as a result from 2004 to 2008. However, the EIDS was surrounded by both opacity with respect to the targeting of benefits to “insider” firms (Roll 2013), and weakness in measuring actual impact and costs and benefits of EIDS subsidies. For example, the composition of total FDI inflows into Egypt (and other MENA countries) is mostly concentrated in real estate and min- ing, which together account for 75 percent of total FDI. The high share of FDI flows into real estate, primarily from GCC countries, relativizes the importance for economic development. This is because capital accumulation in this sector typically has very limited scope for technology spillovers, expanding production capacities, or generating employment effects beyond construction periods. 3 In addition, on other dimensions of private sector growth and structural transforma- tion, the effects of EIDS were more ambiguous. Symptoms of market failure—limited research and development, insufficient coordination of complementary economic activities—seemed to barely change as a consequence of EIDS. In 2004, for example, total R&D spending was an almost imperceptible 0.27 percent of GDP; by 2008, it had actually declined to 0.23 percent. While the activities under EIDS were con- sistent with efforts to solve coordination failures, the program was not set up either to identify market failures or to evaluate whether it cor- rected them.4 The absence of clarity in the targeting of these subsidies and lack of rigorous tracking of their efficacy raised questions of privileges to specific firms. What explains the mixed results for EIDS? The most plausible reasons for greater exports were simply large government subsidies. Under EIDS the government made substantial financial transfers to ben- eficiary firms, particularly export subsidies (up to 15 percent of the value of goods) and modernization (up to 95 percent of the costs). These sub- sidies were likely sufficient to offset significant public policy distortions Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 87 in finance, human capital, and administrative interference. On the other hand, the subsidies were also large enough to yield significant rents for beneficiaries. Given the absence of rigorous tracking of their efficacy, this raised questions of privileges to specific firms in the distribution of EIDS benefits. For example, when tariff rates were reduced in Egypt at the end of the 1990s, Egypt apparently responded by increasing the use of nontariff technical import barriers.5 A new World Bank database allows measuring NTMs in various countries. Figure 3.1 illustrates the decline in average weighted tariffs from about 16.5 percent in 1995 to 8.7 percent in 2009— but also shows a steady and offsetting increase in NTMs. Of the 53 dif- ferent NTMs in place in Egypt in 2009, almost half (24) were introduced or amended around 2000, and 21 percent between 2005 and 2009. Of these, most were issued by the Ministry of Industry and Trade, which was headed at the time by a prominent businessman. As a result, Egypt had one of the highest NTM frequencies in the world in 2010 (Malouche, Reyes, and Fouad 2013; see also figure 4.5). Even if discrete policy initiatives were well designed and effectively implemented—a disputed assumption—the broader policy framework in Egypt did not constitute a successful industrial policy. While political connections evidently did not lead to broad benefits for all Egyptian industry, they delivered substantial benefits to the connected firms FIGURE 3.1 The Evolution of Average (Weighted) Tariffs and NTMs on Imports, 1995–2010 17 60 16 55 15 Number of NTMs on imports 50 14 Tariffs, in percent 45 13 40 12 35 11 10 30 9 25 8 20 1995 1998 2001 2004 2007 2010 Average tariffs (weighted) Number of NTMs on imports Source: Data are from WITS. Note: Rate reflects most-favored nation tariffs. The NTMs data provides either the year when an NTM has been introduced or the latest year in which it is has been substantially revised. 88 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa themselves. Diwan and Chekir (2012) identify 22 politically connected firms among the 116 largest Egyptian firms traded on the Egypt Stock Exchange. Following the Arab Spring revolution, they estimate the value of connected firms dropped 23 percentage points more than noncon- nected firms—that is, 23 percent of the expected future returns to invest- ments in these firms were contingent on political connections. Industrial Policy in Morocco Dramatic movements in the exchange rate, however, have consistently dwarfed the impact of other industrial policy initiatives targeting export. Moroccan industrial policy has long vacillated between providing selec- tively targeted benefits and benefits to all exporters. In the 1980s, Morocco sought to increase manufacturing with tariffs and licenses. However, over the same period the currency experienced a large, 40 percent real depre- ciation. This likely played the largest role in the significant rise in exports and manufacturing that occurred over the period (Achy 2013). The gov- ernment liberalized in the 1990s, reducing tariffs, margin controls, and other licensing requirements; it halted direct credits for exporters and increased the role of the market in the allocation of credit. Neither manu- facturing nor the economy in general prospered during this time—but this was likely the result of the 22 percent appreciation of the exchange rate over the period. In the 2000s the government began to use several selective investment promotion schemes to encourage job creation, export growth, and struc- tural transformation. Several investment promotion and tax exemption programs were used in the early 2000s to stimulate investment and struc- tural transformation. The largest was the Hassan II Fund for Economic and Social Development, which provided investment subsidies amounting to about US$560 million (4.5 billion dirham), mostly to textile manufac- turers and automotive suppliers. Starting in the mid-2000s many existing instruments were redefined to fit within a more comprehensive industrial policy program called Plan Emergence. It focused on the modernization of the industrial sector and offshoring. Eight sectors were initially selected: agro-food industry, seafood industry, textiles, automotive, aeronautics, electronics, and offshoring services particularly in French and Spanish languages. Investment incentives were granted to foreign as well as domes- tic firms. Again, as was the case throughout the region, market failures were not defined, nor were the efficiency effects of the subsidies ever evaluated. Moroccan industrial policy was not accompanied by significant admin- istrative reforms. A group of prominent and politically loyal business firms also enjoyed the capacity to act collectively, through the business Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 89 TABLE 3.1 The Cost of Industrial Policy in Morocco, 2010 millions of Moroccan Dirham Industrial policy measure Estimated cost, 2010 Value-added tax exemptions/rebates for capital goods 102 Tax exemptions for exporters 2,502 Tax exemptions for new enterprises in Tangiers, other targeted locations 697 Tax exemptions for locating in free export zones 55 Customs exemptions for capital goods imports by large investors 283 Auto industry customs exemptions 365 Hassan II Fund for Economic and Social Development 900 Total 4,904 Source: Adapted from Achy 2013, table 6. Note: US$ 1 = 8 Moroccan Dirhams. organization the Confédération Général des Entreprises du Maroc (CGEM). They saw their interests hurt by the removal of tariff protections, but when they used the CGEM to resist these reforms, the government responded by expanding the ranks of the CGEM to include more small and medium-sized firms. The government also began an anti-corruption campaign that targeted some in the business community (Achy 2013). Concurrently, and as in Egypt, the government also brought business representatives into the legislature. In addition, the reorganized CGEM began to call publicly and insistently for a level playing field in economic policy, and to act autonomously of the government. The degree to which this ability to act autonomously also protected member firms from oppor- tunistic changes in government policy is, however, unclear. One reason for the modest effect of these programs is that they were small. None of these policy initiatives seemed to have a perceptible effect on structural transformation. Achy (2013) catalogs all of the subsidies and their cost. In 2010, they amounted to approximately US$612 million, less than 0.7 percent of GDP (see table 3.1). Even if exceedingly well-targeted to market failures and credibly implemented with respect to time-bound goals, the industrial policy program in Morocco was small relative to the spending associated with industrial policy in East Asia. Industrial Policy in Syria, Jordan and Tunisia In Syria, the 10th five-year development plan (2006–10) emphasized the more rapid growth of manufacturing exports. Accordingly, similar to the EIDS effort in Egypt, investment and export promotion agen- cies were created, as were “industrial cities,” meant to support the clustering of manufacturing firms. In cooperation with UNIDO, 90 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa the government launched the Industrial Modernization and Upgrading Program, which focused its support on the textile and clothing sectors. Nevertheless, the program did not respond to any explicit analysis of market failures, nor was there an evaluation of the efficiency gains from allocating subsidies to these sectors rather than to other sectors, or to allocating no subsidies at all. Chahoud (2011) found little evi- dence that these initiatives were broadly implemented during the period. Similar to Syria, Jordan created a variety of programs to support indus- try, with responsibilities dispersed across several ministries. The Ministry of Trade and Industry created a strategy to support small and medium- sized enterprises, the Jordan Investment Board, which was charged with improving the business environment and, especially, allocating tax incen- tives to investors; the Development and Free Zones Commission was created to develop four regional development zones, meant to target spe- cific industries with substantial tax and other benefits. Even the Central Bank adopted a policy of reducing reserve requirements for private banks by an amount equal to their SME loans. Among all these programs in Jordan, perhaps the most significant has been the least systematic: tax incentives granted to selected firms and industries by the Council of Ministers. These are issued without transpar- ent conditions or evaluation procedures, and outside of a bureaucratic apparatus that could monitor the contribution of beneficiary firms towards growth or employment. BOX 3.2 Are GCC Countries an Exception? The GCC provides some cases of what appear to be successful industrial policy interventions. One class of successes relates to the energy sector development. The GCC countries entered the 1970s almost entirely concentrated in crude oil production operated by international companies, and with basic needs for access to services and infrastructure still unmet. In that decade, Saudi Arabia embarked upon a strategy to develop its own technical capacity in oil production along with facilities for oil refining and petrochemicals. Among the most ambitious of these interventions was the creation of two industrial cities, in Jubail (on the Gulf Coast) and Yanbu (on the Red Sea). These cities are governed by a Royal Commission (set up in 1975), which operates outside the administrative ministry structure and has complete autonomy over spatial planning, regulation, and investment in the cities. (continued on next page) Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 91 BOX 3.2 Continued The intention of the two cities was to transform the energy sector by promoting a cluster of subindustries related to petroleum products and petrochemicals, including associated logistics. All major production operations in the industrial cities are owned by Saudi Arabian Aramco, Saudi Arabian Basic Industries (Sabic, a government-created petrochemicals company), or joint ventures of one of these two companies with international partners. As a result of these efforts, Saudi Arabia now has a broad-based hydrocarbons sector, in which its massive oil endowment is complemented by a downstream value chain. Of course, this industrial capacity reflects a policy decision to provide industry with oil and gas inputs at below export price. The key to avoiding dissipation of this cost advantage lies in the effective commercialization of these companies through professional management, insulation from political pressures, and exposure to international best practices by forcing foreign investors into joint ventures (Hertog 2008). Whereas the Saudi Arabian example can be linked to its energy endowment, Dubai presents a more complex case where a services industry specialization did not have an obvious starting point. Instead a few critical decisions made by the leadership—dredging Dubai Creek to facilitate bigger ships; establishing a free zone around the new port at Jebel Ali to encourage transit and assembly activity; building up the airport and the airline; and encouraging foreign investment in finance and real estate (not least through liberal visa policies)—combined to set in motion a sustained boom and an acquired comparative advantage in logistics. While virtually every element of this strategy was implemented by state-owned companies, emirates were in constant competition with other emirates. As in Saudi Arabia, these companies were professionalized and run on a commercial basis. The context of Dubai provided further discipline. The emirate was in constant competition with other emirates and some decisions came as competitive responses to them. For example, Jebel Ali port was triggered by Sharjah’s initial moves to attract container traffic and Abu Dhabi has mimicked elements of the Dubai logistics strategy. Finance provides an additional lever. As a subnational entity without its own large oil resources, Dubai companies had to fund themselves through operations or debt—the latter forcing some analysis of viability and profitability from banks or securities markets. While banks were themselves closely linked to the government, they were run along sufficiently commercial principles to induce some genuine economic pressures on the SOEs. The irony was that the growth strategy, initially presented as diversification, was in fact linking the various facets of the growth closely together, as the Dubai debt crisis of 2008 showed. Nonetheless, the logistics network saw little adverse impact even at the peak of the crisis, indicating its resilience. At least part of the apparent success of GCC countries’ industrial policies is explained by the sheer size of the programs made possible by oil revenues. This contrasts starkly with the initiatives in other MENA countries. 92 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa In contrast with the Korean experience subsequently described, Jordanian benefits were uncoordinated, relatively small, and not condi- tioned on performance. In addition, there was no effort to identify market failures. Instead, government policy sought, as is often the case, simply to create industrial activity in sectors or regions where there was little. It did not distinguish whether there was little activity because of market failure or because of a simple lack of comparative advantage. Industrial policy in Tunisia took the form of special regulatory regimes for exporters, including generous tax and tariff rules. A central feature of Tunisian industrial policy was the formal offshore regulatory framework. For nonexporters, the firms in the onshore economy, the regulatory and tax environment did little to promote competition and innovation. On the contrary, they established significant barriers to entry of foreign or domestic firms, especially in service sectors where most of the politi- cally connected firms close to the Ben Ali family operated. The protec- tion of rents in service sectors likely also reduced the quality of backbone services provided to the rest of the economy (creating weak links), poten- tially also constraining productivity in the offshore economy despite the generous tax and tariff exemptions. What Did Successful Countries Do? The Case of the Republic of Korea Successful firms emerged despite the fact that the political challenges of promoting private sector job creation and structural transformation in Korea in the 1960s echoed those of the MENA region in the 2000s.6 The 1960s in Korea were a period of significant political unrest. Student demonstrations and military coups drove regime change, the president faced few institutional controls on his authority, and top officials earned significant rents (Kang 2002). Even as late as 1982, the first year for which governance indicators are available, Korea looked little different than Egypt in 2010, according to the International Country Risk Guide indicators of Political Risk Services. In sharp contrast to MENA and such programs as EIDS in Egypt, however, the firms supported by Korean industrial policy rose to become world-class producers. One of the outstanding success stories of Korean industrial policy was the for- merly state-owned Pohang Steel Company (POSCO). What explains these different outcomes, despite similar governance challenges and political risks? Ironically, in one important respect, Korean policies appear very much like those undertaken in MENA: selected industries received support with little attention to the identification of market failures or cost-benefit Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 93 analysis. However, Korean policies were substantially more generous, tightly linked to the achievement of private sector growth and structural transformation goals, and supported by significant organizational changes in both the public and private sectors. The magnitude of government intervention to drive private sector growth and structural transformation in Korea is well known. Private investors shy away from large commitments of capital in environments where leaders can predate on their investments with impunity. The poli- cies that Korea pursued in the 1960s and 1970s had precisely the effect of offsetting the reluctance of private investors to commit capital. First, President Park bypassed private investment altogether and placed heavy reliance on state-owned enterprises; government investment is naturally free from the hazards of government predation. The government estab- lished more than 20 major state enterprises in capital-intensive sectors (electricity, airlines, shipbuilding, steel, and so on). Second, private firms in priority sectors received massive direct and indirect subsidies, ranging from direct cash payments and tax exemptions to favorable import and foreign exchange regimes. These subsidies compensated investors in priority sectors for the political risks they incurred in committing sub- stantial amounts of capital to the private sector growth and structural transformation agenda. The government’s massive infrastructure invest- ments also effectively raised the private return to investment. From 1960 to 1970, Korea dedicated one-third of gross domestic investment to infra- structure and dramatically increased electricity generation and installed telephones. Why did these policies succeed in Korea? Three responses to this question are most plausible. First, by implementing policies to stimulate activities where none pre- viously existed, industrial policy in Korea was effectively, if not intention- ally, more likely to address market failure. Subsidies aimed at—and were conditioned on—creating economic activities in areas where there was none, most famously in the heavy and chemical industries. In addition, the state aggressively funded information acquisition—again, at least potentially addressing a market failure. For example, the state funded 97 percent of research and development expenditures in Korea in the early 1960s (Evans p. 147). In contrast, Galal and El-Megharbel (2005) show that industrial policy in the MENA region, including Egypt’s EIDS initiative in the mid-2000s, did not effectively target new markets or products—the ones most exposed to market failures. Second, the Republic of Korea credibly linked subsidies to export performance; even those that benefited insiders and cronies. In contrast with MENA, the implementation of industrial policy was conditional on firm success in pursuing private sector growth and structural transforma- tion. Not only were subsidies contingent on firms entering new activities, 94 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa but they had to succeed in those activities in order to continue to receive subsidies. That is, even if subsidies might also have been disproportionally channeled to politically connected firms in East Asian countries, these firms still had to meet the performance targets aligned with economic growth. Time-bound goals are central in order to provide entrepreneurs with incentives to innovate and invest. However, if political incentives to pur- sue growth are weak, friends of the regime are unlikely to regard dead- lines as credible. They will anticipate that governments will prefer to extend deadlines in exchange for rents, weakening their incentives to innovate. More generally, the efficacy of industrial policy hinges on entrepreneurs’ confidence that successful firms will not confront an abrupt and opportunistic change in the rules of the game (higher taxes, more intense regulation, and predatory behavior by officials).7 The greater this threat, the less credible are government promises and the larger must be the industrial policy subsidies that governments use to accelerate growth. In contrast, the policy benefits offered to firms in MENA were not conditioned on concrete goals, such as export success. Instead, the policy regime left open the possibility that support to firms could continue indefinitely. Third, the government made public and private sector organizational reforms to ensure the successful implementation, and the credibility, of industrial policy. On the one hand, it is technically difficult to design subsidies, their timing and their expiration; on the other hand, the private sector response to subsidies is greater to the extent that private firms trust in the credibility of future policies. President Park mandated wholesale changes in government and in the industrial structure of the economy to improve the government’s implementation capacity, and to make it costly for him to act opportunistically. In the narrow pursuit of better industrial policy, he established a super-ministry, the Economic Planning Board, to consolidate functions—previously scattered across various ministries— related to the formulation and implementation of industrial policies. More broadly, though, he substantially moderated the tendency, manifest over the years 1948–60, to treat the public administration as a spoils sys- tem, where civil service positions were used to reward political allies and supporters. Instead, the public administration reform imposed to practi- cally all positions the requirement that appointments be made on the basis of open, competitive examinations; increased the difficulty of those examinations8; linked promotions strictly to job performance; and pro- vided civil servants with job security. The administrative reforms imme- diately improved the capacity of the civil service to implement industrial policy: expertise was higher and promotion systems were linked more transparently to success in the public sector mission, which was private sector growth, jobs, and structural transformation. Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 95 Administrative reforms also improved the credibility of the presiden- tial commitment to industrial policy. They increased the ability of the bureaucracy to resist opportunistic policy reversals by the president. On the one hand, the elite Economic Planning Board, with widely recognized levels of expertise, gave the civil service an informational advantage with respect to the president. On the other hand, consistent with Gehlbach and Keefer (2011), the administrative reforms improved the ability of officials and firms to act collectively in the event that President Park reneged on his commitments. In contrast, before the civil service reforms, there was little horizontal cohesion among civil servants, who received their jobs through connections with higher level patrons. Gehlbach and Keefer (2011) argue that these institutional changes are sufficient to generate credible commitments. They also document similar changes undertaken by Deng Xiaoping in China, when he changed state and party organizations to support increased investment and faster economic growth. For example, promotions in the Chinese public sector (for example, from county executive to provincial executive) are contingent on achieving economic growth in their jurisdiction. These administrative reforms contrast sharply with the MENA experi- ence. The Social Fund for Development was once one of the most effi- cient and transparent agencies in Egypt (Loewe 2013). However, its preeminence faded in the face of political pressure to use the fund as a source of patronage jobs. This weakened the capacity of the government to implement industrial policy, but it also undermined the credibility of its policies, since bureaucracies organized around patronage are less effec- tive checks on opportunistic behavior by leaders. Fourth, the emerging industrial organization of the country also sup- ported collective action by the private sector. The Korean government famously encouraged very large industrial enterprises, the Chaebols. Large conglomerates, each representing substantial shares of total indus- trial employment in the country, and each the potential source of rents and campaign contributions to politicians, could more easily defend their interests before the state. For example, the nine Chaebols that received the plurality of bank loans in 1964 all had family members in high posi- tions of the ruling party or the bureaucracy (Kang, 189). Kang (190–192) argues that the arrangements between the Chaebols and the government allowed each to hold the other hostage—and, therefore, to make credible commitments. The top 20 Chaebols accounted for nearly 15 percent of nonagricultural GDP in 1975, but they were also heavily indebted, with debt-equity ratios approaching on the order of 350 percent. They needed the government, and so had every incentive to fulfill their commitments to pursue export growth, and to provide private financing to government. 96 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa However, the government needed them and could not let large swaths of the economy go out of business. Despite similarities between MENA and Korea in the linkages between large businesses and high positions in the ruling party, the political com- mitment to economic growth was more deeply rooted in Korea, taming the extent of rent-seeking. The prominence of individual families in the economic life of low and middle-income countries is a common occur- rence, and also pervasive in the MENA region. Korea appears to stand out, though, in two important ways. The “mutual hostage taking” char- acterized by Kang was more extensive and pervasive than in MENA. More important, the political commitment to economic growth was more deeply rooted, taming the extent of rent-seeking that typically accompa- nies oligarchic industrial structures. In this regard, all observers agree that the leadership in the Republic of Korea had a single-minded commitment to economic growth. From the Republic of Korea and China to Malaysia and Singapore, the govern- ments of East Asia structured their bureaucracies and ruling parties around the goal of economic growth. The political imperative of generat- ing growth motivated leaders to embrace organizational reforms that substantially reduced their discretion over the decisions of the bureau- cracy. Indeed, this commitment is the reason that the institutional reforms (civil service reform, Chaebols, infrastructure, and so forth) did not col- lapse into patronage, rent-seeking, and stagnation, as in other parts of the world. For example, when the political process allows bureaucrats to focus more on collecting rents from industrial policy than on using it to transform the economy, private-public dialogue is likely to yield corre- spondingly less useful information and leads potentially to counter- productive policies. What explains this commitment? This question is not unique to Korea. It also arises in the case of other East Asian “miracles.”9 The question persists because there is no systematic explana- tion of the unusually focused dedication of Park Chung Hee, Deng Xiao Pen, and Lee Kwan Yu to the goal of economic growth. Most explana- tions plausibly refer to the devastation of war, the need to support a large military, the tapering of aid, and concerns about public support (despite the nondemocratic nature of these regimes). Lessons from East Asia Are More Difficult to Implement than Is Commonly Understood East Asia implemented its industrial policies much differently than MENA and in ways that leaders in many countries often resist. The modalities these countries used to implement industrial policy imposed Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 97 considerable limits on the discretion of leaders. These also included organizational reforms that restricted their discretion and strength- ened the credibility of their commitments to the reforms. In addition, the pursuit of growth as a strategy of gaining political support neces- sarily substituted for other strategies. However, this focus on growth came at a cost. Comparing only the within-sector benefits of industrial policy (abstracting from important potential spillover benefits on other sectors), the resources used for private sector growth and struc- tural transformation were, in many cases, diverted from uses that might have delivered greater welfare to citizens, including Koreans’ own consumption. For example, East Asian leaders could have preferred large consumer subsidies to build legitimacy. However, pri- vate sector growth and structural transformation, Korean style, is expensive and incompatible with the subsidies common in the MENA region. Other attributes of industrial policy in East Asia are, however, worthy of replication. Industrial policies • offset governance and political risks; • were complemented by infrastructure construction and ample support for human capital acquisition; • focused on activities that were entirely absent in the economy; • were accompanied by far-reaching organizational reforms in the public sector; • were implemented in an environment of a single-minded focus on growth; • tightly linked subsidies to the success in more competitive export markets; and • were applied at the sector, rather than the firm level. The firm-directed industrial policies common in the MENA region distorted competition and growth. Industrial policy in MENA had a ten- dency to privilege individual (connected) firms instead of benefitting all firms (and new entrants) in targeted sectors.10 These firm-specific poli- cies concentrated benefits on privileged firms, not sectors. These policy privileges provide these firms with potentially large exogenous cost advantages over their competitors in the same sector. Thus, the Schumpeterian growth framework predicts that they lead to less neck- and-neck competition and hence growth; that is, they drive nonconnected firms out of the market and suppress the incentives to innovate (to escape competition) for all firms in the sector. In contrast, Aghion et al. (2012) 98 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa indicate that industrial policy can promote productivity growth when it favors competition by reducing costs for all firms and entrants in a sector. The authors argue that industrial policy in China has done precisely this. Moreover, they show that sectors in China that benefited from more uni- form intra-sector subsidies exhibited greater productivity growth. The even-handed, effective application of the policy requires that gov- ernment decision makers be relatively immune to the influence of vested interests, at least in fast-growing modern sectors. In the absence of a single-minded focus on growth, the political cost of catering to vested interests is low, political incentives are correspondingly higher to privi- lege some firms over others, and to pursue industrial policy even when it has no demonstrated positive effect on development. The single-minded focus on growth, however, might be the most difficult to replicate as it implicitly requires a new social contract between government and citizens. Much is made of the social compact in MENA, one that trades government employment and consumer subsidies for limitations on expressions of citizen voice. These same limitations were pervasive in East Asia, but the social compact took a different form, emphasizing jobs and productivity growth. Moreover, the organizational changes in the public sector have been consistently some of the hardest for MENA governments to accommodate, and yet played an essential role in the success of industrial policies in East Asia. Policy makers can choose an alternative and potentially cheaper strat- egy to accelerate private sector job creation and structural transforma- tion, by reducing the impact of policy failures before seeking to address market failures; or by using expensive subsidies to offset the costs of both. Government policy failures can rival or exceed market failures as obsta- cles to job creation and structural transformation. In many countries, however, industrial policies such as those followed in East Asia are too expensive or too difficult to implement credibly. Moreover, government- induced market failures in MENA have arguably been at least as signifi- cant a barrier to growth as information asymmetries and coordination difficulties in private markets. Notes 1. This section is based on Keefer (2014). 2. TFP grew in the Arab Republic of Egypt at a 3.3 percent annual rate from 1983 to 1990, before dropping to 1.6 percent from 1991 to 2000 and to 1.1 percent from 2001 to 2006. TFP growth in the private sector soared to 5.6 percent in the 1980s, falling to 1.9 percent from 1991 to 2006 (Loayza and Honorati 2007). Avoiding the Pitfalls of Industrial Policy: Program Design in MENA and East Asia 99 3. In contrast, figure 2.1 shows that FDI inflows into China, Brazil, Indonesia, and India were concentrated in manufacturing or high technology services which typically have high potential for spillovers in technologies, production capacities, and employment. 4. In discrete cases, however, government efforts to solve coordination failures may have succeeded. Loewe (2013) points in particular to the marble sector and fashion industry. 5. The World Bank database on NTMs provides either the year when a particular NTM has been introduced or the latest year in which it is has been substantially revised. Unfortunately, the database does not distinguish between the two. 6. In 1980, Korea’s real purchasing parity power-adjusted income per capita ($5,543) was the same as Egypt’s in 2010 ($5,760). 7. These problems of credible commitment are pervasive. In monetary policy, for example, governments have an incentive to deviate from low inflation policies to reduce government debt burdens. They increase the credibility of their low inflation commitments by increasing the independence of central banks. 8. Only about 4 percent of those filling the higher entry-level positions had taken the civil service exam (Evans, 52). Under Park, the civil service became more strictly meritocratic, such that approximately 20 percent of those tak- ing high entry level positions had passed the civil service exam. The exam also became more difficult. One sign of its difficulty: between 1963 and 1985, 157,000 persons took the civil service exam and 2,600 passed it. 9. Outside of East Asia, Rodrik and Subramanian (2005) argue that Indian growth was driven by a change in attitude of Indian leaders: they began to see growth as a viable strategy for political survival. 10. Some have argued that industrial policies should be “horizontal,” applying to all sectors. This has the appeal of at least superficially preserving a “level playing field.” However, for two reasons horizontal policies may be incom- patible with this goal. First, identical policies have heterogeneous effects across sectors (cheap capital or energy favor capital- or energy-intensive in- dustries). Second, horizontal policies have a limited economic rationale to the extent that market failure drives industrial policy, since market failures are heterogeneous across sectors. Sectoral policies could, however, be broadly targeted. For example, countries could promote an export sector, comprised of many different lines of economic activity, by undervaluing their exchange rates. References Achy, Lahcen. 2013. “Structural Transformation and Industrial Policy in Morocco.” Economic Research Forum Working Paper 796, Economic Research Forum, Cairo. Aghion, P. M., Dewatripont, L. Du, A. Harrison, and P. Legros. 2012. “Industrial Policy and Competition.” Mimeo (April 24). Chahoud, Tatjana. 2011. “Syria’s Industrial Policy.” Working paper, Deutsches Institut fuer Entwicklungspolitik (DIE), Bonn. 100 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Diwan, I., and H. Chekir. 2012. “State-Business Relations in Mubarak’s Egypt.” Mimeo, World Bank, Washington, DC. Galal, Ahmed, and Nihal El-Megharbel. 2005. “Do Governments Pick Winners of Losers? An Assessment of Industrial Policy in Egypt.” Working Paper 108, The Egyptian Center for Economic Studies, Cairo. Gehlbach, Scott, and Philip Keefer. 2011. “Investment without Democracy: Ruling-Party Institutionalization and Credible Commitment in Autocracies.” Journal of Comparative Economics 39 (2): 123–278 Harrison, Anne, and Andres Rodriguez-Clare. 2010. “Trade, Foreign Investment, and Industrial Policy for Developing Countries.” In Handbook of Development Economics 5: 4039–214. Hausmann, Ricardo, and Dani Rodrik. 2006. “Economic Development as Self- Discovery.” Journal of Development Economics 72. Hertog, Steffen. 2008. “Petromin: The Slow Death of Statist Oil Development in Saudi Arabia.” Business History 50 (5): 645–67. Kang, David C. 2002. “Bad Loans to Good Friends: Money Politics and the Developmental State in South Korea.” International Organization 56 (1): 177–207. Keefer, Philip. 2014. “Industrial Policy and MENA—Lessons from Research and East Asia.” Mimeo, World Bank, Washington, DC. Loayza, Norman, and Maddalena Honorati. 2007. “Investment and Growth in Egypt.” Mimeo, World Bank, Washington, DC. Loewe, Markus. 2013. “Industrial Policy in Egypt 2004–2011.” DIE Discussion Paper 13/2013, Deutsches Institut fuer Entwicklungspolitik, Bonn. Malouche, Mariem, José-Daniel Reyes, and Amir Fouad. 2013. “New Database of Nontariff Measures Makes Trade Policy More Transparent.” Mimeo, World Bank, Washington, DC. Murphy, Kevin, Andrei Schleifer, and Robert W. Vishny. 1989. “Industrialization and the Big Push.” Journal of Political Economy 97 (5): 1003–1026. Rodrik, Dani. 2008. “Normalizing Industrial Policy.” Commission on Growth and Development, Working Paper 3, World Bank, Washington, DC. Rodrik, Dani, and Arvind Subramanian. 2005. “From Hindu Growth to Productivity Surge: The Mystery of the Indian Growth Transition.” IMF Staff Papers 52 (2): 193–236. Roll, Stephan. 2013. “Egypt’s Business Elite after Mubarak: A Powerful Player between Generals and Brotherhood.” SWP Research Paper 8 (September), Stiftung Wissenschaft und Politik, Berlin. CHAPTER 4 Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA This chapter shows that policies in MENA have often been captured by a few politically connected firms. This has led to a policy environment that created privileges rather than a level playing field, undermining competition, the ability of all entrepreneurs to pursue opportunities on an equal footing, and job cre- ation. The analysis builds on new data and information on first-tier politically connected firms in the Arab Republic of Egypt and Tunisia that became avail- able after the Arab Spring—and from more qualitative evidence from other countries in the region—that allow us for the first time to provide direct quan- titative evidence on how firm privileges affect competition, the level playing field, and job growth in the region. Taken together, the findings shed light on the entire microeconomic transmission mechanism, from privileges to limited competition and unleveled playing fields, to weak firm dynamics and slow aggregate job growth. This chapter provides evidence that many policies in MENA favor privi- leges over innovation and jobs. In the Schumpeterian growth framework, influential political connections provide firms with an outside option to escape competition by tilting regulations towards their favor instead of innovating. Aghion et al. (2001) predict that growth declines if a few col- luding market leaders have sizeable cost advantages, which are unbridge- able by competitors operating in the same sector. Chapters 2 and 3 document examples of policies in MENA that favor specific types of firms over others. If these privileges are large enough, the model predicts that sectors end up with a few colluding, politically connected market leaders; a potentially large number of unproductive micro firms; and most impor- tant, lower productivity and job growth. The more widespread these firm-specific privileges across sectors, the lower are aggregate growth and job creation. 101 102 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa The literature on Arab capitalism contains rich analyses of how auto- crats granted exclusive privileges to business elites allowing them to dom- inate the business sector in exchange for support for the regime. Qualitative research has documented barriers to entry that excluded opponents and provided privileges to a small coterie of friendly capitalists (Henry and Springborg 2010; Heydemann 2004; King 2009; Owen 2004). In the Arab Republic of Egypt, observers argue that cronyism thrived in the “businessmen” cabinet headed by Ahmad Nazif from 2004 to 2011 (Kienle 2001; Sfakianakis 2004). In Tunisia, the Ben Ali and Trabelsi families monopolized business opportunities and even expropri- ated the real estate and business holdings of wealthy elites. Similar stories about favoritism and insiders abound in the Syrian Arab Republic, Libya, the Republic of Yemen, and Algeria, where political cronies seem to control large chunks of the private sector (Alley 2010; Haddad 2012; Tlemcani 1999). However, previous work came short of providing quan- titative evidence associating the privileges to specific policies or showing their impact on economic performance. We use novel data from Egypt and Tunisia to test whether political connections lead to large privileges, and hence lower competition and growth. Chapter 2 analyzed several policies in MENA that benefit specific types of firms, potentially distorting neck-and-neck competi- tion; these include energy subsidies to industry, licenses, access to land, and biased regulatory enforcement. Two novel data sets on politically connected firms in Mubarak’s Egypt and Ben Ali’s Tunisia allow quantifying for the first time whether these policies disproportionally benefitted connected firms. They also allow us to quantify for the first time if the presence of politically connected firms changes sectors’ market structures and aggregate job growth in line with the predictions of Aghion et al. (2001). Moreover, this chapter provides evidence from other MENA countries and discusses to which extent privileges are a regional phenomenon. In “Privileges to Politically Connected Firms Undermine Competition and Job Creation: Evidence from Egypt and Tunisia” section, we use novel data sets on first-tier politically connected firms in Egypt and Tunisia to quantify their economic impact in both c ountries. 1 “The Available Qualitative Evidence Points to Similar Mechanisms of Policy Privileges in Other MENA Countries” section presents more qualitative evidence on policy privileges in other MENA countries. In “The Extent to which Political Connections Hampered Competition Differed in MENA and East Asia” section, we highlight potential factors that explain why private sector and jobs outcomes were different in MENA than East Asia, in spite of the presence of politically connected firms in both regions. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 103 Privileges to Politically Connected Firms Undermine Competition and Job Creation: Evidence from the Arab Republic of Egypt and Tunisia The governments in Egypt and Tunisia erected barriers to entry and competition even as they engaged in economic liberalization. In Egypt, President Hosni Mubarak’s son, Gamal, working closely with a group of economic experts and ambitious businessmen, shifted the country’s policies in the early 2000s towards accelerated privatization and financial sector and trade reforms. Insider firms were able to cap- ture the opportunities that emerged with the modernization of the economy2; these included massive real estate and construction projects, tourism at coastal areas, the oil and gas sectors, the banking sec- tor, telephony, and local distribution of international consumer brands. Government decisions were key in all of these areas—for example, connected families invested in specific manufacturing or mining sectors such as cement or oil and gas where each new factory required government approval; they obtained privileged access to state procurement contracts or exclusive licenses to distribute interna- tional brands in Egypt, shielding them from domestic competition; they entered the real estate, tourism, and transport sectors by acquiring large sections of prime land from the government, reportedly, involv- ing closed and nontransparent deals.3 In fact, connected businessmen were well placed to influence these decisions: they were not only personally well connected with the political leadership, but they them- selves also occupied important post in government, the ruling party, parliament, and various influential boards and committees. In Tunisia, the Investment Law was amended several times in the 2000s to provide incentives for private sector investments in the offshore economy but, at the same time, also to protect connected firms from competition in the onshore economy. The amendments included gener- ous tax breaks for firms operating in the offshore economy. In addition, it stipulates the freedom to invest for both foreign and domestic entities. However, it also contains provisions that restrict this freedom, including authorization requirements and FDI restrictions in the onshore econ- omy, which allow the government to control the entry of selected firms in some lucrative service activities. The Ben Ali family’s business interests in these services sectors were not a secret. In part, however, because Tunisia registered stable positive growth rates hovering around 4–5 percent per year, Ben Ali also had a somewhat favorable external image. The World Economic Forum repeatedly ranked Tunisia as the most competitive economy in Africa, and the IMF and the World Bank heralded Tunisia as a role model for other developing countries. Yet, at 104 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa the same time there were few formal sector jobs and perceptions of corruption was high. This section demonstrates that policies in both countries have often been captured by a few privileged firms, thereby limiting competition, distorting the playing field, and curtailing job creation. First, we discuss our measures of political connectedness and highlight the characteristics specific to connected firms. Second, we document that politically con- nected firms profited disproportionately from policy privileges in Egypt and Tunisia, distorting the playing field in both countries. The evidence implies that business regulations in Egypt and Tunisia were abused as a rent creation vehicle for friends and family of the two former presidents. Third, we show in more detail for Egypt that the presence of connected firms reduced the dynamism and growth opportunities for the rest of the economy; i.e., firm entry is lower in sectors where connected firms are already present and aggregate employment growth declines once connected firms enter new, previously unconnected sectors. The results suggest that distortive policies, such as authorization requirements, energy subsidies to industry, trade protection, and burdensome regula- tion benefit a small group of “profitable” firms, but reduce the total number of jobs created in Egypt and Tunisia. Notably, most of these business regulations are still in place. Identifying the Politically Connected Firms and Their Economic Significance Who Are They? To examine the economic effects of insider privilege, we need both a data set of politically connected firms and information about firm perfor- mance. In Tunisia, we use government data on 214 Ben Ali firms confis- cated by the Tunisian authorities in the aftermath of the Jasmine revolution. The confiscation involved 114 individuals, including Ben Ali himself, his relatives, and his in-laws, and involved the period from 1987 until the outbreak of the revolution. The seized assets included some 550 properties, 48 boats and yachts, 40 stock portfolios, 367 bank accounts, and approximately 400 enterprises, not all of which operate in Tunisia. The confiscation commission estimates that the combined total value of the confiscated assets of the Ben Ali clan is approximately US$13 billion, about one quarter of Tunisian GDP in 2011. We obtained a list of 252 confiscated firms from the Tunisian authorities, of which we were able to identify 214 firms with available data in the Tunisian annual firm census (Tunisian Business Register).4 The census contains information on the size, age, location, and legal form of all private nonagricultural registered firms in Tunisia, including one-person firms without paid employees. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 105 The census data are further merged with administrative data from the tax authorities, containing balance sheet information, and information on business regulations from the Tunisian Investment Law from 1993 to 2010. To identify politically connected individuals in Egypt, we followed Fisman (2001) and interviewed managers of banks and private equity funds, lawyers, and nongovernmental organizations (anti-corruption organizations) after the fall of Mubarak in 2011 to create a list of politically connected businessmen. We confirmed the representativeness of this list in two ways. First, we matched this list with the names of businessmen whose assets were frozen immediately after the regime change. Second, we pruned the list to include only those businessmen who had political posts in the ruling party or in the government, or whose immediate family members did. We also had sufficient information to identify long-term friends of the Mubarak family; these were also identified as connected businessmen.5 We matched this list with firm data from the OECD Orbis database, which includes information on the board members, managing directors, and major shareholders for 854 firms that are currently or were formerly traded on a stock exchange.6 We were able to unambiguously match the names of the 32 businessmen identified in step one with board members, managers, and major shareholders of 104 firms. Several of the connected firms in Egypt are holding companies and investment funds. Using the Internet, we identified the names of all subsidiaries—up to two tiers—of these 104 firms, and matched these subsidiaries with firms in the Orbis database. This process identified 469 firms that are unambiguously controlled, directly or indirectly, by a connected businessman. Of these firms, 47 have at least one politically connected businessman as a general manager (CEO) and in 334, at least one connected businessman or firm was unambiguously identified to have an ownership stake. In addition, in 172 firms a private equity fund owned by at least one politically connected businessman an ownership stake.7 Moreover, politically connected firms are widely spread across the 320 nonfarm, nongovernment four-digit ISIC Rev.4 sectors: about half (49 percent) of the sectors include connected firms (186 out of 372). Within manufacturing, where 41 percent of the connected firms operate, they are present in 58 percent of the four-digit industries (73 out of 126). We combine the information on politically connected firms in Egypt with four sources of data. First, the Orbis database has firm characteristics— including firm names—and balance sheet variables for a panel of over 20,000 establishments between 2003 and 2012, which allows us to com- pare the performance of connected and unconnected firms.8 While pro- duction data on small enterprises are frequently missing in Orbis, the data on medium and large establishments, the right comparison group for 106 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa politically connected firms, are comprehensive. Second, establishment census data from the department of statistics in Egypt (Central Agency for Public Mobilization and Statistics) do not contain firm names, but they do allow us to estimate how the dynamics across detailed four-digit sectors change depending on the presence of connected firms. The cen- sus includes employment and firm characteristics of over two million nonfarm economic establishments in 1996 and 2006. Third, World Bank Enterprise Survey (WBES) data allows us to assess correlations between the presence of connected firms and perceived policies.9 Fourth, to inves- tigate whether connected firms benefitted from state-supported barriers to entry or energy subsidies, we use information on nontariff barriers to trade (NTMs) from the World Bank (WITS), and UN data on the energy intensities of manufacturing industries. We only observe a subset of politically connected firms in Egypt and Tunisia: those private sector firms with first-tier political connections to the Mubarak or Ben Ali family. However, there are other connected firms. Reportedly, the most important group of firms is controlled directly or indirectly by the Egyptian army, which operates businesses in tourism, construction, white goods, vehicles, fertilizer, mineral water, olives, and bread. Most of these businesses initially were financed by the sale of government land in Cairo and on the seaside (Loewe 2013). Similarly, the sample of 214 connected firms in Tunisia is most likely skewed towards the largest and economically most relevant firms since these are easier to identify. Where Are They? A direct comparison of the distribution of politically connected firms across countries would suggest that this phenomenon was more widespread in Egypt. While the number of connected firms should be regarded as a lower bound in both countries, we observe fewer connected firms in Tunisia (214) relative to Egypt (469). Moreover, connected firms were substantially larger and economically more significant in Egypt: connected firms employ on average 941 workers in Egypt 66 relative to workers in Tunisia; they accounted for about 7 percent of total private sector employment in Egypt relative to about 1 percent in Tunisia.10 These disparities might originate from the different nature of the data. The nature of political connections is also different, since the confiscation commission in Tunisia focused exclusively on firms owned by members of the Ben Ali family. In contrast, the Egypt data also include first-tier Mubarak associates—connected businessmen with influential political posts, whose assets were also confiscated in 2011. It is unclear to which extent first-tier political connections beyond the extended Ben Ali family played a role in the Tunisian economy. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 107 The presence of politically connected firms appears to be more wide- spread across various economic activities in Egypt, especially among manufacturing industries. Despite the differences in the nature of the data, the distribution of connected firms across sectors exhibits some similarities (table 4.1). In both countries, connected firms are concen- trated in real estate, business services, tourism, wholesale and retail trade, mining, telecommunications, and transport services. In Egypt, however, their activities reach far beyond these sectors. Politically connected firms operate in 49 percent of all nonfarm, nongovernment four-digit sectors (186 out of 372). In contrast, Ben Ali firms operate in only 14 percent of all five-digit sectors in Tunisia (45 out of 321). In particular, the concen- tration of politically connected firms in various Egyptian manufacturing industries is striking: 42 percent of politically connected firms operate in manufacturing in Egypt relative to 13 percent of connected firms in Tunisia (table 4.1). Manufacturing industries are typically considered harder to protect from (international) competition. In fact, the subse- quent analysis shows that, in Tunisia, policies protecting connected firms from competition focused on service sectors in the onshore economy while such restrictions were absent for manufacturing firms in the offshore economy. For example, Ben Ali firms dominated the telecommunications TABLE 4.1 Number of Politically Connected Firms, by Economic Sectors Sector Egypt, Arab Rep. Tunisia Mining 12 8 Manufacturing 193 31 Food and beverages 33 9 Textiles and apparel 22 2 Chemicals 28 3 Base metals 19 2 Machinery and equipment 27 2 Other manufacturing 64 13 Utilities 18 0 Construction 36 9 Services 388 166 Wholesale trade 91 38 Retail trade 25 3 Transport 13 16 Hotels and restaurants 43 7 Finance 53 8 Real estate and construction 138 59 Other services 25 35 Total 647 214 Source: World Bank calculation. Note: The last row represents the total number of politically connected firms operating in each four-digit sector. In Egypt, it amounts to 647 because several of the 469 connected firms operate in more than one four-digit sector. In Tunisia, we observe only one (i.e., the main) sector for each firm in the data. 108 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa and air transport sectors, and were also important players in other trans- port sectors and real estate, all sectors in which entry is highly regulated. In contrast, policies favoring connected businessmen in Egypt (e.g., trade protection or energy subsidies to industry) also profited selected manufacturing firms. The stronger presence of first-tier politically connected firms across sectors in Egypt over the past decade might indicate that the regime aimed to tighten control of the economy (i.e., the recipients of extracted rents and their potential use for political financing). By 2010, the activity of politically connected firms in Egypt was not constrained to more mature traditional sectors but they also operated in some younger modern sectors (e.g., manu- facture of pharmaceuticals or plastics). Moreover, table G.2 in Appendix G shows that connected firms entered various new sectors between 1997 and 2006 which had been open (i.e., not connected) before (e.g., manufacturing of batteries or computer programming services) while they did not enter others (e.g., manufacturing of optical instruments or specialized design services). Thus, there is also substantial variation of the presence of politically connected firms even across four-digit industries within the same two-digit sector. These attributes of the distribution of connected firms across sectors in Egypt aid in the empirical identification of the impact of political connec- tions on four-digit sector outcomes in the subsequent analysis. Politically Connected Firms Are Highly Profitable The few politically connected firms in both countries accounted for the lion’s shares of profits. Profits are measured as operating profits declared to the tax authorities.11 Among medium and large establishments in Egypt, politically connected firms accounted for only 11 percent of total employment, but 60 percent of total net profits.12 The average net profits were 13 times higher for the 49 connected establishments included in the available data, indicating that at least some of the politically connected firms make excessively high profits (table 4.2). Among all firms in Tunisia, the 214 confiscated Ben Ali firms appropriated 21 percent of all net private sector profits in 2010.13 In contrast, they accounted for only about 1 percent of all wage jobs. Ben Ali firms also report significantly higher profits when looking at within-sector comparisons (table 4.2, last column). The potential advantages of connected firms that lead to their higher profits are specific to the individual firm, or to the product it sells. Table 4.2 reports the descriptive statistics among politically connected and other firms in Egypt and Tunisia. Politically connected firms are significantly larger than other firms, both in terms of employment and output. The fourth column reports the difference in performance between connected and other firms that operate in the same two-digit sectors. It shows that the performance differences are not specific to the broader Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 109 TABLE 4.2 Within-Sector Differences, Politically Connected and Other Firms PC vs. other PC vs. other No. of PC No. of other PC vs. other establishments, within establishments, within establishments establishments establishments two-digit sector four-five-digit sector Egypt, Arab Rep. Ln(Employment) 436 19,375 1.40** 1.02** 0.97** Ln(Revenues) 67 611 1.61** 1.59** 1.50** Ln(Profits) 49 239 1.43** 1.37* 1.29 Ln(Profits/Rev) 47 236 1.88** 2.17** 1.02 Tunisia Ln(Employment) 114 81,180 1.61** 1.49** 1.05** Ln(Revenues) 81 250,340 5.17** 4.27** 2.38** Ln(Profits) 94 93,098 −1.41** 0.10 1.10** Ln(Profits/Empl) 64 41,760 −0.08 0.88** 0.01 Source: World Bank calculation. Note: Data are from the Orbis database (Egypt) and the firm census (Tunisia). In contrast to the firm census in Tunisia, the Orbis data for Egypt primarily include medium and large establishments, which are the correct comparison groups when comparing politically connected and unconnected establish- ments. The statistics in Egypt show the results for the broadest measure of political connections, which also include firms that received significant invest- ments from politically connected private equity funds. Columns 3–5 report the coefficient and t-statistic on the politically connected dummy variable, from an OLS regression of the performance variable (e.g., Ln(employment)) on the dummy variable which is equal to 1 for politically connected firms and 0 otherwise. In the fourth (fifth) column, we also include two(four/five)-digit sector dummies so that the connection dummy coefficient measures the difference between connected and unconnected firms operating within the same two(four)-digit sector. * and ** indicate that the coefficients are sig- nificant at the 5 percent and 10 percent level, respectively. Note that to account for negative profits, we use a transformation of the log profits measure ( ) ( ) that also accounts for negative profits, notably log Profits + Profits 2 + 1 . Similarly, ln(Profits/L) is constructed as log Profits + Profits 2 + 1 − lnL. sectors in which firms operate. In other words, if connected firms receive preferential benefits or treatment, these must not be sector specific, but rather specific to the connected firm or the individual product it sells. The last column shows that after controlling for detailed four-digit sectors (product classes), politically connected firms in Egypt cease to have significantly higher profit margins relative to other firms, suggesting that portions of their higher profits originate from characteristics specific to the product classes they are selling. The significantly larger net profits in Egypt were systematically related to the survival of the regime. Figure 4.1 plots the evolution of the differ- ences in (log) net profits between politically connected and other large firms from 2003 to 2011. After the fall of the Mubarak regime on February 11, 2011, the positive profits differential of politically connected firms sud- denly disappeared.14 The finding suggests that the larger profits of politi- cally connected firms originated from firm-specific factors directly related to the existing political regime, such as firm-specific privileges in the form of subsidies or trade protection, rather than the greater entrepreneurial skills of the managers, which are independent from regime shifts. The fact that the profit differential between connected and unconnected firms dis- appears shortly after the fall of Mubarak also corroborates the quality of our empirical measurement of politically connectedness in Egypt.15 110 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 4.1 The Evolution of Net Profit Differentials between Connected and Other Firms, 2003–11 6 4 Difference in net profits, in log 2 0 –2 –4 –6 2003 2004 2005 2006 2007 2008 2009 2010 2011 5% confidence interval lnProfits difference 95% confidence interval Source: World Bank calculation. Note: Data are from Orbis establishment database and establishment census. Connected Businessmen Capture Policies to Secure a Range of Privileges Trials of leading businessmen since the Arab Spring have shed light on the potential mechanisms through which privileges were granted to connected firms; they revealed several common practices to favor politically connected firms in Egypt and Tunisia, including land appropriation at below-market prices; the manipulation of government regulations to stifle competition; and privileged access to subsidized energy and state procurement con- tracts.16 Our newly constructed data allows for the first time to uncover empirically the main policy privileges granted to connected firms and show how they tilt the level playing field and affect competition. Lastly we show that these privileges led to large profits of connected firms but came at a high cost for private sector growth and job creation. Politically Connected Firms Are Insulated from Competition through Various Entry Barriers The Investment Law in Tunisia requires prior authorization from the government in order to operate legally for a number of activities; includ- ing fishing, tourism (travel agencies), air transport, maritime transport and road transport, telecommunications, education, the film industry, real estate, marketing, and health-related industries. If not administered Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 111 equitably, these authorization requirements can be abused to create market power and stifle competition, both from prospective entrants and incumbents. Anecdotal evidence suggests this happened in the case of the closing of the Bouebdelli School, a highly respected private school from which many of Tunisia’s elite have graduated. This school was perceived to be in direct competition with an international school founded by the Ben Ali family. In spite of widespread public protests, the Minister of Education ordered the school closed for failure to comply with registra- tion regulations.17 The Investment Law also stipulates a number of activities for which foreign firms are required to obtain permission from the Investment Commission (CSI), when their foreign equity exceeds 50 percent of capi- tal. These include transport, communications, tourism, education, cultural production, entertainment, construction, real estate, computer services, and a select number of other services. Obtaining such permis- sion is notoriously difficult. Since 2005, the CSI has been processing between two and three applications per year with roughly half of all applications being successful. The list of sectors subjected to restrictions on foreign investment overlaps considerably with those that are sub- jected to government authorization. We note that many other sectors are also subject to government intervention, but not through the Investment Code. Restrictions on foreign entry likely limit foreign competition and can also be used to direct foreign funds to certain domestic firms. The failed entry of McDonald’s into the Tunisian food market is often used to illustrate the Ben Ali family’s hold on specific sectors. The exclusion of McDonald’s from the Tunisian market followed from their unwillingness to grant the sole license to a franchisee with family connections. The government of Tunisia in turn refused to grant authorization to invest.18 Connected firms are more likely to operate in sectors which are protected from competition through entry barriers. Figure 4.2 illustrates that 39 percent of the sectors with at least one Ben Ali firm require previ- ous authorization by the government, relative to 24 percent of noncon- nected sectors.19 Similarly, 43 percent of connected sectors are protected from foreign entry relative to only 14 percent of nonconnected sectors. Moreover, 64 percent of Ben Ali firms are in sectors subject to authoriza- tion requirements and 64 percent are in sectors subject to restrictions on FDI. For nonconnected firms the comparable numbers are 45 percent and 36 percent, respectively.20 Egypt imposed more nontariff barriers to import than most other coun- tries in the world. Tariff rates were reduced in Egypt at the end of the 1990s; at the same time, however, the government increased the use of 112 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 4.2 Authorization Requirements and FDI Restrictions Protect Politically Connected Firms in Tunisia 50 40 (24/56) (22/56) 30 Percent 20 (69/267) 10 (39/267) 0 Authorization FDI restriction Sectors with BA firms (45 out of 332) Sectors without BA firms (287 out of 332) Source: World Bank calculation. Note: Differences between Ben Ali and other firms are measured at the five-digit sector level (no. of restricted sectors/total no. of sectors). The difference in authorization requirements Fisher’s t test probability = 0.04) and FDI restrictions Fisher’s t test probability = 0.00 between connected and nonconnected sectors is significant at the 5 percent level. nontariff technical import barriers (see figure 3.1). As a result, Egypt had one of the highest nontariff measures (NTM) frequencies in the world in 2010 (Malouche, Reyes, and Fouad 2013). Most NTMs in Egypt are “Class B” NTMs, legal technical barriers to import, including license or registration requirements for importers; regulations on production and dis- tribution processes; traceability; and product quality requirements. They are imposed on 65 percent (96 out of 147) of the four-digit manufacturing industries. All of these restrictions make it harder for foreign companies to sell their goods and services in Egypt and thus can be abused to create market power and protect domestic firms from foreign competition. Politically connected firms in Egypt are more likely to sell products protected from foreign competition. Table 4.5 shows that NTMs dispro- portionally benefitted politically connected firms21; i.e., manufacturing and mining industries in which politically connected firms are present are more likely to be protected from import competition by NTMs than sectors without politically connected firms. Politically connected firms are also more likely to be protected by NTMs at the individual establish- ment level; 82 percent of all politically connected manufacturing and mining establishment sell products that are protected by technical non- tariff import barriers. In contrast, only 56 percent of all manufacturing or mining establishments in Egypt in 2006 operated in these sectors. The gap in trade protection between politically connected and other firms increases substantially with the number NTMs imposed on a single product class. Table 4.5 shows that 82 percent of connected firms, but only 27 percent of all firms, sell products that are protected by at least two Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 113 technical import barriers. Seventy-one percent of connected firms, but only 4 percent of all firms, sell products that are protected by at least three technical import barriers. These benefits accrued to connected firms despite the fact that, at the same time, Egypt was acclaimed for its efforts to reverse decades of state control of the economy. Politically Connected Firms Enjoy Privileged Access to Subsidized Inputs and Assets Both regimes appear to have moved away from the most visible modes of support for connected firms in the 2000s. The most direct way to subsi- dize firms connected to or owned by members of the political regimes are probably direct fiscal transfers through tax breaks or directed lending by state banks. However, these are also the most visible channels for outside observers. Both regimes aimed to attain an image of a reformer, business- friendly government in the late 1990s and 2000s partly because of reform pressure accompanying IMF programs. Obvious tax evasion or direct fiscal transfers to politically connected firms might have made it more difficult for the regimes to maintain that image potentially making them more vulnerable. We do not find evidence that fiscal advantages disproportionally ben- efitted politically connected firms in Tunisia. The Tunisian Investment Law also stipulates that firms engaging in particular activities are eligible for special fiscal incentives. While these fiscal transfers seem to be more frequent for activities conducted by Ben Ali firms, we cannot reject the null hypothesis that they are equally prevalent in sectors in which con- nected firms were active as in sectors in which they were not. Thus, in contrast to authorization requirements and FDI restrictions, the special TABLE 4.3 Politically Connected Firms and All Firms Protected by Nontariff Trade Barriers in the Arab Republic of Egypt Firms Sectors Number of class B NTMs per industry PC firms (%) All firms (%) Pearson χ2 test ( p) PC sectors (%) Non-PC sectors (%) Pearson χ2 test ( p) At least 1 82 56 0.00 76 55 0.01 At least 2 82 27 0.00 76 52 0.00 At least 3 71 4 0.00 59 38 0.01 At least 4 26 3 0.00 22 7 0.01 At least 5 18 3 0.00 15 5 0.05 At least 6 15 2 0.00 14 5 0.08 At least 7 13 0 0.00 9 3 0.09 At least 8 10 0 0.00 5 1 0.37 Source: World Bank calculation; WITS Comtrade. Note: Because of small samples, we use the Fisher test to test for the significance in differences between PC and non-PC sectors for all comparisons with more than five NTMs per industry. The statistics show the results for the broadest measure of political connections, which also include firms that received significant investments from politically connected private equity funds. 114 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa fiscal incentives appear not to be the preferred way of supporting politi- cally connected firms in Tunisia. After the financial crises in the late 1990s in Egypt, advantages of con- nected firms shifted away from directed lending by state banks to other policy areas. The most commonly documented advantage enjoyed by connected firms is access to capital. This was also the case in Egypt before the financial crisis at the end of the 1990s, when connected firms enjoyed privileged access to credit from state-owned banks. After the banking cri- sis in the late 1990s, however, policy reforms circumscribed the activities of state banks and opened the financial sector for (foreign) private banks. As indicated earlier, we still find that connected firms in Egypt absorb most bank loans. However, interviews with foreign banks operating in Egypt suggest that private banks compete to lend to these firms as they are the most profitable in the country. Thus, the concentration of bank loans among connected firms in Egypt appears to be an equilibrium out- come of a system of privileges guaranteeing higher profits for connected firms rather than a direct policy privilege. Instead, the available evidence suggests that privileges shifted to other more subtle mechanisms such as energy subsidies to industry, land deals, trade protection through NTMs, or discretion in rule enforcement. Politically connected firms in Egypt benefit disproportionally from energy subsidies. Chapter 2 documented that large establishments are more likely to benefit from the generous energy subsidies to industry in Egypt. Figure 4.3 shows that among large firms, the few politically connected ones are much more likely to operate in energy-intensive industries. That is, 45 percent of all connected establishments operate in energy-intensive industries, compared with only 8 percent of all establishments. In contrast, there is no statistical difference between the number of connected firms and all establishments operating in low or moderate energy-intensive industries. Likewise, at least one connected firm operates in 81 percent of all high energy-intensive industries. In contrast, connected firms are present in only 43 percent of low energy- intensive industries, and entirely absent in 57 percent. Firms operating in sectors with more connected firms are more likely to have access to government land. In the manufacturing sector, access to land includes access to industrial zones, which guarantee several benefits relative to competitors outside of these zones, including tax exemptions from corporate taxes or customs duties, better infrastructure, and more streamlined regulations.22 In the following, we test whether firms in sec- tors with a higher intensity of political connections in Egypt are more likely to obtain land from the government and/or be located in an indus- trial zone. To do this we employ the WBES data between 2004 and 2008, which contains information for all of these variables for about 3,000 firms Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 115 FIGURE 4.3 Share of Politically Connected Firms in High and Low Energy- Intensive Sectors in the Arab Republic of Egypt 60 Χ2 test probability = 0.000 Χ2 test probability = 0.173 50 40 Percent 30 20 10 0 High energy-intensive Low energy-intensive Politically connected firms All firms Source: World Bank calculation. Note: The difference between politically connected and all other firms is significant at the 1 percent level in high energy-intensive industries but not significant in low energy-intensive industries. The percentage of firms in medium energy-intensive sectors has been excluded. The statistics show the results for the broadest mea- sure of political connections which also include firms that received significant investments from politically connected private equity funds. in Egypt. Firm responses to the WBES are anonymous, so we cannot distinguish connected and unconnected firms directly. However, as with NTMs and energy subsidies, we can identify the detailed four-digit in- dustries in which politically connected firms are active by supplementing the WBES data with the information on the number of politically connected firms per four-digit sector.23 Descriptive statistics show that firms in politically connected sectors (i.e., with at least one connected firm) are 11–14 percent more likely to have acquired land from the gov- ernment and 7–11 percent more likely to be located in an industrial city (table 4.4). In the following, we test more systematically if politically con- nected firms benefitted disproportionally from government relations us- ing regression analysis.24 We emphasize that all results reflect the most conservative empirical tests, since we only compare differences in the impact of the intensity of political connections among firms located in the same two-digit manufacturing sector (e.g., textiles), but in different four- digit subsectors (which vary in the number of politically connected firms). We find that with each additional politically connected firm in a four- digit manufacturing sector, the probability of obtaining land from the government increases by 1.8 percentage points. Thus, assuming linearity, sectors with five connected firm owners are 9 percentage points more likely to have obtained land from the government than sectors without connected firm owners, which is a significant effect. 116 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa TABLE 4.4 Government Relations and Competition in Sectors with Politically Connected Firms versus Nonconnected Sectors in the Arab Republic of Egypt Sectors with Sectors with Sectors with politically politically any politically connected All other connected All other connected All other CEOs sectors owners sectors firm sectors Government relations Share of firms acquired land from government (%) 48 37 44 33 44 30 Share of firms in industrial city (%) 47 36 42 33 41 34 Share of firms with bank loan (%) 21 17 19 17 19 13 Waiting days for construction permit 595 642 608 681 610 696 Coefficient of variation (construction permit) 0.56 0.45 0.54 0.33 0.53 0.30 Number of tax inspections per year 4.6 5.7 5.1 5.7 5.3 5.2 Coefficient of variation (tax inspections) 1.34 1.32 1.35 1.25 1.34 1.27 Number of inspections by municipal authorities 1.6 1.9 1.7 2.0 1.6 2.5 Coefficient of variation (municipal inspections) 2.23 2.19 2.31 1.92 2.23 2.03 Share of firms’ total sales to government 21 16 19 14 19 12 Competition Share of firms <10 domestic competitors (%) 36 29 32 30 32 29 Source: World Bank calculation. Note: Data are from WBES 2004–08 and number of politically connected firms in Egypt. Politically connected four-digit sectors have at least one politically connected firm while all other sectors include zero connected firms depending on the type of political connection. The types of political connections are ranked according to their restrictiveness. The incentive of the connected individual to leverage connections on behalf of the firm is strongest if he is the CEO (almost all connected CEOs also own at least part of their companies). It is less strong for politically connected owners and weakest for any type of connected firms for which we also include firms which received significant investments from connected private equity funds. Trials of leading businessmen after the fall of Mubarak in Egypt cor- roborate our empirical finding that connected firms profited from cheap access to prime land. There is plenty of anecdotal evidence that politically connected firms in Egypt have superior access to land and credit. Reportedly, the government not only sold the land but also guaranteed to connect the land with the necessary electricity, telecommunication, and transport infrastructure; this practice immediately increased the value of land, which the businessmen used as collateral to get bank loans far exceeding the initial purchase value of the land. The past practice of selling prime land below market value in closed deals also became appar- ent in the emergence of numerous court disputes filed against major real estate developers after the regime change in 2011. These trials aimed to force these real estate firms to revalue past land deals with the state and pay the difference. Several of these disputes have been settled outside courts in recent months (Ahram Online, various issues). Large firms are more likely to be located in an industrial zone if they operate in politically connected industries (containing a higher number of connected firms). Figure 4.4, panel a, illustrates how the probability that a large firm with at least 100 employees is located in an industrial zone increases with the number of firms managed by a politically Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 117 FIGURE 4.4 Large Firms in Politically Connected Industries Are More Likely to Be Located in an Industrial Zone a. Large firms only b. Large vs. small firms Probability to operate in an industrial zone Probability to operate in an industrial zone 1.0 0.8 5% CI 0.8 0.6 0.6 0.4 0.4 95% CI 0.2 0.2 0 1 2 3 0 1 2 3 Number of firms with connected CEOs Number of firms with connected CEOs Large firms Small firms Source: World Bank calculation. Note: Data are from WBES 2004–08 and number of politically connected firms in Egypt. Large firms have at least 100 employees. The graph illustrates how the probability that a large firm with at least 100 employees operates in an industrial zone increases ‘the number of firms with a politically con- nected CEO across four-digit industries. The number of firms with politically connected CEO across four-digit industries in the sample ranges from 0 to 3. It is based on a probit regression of a dummy variable if the firm is located in an industrial zone and the number of firms with a politically con- nected CEO within a four-digit sector. The regression controls for firm level size, age, export shares, and two-digit sector dummies. We also include interaction terms between firm size categories (small versus large) and the number of connected firm per four-digit sector. connected CEO across four-digit industries. Note that this result is likely to be driven by the connected firms in these sectors, since large firms are generally much more likely to be politically connected.25 It shows that approximately 41 percent of all large firms in four-digit sectors without connected firms operate in industrial zones. This share increases to about 58 (respectively 62 percent) in sectors with one firm (respectively three firms) managed by a politically connected CEO. Among politically connected industries, large firms are more likely to be located in an industrial zone than small firms. Figure 4.4b illustrates how the probabilities that large and small firms operate in an industrial zone increases with the number of firms led by a politically connected CEO across four-digit industries. Given that the majority of connected firms in our sample are large, the results strongly indicate that it is the connected firms within four-digit sectors that are located in industrial zones. Reportedly, the identification of activities benefitting from tax exemp- tions in special economic zones was also driven by vested interests; for example, the list of sectors eligible for tax exemptions was expanded to include media companies after the construction of a new media complex (including the media company’s offices, hotels, theatres, and so forth) of a politically connected businessman. The complex was declared a special economic zone shortly after, allowing him to benefit from tax exemptions (Ahram Online, various issues). 118 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Politically Connected Firms Benefit from Discretionary Policy Implementation Connected firms disproportionately benefit from the enforcement of rules. Politically connected firms also used their connections to minimize their regulatory burden and the threat of predatory behavior by govern- ment officials, relative to the burden and threats faced by their competi- tors. To analyze this situation, we again employ the WBES data, which contains firms’ assessments of the implementation of various government policies and regulations. Following Hallward-Driemeier et al. (2010), we also examine within-industry variations of firm reports regarding the regulatory environment. Descriptive statistics show that firms in politi- cally connected sectors (i.e., with at least one connected firm) report much lower waiting times for construction permits. For example, for the most conservative measure of political connections, table 4.4 shows that firms in connected sectors wait on average 47 days less.26 In the following, we test more systematically if politically connected firms benefitted disproportionally from discretion in rule enforcement using regression analysis.27 The data show that, for the most conservative measure of political connections, an additional firm with a politically connected CEO reduces the average waiting time in a four-digit sector by 51 days. Furthermore, large firms in industries that are less/not connected have to wait substantially longer (between 11 and 48 days, depending on the type of connection) than large firms in sectors with more politically connected firms. Given that politically connected firms are much more likely to be large relative to the average firm in the WBES, the finding suggests that connected firms have access to fast-track enforcements relative to other large firms in the same two-digit (but different four-digit) manufacturing sector. The data indicate that sectors with more politically connected firms exhibit a significantly higher coefficient of variation in the waiting days for construction permits, consistent with the argument that con- nected firms are able to access fast-track regulatory services while uncon- nected firms in the same four-digit industry are not. Discretionary enforcement of rules can also be used to keep out or weaken potential competitors. For example a businessman, that pursued new investment abroad, recalls his encounter with a prominent local entrepreneur (World Bank 2009): “As I was going through the investment process in that country and had already transferred the initial capital, I was contacted by a local entrepre- neur whom I knew was close to the country’s leadership. He offered to take part in my venture with a 25 percent share, bringing in a free land plot . . . and assurances that the investment would proceed smoothly with ‘no administrative hassle’. I knew what that meant, of course, and the risk it Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 119 involved for my control of the enterprise. Fortunately, I knew of ongoing investment in my own country. I made him understand that we would both gain in our respective countries to have our investments proceed smoothly, but also by staying away from each other’s businesses as I could also make things difficult for him in my country, thanks to my own connections.” Firms that are not politically connected appear to be more frequently targeted by government inspections. Reportedly, using political connec- tions to increase the number of inspections by government officials for a direct competitor is a mechanism to prevent a potential competitor from growing. The WBES for Egypt contain information on the number of tax and other inspections. Table 4.4 shows that, on average, firms in sectors with at least one politically connected CEO are inspected by tax officials 4.6 times a year. In contrast, the frequency of tax inspections increases by 24 percent (to 5.7 times a year) for firms in sectors without a connected CEO. Similarly, the frequency of inspections by the municipality is about 20 percent higher in sectors without politically connected firms. What is more, the dispersion (coefficient of variation) of reported inspections across firms is significantly higher in the connected sector (table 4.4). Thus, some firms received very few inspections while others are inspected frequently when connected firms are present in the sector. While we do not directly observe if connected firms report very few inspections, it is likely that the nonconnected firms are the ones targeted by government officials more frequently. The finding is also consistent with the notion that nonconnected small firms or firms in the informal sector stay small (under the radar of large connected competitors) in order to avoid being targeted by anti-competitive actions or government scrutiny. BOX 4.1 Did Ben Ali Firms Dictate Amendments to the Investment Law in the 2000s? The establishment of new entry barriers was NAT 96 level, which is the five-digit sector more likely in sectors hosting Ben Ali firms. level. Table B4.1.1 summarizes changes made The list of activities which are subject to to the Tunisian investment code between authorization requirements and FDI restric- 1994 and 2010 through 22 decrees issued by tions changed since 1993; they were supple- Ben Ali himself. These decrees introduced mented by 22 subsequent presidential new authorization requirements pertaining decrees, resulting in 73 amendments at the to 45 sectors and new FDI restrictions in (continued on next page) 120 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa BOX 4.1 Continued TABLE B4.1.1 Correlation between New Barriers to Entry and the Presence of Ben Ali Firms New regulations and presence of Ben Ali firms Ben Ali presence New Authorization requirements New FDI restrictions N n % Fisher’s F-test n % Fisher’s F-test At least one firm 451 7 1.55 Table Pr 0.0195 9 2.00 Table Pr 0.0195 None 5058 38 0.75 0.046 19 0.38 0.000 All 5509 45 p = 0.0961 28 p = 0.000 New regulations and entry of Ben Ali firms Ben Ali entry in the same or the subsequent year New Authorization requirements New FDI restrictions N n % Fisher’s F-test n % Fisher’s F-test At least one entry 168 4 2.35 Table Pr 0.0195 3 1.76 Table Pr 0.0195 None 5031 41 0.82 0.043 25 0.50 0.049 All 5199 45 p = 0.0582 28 p = 0.0619 Source: World Bank calculation. Note: The test for equality is Fisher’s exact t test. It tests the null hypothesis that the introduction of new regula- tions referred to in the column heading pertaining to narrowly defined five-digit sectors is independent of the presence (top row) and start-up (bottom row) of connected firms within such sectors. The entry indicator is a sector-level binary indicator taking the value 1 if a Ben Ali firm entered in the same or following year. 28 sectors. Table B4.1.1 (upper panel) shows manipulated by the Ben Ali family. For that connected firms were present in seven example, Décret n° 96-1234, issued in 1996, (nine) of the 45 (28) sector-years in which amended the investment code by introduc- new authorization requirements (FDI restric- ing authorization requirements for firms tions) were imposed. The null hypotheses engaging in the handling and transfer of that the likelihood of new FDI restrictions goods in ports and the towing and rescue of (authorization requirements) does not depend ships. The decree also introduced restric- on the presence of connected firms is rejected tions on FDI for firms involved in the at the 1 percent (10 percent) significance transport of red meat. In the same year, a level. While the number of observations is shipping and logistics company focused on again small, the data also reject the null the transport of refrigerated products was hypothesis of independence between the established by a member of the Ben Ali startup of new Ben Ali firms and the intro- family. Moreover, immediately after the duction of new authorization requirements entry of a politically connected firm into and FDI restrictions at the 10 percent signifi- the cement sector, Décret n° 2007–2311 was cance level (table B4.1.1, lower panel). introduced stipulating that government au- A few anecdotal clues support the view thorization was required for firms produc- that the investment code has been actively ing cement. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 121 Privileges to Politically Connected Firms Undermine Private Sector Development and Job Creation The theory developed in Aghion et al. (2001) points to an indirect empiri- cal strategy for assessing whether the advantages of political connections constitute a drag on growth. First, if political connections are a drag on growth, it must be the case that the policy privileges of the politically connected firms drive a wedge between the prices of inputs and outputs that they face compared with the prices encountered by unconnected firms. If this is the case, the policy privileges that connected firms receive should account for their better performance relative to unconnected firms. The subsequent evidence shows that this is the case. Policy Privileges Explain the Superior Profits of Connected Firms Connected firms in Egypt are more profitable because they benefit more from trade protection and energy subsidies. The joint distribution of NTMs, energy subsidies, and politically connected firms across four-digit industries accounts for the entire profitability differential between con- nected and other firms. That is, politically connected firms are signifi- cantly more profitable than unconnected firms if their products are protected from import competition, but are not so otherwise. We find similar results once we account for the joint distribution between political connections and energy subsidies in high energy-intensive industries.28 These results indicate that nontariff barriers and energy subsidies are targeted to connected firms. These barriers and subsidies appear to exclude unconnected firms operating in the same sectors. For example, some barriers to entry limit the ability of unconnected domestic firms to benefit from the privileges granted to connected firms. In the case of energy subsidies, firms are required to obtain a government license to build a factory in energy-intensive sectors such as steel and cement. This license was issued by the Ministry of Industry and Trade, or the Ministry of Investment, and had to be renewed annually. The licensing procedure favored politically connected firms, which were both more likely to get the license and less likely to be exposed to predatory behavior (the non- renewal of a license after they had undertaken large sunk investments). In the very profitable energy-intensive and trade-protected cement and steel sectors, by 2010 only a few connected firms had obtained the license guaranteeing access to energy subsidies. In the case of NTMs, some of these measures also required explicit licenses to import specific interme- diates from foreign manufacturers (as in the automobile industry). Table 4.6 shows that connected firms are significantly more likely to benefit from authorization requirements for importing. Moreover, enforcement of NTMs requires government action, which has been 122 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa shown to be uneven across firms operating in the same sector when con- nected firms are present.29 Connected firms in Tunisia are more profitable because legal barriers to entry guaranteed their market power. Ben Ali firms only have higher market shares and value added per worker if they are protected by autho- rization requirements and FDI restrictions.30 In sectors covered by the Investment Code but not subject to these regulatory requirements, the differences are statistically negligible once the larger size of connected firms is taken into account. On average, the market shares of Ben Ali firms exceed that of nonconnected firms in sectors with authorization requirements and FDI restrictions by 4 percentage points and 6.4 percentage points, respectively. These are sizeable differences consid- ering that the average market share of nonconnected firms in sectors subject to authorization requirements is only 0.27 percent. Notably, Ben Ali firms are also significantly larger in sectors with entry restrictions. Ben Ali firms employ 137 percent and 285 percent more salaried employ- ees than nonconnected firms when authorization requirements or FDI restrictions are present, respectively. Moreover, we find that the growth differences in these variables between Ben Ali and other firms also fluctu- ate systematically with the prevalence of regulations. Business regulations helped generating higher profits for Ben Ali firms. Ben Ali firms are especially more profitable than their peers in sec- tors subject to authorization requirements and FDI restrictions. In sec- tors not subject to these restrictions, however, Ben Ali firms make significantly lower profits than their competitors. These results suggest regulatory capture by connected firms. Given our findings that political connections in Egypt and Tunisia translate into large policy privileges, we also expect to find that the presence of connected firms affects competition and firm dynamics as predicted in Aghion et al. (2001). Sectors including politically connected firms should see less firm entry and weaker competition among firms. Likewise, sectors dominated by these firms should have a more skewed firm distribution, characterized by a large connected market leader and a potentially large number of small or informal micro firms using vintage technologies to serve local market niches. In the following, we use our newly constructed data set for Egypt to present empirical evidence con- sistent with these predictions. Politically Connected Firms Are Insulated from Competition Large firms in connected sectors—those with more connected firms— report fewer domestic competitors. The analysis is based on approxi- mately 3,000 firms from the WBES data for Egypt which report their number of domestic competitors. Descriptive statistics show that firms in Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 123 politically connected sectors (i.e., with at least one connected firm) are more likely to report fewer than 10 competitors in the domestic market (table 4.4). In the following, we test more systematically if firms in politi- cally connected sectors report less competition using regression analy- sis.31 The information is observed at the firm-level, allowing us to test for complementarities between the effect of political connections at the four- digit sector level and specific characteristics of firms in these sectors, such as their size. This is important because large firms in the WBES data are much more likely to be connected. Thus, the data make possible measur- ing the intensity of domestic competition faced by large firms in con- nected sectors relative to other large firms in less/nonconnected sectors. In other words, the competition results are much more likely to be driven by the politically connected firms (even though we cannot identify them directly in the WBES) when we focus on the subgroup of large firms across all sectors. The findings confirm that large firms report fewer domestic competitors when they operate in more connected manufactur- ing sectors. Moreover, within more connected sectors, large firms are more likely than small firms to report fewer domestic competitors. Taken together, large firms in connected sectors report facing fewer domestic competitors. In sum, the findings suggest that connected manufacturing firms are more likely to be protected from domestic competition than other large firms. Privileges to Politically Connected Firms Suppress the Firm Dynamics Associated with Job Creation Low rates of entry in sectors dominated by connected firms—despite higher rents in these sectors—are further evidence that connected firms benefit from barriers to entry. We expect to find that the presence of politically connected firms discourages the entry of unconnected firms, as the latter cannot compete with the connected firms’ privileges. Thus, unconnected firms would have to specialize in unproductive local market niches in these sectors. While the counterfactual of firm entry in the absence of connected firms in the same sectors is not observable, our empirical strategy is to compare firm dynamics across detailed four-digit sectors, which differ in their intensity of political connections in a given year and over time. The cross-sector comparison can be biased because of an endogenous selection effect of connected firms into sectors with spe- cific characteristics, such as growth opportunities associated with their maturity. The findings in the previous sections help us assess the potential direction of such bias. First, the sizeable rents from energy subsidies, trade protection, and the use of prime land should attract substantial entry into these sectors, implying that the observed correlation between politi- cal connections and firm entry is biased downward. Second, the analysis 124 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa has shown that the presence of connected firms is relatively broad-based across economic activities; including manufacturing and modern service sectors (machinery and ICT services) with arguably higher sector-specific growth opportunities (table 4.1 and table G.2 in appendix G). Thus, we argue that there is sufficient variation in the distribution of connected firms across sectors with high- and low-growth potential in Egypt in the 2000s to detect whether firm dynamics vary across sectors depending on the presence of connected firms. In addition, we control for sector-specific characteristics that are correlated with sectors’ growth opportunities in all estimation specifications (for example, average size and age of establish- ments in a sector and sector dummies). Thus, we only use four-digit sectors with comparable characteristics to empirically identify the impact of political connections on establishment entry.32 Firm entry is lower in connected sectors. Table 4.5 summarizes the descriptive statistics of selected variables from the establishment census across four-digit sectors with at least one politically connected firm and all other sectors with zero connected firms. Entry rates into sectors with at least one politically connected firm were 0.8 percentage points lower in 2006 (based on the most restrictive measure of sectors with politically connected CEOs); this corresponds to 11 percent lower entry rates in connected sectors. The difference is even larger for employment weighted entry rates which are 28 percent lower in connected sectors. Moreover, connected sectors had a higher share of old establishments in 2006 pointing to either lower firm entry or exit in previous years. As discussed in chapter 3 there is evidence that the extent of privileges to politically connected firms increased between 1996 and 2006.33 Thus, if privileges to politically connected firms discourage firm entry (of unconnected firms), we expect declining firm entry rates in connected sectors between 1996 and 2006. Table 4.5 shows that this was indeed the case. Entry rates into unconnected sectors increased significantly between 1996 and 2006 but hardly changed in sectors with at least one connected firms over the same period. Likewise, the share of young firms increased more rapidly in connected sectors without connected firms. The results are robust when controlling for sector-specific char- acteristics (e.g., average size and age of establishments in a sector as well as one- or two-digit sector dummies). For example, an increase in the number of firms with a connected CEO in a four-digit sector from zero to one increases the share of old establishments in that sector by 1.7 percentage points after controlling for average firm size and two-digit sector dummies. Thus, either entry or exit between 1996 and 2006 has been significantly lower in politically connected four-digit subsectors relative to not/less connected four-digit subsectors belonging to the same two-digit sector. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 125 The findings suggest that the presence of politically connected firms crowds out the type of firms that have the highest potential for job creation. The presence of connected firms appears to discourage new (unconnected) entrepreneurs to enter as they cannot compete with the connected firms’ privileges. Chapter 1 suggests that this decline in the share of young firms reduces job growth. The presence of political connected firms tends to push the majority of unconnected firms towards unproductive small-scale, potentially infor- mal activities. Table 4.5 shows that the coefficient of variation (standard deviation divided by the mean) and skewness in the establishment size distribution are almost twice as high and 50 percent higher in sectors with at least one politically connected firm, respectively. Both measures also increased substantially in politically connected sectors between 1996 and 2006 but hardly changed or even declined in unconnected sectors. Note that a higher coefficient of variation implies fewer medium-size establish- ments since either the share of micro or of large establishments increased (or both); given that the distribution of employment across establish- ments is right-skewed, i.e., characterized by many micro and few large establishments, a higher skewness in the establishment size distribution implies that the employment share of micro establishment increased or the employment share of large establishment declines. Taken together, the simultaneous increase in the coefficient of variation and the skewness TABLE 4.5 Firm Dynamics in Sectors with Politically Connected Firms versus Nonconnected Sectors in the Arab Republic of Egypt Sectors with Sectors with Sectors with any politically All other politically All other politically All other connected CEOs sectors connected owners sectors connected firm sectors Level effects, 2006 Entry rate (%) 6.5 7.3 7.0 7.4 6.9 7.6 Entry rate, employment weighted (%) 3.6 4.6 4.4 4.5 4.2 4.8 Share old establishments (age 11–30 years) (%) 26.1 24.0 24.5 24.1 25.3 23.2 Coefficient of variation (empl) 2.6 1.5 2.0 1.4 2.0 1.3 Skewness (empl) 8.9 5.8 7.3 5.4 7.4 5.0 Dynamic effects, 1996–2006 Growth entry rate (decade) 0.1 1.2 0.2 2.3 0.2 2.9 Growth entry rate empl-weighted (decade) 0.4 3.6 2.0 4.1 2.0 5.0 Change share young establishments (age ≤10) 5.7 9.1 7.5 9.2 7.4 13.5 Change coefficient of variation (empl) 2.7 0.5 1.8 −0.3 1.7 −0.4 Change skewness (empl) 7.5 3.3 6.2 0.9 6.1 0.5 Source: World Bank calculation. Note: Data are based on the Egyptian establishment census in 1996 and 2006 and number of politically connected firms. Politically connected four-digit sectors have at least one politically connected firm while all other sectors include zero connected firms depending on the type of political connection. The incentive of the connected individual to leverage connections on behalf of the firm is strongest if he is the CEO (almost all connected CEOs also own at least part of their companies). It is less strong for politically connected owners and weakest for any type of connected firms for which we also include firms which received significant investments from connected private equity funds. 126 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa in a four-digit sector thus indicate that the employment share of micro establishments increased in connected sectors while the employment shares of medium and large establishments declined; this is consistent with the predictions of Aghion et al. (2001). Since most micro enterprises in Egypt are informal, the presence of political connections appears to push the majority of unconnected firms towards informal activities.34 These findings suggest that unconnected firms are not able to compete with politically connected firms in the same sector because they do not receive the same policy privileges. Instead, unconnected firms in these sectors are forced to cater to local market niches involving typically small- scale, potentially informal activities. If these activities are also less productive the result signals a higher misallocation of labor across firms in political connections sectors. In that case, the dynamic impact of privi- leges to politically connected firms on the firm size distribution comes with a loss in aggregate productivity, because of a less efficient allocation of resources. Although comparable evidence is not available for Tunisia because of data limitations, the Schumpeterian growth framework suggests that major policy privileges, such as those granted in the form of entry bar- riers, also distort competition and firm dynamics in Tunisia. Figures 1.19 and 1.12 documented that firm turnover in Tunisia is low and job cre- ation is skewed towards small-scale, unproductive activities especially in the service sectors. Both stylized facts are consistent with the predic- tions of the adopted Schumpeterian growth framework; i.e., they are symptoms of a lack of private sector competition. That is, Aghion et al. (2001) predict that the large cost advantages of Ben Ali firms resulting from the biased legislation limit neck-and-neck competition among firms, reducing their incentives to adopt new (foreign) technologies. All together, the findings in this section suggest that at least in part the distortions to firm dynamics and competition in Tunisia documented in chapter 1 originate from legislative barriers to entry that benefitted a few connected firms. Entry Barriers in Backbone Services Has Likely Limited Growth in Downstream Manufacturing Industries Barriers to entry and competition are expected to have reduced the quality of services provided by the few firms authorized to operate in these sectors in Tunisia. The entry barriers translate into sizeable cost advantages for the few connected firms authorized to operate in these sectors. They lead to a monopolistic market structure that helps the few connected firms shielded from competition to achieve abnormally high profits. Aghion et al. (2001) show that the resulting market structure discourages the incentives of market leaders to improve the quality of Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 127 their services; hence, it is expected to reduce the aggregate service sector performance. While services are an important part of the economy, the results do not directly explain why the generous tax breaks provided to manufactur- ing firms in the offshore economy in Tunisia did not generate more growth and jobs. While the Schumpetarian growth framework explains the distorted dynamics and firm performances in protected service sec- tors, it falls short of explaining the modest productivity and job growth of manufacturing firms in the offshore economy. Manufacturing firms in the offshore economy benefitted from generous tax incentives. For example, the investment code stipulates that offshore firms—those that export at least 70 percent of their output (Articles 10 and 16 of the code)— do not have to pay profit and turnover taxes. Moreover, they usually did not have to compete directly with Ben Ali firms. The tax incentives have helped Tunisia attract foreign investors in spite of the onshore sector being highly protected and largely closed to foreign competition, as sub- sequently discussed. The protection of Ben Ali firms from competition in Tunisia’s onshore economy is likely to have reduced the quality of backbone services provided to downstream manufacturing firms, limiting their growth.35 The theory of weak links (Jones 2011; Kremer 1993) highlights that the performance of manufacturing firms cannot be analyzed in isolation from the performance of nontradable service sectors. Weak performances in backbone service sectors lead to lower quality services provided to firms in downstream using industries. Hence, despite the generous tax regime, productivity and job growth in the downstream manufacturing industries that use these lower quality services can be limited. In fact, the results in chapter 2 show that FDI in services led to significant jobs spillovers in downstream using sec- tors in Jordan. Given the sizeable entry barriers in backbone service sectors in Tunisia because of the presence of Ben Ali firms—which primarily oper- ated in service sectors—we would anticipate the potential impact of weak links in Tunisia to be significant. The recent work of Marotta, Ugarte, and Baghdadi (2014) supports this hypothesis showing that weak links led to lower levels of productivity per worker in Tunisia. The Presence of Connected Firms Reduces Aggregate Job Creation The findings thus far provide ample indirect evidence that privileges lead to firm dynamics associated with lower aggregate job growth. All of these findings—the higher profitability of connected firms due to granted pol- icy privileges and the adverse impact of their presence on competition, entry, and employment in medium and large firms—are consistent with the empirical hypotheses derived from the Schumpeterian growth model of Aghion et al. (2001). They suggest that aggregate employment growth 128 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa would have been higher if the intensity of political privileges declined. This would necessitate a decline in the intensive margin, measured by the number of firms with a strong political influence within sectors, and the extensive margin, measured by the expansion of politically connected firms into new, initially unconnected sectors. We cannot observe directly if employment growth in connected sec- tors would have been higher in the absence of political connections. Employment growth in politically connected sectors between 1996 and 2006 was comparable to other sectors. Thus, if connected firms indeed have positive employment growth, the effect is offset by the negative employment growth of unconnected firms in these sectors. Still, drawing conclusion from directly comparing employment growth among con- nected and unconnected sectors has limitations since we do not observe simultaneous changes in various other determinants of employment growth in our data. Instead, we would like to measure to what extent employment growth in connected sectors would have been higher in the absence of politically connected firms. This relevant counterfactual is of course not directly observable. The nature of our data, however, provides a quasi-experimental set- ting which allows determining the aggregate employment impact of the entry of politically connected firms into new, previously unconnected sectors. We do observe the year in which politically connected firms entered into new sectors. Therefore, we can observe when connected firms enter into sectors which were previously unconnected. There are 41 such sectors: 18 service sectors, 16 manufacturing, 8 utilities, and 4 mining sectors. These include several sectors with high growth poten- tial in Egypt, such as manufacture of primary cells and batteries, television and radio receivers, wholesale of solid, liquid and gaseous fuels, inland water trans- port, legal activities, and advertising. We test whether aggregate employment growth over a 10-year period between 1996 and 2006 declined after the entry of politically connected firms into new, previously unconnected (open) sectors. Holding all else constant, entry always increases employment in the sector regardless of whether the entrant is connected or not. Thus, we expect that the entry of connected firms leads to sector employment growth, unless the adverse impact of connected firms on the growth opportunities of their uncon- nected peers leads to their exit or shrinkage. In contrast, we do not expect to observe the latter adverse effect (or at least expect it to be less pro- nounced) when connected firms enter into sectors which were already dominated by privileged connected firms in previous years. Therefore, negative aggregate employment growth after the entry of connected firms into previously unconnected sectors implies that the decline in employment in unconnected firms—which cannot compete—outweighs Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 129 BOX 4.2 Political Connections and Patronage in the Republic of Yemen Large government spending on mainte- water-consuming narcotic leaf, Qat. Second, nance of oil infrastructure has benefitted a the dominance in the food import market of small group of Yemeni companies and a small number of private and public players individuals.a Revenues from oil exports sub- with ties to the regime of former President stantially contributed to economic growth Saleh. Reforming the water sector has and imported goods subsidies in the years proved to be extremely difficult as the direct before 2011. However, the government beneficiaries of Qat production are the Saleh spent US$ billions over the past three family and other landowners with significant decades on maintenance of oil infrastructure stakes in the political regime. Moreover, the because of the cost-inflation actions of a Republic of Yemen has to import nearly all small, well-connected business elite. These of its wheat and rice, the two most important elites are local intermediaries that connect staples of national diet. Major importers are foreign oil companies with local govern- a military-run firm (Yeco), and three private ments; they have either close ties to former entities of which former President Saleh is President Saleh and his subordinates, or to a shareholder. These few connected firms powerful tribal sheikhs. Owners of domi- reportedly influence the regulations in the nant oil-related service providers in engi- sector to their own favor. neering and construction, transport and The lucrative telecommunications sector logistics, facilitation, and security sectors are in the Republic of Yemen has been beset relatives of, or closely connected to, the for- with government monopoly, privileges to mer president, military generals, and minis- politically connected firms, and opacity since ters. The most lucrative aspects of the energy market liberalization began in the 1990s. sector are oil exports and fuel imports, which The state-run public telecommunication in turn are controlled by powerful persons corporation (PTC) has been the sole pro- including the former president, sheikhs, and vider of broadband Internet in the country. military commanders. Their behavior leads While the telecom market appears to be to inflated production costs, lost revenues, competitive when looking at the market diesel smuggling, and likely diminish the shares of three major private and public multiplier effect of investment in the operators, most entrants in the sector were sector. linked with the former president’s family and A handful of firms connected to the mili- his close connections. The first two private tary or the former president control the mobile licenses were granted to business production of water-consuming Qat and the groups supported by, or financially and lucrative food import market. Insecure food personally connected to, former President and water supplies are chronic issues in the Saleh. The third and last private mobile Republic of Yemen. Two problems worsen license was awarded to a company whose the situation. First, the production of a ultimate owner remains opaque and is widely (continued on next page) 130 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa BOX 4.2 Continued seen as an attempt by the Saleh family to take system was small, underdeveloped, poorly a stake in the lucrative telecom market. regulated, and limited to a small group of Furthermore, former President Saleh and his elite actors, all of whom had a close relation- relatives are widely rumored to own shares in ship with former President Saleh. The bank- the sole public mobile operator, Yemen ing system is accessible only to the tiny Mobile. It has increased its share of subscrib- middle class and wealthy elites. This restrains ers as a result of substantial government the growth potential of nonconnected firms. assistance. Some of these supports include While the private sector accounts for the applying lower tariffs; privileged access to largest share of formal banking, its few major private infrastructure networks built by other financial institutions were founded by elites operators; compelling entire ministries to with strong connections to Saleh’s family. use Yemen Mobile’s services; and direct The central bank’s upper management is intervention from the former president by well-respected by international institutions, refusing import and export licenses crucial to but the bank itself is reputedly used to laun- the day-to-day business of other operators. der the profits of illicit activities. Moreover, Given the importance of the lucrative sector, it is hamstrung by poor government fiscal it remains to be seen whether the state will position and limited foreign currency allow fairer competition among current and reserves. The informal banking system, on future players in the market. the other hand, is reportedly as large as its The structure of Yemen’s financial sector formal counterpart; it serves as the source of in 2012 privileges a small group of politically microfinance, for example, for firms in food connected firms. Yemen’s formal banking production and water merchants. Note: a. The following analysis is based on a series of papers analyzing political patronage in different economic sectors in the Republic of Yemen produced by the Chatham House (2013). any positive job creation by the connected firm(s).36 We test this hypoth- esis in a difference-in-difference estimation (also controlling for other industry-specific characteristics correlated with job growth).37 We find that aggregate employment growth declines by about 1.4 percentage points annually when connected firms enter new, previously unconnected sectors. The economic impact is large. The magnitude of the corresponding coefficient suggests that aggregate employment in these sectors shrinks by 25 percent over the 10-year period from 1996 to 2006. Note that the connected firms did not necessarily enter directly in 1997, so the employment growth might have been positive in earlier years, but then declined substantially because of the sudden presence of the connected firm with access to policy privileges, itself guaranteed a large cost advantage over the existing competitors or potential new, unconnected entrants. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 131 These findings provide quantitative evidence that the growth impact of connected firms’ entry is more than offset by their adverse impact on the growth opportunities of the majority of unconnected firms that stop growing or exit. As a consequence, political connections reduce aggregate employment growth in this sector. This conclusion is consistent with the indirect evidence that political privileges lead to firm dynamics associated with lower aggregate job growth presented earlier. It is also consistent with the prediction of the model of Aghion et al. (2001), who show that less neck-and-neck competition because of large exogenous cost advan- tages of market leaders reduce aggregate long-term growth. In the case of Egypt and Tunisia, such large exogenous cost advantages are granted by policy privileges such as licenses requirements, trade protection, energy subsidies, access to prime land, or biased regulatory enforcement. Even though these policy privileges might help the few benefitting firms grow and create jobs, we show that the aggregate employment impact is nega- tive because of the adverse effects of such policies on competition, and thus on the growth opportunities for the large majority of firms, which are unconnected. Available Qualitative Evidence Points to Similar Mechanisms of Policy Privileges in Other MENA Countries The results so far show that politically connected firms in Egypt and Tunisia received large privileges that distorted competition and thus firm dynamics associated with job creation. What is more, the evidence in Tunisia suggests that firms connected to Ben Ali used their political influ- ence to affect the regulatory environment to their favor. In addition, there is direct evidence in Egypt that the presence of firms connected to Mubarak led to similar capture and also to lower aggregate job creation. This section argues that policy privileges and their adverse impact on regulations, competition, firm dynamics, and ultimately job creation are also frequent in other countries of the MENA region. We do not have comparable detailed data listing politically connected firms for other MENA countries. However, there is ample qualitative evidence from other countries in the region which we review. The section points out that the system of closed deals between the state and businesses in Egypt and Tunisia are not outliers, but rather representative of the way business is conducted in MENA. The frequent use of nontariff measures in Egypt documented in “Privileges to Politically Connected Firms Undermine Competition and Job Creation: Evidence from the Arab Republic of Egypt and Tunisia” 132 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE 4.5 Nontariff Barriers Are Frequently Imposed in MENA Resource-rich, labor-importing countries Resource-rich, labor-abundant countries Resource-poor, labor-abundant countries Sub-Saharan Africa South Asia Middle East and North Africa Latin America and Caribbean Europe and Central Asia East Asia and Pacific 0 5 10 15 20 25 30 35 40 Uniform tariff equivalent, % Nontariff barriers Tariffs Source: from Malik (2013). Note: Figure shows the average tariff equivalent value of tariffs and nontariff measures by region of the world and categories of MENA countries by level of endowment. section is representative for oil-importing MENA countries. Malik (2013) indicates that nontariff measures are more frequently used in MENA countries than in other regions and argues that they are likely exploited to protect connected firms from import competition (figure 4.5). MENA Lags Behind Other Regions in Governance and Corruption Indicators, Especially in Corruption in Defense and Military Involvement in Business The relative prevalence of the role of privileges in MENA can also be characterized through a number of qualitative governance indices, espe- cially in regards to the military sector. For example, the Transparency International (TI) Government Defense Anti-Corruption Index analyzes corruption risk in defense establishments worldwide. This index assesses and compares levels of corruption risk and vulnerability across countries. TI assessed 82 countries in 2012 and classified each country in a category38 from A to F, with A being the lowest corruption risk and F the highest. The countries included in this index accounted for 94 percent of global military expenditure in 2011. TI’s evidence suggests that poor rankings are associ- ated with patronage networks. The report found that networks based on Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 133 close family ties between the military and businesses and restrictions on pub- lic debate and civil society freedom are features of most MENA countries. All of the MENA countries assessed have high to critical risk of corruption (categories D, E, or F). Out of these 18 MENA countries, twelve were placed in category E and F, corresponding to very high or critical corrup- tion risk (33 percent of all countries); these include Egypt, Algeria, Libya, Syria, and the Republic of Yemen, along with non-MENA countries like Angola. Three were ranked D+: Kuwait, Lebanon, and UAE, along with countries like India, Israel, and Thailand (18 percent of all countries sur- veyed); and two were placed in category D-: Jordan and West Bank and Gaza, along with countries like China, Pakistan, the Russian Federation, and Turkey (18 percent of all countries surveyed). Figure 4.6 lists the remaining MENA countries according to their ranking. Patronage networks between the military and business are common features in most MENA countries. Looking at the financial risk subindex FIGURE 4.6 Transparency International: Defence Anti-Corruption Index D+ C Kuwait Lebanon United Arab Emirates Brazil India Chile Israel France Spain Jordan Norway West Bank and Gaza United States Turkey B China D– Korea, Rep. Pakistan Algeria A Egypt, Arab Rep. Bahrain Libya Iran, Islamic Rep. Syrian Arab Republic Iraq Yemen, Rep. Morocco Angola Oman F Qatar Saudi Arabia E Source: Transparency International: Government Defence Anti-Corruption Index. 134 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa of TI’s Government Defence Anti-Corruption Risk Index allows us to refine our qualitative assessment of MENA countries. Countries in the report were categorized into five risk areas: political, financial, personnel, operations, and procurement. Financial corruption risks are linked to the abuse of large, potentially secretive defense budgets and asset disposal and links to businesses. Countries were asked 5 questions (2 for asset disposal and 3 regarding links to businesses); scores were associated according to the responses.39 TI reports that military institutions’ commercial inter- ests (military ownership of businesses) creates substantial conflicts of interest and thus an increased risk of corruption. The results for MENA are summarized in table 4.6. Military-owned businesses are common in 11 MENA countries (out of 18). For example, in Jordan, TI reports that in recent years the line between business and the military has become blurred with the government’s efforts to focus more on profit-generating activities. Anecdotal evidence indicates that this closer relationship between business and military actors has not been accompanied by adequate controls. There is no evidence of military institutions owning commercial businesses at a significant scale in Morocco, Tunisia, and West Bank and Gaza (equivalent to only 1 percent of the defense budget or less). Still, there are reports of military personnel engaging in unauthorized private enterprise in Morocco. In Tunisia, while the armed forces did not appear to own businesses or engage in illicit economic activities, security forces exploited their political power to own commer- cial businesses and attain licenses and other privileges during the previous regime. Military-owned businesses exist and are lacking scrutiny in the Islamic Republic of Iran and Iraq. There is a severe lack of institutional controls to contain corruption in defense in a number of MENA countries. There is no evidence of institutional activity and transparency to prevent corruption in the dis- posal of assets for defense. Military-owned businesses40 are prevalent in each country, and are characterized by a complete lack of transpar- ency and absence of any form of oversight. For example, the military in Egypt has considerable economic interests and assets, estimated at between 10 and 40 percent of the country’s economy, according to TI. The profits of these firms are deemed “national secrets.”41 In Algeria, an anti-corruption law attempts to prohibit participation of the mili- tary in corrupt private enterprises, but this type of illicit activity is still common practice because of the lack of implementation of this law. In Syria, there is no evidence that military-owned businesses are subject to any scrutiny or auditing processes. The entire budget of the military is “off-budget.” Defense and security institutions have ownership of several commercial businesses, which are not independently scrutinized. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 135 TABLE 4.6 Financial Corruption Risk Subindex: Asset Disposal and Links to Business, MENA Countries Asset disposal Links to business Asset disposal Asset disposal Mil. owned Mil. owned business Unauthorised private Country name controls scrutiny businesses exist scrutiny enterprise Band D+ Kuwait 4 4 4 .. 2 Lebanon 0 0 2 2 2 United Arab Emirates 2 0 3 2 4 Band D− Jordan 1 2 1 1 2 West Bank and Gaza 1 2 1 1 1 Band E Bahrain 1 1 4 .. 3 Iran, Islamic Rep. 0 1 1 1 0 Iraq 0 0 0 0 1 Morocco 1 0 4 .. 1 Oman 0 1 4 .. 4 Qatar 1 0 2 0 0 Saudi Arabia 2 0 2 1 0 Tunisia 1 0 .. 4 2 Band F Algeria 0 0 0 0 2 Egypt, Arab Rep. 0 0 0 0 2 Libya 0 0 .. 0 0 Syrian Arab Republic 0 0 0 0 0 Yemen 0 0 0 0 0 Q22: How effective are controls over the disposal of assets, and is information on these disposals, and the proceeds of their sale, transparent? Q23: Is independent and transparent scrutiny of asset disposals conducted by defence establishments, and are the reports of such scrutiny publicly available? Q30: Do national defense and security institutions have beneficial ownership of commercial businesses? If so, how transparent are details of the operations and finances of such businesses? Q31: Are military-owned businesses subject to transparent independent scrutiny at a recognised international standard? Q32: Is there evidence of unauthorised private enterprise by military or other defence ministry employees? If so, what is the government’s reaction to such enterprise? Hint: Such enterprises may operate under the pretence of being part of official military activity. Source: Transparency International: Government Defence Anti-Corruption Risk Index. Note: 4 = High transparency; strong, institutionalized activity to address corruption risks, 3 = Generally high transparency; activity to address corruption risks, but with shortcomings. 2 = Moderate transparency; activity to address corruption risks with significant shortcomings. 1 = Generally low transpar- ency; weak activity to address corruption risks. 0 = Low transparency; very weak or no activity to address corruption risks. Qualitative evidence exemplifies how military businesses can use their connections to stifle competition. A former entrepreneur from a large country in the region founded a larger investing in the dairy and meat sector company in his country. After he entered the market he learned that his company would potentially compete regionally with an 136 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa incumbent business run by a military general. He describes his experience as follows: We (my father, brother, and I) decided to invest in cattle production in our country. We pooled our money together and invested about 300,000 euros to develop our cattle production since there is a supply shortage. Our live cargo arrived at port in containers and was not released to us as we were told that we were missing crucial documents for customs clearance. This was evidently a new procedure we had not heard of before. We learned that a Military General had cornered the market in that part of the country and had decided he would not sustain any competition. The amount of red tape and delay to get the approval took more than three weeks with the cattle sitting in their containers. When the paperwork finally arrived and we were cleared to take our live cargo out of the port, the cattle headcount had dropped to 15 cattle from 100 and that was the end of our business venture. BOX 4.3 The Islamic Republic of Iran: Privatizations without the Private Sector Between 2006 and 2010, Iran engaged in a firms, pension funds, state-linked investment large and wide-ranging privatization pro- and holding companies, endowed founda- gram with a goal of privatizing 80 percent of tions, and recipients of the “Justice Shares” the public sector. The program had the bless- program. Harris (2013) shows how different ings of Ali Khomeini, Iran’s leader and political economy factors have shaped the supreme jurist, who formulated the pseudo-privatization process in Iran and the 80 percent privatization goal.a By late 2009, distribution of privatized assets to various the government had divested over 800 tril- constituencies between 2006 and 2010. lion rials (about US$80 billion) in more than “Justice Shares” and the social politics of 370 state-owned enterprises (SOEs), includ- privatization. Following his election in 2005, ing petrochemical plants, fuel refineries, air- President Mahmoud Ahmadinejad lines, banks, insurance companies, announced that the SOE privatization pro- telecommunication companies, and so forth. gram, legitimated by Iran’s supreme leader However, in 2010 an Iranian parliamentary through an executive order, would move for- commission on privatization found that ward but with the benefits distributed to the among all the SOE assets divested since 2006, people via a program called “Justice Shares.” only about 13 percent of the shares went to The program was designed such that the bot- the private sector. The remainder of the tom six income deciles of the population shares was transferred to what constitutes the were eligible to buy “justice shares” of the pseudo or parastatal state, including military privatized SOEs; the bottom two income (continued on next page) Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 137 BOX 4.3 Continued deciles were able to buy shares at half their 2001, for example, the government trans- face value, and the third to sixth deciles were ferred assets worth US$400 million to the permitted to buy “justice shares” at full price SSO to cover mandated obligations to pen- (payable over 10 years). However the pro- sioners. In 2011, the SSO claimed that the gram was expanded to various groups while it fund was owed nearly US$24 billion by the was implemented; these included low-income government, pointing to a high likelihood villagers and nomads, public sector retirees, of future demands for pseudo-privatization beneficiaries of the Imam Khomeini Relief from the SSO and other pension funds. Committee and other welfare organizations, Privileged access to SOE privatization for the and families with martyr status. These groups military. The military establishment (retired represented already existing categories of and acting) benefitted largely from privi- beneficiaries within the Iranian welfare leged access to privatized SOEs from 2006 system. The Iranian Parliament Research to 2009. Harris (2013) documents how large Center found that among 264 privatized divestment scandals involving privileged SOEs initially valued at US$54 billion, access to privatization for the military made over 68 percent of shares went to “justice front pages amid post-election street dem- shares.” Harris (2013) further argues that onstrations in 2009. For example, 51 per- Ahmadinejad’s push for privatization consti- cent of the Telecommunications Company tuted a strategic component of the president’s of Iran was sold to a conglomerate linked to public relations campaign against its critics. the Islamic Revolutionary Guards Corps Pension funds and pseudo-privatization: Cooperative Foundation, a large investment Harris (2013) suggests that pension financ- company and service contractor. The auc- ing in Iran created a sizable interest group tion was limited to only two bidders, with for pseudo-privatization, namely the Iranian the second linked to the Basij (voluntary middle class and formal labor force. Fiscal militia) investment cooperative. Hence, two pressures because of an overly generous sys- military parastatals were competing for a tem pushed the Social Security Organization major share in the lucrative domestic tele- (SSO) to become more active in the acquisi- com market. The International Exposition tion of SOEs, both in the stock market and Center was also transferred to the Armed in negotiations over government debt. In Forces Social Security Organization.b Notes: a. This section is based on Harris (2013). b. Harris (2013) also documents how the engineering arm of the Islamic Revolutionary Guards Corps, Hatam al-Anbiya, (which emerged dur- ing the Iran-Iraq war and was subsequently involved in postwar reconstruction) has over the past decade (along with its subcontractors) replaced foreign firms in the development of oil and gas fields, pipeline projects and highway and tunnel construction. Public perceptions of corruption in business are strongly correlated with perceptions of government corruption in MENA. The favors exchanged between business and political elites include official bribes, illegal funding of political campaigns, and the manipulation of the financial markets for the benefit of both firm and government insiders. 138 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa These favors have sometimes also been documented in the media, influencing public opinions. figure 4.7 reveals the consequences: public perceptions of corruption in business are strongly correlated with per- ceptions of government corruption. As a result, popular perceptions about business elites became negative in the region in the years before the recent uprisings. For example, a Pew survey reveals that in 2010 corruption was the top concern of Egyptians, with 46 percent listing it as their main concern, even ahead of a lack of democracy or poor eco- nomic conditions. Also, changes in the corruption ratings of MENA countries in the overall Transparency International corruption index confirm popular perceptions. In 2005, Egypt ranked 70, Tunisia ranked 43, Libya ranked 117, and the Republic of Yemen ranked 103, out of 158 rankings on TI’s Corruption Perceptions Index (CPI). Perceived corruption increased markedly in the following three years. In 2008, Egypt dropped to 115, Tunisia to 62, Libya to 126 and the Republic of Yemen to 141, out of 180 rankings on the CPI. Governance indicators suggest that MENA lags behind other regions. The World Bank Governance Indicators measure government effective- ness, regulatory quality, the rule of law, and control of corruption. Figure 4.8 reports the relative performance of MENA countries. MENA countries are typically ranked in the bottom 40 percent worldwide in all four dimensions. FIGURE 4.7 Perceptions of Corruption in Government and Business, Middle East and North Africa, 2011 1 Lebanon Iran, Islamic Rep. Morocco Tunisia Egypt, Arab Rep. 0.8 Palestine Yemen Syria Corruption in business Jordan Algeria 0.6 Bahrain Saudia Arabia 0.4 0.2 Qatar 0.2 0.4 0.6 0.8 1 Corruption in government Source: Transparency International Corruption Perception Index; in Diwan (2012). Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 139 FIGURE 4.8 Worldwide Governance Indicators Libya Syrian Arab Republic Iran Egypt, Arab Rep. West Bank and Gaza Morocco Jordan Norway 0 20 40 60 80 100 Percentile rank Control of corruption Regulatory quality Government effectiveness Source: World Bank’s Worldwide Governance Indicators 2012. Note: Normalized to 1–100. What Explains the Different Outcomes in MENA and East Asia? The Extent to Which Political Connections Hampered Competition Differed in MENA and East Asia The analysis suggests that privileges limit job creation in MENA. The report provides novel empirical evidence on how business regulations in MENA countries are distorted to protect the interests of a few politically connected firms. The results further suggest that these political privileges tend to reduce competition and job creation. However, the occurrence of politically connected firms is not specific to MENA economies. There is also evidence that politically connected 140 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa firms were common among East Asian countries at the time when their economies started to grow at double-digit rates (see the discussion on the Republic of Korea in chapter 3). What is more, the governance frame- work of East Asian countries at the time appears to be comparable to governance levels among MENA countries. How can we explain the dif- ferent experiences of these two regions? A comprehensive answer to this important question is beyond the scope of this report. Nevertheless, the theoretical and empirical framework employed in this chapter point to potential explanations. The extent to which political connections hampered competition differed in both regions. Chapter 3 provides two different arguments that politically connected firms in East Asia indeed faced more competitive pressures, forcing them to become more cost efficient and grow. First, there is evidence that political connections were not sufficient for East Asian firms to escape competition. The previous analysis has shown that politically connected firms in Egypt and Tunisia were able to trans- form their connections into firm-specific privileges. They found ways to exclude their competitors from access to these privileges and made higher profits. Chapter 3, however, suggests that government support in the form of subsidies, credit, and other means in East Asian countries was granted at the industry rather than the firm level. Thus, politically connected firms still encountered higher domestic competition and higher firm entry into their sectors once they made high profits (Aghion et al. 2012). Second, Asian countries credibly linked privileges to performance tar- gets; even those that benefitted insiders and cronies. Chapter 3 docu- ments that in Korea a few large businesses families controlled large parts of the economy. These families were often also politically connected through family members in high positions in the ruling party or the bureaucracy (Kang, p. 189). “Privileges to Politically Connected Firms Undermine Competition and Job Creation: Evidence from the Arab Republic of Egypt and Tunisia” section reveals similar structures in Mubarak’s Egypt.42 Nevertheless, chapter 3 provides evidence that in East Asia politically connected firms still had to meet performance (export) targets to continue to benefit from industrial policies. Taiwan, China, provides an example of the enforcement of perfor- mance targets in East Asia. It conditioned its sector subsidies on perfor- mance criteria, such as export growth, and performance was regularly reviewed. One target sector, the video industry, fell slightly below its tar- get growth and the government withdrew support. As a result, three large firms went bankrupt and in contrast to other East Asian countries, the industry never developed in Taiwan, China. However, the example sent a clear signal to firms in all other sectors that benefitted from industrial policy support. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 141 East Asia’s export orientation exposed firms to competition in highly contested global markets. Even if politically connected firms faced only a few domestic competitors, which was the case initially in Korea, they had to meet credible performance targets to continue to benefit from govern- ment support. To a certain extent, this policy offset the initial lack of domestic competition. In other words, East Asian governments imported competition through their focus on exports. The destination of exports may also have mattered. East Asian firms targeted highly contested export markets in the US and EU. In contrast, manufacturing exporters in the Middle East often target local market niches in other Middle Eastern or African markets, which are typically less contested. For example, pharma- ceutical companies in Jordan are the only foreign firms that are allowed to sell medicines in Algeria. In other words, sector-specific policies in East Asia tended to offset gov- ernance challenges whereas in MENA sector-specific policies may have reinforced those challenges. Thus, while the overall governance framework was comparable in both regions, there is evidence that East Asian countries designed industrial policy to mitigate policy distortions in the few targeted sectors, while firm-directed industrial policies worsened policy distortions in MENA. Moreover, chapter 3 argues that the costs of catering to vested interests for government officials were higher in East Asia because bureau- crats were committed to, and benefitted directly from economic growth. In a Schumpeterian world, the impact of privileges to politically con- nected firms on growth also depends on countries’ barriers to innovation. The process of foreign technology adoption is costly and risky. Therefore, firms are likely to use cheaper options to escape competition if they exist. Political connections provide such an option. More specifically, in the Schumpeterian growth framework, firms are more likely to use their con- nections if the expected costs of seeking policy protection are lower than the costs of innovating. The argument essentially indicates that the costs to lobby for policy protection were higher in East Asian countries because of their industrial policy design and complementary reforms of the public sector. At the same time, however, firms are also more likely to rely on their political connections to escape competition if they face higher bar- riers to innovate. Thus, for any given level of governance, growth in a country is more likely to suffer from privileges if firms’ costs to innovate are higher. The adopted Schumpeterian growth framework predicts that, among two regions (such as MENA and East Asia) with the same level of governance, the adverse impact of privileges on growth is stronger in the region where firms’ face higher costs to innovate. Given a higher regula- tory burden for firms to innovate and MENA countries’ weaker integra- tion into global markets (through trade or FDI), we should expect higher costs for MENA firms. 142 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Notes 1. The methodologies and additional country-specific analysis are described in detail in the corresponding companion papers of this report including Diwan, Keefer, and Schiffbauer (2014); Rijkers, Freund, and Nucifora (2014). 2. See chapter 3 for a detailed review of Egypt industrial policy program at the time. 3. The Egyptian military implicitly or explicitly agreed on all government land sales, as they had a de facto veto right to any land deal. The Egyptian Minister of Defense can intervene to block a land deal—especially in coastal areas—if the land is considered strategically important by the military. 4. The group of Ben Ali firms is highly. While three connected firms feature in the list of the 10 largest firms in Tunisia, 100 connected firms did not report using any paid laborers at any point in time. Some such firms may have served as shell companies for money laundering or to benefit from tax breaks. 5. Out of the 32 PC businessmen, 18 had high political posts after 2002 (either in the ruling party or in the government) and controlled 307 of the 469 firms we ultimately identified as connected. Among the other 14 businessmen, the most important ones are long-term friends of Hosni Mubarak from his military period or cofounders of a large investment bank partly owned by a Cyprus registered company said to be owned by the Mubarak family. 6. Many large firms were listed at stock exchanges in Egypt, since gains from selling shares of listed companies are exempted from taxation. Reportedly, several politically connected firms exploited this legal tax loophole to avoid paying taxes for takeovers; that is, instead of selling firms directly, which is taxable, the transaction was conducted as an untaxed market transaction by first listing the company for sale at the stock exchange (Ahram Online, vari- ous issues). 7. Note that these types of political connections can be ranked according to their restrictiveness. The incentive of the connected individual to leverage connections on behalf of the firm is strongest if he is the CEO of the com- pany (almost all politically connected CEOs also own at least part of their companies). It is less strong for politically connected owners and weakest for any type of connected firms for which we also include firms which received significant investments from politically connected private equity firms. Of course, it also matters how “close” the political connection is to the business- man. However, we do not have information to distinguish between different types of connections, as all connected businessman are considered to have first-tier political influence over regulations and their implementation. 8. Employment is observed for about 20,000 establishments, while operating revenues and profits are only available for about 700 and 400 large establish- ments, respectively. 9. We pool all available surveys for Egypt between 2004 and 2008 in order to maximize the representativeness of the perceived policy data at the sector level. Overall, there are more than 4,200 firms which are aggregated into 90 (ISIC Rev. 3.1) four-digit sectors. We exclude sectors for which we observe less than 4 firms, which produces on average 38 firms per four-digit sector. 10. See Rijkers, Freund, and Nucifora (2014) and Diwan, Keefer, and Schiffbauer (2014) for details. The total share of employment in Egypt is calculated as Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 143 the approximate total number of employees in connected firms in 2010 (550,000) relative to the total number of about 7.5 million private sector employees. 11. As in any country, we expect that several firms underreport their output, employment, and profits. It is difficult to assess if connected firms are more or less likely to underreport. 12. The Orbis data for Egypt primarily includes medium and large establish- ments which are the correct comparison groups when comparing politically connected and unconnected establishments. Large firms are well-distributed among connected and unconnected establishments with available data. 13. The high share of net profits is in part the result of many firms reporting losses. When only firms reporting positive profits are considered, Ben Ali firms account for about 7 percent of all profits. 14. Longer time series data for profits are not available in Orbis. We note that the precision of estimated profit differential in 2003 and 2004 is low because of the few available observations. 15. We note that most of the regulatory privileges favoring connected firms (e.g., energy subsidies to industry or trade protection) are still in place until today. Thus, the decline in the profit differential for connected firms immediately after Mubarak’s fall might reflect that other policy privileges (temporarily) disappeared (e.g., implementation bias) or that the new regime at least initially made it generally more difficult for these businessmen to operate in Egypt. 16. In the following, we document only selected channels of policy privileges for which we have available data in Egypt and Tunisia. For example, we neither have sufficiently detailed data on licenses requirements and FDI restrictions for specific sectors in Egypt nor on input subsidies in Tunisia. 17. Wikileaks cables 09TUNIS372_a and 07TUNIS1489-a: see https:// wikileaks.org/plusd/cables/09TUNIS372_a.html, https://wikileaks.org /plusd/cables/07TUNIS1489_a.html, accessed February 23, 2013. 18. Wikileaks cable 08TUNIS679_a, https://wikileaks.org/plusd /cables/08TUNIS679_a.html, accessed February 23, 2013. 19. One issue we encountered was matching the activities listed in the Invest- ment Code to specific five-digit sectors, which do not perfectly overlap. In some cases, the Investment Code provides a more detailed description of activities, whereas in others, the code is more general than the Tunisian NAT 96 classification that we use. With the help of officials at the Tunisian Institut National de la Statistique we created a correspondence between activities and sectors, but in some cases multiple activities were mapped to the same sector and vice versa. As a consequence, it is possible for some sectors to be subjected to several regulations of the same kind. 20. Note that the number of observations on these variables is limited to 64 because we confine attention to enterprises operating in sectors in which the investment code is binding; this reduces the nonconnected firm sample to 70,259. This amounts to about 55 percent of the full sample for both con- nected and nonconnected firms. The regressions are also confined to this group of firms. 21. In order to test this hypothesis, we first match data on NTMs (at the six-digit product level harmonized system classification) from the World Bank data set with the Orbis data (which is at the four-digit industry level). The NTM 144 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa measures are available for tradable goods, corresponding broadly to the manufacturing and mining industries. We therefore limit the analysis of NTMs to these 147 sectors. Our data includes 200 politically connected firms operating in at least one of these sectors. 22. Industrial zones in Egypt include qualified industrial zones (QIZs) which guarantee firms duty and quota free exports to the U.S. Abdel-Latif and Nugent (2010) review the impact of QIZs in Egypt and find that large firms disproportionally benefit from the QIZ agreement: in the 17 industrial zones hosting QIZ factories, 88 percent of exports are concentrated in firms with more than 500 workers. Textiles and garments account for 89 percent of QIZ exports, followed by plastics and chemicals. 23. The WBES data include firm-level data for 95 four-digit (ISIC Rev. 3.1) sectors, including 84 manufacturing and 11 services sectors. All of the 11 four-digit services sectors include multiple connected firms (in hotels and restaurants, retail and wholesale trade), so we restrict the analysis to the four- digit manufacturing sectors including 3,040 firms. 24. We use the following regression model: Polis = bc connecteds + bs Sizeis + bcs connecteds * Sizeis + bXXis + bstS + eis. The dependent policy variable Polis is a dummy variable for firm i in the four-digit sector s. It is 1 if the firm bought land from the government or it is located in an industrial zone, respectively, and zero otherwise; connected measures the number of politi- cally connected firms by type in the four-digit sector s. Size is the dummy variable Small, which is equal to 1 if the firm has less than 100 employees and zero otherwise. Xis is a matrix of firm level control variables: age, export share, S is a matrix of two-digit sector dummies. If we include the dummy variable “Small” for the “Size” variable, bc measures if large firms’ access to land is different in sectors with more connected firms while mea- sures if large firms’ access differs from small firms’ access in sectors with more politically connected firms relative to sectors without (or fewer) political connections. See Diwan, Keefer, and Schiffbauer (2014) for more details. 25. Overall, 85 percent of manufacturing firms with available employment data have at least 100 employees. In contrast, among all manufacturing firm in the WBES, only 33 percent have at least 100 employees (on average, we observe about 12 large firms in a four-digit manufacturing sector in the WBES data). Thus, large firms in the WBES data are much more likely to be politically connected. We also tested for differences in firm age between connected and unconnected sectors. However, the age distribution of politically connected firms and all firms in the WBES data are very similar; the median age among the former is 18, and among all WBES firms it is 19. 26. Table 4.4 also shows that the share of output directly sold to the government is 5–7 percent higher for firms in politically connected sectors indicating that connected firms have preferential access to government procurement contracts. 27. The following results are based on a regression model analog to the ones testing for access to government land but using waiting time for construction permits as the dependent variable. See Diwan, Keefer, and Schiffbauer (2014) for more details. 28. The results are based on regression analysis, including interaction terms between the number of NTM restrictions (at the industry level) or a dummy Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 145 for high energy intensive industries and a dummy variable indication if a firm is politically connected or not. This framework allows testing the hypothesis that connected firms outperform their competitors when trade protection or energy subsidies are prevalent. The sample is confined to 2003–11 because of lack of profits and output data in earlier years. See Diwan, Keefer, and Schiffbauer (2014) for more details. 29. In some sectors we observe several politically connected firms, which could in principle lead to competition among them. Instead, however, we observe a web of intertwined ownership structures and co-investments among politically connected firms. For example, the 6 (10) most intertwined busi- nessmen together control stakes directly or indirectly in 240 (322) firms. In addition, 85 firms (18 percent) managed or owned by a connected business- man received significant investments from private equity funds controlled by other politically connected investors. Thus, collusion among politically connected firms is much more likely. 30. The results are based on regression analysis, including interaction terms between authorization requirements and FDI restrictions (at the sector level) and a dummy variable indication if a firm is politically connected or not. This framework allows testing the hypothesis that Ben Ali firms outperform their competitors when regulatory restrictions are prevalent. The sample is con- fined to activities covered by the investment code. Only firms which report hiring paid workers at some point during the year are included; we exclude the self-employed and those without employees. The sample is confined to 2000–10 because of lack of profits and output data in earlier years. See Rijkers, Freund, and Nucifora (2014) for more details. 31. The following results are based on a regression model analog to the ones testing for access to government land but using the following dependent variable instead: a dummy variable which is equal to one if a firm reports less than 10 competitors in domestic markets and zero otherwise. See Diwan, Keefer, and Schiffbauer (2014) for more details. Comparable data is not available for Tunisia. 32. The analysis is based on the establishment census, which includes more than 2,000,000 establishments across all nonagriculture, nongovernment economic sectors in 1996 and 2006. The entry rates and parameters of the distribution of employment across establishments (coefficient of variation, skewness, and share of micro establishments) are computed at the four-digit sector level and then matched with our information on the number of politically connected firms per sector. Employment weighted entry rates are weighted by the number of employees in entering firms relative to the total number of employees in the four-digit sector. The descriptive statistics are summarized in table 4.6. Moreover, we estimate the effects of the number of connected firms (connected) in the four-digit sector s on (changes in the) measures of firm dynamics, e.g., entry, (Z) for sector s, controlling for the average log of the number of employees and the average establishment age (X) as well as sector dummies at the one- or two-digit level b: Zs,2006 = bc connecteds, 2006 + bXlnXs, 2006 + SB + es, 2006. 33. First, this period witnessed more widespread political connections across sectors. Second, state-business relations intensified as several well-connected businessmen took high political posts, allowing them to directly steer economic policies (Demmelhuber and Roll 2007; Roll 2013). 146 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa 34. The results are robust when controlling for sector-specific characteristics (e.g., average size and age of establishments in a sector as well as one- or two-digit sector dummies) in a regression framework. 35. Arnold, et al. (2012) document that service trade liberalizations—the removing restrictions on FDI in India’s service sectors over the previous decade—led to large productivity gains in downstream using manufacturing industries. 36. The estimation procedure and results are outlined in detail in Appendix E. 37. The estimation is based on establishment census data from 1996 and 2006, including more than 2,000,000 establishments in all nonagriculture, non- government economic sectors. We control for specific sector characteristics such as establishment size and age, and for broad sector dummies in all esti- mation specifications. 38. These bands are based on scores from an assessment consisting of 77 questions— for each question, the government was scored from 0 to 4. TI considered a range of institutions in each country: the defense and security ministries, and armed forces in each country, including any other government institutions with the po- tential to influence levels of corruption risk in the sector. 39. 4 = high transparency; strong, institutionalized activity to address corruption risks. 3 = generally high transparency; activity to address corruption risks, but with shortcomings. 2 = moderate transparency; activity to address corruption risks with significant shortcomings. 1 = generally low transparency; weak activity to address corruption risks. 0 = low transparency; very weak or no activity to address corruption risks. 40. Civilian businesses and defense companies owned, in whole or part, by the government defense establishment or the armed forces. This does not include private businesses lawfully owned by individuals in the defense estab- lishment. 41. This also explains why we were unable to obtain information on politically connected military firms in sections “Privileges to Politically Connected Firms Undermine Competition and Job Creation: Evidence from the Arab Republic of Egypt and Tunisia” and “The Available Qualitative Evidence Points to Similar Mechanisms of Policy Privileges in Other MENA Countries.” 42. One might argue that the political influence of connected businessmen was still stronger in Egypt, since some of these businessmen were ministers and did not have to rely on the influence of family members to direct economic policies. References Aghion, P. M., Dewatripont, L. Du, A. Harrison, and P. Legros. 2012. “Industrial Policy and Competition.” Mimeo (April 24). Aghion, P., C. Harris, P. Howitt, and J. Vickers. 2001. “Competition, Imitation and Growth with Step-by-Step Innovation.” The Review of Economic Studies 68 (3): 467–92. Alley, April Longley. 2010. “The Rules of the Game: Unpacking Patronage Politics in Yemen”. Middle East Journal 64 (3): 385–409. Privileges Instead of Jobs: Political Connections and Private Sector Growth in MENA 147 Arnold, Jens, B. Javorcik, M. Lipscomb, and A. Mattoo 2012. “Services Reform and Manufacturing Performance—Evidence From India.” Policy Research Working Paper 5948, World Bank, Washington, DC. Demmelhuber, T., and S. Roll. 2007. “Maintenance of Domination in Egypt: The Role of Reforms and Economic Oligarchs.” Stiftung Wissenschaft und Politik (SWP), Deutsches Institut für Internationale Politik und Sicherheit. Diwan, Ishac. 2012. “A Rational Framework for the understanding of the Arab Revolutions.” CID Working Paper 237, Center for International Development, Harvard University, Cambridge, MA. Diwan, I., and H. Chekir. 2012. “State-Business Relations in Mubarak’s Egypt.” Mimeo. Diwan, Ishac, Philip Keefer, and Marc Schiffbauer. 2014. “On top of the Pyramids: Cronyism and Private Sector Growth in Egypt.” Working Paper, World Bank, Washington, DC. Faccio, M. 2007. “Politically Connected Firms.” American Economic Review 96: 369–86. Fisman, R. 2001. “Estimating the Value of Political Connections.” American Economic Review 91: 1095–02. Haddad, Bassam. 2012. Business Networks in Syria: The Political Economy of Authoritarian Resilience. Stanford, CA: Stanford University Press. Hallward-Driemeier, M., G. Khun-Jush, and L. Pritchett. 2010. “Deals versus Rules: Policy Implementation Uncertainty and Why Firms Hate It.” Working Paper 16001, NBER, Cambridge, MA. Harris, Kevan. 2013. “The Rise of the Subcontractor State: Politics of Pseudo- privatization in the Islamic Republic of Iran.” International Journal of Middle East Studies 45: 45–70. Henry, Clement Moore, and Robert Springborg. 2010. Globalization and the Politics of Development in the Middle East. Cambridge University Press. Heydemann, Steven. 2004. Networks of Privilege in the Middle East: The Politics of Economic Reform Revisited. New York: Palgrave Macmillan. Jones, C. 2011. “Intermediate Goods and Weak Links in the Theory of Economic Development.” American Economic Journal: Macroeconomics 3 (2): 1–28. Kienle, Eberhard. 2004. “Reconciling Privilege and Reform: Fiscal Policy in Egypt, 1991–2000.” In Networks of Privilege in the Middle East: The Politics of Economic Reform Revised, edited by Steven Heydemann, 281–96. New York, NY: Palgrave Macmillan. King, Stephen Juan. 2009. The New Authoritarianism in the Middle East and North Africa. Bloomington: Indiana University Press. Kremer, Michael. 1993. “The O-Ring Theory of Economic Development.” The Quarterly Journal of Economics 108 (3): 551–75. Loewe, Markus. 2013. “Industrial Policy in Egypt 2004–2011.” DIE Discussion Paper 13/2013, Deutsches Institut fuer Entwicklungspolitik, Bonn. Malik, Adeel. 2013. “The Economics of the Arab Spring.” World Development 45: 1–13. Malouche, Mariem, José-Daniel Reyes, and Amir Fouad. 2013. “New Database of Nontariff Measures Makes Trade Policy More Transparent.” Mimeo, World Bank, Washington, DC. Marotta, Daniela, Christian Ugarte, and Leila Baghdadi. 2014. “Weak Links in Tunisia.” Mimeo, World Bank Washington, DC. 148 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Owen, Roger. 2004. “State, Power and Politics in the Making of the Modern Middle East.” Routledge. Rijkers, Bob, Caroline Freund, and Antonio Nucifora. 2014. “All in the Family: State Capture in Tunisia.” Working Paper, World Bank, Washington, DC. Roll, Stephan. 2013. “Egypt’s Business Elite after Mubarak: A Powerful Player between Generals and Brotherhood.” SWP Research Paper 8 (September), Stiftung Wissenschaft und Politik, Berlin. Sfakianakis, John. 2004. “The Whales of the Nile: Networks, Businessmen and Bureaucrats during the Era of Privatization in Egypt.” In Networks of Privilege: Rethinking the Politics of Economic Reform in the Middle East, edited by S. Heydemann. New York: Palgrave Macmillan. Tlemcani, Rachid. 1999. “Etat, Bazar, et Globalisation: L’Aventure de l’Infitah en Algerie”. Algiers: Les Editions El Hikma. UNIDO. 2010. “Compilation of Energy Statistics for Economic Analysis.” Development Policy and Strategic Research Branch Working Paper 01/2010, UNIDO, Vienna. World Bank. 2009. “From Privilege to Competition: Unlocking the Private-Led Growth in the Middle East and North Africa.” MENA Development Report. Washington, DC: World Bank. Implications for Policy This report shows that the factors holding back formal sector job growth in MENA, such as weak firm entry and exit and low productivity growth, are rooted in a policy environment that favors a few dominant market players and insulates them from competition. The various privileges that such firms capture come at a heavy cost to job creation. Few new firms enter these markets and when they do, they are excluded from these privi- leges and do not grow. Aggregate job creation is therefore weak, many people stay out of the labor market, or are obliged to find employment in small-scale and low productivity activities. The roadmap to more jobs in MENA countries cannot therefore stop at a destination that includes only improved supply-side policies— education, wages, job training, and so forth. It must also encompass significant reforms to stimulate labor demand. The findings in this report point to certain critical elements of this roadmap including: (a) removing the costly policies identified in this report; (b) promoting competition, open markets, and equal opportunities for all entrepreneurs; and (c) most critically, ensuring that going forward, policies and the policy administra- tion are aimed at leveling, rather than tilting the playing field. In the rest of this section we discuss these components of the roadmap; however, the specific details will depend on each country case and should include addi- tional policy areas not covered in this report. First, governments in MENA should reform policies that unduly con- strain competition and the ability of entrepreneurs to pursue opportuni- ties on an equal footing. Chapter 1 suggest that if MENA governments want to pursue private sector development programs targeting specific types of firms, they would be well advised to focus on firm age and not firm size as the primer targeting criterion. Chapter 2 identifies a num- ber of policies that lower competition, tilt the playing field and reduce firm entry, productivity growth, and ultimately, job creation in 149 150 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa MENA countries. These policies include energy subsidies to industry; exclusive license requirements to operate in specific sectors; legal barriers to FDI; trade barriers, including nontariff measures; administrative bar- riers to entry and firm growth; and barriers in access to the judiciary, land, and industrial zones. Moreover, several other policies not analyzed in this report but potentially as important in maintaining a level playing field, should also be considered when dealing with specific country cases, such as barriers to firm entry and exit resulting from restrictive hiring and fir- ing laws, cumbersome bankruptcy laws, and so forth. Second, policymakers should ensure that state interventions that affect competition and equality of business opportunity for all entrepreneurs are enforced uniformly across firms. The implementation of laws is not consistent for all firms, even when they operate in the same sector. The predominance of not only arbitrary, but predatory decision making in the exercise of administrative discretion discourages entrepreneurship and reduces competition. The even-handed enforcement of laws and regula- tions demands that public officials have incentives to exercise discretion fairly and transparently, in pursuit of the legitimate aims of public policy. Such incentives are more likely to exist when laws and regulations are clear; policy implementation is simple and predictable; entry and promo- tions into the administration are based on merit instead of political con- nections; and when merit is judged on the basis of potential or actual contributions to the legitimate goals of public policy. Third, if MENA governments want to pursue state-led development policies, they would be wise to avoid past mistakes and ensure that these new industrial policies—and the administrative structure that implements them—minimize the scope for capture, promote competition, and tightly link support to performance. Chapter 3 lists elements of industrial policy design and complementary institutional changes that appear crucial to making industrial policy work. These include far-reaching organizational reforms in the public sector; a focus on correcting market failures and on new economic activities where market failures are more likely to have a binding influence; an evaluation system in which the performance of both policies and officials is judged by their effects on economic growth and job creation; and ensuring that all firms and potential entrants in the tar- geted sector have access to these specific interventions based on their performance. One critical aspect of this reform agenda is to create the institutions necessary to prevent future capture, thus safeguarding competition and equal opportunities to all entrepreneurs. While several distortive policies could be removed relatively quickly given the political will, the likelihood that other existing or new policies could be captured or serve privileged firms and undermine competition and open markets is high. Faced with Implications for Policy 151 external pressure to reform the economy, the Arab Republic of Egypt substantially reduced import tariffs in the early 2000s. However, when one source of privilege disappeared, another emerged: tariffs declined, but nontariff barriers dramatically increased. Table 4.3 showed that these nontariff barriers disproportionally benefitted politically connected firms. To prevent this, policymakers need to build institutions that pro- mote competition and prevent future capture of policies; such institu- tions include, but are not limited to, a strong, well-organized and highly competent public administration. A strong public administration is necessary to implement the policy changes necessary to build open mar- kets that are resilient to the risk for capture. These policy changes include a strong competition law and an independent competition authority; appropriate procurement laws and implementation; an independent judiciary, and so forth. Another component, just as important, is to ensure policy making is transparent and open, with processes that allow citizens to participate. Transparency is not a panacea, but it is hard to conceive how institutions that safeguard open markets and competition can emerge without citizen access to information on proposed and ratified laws and regulations; citi- zen input into policy design and evaluation; citizen knowledge of politi- cians’ stakes in firms that benefit from government policies; and citizen awareness of the beneficiaries of subsidies, procurement tenders, public land transactions, privatizations, and so forth. Last, this report points to a decision-making guide that summarizes the foregoing, which governments can use as a framework when design- ing and implementing policies. It is also important when using this frame- work to recognize that policymaking faces risk and uncertainty as to which policies will work and achieve its objectives. The decision-making guide is aimed to maximize the likelihood of success given inherent uncertainties and maximize the positive impact of policies on growth and jobs by ensuring that they respond to real obstacles to job growth while minimizing the risk for capture. The decision-making guide suggests that any development policy should pass the following questions; a negative response to one of the subsequent questions raises a red flag indicating that the proposed policy could lead to inefficient and inequitable out- comes ultimately failing to create jobs: 1. Does the policy seek to provide a good or service currently not avail- able in the country in an economically efficient and sustainable manner? 2. Do all potential market participants have equal access to the benefits of the policy? 152 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa 3. Are the benefits of the policy reversible if rigorous performance measures are not met? 4. Will the bureaucracy and courts implement the policy accurately, fairly, transparently? Policy Making Affecting the Private Sector: Decision-Making Flow Chart Policy designed to produce goods and services currently not produced domestically? Yes No Absent market failures, does No No Policy designed to encourage country have a comparative those excluded from the advantage in production? market to participate? Yes Policy is likely to Yes be inefficient and No Will the policy address a clear inequitable Will all or nearly all benefits of market failure that prevents the interventions flow to those production? currently excluded from the market? No Yes No Yes Policy is likely to Will benefits be conditioned Will all potential create jobs No on rigorous measures of Identify market failures entrepreneurs in the performance? that prevent country targeted group or from exploiting its sector be eligible for comparative advantage Yes the policy benefit? No Yes Will the bureaucracy and courts Yes implement policy accurately, fairly, transparently? APPENDIX A Economic Growth and Structural Transformation Our analysis is based on data on employment, value added, and labor productivity by sector for a panel of 35 countries, of which seven countries are from the Middle East and North Africa (MENA) region. Nominal value added is converted to value added in constant purchasing power parity (PPP) terms using the PPP conversation factor from Penn World Tables Revision 7.1. Labor productivity for sector i is calculated as value add in constant PPP terms of sector i divided by employment of sector i. The data set expands data used by Timmer and de Vries (2009) and McMillan and Rodrik (2011), by adding newly compiled statistics for MENA countries. Our time series on value added, price deflators, and employment by sector for MENA countries is constructed from available statistical sources. We follow the methodology developed by Timmer and de Vries (2009), also used by McMillan and Rodrik (2011), using national data which tends to be harmonized in terms of industry classifications. Gross value added in current and constant prices is taken from the national accounts of the various countries. In recent years, value added series have been compiled according to the 1993 United Nations System of National Accounts (UN SNA, see UN 1993). So, international compa- rability is, in principle, high. We follow Timmer and de Vries (2009) and McMillan and Rodrik (2011) as closely as possible to construct a data set for the main 9 sectors according to the definition of 2nd revision of the internal standard industrial classification (ISIC, rev. 2). These nine sectors are 1) agriculture, hunting, forestry and fishing; 2) mining and quarrying; 3) manufacturing; 4) public utilities (electricity, gas, and water); 5) construction; 6) wholesale and retail trade, hotels and res- taurants; 7) transport, storage and communications; 8) finance, insur- ance, real estate and business services; and 9) community, social, personal, and government services. We also use data from population censuses as 153 154 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa well as labor surveys to estimate sector employment, which captures here all persons employed in a particular sector, independent of their formality status or whether they are self-employed. We received employment data from the Central Agency for Public Mobilization and Statistics for the Arab Republic of Egypt using the ISIC Rev. 3 classification for 1998, 2006 and 2012 and linearly projected employment data for the periods 1999–2005 and 2007–11. Employment for Morocco comes from the population census, for Tunisia from the National Employment Survey (Enquête Nationale de l’Emploi), and includes public and informal employment. We complement this data with information on the level of education of workers by sector for Tunisia, Morocco, and Jordan. Labor productivity growth in terms of change in output per worker can be decomposed into within-sector change and changes across sectors, or structural change. Structural change captures the contribution of real- location of labor (or change in sector weights) to growth. This can be written as: Δyt = ∑Nsi,t–k Δyit + ∑Nyi,t Δsit where Δyt is the change in aggregate labor productivity between t and t−k, Sit is the employment in sector i at time t and yit is the productiv- ity level in sector i at time t. The first term is the “within” component and the second term the “across” component (figure 1.13). Economy- wide labor productivity is thus decomposed into two parts. The first FIGURE A.1 Real GDP Per Capita Growth Decomposition Per capita GDP growth Change in Change in output Change in employment rate per worker demographic structure Changes across Changes within sectors (structural sectors change) Economic Growth and Structural Transformation 155 FIGURE A.2 Long-Term Structural Change in Four MENA Countries a. Egypt, Arab Rep. b. Jordan 2000–07 2000–07 1990–2000 1995–2000 1982–90 –15 –10 –5 0 5 –1 0 1 2 c. Morocco d. Tunisia 2005–10 2005–10 2000–05 2000–05 0 0.5 1.0 1.5 2.0 0 0.5 1.0 1.5 Within Structural change Source: World Bank data. component measures the change in labor productivity that is due to changes in sectoral labor productivity, and it captures how labor pro- ductivity evolved under constant employment shares across sectors. The second component captures the impact of structural change on labor productivity development. It measures the counter-factual pro- ductivity level that was reached if sectoral productivity levels remained unchanged and only shifts in labor across sectors change productivity. Marginal Productivity of Labor The aforementioned analysis of structural change has been based on average productivity. To pass judgment on whether this change was welfare improving and growth promoting, however, would require a 156 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE A.3 Structural Change, by Sector, 2000–05 a. Egypt, Arab Rep. b. Jordan 600 1,000 Contribution to labor productivity change Contribution to labor productivity change 400 800 200 600 0 400 –200 200 –400 0 –600 –200 R N N PU N T C CS E GS R R N N PU N T C E GS R WR FIR WR FIR TS TS AG HE AG HE MI MA CO MI MA CO PS PS OT OT CS c. Morocco d. Tunisia 400 Contribution to labor productivity change 1,000 Contribution to labor productivity change 300 500 200 100 0 0 –100 –500 R N N PU N T C CS E GS R R N N PU N T C CS E GS R WR FIR WR FIR TS TS AG HE AG HE MI MA CO MI MA CO PS PS OT OT e. Saudi Arabia f. Syrian Arab RePublic utilitiesblic 2,000 Contribution to labor productivity change 2,000 Contribution to labor productivity change 1,000 0 0 –1,000 –2,000 –2,000 –4,000 –3,000 R N N PU N T C CS E GS R R N N PU N T C CS E GS R WR FIR WR FIR TS TS AG HE AG HE MI MA CO MI MA CO PS PS OT OT g. West Bank and Gaza Contribution to labor productivity change 600 400 200 0 –200 –400 R N N PU N T C CS E GS R WR FIR TS AG HE MI MA CO PS OT Within Structural change Note: AGR = agriculture; MIN = mining; MAN = manufacturing; PU = public utilities; CON = construction; WRT = wholesale and retail trade; TSC = trans- ports and communication; FIRE = financial and business-oriented services; CSPSGS = community and family-oriented services; OTHER = other sector. Economic Growth and Structural Transformation 157 more in-depth analysis.1 One important step in this direction is to look at marginal productivity across sectors. Under perfect competition, marginal labor productivity—not average productivity—should be equalized across sector. Assuming a constant returns production func- tion, since labor share are not necessarily negatively correlated with average productivity, large gaps in average productivity may reflect large gaps in marginal labor productivity. There are some cave- ats though. For example, high average labor productivity in capital- intensive sectors, such as mining, may simply reflect that the labor share is low. The marginal productivity of labor can be calculated by estimating the labor share of income. Using World Bank I2D2 data, we calculated the income share of labor using wage data for Tunisia and Egypt, the only two countries with reliable wage data.2 In a perfectly competitive market, wages equal the marginal product of labor. Labor markets are often not perfectly competitive, for example, in the presence of union- ization or indexed contracts. Moreover, in many developing countries some workers, such as those in the agricultural sector household employees are only paid partially in wages. Using wages to calculate labor’s share of income automatically leads to an exclusion of self- employed. To eliminate biases arising from unobserved heterogeneity, the data is narrowed down to a subset of workers. The marginal labor productivities are calculated for single males aged 30–34 years with elementary education. The wage data is adjusted for the rural-urban price differential. The share of labor force in paid employment is particularly low in agriculture. In Egypt, only 12.1 percent of the agriculture labor force was in paid employment in 2006, declining from 14.3 in 1998 (table A.1). In Tunisia, agriculture is also the sector with the lowest share of the labor force in paid employment, though at 28.2 percent in 2001. The sectors with the highest share of the labor force in paid employment are public utilities and mining. Gaps in marginal productivities measured by average wages across sectors are smaller than gaps measured by value added per worker, but sectoral differences remain significant. In 1998 in Egypt, the gap between the highest productivity sector (mining) and the lowest pro- ductivity sector (community services) was 57, while the difference in raw wages between the minimum (agriculture) and maximum (financial intermediation) wage was only 2.2. The difference becomes even smaller when controlling for individual characteristics, shrinking to 1.5. However, this gap between wages has increased in Egypt. In 2006, an individual with the same characteristics and education would have earned nearly 2.5 half times more if she would have moved from agri- culture (the lowest wage sector) to mining (the highest wages sector). TABLE A.1 158 Estimates of Labor’s Share and Marginal Productivities Using Harmonized Household Survey Data from the World Bank (I2D2) Average wages Average wages Average wages (controlling for individual (controlling for additional Value added per capita (raw differences) characteristics)a individual characteristics)b Employment in Implied Implied Implied Labor force In 2005 national accounts In 2005 PPP labor In 2005 PPP labor In 2005 PPP labor in paid Employment Country Year Sector Code PPP dollars share % dollars share (pc), % dollars share (pc), % dollars share (pc), % employment, % share in I2D2, % Egypt, 1998 Agriculture agr 6,320 29.7 1,447 22.9 1,394 19.3 1,448 19.4 14.3 40.5 Arab Rep. Mining min 298,623 0.5 2,681 0.9 1,642 0.9 1,694 0.9 95.5 0.2 Manufacturing man 12,642 11.5 2,281 18.0 1,837 19.5 1,899 19.5 81.3 11.9 Public utilities pu 21,056 0.8 2,737 13.0 1,734 13.7 1,797 13.7 100.0 0.7 Construction con 7,888 6.9 2,188 27.7 1,929 29.4 2,002 29.4 87.8 5.3 Commerce wrt 19,262 10.5 2,032 10.5 1,835 13.6 1,898 13.6 36.0 11.9 Transports and communications tsc 16,908 4.3 2,586 15.3 1,823 16.7 1,882 16.7 81.7 4.5 Financial and business-oriented services fire 14,529 1.4 3,247 22.4 1,929 24.3 2,002 24.3 84.2 1.5 Community and family-oriented services cspsgs 5,221 34.5 1,856 35.6 1,304 37.8 1,351 37.8 95.3 23.5 2006 Agriculture agr 6,166 27.3 1,646 26.7 1,654 22.3 1,712 22.3 12.1 38.4 Mining min 229,672 0.8 5,623 2.4 3,935 2.7 4,083 2.7 89.6 0.2 Manufacturing man 8,138 12.7 2,631 32.3 2,104 33.6 2,169 33.6 73.2 11.6 Public utilities pu 25,656 0.8 3,520 13.7 2,163 14.3 2,235 14.3 98.7 0.7 Construction con 6,287 7.8 2,667 42.4 2,294 44.9 2,366 44.9 82.0 6.3 Commerce wrt 18,278 11.6 2,397 13.1 2,267 16.9 2,337 16.9 45.3 14.5 Transports and communications tsc 16,420 4.7 3,379 20.6 2,592 21.5 2,675 21.5 79.2 5.8 Financial and business-oriented services fire 13,978 1.4 3,680 26.3 2,393 29.4 2,475 29.5 77.7 2.2 Community and family-oriented services cspsgs 5,137 33.0 2,583 50.3 1,697 53.3 1,754 53.3 96.9 19.6 Tunisia 2001 Agriculture agr 9,036 21.8 249 2.8 249 2.5 215 2.6 28.2 22.0 Mining min 109,239 0.6 782 0.7 406 0.8 353 0.8 95.6 0.5 Manufacturing man 12,302 18.2 380 3.1 315 3.4 272 3.4 80.7 20.5 Public utilities pu 34,167 0.6 408 1.2 310 1.1 268 1.1 90.3 12.8 Construction con 6,158 11.9 468 7.6 283 8.3 245 8.3 54.6 3.4 Commerce wrt 12,842 17.3 417 3.3 287 3.5 248 3.5 50.7 14.1 Transports and communications tsc 30,666 5.8 610 2.0 342 1.9 296 1.9 64.6 5.4 Financial and business-oriented services fire 69,077 0.8 837 1.2 365 1.3 314 1.3 98.0 0.8 Community and family-oriented services cspsgs 13,169 18.7 592 4.5 308 4.8 267 4.8 93.5 20.3 Source: I2D2; World Bank calculation. a. Regression results controlling for urban location, gender, age (6-year intervals), marital status, occupation and education level (no education, primary, secondary and postsecondary). b. Adds household size and number of working members to the previous controls. Economic Growth and Structural Transformation 159 A 30–34 year-old male Tunisian worker with primary education, work- ing in an elementary education and living in 6-person household with three working members would have increased his wage 1.6 fold in 2001 by moving from agriculture (the lowest wage sector) to mining (the highest wage sector). The Productivity in Modern Formal Sectors Converges but Their Labor Shares Remain Small The decline of the employment share in manufacturing of several MENA countries could potentially slow down their rate of conver- gence. According to Rodrik (2013), labor productivity in formal manufacturing in poorer countries tends to converge to that in high income countries independent of institutions, education, or other growth determinants. Rodrik (2013) argues that this unconditional convergence in formal manufacturing, however, does not imply unconditional aggregate income convergence because of (a) the lack of unconditional convergence in the rest of the economy; and (b) very small and in some developing countries declining (formal) manufac- turing labor shares. Manufacturing labor productivity in MENA’s formal manufactur- ing sector is converging at the same rate as the rest of the world inde- pendent of MENA policies or institutions. Table A.2 (left panel) reports the results of the two main estimation specifications from Rodrik (2013).3 It shows a convergence rate of 2.9 percent implying that industries that are a tenth of the way to the technology frontier (roughly the bottom 20 percent of industries in the worldwide sample) TABLE A.2 Manufacturing Labor Productivity Growth Rates Rodrik (2013) Did the speed of convergence differ in manufacturing productivity in MENA? Oil- Syrian Arab Egypt, Arab All countries Countries Observations MENA importing Oil Jordan Republic Rep. Morocco Tunisia Baseline –0.029** 118 2,122 –0.041 –0.037 –0.044 –0.033 –0.039 –0.064** –0.005** 0.195** (–6.95) (–1.51) (–0.75) (–1.05) (–0.57) (–0.56) (–3.21) (3.30) (3.49) Post-1990 –0.029** 104 1,861 –0.037 –0.039 –0.026 –0.033 –0.144** –0.064** –0.005** (–7.14) (–1.01) (–0.95) (0.21) (–0.58) (–8.32) (–3.33) (3.43) Source: World Bank calculation. Note: Columns 2–4 replicate the baseline finding of Rodrik (2013). Columns 5–12 show the convergence rate in manufacturing labor productivity different MENA countries and MENA country groups. Each cell is based on a regression of growth on initial productivity including year-industry dum- mies and a region dummy as well as the interaction term of the region dummy with initial productivity. The coefficient shows the compound conver- gence coefficient (baseline-coefficient + interaction term). Standard errors are clustered at the country level in all specifications. Significance level: * = 10%, ** = 5%. 160 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa experience a convergence boost in their labor productivity growth of 6.7 percentage points per year. In table A.2, we test whether the con- vergence rate was different in the MENA region. Therefore, we include a region dummy and its interaction term with log initial labor productivity in the corresponding estimation specifications. 4 The coefficient of the interaction term measures whether the convergence rate was differed from the convergence rate across all other countries. The results show that the convergence rate in the MENA region over- all was the same as in the rest of the world. There is some evidence that the pace of convergence is slightly lower in oil-exporting countries. Moreover, unconditional convergence of formal manufacturing labor productivity tended to be faster in Egypt and the Syrian Arab Republic but slower in Morocco and Tunisia; in the latter case it did not con- verge at all. Unconditional convergence in formal manufacturing in MENA did not lead to aggregate productivity convergence because of the very small and in some countries declining labor share of formal manufacturing in the region. The average labor share of formal manufacturing was as low as 2 percent in Syria, 5 percent in Morocco, and 7 percent in Egypt and Jordan, respectively. Moreover it was declining in Morocco and Egypt between 1995 and 2005. Do We Observe Specific Patterns of Specialization in Formal Manufacturing in MENA? We analyze the pattern of specialization and performance of formal sector manufacturing in MENA countries in more detail through the lens of the product space. The product space illustrates the relatedness between products whereby distances between two products represent the similarity between their production structures. Hidalgo et al. (2007) argue that the assets and capabilities needed to produce one good are imperfect substitutes for those needed to produce other goods; in part because the production processes of two goods require similar technology, (intermediate) inputs, or machinery. The authors derive an empirical measure for the relatedness between every pair of 775 four-digit SITC products and show that countries that manufac- ture more “connected” goods are better positioned to diversify in new (related) products. Figure A.4 presents the raw product space which is a graphical illustration of this measure of the relatedness between products. It reveals the existence of a densely connected industrial core (center) and peripheral clusters, garments (left), textiles (left), or elec- tronics (lower right). Economic Growth and Structural Transformation 161 FIGURE A.4 The Product Space Node color Oil Cereals Node size [Learner classification] Fishing [Millions of dollars] Petroleum Tropical Forest 0.3 2 8 40 2000 agriculture Raw materials products Forest products Mining Link color Garments Tropical agriculture [Proximity] Vehicles Animal agriculture ф>0.65 Metallurgy Cereals Electronics ф>0.55 Labor intensive Textiles Chemicals Capital intensive Animal ф>0.4 agriculture Machinery ф<0.4 Chemicals Source: Sahnoun and Schiffbauer 2013. The comparison with the evolution of the production structures in East Asia reveals a lack of cluster formation among related products or manufacturing subsectors in MENA. Figure A.5 illustrates the prod- uct space among lower middle income countries (LMIC) of different regions of the world today and 30 years ago. While the product space itself is the same for all countries (by definition), countries or regions differ in the specialization of products that they successfully export. We follow the authors to use the revealed comparative advantage (RCA) as the measure of export specialization; products in which a country or region has an RCA in exporting are depicted as “black squares.” Figure A.5 shows that low middle income countries in MENA had a comparative advantage in exporting oil and agricultural products (upper sparse part of the product space) 30 years ago. Over time, they developed a comparative advantage in processed food (upper left), garments (left), and base metal product clusters (middle left). Thus, MENA (LMIC) countries’ prospects to further diver- sify have improved over the last 30 years. However, a comparison with the evolution of the product spaces among LMICs in East Asia or Latin America reveals that the speed diversification in MENA 162 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa FIGURE A.5 Product Space in Selected Regions, 1976–78 and 2007–09 a. MENA b. Sub-Saharan Africa c. East Asia and Pacific d. Latin America and Caribbean LMIC, 1976–78 LMIC, 1976–78 LMIC, 1976–78 LMIC, 1976–78 e. MENA f. Sub-Saharan Africa g. East Asia and Pacific h. Latin America and Caribbean LMIC, 2007–09 LMIC, 2007–09 LMIC, 2007–09 LMIC, 2007–09 Source: Sahnoun and Schiffbauer 2013. manufacturing has been lagging. For example, LMICs in East Asia developed export clusters in garments, textiles, electronics, and motor vehicle parts. In contrast, MENA countries did not develop produc- tion clusters among related products or manufacturing subsectors (apart from garments). Notes 1. Not all structural change is good. For example, productivity may be higher in sectors with monopoly power. A reallocation to these sectors would con- tribute positively to structural change but would not necessarily promote growth or enhance welfare (for a more detailed discussion, see Maloney 2012). 2. See also McMillan (2013) “Measuring the Impact of Structural Change on Labor’s Share of Income,” unpublished manuscript. 3. We would like to thank Danny Rodrik for sharing the original data and Stata codes of Rodrik (2013) with the authors. We added regression specifi- cations to test for differences in the speed of unconditional manufacturing convergence in Latin America. All potential errors are the responsibility of the authors. Economic Growth and Structural Transformation 163 4. Each cell of the table reflects the coefficient (and t-value) of a regression. In all cases, the dependent variable is the (compound annual) growth rate of labor productivity for two-digit manufacturing industries. The explanatory variables are the log of initial labor productivity and industry-year fixed effects. The baseline estimation specification consists of a pooled sample that combines the latest 10-year period for each country with data maximizing the number of countries covered (118). Because each country enters with around 20 industries, the total number of observations is 2,122. The second specification restricts the sample to post-1990 10-year periods while the third is a pure cross section for 1995–2005. APPENDIX B Firm Censuses and Surveys: Countries, Time, and Sector Coverage The Arab Republic of Egypt: establishment census; all nonagricultural sectors; no size restrictions in repeated cross sections, 1996, 2006; manufac- turing establishment census for firms with at least 10 employees, panel 2007–11. We use two main establishment census data sets. The data are obtained from the department of statistics in Egypt (Central Agency for Public Mobilization and Statistics). First, the establishment census covers information on employment and firm characteristics of over 2,000,000 and 2,400,000 (nonfarm) economic establishments for the repeated cross- sections in 1996 and 2006, respectively. It covers all economic establish- ments with a fixed location independent of their size; it includes self-employed. Second, we use the annual industrial production survey between 2007 and 2011 also obtained from the Central Agency for Public Mobilization and Statistics. It includes all surviving establishments with at least 10 employees in manufacturing and mining as well as a sample of smaller establishments. The data are in panel format so that we are able to follow individual firms over time. However, given that we only observe firm exit in 2011, we cannot compute firm turnover between 2007 and 2010. The industrial production survey includes various production vari- ables such as value added and capital (fixed assets) allowing us to compute firm productivity. We used two-digit sector output and added value price indices to deflate production and added value. For more details, see Hussain and Schiffbauer (2014). Jordan: establishment census; all nonagricultural sectors; no size restric- tions; repeated cross section, 2006, 2011; panel data with sampling weights available for 15,470 establishments. The establishment census data are obtained from the Department of Statistics in Jordan. The census covers information on employment, capi- tal, and firm characteristics of about 150,000 (nonfarm) economic 165 166 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa establishments in 1996 and 2006. Information on establishments’ output (revenues) is not available (establishments only report if their revenues are within a certain range). The census covers all nonagricultural economic establishments with a fixed location independent of their size; it includes self-employed. Panel data are available for a subsample of 15,470 establishments which are observed in both years. Thus, we do not observe firm exit. Sampling weights are available for these firms allowing us to compute changes in the variables between 2006 and 2011 represent- ing all establishments (operating in both years). We used two-digit sector output and added value price indices to deflate production and added value. For more details, see Al-Kadi (2014). Lebanon: firm census; all nonagricultural sectors formal firms with a unique tax identifier; no size restrictions; panel 2005–10. The firm census data are obtained from the department of statistics in Lebanon. It includes only firms that to valid tax identifier with the federal tax administration. The census covers information on employment, value added, capital, wages, and firm characteristics of about 150,000 (nonfarm) economic establishments from 2005–10; it includes private sector estab- lishments with a fixed location independent of their size (including self- employed). We used two-digit sector output and added value price indices to deflate production and added value. Morocco: manufacturing firm census, mostly firms with at least 10 employees; panel 1996–2006. The database used originates from the yearly survey conducted by the Ministry of Industry and Trade. This survey covers all manufacturing firms with at least 10 employees or with an annual turnover that exceeds 100,000 MAD (about US$11,000). It collects firm level data on a set of variables such as turnover, output, value added, exports, gross labor cost, and the number of permanent and temporary employees. It does not include capital (fixed assets). The survey has almost universal coverage of manufacturing firms across all sectors and areas of the country, with approximately 90% of firms responding.1 On average 7,082 firms were interviewed each year during the sample period of which 536 were new firms and 412 had exited. We used two-digit sector output and added value price indices to deflate production and added value. For more details, see Sy (2013). Tunisia: firm census; all nonagricultural sectors; no size restrictions; panel 1997–2012. We use two main firm census data sets. First, the Tunisian registry of firms, the Répertoire National des Entreprises, 1996–2010 collected by Firm Censuses and Surveys: Countries, Time, and Sector Coverage 167 the Tunisian Institut National de la Statistique. The Répertoire National des Entreprises draws on information from a host of constituent administrative databases including from the social security fund (Caisse Nationale de la Sécurité Sociale—CNSS), which is the source for the employment data, as well as from Tunisian Customs, the Tunisian Ministry of Finance, and the Tunisian Investment Promotion Agency (l’Agence de Promotion de l’Industrie et de l’Innovation), containing data on all firms registered with the tax authorities (see Institut National de la Statistique 2012) for detailed information on its construction). It has information on inter alia the employment, age and main activity of all registered private nonagricultural firms, except cooperatives. The census covers all nonagricultural private sector firms with a fixed location inde- pendent of their size; it includes self-employed. In 2010, the census data contained information on 102,660 firms with employees and an additional 501,746 firms without paid employees (the registered self-employed). Second, the Répertoire National des Entreprises was merged with confi- dential profit and turnover data from the Tunisian Ministry of Finance including private firm tax records for the period 2006 through 2010. The smaller sample of merged data includes production variables such as value added and profits but not capital (fixed assets). We used two-digit sector output and added value price indices to deflate production and added value. For more details, see Rijkers et al. (2014). West Bank and Gaza: establishment census; all nonagricultural sectors; no size restrictions in repeated cross sections, 2004, 2007, 2012; manufacturing establishment census for firms with at least 10 employees, panel 2004–12. We use two main establishment census data sets. The data are obtained from the department of statistics in the West Bank (Palestinian Central Bureau of Statistics). First, the establishment census covers information on employment and firm characteristics of over 80,000 (nonfarm) economic establishments in 2003, 2007, and 2012. Information on establishment age (i.e., year of creation) is not available. It covers all economic establish- ments with a fixed location independent of their size; it includes self- employed. Second, we use the annual industrial production survey between 2004 and 2012. The data are in panel format so that we are able to follow individual firms over time. The industrial production survey includes vari- ous production variables such as value added and capital (fixed assets) allowing us to compute firm productivity. We used two-digit sector out- put and added value price indices to deflate production and added value. Turkey: employment and firm characteristics for all firms; no size restrictions in repeated cross sections 2005–10; annual panel 2005–10 with production variables for all firms with at least 20 employees. 168 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa The Annual Industry and Service Statistics (AISS) provides detailed information on revenue, costs, employment, investment, sector of activity, and the region of location. The census covers more than 2,400,000 nonagricultural private sector firms with a fixed location independent of their size; it includes self-employed. The AISS does not cover the following sectors: Agriculture, hunting and forestry (A), Fishing (B), Financial Intermediation (J), Public administration and defense; compul- sory social security (L), Other community, social and personal service activities (O), Activities of households (P), Extraterritorial organizations and bodies (Q). The AISS data set covers production variables for all firms with 20 or more employees, and a representative sample of small firms with 1–19 employees (AISS provides sampling weights). However, all firms with more than one plant (regardless of number of employees) are covered if they are in one of the following sectors: mining and quar- rying (C), electricity, gas and water supply (E) and transport, storage and communications (I). The AISS data set does not contain information on physical capital stocks. We use depreciation allowances to impute capital stocks at the firm level. We used two-digit sector output and added value price indices to deflate production and added value. For more details, see Atiyas and Bakis (2014). Note 1. The high response rate can be attributed to the rigorous manner in which the survey is conducted. Each year, firms are sent (via post) a questionnaire to complete. Firms failing to complete this questionnaire are then visited by officials from the Ministry of Industry, Commerce and Productivity (MICP) in order to conduct a face-to-face interview. APPENDIX C Share of Employment in Large Firms among State-Owned Enterprises and Foreign Firms Jordan’s, and to a lesser extent Tunisia’s, relatively high concentration of employment in large firms is in part explained by higher inflows of foreign direct investment. That is, 19 percent of all large firms in Jordan and Tunisia are foreign owned (figure C.1).1 These firms account for 30 and 19 percent of employment generated by large establishment in both countries, respectively.2 The contribution of large domestic private sector firms to total employment in economic establishments in the Arab Republic of Egypt FIGURE C.1 Number of Firms and Jobs in Foreign, Domestic Private, or Public Establishments a. Firms b. Jobs 100 100 80 80 60 60 Percent Percent 40 40 20 20 0 0 za p. an ia y za p. n ia y rke rke da nis nis Re Re Ga Ga d Jor Jor Tu Tu Tu Tu rab rab d d an an t, A t, A k k an an yp yp st B st B Eg Eg We We Foreign Domestic private Public Source: Calculation based on census data. Note: Large firms have at least 100 employees. Establishments are defined as public or foreign if at least 10 percent of the capital was owned by the state or foreign owners, respectively. In the Arab Republic of Egypt, we added establishments that are foreign branches according to their legal status. In West Bank and Gaza, we added establishments with a legal status of a domestic nongovernmental organization to public establish- ments; these are the majority of public establishments in West Bank and Gaza. 169 170 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa and West Bank and Gaza is marginal even by regional standards. These firms accounted for less than 10 percent of total employment in Egypt and West Bank and Gaza in the late 2000s (figure C.1). Figure C.1 shows that among the few large establishments in Egypt in 2006 only about half were domestic private sector firms. Furthermore, state-owned enterprises in Egypt still accounted for 29 percent of total employment in large establishments with at least 100 employees. In West Bank and Gaza, almost all employment attributed to the public sector is in nongovern- mental organizations. Taken together, figure 1.6 and figure C.1 reveal that the share of jobs in large domestic private sector establishments in Egypt and West Bank and Gaza is small compared with regional peers. Notes 1. SOEs in Tunisia are excluded; they are, however, relatively few in number. Establishments are defined as public or foreign if at least 10 percent of the capital was owned by the state or foreign owners, respectively. In Egypt, we added establishments that are foreign branches according to their legal status. In West Bank and Gaza, establishments with the legal status of an nongovernmental organization are tallied with the number of public establishments (which are the majority of public establishments in West Bank and Gaza). 2. The evolution of foreign direct investment in Jordan and its effect on domestic employment is analyzed in detail in Chapter 2. APPENDIX D Employment Growth over Firms’ Life Cycles: Manufacturing Sector FIGURE D.1 Manufacturing: Employment Growth over the Life Cycle 8 Employment (1 for age 0–4) 6 4 2 0 0–4 5–9 10–14 15–19 20–24 25–29 30–34 35–39 ≥40 United States Turkey Jordan Tunisia Egypt, Arab Rep. Lebanon Source: Calculations based on census data. Note: The figure shows the average number of employees for different age-cohorts across establishments in manufacturing (weighted by employment share of 4-digit sectors following Hsieh and Klenow (2012). The average number of employees in each age cohort has been normalized to 1 for the youngest age category (age 0–4 years). The analysis for Turkey, the Arab Republic of Egypt, and Lebanon is based on census data in 2006, for Tunisia in 2009, for Jordan in 2011, and for the United States in 2002. Results for Jordan and Tunisia are similar for other years (2006, 2010, or 2012). We excluded the two largest firms in the oldest age category in Jordan (the dotted line shows the average size of firms when including these outliers). 171 APPENDIX E FDI Inflow and Employment in Jordan: Regression Analysis TABLE E.1 Employment Spillovers from FDI, by Firm Characteristics (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) Variable All Manufacturing Services Small Large Old Young All Manufacturing Services Foreign share 06 0.000 0.000 −0.001 −0.001* −0.001 −0.002 0 0.000 0.000 −0.001 −0.001 −0.001 −0.001 −0.001 −0.001 −0.001 −0.001 −0.001 −0.001 −0.001 Horizontal 06 −0.148** −0.048 −0.157 −0.157** −0.331* −0.338** −0.087 −0.117 −0.053 −0.171 −0.073 −0.093 −0.127 −0.077 −0.188 −0.158 −0.083 −0.073 −0.094 −0.126 Backward 06 0.111 −0.057 0.400** 0.078 0.157 0.006 0.163* −0.075 −0.086 −0.166 −0.079 −0.205 −0.162 −0.084 Forward 06 0.023 −0.13 0.623** 0.027 −0.033 0.149 −0.033 −0.08 −0.097 −0.224 −0.084 −0.217 −0.187 −0.088 Backward services 06 0.171** −0.022 0.605** −0.086 −0.135 −0.239 Forward services 06 0.714** 0.347 1.076** −0.251 −0.736 −0.301 Backward manufacturing 06 −0.07 −0.079 1.282* −0.092 −0.096 −0.704 Forward manufacturing 06 −0.084 −0.142 1.120* −0.084 −0.099 −0.596 Constant 0.315** 0.264 −0.932** 0.204** −0.119 −0.107 0.068 0.276** 0.230 −1.132** −0.069 −0.196 −0.217 −0.061 −0.219 −0.163 −0.08 −0.070 −0.200 −0.262 Observations 15,465 2,637 12,828 14,605 860 3,464 12,001 15,465 2,637 12,828 R2 0.01 0.019 0.013 0.011 0.085 0.023 0.012 0.012 0.019 0.013 Industry fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Region fixed effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Significance level: * = 10%, ** = 5%. 173 APPENDIX F Quality of Business Environment and Jobs in Morocco: Data, Methods, and Main Findings The data used in this paper derive from two sources: the Moroccan Annual Census, and detailed surveys conducted by the World Bank. The Moroccan Annual Census of Manufacturing covers the period 1997–2004. This annual census covers all manufacturing firms with no size limitation. It contains information on sales, value added, output, exports, employ- ment, date of creation, location, investment, and four-digit industry code using the Moroccan Nomenclature of Economic Activities (NMAE). For a subset of firms we also have access to three much more detailed data sets: (1) FACS which contains production data for the years 1998 and 1999 (with some data for 1997), and business environment data for 1998, (2) Investment Climate Assessment (ICA)-2004 which contains produc- tion data for the years 2000–02 and business environment data for 2000 and (3) ICA-2007 which contains production and business environment data for 2002 and 2005. The firms included in FACS, ICA-2004 or ICA-2007 are all contained in the Census. In order to correct for the pos- sible over or underrepresentation of firms in the ICA and FACTS samples, we weight each firm surveyed in the FACS and ICA by the share of the corresponding firm type in the census. The weights are defined on the basis of the 10 NMAE industries, 10 regions, and three size classes used in the regressions. Thus, the results can be interpreted as benign representa- tive for the manufacturing sector in Morocco. The FACS and ICA surveys cover food, textiles, garment, leather, chemicals, wood & paper including publishing, rubber & plastics, metals & mechanical, and electrical & electronic industries. The surveys include firm location with seven distinct geographical areas identified.1 The data con- tains considerable detail on production variables, firm characteristics, and features of the business environment. We use two-digit NMAE production price index and investment price index to deflate production, value-added, and investment. The capital stock is available only for years 1997–2002 and 2005 in the FACS and ICA databases and for 2003–04 in the Census. 175 176 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa To obtain the stock of capital for the remaining years of the sample, we use the available data on investment and apply the perpetual inventory method taking a depreciation rate of physical capital of 5 percent. After cleaning of the data set we end up with an unbalanced panel containing 35,534 observations and 6,119 firms.2 Each firm appears in the sample for at least three consecutive years and at most 9 years (see Table F.1). The estimation function is based on the theoretical model of firm growth proposed by Evans (1987). We model plant growth as a function of age and size. For employment growth we employ the job creation rate following Davis, Haltiwanger, and Schuh (1996), which accounts for the employment growth that occurs in the year that a firm was created, and the employment destruction that occurred when a firm exits. Given the aim of our analysis, in addition to age and size, we control also for trade orientation adding the average firm export share, a dummy variable cap- turing whether the firm has any foreign ownership, the productivity level and the level of competition at the four digit level. The core regression equation therefore takes the following form: JCRi (t , t + n ) = β1 + β 2 ln S i + β3Age i + β 4 AgeSQi + β5Trade i + β6 ForeignOwni + β7TFPi + β8 Herf i + d r + d s + u i i where JCR is the computed job creation rate; ln_S and Age refer to the logarithms of beginning-of-the period total employment and age; AgeSq is the squared age and captures the nonlinear relation between this vari- able and firm growth. Trade is computed as the average firm export share (exp_share) over the time period and ForeignOwn is a dummy variable which is equal to 1 if the firm reports a positive share of foreign capital at the beginning of the period. Estimates of total factor productivity (TFP) are derived at the firm level in the presence of endogenous input choices and selection issues using investment as a proxy for unobservable firm productivity. The estimates are based on both the semi-parametric method developed by Olley and Pakes (1996, henceforth OP), as well as the improvements suggested by Ackerberg, Caves, and Frazer (2007, henceforth ACF). For the degree of competition (Herf) we use the Moroccan census to compute Herfindahl indices at both the three digit and four digit level and explore the sensitivity of the results to these alter- natives. Industry (ds) and regional (dr) dummies are added. We then consider the role of the financial, policy and business constraints and competition variables, and this is done in two ways. In a first stage, we run a series of regressions where we sequentially and separately include each of the finance, policy, business, and competition variables in order to assess whether in aggregate these appear to be correlated with employment growth. In a second stage, we again take each of these variables, and Quality of Business Environment and Jobs in Morocco: Data, Methods, and Main Findings 177 consider whether different “types” of firms show a different relationship between each variable and employment growth. Here we divide firms into different types in various ways: by size, by age, by export-share intensity; by foreign ownership status, by level of productivity, and by the degree of com- petition in Morocco as measured by the Herfindahl index. (Table F.2 sum- marizes the main findings; it only reports the actual coefficients of the interaction terms if they are significantly different from zero [statistically.]) TABLE F.1 List of Regulatory Policy Variables Name Description Type Source Long-term credit cost The interest rate on long-term domestic debt 440 C FACS Access to bank credit To what extent is this an obstacle to the growth of your firm? 684 0–4 ICA Equivalent fiscal treatment Do firms in your sector face equivalent fiscal treatment? 668 0–1 FACS Dispute resolution To what extent is this an obstacle to the growth of your firm? 830 0–4 ICA Judicial system To what extent is this an obstacle to the growth of your firm? 684 0–4 ICA Wait permit What is the average time taken to obtain a construction permit? 660 C ICA No. of permits for enterprise creation If firm was set up in 1999 how many permits were needed? 575 C FACS Admin constraints No of permits needed each year to operate? 488 C FACS Corruption To what extent is this an obstacle to the growth of your firm? 822 0–4 ICA Unfair informal sector competitors To what extent is this an obstacle to the growth of your firm? 684 0–4 ICA No. of competitors For your principal product how many competitors do you have? 640 C FACS Extent of foreign competitors Are there any foreign firms among your competitors on Morocco? 667 0–1 FACS Note: C = continuous. TABLE F.2 Job Growth Regression with Coefficients of the Policy and Environment Variables Coefficient without Small firms Medium Large firms Startups Young firms Old firms Dependent variable: Job growth interaction (<=15) (10–100) (>100) (<=3) (4–10) (>10) Regulatory constraints Equivalent fiscal treatment 0.024 0.311** −0.132** Dispute resolution 0.010 0.053** 0.260** −0.173** Judicial system −0.005 0.275** −0.193** −0.048** Wait permit −0.047** −0.050** −0.048** −0.041** −0.083** −0.063** No. of permits firm creation −0.011 0.088** −0.066** −0.033* Admin constraints 0.010 0.019* −0.069** 0.088** Corruption −0.001 −0.059** 0.058** 0.238** −0.207** Competition Unfair informal comp −0.004 0.031** −0.030** 0.134** −0.112** No. of domestic competitors 0.001 0.012** 0.012** −0.002* Extent of foreign comp −0.058** 0.027* −0.086** −0.110*** −0.055** −0.086** Finance constraints Long-term credit cost −1.66** −0.958** −1.81** −2.26** 0.487* −2.16** −2.42** Access to bank credit −0.020 0.067** −0.044** −0.039** 0.152** −0.121** −0.061** Note: Coefficients of all variables are shown in the first column. However, coefficients of the interaction terms in subsequent columns are only shown if significant. Significance level: * = 10%, ** = 5%, *** = 1%. 178 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa Notes 1. Grand Casablanca, Tanger-Tetouan, Rabat-Sale-Zemmour, Fes-Boulmane- Meknes, Oriental, Chaouia-Ouardigha, and Agadir. 2. The procedure used is close to Fernandes’ one (2008) but less restrictive. Our cleaning has been realized in two steps. In a first one, we have eliminated from the sample (a) firms that have never reported any sales or material costs (costs of raw materials), (b) observations when exports are bigger than sales, and (c) observations with year-to-year growth rates in any of 3 ratios (sales to total workers, material costs to total workers and capital to total workers) larger (smaller) than 500% (−500%). These year-to-year growth rates are calculated with the constant variables. In a second step, we have always kept the firms who exist at less three consecutive years and we have dropped observations when we have one isolated year. APPENDIX G Political Connections and Private Sector Growth in the Arab Republic of Egypt We use this macroeconomic quasi-experimental setting, to test whether aggregate employment growth over a 10-year period between 1996 and 2006 declined after the entry of politically connected firms into initially unconnected (open) sectors. Therefore, we use the following difference- in-difference estimation specification, whereby ΔYst measures employ- ment growth of the four-digit sector s between 1996 and 2006, PCEntry indicates the entry of politically connected firms between 1997 and 2007, NPC are sectors without connected firms before 1997, X is a matrix of con- trol variable (employment and age), and S a matrix of sector dummies: ΔYs,2006–1996 = βEPCEntrys,1997–2006 + βNNPCs,1996 + βEN(PCEntrys,1997–2006 * NPCs,1996) + βXXs,1996 + S + εs,2006 (G.1) Holding all else constant, entry always increases employment in the sector regardless of the fact that the entrant is connected or not. Thus, we expect that the entry of connected firms leads to sector employment growth, unless the adverse impact of connected firms on the growth oppor- tunities of their unconnected peers leads to their exit or shrinkage. In con- trast, we do not expect to observe the latter adverse effect (or at least expect it to be less pronounced) when connected firms enter into sectors which were already dominated by connected firms in previous years. Therefore, negative aggregate employment growth after the entry of connected firms into previously unconnected sectors implies that the decline in employ- ment in unconnected firms (which cannot compete) outweighs any positive job creation of the connected firm(s).1 Table G.2 shows that several sectors across the economy that previously had no connected firms have experi- enced the entry of connected firms in the time period of interest. Table G.1 summarizes the findings of the difference-in-difference estimation. Columns 2 and 3 show the results for our most conservative 179 180 Jobs or Privileges: Unleashing the Employment Potential of the Middle East and North Africa TABLE G.1 Employment Growth Declines after Politically Connected Firms Enter Initially Unconnected Sectors Employment growth, 1996–2006 CEO Owner Broad Entry PC 32.2* 36.1** 7.15 10.3 4.83 4.40 (1.95) (2.09) (0.84) (1.24) (0.99) (0.77) Not connected before 1996 −6.32 15.1 −10.5 (−0.58) (0.82) (−0.67) (Entry PC)* −24.8** −18.7** −14.96 (not connected before 1996) (−2.17) (−3.47) (−0.97) ln(empl) −.418** −.401** −.420** −.382** −.420** −.376** (−2.44) (−2.17) (−2.37) (−2.16) (−2.34) (−2.62) Age 12.5 12.6 12.4 12.3 12.4 12.9 (1.57) (1.56) (1.51) (1.53) (1.51) (1.55) No. of sectors 224 224 224 224 224 224 R2 0.161 0.163 0.155 0.159 0.048 0.160 Sector dummies 1-dig 1-dig 1-dig 1-dig 1-dig 1-dig Significance level: * = 10%, ** = 5%. TABLE G.2 Entry of Connected Firms into Initially Unconnected Sectors, 1997–2006 Sector name 2-digit ISIC Rev. 3.1 4-digit Sector name 4-digit Other mining and quarrying 1410 Quarrying of stone, sand, and clay 1429 Other mining and quarrying n.e.c. Manufacture of food products and beverages 1551 Distilling, rectifying, blending of spirits 1552 Manufacture of wines 1553 Manufacture of malt liquors and malt 1554 Manufacture of soft drinks and mineral water Manufacture of chemicals and chemical products 2412 Manufacture of fertilizers Manufacture of basic metals 2720 Manufacture of basic precious metals Manufacture of electrical machinery 3140 Manufacture of primary cells and batteries Manufacture of radio, TV, and communication equ. 3230 Manufacture of television and radio receivers Manufacture of furniture 3691 Manufacture of jewellery and related articles Recycling 3710 Recycling of metal waste and scrap 3720 Recycling of nonmetal waste and scrap Electricity, gas, steam, and hot water supply 4010 Electricity production, transmission, and distribution 4020 Manufacture and distribution of gas Collection, purification, and distribution of water 4100 Collection, purification, and distribution of water Wholesale trade and commission trade 5131 Wholesale of textiles, clothing, and footwear 5141 Wholesale of solid, liquid, and gaseous fuels 5152 Wholesale of electronic and telecommunications parts Retail trade 5211 Retail sale in nonspecialized stores with food Water transport 6120 Inland water transport Insurance and pension funding 6601 Life insurance Renting of machinery and equipment 7111 Renting of land transport equipment Other business activities 7411 Legal activities 7430 Advertising Political Connections and Private Sector Growth in the Arab Republic of Egypt 181 measure, firms managed by a political connected CEO. We find that entry of connected firms into initially already connected sector increased employment growth, potentially because of the direct positive employ- ment impact of the new connected entrant. Most important, however, we find that aggregate employment growth declines once connected firms enter new, initially unconnected sectors; the corresponding coefficient is significant at the 5 percent level. The economic impact is large. The mag- nitude of the corresponding coefficient suggests that aggregate employ- ment in these sectors shrinks by 25 percent over the 10-year period 1996–2006. Note that the connected firms did not necessarily enter directly in 1997 so that employment growth might have been positive in earlier years but then declined substantially because of the sudden pres- ence of the connected firm with access to policy privileges guaranteeing a large cost advantage over the existing competitors or potential new (unconnected) entrants. The negative aggregate employment growth effect after the entry of connected firms into new unconnected sectors is comparably large and significant at the 5 percent level when we restrict our definition of cronyism to firms owned by politically connected busi- nessmen (column 5). For the broadest measure of connectedness, which also includes firms that received investments from connected private equity funds, the relevant coefficient of the interaction term is still nega- tive and of comparable magnitudes but not significant at conventional levels. Note 1. We do typically not observe if other first-tier politically connected firms operated in these “unconnected” sectors but exited before 2006. Thus, we have to assume in this macroeconomic quasi-experiment that, if unobserved first-tier connected firms which were forced to exit before 2006 existed, they did not operate in these “unconnected” sectors. All available evidence, how- ever, suggests that policy privileges granted to the private sector expanded rather than declined between 1996 and 2006 (see Demmelhuber and Roll, 2007; Roll, 2013) making the exit of unconnected firms less likely. ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. • 10 trees The Office of the Publisher has chosen to • 5 million Btu of print Jobs or Privileges: Unleashing the total energy Employment Potential of the Middle East • 1,335 lb. of net and North Africa on recycled paper with greenhouse gases 30 percent postconsumer fiber in accordance • 3,763 gal. of waste with the recommended standards for paper water usage set by the Green Press Initiative, a non- • 784 lb. of solid profit program supporting publishers in waste using fiber that is not sourced from endan- gered forests. For more information, visit www.greenpressinitiative.org.