DIREC TIONS IN DE VELOPMENT Human Development 90271 Striving for Better Jobs The Challenge of Informality in the Middle East and North Africa Roberta Gatti, Diego F. Angel-Urdinola, Joana Silva, and András Bodor Striving for Better Jobs Direc tions in De velopment Human Development Striving for Better Jobs The Challenge of Informality in the Middle East and North Africa Roberta Gatti, Diego F. Angel-Urdinola, Joana Silva, and András Bodor © 2014 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, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http:// creativecommons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Gatti, Roberta, Diego F. Angel-Urdinola, Joana Silva, and András Bodor. 2014. Striving for Better Jobs: The Challenge of Informality in the Middle East and North Africa. Directions in Development. Washington, DC: World Bank. doi:10.1596/978-0-8213-9535-6. 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. Responsibility for the views and opinions expressed in the adaptation rests solely with the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third- party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to 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 (paper): 978-0-8213-9535-6 ISBN (electronic): 978-0-8213-9536-3 DOI: 10.1596/978-0-8213-9535-6 Cover photo: © Toufik Berramdane. Used with permission. Further permission required for reuse. Cover design: Naylor Design Library of Congress Cataloging-in-Publication Data Striving for better jobs : the challenge of informality in the Middle East and North Africa / Roberta Gatti . . . [et al.]. p. cm. — (Directions in development) Includes bibliographical references. ISBN 978-0-8213-9535-6 (alk. paper) — ISBN 978-0-8213-9536-3 1. Informal sector (Economics)—Africa, North. 2. Informal sector (Economics)—Middle East. 3. Labor market—Africa, North. 4. Labor market—Middle East. I. Gatti, Roberta. II. World Bank. HD2346.A56S77 2012 331—dc232012011909 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Contents Preface xv Acknowledgments xix Contributors xxi Abbreviations xxiii Overview 1 Background 2 Understanding Informality 3 Informality in MENA: Levels and Trends 5 Who Are Informal Workers? 8 Informality among Firms 16 Informality in MENA: Exclusion or Choice? 18 Barriers to Coverage and Policy Options 21 Notes 30 References 30 Chapter 1 How Large Is Informality and Why Do We Care? 33 What Is Informality? 33 How to Measure Informality 35 Why Does Informality Matter? 42 Informality in MENA 46 What Causes Informality? 54 Conclusions 64 Annex 67 Notes 72 References 73 Chapter 2 Profile and Micro-Determinants of Informality 77 Introduction 78 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 v vi Contents Macroeconomic Context 80 Informality Profile 85 Determinants of Informality 99 Conclusions 109 Annex 114 Notes 122 References 122 Chapter 3 Informality and the Firm 125 Introduction 126 Description of Informality among Firms 126 Main Costs and Benefits of Informality for Firms 135 Stylized Facts on Informality among Firms in MENA 142 Conclusions 163 Annex 166 Notes 181 References 183 Chapter 4 Informality: Choice or Exclusion? 187 Introduction 188 Informality and Wages 189 Job (Im)Mobility 200 What Are Informal Jobs Like? 211 Self-Employment: A Choice? 215 How Much Are Workers Willing to Contribute to Social Security? 217 Networks and Intergenerational Persistence 218 Conclusions 221 Annex 224 Notes 228 References 229 Chapter 5 Barriers to Coverage and Policy Options 231 Part 1. Addressing Institutional Barriers and the Low Productivity Trap 232 Part 2. Extending Social Insurance Coverage 265 Extending Social Insurance Beyond the Existing Mandatory System: Guiding Tree with ­ Application to MENA 294 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Contents vii Extending Coverage beyond an Existing Social Insurance System Mandatory ­ 299 Annex 316 Notes 317 References 321 Boxes 1.1 Defining Informality 35 1.2 Main Definitions of Relevant Variables and Survey Questions from Data Available in MENA 39 2.1 Informality in Tunisia 2005–2009 87 2.2 Informality Trends and the Size of the Public Sector (Egypt, 1998–2006) 110 3.1 Job Creation and Productivity: Small versus Large Firms in Morocco 131 4.1 Cognitive and Noncognitive Tests 197 5.1 Employment Protection Legislation (EPL) 239 5.2 Minimum Wages 244 5.3 MENA Case Studies of Productivity-Enhancing ­Programs in ­Urban Areas 263 5.4 Proposal for Expansion of Health Insurance in Lebanon through a ­ Targeted MDC 273 5.5 Integration and Portability of Social Insurance Schemes 278 5.6 Social Insurance Reform in Jordan 284 5.7 Coverage Rates and Vulnerability Due to Informality 290 5.8 Political Economy Considerations for Coverage Extension 294 5.9 Targeting Methods 301 5.10 The Design of MDC Schemes in India, Mexico, and China 304 5.11 Implementation of the Indian Rashtriya Swasthya Bima Yojana 310 Figures O.1 Informality and Economic Development 4 O.2 Informality in MENA Compared to Other Regions 7 O.3 Informality Rates for Selected Non–Gulf Cooperation Council Members 9 O.4 Annual Growth Rates of Informality 10 O.5 Informality Rates by Quintile of Per Capita Consumption for Selected Countries 11 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 viii Contents O.6 Employment Status by Age for Selected Countries, Urban Areas Only 12 O.7 Informality Rates by Highest Educational Level Completed 14 O.8 Informality and Firm Size 15 O.9 Unregistered Firms, by Region 17 O.10 Highest Level of Education of Managers in Formal and Informal Manufacturing Firms in Egypt, 2009 18 O.11 Transitions of Originally Informal Salaried Private Sector Male Workers 19 O.12 Estimated Formality Premium by Gender in Different ­Countries 20 1.1 Correlation among Most-Used Informality Indicators 40 1.2 Correlations among Different Definitions of Labor Informality (Worker’s Side) 42 1.3 Informality and Economic Development 44 1.4 Effect of Informality on Growth 45 1.5 Employment and Unemployment Rates in MENA 47 1.6 Economic Growth and Employment Growth 49 1.7 Annual Population Growth Rates (%) (2005–2009) 50 1.8 Sectoral Contribution to Annual Employment Growth in Typical MENA Country and Selected Other Countries, Average for 2000s 50 1.9 Prevalence of Informality in MENA versus Other Regions 52 1.10 Informality Rates for Selected Non-GCC Economies 53 1.11 Annual Growth Rates of Informality 55 1.12 Correlation between Informality and Governance/Regulation 57 1.13 Correlation between Informality and Education/Demographic Factors 58 1.14 Correlation between Informality and Government Size 60 1.15 Correlation between Informality and Indirect Labor Costs 61 1.16 Percentage Difference in Education and Youth Bulge with Respect to the World Median 62 1.17 Percentage Difference in Rural Population and Agricultural Output with Respect to the World Median 63 1.18 Percentage Difference in Public Employment with Respect to the World Median 65 2.1 Employment Composition by Sector (Using Latest Available Year) 80 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Contents ix 2.2 Yearly GDP Per Capita Growth Rate 81 2.3 Growth in Employed and Working-Age Population in Non-GCC MENA Countries, 1998–2009 82 2.4 School-to-Work Transition (for Ages 15–35) in MENA 83 2.5 Employment Growth between 2000 and 2007 for Selected Countries 84 2.6 Annual Growth of Informality for Selected Economies 86 2.7 Informality Rates by Quintile of Per Capita Consumption for Selected Countries 88 2.8 Size of Private Salaried Formal Sector versus Other Sectors 89 2.9 Distribution of Formal Employment for Selected MENA Countries 90 2.10 Employment Status by Age for Selected Countries, Urban Areas Only 91 2.11 Basic Correlations by Strata (Informality and Individual’s Characteristics) 97 2.12 Informality Rates by Gender 98 2.13 Informality Rates by Age Groups 100 2.14 Informality Rates by Highest Educational Level Completed 101 2.15 Marginal Increase in the Probability of Being “Informal” according to Gender and Marital Status 105 2.16 Informality Rates by Age Group (Syrian Arab Republic and the Arab Republic of Egypt) 107 2.17 Marginal Increase in the Probability of Being “Informal” according to Firm Size (Omitted Category, Small ­ Firms, 2–9 Workers) 111 3.1 Margins of Informality among Micro- and Small Firms in Morocco and Lebanon 128 3.2 Unregistered Firms, by Region 129 3.3 Informality among Registered Firms, by Region 133 3.4 Informal Workers in Formal and Informal Micro- and Small Firms 133 3.5 Average Share of Informal Salaried Workers in Formal and Informal ­Micro-Firms in Morocco by Margins of ­Informality 134 3.6 Where Do the Informal Workers Work? 134 3.7 Obstacles to Formalization 137 3.8 Corporate Tax Rates in Developing Economies 137 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 x Contents 3.9 Difficulty of Redundancy Index 139 3.10 Percent of Firms Indicating Labor Regulations as a Major Obstacle to Business Development 140 3.11 Entry Regulation across Regions and in Selected MENA ­Countries 141 3.12 Enforcement of Tax Regulations and Bribes during Tax Inspections 141 3.13 Percentage of Underreported Sales and Workers, by Economy 142 3.14 Worker and Sales Underreporting by Firm Size 144 3.15 Association between Doubling Employment Size and Worker ­Underreporting 145 3.16 Association between Doubling Labor Productivity and Worker Underreporting 146 3.17 Worker Underreporting by Manager’s Education Level 146 3.18 Changes in Worker Underreporting Associated with a Change in M Education Level ­ anager’s ­ 147 3.19 Worker Underreporting and the Workforce Education Level 148 3.20 Change in Worker Underreporting When Workforce Education Increases 149 3.21 Relative Provision of Worker Training in MENA Economies and ­ Informality 150 3.22 Firms’ Reported Reasons for Registering or Acquiring a License or Tax Card 153 3.23 Highest Education of Managers for Formal and Manufacturing Firms in Egypt Informal ­ 154 3.24 Highest Education of Male Workers in Formal and Informal Manufacturing Firms in Egypt 155 3.25 Worker Categories in Micro-Firms in Morocco 156 3.26 Distribution of Workers by Hourly Wage Quintile in ­ Micro-Firms in Morocco 157 3.27 Distribution of Hourly Earnings of Workers in Micro-Firms in Morocco 158 3.28 Distribution of Workers’ Age, by Type of Informal Employment 159 3.29 Likelihood of Earning “Satisfactory Income” in Registered versus Unregistered Micro-Firms in Morocco 161 3.30 Cost Categories among Micro-Firms in Morocco 161 3.31 Sources of Investment among Micro-Firms in Morocco 162 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Contents xi 3.32 Obstacles to Credit Requests among Micro-Firms in Morocco 163 4.1 Kernel Density Plots of Hourly Wages by Employment Sector 191 4.2 Estimated Formality Premiums by Gender in Different Countries 193 4.3 Oaxaca Decomposition of the Differences in Wages 195 4.4 Distribution of Labor Market Transitions in Egypt 203 4.5 Contribution Density and the Distribution of Contributory and N ­ oncontributory Spells in Jordan 208 4.6 Distribution of the Length of Social Insurance Contributory and N ­ oncontributory Spells in Jordan and Chile 209 4.7 Lack of Access to Job Benefits between Informal and Formal Sectors (% of Workers without Access) 213 4.8 Net Wages by Employment Status (Lebanon, 2010) 215 5.1 Correlation across Countries between the Size of the Informal ­Sector and the Ease of Doing Business 234 5.2 Obstacles to Formalization in Egypt and Morocco 236 5.3 Share of Firms Identifying Labor Regulations as a Major Constraint in Doing Business in MENA (%) 240 5.4 Ratio of Minimum Wages to the Average Value Added per Worker in ­Selected Countries in 2010 246 5.5 Hourly Wage Distribution and Minimum Wages in Egypt, the Republic of Yemen, and Morocco 247 5.6 Severance Pay for Redundancy Dismissal (Average for Workers with 1, 5, and 10 Years of Tenure, in Salary Weeks) 250 5.7 Percent Tax Wedge in Various Countries 253 5.8 The Relationship between Educational Attainment and the ­Unemployment Rate in Urban Egypt and the Republic of Yemen 258 5.9 A Framework for Social Insurance/Coverage Extension Policies: Objectives, Social and Economic Risks, Policies, Instruments, and Behavioral Impacts ­ 267 5.10 Old-Age Pension Design and Coverage: Options Ranging from Pure Earnings-Related Pensions to Pure Noncontributory Flat Pensions 270 B5.4.1 Subsidized and Individual Contributions by Income Quintile 274 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 xii Contents 5.11 Path Dependency in the Evolution of Pension Systems: Degree of Integration of Civil Service Pension Schemes in MENA and Other Regions 277 5.12 Mandatory Pension Design Types across Regions of the World 280 5.13 Internal Rate of Return under Various Pension System Participation Strategies in Egypt, Jordan, and Syria ­ 287 5.14 Internal Rate of Return Patterns of the Egyptian Pension ­ System before and after Recently Legislated Pension Reform and of Hypothetical DB Reform in Syria 289 5.15 Guiding Tree for Introducing Coverage Extension Strategies 296 5.16 A Framework for Implementing Social Protection Programs 298 5.17 Preferences and Perceptions of Informality among Workers in Lebanon and the Syrian Arab Republic 307 5.18 Implications of Rational and Behavioral Economics Theories Explaining an Individual’s Savings Behavior 313 Tables 1.1 Summary of Data and Definitions Used in This Report 38 1A.1 Sample Economies in the Informality Regression 67 1A.2 Definitions and Sources of Variables Used in Regression Analysis 70 2.1 Economic and Demographic Factors for Selected Non-GCC Countries 79 2.2 Composition of Employment Growth in the Services Sector for Selected Countries, 2000–2007 85 2.3 Informality Rates for Selected Countries 86 2.4 Informality Rates by Employment Status 94 2.5 Self-Employed versus Informal Wage Earners (Basic Characteristics) 95 2.6 Informality Rates by Sector of Employment 102 2.7 Marginal Increase in the Probability of Being “Informal” ­ according to the Characteristics of the Worker (Nonagricultural Employment Only) 104 2A.1 Description of the Data Used for the Micro-Analysis 114 2A.2a Informality Profile (All Workers) 118 2A.2b Informality Profile (Private Sector Workers) 120 3.1 Firm Size and Worker Informality 130 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Contents xiii 3.2 Benefits and Costs of Formality 136 3.3 Informality among Formal Firms: Matched Employer-Employee Data, Syrian Arab Republic, 2011 142 3.4 Informality, Connectivity, and Networks among Firms in MENA ­Economies 149 3.5 Correlation between Worker Underreporting and Perceptions of Labor Regulations and Tax Rates 151 3.6 Informal Employment in Micro-Firms in Morocco 157 3.7 Differences between Informal and Formal Salaried Workers in Micro-Firms in Morocco 160 3A.1 Informality Profile among Firms 166 3A.2 Underreporting, Firm Size, and Productivity 168 3A.3 Underreporting and Manager’s Education 170 3A.4 Underreporting, Regulations, Enforcement, and Corruption 172 3A.5 The Effect of Firm Size and Productivity on the Association between Regulations and Underreporting 174 3A.6 Description of Data Sources 176 3A.7 Descriptive Statistics on Moroccan Micro-Firms, Their Workers, and Their Owners 177 3A.8 Descriptive Statistics on Egyptian Micro-Firms, Their Workers, and Their Owners 179 4.1 Average Hourly Wages in Local Currency 190 4.2 Formality Premium, Lebanon: Ordinary Least-Squares (OLS) Estimates for Private Sector Formal and Informal Workers 200 4.3 Formality Premium, Syrian Arab Republic: Ordinary Least-Squares (OLS) Estimates for Private Sector ­ Formal and Informal Workers 201 4.4 Labor Market Transitions in Egypt (1998–2006) 205 4.5 Labor Market Transitions in Egypt (2008–2009) 206 4.6 Average Level of Earnings Increase When Moving from Formality to Informality 207 4.7 Individuals Who Want to Change Jobs 211 4.8 Differences in Job Quality between Formal and Informal Jobs 212 4.9 Job Conditions among Moroccan Youth 214 4.10 Preferences for Wage Earning versus Self-Employment ­(Lebanon, 2010) 216 4.11 Reasons Why Informal Workers Do Not Want to Contribute to Social Security 217 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 xiv Contents 4.12 How People Find Jobs 218 4.13 Relationship to Owner 219 4.14 Determinants of Informality: Networks in Morocco 220 4A.1 Estimating the Formality Premium in Lebanon 224 4A.2 Estimating the Formality Premium in Syria 225 4A.3 Estimating the Formality Premium in Egypt 226 5.1 Employment Regulations and Skills and Education of Labor as a Major or Severe Obstacle for Expanding ­Employment 239 5.2 Arrangements for Fixed-Term Contracts around the World in 2010 242 5.3 Firing Regulations around the World, 2010 248 5.4 Existence of Unemployment Protection Legislation around the World 251 5.5 Contribution Rates for Social Security Programs, 2009/2010, and Tax Wedge on Average Wages in Private Sector 253 5.6 Legal Coverage for Self-Employed Workers and Non-GCC MENA Economies Agricultural Workers in ­ 279 5.7 Pension Design Parameters of Mandatory Formal Sector Schemes in MENA Retirement Age, Minimum Vesting Period, and Early Retirement Provisions 282 5.8 Average Wage Measures in the Pension Benefit Formula in Selected MENA Countries 286 5.9 Existing Country Conditions Guiding Indicators (Score Card) 299 5A.1 Country Conditions in MENA Countries 316 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Preface Throughout the Arab World, many are striving for the opportunity to realize their potential and aspirations in a region that is rich in both human and physical capital. Although in several countries in the region economic growth has been sustained for several decades, this process has failed to create enough opportunities in the form of good jobs. This report addresses one of these potential margins of exclusion: informal employ- ment and the vulnerabilities and lack of opportunity associated with it. This report has two main objectives. The first is to provide an under- standing of the extent, determinants, and challenges posed by informal employment, bringing together new evidence, data, and country-specific analyses. The second objective is to open up and inform a debate on feasible policy options to better manage vulnerability through social security cov- erage extension, promote fulfilling work, and support an inclusive devel- opment process. The informality phenomenon in the Middle East and North Africa (MENA) region involves the vast majority of workers in the private sector, and so the evidence in this report suggests the need to rethink policy making, especially concerning labor markets and social security reform, because many of the intended beneficiaries operate beyond the reach of legislative reform. The observed patterns in MENA indicate that informal employment is consistently associated with overall worse conditions, lower pay, and lower productivity than formal employ- ment. This report analyzes the constraints that underpin the formal- informal divide and discusses policy options that can support a broad-based development process. This report was completed in the aftermath of the Arab Spring. The region is undergoing important institutional t ­ransitions—including contin- uous instability in some countries—that are complex and outside the scope of this report. This report focuses primarily on options for technical reform. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 xv xvi Preface At the same time it stresses the importance of an inclusive social dialogue open to a broader set of actors, including those whose voices were tradi- tionally not heard, such as youth, women, the unemployed, and informal workers. A widely accepted definition describes informality as an activity that is unregulated by formal institutions such as labor laws, registration, and taxation. Although many angles to informality can be identified, this report looks at informality through the human development lens and focuses particularly on informal employment. As such, the working definition adopted herein defines informality as a lack of social security coverage (usually understood as pensions, or if a pension system is not existent, as health insurance). The report aims to characterize the mag- nitude of the informality phenomenon in the MENA region compared to other regions of the world and, to the extent possible, its evolution over time and macroeconomic determinants. This is discussed in chap- ter 1. Regional and cross-country comparisons are useful, but a more country-specific analysis is needed to explain the ample heterogeneity observed. Thus the analysis in chapter 2 uses all available household- level datasets to create a detailed profile of informal workers. Informality among firms matters for job quality and workers’ access to risk-sharing mechanisms and is a key piece of this puzzle. An in-depth study of the dynamics of firms is beyond the scope of this work. However, data from firm-level surveys are used to describe the charac- teristics of informality in firms, both in micro-firms that operate out- side the regulatory framework as well as in larger formal firms that do not fully comply with social security contributions and taxes (­chapter 3). Data on earnings, job mobility patterns, attitudes, and self-rated job satisfaction are used to provide a full characterization of the quality of informal jobs (chapter 4). The resulting evidence corroborates the view that important segmentations exist across formal and informal job sec- tors, and that much of the observed informality is likely due to segmen- tation and exclusion factors, which in some countries are linked to the prominent role of the public sector as employer. Chapter 5 discusses determinants of informality and links them to options for reform. Although a “one size fits all” approach is clearly not suitable to address the heterogeneity of informality in the region, policy options in five complementary areas emerge as the key to promoting social security coverage extension, better job quality, and overall a more inclusive development process: (1) creating a level playing field for small and large firms to compete; (2) addressing regulatory barriers in labor Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Preface xvii ­ arkets; (3) realigning incentives, pay, and benefit packages in the pub- m lic sector; (4) redesigning key features of pension systems and e ­ xtending social security coverage; and (5) increasing access to skills-upgrading instruments for informal workers. Data availability defined the geographic scope of this report. With the exception of the evidence presented in chapter 1, which relies in part on standardized International Labour Organization data and thus allows for full coverage, this report focuses on the countries in the MENA region for which household-level data with information on social security cover- age were available: the Arab Republic of Egypt, Iraq, Jordan, Lebanon, Morocco, the Syrian Arab Republic, and the Republic of Yemen. With the exception of household data for Egypt and administrative data kindly provided by the Jordan Social Security Corporation, none of these sur- veys’ data track individuals over time. As a result, the scope to draw broad inferences at the regional level about the dynamics of informality as well as to identify causal determinants is limited. Moreover, some of the diag- nostics are based on data that are a few years old. However, the behav- ioral patterns identified are likely to continue to be valid, especially if linked to slow-changing institutional determinants. Additionally, with the exception of Egypt, only cross-sectional data were available for firms. The large majority of firm-level data are drawn from cross-sectional Investment Climate Surveys, which sample only registered firms. Data surveying informal/micro-firms were available only for Egypt and Morocco. To overcome these constraints, this project cosponsored four new sur- veys that were fielded in 2009 and 2010: the Egypt and Morocco youth surveys, the Lebanon Labor Force Survey, and the Syria matched employer-employee module. In these four surveys, questions were included about social security coverage, job characteristics, workers’ atti- tudes, willingness to contribute to social security, and the like. The Lebanon and Syria surveys also included innovative modules reporting the results of direct cognitive and noncognitive skills tests, which are linked to informality for the first time in this literature. Notwithstanding these important additions to the database for the report, data availability and access presented a considerable constraint. An important message emerging from this work is that generating more and better data, promot- ing open access to them, and especially investing in longitudinal surveys are likely to have large payoffs in terms of a deeper understanding of labor markets in this region. The lack of experimental evidence on how firms and individuals respond to incentives to formalize also imposes a limit on the ability to make inferences about the possible impact of different Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 xviii Preface intervention in the context of MENA. This is another area in which investments can have high returns. This report builds on and adds to the growing body of literature addressing the question of informality worldwide and in the MENA region. With its broad country coverage and intensive use of all regional household- and firm-level data accessible to the World Bank, this study provides a unique source for cross-country comparisons and benchmark- ing of informal employment in MENA. The use of newly collected infor- mation on workers’ attitudes, benefit structure, and working conditions provides a full characterization of the quality and features of informal jobs. Moreover, introducing direct measures of cognitive and noncogni- tive ability allows researchers, for the first time, to reduce much of the usually unexplained variation in determinants and returns to informality. By linking this systematic and innovative empirical analysis with a detailed description of constraints and regulatory barriers, especially those associated with labor markets and pension institutions, the report pro- vides a framework for policy options that can be adapted to the different country contexts. Because of the multifaceted nature of informality, this report touches upon many agendas, including labor market regulation, public sector reform, social insurance design, private sector dynamics, and skills upgrading. Although their in-depth treatment is clearly beyond the scope of this work, these issues are explored from the specific angle of their effect on informality. As such, this report complements numerous recent and ongoing analytical efforts, including the World Bank MENA Private Sector Flagship Report, From Privilege to Competition; the MENA Flagship Report, Jobs for Shared Prosperity: Time for Action in the Middle East and North Africa; Country Economic Memorandums; and country- specific labor market studies in the region. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Acknowledgments This report is the product of the collaborative effort of a core team led by Roberta Gatti and comprising Diego Angel-Urdinola, Joana Silva, and András Bodor (principal authors); Kimie Tanabe (data management); and May Wazzan, with important contributions from Mohamad Alloush, Carole Abi Nahed Chartouni, Anne Hilger, Arvo Kuddo, and Matteo Morgandi. Many colleagues have provided useful inputs at different stages of this report, including Surani Abeysekera, Rita Almeida, Luca Etter, Alvaro Forteza, Maros Ivanic, David McKenzie, Jaime Saenz, Amina Semlali, and Andrew Stone. Background papers were written for this report by Mohamad Alloush, Diego Angel-Urdinola, Mehdi Benyagoub, András Bodor, Carole Abi Nahed Chartouni, Yoonyoung Cho, Yohana Dukham, Mario di Filippo, Norman Loayza, Daniela Marotta, Shaumik Paul, Amina Semlali, Joana Silva, Kimie Tanabe, Tomoko Wada, and George Wellner. Amy Gautam edited the report. We are grateful to the participants of the consultations held in Cairo (September 2010), at the International Labour Organization (ILO)– World Bank labor market course in Turin (October 2010), and in Tunis (March 2011) for their insights into the nature of informal employment in the Middle East and North Africa region. We thank our peer reviewers, Ahmed Galal, William Maloney, and Milan Vodopivec, for their guidance; Gordon Betcherman and William Maloney for useful discussions; and Jean- Pierre Chauffour, Prof. Alia Abdel Monem El Mahdi, Sherine El-Shawarby, Caroline Freund, Santiago Herrera, Robert Holzmann, Ruslan Yemtsov, and the participants of the brainstorming workshops that were held at the World Bank and at the International Monetary Fund in Washington, DC, in October 2009, October 2010, and May 2011, for their valuable comments. Through this report, four new surveys were cosponsored (Egypt youth, Morocco youth, Lebanon Labor Force Survey, Syria employer-employee matched module). We acknowledge the productive collaboration to struc- ture the questionnaires with Gloria La Cava, Yaa Oppong, David Robalino, Haneen Sayed, and Tara Vishvanath. We are grateful to the Social Security Corporation of Jordan for sharing data on social security contributions. This work was supported by the Arab World Initiative. We thank Shamshad Akhtar and Steen Jorgensen for their strategic guidance. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 xix Contributors The authors of the chapters are as follows: Preface and Overview Roberta Gatti Chapter 1  Diego F. Angel-Urdinola with Anne Hilger and Surani Abeysekera Chapter 2  Diego F. Angel-Urdinola with Kimie Tanabe and Anne Hilger Chapter 3 Joana Silva Chapter 4 Roberta Gatti Chapter 5  András Bodor and May Wazzan with Arvo Kuddo and Roberta Gatti Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 xxi Abbreviations ALMP active labor market program CAPMAS Central Agency for Public Mobilization and Statistics DB defined benefit DC defined contribution ECA Europe and Central Asia EPL employment protection legislation ERF Economic Research Forum GCC Gulf Cooperation Council GDP gross domestic product ICA Investment Climate Assessment ID identification ILCS International Conference of Labor Statisticians ILO International Labour Organization IRA individual retirement account IV instrumental variable LAC Latin America and the Caribbean MDC matching defined contribution MEAS Micro- and Small Enterprises Survey MENA Middle East and North Africa MIMIC Multiple Indicator–Multiple Cause MSE micro- and small enterprise NDC notional defined contribution NGO nongovernmental organization NPS New Pension Scheme NSSF National Social Security Fund Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 xxiii xxiv Abbreviations OECD Organisation for Economic Co-operation and Development OLS ordinary least squares PAYG pay-as-you-go PF provident fund PMT proxy means testing RSBY Rashtriya Swasthya Bima Yojana SA South Asia SEA Southeast Asia SME small and medium enterprise SSA Sub-Saharan Africa SSC Social Security Corporation TPI third-party insurer UI unemployment insurance UN United Nations WAP working-age population Currency values are given in U.S. dollars unless otherwise indicated. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview Informality is a complex phenomenon, encompassing unpaid workers and workers without pension or health insurance coverage, small or micro- firms that operate outside the regulatory framework, and large registered firms that might decide to partially evade corporate taxes and social secu- rity contributions. The report looks at informality through a human devel- opment angle and focuses specifically on informal employment. In line with this approach, the working definition for informality adopted in the report is “lack of social security coverage” (usually understood as pensions, or if a pension system does not exist, as health insurance), which captures well the vulnerability associated with informal employment. Although a great deal of heterogeneity is seen across countries, informality is wide- spread, and some Middle East and North Africa (MENA) countries are among the most informal economies in the world. Evidence from this and other studies indicates that informal workers in most MENA countries are engaged in low-productivity jobs, more so than in comparator countries. Informal workers in the region are generally paid less for otherwise similar work in the formal sector, and their self-reported work satisfaction is low. Moreover, the data underscore the presence of important mobility barriers between formal and informal employment. Although extending social security coverage is likely to bring about welfare improvements in and of itself, countries will find it necessary to remove existing barriers to the creation of high-quality jobs to promote long-term inclusive growth and, thus, formality. Policies that can support this process include those that 1)  Create a level playing field for small and large firms to compete 2) Move toward labor regulations that promote labor mobility and provide support to workers in periods of transition ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 1   2 Overview Realign incentives, pay, and benefit packages in the public sector 3)  Reform pension and existing social insurance systems and introduce 4)  new instruments for coverage extension and Enhance the productivity of informal workers through training and 5)  skills upgrading. Background Economic growth has been sustained for many years precrisis in the region, but this has not resulted in the creation of an adequate number of jobs and has succeeded, at best, in generating low-quality, informal jobs. The report addresses one margin of exclusion: informal employment and the vulnerabilities and lack of opportunities associ- ­ ated with it. Informality in the MENA region involves the vast majority of work- ers outside of the public sector, and because many of the intended beneficiaries operate outside the reach of existing systems, the report’s findings point to the need to rethink policy making, especially with regard to those policies pertaining to labor market and social security reform. The report analyzes the constraints that prevent informal workers from becoming formal and discusses policy options to effec- tively address these constraints. The current transitions in MENA are accompanied by important challenges, but they also provide new opportunities to improve living standards and equity. Achieving this will require, above all, a new and inclusive social dialogue open to a broader set of actors, particularly to those whose voices have traditionally not been heard, such as youth, women, the unemployed, ­ and informal workers. This Overview first describes the magnitude of the informality phe- nomenon in the MENA region relative to other regions of the world, its macroeconomic determinants, and, to the extent possible, its evolu- informal tion over time. The first section provides a detailed profile of ­ workers in the region. The next section describes the characteristics of informality in micro-firms that operate outside the regulatory frame- work and in larger firms that do not fully comply with social security contribution requirements and tax obligations. This Overview then discusses earnings, job mobility patterns, attitudes, and workers’ self- rated job satisfaction to provide a better understanding of the quality of jobs in the informal sector. The evidence corroborates that important segmentations exist between formal and informal jobs, and that much Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 3 of the observed informality is likely due to barriers to mobility into better jobs. In some countries, these barriers are linked to the promi- nent role of the public sector as the main employer in the formal sector. The last section discusses policy options for effectively expanding cov- erage of health insurance and pension systems and promoting the cre- ation of better quality jobs. Understanding Informality This report looks at informality through a human development angle and focuses particularly on informal employment. Informality is a complex phenomenon, comprising unpaid workers and workers without social security or health insurance coverage, small or micro-firms that operate outside the regulatory framework, and large registered firms that might partially evade corporate taxes and social security contributions. Three indices are commonly used to measure informality: (1) the Schneider Index,1 which uses a broad set of country correlates to estimate the share of production not declared to tax and regulatory authorities, (2) the share of employed workers without social security coverage, and (3) the prevalence of self-employment. Defining informality as “lack of social security coverage” captures well the vulnerability associated with ­ informal employment and, as such, is the working definition in this analysis. Informality is a fundamental characteristic of underdevelopment and has costs and benefits. As depicted in figure O.1, a larger informal sector is often associated with lower gross domestic product (GDP) per capita. For most countries in the region, the share of informal to total employ- ment is broadly aligned with the level that their economic development would predict. Informal employment is associated with costs and negative externalities that may be a source of economic slowdown. For ­ example, a large informal sector implies that a heftier tax burden will fall on the formal sector. This may hold back new and productive firms, precisely those firms that have the potential for driving growth in a ­ dynamic economy. In addition, informal activities use and congest public infrastructure without contributing to the tax revenues necessary to maintain them. Moreover, workers involved in the informal sector often lack a written contract regulating their work relationship and/or pension and health insurance coverage. This exposes them to significant occupa- tional risks, including a potentially less safe and decent work environ- ment, as well as insufficient mechanisms to cope with risks related to Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 4 Overview Figure O.1 Informality and Economic Development % of workers not contributing to a pension scheme 100 NER GNB CAF NPL NGA SLE RWA BGD GMBSEN ZMB LSO KENPAK GHA UGA VNMIND YEMBTN BOL DJI IDN NICHNDPRY 80 SYR WBG MAR JAM CHNGTMCOL PHLECU PER SLV THA MDV CPV DOM VEN KGZ JOR IRN LBN MEX ARG LKA GEO AZE KAZ BIH DZA 60 VCT YUG GUY ALB TUN TUNTUR MUS MKD BRA URY EGYROM CRI CHL TTO 40 MDA MNG ARM DMA MYS LCA LBY UKR GRD SGP BLZHRV HUN BRB SVK 20 LTU KOR GRC POL CZE ATG DEU AUT FIN FRA CAN ISL SWE IRL EST PRT LVA BRN ITA ESP NLD AUS GBR NOR USA KNA BHR NZL DNK BEL JPN 0 CHE 4 6 8 10 12 GDP per capita, constant 2000$, in ln 60 CMR % of self-employment in total employment IDN ETH MDG VNM HND KHM PRY COL DOM 40 BGD ZMB BOL PAKNIC GTMPER PHL SYR IRN JAM ECUTHAMDV DMA YEM LKA MNG CPV SLV VEN LBN LCA DZA PAN GRC WBG EGY MAR BLZ BRACHL MEX URY KOR NAM CRI FJI TUN ARGPRT ITA 20 MYS TTOCYP TWN ZAF MUS NZL IRL ESP CAN MLT PRI BRB IMY SGP BEL ISL CHE OMN ISR AUS GBR NLDSMR HKG AUT FRA DEU JPN MAC ABW DNK USA NOR LSO 0 4 6 8 10 12 GDP per capita, constant 2000$, in ln Source: Loayza and Wada 2010. Note: GDP = gross domestic product. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 5 old-age and health problems. Nevertheless, informal employment has benefits for the labor market and population, because it provides employ- ment to a significant portion of the population, may be a source of inno- vation and entrepreneurship (many firms start operating informally when they are small and formalize as they grow), and serves as a safety net in periods of transition (such as during times of economic downturn). Informality can be the result of exclusion or of a rational choice by entrepreneurs or workers who decide to opt out of the formal labor ­ market. Traditionally informality has been considered the result of invol- untary exclusion of workers due to high and distortive regulation (De Soto 1989). Other factors can also affect workers’ and firms’ deci- sions to work informally, including myopia (that is, lack of awareness about retirement needs and health risks) and limited information, both of which might be associated with exclusion margins, such as geographic location and low human capital (that is, poor people in remote areas and/ or uneducated agricultural workers may not have access to social security systems but would join them if they could). Some workers and firms decide to opt out of the formal economy as a result of a rational cost- benefit evaluation (Perry and others 2007). These reasons for informality are not mutually exclusive and likely coexist in a continuum. In all cases, widespread informality implies high vulnerability, and so a clear rationale for government intervention can be found. On the one hand, when involuntary exclusion from formal employment is predominant (for ­ example, small firms can survive only if they avoid paying taxes, or ­ salaried workers would like access to social security but cannot find jobs with coverage), then policies that reduce existing barriers are needed to improve welfare. Such policies might include rationalizing regulations to create a level playing field for small and large firms to compete fairly, as well as tackling labor market reform. On the other hand, when more workers and firms voluntarily opt out of the regulatory system, policy interventions that rebalance the perceived costs and benefits of participat- ing in the formal sector are needed. Such policies can focus on improving the quality and outreach of key public services, enhancing communication and transparency about social services, and redesigning and extending instruments for better m­ anagement of health- and age-related risks. Informality in MENA: Levels and Trends The MENA region displays lower employment and higher unemploy- ment rates than any other region in the world. Although economic Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 6 Overview growth was sustained in the years before the 2008 financial crisis and employment growth was positive (and particularly high for women), high population growth has continued to put pressure on labor markets to absorb an increasing number of new entrants. Moreover, in the last decade, jobs were created mainly in low value-added service sectors, most of which are associated with high rates of informal employment, such as construction, commerce, and transport. This process may have prevented the benefits of economic growth from being shared fully among poorer segments of the population and has led to increasing informality. A typical country in MENA produces about one-third of its GDP and employs 65 percent of its labor force informally. This means that around two-thirds of all workers in the region may not have access to health insurance and/or are not contributing to a pension system that would provide them with income security after retirement. From a fiscal per- spective, about one-third of total economic output in the region remains undeclared, with considerable implications for government revenue. As illustrated by figure O.2, the typical MENA country is more informal than the typical developed country and countries in Europe and Central Asia (ECA) when considering the percentage of labor for not contribut- ing to social security and self-employment. It is, however, less informal than the typical country in Latin America and the Caribbean (LAC), East Asia and Pacific (EAP), South Asia (SA), and Sub-Saharan Africa (SSA). Important variations are found in the prevalence of informality across non-GCC countries, depending, among other factors, on the availability of natural resources and labor and on the size of the public sector. Non-GCC countries are quite heterogeneous in terms of size, availability of resources and labor, economic development, and demographic structure, all factors that influence the size of the informal economy, which, as a result, varies significantly (see figure O.3 and chapter 2 for a more thorough discus- sion). In general, resource-rich/labor-abundant economies (such as the Islamic Republic of Iran and the Syrian Arab Republic) tend to display high informality rates in the region as proxied by the share of the labor force not contributing to social security (between 66 and 73 percent) and by the share of self-employment to total employment (between 32 and 54 percent). However, their share of undeclared output to total GDP (at around 20 percent) is comparable to that of GCC countries. This occurs because these countries generally have few, but large, formal firms (many in the energy sector) that are capital intensive, resulting in lower informal- ity in production than in labor (Loayza and Wada 2010). On the contrary, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 7 Figure O.2 Informality in MENA Compared to Other Regions % of the labor force not contributing to social security MENA, non-GCC countries 65.0 developed countries 9.3 33.2 ECA LAC 73.6 EAP 75.0 SSA 93.2 0 20 40 60 80 100 self employment (% of total employment) MENA 32.4 MENA, GCC countries 3.8 MENA, non-GCC countries 36.5 developed 13.4 ECA 22.7 LAC 33.1 EAP 56.3 SA 71.1 SSA 80.8 0 20 40 60 80 100 undeclared output as % of GDP (Schneider Index) MENA 27.7 MENA, GCC countries 20.1 MENA, non-GCC countries 36.2 developed countries 16.6 EAP 32.8 SA 37.0 ECA 38.9 SSA 42.2 LAC 43.9 0 20 40 60 80 100 Source: Schneider et al., 2010 for Schneider Index; WDI for self employment and pension scheme. Note: EAP = East Asia and Pacific; ECA = Europe and Central Asia; GCC = Gulf Cooperation Council; LAC = Latin America and the Caribbean; SA = South America; SSA = Sub-Saharan Africa. The periods covered are the latest years in 2000–10 for pension scheme, 2000–11 for self- employment, and 1999–2007 for the Schneider index, respectively. Data on social security contributions in GCC countries are only available for Bahrain, 2007 ­ (20 percent), and Qatar, 2008 (4.4 percent) (Palleres-Miralles, Romero, and Whitehouse 2012). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 8 Overview resource-poor/labor abundant economies (such as the Arab Republic of Egypt, Tunisia, and Morocco) display a high share of undeclared output (between 35 and 40 percent of GDP) and a lower share of the workforce ­ ecurity (between 45 and 76 percent), which is not contributing to social s consistent with a higher share of medium-size (and semiformal) labor- intensive firms and, in some cases, significant public sector employment. Informality in MENA has been rising in recent years. Data indicate that informality has been increasing rapidly in the entire developing world, as proxied by the Schneider Index and the share of self- employment to total employment. The increase in the share of unde- clared/informal output to total GDP in recent years has been a global trend (figure O.4). Who Are Informal Workers? Informality is highest among the working poor. Overall, in MENA infor- mality decreases as wealth increases. In countries such as Lebanon, the share of informal employment is significantly lower among the wealthiest segments of the population. Nevertheless, in some MENA countries, informality is so widespread that it remains significant even among the wealthier segments of the population. In the Republic of Yemen and Syria, for instance, more than two-thirds of all workers who belong to the richest households work informally (figure O.5). In the Arab Republic of Egypt, the Republic of Yemen, and, to a lesser extent, Morocco, an inter- esting pattern emerges, whereby informality decreases with income/ wealth but unemployment increases with income/wealth. This ­ correlation suggests that in these countries, programs targeted at informal workers would be addressing the working poor, but interventions targeted at the unemployed (although a clearly visible and often vocal category) would reach those who can afford to be unemployed, that is, those who are relatively better off. Over the life cycle, in some countries an important transition is made from informal employment into public sector employment as young individuals reach prime-age adulthood. Figure O.6 illustrates employment patterns by age for urban workers in a selected group of countries. Informality rates are very high among youth between the ages of 15 and 24. After age 24, informality decreases rapidly until individuals reach 40 to 45 years of age. After age 45, informality rates fall to between 20 and 30 percent. This rapid decrease in informality rates goes hand in hand with a rapid rise in public employment, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 9 Figure O.3 Informality Rates for Selected Non–Gulf Cooperation Council Members % of the labor Libya 31.5 force not contributing Egypt, Arab Rep. 44.9 to social security Tunisia 51.4 Jordan 61.6 Algeria, Rep. 63.3 Iraq 64.4 Lebanon 65.5 Iran, Islamic Rep. 65.8 Syrian Arab Republic 73.2 Morocco 76.2 West Bank and Gaza economy 86.0 Yemen, Rep. 89.6 self-employment Lebanon 15.6 (% of total Yemen, Rep. 32.4 employment) Tunisia 34.2 West Bank and Gaza economy 35.6 Morocco 37.3 Egypt, Arab Rep. 37.8 Iran, Islamic Rep. 46.8 Syrian Arab Republic 53.9 Algeria, Rep. 67.1 Syrian Arab Republic 19.5 Iran, Islamic Rep. 19.5 undeclared output as % of GDP Jordan 20.3 (Schneider Index) Yemen, Rep. 27.7 Lebanon 35.1 Egypt, Arab Rep. 35.3 Algeria, Rep. 35.7 Libya 36.7 Morocco 37.9 Tunisia 39.6 0 20 40 60 80 100 Source: Processed from Loayza and Wada 2010. Note: Time periods are as follows: Schneider Index, average 1999–2007; self-employment, 1999–2007; not contributing to social security (S.S.), 2000–2007. ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 10 Overview Figure O.4  Annual Growth Rates of Informality annual growth rate, Schneider Index (2000–2007) LAC 1.39 ECA 0.76 MENA 0.69 SSA 0.64 developed 0.62 countries SA 0.55 EAP 0.24 0 0.5 1 1.5 annual growth rate of self employment (2000–2011) SSA 7.2 SA 7.2 EAP 5.2 MENA 2.1 OECD –1.5 LAC –1.8 ECA –4.3 –6.00 –4.00 –2.00 0.00 2.00 4.00 6.00 8.00 Source: Processed from Loayza and Wada 2010. Note: EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; SA = South America; SSA = Sub-Saharan Africa. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 11 Figure O.5 Informality Rates by Quintile of Per Capita Consumption for Selected Countries 100 95.9 90 86.1 82.8 82.5 80 70 60 58.7 percent 50 40 35.1 30 20 10 0 Yemen, Rep. Syrian Arab Republic Lebanon poorest Q2 Q3 Q4 richest Source: Angel-Urdinola and Tanabe 2011. Note: Q = quintile. s ­ uggesting that a large portion of younger informal workers enter into public sector jobs as they become adults. These trends are very differ- ent from those observed in Latin America. Still, one has to be cautious in interpreting these results as they are likely to reflect vintage effects since, especially in countries such as Egypt, more public sector jobs were available to earlier cohorts of workers. For example, in countries such as Mexico, although informality rates also decrease with age, the observed transition occurs not between informality and public employ- ment, but between informality and self-employment, with many young individuals becoming entrepreneurs and thus contributing to a more dynamic creation of employment in the private sector (Perry and others 2007). Lower levels of education are strongly and linearly associated with higher rates of informality in most countries, an association that seems to be driven primarily by more educated workers joining the public sector (particularly in Egypt, Jordan, Morocco, and Syria). Informality rates among workers who completed primary and/or basic education (account- ing for at least 50 percent of all the employed in most countries in the region) are generally much higher than among those workers who com- pleted secondary and/or tertiary education. Consistent with the transi- tions depicted in figure O.6, differences in informality rates by age and education are mainly due to the life-cycle movement of older and more Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 12 Overview Figure O.6 Employment Status by Age for Selected Countries, Urban Areas Only Egypt, Arab Rep., 2006 70 employment rate (%) 60 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group Jordan, 2010 70 60 employment rate (%) 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group Iraq, 2007 70 employment rate (%) 60 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group public private formal private informal self-employed figure continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 13 Figure O.6  Employment Status by Age for Selected Countries, Urban Areas Only (continued) Yemen, Rep., 2006 70 60 employment rate (%) 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group Morocco (Urban), 2010 70 60 employment rate (%) 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group public private formal private informal self-employed Source: Angel-Urdinola and Tanabe 2011. educated workers from informal employment to public sector e ­ mployment. In countries where the bulk of formal jobs are in the public sector (such as the Republic of Yemen and Iraq), differences in informality rates across education levels in the private sector are negligible (figure O.7). Informality is more prevalent among workers in small firms. In the private sector, informal workers are mostly employed in small firms, with a distribution that is particularly skewed toward firms with fewer than five workers and that are most likely engaged in low-productivity activities (figure O.8). Controlling for other observable characteristics, ­ workers in medium-size (10 to 50 workers) and large-size (more than Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 14 Overview Figure O.7 Informality Rates by Highest Educational Level Completed all workers 100 97.4 90.9 90 86.2 82.2 77.8 75.7 80 71.9 70 60 percent 50 40.7 40 36.5 28.4 30 25.4 23.7 24.3 22.0 20 10 0 Iraq Jordan Egypt, Lebanon Morocco Syrian Arab Yemen, Arab Rep. (Urban) Rep. Rep. private sector workers 99.4 97.5 99.9 100 93.5 95.5 94.9 88.2 88.3 90 80.4 80 70 61.9 60 50.5 50.7 percent 50 40.2 40 34.3 30 20 10 0 Iraq Jordan Egypt, Lebanon Morocco Syrian Arab Yemen, Arab Rep. (Urban) Rep. Rep. primary or below preparatory/sec. general secondary vocational tertiary education Source: Angel-Urdinola and Tanabe 2011. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 15 Figure O.8 Informality and Firm Size 100 90 proportion of firms by size (%) 80 70 60 50 40 30 20 10 0 al al al al al al al al rm rm rm rm rm rm rm rm fo fo fo fo fo fo fo fo in in in in Egypt, Arab Rep. Jordan Lebanon Morocco <5 workers 5–9 workers 10–50 workers >50 workers Sources: Lebanon Labor Force Survey 2010; Morocco Youth Survey 2010; Egypt Labor Market Panel Survey ­(ELMPS) 2006. Note: In Morocco 10–50 workers actually incorporates 10–100 workers, and >50 captures firms with 100+ employees only. 50 workers) firms are, respectively, 16 to 21 and 17 to 53 percent less likely to work without social security coverage than workers in small-size (five to nine workers) firms. The relationship between gender and informality is quite complex. It should be noted that despite positive improvement in female employ- ment rates in the last decade, labor force participation among women in the MENA region remains the lowest in the world, and among women who participate, unemployment is quite high, especially among the most educated. In countries such as the Republic of Yemen and Morocco, where agricultural employment constitutes an important share of overall employment, being a woman—other things being equal—is associated with a higher probability of working in the informal sector, because women are often employed in unpaid or subsistence agriculture. When public employment constitutes a significant share of overall employment, as in Egypt, Iraq, and Syria, being a woman is associated with a lower probability of working in the informal sector, because women who par- ticipate in the labor forces in these countries (generally those with higher Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 16 Overview working levels of education) self-select into public sector jobs, which have ­ hours, benefits, and overall safety characteristics that are perceived as more desirable. Informality among Firms Firms can operate informally along various margins. For wage workers, working informally is the result of factors that go beyond individual skills and attitudes and relates to labor demand features such as the types and availability of jobs and workers’ relative bargaining power with the hiring firm. Hence, understanding why firms operate informally is an important dimension of informal employment. In particular, small firms might not register their activity to avoid paying taxes and/or to avoid paying social security contributions for their workers (in full or in part). Registered firms (usually larger firms) might only partially report their sales and underreport workers and/or workers’ wages to avoid paying taxes and/or social security contributions. Informality is prevalent among firms in MENA. In the region many firms never formalize, and even those that formally register still operate informally for a significant amount of time. Among existing formal firms, MENA has the highest share of firms starting up in the informal sector (25 percent compared to less than 10 percent in the LAC and ECA regions). Also, firms in the region have, on average, a longer operating period without formalization than other developing regions (four years versus less than one year in the LAC and ECA regions; see figure O.9). Small enterprises, which account for a large share of all private sector jobs, are mostly unregistered and often employ workers informally. Among larger registered firms (with 10 or more employees), about one- fifth of sales and workers are not reported. A great deal of heterogeneity exists in the region, but high taxes are consistently identified by entre- preneurs as the top constraint to formalization, together with the cost of registration, bureaucratic complexity, and administrative requirements. In many economies, including Algeria, Djibouti, Iraq, and the West Bank and Gaza economy, there seems to be ample room for simplification of business entry regulations and procedures. A strong association emerges between informality and low productiv- ity among firms. When firm informality is measured as the share of workers and sales that are underreported for social security and tax purposes, respectively, firm size, low productivity, and a manager’s education seem to be its key determinants. Although the available data ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 17 Figure O.9  Unregistered Firms, by Region 30 5 25 24.4 4.06 4 number of years 20 16.13 3 percent 15 2 10 7.46 0.9 0.95 5 1 0.54 3.45 0 0 Middle East and East Asia and Latin America Eastern Europe North Africa Pacific and the Caribbean and Central Asia % of firms that started operations without formal registration number of years firms operated without formal registration Source: World Development Indicators Dataset. do not allow identification of a causal relationship between firm infor- mality and the aforementioned determinants, these patterns are observed consistently in other regions. For instance, firm size has a significant association with informality and productivity: In Morocco, data suggest that labor productivity in smaller firms (nine employees or less) is one- half the productivity of firms with more than 100 employees. In addition to affecting productivity, firm size is strongly associated with ­ underreporting wages to the social security administration. Specifically, among otherwise similar firms (for example, similar sector, product, or years in operation) doubling firm size is associated with a significant reduction in worker underreporting (such as by 10 percent in the Republic of Yemen and by 35 percent in Jordan). The education level of managers and workers is also a strong predictor of underreporting workers to social security and/or tax authorities. ­ A large education gap exists between managers of informal and ­ formal firms. Consistent with patterns in other countries, informal micro-firms in MENA typically have managers and a workforce with lower levels of education. For instance, among small firms in Egypt, the managers who have completed secondary school or have attained share of ­ university education is about 45 percent in informal firms versus Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 18 Overview 85 ­ ­ ormal firms (figure O.10). Workers in informal firms also percent in f tend to be less educated. Indeed, insufficient skills and education of the working force are perceived as serious constraints to employment and output growth in informal firms. Data from matched employer- employees surveys for small informal firms show that informal salaried workers are younger, are more likely to work longer hours, and earn a relatively lower hourly wage than formal workers. Informality in MENA: Exclusion or Choice? Workers in MENA face low mobility across informal and formal jobs. In instances where the data allowed for analysis, mobility from informal to formal jobs was found to be extremely limited. For example, between 2008 and 2009, an informal worker in Egypt had a 4 percent chance of moving to a private sector formal job and a 5 percent chance of moving to a public sector job (see figure O.11). The implied average duration of informal salaried work is about three years in Egypt and job ­ Figure O.10  Highest Level of Education of Managers in Formal and Informal ­Manufacturing Firms in Egypt, 2009 100 90 80 70 60 percent 50 40 30 20 10 0 informal formal formal formal small medium large did not complete primary/did not enter school completed primary preparatory/incomplete secondary completeted secondary (incl. vocational) some university degree or higher Source: World Bank 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 19 Figure O.11 Transitions of Originally Informal Salaried Private Sector Male Workers 100 90 80 70 60 percent 50 40 30 20 10 0 1998–2006 2008–2009 unpaid self-employed public informal salaried private formal salaried private unemployed out of labor force Source: Silva and others 2010. is reported to be about four and one-half years in Lebanon and Syria, both longer than the duration observed in countries such as Mexico, where workers move out of informal salaried jobs within two years (see Maloney 1999). Monetary returns to skills differ substantially across informal and formal wage employment. Informal workers earn lower salaries than formal workers with similar skills: The estimated premium associated with formal jobs varies from 10 percent for all workers (15 to 65 years of age) in Syria to more than 50 percent among Moroccan youth (15 to 34 years of age). The formality wage gap persists even when Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 20 Overview differences in individual characteristics (for instance, age, education, and measured ability) and firm type are explicitly accounted for (see ­figure O.12). Most measures of job quality, including amount of annual leave and other benefits as well as access to training, suggest poorer working ­ conditions in informal jobs. This is also reflected in direct and indirect measures of job satisfaction. For example, Moroccan youth working in informal jobs reported being significantly less satis- fied than youth w ­ orking in formal jobs. In Egypt and Lebanon, infor- mal workers are uniformly more likely than formal workers to want to change jobs or to be searching for a new job. In the absence of well-developed public and private formal labor intermediation systems, social networks matter for finding jobs. More than 70 percent of workers in Lebanon and Syria reported having found jobs predominantly through personal connections. This is even more common in the informal sector. Such a strong reliance on personal social networks (which are often formed within homogenous socioeconomic strata) is likely to limit the size of the talent pool from which firms can select their workforce. There is also evidence that networks matter inter- generationally for finding employment. All else being equal, the chance Figure O.12 Estimated Formality Premium by Gender in Different Countries 80 70 60 50 percent 40 30 20 10 0 Egypt, Arab Morocco Lebanon, 2010 Syrian Arab Rep., 2006 (youth), 2010 Republic, 2010 total males females Sources: Angel-Urdinola and Tanabe 2010; Alloush and others 2011. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 21 of a worker holding a formal job in Morocco increases significantly if his or her father also has a formal job. Low mobility across informal and formal jobs, persistent wage gaps, and low self-reported satisfaction in informal jobs all suggest the pres- ence of labor market segmentations.2 These patterns dispel the notion that informal jobs in MENA primarily reduce asymmetric information for new entrants to the labor market. Instead, they support the view that segmentations and important margins of exclusion exist along the formal and informal sectors, implying that relevant labor market out- comes (including unemployment rates and earnings) are not arbitraged across the two sectors. In some countries, such as Egypt, these differences are predominantly driven by the public/private sector ­ employment divide. Barriers to Coverage and Policy Options The observed patterns indicate that economies in the MENA region gen- erate a limited number of good jobs, with a large and growing i ­nformal sector characterized by overall worse working conditions and lower pro- ductivity than in the formal private sector. A complex set of institutional barriers and distorted incentives account for this disequilibrium. As a result, promoting a more inclusive growth process, with the creation of decent jobs and reduced vulnerabilities, will require a set of coordinated policies spanning the business environment, labor market regulations, civil service reform, pension system design, and skills-upgrading interven- tions. Importantly, the observed high prevalence of informality indicates a need to rethink policy making. A vast majority of economic agents oper- ate outside the realm of regulation, and until recently in many countries, those outsiders (including the unemployed, youth, women, and informal workers) may not have been adequately represented in the social dia- logue and consensus-building process around reform. Five key distinct and complementary policy options, each linked to the drivers of informality in the MENA region, are identified: (1) creating a level playing field for small and large firms to compete, (2) moving toward labor regulations that promote labor mobility and provide income support to workers in periods of employment transition, (3) realigning incentives, pay, and benefit packages in the public sector, (4) reforming pension and existing social insurance systems and introducing new instru- ments for coverage extension, and (5) enhancing the productivity of informal workers through training and skills upgrading. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 22 Overview Designing Reforms that Foster Competition Although MENA countries have moved progressively away from state- led growth, their private sectors, with few exceptions, still suffer from limited dynamism and stagnating private investments. Even if private sector reforms have accelerated in recent years (as measured by Doing ­ Business indicators; International Finance Corporation and World Bank 2012), policy gaps still remain. In particular, high taxes and compliance costs are consistently seen as binding constraints to firms’ growth and formalization. Importantly, policy uncertainty and discretion in imple- menting the law constrain investment and reinforce the role of existing firms, who have a vested interest in protecting their rents and limiting competition (World Bank 2009). In such an environment, the costs of joining the formal economy are often prohibitively high for small enter- prises that opt to operate “below the radar” and consequently face limited opportunities to grow, exploit economies of scale, or connect to larger markets and access know-how. Informality appears to be intertwined with the development process and can be largely explained as the private sector’s response to restrictive regulation. It should be noted that informality is preferable to a fully formal economy where firms or workers are unable to comply with or are able to circumvent regulations. Policies solely based on strict enforcement of regulations are likely to increase formalization in the short term but are also likely to increase unemployment and could eventually reduce growth. Instead, formalization would be better achieved as the result of a development process whereby growth becomes more inclusive, incen- tives to create more good jobs are fostered, and rents are more evenly distributed among the outsiders (informal workers and the unemployed). This process can begin by improving the regulatory framework for busi- nesses, including simplifying regulations, decreasing barriers to entry (such as business registration and high corporate taxes), and consistently enforcing regulations to induce firms to compete fairly. Reforms to entry regulations have been shown to have positive, albeit moderate, effects on formalization. Possible reforms options include (1) reducing the costs of registration and the number of procedures and minimum capital requirements, (2) providing information on procedures and benefits of becoming a formal business and training entrepreneurs (for example, on filling forms), and (3) facilitating registration by estab- lishing one-stop shops for registration. Although no evidence is at hand on the likely impact of these reforms in the MENA region, relevant inter- national data are available. For example, in Mexico, simplifying the Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 23 ­ rocess of business registration had a moderate impact on formalization p of existing firms, but formalization was substantially increased by creation of new businesses by former wage earners. Overall, the evidence suggests that it might be difficult to promote formalization for a large number of existing informal firms without a change in the associated benefits or without reducing other compliance costs such as taxation. Small firms in Egypt and Morocco identify taxation as a significant constraint to formalization. Lowering the corporate tax rate can impact tax revenue through three main channels: (1) existing formal firms may invest more and earn more income on which they pay taxes, (2) existing informal firms may be induced to formalize and start paying taxes, and (3) new firms may be induced to operate formally. Evidence from other regions suggests that the net effect on tax revenue is likely to depend on whether a reduction in the tax rate is accompanied by additional enforce- ment and a reduction in exceptions. In many MENA countries, investment climate reforms have acceler- ated in recent years. However, consistent implementation of reforms matters greatly for their success. According to recent evidence (World Bank 2009), as a consequence of reforms related to the investment ­ climate, private investment in the MENA region increased by only 2 ­percent of GDP, compared to between 5 and 10 percent in Asia, Eastern Europe, and Latin America. The same report estimates that the number of registered businesses per 1,000 people in MENA is less than one-third of that reported in Eastern Europe and Central Asia, and the average business in MENA is 10 years older than that found in East Asia or Eastern Europe (suggesting slower firm entry and exit). Close to 60 ­percent of business managers surveyed did not think that the rules and regulations were applied consistently and predictably, and policy uncer- tainty, unfair competition, and corruption were identified as major ­ concerns for investors. Reforms are unlikely to produce the expected results without a credible commitment to more transparency and fairness in the relationship between businesses and the government. Discretionary enforcement of regulation acts as a strong deterrent to small entrepre- neurs who start their businesses informally but are then forced to stay small to escape controls. Staying small may, in turn, make it prohibitively costly to formalize over time. Realigning Incentives, Pay, and Benefits in the Public Sector In some countries the public sector continues to employ a large share of workers. For example, in Egypt, Iraq, and Syria, up to one of every three Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 24 Overview workers is employed by the public sector. Public sector jobs often offer higher wages than private sector jobs, as well as more job security and generous social security coverage (Bodor, Robalino, and Rutkowski 2008). Such advantageous conditions are, not surprisingly, very attractive to workers. In a recent survey, 70 percent of youth in Egypt indicated that “it is best to work in the public sector.” In Syria, only a quarter of those officially in the queue for public employment accepted a recent offer of training and private sector placement, because registering in social security rosters would imply removal from the public sector queue. Many of these workers then enter into informal employment as a means of supporting themselves while waiting for a public sector job. The generosity of public sector employment conditions (including pay, benefits, and job security) thus contributes to higher informality and important segmentations, making the need for civil service reform even more pressing. Moreover, ­ the prospect of obtaining stable and remunerative employment in the public sector distorts skills formation, because a number of workers acquire education (especially in the humanities) that is not necessarily in line with private sector demands. Reforming the civil service is a key priority. In the short run, elimi- nating institutionalized queuing for public sector employment can reduce the segmentation between public and private employment and foster labor mobility. Queuing for public sector jobs contributes deci- sively to unemployment among graduates in the region and fosters informality. Moreover, wage scales in the public sector currently reward education level and tenure, but not performance. In several MENA countries, placement depends on connections and is virtually irrevers- ible, because dismissing a civil servant is a difficult and complex process. Linking the public sector wage scale to competence and performance and easing the rigidity of civil service contracts may reduce the gap between public and private sector employment, thus limiting the ­ incentive to work informally while queuing. Moving toward Labor Regulations That Promote Labor Mobility and Provide Support to Workers in Periods of Employment Transition Labor market regulation is perceived as an important constraint by more than one-third of employers in MENA, the highest share among all regions in the developing world. Employment protection legislation (EPL) includes a wide array of regulatory items, such as minimum wage, hiring and firing regulations, and provisions for fixed-term contracts. In MENA, legislated minimum wages are generally low, are not well Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 25 enforced, and do not constrain employment outcomes significantly. However, in some countries, such as Tunisia, a centralized and rigid wage-setting process may significantly contribute to informality and higher unemployment. Moreover, in some countries in the region, firing regulations are extremely restrictive, requiring notification and approval of a third party for a single worker’s dismissal. In parallel, severance pay is often generous. Such strict EPL reinforces segmentations between insiders (workers with formal, protected jobs) and outsiders (the unem- ployed and informal or fixed-term workers), because many firms ­ circumvent regulations and resort to employing informal workers to business needs. fulfill their ­ MENA countries might improve compliance and employment out- comes by easing some labor legislation provisions. Although certain pro- visions in labor legislation in some MENA countries might be rigid in theory, in reality they are widely evaded. A shift toward a richer and more flexible set of labor contracts (including more fixed-term contracts and fewer open-ended contracts) would provide opportunities for young workers and new entrants to join the formal sector through flexible work- ing arrangements with social insurance coverage. Policy reforms that ease regulations and make them more realistic to comply with should be supplemented with social protection system reforms that better protect the income position of workers and their employment transitions.3 For example, recent experience shows that moderately strict EPL, when com- bined with a well-designed system of unemployment benefits and a strong emphasis on active labor market programs, can help create a dynamic labor market while also providing adequate employment secu- rity to workers (OECD 2008). Adequate safety nets could also play an important role in protecting workers from sudden job loss, helping workers transition between jobs, and preventing more people from ­ ­ slipping into poverty. The newly legislated unemployment insurance schemes in Jordan and Egypt provide an example of reforms that should be considered by other countries in the MENA region. Keeping the cost of labor at a realistic level via affordable social secu- rity contributions and relaxing wage rigidities are likely to reduce infor- mality. In general, institutionalized minimum wages in the MENA region are neither high (with the exception of Morocco) nor binding. Yet cen- tralized wage-setting mechanisms, as in the case of Tunisia, contribute to informality by setting artificially high wage floors for certain occupations and skill levels. In addition to keeping minimum wages at low levels that can be realistically enforced, wage-setting mechanisms in the MENA Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 26 Overview region should be anchored to changes in productivity. In countries where minimum wages are high (whether economy wide or sector specific), and where it is not politically feasible to reduce them, governments could consider reducing the minimum wage for youth to improve transitions of new labor market entrants into formal employment, while maintaining higher minimum wages to protect well-established workers. Similarly, labor taxes in MENA are not high by international standards, with the exception of Morocco and Egypt. The wedge between the employer’s labor cost and the worker’s take-home pay (that is, the tax wedge) could be further reduced through social insurance reforms that lower the social security contribution rates or by shifting a portion of the labor taxes toward other general revenue sources, such as consumption or property taxes. Engaging in a more inclusive social dialogue is key to sustaining these reforms. There are important political economy aspects to labor market reform. In particular, the traditional tripartite structure that convenes government actors, trade unions, and employer representatives is likely to favor the status quo of protective regulation for employed, unionized workers. Including representation from outsiders would likely shift the dialogue toward facilitating entry into labor markets, improving mobility, and promoting a more equitable distribution of returns across less favored segments of the population. Enhancing the Productivity of Informal Workers through Training and Skills Upgrading The low-productivity dimension of informality is especially notable in MENA’s poorer countries and rural areas, where workers with limited literacy and education are engaged in micro-entrepreneurship and low-yield agricultural work. Effectively increasing productivity in the ­ informal sector, particularly in rural areas, is a complex undertaking that ­ requires creating opportunities that promote human capital and connect people to markets. Well-designed programs aimed at increasing produc- tivity in the informal sector through training and skills upgrading are a potentially important intervention to promote inclusive growth in the medium term. Privately and publicly provided training programs intended to increase workers’ employability are abundant in the region. However, these programs tend to be delivered through in-class trainings targeted to urban, unemployed workers and rarely reach informal workers (overwhelmingly the working poor), especially in rural areas. ­ With limited access to these types of training opportunities, many Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 27 ­nformal workers are unable to enhance their skills and thus remain in i low-paying, low-productivity jobs. Emerging evidence from skills upgrading interventions sheds light on several success factors for informal, rural employment. How training is delivered matters. For example, programs such as the Barefoot College in India and the Agriculture and Fishing Fund in the Republic of Yemen suggest that the very modalities with which training is delivered (such as hands-on, community-based, or combining learning with earnings) are key factors for success. Training might be most effective if provided with placement services such as job search assistance and soft skills training. Low levels of literacy, scarce information, inflexible program schedules, traveling restrictions, and language barriers all limit access to and useful- ness of traditional, class-based training programs for informal workers, especially in rural areas (World Bank forthcoming). Facilitating the setup of training cooperatives might allow small firms to overcome the high fixed costs of structuring and delivering on-the-job training. Because informal workers are overwhelmingly employed in small enterprises or hold less stable jobs in larger companies, they are less likely to receive on-the-job training than formal workers. Providing incen- tives to firms (such as through training cooperatives) and workers (such as by providing vouchers) to engage in training could promote skills upgrading. To make these interventions more effective, reorienting and tailoring the delivery and design of training toward the particular needs of informal workers is necessary. Second-chance programs, traditional apprenticeships, and training specifically designed for the self-employed and micro-entrepreneurs are examples of interventions likely to be effective in the MENA region. ­ Reforming Existing Social Insurance Systems and Introducing New Instruments for Coverage Extension Coverage extension through social insurance systems is an important yet complex undertaking. In MENA expansion of social insurance schemes, particularly pensions, beyond the civil service has been achieved in all countries, but it has happened in a fragmented manner, and coverage remains limited. For example, in many countries, including Djibouti, Iraq, Morocco, and Syria, and in the West Bank and Gaza economy, no legal coverage instruments are available for self-employed and agricultural workers. No pension schemes exist for agricultural workers in Algeria, the Islamic Republic of Iran, Iraq, Libya, the West Bank and Gaza economy, and the Republic of Yemen. Although the existing pension ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 28 Overview systems in the MENA region have limited coverage, they are quite gener- ous. Internal rate-of-return computations indicate returns of between 6 and 17 percent, significantly higher than alternative investment instru- ments. Workers with information about these returns indicate that they want to participate in the pension system; this suggests that lack of par- ticipation in social security might be largely driven by constraints to accessing the system (that is, exclusion). Given the generosity of existing pension ­ systems, policy makers might be reluctant to extend coverage at these favorable conditions beyond the current (limited) set of beneficia- ries, to avoid amplifying the implicit pension debt. This justifies sequenc- ing coverage extension, where reforms aimed at improving the sustainability of current systems will need to precede coverage extension. Consequently, coverage ­ extension can take different forms, including extending provisions of current mandatory systems, improving the design of existing social insurance schemes to promote coverage, or extending coverage beyond existing mandatory social insurance systems and tradi- tional professions through new instruments. The latter, depending on country parameters and social preferences, could be based on providing universal or targeted benefits (that is, adopting a noncontributory and/or a­contributory approach). Important links are found between coverage outcomes and the design of social insurance schemes in the MENA region, particularly for pen- sions. Many countries in the region have defined benefit pension systems with design elements that are not in line with international best practices. Often these features contribute to low participation (coverage) rates; they include limited legal coverage, short minimum vesting periods that promote gaming of the system, generous early retirement provisions that distort incentives to participation, and the use of average wages in the last few years of service as a basis for calculating benefit amounts, which makes these schemes more expensive. Addressing these issues through reforms to ensure adequate, affordable, and sustainable benefits appears to be an important precondition for effective coverage extension to a large and increasing share of the population. Moving toward a defined contribution system is likely to achieve some of these objectives. Moreover, defined contribution systems are more useful for providing a platform to integrate different risk management instruments to cover risks such as disability, unemployment, and work injury, in addition to those associated with old age. Adequate coverage extension in MENA is not likely to be achieved if efforts do not go beyond providing traditional social insurance and do not Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 29 specifically target informal workers and those outside the labor force. Though this notion might be considered controversial, access to social insurance coverage could be designed to be unconditional on a formal employment relationship for some workers, particularly for the self- employed or those with irregular earnings and working patterns. Moreover, some workers might have too limited a savings capacity to contribute meaningfully to social security. Governments that opt to extend social insurance coverage beyond the existing mandatory social insurance system by targeting groups of informal workers are faced with a set of decisions related to strategy, design, and implementation. Key decisions pertaining to coverage extension strategies should be informed by an assessment of the relative appropriateness of contributory or non- contributory programs (in the light of existing country conditions), fiscal sustainability, and an evaluation of universal versus targeted subsidies. Matching defined contribution schemes are one promising option for extending coverage. These schemes are voluntary defined contribution savings mechanisms offering old-age or other social insurance benefits, where the government or employer provides incentives to enroll by matching individual contributions at a given rate and threshold. Piloting these schemes targeted to informal workers with limited but positive sav- ings capacity could provide policy makers with important insights on effective levers to increase coverage. In the short run, noncontributory schemes (such as social pensions) are still likely to play an important supplemental role. Middle-income countries such as Lebanon, Syria, and Jordan, where poverty-based targeting systems are emerging, might con- sider targeted contributory interventions. In poorer countries such as Djibouti and the Republic of Yemen, means-tested noncontributory strat- egies may be more appropriate, because contributory schemes are not yet feasible given current country conditions. Going beyond the traditional assumptions of agent rationality is impor- tant for promoting voluntary participation of informal workers in social insurance schemes, especially where contributions are required. Many of the observed behaviors in experimental settings contradict the standard economic models’ predictions based on the assumption of agent r­ ationality. In particular, behaviors such as procrastination and loss aversion could deter voluntary participation in savings schemes. Therefore, program fea- tures such as auto-enrollment, default options, and auto-escalation of ­ savings could induce participation. Addressing the particular needs of informal workers in the design of programs could also mean allowing for flexible contribution schedules, amounts, and withdrawal conditions. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 30 Overview Moreover, implementation arrangements that aim at minimizing transaction costs can also increase participation in these programs. Finally, ­ complementary financial literacy and general interventions to raise aware- ness of the benefits of social insurance schemes should be considered. Notes 1. Schneider (2004) developed a Multiple Indicator-Multiple Cause (MIMIC) model to quantify the size of the informal economy as a percentage of GDP. 2. A labor market is considered segmented if it consists of various subgroups with little or no crossover capability. 3. The International Labour Organization, European Union, and Organisation for Economic Co-operation and Development have embraced the concept of “flexicurity,” combining flexible regulation, safety nets (such as unemploy- ment insurance), and active social policies. One component of flexicurity policies is flexible and reliable contractual arrangements (from the perspec- tive of the employer and the employee, of “insiders” and “outsiders”) through modern labor laws, collective agreements, and work organization. References Alloush, M., C. Chartouni, R. Gatti, and J. Silva. 2011. “Informality and Exclusion: Evidence from Lebanon and Syria.” Mimeo. Washington, DC: World Bank. Angel-Urdinola, D., and K. Tanabe. 2011. “Micro-Determinants of Informal Employment in the Middle East and North Africa Region.” Mimeo. Washington, DC: World Bank. Bodor, A., D. Robalino, and M. Rutkowski. 2008. “How Mandatory Pensions Affect Labor Supply Decisions and Human Capital Accumulations: Options to Bridge the Gap between Economic Theory and Policy Analysis.” Washington, DC: World Bank. De Soto, H. 1989. The Other Path: The Invisible Revolution in the Third World. New York: HarperCollins. Elbadawi, I., and N. Loayza. 2008. “Informality, Employment and Economic Development in the Arab World.” Journal of Development and Economic Policies 10 (2): 25–75. International Finance Corporation and World Bank. 2012. Doing Business. http:// www.doingbusiness.org. Loayza, N., and T. Wada. 2010. “Informal Labor in the Middle East and North Africa: Basic Measures and Determinants.” Mimeo. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Overview 31 Maloney, W. F. 1999. “Does Informality Imply Segmentation in Urban Labor Markets? Evidence from Sectoral Transitions in Mexico.” World Bank Economic Review 13 (2): 275–302. OECD (Organisation for Economic Co-operation and Development). 2008. Employment Outlook. Paris: OECD. Perry, G., W. Maloney, O. Arias, P. Fajnzylber, A. Mason, and J. Saavedra-Chanduvi. 2007. Informality: Exit and Exclusion. Washington, DC: World Bank. Radwan, S. 2007. “Good Jobs, Bad Jobs, and Economic Performance: A View from the Middle East and North Africa Region, Employment and Shared Growth.” In Employment and Shared Growth: Rethinking the Role of Labor Mobility for Development, ed. Pierella Paci and Pieter Serneels, 37–51. Washington, DC: World Bank. Schneider, F. 2004. “The Size of the Shadow Economies of 145 Countries All over the World: First Results over the Period 1999 to 2003.” DP 1431. Bonn: IZA. World Bank. 2009. “From Privilege to Competition: Unlocking Private-Led Growth in Middle East and North Africa.” Washington, DC: World Bank. ———. 2010. “Egypt Investment Climate Assessment 2009: Accelerating Private Enterprise-Led Growth.” Washington, DC: World Bank. ———. Forthcoming. “Challenges to Enterprise Performance in the Face of the Financial Crisis-Eastern European and Central Asia.” Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Cha p t e r 1 How Large Is Informality and Why Do We Care? SUMMARY: This chapter provides general background on the basic measures of the prevalence of informal employment in the Middle East and ­ North Africa (MENA). Informality is best understood as a complex, multi- faceted phenomenon, which is determined by the relationship that the state establishes with private agents through regulation, monitoring, and provision of public services. Results indicate that some countries in the MENA region are among the most informal economies in the world. A typical country in MENA produces about one-third of its GDP and employs 65 percent of its labor force informally (using the Schneider Index and the share of the labor force without social security coverage, respectively). These stylized facts indi- cate that more than two-thirds of all workers in the region may not have access to health insurance and/or are not contributing to a pension system that would provide income security after retirement. At the same time, from a fiscal perspective, these results indicate that about one-third of total eco- nomic output in the region remains undeclared and, therefore, not registered for tax purposes. What Is Informality? There are many angles to informality. In his classic study, De Soto (1989) defines informality as the collection of firms, workers, and activities that operate outside the legal and regulatory framework. Overall informality can be studied through three main lenses: a firm-based productivity perspective, an employment perspective (workers), and a fiscal perspective (untaxed activities). Informality is a heterogeneous concept, ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 33 34 How Large Is Informality and Why Do We Care? comprising different situations such as the unregistered small firm, the street vendor, and the large registered (and hence “formal”) firm that employs some of its workers without offering written contracts or access to social security benefits: • Firms: A generally accepted definition of “firm” informality was proposed by the Delhi Group in 1997. According to their definition, ­ the informal sector includes private unincorporated enterprises (or quasi-unincorporated), which produce at least some of their goods and services for sale or barter, have fewer than five paid employees, are not registered, and are engaged in nonagricultural activities (ILO 2002). Studies generally measure informality at the firm level through a direct (micro-) measurement based on individual surveys, such as the World Bank’s Enterprise Surveys, that explicitly ask the firm’s owner or man- ager for information on the years in which the firm started its ­operations and legally registered. A discrepancy between the two is typically con- sidered the period during which the firm operated informally (Angel-Urdinola, Reis, and Quijada 2009). ­ • Workers: According to this definition, the informal sector is proxied by informal employment in either formal (small unregistered or unincor- porated firms) or informal enterprises, with “informal employment” refereeing to the absence of benefit from or ­registration to social secu- rity or the absence of a written contract (see box 1.1. for more details on defining informality). Hence, informal ­ sector employment includes enterprises comprising (1) contrib- informal employment in formal sector ­ uting family members and (2) employees; informal employment in informal sector enterprises comprising (1) own-account workers, 1 ( 2) employers, (3) contributing family workers, (4) employees, ­ and (5) members of producers cooperatives; and informal employ- ment in households producing goods exclusively for their own final use and households employing paid domestic workers comprising (1) own-account workers and (2) employees. • Untaxed Activities: From a fiscal point of view, a large informal sector constitutes a set of activities that are “hidden” for tax purposes. The United Nations System of National Accounts (EC and others 2008) ­ ctivities, distinguishes between four types of such activities: (1) informal a which are undertaken “to meet basic needs”; (2) underground activities, which are deliberately concealed from public authorities to avoid either Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 35 Box 1.1  Defining Informality The concept “informal employment” was adopted by the 17th International ­ Conference of Labor Statisticians (ILCS) in 2003. Hart (1973) was the first author to make the distinction between formal and informal employment based on the difference between wage-earning employment (formal) and self-employment on a ­ permanent and regular basis, either alone or in a partnership (informal). The ­ guidelines of the definition refer to the characteristics of the job, and the criteria are absence of benefit from or registration for social security or, more strictly, the absence of a written contract. The guidelines provide more detailed criteria that are to be determined “in accordance with national circumstances and data availability.” Although informal employment is a job-based concept, the category of “informal sector employment” is an enterprise-based concept as defined by the 15th ICLS in 1993. A set of characteristics of the economic units is used: (1) legal status (unincor- porated individual enterprises owned by households) and (2) nonregistration of the unit or nonregistration of its employees or size of the unit (fewer than 5 or 10 ­permanent employees). Because this enterprise-based definition was missing the dramatic increase of informal jobs outside the traditional enterprises (especially the flexible and unprotected jobs directly or indirectly created by formal enterprises confronted with strong competition), the broader concept of informal employment was adopted. “Self-employment” comprises own-account workers, employers, and contributing family workers, as defined in the International Classification of Status in Employment. Although it includes categories such as professionals, ­ self-employment can be used as a proxy for the main component of informal employment, especially as it has been collected in household surveys and popula- tion censuses, and data are available. Source: World Bank 2009. the payment of taxes or compliance with certain regulations (for exam- ple, most cases of tax evasion and benefit fraud); (3) illegal activities, which generate goods and services forbidden by the law or which are unlawful when carried out by unauthorized ­ producers; and (4) house- services for own-consumption. hold activities, which produce goods and ­ How to Measure Informality If defining informality is a complex task, its measurement is even more daunting. Given that it is identified by working outside the legal and Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 36 How Large Is Informality and Why Do We Care? regulatory frameworks, informality is best described as a hidden, unob- served variable. Hence, accurate and complete measurement is not feasi- ble, but an approximation is possible using indicators that reflect its various aspects. To provide an estimate of the magnitude of informality in a country, it is better to use a variety of different indicators that, taken together, can provide a more robust approximation to informality. Several direct and indirect methods are used to measure informality ­ (Angel-Urdinola, Reis, and Quijada 2009): • Survey methods: Informality could be measured directly using individual surveys, such as the World Bank’s Enterprise Surveys, which explicitly ask the firm’s owner or manager for information on the years in which the firm started its operations and was legally registered. A discrepancy between the two is typically considered as the time when the firm oper- ated informally. In some household or labor force surveys, interviewees are asked whether they have signed a formal contract in their current employment, or whether they are affiliated with the social security administration (meaning that they, or their employer, are contributing to a pension plan or another social protection program). A potential problem with direct measures is that the interviewee’s answer depends heavily on the phrasing of the question, and many interviewees might be reluctant to reveal their behavior (for example, in the case of firms). • Tax audits: Tax-audit data can be used to determine the percentage of firms audited that evade taxes and to quantify the amount of tax under- reporting as informal activity (as well as a firm’s legal status). The short- coming is that tax audits are not conducted randomly typically, and hence the information is not representative of the population of firms (Perry and others 2007). • National Accounts: An indirect method commonly used to estimate the size of the informal economy is the difference between aggregate income and aggregate expenditures from the National Accounts. This measure has the advantage of being conceptually simple. However, it has been used mainly in developed countries, because it requires inde- pendent calculations of aggregate income and expenditure. • Multiple Indicator-Multiple Cause model: Another method that has been used recently is the Multiple Indicator-Multiple Cause (MIMIC) model, popularized by Schneider (2004), who applied it for 145 countries. This model assumes that although informal activity is not observable, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 37 magnitude can be represented by a latent variable (in index form), and its ­ both its causes and effects can be observed and measured. This latent variable is then used in a set of two equations: First, the latent variable is the dependent variable, and its causes are the explanatory variables; sec- ond, the effects of informality are modeled as a function of the latent variable. The set of equations is then simultaneously estimated, and the fitted values of the latent variable are used to compute an estimate of the size of the informal sector as a share of GDP. This technique has been ­ criticized because of the lack of theoretical support for the equations that are supposed to capture the causes and effects of informal activity. Nev- ertheless, its use remains widespread, probably as a consequence of the aforementioned difficulties to obtain broad estimates of informality. Three indicators are commonly used to measure informality: the Schneider Index, lack of social security coverage, and prevalence of self- employment (Loayza and Wada 2010). The Schneider Index combines the MIMIC method, the physical input (electricity) method, and the excess currency-demand approach for the estimation of the share of pro- duction that is not declared to tax and regulatory authorities.2 Lack of social security coverage is often estimated as the fraction of the labor force (or employment) that does not contribute to a retirement pension scheme, as reported in the World Bank’s World Development Indicators. The prevalence of self-employment is often computed as the ratio of self- employment to total employment (as reported by the ILO). Additional measures, such as the share of total employment without health insur- ance and the share of unpaid workers in total employment, are also com- monly used proxies of informality (Angel-Urdinola and Tanabe 2011). Table 1.1 and box 1.2 present the different definitions used for MENA countries included in the study. Although each indicator has its own conceptual and statistical short- comings as a proxy for informality, taken together they provide a robust approximation to the issue. Cross-country scatter plots of the three mea- sures of informality against each other provide evidence for the significant correlation among the three indicators, with correlation coefficients ranging from 0.10 to 0.91 (which are high enough to represent the same phenom- enon but not so high as to make them mutually redundant) (figure 1.1).3 This report addresses informality from a worker’s perspective and focuses specifically on informal employment. Informal employment can be captured in many ways. Informal employment is often proxied as (1) the share of unpaid employment, (2) the share of ­ self-employment, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 38 Table 1.1 Summary of Data and Definitions Used in This Report Egypt, Iran, Syrian West Bank Arab Islamic Arab and Gaza Yemen, Algeria Djibouti Rep. Rep. Iraq Jordan Lebanon Libya Malta Morocco Republic Tunisia economy Rep. Not contributing to so- X X X X X X X X X X X X X X cial security X X X X X Percentage of self-em- X X X X X X X X X X ployed X X X X X X Schneider Index X X X X X X X X X Share of employment without access to health insurance X X X Share of unpaid w­ orkers ­ mployment) (to total e X X X X X Source: See annex table 1A.2 for information on data sources. Note: For a detailed description of the micro-surveys used in this study, see the annex in chapter 2. How Large Is Informality and Why Do We Care? 39 Box 1.2 Main Definitions of Relevant Variables and Survey Questions from Data Available in MENA Working-age population (WAP) is composed of individuals in the age group 15 to 64 years old. Employed are those individuals in the WAP who satisfy any of the following ­ conditions: (1) the person did any work for at least one hour during the previous week (with a remuneration that could have been monetary or in-kind); (2) the person was “temporarily absent” from work because of holidays, illness, paid ­ absence for education or training, maternity leave, strike, work dispute, seasonal work, and the like. Informal workers as defined in this report are those individuals in the WAP (or, in the ILO dataset, participate in the labor force) who are employed and do not make contributions to social security. In different countries, depending on data availabil- ity, different questions were used to define informal workers: Iraq: Is this job included in retirement systems and social security? (social security includes pension and medi- cal benefits); Morocco: What is your status as regards the social and medical security? 1 = member; 2 = beneficiary; 3 = not covered (social security and medical insurance comes together); Egypt: Do you have social security? (Almost all people who have social security are covered by health insurance); the Republic of Yemen: Benefit offered for this job? (Pension) (Among those covered by pension systems, only 29 ­percent have health insurance); Syria: Covered by social security or insurance and salaries? (Social security includes pension and medical benefits); Lebanon: Is worker benefiting from health insurance? Yes (1 = national security fund; 2 = government employees; 3 = army and security forces; 4 = municipality; 5 = private insurance at employer’s expense; 6 = mutual fund; 7 = private through an institution, syndicate, or committee). Indicator Calculation Description Labor force LF ÷ WAP Share of WAP in the participation ­ (U+E) ­labor market rate (%) Employment E ÷ WAP Share of employed in rate (%) the WAP Unemployment U ÷ LF Share of unemployed rate (%) (U+E) in the labor force Source: Based on Labor Force Surveys. Note: E = employed population; LF = population in the labor force (employed and unemployed); U = unemployed population; WAP = population between 15 and 64 years old. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 40 How Large Is Informality and Why Do We Care? Figure 1.1 Correlation among Most-Used Informality Indicators 100 Yemen, Rep. Syrian Arab Republic % of workers not contributing West Bank and Gaza 80 Morocco to a pension scheme Lebanon Algeria Iran, Islamic Rep. 60 Tunisia Egypt, Arab Rep. 40 20 0 0 20 40 60 80 % of self-employment in total employment 100 Yemen, Rep. % of workers not contributing 80 Syrian Arab RepublicMorocco to a pension scheme Jordan Lebanon 60 Iran, Islamic Rep. Algeria Tunisia Egypt, Arab Rep. 40 20 0 0 20 40 60 80 Schneider index 100 80 % of self-employment in total employment Iran, Islamic Rep.Lebanon 60 Syrian Arab Republic Yemen, Rep. Algeria Egypt, Arab Rep. Morocco 40 Tunisia 20 0 0 20 40 60 80 Schneider index Source: Processed from Loayza and Wada 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 41 (3) the share of total employment not contributing to social security (Gasparini and Tornarolli 2006; Loayza and Rigolini 2006; World Bank 2009), and (4) the share of employees without a contract. Figure 1.2 displays the correlations among these different measures for countries where data are available. As illustrated by the correlations, not having access to social security is positively correlated with all other measures (and especially with not having access to health insurance), although the magnitude of the correlation varies by country and by strata (urban or rural). The strong and positive association between contributing to social security and having access to health insurance (in most cases both come bundled in the social security package) suggests also that informal workers are generally not covered against health risks.4 As seen in ­ figure 1.2, unpaid work and self-employment are negatively correlated. ­ This indicates that some features of informality among the ­ self-employed are quite unique from those of unpaid workers. However, both of the aforementioned definitions are positively correlated with not having access to social security, indicating that the “share of overall employment not contributing to social security” is able to capture some of the features of informality inherent to these two different groups. In the context of this report, the standard definition of informality will be the share of the labor force/employment that does not contribute to social security. In characterizing informal employment from the worker/ social protection perspective, one core definition relates to the absence of workers’ coverage by traditional social security programs, most notably health insurance and pensions, but often also other benefits available to workers by virtue of their labor contract (Perry and others 2007). To ensure comparability across countries, and based on data availability, informal employment is defined herein as the share of overall labor force/employ- ment not contributing to social security (and therefore not covered by a pension system and, in most cases, not covered by health insurance). Various advantages are found by selecting this definition. The first is data availability. As presented in table 1.1, data used to proxy alternative definitions of work- ers’ informality, such as health insurance coverage, lack of contract, and unpaid employment, are not available for many countries in the MENA region. Second, the lack of social security coverage captures well different aspects of informal employment and thus is more likely to portray the com- mon features of informal employment. In particular, as seen in figure 1.2, unpaid work and ­ self-employment are negatively correlated, suggesting that some features of informality among the self-employed are quite different when compared with those of unpaid workers. However, both of these Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 42 How Large Is Informality and Why Do We Care? Figure 1.2 Correlations among Different Definitions of Labor Informality ­(Worker’s Side) no social security Morocco and no health Iraq insurance Yemen, Rep. Egypt, Arab Rep. no social security Syrian Arab Republic and unpaid workers Morocco Iraq Lebanon Iraq Syrian Arab Republic no social security Lebanon and self-employment Morocco Egypt, Arab Rep. Yemen, Rep. Iraq Lebanon self-employment Egypt, Arab Rep. and unpaid work Syrian Arab Republic Morocco Yemen, Rep. –0.80 –0.60 –0.40 –0.20 0.00 0.20 0.40 0.60 0.80 1.00 rural urban Source: Angel-Urdinola and Tanabe 2011. definitions are positively correlated with not having access to social security, indicating that this definition is able to capture some of the features of informality inherent to these two very different groups. Finally, the defini- tion has been used often in other regional studies on informality, which allows for comparability between MENA and other world regions (ILO 2002; Perry and others 2007; World Bank 2009). Why Does Informality Matter? There are costs and benefits to informality. In many countries the informal sector often represents a very resourceful part of the economy and is a source of creativity and dynamism. However, important costs associated with informality, such as unprotected work, low firm are ­ productivity, and tax evasion. A traditional view of informality argues that, in general, workers and firms in the informal sector would prefer to be formal (such as by registering with the state, paying taxes, or Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 43 being affiliated with social security), but regulatory and administrative barriers prevent them from doing so. As argued by Perry and others (2007), however, considerable evidence in Latin America suggests that the informal sector is fairly heterogeneous, with workers and firms that have been excluded from the formal economy coexisting with others that have opted out on the basis of implicit cost-benefit analyses. This latter concept of “exit” posits that at least some of those in the informal sector are there as a matter of choice. Specifically, some workers and firms, upon making some implicit or explicit assessment of the benefits and costs of formality, actually prefer operating informally. A wide range of degrees to which exit or exclusion holds in any economy are found. Hence, these two perspectives are complementary characteriza- tions rather than competing hypotheses. Informality is a fundamental characteristic of underdevelopment. As depicted in figure 1.3, a larger informal sector is often associated with lower GDP per capita. For most countries in the region, the share of informal to total employment is mostly aligned where it should be given the level of economic development. A two-way relationship can explain the strong and negative association between informality and economic development. On the one hand, widespread informality induces firms to remain suboptimally small, use irregular procurement and distribution channels, and constantly divert resources to mask their activities or bribe officials. As such, informality may be a source of economic retardation because it is associated with misallocation of resources and loss of the advantages of legality (such as police and judicial protection, access to formal credit institutions, and participation in international markets) (Loayza and Wada 2010). On the other hand, low levels of economic development affect the capacity and quality of institutions to enforce regulation, collect taxes, and provide services that firms deem worth pay- ing for through taxation (such as social security and infrastructure). At the same time, a powerful spurious effect could drive this association, because informality and low income could result from policies and insti- tutions that affect negatively both. Measurement issues might also affect this correlation, because in countries with higher informality more GDP escapes formal measurement. Informality also has positive implications for the labor market, because it provides employment to a significant share of the population, may be a source of innovation and entrepreneur- ship (many firms start operating informally when they are small and formalize as they grow), and serves as a safety net in periods of transition. As such, at a given level of distortion, it is preferable to a fully formal Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 44 How Large Is Informality and Why Do We Care? Figure 1.3 Informality and Economic Development % of workers not contributing to a pension scheme 100 NERCAF GNB NPL NGA BGD GMB SLE RWA SEN ZMBLSO PAK GHAKEN UGA IND YEM BTN BOL VNM DJI IDN PRY NICHND SYR JAM 80 WBG MAR CHN PHL GTMCOL PER ECUSLV CPV THA MDV DOM VEN KGZ JOR IRN LBN MEX ARG LKA GEO AZE BIH KAZ DZA 60 VCT GUYYUG ALB TUN TUNTUR MUS MKD BRA URY EGYROM CRI CHL TTO 40 MDA MNG ARM DMA MYS LCA LBY UKR GRD SGP BLZHRV HUN BRB SVK 20 LTU KOR GRC POL CZE ATG DEU AUT FIN FRA CAN ISL SWE IRL EST PRT LVA BRN ITA ESP NLD AUS GBR NOR USA KNA BHR DNK NZL BEL JPN 0 CHE 4 6 8 10 12 GDP per capita, constant 2000$, in ln 60 CMR % of self-employment in total employment IDN ETH MDG VNM HND KHM DOM PRY COL 40 BGD ZMB BOL PAKNIC GTMPER PHL SYR IRN JAM MDV ECUTHA DMA YEM LKA MNG CPV SLV PAN VEN LBN LCA DZA BRA GRC WBG EGY MAR BLZCHL MEX URY KOR NAM CRI FJI TUN ARG PRT ITA 20 TWN MYS TTOCYP ZAF MUS NZL ESP CANIRL MLT PRI BRB IMY SGP BEL ISL CHE OMN ISR AUS GBR NLD HKG AUT FRA DEU SMRJPN MAC ABW DNK USA NOR LSO 0 4 6 8 10 12 GDP per capita, constant 2000$, in ln Source: Loayza and Wada 2010. Note: GDP = gross domestic product. economy where agents are unable to circumvent regulation-induced rigidities and would not otherwise be able to exist without the flexibility provided by the informal sector. The evidence suggests that higher levels of informality are associated with lower levels of economic growth. Loayza and Wada (2010) conducted a simple regression analysis of the association between informality and growth for all countries for which data are available (that is, using the Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 45 Schneider Index, self-employment as a share of total employment, and share of the labor force not contributing to social security). Using all avail- able informality proxies, the authors find that higher informality is associ- ated with lower economic growth. The dependent variable is the average growth of GDP per capita over 1985–2005. Loayza and Wada (2010) consider a period of about 20 years for the measure of average growth to achieve a compromise between merely cyclical, short-run growth (which would be unaffected by informality) and very long-run growth (which could be confused with the sources, rather than consequences, of informal- ity). They include a proxy for the overall capacity of the state as a control variable (level of GDP per capita and initial ratio of government expendi- tures to GDP) in the regression. The explanatory variables of interest are the three informality indicators, considered one at a time. Regression analy- sis uses both ordinary least-squares (OLS) and instrumental-variable (IV) methods. It should be noted, however, that results from cross-country regressions often encounter skepticism because of the many potential sources of spurious correlation. The authors’ estimates indicate that con- trolling for the country’s initial level of expenditures and GDP, an increase of one standard deviation in any of the informality indicators leads to a decline of 0.7–1 percentage points in the rate of GDP per capita growth (figure 1.4). Loayza, Oviedo, and Serven (2005) identify informality as one Figure 1.4 Effect of Informality on Growth (Decline in GDP growth per standard deviation increase of each indicator) –1.8% % of workers not contributing –2.0% to social security –5.6% % of self-employment in –6.0% total employment –6.2% Schneider Index –7.9% –10.0% –8.0% –6.0% –4.0% –2.0% 0.0% OLS IV Source: Processed from Loayza and Wada 2010. Note: IV = instrumental variable; OLS = ordinary least-squares. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 46 How Large Is Informality and Why Do We Care? of the main channels through which regulation affects growth, with a heavier regulation burden reducing growth and inducing higher informality. Important adverse consequences of labor informality on other human development outcomes are seen, such as less protection at old age, lower access to and affordability of “quality” health care, and reduced employ- ment quality (high working hours, unsafe conditions, and low pay). Social insurance systems in most MENA countries are historically based on Bismarckian principles, whereby pension, health, and disability benefits are linked to employment in the formal sector. As a result, formal sector actors (public and private sectors, employers and employees) have contributed to social security programs and in return have been covered ­ by relatively generous, multidimensional benefit packages. Those outside the formal sector, in both urban and rural areas, have traditionally had limited or no access to formal risk management instruments or other government benefits. At the same time, noncontributory antipoverty and social protection programs developed recently in MENA remain limited in scope and reach, in most cases, a very narrow portion of society. In this sense, nearly all MENA countries are characterized by “truncated welfare systems.” Informal workers in MENA are likely to use inadequate mecha- nisms to cope with social risks, such as informal and less effective social safety net arrangements or by selling assets or withdrawing children from school (which can have long-term adverse consequences). Emerging evi- dence also finds that informality is associated with lower life expectancy and health outcomes. Using data from the Republic of Yemen, Cho (2011) finds that, controlling for other factors, individuals living in house- holds having a head working in the formal sector are associated with better health outcomes: lower malnutrition for children, low o ­ ut-of-pocket expenses, and low adult morbidity. Informality in MENA Brief Macroeconomic Background On average, the MENA region displays lower levels of employment and higher levels of unemployment than any other region in the world. Some basic macroeconomic trends comparing the MENA region to other regions in the world and, within MENA, non-GCC to GCC members are presented to provide context to the analysis that follows.5 The level of economic development and other macroeconomic variables, such as recent economic growth and employment composition, are likely to be important factors for understanding a country’s profile and determinants Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 47 Figure 1.5 Employment and Unemployment Rates in MENA employment rates (%) 2008 MENA 50.2 GCC countries 64.7 non-GCC 43.0 ECA 53.0 developed countries 54.2 SA 56.5 world 60.3 LAC 60.5 SSA 65.4 SEA & Paci c 65.5 East Asia 70.0 0 20 40 60 80 100 unemployment rates (%) 2008 MENA 10.0 GCC countries 3.4 non-GCC 12.9 East Asia 4.3 SA 4.9 SEA & Paci c 5.5 World 5.9 developed countries 6.1 LAC 7.2 SSA 7.6 ECA 9.0 0 5 10 15 20 Source: World Development Indicators dataset. Note: ECA = Europe and Central Asia; GCC = Gulf Cooperation Council countries; LAC = Latin America and the Caribbean; SA = South Asia; SEA = Southeast Asia; SSA = Sub-Saharan Africa. of informality. Figure 1.5 illustrates the patterns of employment in MENA (subdivided into non-GCC and GCC countries) versus other regions. Differences between GCC and non-GCC countries are notable. GCC countries display high employment rates and low unemployment rates in comparison to the rest of the world. On the other hand, non- GCC countries show low employment rates (mainly because of low levels of female employment) and by far the highest unemployment rates in the world (at 13 percent in 2008). In recent years, MENA countries have displayed positive economic growth and rapid employment growth, especially for women. Although annual GDP per capita growth rates among GCC and non-GCC c ­ ountries have been rather similar in recent years and generally above the world Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 48 How Large Is Informality and Why Do We Care? average, annual employment growth rates have been significantly larger for GCC than for non-GCC countries (suggesting that GCC countries show higher employment elasticities to growth) (figure 1.6). In the MENA region, as in many other regions in the world, employment growth rates for the period 1998–2008 have been higher for women than for men. In particular, during this period, the growth rate of female (male) employ- ment was 6.4 percent (4.2 percent) in GCC countries and 3.8 percent (2.9 percent) in non-GCC countries. Despite such rapid increases in female employment, female employment levels in the region remain the lowest in the world. At the same time, population growth in the MENA region has been faster than in other regions. Annual population growth of 1.9 and 2.7 percent per year in GCC and non-GCC countries, respectively (­figure 1.7), is among the highest in the world (and equivalent to those in ­ Sub-Saharan Africa). While MENA countries created jobs at a higher pace than other countries in the world, the level of employment creation did not keep pace with population growth. Unemployment rates in GCC countries remain low at the time this report was written, but rapid population growth (especially if these countries continue to receive flows of interna- tional migrants) poses important potential risks for increasing unemploy- ment and informality. MENA countries differ significantly from other middle-income coun- tries in the composition of employment generation (World Bank 2011). In the typical country in MENA, agriculture contributes to a larger extent to employment growth than in other middle-income countries. In other regions, there was a shift of labor away from agriculture towards services. In the typical MENA country, the contribution of manufacturing and private communication sectors services, especially the trade, tourism, logistics and ­ (figure 1.8) is low. These sectors were the main contributors to employ- ment growth in comparable middle income countries (such as Malaysia or Indonesia). Hence, jobs created in the typical MENA country were low- quality jobs, skewed toward low value-added sectors, which reflects the lack of structural transformation. In GCC countries, employment growth took place largely in the construction sector (jobs taken up by expatriates) and the creation of higher paid public sector jobs directed to nationals. Informality Levels and Trends A typical country in MENA produces about 28 percent of its GDP and employs 65 percent of its labor force informally. Using available data, Loayza and Wada (2010) assess the prevalence of informality in MENA Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 1.6 Economic Growth and Employment Growth annual GDP per capita growth rate (2005–2009) total annual employment growth (1998–2008) MENA 3.3% MENA 3.3% GCC countries 3.4% GCC countries 4.3% non-GCC 3.2% non-GCC 3.1% developed countries 0.0% developed countries 0.9% ECA 0.9% East Asia 1.0% SEA & Paci c 1.4% ECA 1.1% world 1.9% world 1.7% LAC 2.2% SEA & Paci c 2.0% SSA 2.7% SA 2.2% SA 5.6% LAC 2.5% East Asia 6.2% SSA 3.0% 0% 1% 2% 3% 4% 5% 6% 7% 0% 1% 2% 3% 4% 5% 6% 7% Source: World Development Indicators dataset. Note: ECA = Europe and Central Asia; GCC = Gulf Cooperation Council; LAC = Latin America and the Caribbean; SA = South Asia; SEA = Southeast Asia; SSA = Sub-Saharan Africa. 49   50 How Large Is Informality and Why Do We Care? Figure 1.7  Annual Population Growth Rates (%) (2005–2009) MENA 1.9 GCC countries 2.7 non-GCC 1.8 ECA 0.4 developed countries 0.7 EAP 0.7 LAC 1.2 world 1.2 SA 1.5 SSA 2.5 0 0.5 1 1.5 2 2.5 3 Source: World Development Indicators dataset. Note: ECA = Europe and Central Asia; GCC = Gulf Cooperation Council; LAC = Latin America and the Caribbean; SA = South Asia; SSA = Sub-Saharan Africa. Figure 1.8 Sectoral Contribution to Annual Employment Growth in Typical MENA Country and Selected Other Countries, Average for 2000s 140 120 100 80 60 40 20 0 –20 –40 Typical MENA Malaysia Indonesia Brazil country Agriculture Manufacturing Trade, Tourism, Logistics and Communication Gov't admin and Social services Mining and Utilities Construction Financial and Real estate services Source: World Bank 2013. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 51 using three commonly used proxies of informality: (1) the Schneider Shadow Economy Index, (2) the share of the labor force not contributing to social security, and (3) the share of self-employment to total employ- ment (figure 1.9). These stylized facts indicate that more than two-thirds of all workers in the region may not have access to health insurance and/or are not contributing to a pension system that provides income security after retirement age. At the same time, from a fiscal perspective, these results indicate that about one-third of total economic output in the region remains undeclared and therefore, not registered for tax purposes. The degree of informality in the median country in MENA is lower than that of most developing countries, but much higher than that of the median developed country. As illustrated by figure 1.9 and using the lack of coverage measure, the typical MENA country has a labor informality rate larger than ECA, and lower than LAC, EAP, SA, and SSA. Using the Schneider Index, the median country in MENA has a lower informal production than the other developing regions. Using the share of self- employed, a typical MENA country is as informal as a typical country in LAC, more informal than a typical country in ECA, and less informal than a typical country in EAP, SA, and SSA. Important variations are found in the prevalence of informality across non-GCC countries, depending, among other factors, on the availability of natural resources and labor. Non-GCC countries are quite heteroge- ­ conomic devel- neous in terms of size, availability of resources and labor, e opment, and demographic structure, all factors that influence the size of the informal economy (see chapter 2 for a more thorough discussion). The prevalence of informality varies significantly across countries in this group (figure 1.10). In general, resource-rich/labor-abundant economies (such as the Islamic Republic of Iran and the Syrian Arab Republic) tend to display high informality rates in the region as proxied by the share of the labor force not c ­ontributing to social security (between 66 and 73 percent) and by the share of self-employment to total employment (between 32 and 54 percent). However, their share of undeclared output to total GDP (at around 20 percent) is comparable to that of GCC countries. This occurs because this group of countries generally has few, ­ but large, formal firms (many in the energy sector) that are capital inten- sive, thus rendering lower informality in production than in labor (Loayza and Wada 2010). On the contrary, resource-poor/­ labor-abundant econo- mies (such as Egypt, Tunisia, and Morocco) display a high share of unde- clared output (between 35 and 40 percent of GDP) and a lower share of the workforce not contributing to social security (between 45 and Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 52 How Large Is Informality and Why Do We Care? Figure 1.9 Prevalence of Informality in MENA versus Other Regions % of the labor force not contributing to social security MENA, non-GCC countries 65.0 developed countries 9.3 33.2 ECA LAC 73.6 EAP 75.0 SSA 93.2 0 20 40 60 80 100 self employment (% of total employment) MENA 32.4 MENA, GCC countries 3.8 MENA, non-GCC countries 36.5 developed 13.4 ECA 22.7 LAC 33.1 EAP 56.3 SA 71.1 SSA 80.8 0 20 40 60 80 100 undeclared output as % of GDP (Schneider Index) MENA 27.7 MENA, GCC countries 20.1 MENA, non-GCC countries 36.2 developed countries 16.6 EAP 32.8 SA 37.0 ECA 38.9 SSA 42.2 LAC 43.9 0 20 40 60 80 100 Source: Schneider et al., 2010 for Schneider Index; WDI for self employment and pension scheme. Note: EAP = East Asia and Pacific; ECA = Europe and Central Asia; GCC = Gulf Cooperation Council; LAC = Latin America and the Caribbean; SA = South Asia; SSA = Sub-Saharan Africa. The periods covered are the latest years in 2000–10 for pension scheme, 2000–11 for self- employment, and 1999–2007 for the Schneider index, respec- tively. Data on social security contributions in GCC countries are only available for Bahrain, 2007 (20 percent), and Qatar, 2008 (4.4 percent) (Palleres-Miralles, Romero, and Whitehouse 2012). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 53 Figure 1.10 Informality Rates for Selected Non-GCC Economies % of the labor Libya 31.5 force not contributing Egypt, Arab Rep. 44.9 to social security Tunisia 51.4 Jordan 61.6 Algeria, Rep. 63.3 Iraq 64.4 Lebanon 65.5 Iran, Islamic Rep. 65.8 Syrian Arab Republic 73.2 Morocco 76.2 West Bank and Gaza economy 86.0 Yemen, Rep. 89.6 self-employment Lebanon 15.6 (% of total Yemen, Rep. 32.4 employment) Tunisia 34.2 West Bank and Gaza economy 35.6 Morocco 37.3 Egypt, Arab Rep. 37.8 Iran, Islamic Rep. 46.8 Syrian Arab Republic 53.9 Algeria, Rep. 67.1 Syrian Arab Republic 19.5 Iran, Islamic Rep. 19.5 undeclared output as % of GDP Jordan 20.3 (Schneider Index) Yemen, Rep. 27.7 Lebanon 35.1 Egypt, Arab Rep. 35.3 Algeria, Rep. 35.7 Libya 36.7 Morocco 37.9 Tunisia 39.6 0 20 40 60 80 100 Source: Processed from Loayza and Wada 2010. Note: Time periods are as follows: Schneider Index, average 1999–2007; self-employment, 1999–2007; not contributing to social security (S.S.), 2000–2007. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 54 How Large Is Informality and Why Do We Care? 76 percent), which is consistent with a higher share of medium size (and semiformal) labor-intensive firms. Informality in the MENA region has been rising in recent years, as proxied by the Schneider Index and the share of self-employment to total employment. However, the increase in the share of undeclared/ informal output to total GDP in recent years has been a global trend (figure 1.11). Within the region, important differences are seen in the growth rate of informality between GCC and non-GCC countries. First, informality, as proxied by the Schneider Index, has been growing faster in non-GCC countries than in GCC countries. Second, although self- employment has been rising in non-GCC countries, it has been declin- ing rapidly in GCC countries. It is worth noting that ­ self-employment in non-GCC countries (at 37 percent) remains somewhat lower than in other developing regions such as South Asia and Sub-Saharan Africa (71 and 81 percent, respectively) (Loayza and Wada 2010). This report focuses its analysis and discussion on non-GCC MENA countries. A discussion of the aggregate trends in GCC and non-GCC countries is important to provide a comprehensive overview of informal- ity throughout the region, but the different dynamics in labor markets in GCC and non-GCC MENA economies and the lack of micro-data make in-depth analysis of the informality phenomenon in GCC countries out- side the scope of this report. Hence, all results from this point forward focus only on non-GCC countries, and any references to “MENA” need to be understood to refer to only non-GCC countries. What Causes Informality? Informality can be the result of exclusion or of a rational exit in opting out of the coverage system. The option to participate in the informal sector reflects cost and benefit considerations but is not always voluntary or desirable (Loayza and Wada 2010). From the perspective of a worker or a firm, joining the informal sector can be seen as either “exiting” the legal framework or “being excluded” from it. In the former case the informality option implies a voluntary decision, whereas in the latter it is derived from segmentation or segregation. In both cases, however, what matters is that informality in the economy results from a combination of factors affecting the potential gains, costs, and restrictions related to legally established firms and workers. Formality entails costs of entry (in the form of lengthy, expensive, and complicated registration procedures) and costs of perma- nence (including payment of taxes, compliance with mandated labor Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 55 Figure 1.11  Annual Growth Rates of Informality annual growth rate, Schneider Index (2000–2007) LAC 1.39 ECA 0.76 MENA 0.69 SSA 0.64 developed 0.62 countries SA 0.55 EAP 0.24 0 0.5 1 1.5 annual growth rate of self employment (2000–2011) SSA 7.2 SA 7.2 EAP 5.2 MENA 2.1 OECD –1.5 LAC –1.8 ECA –4.3 –6.00 –4.00 –2.00 0.00 2.00 4.00 6.00 8.00 Source: Processed from Loayza and Wada 2010. Note: EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; SA = South Asia; SSA = Sub-Saharan Africa. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 56 How Large Is Informality and Why Do We Care? benefits and remunerations, and observance of ­ environmental, health, and other regulations). The benefits of formality potentially consist of police protection against crime and abuse, recourse to the judicial system for conflict resolution and contract enforcement, access to legal financial insti- tutions for credit provision and risk diversification, and, more generally, the possibility of expanding markets both domestically and internationally. At least in principle, formality also reduces the need to pay bribes and prevents incurrence of penalties and fees to which informal firms are likely subjected. Informality is more prevalent when the regulatory framework is burdensome, the quality of government services to formal firms is low, and the state’s monitoring and enforcement power is weak (Friedman and others 2000; Johnson and others 1997; Loayza and Rigolini 2009; Schneider and Enste 2000). According to Galal (2005), the current regula- tory framework in Egypt, for instance, discourages formality, leading to an annual loss in tax revenues equivalent to 1 percent of GDP. The prevalence of informality depends on structural characteristics of countries, such as governance, productivity, and economic composition. Loayza and Wada (2010) analyze the structural factors that influence informality: (1) a country’s governance structure and regulatory frame- work (which will affect the opportunity cost of informality), (2) labor productivity, demographics, and employment composition, and (3) the size of the public sector. To illustrate the relative importance of these determinants, Loayza and Wada (2010) first calculate a series of correla- tions between proxies for the aforementioned factors and the level of informality (as proxied by the share of the labor force not contributing to social security) for all countries where data are available. The authors find that, remarkably, the majority of correlation coefficients between these structural factors and informality are statistically significant, ranging between 0.29 and 0.83. Second, the authors use cross-country regression analysis to evaluate the importance of each proposed explanation regard- ing the causes of informality. The authors’ regression results indicate that law and order, business regulatory freedom, education, and sociodemo- graphic factors are all remarkably robust determinants of informality, with informality decreasing as these factors improve. Similarly, informal- ity decreases when the production structure shifts away from agriculture and when demographic pressures from youth and rural populations decline. The main results are summarized as follows: • Governance and regulation: A country’s governance structure will directly affect the opportunity cost of informality. The prevalence of Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 57 informality is influenced by the extent to which countries enforce law and regulation (more enforcement is associated with more compliance, especially if penalties are costly) and by the level of labor market regulatory freedom (more regulation/higher labor costs provide incen- ­ tives for firms to bypass regulation) (figure 1.12).6 Controlling for other factors, Loayza and Wada (2010) find that a 1 percent increase in the index of business regulatory freedom (index of law and order) is associated with 5 to 6 (2 to 3) percent lower informality worldwide. Figure 1.12 Correlation between Informality and Governance/Regulation 100 correlation: –0.74*** YEM N = 95 SYR % of informality in labor force 80 IRQ MAR LBN JOR 60 DZA IRN TUN EGY 40 LBY 20 0 1 2 3 4 5 6 law and order (index: higher, better) 100 correlation: –0.70*** SYR N = 99 80 % of informality in labor force MAR JOR 60 DZA IRN TUN EGY 40 20 0 3 4 5 6 7 8 business regulatory freedom (index: higher, less regulated) Source: Processed from Loayza and Wada 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 58 How Large Is Informality and Why Do We Care? • Productivity, demographics, and employment composition: A higher level of education is associated with lower informality because investments in human capital increase productivity and hence make business regu- lations less onerous and formal returns potentially larger. Other things equal, Loayza and Wada (2010) find that an additional year of educa- tion of the labor force is associated with a 5 to 6 percent lower level of informality worldwide (figure 1.13). Furthermore, a production Figure 1.13 Correlation between Informality and Education/Demographic Factors 100 correlation: –0.82*** SYR YEM IRQ MAR N = 84 % of informality in labor force JOR DZA 50 TUN EGY 0 –50 0 1 2 3 4 5 average years of secondary schooling 100 DJI YEM % of informality in labor force Iraq SYR WBG MAR LBN JOR IRN DZA 50 TUN EGY correlation: 0.83*** N = 110 0 –2 –1 0 1 2 sociodemographic factors, standarized Source: Processed from Loayza and Wada 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 59 ­tructure tilted toward agriculture, rather than toward the more s ­ complex processes of industry, favors informality by making legal ­protection and contract enforcement less relevant and valuable. Finally, a demographic composition with larger shares of youth or rural popu- lations is likely to increase informality by making monitoring more difficult and expensive, by placing bigger demands on resources for ­ training and acquisition of abilities, by creating bottlenecks in the ini- tial school-to-work transition, and by making the expansion of formal public services more problematic (Fields 1990; ILO 2004; Schneider and Enste 2000).7 • Size of the public sector: Informality may also be affected by the size of public sector. Informality tends to be less prevalent in economies with a larger state presence (figure 1.14). In fact, in economies heavily dominated by the state, informality is practically nonexistent, as in the former Soviet Union and other communist countries in Eastern Europe. The influence of the state on informal employment can be direct, through the absorption of labor and economic production. This is especially important for the MENA region, given the state’s traditionally important role as source of formal employment. The influence can also be indirect, through the links that the government establishes with private firms, by requiring them to register officially and comply with its regulations. Furthermore, given a sufficient level of quality, the size of the government can affect the state’s ability to monitor and enforce formal taxes and regulations.8 • Labor costs: Demand for formal employment will be determined by direct and indirect labor costs relative to informal workers. Higher indi- rect costs of labor, generally in the form of stricter employment protec- tion legislation and/or high labor taxes, are associated with higher levels of informality (Botero and others 2004; Djankov and Ramalho 2009; Grubb and Wells 1993). This is likely to be important in MENA, where, compared with international benchmarks, firing regulations are rather strict and labor taxes are rather high (Angel-Urdinola and Kuddo 2010). Figure 1.15 illustrates proxies for indirect labor costs (labor taxes and severance pay for redundancy dismissals) using available data from the Doing Business 2011 dataset and compares a selected group of non- GCC countries with respect to the world median (represented by the horizontal and ­vertical lines in the figure). Some economies in the region that display high levels of informality (such as Morocco, Syria, ­Lebanon, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 60 How Large Is Informality and Why Do We Care? Figure 1.14 Correlation between Informality and Government Size 100 YEM DJI SYR % of informality in labor force 80 MAR WBG IRN LBN JOR DZA 60 TUN 40 EGY correlation: –0.43*** LBY N = 114 20 0 5 10 15 20 25 30 government expenditure (% of GDP) 100 WBG SYR % of informality in labor force 80 MAR IRN JOR 60 EGY 40 20 correlation: –0.29*** 0 N = 65 0 20 40 60 80 public sector employment (% of total employment) Source: Processed from Loayza and Wada 2010. West Bank and Gaza economy, and Algeria) are also associated with high levels of labor taxes and with high severance payment schemes. Although some authors find a positive and significant relationship between the generosity of severance pay schemes and informality, the relationship between labor taxes and informality is inconclusive here (that is, some countries from ECA and OECD display high levels of labor taxes and low levels of informality). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 61 Figure 1.15 Correlation between Informality and Indirect Labor Costs 100 DJI YEM SYR WBG % of informality in labor force 80 IRQ MAR JOR LBN 60 DZA IRN TUN 40 EGY 20 0 0 10 20 30 40 50 labor tax and contributions (%) 100 DJI YEM SYR WBG % of informality in labor force 80 IRQ MAR JOR LBN DZA IRN 60 TUN EGY 40 20 0 0 20 40 60 severance pay for redundancy dismissal, in salary weeks Source: Based on Doing Business 2011 data. Note: Lines inside figures represent world medians. Structural Characteristics in MENA Countries MENA’s large youth bulge and educational deficit has important impli- cations for the region’s prevalence of informality. Loayza and Wada (2010) illustrate the importance of structural factors on informality for a selected group of countries in the MENA region where data are avail- able. The authors find that MENA countries, while being quite hetero- geneous, have characteristics that are associated with higher prevalence Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 62 How Large Is Informality and Why Do We Care? of informality. In particular, MENA countries have a large youth bulge and still display important education deficits. As mentioned before, countries with a larger youth to total population ratio are associated with higher informality levels. The right-hand panel of figure 1.16 illus- trates the percentage difference between the youth bulge in MENA versus the world’s median for a selected group of countries. Results indicate that with the exception of Lebanon, the youth bulge for all other countries in the region is larger than the world’s median (ranging Figure 1.16 Percentage Difference in Education and Youth Bulge with Respect to the World Median average years of education of labor force Jordan 33.9 Iran, Islamic Rep. 11.4 Egypt, Arab Rep. 7.6 Algeria –1.8 Syrian Arab Republic –11.9 Morocco –15.6 Tunisia –16.4 Iraq –46.5 Yemen, Rep. –77.8 –100 –50 0 50 % di erence youth ages 15–24 (% of total population) Iran, Islamic Rep. 20.3 Syrian Arab Republic 16.5 Yemen, Rep. 15.9 Algeria 14.9 Libya 13.1 West Bank and Gaza economy 13.0 Iraq 9.0 Djibouti 8.9 Egypt, Arab Rep. 7.7 Jordan 7.1 Morocco 6.6 Tunisia 4.7 Lebanon –4.3 –10 0 10 20 30 % di erence Source: Processed from Loayza and Wada 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 63 from 5 percent in Tunisia to 20 percent in the Islamic Republic of Iran). At the same time, some countries in the region, such as Iraq and the Republic of Yemen, display important education deficits, which are associated with lower levels of productivity (and thus higher levels of informality). Some countries in the region remain very rural by world standards. As illustrated by figure 1.17, the share of total population living in rural areas in countries such as the Republic of Yemen, Syria, Egypt, and Morocco is high by world standards. Not surprisingly, in these economies Figure 1.17 Percentage Difference in Rural Population and Agricultural Output with Respect to the World Median rural population as % of total population Yemen, Rep. 67.8 Egypt, Arab Rep. 32.2 Syrian Arab Republic 12.5 Morocco 0.1 Algeria –13.0 Tunisia –18.7 Iran, Islamic Rep. –21.8 Iraq –24.8 West Bank and Gaza economy –34.6 Jordan –55.3 Libya –59.2 Djibouti –65.9 Lebanon –68.7 –100 –50 0 50 100 % difference agriculture as % of GDP Syrian Arab Republic 95.9 Egypt, Arab Rep. 36.8 Morocco 35.0 Yemen, Rep. 26.6 Tunisia 2.6 Algeria –17.7 Iraq –31.4 Lebanon –36.7 West Bank and Gaza economy –39.4 Jordan –76.4 –100 –50 0 50 100 150 % difference Source: Processed from Loayza and Wada 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 64 How Large Is Informality and Why Do We Care? the share of agricultural to total economic output tends to be a larger component of the total economy. Because the agricultural sector remains largely informal in developing countries, countries with relatively large agricultural sectors will likely display higher informality rates than countries with relatively small agricultural sectors (such as Jordan and Lebanon). Finally, the public sector still plays a major employment and economic role in many countries in the region. Almost by definition, the large majority of employment in the public sector is formal. As such, control- ling for other factors, countries with relatively larger public sectors are likely to be associated with lower prevalence of informality. Figure 1.18 indicates that some countries in the region (such as Egypt, Jordan, and Syria) have public sectors that are quite large by international standards. Conclusions Informality is a fundamental characteristic of underdevelopment. Higher levels of informality are associated with lower levels of economic growth. Widespread informality implies that a large number of people and economic activities operate outside the legal-institutional framework. ­ International evidence indicates that widespread informality induces firms to remain suboptimally small, use irregular procurement and distri- bution channels, and constantly divert resources to mask their activities or bribe officials. Conversely, formal firms will be able to use resources with less regulatory restrictions, which enables them to be less labor intensive compared with their country’s labor endowments. Informality may also be a source of further economic retardation because it is associ- ated with misallocation of resources and entails losing the advantages of legality, such as police and judicial protection, access to formal credit institutions, and participation in international markets. At the same time, informality constitutes an important source of employment growth and of dynamism in many economies. Important adverse consequences of labor informality on human development outcomes are found. Informal workers in MENA have limited or no access to formal risk management instruments to cope with social risks. Social insurance systems in most MENA countries his- torically have been based on Bismarckian principles, in which formal sector actors (public and private sectors, employers and employees) contributed to social security programs, in return for relatively generous, multidimensional benefit packages, often including health insurance, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 65 Figure 1.18 Percentage Difference in Public Employment with Respect to the World Median public employment (% total) Egypt, Arab Rep. 115.3 Jordan 48.9 Syrian Arab Republic 23.7 West Bank and Gaza economy 21.9 Iran, Islamic Rep. –14.5 Morocco –56.1 –100 –50 0 50 100 150 % difference public wage bill (% GDP) West Bank and Gaza economy 95.6 Jordan 43.2 Libya 27.9 Morocco 20.8 Lebanon 4.9 Yemen, Rep. 3.1 Algeria –12.7 Syrian Arab Republic –15.3 Iran, Islamic Rep. –16.3 Egypt, Arab Rep. –21.1 –50 0 50 100 150 % difference Source: Processed from Loayza and Wada 2010 and WDI dataset. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 66 How Large Is Informality and Why Do We Care? old-age pension, d ­ isability and workers’ insurance, and, in some cases, housing, child care, and sports and recreation benefits. Those outside the formal sector, in both urban and rural areas, have had limited or no access to formal risk management instruments or other government benefits. As a result, informal workers cope with many risks though informal safety net a ­ rrangements. Although not as high as in other developing regions, informal employ- ment is rather prevalent in most MENA countries. Moreover, in some of them, informality is nearly as high as in the most informal countries in the world. This observation is important because a large informal sector denotes misallocation of resources (labor in particular) and inefficient utilization of government services, which can jeopardize the countries’ growth and poverty-alleviation prospects. A typical country in MENA produces about one-third of its GDP and employs 65 percent of its labor force informally (using the Schneider Index and the share of the labor force without social security coverage, respectively). These are remark- able statistics, especially from a human development perspective. These stylized facts indicate that more than two-thirds of all workers in the region may not have access to health insurance and/or are not contribut- ing to a pension system that would provide income security after retire- ment age. At the same time, from a fiscal perspective, these results indicate that about one-third of total economic output in the region remains undeclared and, therefore, not registered for tax purposes. Important variations are seen in the prevalence of informality across non-GCC MENA countries, depending on the availability of natural resources and labor. Resource-rich/labor-abundant economies (such as the Islamic Republic of Iran, Syria, and the Republic of Yemen) tend to display the highest labor informality rates in the region (as proxied by the share of the labor force not contributing to social security and the share of self- employment to total employment) but low undeclared output to total GDP (explained by the capital intensity of their production). Resource- poor/labor-abundant economies (such as Egypt, Jordan, Tunisia, and Morocco) display a high share of undeclared output and a lower share of the workforce not contributing to social security, likely reflecting their higher share of medium size (and semiformal) labor-intensive firms. Still, not all is negative about informality: The informal sector is likely to be an important mechanism for employment transition, especially among youth. Given that informality entails misallocation of resources and lower growth, should it be reduced at all costs? The answer to this important policy question is “no.” Although informality is suboptimal Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 67 with respect to a situation of streamlined regulations and adequate provi- sion of public services, it is indeed preferable to a fully formal economy that is unable to circumvent its regulation-induced rigidities. Also, the informal sector could be an important source of economic dynamism and may provide some needed flexibility for small firms to grow and operate productively. The crucial policy implication is that the mechanism of formalization matters greatly for its effects on employment, efficiency, and growth. If formalization is based purely on enforcement, it will likely lead to unemployment and low growth. On the other hand, if it is based on improvements in both the regulatory framework and the quality and availability of public services, it will bring about more efficient use of resources and higher growth. Also, the informal sector is a mechanism for workers to cope with unemployment risks and/or a sector of employment transition, especially for youth as they acquire experience. Annex Annex Table 1A.1 Sample Economies in the Informality Regression Non-contributor Schneider Shadow Country to pension scheme Self-employment Economy Index code Economy (74 economies) (62 economies) (93 economies) DZA Algeria √ √ √ ARG Argentina √ √ √ AUS Australia √ √ √ AUT Austria √ √ √ BGD Bangladesh √ √ √ BEL Belgium √ √ √ BOL Bolivia √ √ √ BWA Botswana √ BRA Brazil √ √ √ BGR Bulgaria √ BFA Burkina Faso √ √ CMR Cameroon √ √ CAN Canada √ √ √ CHL Chile √ √ √ CHN China √ √ √ COL Colombia √ ZAR Congo, Dem. Rep. √ √ √ COG Congo, Rep. √ CRI Costa Rica √ √ √ CIV Côte d’ Ivoire √ (table continues next page) Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 68 How Large Is Informality and Why Do We Care? Annex Table 1A.1  Sample Economies in the Informality Regression (continued) Non-contributor Schneider Shadow Country to pension scheme Self-employment Economy Index code Economy (74 economies) (62 economies) (93 economies) HRV Croatia √ √ DNK Denmark √ √ √ DOM Dominican Rep. √ √ √ ECU Ecuador √ √ √ EGY Egypt, Arab Rep. √ √ √ SLV El Salvador √ √ √ EST Estonia √ √ ETH Ethiopia √ √ FIN Finland √ √ FRA France √ √ √ DEU Germany √ √ √ GHA Ghana √ √ GRC Greece √ √ √ GTM Guatemala √ √ √ GNB Guinea-Bissau √ GUY Guyana √ HTI Haiti √ HND Honduras √ √ √ HKG Hong Kong SAR, √ √ China HUN Hungary √ √ ISL Iceland √ √ IND India √ √ IDN Indonesia √ √ √ IRN Iran, Islamic Rep. √ √ √ IRL Ireland √ √ √ ITA Italy √ √ √ JAM Jamaica √ √ √ JPN Japan √ √ √ JOR Jordan √ √ KEN Kenya √ √ KOR Korea, Rep. √ √ √ KWT Kuwait √ MDG Madagascar √ √ MWI Malawi √ MYS Malaysia √ √ √ MLI Mali √ MEX Mexico √ √ √ MAR Morocco √ √ √ MOZ Mozambique √ NAM Namibia √ √ (table continues next page) Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 69 Annex Table 1A.1  Sample Economies in the Informality Regression (continued) Non-contributor Schneider Shadow Country to pension scheme Self-employment Economy Index code Economy (74 economies) (62 economies) (93 economies) NLD Netherlands √ √ √ NZL New Zealand √ √ √ NIC Nicaragua √ √ √ NER Niger √ √ √ NOR Norway √ √ √ PAK Pakistan √ √ √ PAN Panama √ √ PNG Papua New Guinea √ PRY Paraguay √ √ √ PER Peru √ √ √ PHL Philippines √ √ √ POL Poland √ √ PRT Portugal √ √ SEN Senegal √ √ SLE Sierra Leone √ √ √ SGP Singapore √ √ √ SVK Slovak Republic √ √ SVN Slovenia √ ZAF South Africa √ √ ESP Spain √ √ √ LKA Sri Lanka √ √ √ SWE Sweden √ √ CHE Switzerland √ √ √ SYR Syrian Arab Rep. √ √ √ TZA Tanzania √ √ THA Thailand √ √ √ TGO Togo √ TTO Trinidad and √ √ Tobago TUN Tunisia √ √ √ TUR Turkey √ √ UGA Uganda √ √ GBR United Kingdom √ √ √ USA United States √ √ √ URY Uruguay √ √ √ VEN Venezuela, RB √ √ √ ZMB Zambia √ √ √ ZWE Zimbabwe √ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 70 Annex Table 1A.2  Definitions and Sources of Variables Used in Regression Analysis Variable Definition and construction Source Noncontributor to pension Labor force not contributing to a pension scheme as the percentage of total World Development Indicators, various scheme labor force. Average of 2000–2007 by country. years Self-employment Self-employed workers as percentage of total employment. Country ILO, data retrieved from laborsta.ilo.org averages but periods to compute the averages vary by country. Average of 1999–2007 by country. ECA countries are excluded (Loayza and Rigolini 2006). Schneider Shadow Economy Estimated shadow economy as a percentage of official GDP. Average of Schneider 2007; Schneider, Buehn, and index 1999–2007 by country. Montenegro 2010 Law and Order An index ranging 0 to 6 with higher values indicating better governance. PRS group, data retrieved from https:// Law and Order are assessed separately, with each subcomponent www.prsgroup.com comprising 0 to 3 points. Assessment of Law focuses on the legal systems, and Order is rated by popular observance of the law. Average of 2000– 2007 by country. Business Regulatory Freedom An index ranging 0 to 10 with higher values indicating less regulated. It is Gwartney, Lawson, and Norton 2008, the composed of following indicators: (1) price controls; (2) administrative Fraser Institute. Data retrieved from requirements; (3) bureaucracy costs; (4) starting a business; (5) extra www.freetheworld.com payments/bribes; (6) licensing restrictions; and (7) cost of tax compliance. Average of 2000–2007 by country. Labor Regulatory Freedom An index ranging 0 to 10 with higher values indicating less regulation. It is Gwartney, Lawson, and Norton 2008, composed of following indicators: (1) minimum wage; (2) hiring and firing ­ Fraser Institute, data retrieved from regulations; (3) centralized collective bargaining; (4) mandated cost of www.freetheworld.com hiring; (5) mandated cost of worker dismissal; and (6) conscription. Average of 2000–2007 by country. (table continues next page) Annex Table 1A.2  Definitions and Sources of Variables Used in Regression Analysis (continued) Variable Definition and construction Source Average years of secondary Average years of secondary schooling in the population aged 15 and over. ­ alculations Barro and Lee 1993, 2001, and c schooling The most recent score in each country is used. Sociodemographic factors Simple average of following three variables: (1) youth (aged 10–24) Calculations with data from World Devel- population as a percentage of total population; (2) rural population as a opment Indicators, ILO, and the UN percentage of total population; (3) agriculture as a percentage of GDP. All three variables are standardized before the average is taken. Average of 2000–2007 by country. Public sector employment Public sector employment as a percentage of total employment. Country ILO, data retrieved from laborsta.ilo.org averages but periods to compute the averages vary by country. Average of 2000–2007 by country. Government expenditure General government final consumption expenditure as a percentage of GDP. World Development Indicators, various Country averages but periods to compute the averages vary by country. years Average of 2000–2007 by country. Note: ECA = Europe and Central Asia; GDP = gross domestic product; ILO = International Labour Organization; UN = United Nations. 71   72 How Large Is Informality and Why Do We Care? Notes 1. Own-account workers are workers working on their own account or with one or more partners, holding a job defined as a self-employed job, and not having engaged on a continuous basis any employees to work for them during the reference period (ILO 1993). 2. The currency demand method was developed by Tanzi (1980, 1983) who assumes that transactions in the informal economy take place in the form of cash payments, which are most easily hidden from authorities. Hence, an increase in informality is likely to go hand in hand with an increase in currency demand. The excess demand for currency is estimated econometrically, con- trolling for factors traditionally associated with currency demand (such as income and interest rates) as well as factors associated with an increase of activity in the informal economy (such as direct/indirect tax burden, govern- ment regulation). The physical input (electricity consumption) method takes electricity consumption as a proxy for overall economic activity. Kaufmann and Kaliberda (1996) estimate the growth of the informal economy as the difference between the growth rate of official GDP and the growth rate of electricity consumption. 3. Additional descriptive statistics on these three informality indicators for countries in MENA are presented in the annex. 4. In all countries included in the analysis there is a positive correlation (ranging from 0.17 in the Republic of Yemen and 0.45 in Iraq) between being self- employed and not contributing to social security and a positive correlation (from 0.10 in Lebanon to 0.40 in the Republic of Yemen) between being an unpaid worker and not contributing to social security. At the same time, in all countries included in the sample, there is a negative correlation between being an unpaid worker and being self-employed (from −0.11 in Lebanon to −0.51 in the Republic of Yemen). This indicates that not contributing to social security captures some of the aspects of informality faced by both unpaid workers and self-employed, which are two very different groups. Correlations are available upon request. 5. Economies that belong to the non-GCC MENA group are Algeria, the Arab Republic of Egypt, the Islamic Republic of Iran, Iraq, Jordan, Lebanon, Libya, Morocco, Syrian Arab Republic, Tunisia, and the Republic of Yemen as well as the West Bank and Gaza economy. Countries that belong to the GCC MENA group are Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates. 6. Law enforcement is measured by the Index on the Prevalence of Law and Order obtained from The International Country Risk Guide. The index is a proxy for both the quality of formal public services and the government’s enforcement strength. To measure labor regulation, the Index of Business Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 73 Regulatory Freedom and Labor Market Regulatory Freedom is used, taken from Fraser Foundation’s Economic Freedom of the World report. 7. Labor productivity is proxied by (1) the average years of secondary schooling of the adult labor force (from Barro and Lee 2001) and (2) an index of sociode- mographic factors composed of the share of youth in the population, the share of rural population, and the share of agriculture in GDP, obtained from the WDI dataset. 8. Government size is proxied by (1) the share of employment in the public ­ sector to total employment (from ILO’s LABORSTA dataset) and (2) government consumption expenditure as a percentage of GDP (from WDI). References Angel-Urdinola, D., and A. Kuddo. 2010. “Key Characteristics of Employment Regulation in the Middle East and North Africa.” SP Discussion Paper 1006. Washington, DC: World Bank. Angel-Urdinola, D., J. G. Reis, and C. Quijada. 2009. “Informality in Turkey: Size, Trends, Determinants and Consequences.” Mimeo. Washington, DC: World Bank. Angel-Urdinola, D., and K. Tanabe. 2011. “Micro-Determinants of Informal Employment in the Middle East and North Africa Region.” Mimeo. Washington, DC: World Bank. Barro, R. W., and J.-W. Lee. 1993. “International Comparisons of Educational Attainment.” Journal of Monetary Economics 32 (3): 363–394. ———. 2001. “International Data on Educational Attainment: Updates and Implications.” Oxford Economic Papers 53 (3): 541–563. Botero, J. C., S. Djankov, R. La Porta, F. Lopez-de-Silanes, and A. Shleifer. 2004. “The Regulation of Labor.” Quarterly Journal of Economics 119 (4): 1339– 1382. Cho, Y. 2011. “Informality and Protection from Health Shocks: Lessons from Yemen.” Unpublished manuscript. Washington, DC: World Bank. De Soto, H. 1989. The Other Path: The Invisible Revolution in the Third World. New York: HarperCollins. Djankov, S., and R. Ramalho. 2009. “Employment Laws in Developing Countries.” Journal of Comparative Economics 37 (1): 3–13. EC, IMF, OECD, UN, and World Bank. 2008. System of National Accounts 2008. European Commission, International Monetary Fund, Organisation for Economic Co-operation and Development, United Nations, and World Bank, New York. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 74 How Large Is Informality and Why Do We Care? Fields, G. 1990. “Labour Market Modeling and the Urban Informal Sector: Theory and Evidence.” In The Informal Sector Revisited, ed. David Turnham, Bernard Salomé, and Antoine Schwarz, 49–69. Paris: OECD. Friedman, E., S. Johnson, D. Kaufmann, and P. Zoido-Lobaton. 2000. “Dodging the Grabbing Hand: The Determinants of Unofficial Activity in 69 Countries.” Journal of Public Economics 76 (3): 459–493. Galal, A. 2005. “The Economics of Formalization: Potential Winners and Losers from Formalization in Egypt.” In Investment Climate, Growth, and Poverty, ed. Gudrun Kochendorfer-Lucius and Boris Pleskovic. Washington, DC: World Bank. Gasparini, L., and L. Tornarolli. 2006. “Labor Informality in Latin America and Caribbean: Patterns and Trends from Household Survey and Micro-data.” Mimeo. Washington, DC: World Bank. Grubb, D., and W. Wells. 1993. “Employment Regulation and Patterns of Work in EC Countries.” Economic Studies 21. Paris: OECD. Gwartney, J. D., and R. Lawson, with S. Norton. 2008. Economic Freedom of the World: 2008 Annual Report. Vancouver: Economic Freedom Network. Hart, K. 1973. “Informal Income Opportunities and Urban Employment in Ghana.” Journal of Modern African Studies 11 (1): 61–89. ILO (International Labour Organization). 1993. “Resolutions Concerning International Classification of Status in Employment.” Adopted by the 15th International Conference of Labour Statisticians. Geneva: ILO. ———. 2002. “Decent Work and the Informal Economy: Sixth Item on the Agenda.” 90th International Labour Conference: 2002. Report VII (2A). Geneva: ILO. http://www.ilo.org/public/english/standards/relm/ilc/ilc90/pdf /rep-vi.pdf. ———. 2004. Global Employment Trends for Youth. Geneva: ILO. Johnson, S., D. Kaufmann, A. Shleifer, M. I. Goldman, and M. L. Weitzman. 1997. “The Unofficial Economy in Transition.” Brookings Papers on Economic Activity 2: 159–239. Kaufmann, D., and A. Kaliberda. 1996. “Integrating the Unofficial Economy into the Dynamics of Post Socialist Economies: A Framework of Analyses and Evidence.” Policy Research Working Paper 1691. Washington, DC: World Bank. Loayza, N., A. M. Oviedo, and L. Serven. 2005. “The Impact of Regulation on Growth and Informality Cross-Country Evidence.” Policy Research Working Paper 3623. Washington, DC: World Bank. Loayza, N., and J. Rigolini. 2006. “Informality Trends and Cycles.” Policy Research Working Paper 4078. Washington, DC: World Bank. ———. 2009. “Informal Employment: Safety Net or Growth Engine?” Mimeo. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 How Large Is Informality and Why Do We Care? 75 Loayza, N., and T. Wada. 2010. “Informal Labor in the Middle East and North Africa: Basic Measures and Determinants.” Mimeo. Washington, DC: World Bank. Pallares-Miralles, M., C. Romero, and E. Whitehouse. 2012. “International Patterns of Pension Provision II: A Worldwide Overview of Facts and Figures.” Social Protection and Labor Discussion Paper 1211. World Bank, Washington, DC. Perry, G., W. Maloney, O. Arias, P. Fajnzylber, A. Mason, and J. Saavedra-Chanduvi. 2007. Informality: Exit and Exclusion. Washington, DC: World Bank. Schneider, F. 2004. “The Size of the Shadow Economies of 145 Countries All over the World: First Results over the Period 1999 to 2003.” DP 1431. Bonn: IZA. Schneider, F., A. Buehn, and C. E. Montenegro. 2010. “Shadow Economies All over the World : New Estimates for 162 Countries from 1999 to 2007.” Policy Research Working Paper Series 5356. World Bank, Washington, DC. Schneider, F., and D. Enste. 2000. “Shadow Economies: Size, Causes, and Consequences.” Journal of Economic Literature 38 (1): 77–114. Tanzi, V. 1980. “The Underground Economy in the United States: Estimates and Implications.” Banca Nazionale des Lavoro 135:4. ———. 1983. “The Underground Economy in the United States: Annual Estimates, 1930–1980.” IMF Staff Papers 30:2. World Bank. 2009. “Turkey Country Economic Memorandum: Informality: Causes, Consequences, Policies.” Report 48523-TR. Washington, DC: World Bank. ———. 2013. Jobs for Shared Prosperity: Time for Action in the Middle East and North Africa. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Cha p t e r 2 Profile and Micro-Determinants of Informality SUMMARY: This chapter assesses the main micro-determinants of infor- mal employment in MENA. Analysis in the chapter quantifies the patterns of labor informality (defined as the share of all employment with no access to social security) according to age, gender, educational level, employment sector, profession, marital status, employment status, and geographic area in a selected group of non-GCC MENA countries. Countries in the MENA region are quite heterogeneous in terms of size, economic development, and demographic structure. Results indicate that the sizes of the public and agricultural sectors are perhaps the main correlates of informality in the MENA region. Countries where agricultural employment still constitutes a large share of overall employment (such as Morocco and the Republic of Yemen) display higher levels of overall informality. On the other hand, informality is lower in countries with larger public sectors and more urban- ization, such as Egypt and Lebanon. The existence of a large public sector, which is still associated with generous benefits and better employment quality, creates an important segmentation between public and private ­ employment in many MENA countries. Age, education, and firm size also represent important determinants of informality. Informality rates are gen- erally highest among youth between ages 15 and 24, a group that accounts for 24 to 35 percent of total employment in most MENA countries, and among workers in small firms, whereas informality is generally lower among workers who have attained a university education and who work in the public administration. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 77   78 Profile and Micro-Determinants of Informality Introduction Chapter 2 aims to understand the key micro-determinants of informal employment in MENA. The behavior of informal employment in MENA differs widely between GCC countries and non-GCC countries. Informal employment in GCC countries, as proxied by the share of the labor force not contributing to social security, is rather low (at 6.4 per- cent) and prevalent mainly among the self-employed. In non-GCC countries, using the same proxy, the share of the labor force not contrib- uting to social security is high (at 67.2 percent) and prevalent mainly among wage earners (Angel-Urdinola and Tanabe 2011). The main pur- pose of the chapter is to quantify the patterns of informal employment according to age, gender, educational level, employment sector, profes- sion, marital status, employment status, and strata (that is, urban or rural). Based on availability of micro-data and given the general focus of the report on addressing informality from a human development stand- point, analysis in this chapter is limited to a selected group of non-GCC countries for which the relevant data are available: Egypt, Iraq, Jordan, Lebanon, Morocco, Syria, and the Republic of Yemen. This chapter con- sists of three main sections. The first section provides a brief macroeco- nomic background for these selected countries, focusing on patterns of economic and employment growth. The second section presents the profile of informality, composed of a set of statistics describing the share of workers in the informal sector according to various socioeconomic characteristics. The third section presents the main determinants (or cor- relates) of informality through regression analysis. Countries in the region are quite heterogeneous in terms of size, eco- nomic development, demographic structure, and employment composi- tion. Each of the countries included in the analysis has important economic and demographic factors that are likely to affect the level and characteristics of informal employment, such as the size of the agricul- tural sector compared with the secondary and tertiary sectors (higher levels of agricultural employment are associated with higher labor infor- mality), size of the public sector (a larger public sector is associated with lower levels of labor informality), educational level of the labor force (a better educated labor force is associated with lower levels of labor informality), and age composition of the labor force (countries with younger populations are associated with higher levels of labor informal- ity), among others (table 2.1 and figure 2.1). Important variations are found: Some of the countries are still very rural, and agricultural Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Table 2.1 Economic and Demographic Factors for Selected Non-GCC Countries % employment GDP per capita Agriculture, value % employment % employment in with university   (2000$ constant)a added (% of GDP)a age 15–24a the public sector education % unemployment a Egypt, Arab Rep. 1,785.83 13.22 23.1 30.0 16.8 8.7 Iraq 730.79 8.57 23.2 36.9 12.8 17.5 Jordan 2,244.83 2.58 21.5 35.6 25.8 12.7 Lebanon 5,858.76 6.94 28.7 13.5 17.2 9.0 Morocco 1,718.14 14.64 34.7 11.1 6.6b 9.6 Syria Arab Republic 1,329.85 16.99 32.3 26.8 7.4 8.4 Yemen, Rep. 559.97 14.34 21.9 38.6 5.4 15.0 Source: Household surveys for available years. a. World Development Indicators (WDI) dataset, last available year 1999–2008. b. WDI for 2006. 79   80 Profile and Micro-Determinants of Informality Figure 2.1 Employment Composition by Sector (Using Latest Available Year) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Morocco Egypt, Syrian Iran, Algeria West Bank Jordan Arab Rep. Arab Islamic and Gaza Republic Rep. economy services (%) industry (%) agriculture (%) Source: World Development Indicators (all available non-Gulf Cooperation Council economies). Note: Latest year available in the period 2003–2007. ­ mployment constitutes almost half of overall employment (such as the e Republic of   Yemen and Morocco); others have better educated workers (such as Jordan, Egypt, and Lebanon); and in some countries, the public sector still accounts for a significant share of overall employment (such as Jordan, Egypt, the Republic of   Yemen, Iraq, and Syria). Macroeconomic Context Countries in the region have displayed favorable economic growth in recent years. Some basic macroeconomic trends for the countries included in the analysis are presented to provide context to the analysis that follows. As discussed in chapter 1, the level of economic develop- ment and other macroeconomic variables, such as recent economic growth and employment composition, are likely to be important factors to understand a country’s profile and determinants of informality. This is particularly relevant in MENA, given the social, economic, and cultural heterogeneity of countries in the region. Figure 2.2 illustrates the average yearly economic growth rate of GDP per capita in a selected group of non-GCC countries for the periods 2000 to 2005 and 2005 to 2009. The region’s economic performance was above the world’s average in both time periods. Between 2005 and 2009, annual per capita growth rates showed important variation across countries, with some displaying rapid Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 81 Figure 2.2  Yearly GDP Per Capita Growth Rate 6% 5% 4% 3% 2% 1% 0% ea or n pu a Ye lg ic ab n Jo p. ic ia ab L p. e n ia . Isl Tun o ld ep Re iby M da c Ar no bl am is m er Re Re ag or oc ,R r t, ba W er A yp Le Av Ar n Ira Eg n r ia Sy 2005–2009 2000–2005 Source: World Development Indicators. a. Unweighted average (selected countries). growth (such as Lebanon, Egypt, Jordan, Morocco, and Tunisia), others moderate growth (such as the Islamic Republic of Iran, Libya, and Syria), and some poor growth (such as the Republic of   Yemen and Algeria). Some countries performed much better between 2005 and 2009 com- pared with 2000 to 2005 (mainly Lebanon and the Islamic Republic of Iran), whereas in others, the opposite occurred (such as Algeria and the Republic of  Yemen). While employment growth in the region has been among the highest in the world in the past decade, the level of employment creation has been unable to keep up with population growth (see discussion in ­ chapter 1 and figure 2.3). This demographic dynamic contributed to high unemployment rates, especially among youth, and a difficult school-to-work transition. Joblessness in many MENA countries remains notable, especially among women. School-to-work transition patterns highlight the inci- dence of joblessness and the disadvantaged position of women in MENA. Figure 2.4 illustrates the patterns of school-to-work transition in selected countries. This transition is measured by the length of time between when 50 percent of the population is enrolled in school and when 50 percent is employed. It takes as little as one year in the Republic of Yemen to up to approximately 18 years in Iraq. In devel- oped countries, a comparable process takes on average 1.4 years (Angel- Urdinola and Semlali 2010). Large differences exist in ­ school-to-work Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 82 Profile and Micro-Determinants of Informality Figure 2.3  Growth in Employed and Working-Age Population in Non-GCC MENA Countries, 1998–2009 MENA (non-GCC) 250 200 55% not working 150 Millions 56% not working 100 50 0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 WAP Employed Source: World Bank, based on the ILO’s EAPEP (Economically Active Population, Estimates and Projections) database. Note: GCC - Gulf Cooperation; MENA - Middle East and North Africa. ­atterns by gender. Upon exiting the school system, the transition p majority of women in the countries considered (except for the Republic of Yemen) enter into joblessness (that is, unemployment and/or inactiv- ity), and only a small proportion successfully move into employment. According to the definition used, the school-to-work transition never fully occurs for women in most of the countries. This has important implications for the region’s economic growth potential. First, interna- tional experience indicates that greater economic equality between women and men is associated with poverty reduction, higher GDP, and better governance (Klasen 1999). Recent studies indicate that many economies in MENA display lower participation rates than those pre- dicted given their age and education structures. If female labor force participation in these countries rose to the level predicted by women’s age and education structure, household earnings could increase substan- tially (World Bank 2003). The service sector has been an important source of employment growth in recent years, with crucial implications for informality. A closer look at available data on employment growth by country Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 83 Figure 2.4 School-to-Work Transition (for Ages 15–35) in MENA Egypt, Arab Rep., 2006 Iraq, 2007 100 100 80 80 percent percent 60 60 40 40 20 20 0 0 15 17 19 21 23 25 27 29 31 33 15 17 19 21 23 25 27 29 31 33 Lebanon, 2010 Yemen, Rep., 2006 100 100 80 80 percent percent 60 60 40 40 20 20 0 0 15 17 19 21 23 25 27 29 31 33 15 17 19 21 23 25 27 29 31 33 Jordan, 2010 Morocco (Urban), 2010 100 100 90 80 80 70 percent percent 60 60 50 40 40 30 20 20 10 0 0 15 17 19 21 23 25 27 29 31 33 15 17 19 21 23 25 27 29 31 33 % at school % employed % jobless between 2000 and 2007 reveals several interesting patterns. Data indicate important variation in annual employment growth across ­ countries in the region, from an increase of almost 8 percent per year in Algeria to a decrease of almost 3 percent per year in Syria (left-hand panel of figure 2.5). As illustrated by the right-hand panel of figure 2.5, in many countries, the service sector (mainly commerce and construc- tion) has been an important engine of employment growth, generally followed by the industrial sector. The contribution of agriculture to total employment growth has been important in some economies such as Egypt and the West Bank and Gaza economy, but it has also Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 84 Figure 2.5 Employment Growth between 2000 and 2007 for Selected Countries 10% total annual employment growth share of sectors to yearly employment growth rates 7.8% (2000−2007) (2000−2007) 8% West Bank and 6% Gaza economy 4.4% Syrian Arab Republic 4% 3.3% Morocco 2.9% 2.5% Jordan 2% 1.1% Iran, Islamic Rep. 0% Egypt, Arab Rep. –2% Algeria –2.6% –100% –50% 0% 50% 100% –4% p. p. ria an co bl b om d Re pu ra Re on an Agriculture (%) Industry (%) Services (%) oc rd ge ic y Re n A ic ab Jo or ec nk Al am ria M Ar za Ba Sy Isl t, Ga est yp n, W Eg Ira Source: ILO’s KILM dataset. Profile and Micro-Determinants of Informality 85 Table 2.2 Composition of Employment Growth in the Services Sector for Selected Countries, 2000–2007 percent Egypt, Arab Iran, Islamic Algeria Rep. Rep. Jordan Construction 29.7 28.3 42.8 6.2 Commerce and personal services 44.2 31.0 21.5 35.0 Transport 9.5 19.4 19.2 13.9 Value added/social services 16.7 21.4 16.4 44.8 Source: World Development Indicators and ILO’s KILM dataset. con­tributed to negative growth in employment (that is, to employment destruction) in countries such as Syria, Jordan, and the Islamic Republic of Iran. In Syria, for instance, the ­negative growth in employ- ment between 2000 and 2007 is largely explained by a rapid decrease in agricultural employment. The expansion of the service sector (mainly construction and com- merce) has gone hand in hand with an increase in informality and self- employment in the region. The expansion of employment in the service sector has been an important feature of many economies in MENA in recent years. As presented in table 2.2, the construction and commerce sectors account together for 40 to 70 percent of all employment growth in the service sectors recently. In Morocco, employment growth was 5.1 percent in transport and 3.7 percent in construction for the period 1998 to ­ 2003 (World Bank 2009). These results are likely to have key implications for informality trends because construction and commerce are sectors often associated with high rates of informal employment. Indeed, as illus- trated in figure 2.6, informality (as proxied by the Schneider Index and by the share of self-employment of total employment) has increased in recent years, at a time when employment in the construction and commerce sec- tors has been expanding. Informality Profile This section presents the profile of informality for the countries included in the analysis.1 The profile consists of a set of statistics describing work- ers in the informal sector according to various characteristics, such as socioeconomic status, educational level, age, gender, strata (urban or rural), marital status, and occupation, among others. The informality pro- file is presented for all workers in the sample (including the public sec- tor), and separately for private sector workers. The analysis includes urban Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 86 Profile and Micro-Determinants of Informality Figure 2.6  Annual Growth of Informality for Selected Economies Annual growth rate 2000–2007 1.80 1.61 1.60 1.40 1.28 1.21 1.17 1.20 1.08 1.00 0.86 0.76 0.80 0.58 0.60 0.40 0.31 0.20 0.00 . p. p. n an co sia ep ic ria no Re Re bl oc rd ,R ge ni pu ba ic ab Tu Jo or en Al am Re Le M Ar m Isl ab Ye t, yp n, Ar Eg Ira n ria Sy Source: World Development Indicators and ILO’s KILM dataset. Table 2.3 Informality Rates for Selected Countries percent All workers Urban workers Rural workers Egypt, Arab Rep., 2006 58.3 42.6 70.0 Iraq, 2007 66.9 63.5 76.1 Jordan, 2010 44.2 47.5 28.3 Lebanon, 2010 56.2 48.3 66.7 Morocco, 2010 — 72.7 — Syrian Arab Republic, 2011 71.0 65.9 76.3 Yemen, Rep., 2006 91.4 84.7 94.0 Source: Angel-Urdinola and Tanabe 2011. Note: — = Data not reliable. and rural workers. Many informality studies (Perry and others 2007) exclude rural employment from the analysis because in other regions it is predominantly informal. In MENA countries, this is not necessarily the case because of an important public sector presence (and, thus, formal employment) in rural areas. Although rural employment remains more informal than urban employment (except in Jordan), informality rates in both rural and urban areas are more comparable than in other regions of the world (table 2.3). Informality is a more persistent phenomenon among the poor. As expected, informality generally decreases as wealth increases. Nevertheless, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 87 Box 2.1 Informality in Tunisia 2005–2009 Because of data availability, informality in Tunisia is proxied as the share of overall employment working in the private sector without a contract. Also, analysis could be conducted for urban areas only. Therefore, results for Tunisia cannot be com- pared with other countries in the region. As suggested by the table below, urban informality increased by 6 percent (from 50.5 percent to 53.5 percent) between 2005 and 2009. The increase was more pronounced among men (8 percent increase) and among workers with university education (34 percent increase). Not surpris- ingly, men account for the majority of all informal workers (73.5 percent in 2009) and informality rates are higher among men than women, because many women generally self-select themselves into public sector/formal jobs (Angel-Urdinola and Tanabe 2011). Informality rates are increasing rapidly among young adults (25–34) and adults (35–54). Informality Rates in Urban Areas, 2005–2009 percent % total informal Informality Informality workers, 2009 rate, 2005 rate, 2009 % change Total urban 100.0 50.5 53.5 6.1 Gender Male 73.5 53.7 57.9 7.8 Female 26.5 41.2 41.4 0.3 Age group 15–24 10.8 67.9 70.9 4.4 25–34 32.2 50.4 54.7 8.5 35–54 50.5 45.3 49.3 8.8 55–64 6.5 53.1 52.0 −2.1 Education Primary or below 38.5 69.4 72.6 4.6 Basic 39.0 44.9 50.5 12.5 Secondary 0.8 vocational 53.1 46.0 −13.3 Tertiary 21.8 15.4 20.6 33.5 Source: Angel Urdinola, Brodmann, and Hilger 2011. Note: Results based on 2005–2009 averages. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 88 Profile and Micro-Determinants of Informality in some MENA countries, informality remains significant even among the wealthier segments of the population. In the Republic of   Yemen for instance, more than two-thirds of all workers who belong to the richest households work in the informal sector. Yet, in other countries such as Lebanon, informality rates are significantly lower for the wealthiest seg- ments of the population (figure 2.7). In most countries in the region, the vast majority of formal workers are employed in the public sector. Labor markets in many MENA countries are still influenced by the legacy of a large public sector, which accounts for about 29 percent of overall employment in the Arab world (Elbadawi and Loayza 2008), and civil service in many MENA countries is larger than in other countries with similar levels of income and economic structure. Historically, in countries such as Egypt, the growth in the civil ­ service was the result of a social contract in the 1970s and 1980s whereby the government effectively offered employment guarantees to university graduates and to graduates of vocational secondary schools and training institutes. Despite the fact that employment growth in the public sector has slowed dramatically in recent years, public sector employment (­government and public enterprises) in most countries still accounts for more than 60 percent of all formal sector employment in MENA Figure 2.7 Informality Rates by Quintile of Per Capita Consumption for Selected Countries 100 95.9 90 86.1 82.8 82.5 80 70 58.7 60 percent 50 40 35.1 30 20 10 0 Yemen, Rep. Syrian Arab Republic Lebanon poorest Q2 Q3 Q4 richest Source: Angel-Urdinola and Tanabe 2011. Note: Q = quintile. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 89 figure 2.8). In some countries, such as the Republic of   (­ Yemen and Iraq, formal employment is almost entirely associated with public employ- ment. The public sector remains the main engine of formal employment: In Egypt about 45 percent of all new formal jobs (about 260,000) created in the economy between 1998 and 2006 were in the public sector (Angel-Urdinola and Semlali 2010). Given the weight of the public sector in overall formal employment (figure 2.9), changes in the size of the public sector are likely to affect overall informality trends, especially given that formal private employ- ment growth remains limited. In particular, the creation of formal private sector jobs has not been sufficient to offset the downsizing of the public sector in many countries (Radwan 2007). Radwan argues that formal businesses in many MENA countries face important challenges that restrain their capacity to grow, such as dealing with complex bureaucratic procedures; access to poor infrastructure, credit, and technologies; and high labor taxes. A recent assessment of the private sector in Morocco reveals that excessive regulation of the labor market has pushed much of the economic activity into the informal sector (World Bank 1999). At the same time, the private sector in many MENA countries is primarily com- posed of small and medium enterprises (SMEs), which represent about 95 percent of all registered enterprises. The great majority of SMEs in MENA have fewer than five workers and are characterized by high levels Figure 2.8 Size of Private Salaried Formal Sector versus Other Sectors 100 90 80 % of total employment 70 60 50 40 30 20 10 0 Yemen, Rep. Iraq Syrian Arab Egypt, Morocco Jordan Lebanon Republic Arab Rep. (Urban) private formal private informal public self-employed and employers unpaid Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 90 Profile and Micro-Determinants of Informality Figure 2.9  Distribution of Formal Employment for Selected MENA Countries 100 80 57.3 63.0 68.1 70.4 60 percent 82.5 40 96.5 98.4 20 42.8 37.0 31.9 29.6 17.5 3.5 1.6 0 Morocco Jordan Egypt, Arab Lebanon Syrian Arab Yemen, Iraq (Urban) Rep. Republic Rep. private public of informality; low participation of women; concentration in low-growth sectors; low use of modern technologies; and a low level of product qual- ity, competitiveness, diversification, and innovation. Overall, a large pub- lic sector and restrictive regulations contribute to squeeze the formal private sector and limit its dynamics (see chapter 5, part 1). The private formal sector in the region is still nascent. Employment in the formal private sector is almost nonexistent in the Republic of  Yemen and Iraq, and below 10 percent of total employment in Syria, Egypt, and Morocco, although it is somewhat larger in Jordan and Lebanon. As seen earlier, this reflects a number of factors, including the country’s production structure, the large size of the public sector, which effectively competes for resources and talent with the private formal sector, and the design of pension systems (which in the Republic of Yemen and Iraq do not extend in reality to the private sector). To the extent that formality also reflects higher productivity, more productive workers and especially firms bear the brunt of taxation. This small pri- vate formal sector coexists with a large informal sector that includes both low-productivity firms and workers, but also larger firms that have secured favorable application of regulations through rent seeking. Especially if compared with the latter, one might argue that the small formal private sector bears disproportionately the taxation burden. An important transition is made from informal employment into pub- lic sector employment as young people reach prime age adulthood. Figure 2.10 illustrates employment patterns by age for urban workers in a selected group of countries. Informality rates are very high among youth Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 91 Figure 2.10 Employment Status by Age for Selected Countries, Urban Areas Only Egypt, Arab Rep., 2006 70 60 employment rate (%) 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group Jordan, 2010 70 60 employment rate (%) 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group Iraq, 2007 70 employment rate (%) 60 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group public private formal private informal self-employed figure continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 92 Profile and Micro-Determinants of Informality Figure 2.10  Employment Status by Age for Selected Countries, Urban Areas Only (continued) Yemen, Rep., 2006 70 60 employment rate (%) 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group Morocco (Urban), 2010 70 60 employment rate (%) 50 40 30 20 10 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 Age group public private formal private informal self-employed Source: Angel-Urdinola and Tanabe 2011. between ages 15 and 24. After age 24, informality decreases rapidly until individuals reach prime working age (40 to 45 years). After age 40, infor- mality rates are lower (20 to 30 percent). This rapid decrease in informal- ity rates goes hand in hand with a rapid increase in public employment, which suggests that informal workers enter into public sector jobs as they move from youth into adulthood. Not surprisingly, many individuals in the region queue in the informal sector until they find a job in the public administration. Still, one has to be cautious in interpreting these results as they are likely to reflect vintage effects since, especially in countries Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 93 like Egypt, more public sector jobs were available to earlier cohorts of workers. These trends are very different from those observed in Latin America. In Mexico, for instance, although informality rates also decrease by age, the observed transition occurs not between informality and public employment, but between informality and self-employment (Perry and others 2007). Self-employment is low among youth and young adults but increases rapidly after individuals reach age 50, suggesting a transition from ­ public employment into self-employment as individuals retire from the public administration. As illustrated by figure 2.10, the share of indi- viduals who work in the formal private sector remains almost nonexistent in countries such as the Republic of Yemen and Iraq (at all ­ age groups), which suggests a very limited formal private sector. Even in more dynamic and diversified economies, such as Egypt, Jordan, and Morocco, formal p ­ rivate sector employment accounts for a maximum of 25 percent of overall employment at all age groups. Finally, interest- ing patterns are seen in the trend in self-employment by age. Although overall rates of self-employment remain low, especially among youth, the share of self-employment to overall employment increases steadily as people reach retirement age, suggesting that workers (especially those who decide to retire early) transition from public employment into self-employment. Informality rates are generally higher for wage earners outside the public administration and the self-employed. Results indicate that infor- mality rates among wage earners (who account for 50 percent of overall employment for most countries in the analysis) range between 40 and 60 percent. Informality rates among the self-employed (who account for 15 to 36 percent of all employment for most countries in the analysis) are even higher, from 80 percent in Lebanon to almost 100 percent in Iraq and the Republic of Yemen (table 2.4). The promotion of self- employment/micro-entrepreneurship continues to be a core strategy to boost employment, and so creation of new, flexible, and innovative mechanisms to ensure pension and social security coverage for the self- employed should be developed and implemented (see chapter 5). Self-employed individuals have somewhat different profiles than informal wage earners. As mentioned earlier, self-employed workers in MENA are generally informal because they rarely participate in social security contributing schemes. Nevertheless, this group of workers dis- plays somewhat different characteristics as compared with other informal workers (mainly wage earners). Table 2.5 presents a set of descriptive Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 94 Table 2.4 Informality Rates by Employment Status Egypt, Arab Rep., Morocco (Urban), Syrian Arab Iraq, 2007 Jordan, 2010 2006 Lebanon, 2010 2010 Republic, 2011 Yemen, Rep., 2006 Employment Pop. Pop. Pop. Pop. Pop. Pop. Pop. status share % Inf. share % Inf. share % Inf. share % Inf. share % Inf. share % Inf. share % Inf. Wage worker 68.9 51.7 71.7 56.5 64.3 42.6 65.4 43.1 51.4 67.1 55.6 52.2 23.9 63.7 Employer — — 11.2 94.6 13.2 78.4 4.8 84.0 3.9 98.4 6.7 89.4 — — Self-employed 20.5 100.0 15.4 98.5 9.9 82.1 27.2 80.1 29.6 98.1 24.0 93.4 20.3 100.0 Source: Angel-Urdinola and Tanabe 2011. Note: Inf. = informality; Pop. = population; — = not available. Profile and Micro-Determinants of Informality 95 Table 2.5 Self-Employed versus Informal Wage Earners (Basic Characteristics) Egypt, Arab Rep. Lebanon Informal Informal wage earners Self-employed wage earners Self-employed Average age (in years) 31.8 38.3 31.9 41.8 erage years of education Av (in years) 6.9 6.3 11.4 9.5 Average wage rate (in LC) 2.4 — 4.55 6.01 Time at job — — 5.12 14.1 Gender Men 78.6 71.6 66.1 85.7 Women 21.4 28.4 33.9 14.3 Age group 15–24 32.5 8.8 34.2 7.2 25–34 32.7 31.5 33.9 26.5 35–54 27.5 49.5 25.9 48.7 55–64 7.3 10.1 5.9 17.6 Education Primary or below 56.4 61.3 19.0 33.1 Preparatory/secondary general 7.3 6.6 22.4 30.0 Secondary vocational 30.2 25.5 27.0 23.3 Tertiary education 6.1 6.7 31.6 13.4 Note: LC = local currency; — = not available. statistics for Lebanon and Egypt highlighting important differences in the characteristics of workers in these two groups: • Self-employed are generally older and less educated. Results in table 2.5 indicate that in Egypt and Lebanon, self-employed workers are older than informal wage earners (6 years older in Egypt and about 10 years older in Lebanon, on average). In both countries, the majority of self- employed (almost half) is between 35 and 54 years of age. The share of youth (15–24) who work as self-employed is very low (9 percent in Egypt and 7 percent in Lebanon). Self-employed workers are less educated than informal wage earners. On average, self-employed workers in Egypt (Lebanon) have attained one (two) years of educa- tion fewer than informal wage earners. In Lebanon, only 13.4 ­percent of all self-employed have attained tertiary education compared with 31.6 percent among wage earners. • In Lebanon, self-employed workers have more stable jobs and earn rela- tively higher wages . Despite being less educated, self-employed Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 96 Profile and Micro-Determinants of Informality ­ orkers in Lebanon (which account for about 33 percent of all w employment) earn on average wages that are 30 percent higher than those among informal wage earners, which could reflect experience. Also, self-employed workers claim to have worked in the same job for 14 years (on average) as compared with 5 years among informal wage earners. Age, gender, and education also constitute important correlates of informality. Figure 2.11 presents a basic set of correlations between informal employment and individual characteristics such as age, gender, ­ and years of education. Not surprisingly, results indicate that age and education are negatively correlated with informality (that is, higher age and more education are associated with less informality). The size of the correlation between years of education and informality is large for all countries (from −0.35 in Lebanon to −0.47 in Syria), suggesting an important negative relationship between education and informal employ- ment. Nevertheless, the magnitude of the correlation between age and informality varies across countries and strata. For instance, the negative ­ association between age and informality seems to be larger in Egypt, Syria, and Iraq than in Morocco, the Republic of Yemen, and Lebanon, and generally stronger in urban than in rural areas (figure 2.11 shows results only for urban areas). Being a woman is associated with higher informality rates in some countries and with lower informality rates in others, mainly depending on the overall structure the country’s employment (figure 2.11). In countries where agricultural employment constitutes an important share of overall employment, such as Egypt and the Republic of Yemen, being a woman is associated with higher levels of informality because women are often employed in unpaid/subsistence agriculture. In coun- tries where public employment represents a significant share of overall employment, such as Iraq, and Syria, being a woman is associated with lower levels of informality. Given low overall levels of female labor force participation in these countries, women who participate in the labor force (generally those with higher levels of education) self-select into public sector jobs. Hence, as illustrated in figure 2.12, informality rates are higher among men than among women in Iraq, Syria, Lebanon, Morocco, and Jordan, and higher among women than among men in the Republic of Yemen. However, once the sample is restricted to private sector workers, informality rates between men and women are equally high in most countries. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 2.11  Basic Correlations by Strata (Informality and Individual’s Characteristics) age and informality status as a woman and informality years of education and informality Jordan Yemen, Rep. 0.31 –0.27 Jordan –0.09 Morocco (Urban) Jordan 0.00 –0.35 Lebanon total –0.10 Lebanon Egypt, Arab Rep. –0.05 –0.41 Egypt, Arab Rep. –0.13 Yemen, Rep. Syrian Arab Rep. –0.08 –0.42 Morocco (Urban) –0.15 Iraq Morocco (Urban) –0.09 –0.42 Iraq –0.21 Syrian Arab Rep. Iraq –0.18 –0.43 Yemen, Rep. –0.32 Egypt, Arab Rep. Lebanon –0.22 –0.47 Syrian Arab Rep. Jordan Yemen, Rep. 0.34 –0.29 Lebanon –0.08 Lebanon Jordan 0.00 –0.30 Jordan urban –0.13 Morocco (Urban) Morocco (Urban) –0.10 –0.36 Egypt, Arab Rep. –0.13 Iraq Egypt, Arab Rep. –0.11 –0.41 Yemen, Rep. –0.17 Yemen, Rep. Lebanon –0.20 –0.42 Morocco (Urban) –0.24 Syrian Arab Rep. Iraq –0.31 –0.42 Iraq –0.35 Egypt, Arab Rep. Syrian Arab –0.31 –0.46 Syrian Arab Rep. Rep. Morocco (Urban) Yemen, Rep. 0.28 –0.18 Morocco (Urban) Jordan Syrian Arab Rep. 0.09 –0.25 Jordan –0.04 Yemen, Rep. Egypt, Arab Rep. 0.08 –0.36 Lebanon rural –0.10 Lebanon Iraq 0.07 –0.36 Egypt, Arab Rep. –0.14 Iraq Morocco (Urban) 0.06 –0.38 Iraq –0.18 Syrian Arab Rep. Jordan 0.00 –0.43 Yemen, Rep. –0.25 Egypt, Arab Rep. Lebanon –0.20 –0.46 Syrian Arab Rep. –1.00 –0.50 0.00 –1.00 –0.50 0.00 0.50 1.00 –1.00 –0.50 0.00 Source: Angel-Urdinola and Tanabe 2011. 97   98 Profile and Micro-Determinants of Informality Figure 2.12 Informality Rates by Gender all workers 100 90 80 70 60 percent 50 40 30 20 10 0 an ic q on ) p. . ep an Ira bl Re rd an ,R rb pu Jo ab b (U en Re Le Ar m co ab Ye t, oc yp Ar or Eg n M ria Sy private sector workers 100 90 80 70 60 percent 50 40 30 20 10 0 an ic q n ) p. . ep an no Ira bl Re rd ,R rb pu ba Jo ab (U en Re Le Ar m co ab Ye t, oc yp Ar or Eg n M ria Sy male female Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 99 Over workers’ life cycles, an increase in employment in the public sector mirrors the decline in informal salaried employment. A closer look at informality rates by age group and educational attainment ­ indicates important differences for workers in the private sector. Informality rates are generally highest among young people between the ages of 15 and 24, a group that in most countries in the sample accounts for 24 to 35 percent of total employment. Above age 25, infor- mality rates decrease rapidly up to age 54 in most countries. Informality rates increase again for workers between ages 55 and 64 (which make up 5 to 8 percent of total employment in most countries) as some workers find employment in the informal sector after they retire from ­ their formal jobs. Informality rates among workers who attained pri- mary and/or basic education (who account for at least 50 percent of overall employment in most countries in the region) are generally much higher than among workers who attained secondary vocational and/or tertiary education. Differences in informality rates by age and education are less pronounced for workers in the private sector. Indeed, in some countries such as the Republic of Yemen and Morocco, differences in informality rates by age and educational attainment for workers in the private sector are negligible (figures 2.13 and 2.14). Informality is generally higher in the primary sector, with important implications for countries with a large agriculture sector. Results suggest that the great majority of workers in agriculture and mining activities (which account for as little as 5 percent of employment in Iraq and as much as 30 percent in Morocco) work in the informal s ­ ector (table 2.6). In the tertiary sectors (that is, services), informality rates vary among countries, ranging from 46 percent in Lebanon to 93 percent in the Republic of   Yemen. Among workers employed in the public ­ administration/social services (which in countries such as Egypt, Iraq, and Syria account for as much as one-third of total employment), infor- mality rates are below 20 percent. This is probably explained by the existence (in some countries) of fixed-term contracts in the public sector. Determinants of Informality Although a profile of informality is informative, the main drawback is that it cannot be used to disentangle its determinants. For example, the fact that a group of workers (such as agricultural workers) displays high rates of informality may be due in large part to other characteris- tics of the group (such as the educational level of the group’s members). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 100 Profile and Micro-Determinants of Informality Figure 2.13 Informality Rates by Age Groups all workers 100 97.3 95.2 90.9 89.0 90 84.7 87.1 79.2 80 73.2 69.1 70 63.9 58.8 60 51.2 53.3 percent 49.9 50 40 30 20 10 0 q an p. n ) ic . ep an no Ira bl Re rd ,R rb pu ba Jo ab (U en Re Le Ar m co ab Ye t, oc yp Ar or Eg n M ria Sy private sector workers 99.5 98.5 96.9 99.9 99.9 100 92.4 92.2 91.9 90 87.4 77.5 78.0 80 72.6 69.6 70 64.7 60 percent 50 40 30 20 10 0 q an p. n n) ic . ep Ira no bl Re ba rd ,R pu ba r Jo b (U en ra Re Le m co ,A ab Ye oc pt Ar y or Eg n M ria Sy 15–24 25–34 35–54 55–64 Source: Processed from Angel-Urdinola and Tanabe 2011. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 101 Figure 2.14 Informality Rates by Highest Educational Level Completed all workers 100 97.4 90.9 90 86.2 82.2 77.8 75.7 80 71.9 70 60 percent 50 40.7 40 36.5 28.4 30 25.4 23.7 24.3 22.0 20 10 0 Iraq Jordan Egypt, Lebanon Morocco Syrian Arab Yemen, Arab Rep. (Urban) Rep. Rep. private sector workers 99.4 97.5 99.9 100 93.5 95.5 94.9 88.2 88.3 90 80.4 80 70 61.9 60 50.5 50.7 percent 50 40.2 40 34.3 30 20 10 0 Iraq Jordan Egypt, Lebanon Morocco Syrian Arab Yemen, Arab Rep. (Urban) Rep. Rep. primary or below preparatory/sec. general secondary vocational tertiary education Source: Angel-Urdinola and Tanabe 2011. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 102 Table 2.6 Informality Rates by Sector of Employment Egypt, Syrian Arab Iraq Jordan Arab Rep. Lebanon Morocco (Urban) Republic Yemen, Rep. Pop. Pop. Pop. Pop. Pop. Pop. Pop. Sector a  share % Inf. share % Inf. share % Inf. share % Inf. share % Inf. share % Inf. share % Inf. Industry Primary sector 4.6 52.2 6.4 92.6 25.5 94.1 6.6 94.4 28.4 94.1 31.6 92.4 18.4 95.6 Secondary sector 30.7 78.7 32.7 63.1 21.3 65.8 24.5 75.3 12.6 78.9 6.1 75.5 3.4 86.0 Tertiary sector 27.5 71.8 49.0 75.1 28.0 62.7 68.9 46.0 54.5 82.2 37.7 87.2 40.0 93.3 Public administration/ social services 37.2 12.2 12.0 30.7 25.3 11.0 — — 4.5 18.5 24.6 17.6 38.2 11.7 Source: Angel-Urdinola and Tanabe 2011. Note: Inf. = informality; Pop. = population; — = not available. a. Primary sector (agriculture); secondary sector (manufacturing and construction); tertiary sector (wholesale, transport, services), public administration, and social services (including education and health). Profile and Micro-Determinants of Informality 103 To provide more insights about the determinants or correlates of infor- mality, this section assesses informality through regression analysis using a simple probit regression model. The dependent variable of the regres- sion model is a binary variable that takes a value of one if the worker is employed in the informal sector (that is, if the worker does not contrib- ute to social security) and zero otherwise. Separate regressions are pro- vided for the full sample and for workers in the nonagricultural sector.2 The main independent variables used include (1) strata (an urban dummy), (2) demographic characteristics of the worker (a male dummy, a married dummy, and the worker’s age group), (3) the highest educa- tional level attained by the worker, (4) employment status and sector of the worker, and (5) ownership of the firm where the worker is employed (using a dummy for publicly owned firms). As these are cross-sectional regressions, results should not be interpreted causally. The main results are summarized as follows: • Strata: In many countries, especially in Latin America (Perry and others ­ ctivities 2007), rural employment is mainly associated with agricultural a and mainly informal. In MENA countries, this is not necessarily the case, because of an important public sector presence (and, thus, of for- mal employment) in rural areas. Controlling for other factors, urban workers are only 3 to 12 percent less likely to be employed informally than otherwise similar workers in rural areas in Egypt and Lebanon. Although rural employment remains more informal than urban employment, informality rates in both rural and urban areas are com- parable. Indeed, in some countries such as Iraq, Morocco, and the Republic of Yemen, the difference in the probability of workers’ for- mality between urban and rural areas is small and/or not statistically significant (table 2.7). • Gender: The effect of being male on the probability of working in a formal job varies across countries (figure 2.15 and table 2.7). In Egypt and Morocco, controlling for other factors, being a male worker is asso- ciated with a 4 to 12 percent lower probability of being employed informally, as is generally the case in many developing countries (see Perry and others 2007 for estimates from Latin America). On the other hand, in Iraq, Lebanon, Jordan, and Syria, being male is actually associated with a 6 to 17 percent higher probability of working in the informal sector. This result is probably due to female workers (generally educated ones) participating in the labor force who tend to queue for Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 104 Table 2.7 Marginal Increase in the Probability of Being “Informal” according to the Characteristics of the Worker (Nonagricultural Employment Only) Egypt, Syrian Arab Dependent variable: Arab Rep. Iraq Yemen, Rep. Lebanon Republic Morocco (Urban) Jordan informal employment All Priv. All Priv. All Priv. All Priv. All Priv. All Priv. All Priv. Urban dummy −9.5 −8.0 N.S. N.S. N.S. −1.8 −13.5 −12.1 4.1 1.8 −2.3 −1.5 N.S. N.S. Male dummy −12.0 −9.7 17.1 5.5 N.S. N.S. 13.3 16.0 11.9 3.2 −3.7 −3.4 3.8 13.9 Married −13.9 −11.3 N.S. N.S. −8.0 N.S. −11.5 −10.1 −5.9 −0.9 −1.2 −1.1 N.S. N.S. Age group 25–34 −12.9 −10.3 −7.6 N.S. −16.1 −1.7 −5.7 −6.2 −26.9 −4.8 −2.6 −1.9 −4.7 N.S. 35–54 −29.2 −23.3 −12.7 −1.5 −26.2 −4.7 −12.4 −12.5 −33.8 −7.4 −3.3 −2.3 N.S. N.S. 55–64 −30.3 −35.9 −24.3 N.S. −20.2 N.S. −15.3 −15.4 −36.4 −9.6 −3.2 −2.1 N.S. 15.5 Education Middle school −11.0 −10.4 −8.8 N.S. −17.5 −2.5 −14.6 −12.2 −12.1 −4.5 N.S. N.S. −10.0 −16.7 High school −17.1 −15.0 −22.7 −2.1 −26.7 N.S. −25.7 −25.4 −27.4 −7.9 −14.4 −11.5 −15.4 −36.9 Tertiary −25.2 −33.7 −25.3 N.S. −19.9 −6.6 −33.2 −37.2 −31.8 −17.7 −21.6 −20.6 −17.9 −46.6 Sector Tertiary sectora −4.6 −3.8 N.S. N.S. 25.1 3.7 −13.1 −10.7 23.9 11.8 4.7 4.0 12.4 18.2 Public administration and social services versusb −11.1 −11.1 −14.2 −7.7 N.S. N.S. — — — — −5.8 −6.4 −9.7 −11.7 Ownership Public firm −59.4 — −82.4 — −80.7 — −52.8 — −74.6 — −33.3 — −48.8 Observations 8,752 5,244 12,362 4,361 8,500 3,939 13,373 11,254 32,461 20,471 6,500 6,252 4,441 2,237 Source: Angel-Urdinola and Tanabe 2011. Note: All = all workers; Priv. = private sector workers. All coefficients are multiplied by 100. Coefficients that are not statistically significant are denoted by N.S. Underlined coefficients are significant at a 10 percent confidence level. Other coefficients are significant at a 5 percent confidence level. Omitted categories: Age group: 15–25; Education: primary education or below. Employment sector: secondary sector (manufacturing and construction); Ownership: private firms. — = not available. a. Tertiary sector (wholesale, transport, services). b. Public administration and social services (including education and health). Figure 2.15 Marginal Increase in the Probability of Being “Informal” according to Gender and Marital Status percent Being male and informality Being married and informality Lebanon 16.0 0.0 Jordan Jordan 13.9 0.0 Iraq Syrian Arab Republic 3.2 0.0 Yemen, Rep. private Iraq 0.0 sector –0.9 Syrian Arab Republic Yemen, Rep. 0.0 –1.1 Morocco (Urban) Morocco (Urban) –3.4 –10.1 Lebanon Egypt, Arab Rep. –9.7 –11.3 Egypt, Arab Rep. Iraq 17.1 0.0 Jordan Lebanon 13.3 0.0 Iraq Syrian Arab Republic 11.9 –1.2 Morocco (Urban) all Jordan 3.8 –5.9 Syrian Arab Republic workers Yemen, Rep. 0.0 –8 Yemen, Rep. Morocco (Urban) –3.7 –11.5 Lebanon –12.0 Egypt, Arab Rep. –13.9 Egypt, Arab Rep. –20 –10 0 10 20 –20 –15 –10 –5 0 Source: Angel-Urdinola and Tanabe 2011 (see table 2.7 for more details). 105   106 Profile and Micro-Determinants of Informality formal jobs in the public sector (Angel-Urdinola and Semlali 2010). In the Republic of   Yemen, gender does not seem to be an important determinant of informality. • Marital status: In the MENA region, important associations are found between marriage and labor outcomes. Recent literature shows that having good and stable employment is an important social requirement for individuals, especially young men, to get married. For instance, Assaad, Binzel, and Gadallah (2010) use Egypt’s Labor ­ Market Panel Survey of 2006 (ELMPS 06) to study the role of employ- ment (that is, having a good, fair, or poor job) on the timing of mar- riage. The authors find that having a better job leads men to a faster transition into marriage. These results are consistent with the findings herein, as being married, controlling for other factors, is associated with a 10 to 14 percent lower probability of working in the informal sector in Egypt and Lebanon, and a 2 to 8 percent lower probability in Morocco, the Republic of Yemen, and Syria (figure 2.15 and table 2.7). Jordan, marital status is not significantly associated with informal In ­ employment. • Age: Controlling for other factors, younger workers are more likely to work in the informal sector (table 2.7). Results from Egypt, the Repub- lic of Yemen, and Syria indicate that adults aged 35 and older are 13 to 34 percent less likely to work in the informal sector than youth aged 15 to 24. In Iraq, Lebanon, Jordan, and Morocco, the association between age and informality is less strong, because adults 25 and older are only 2 to 8 percent less likely to work in the informal sector than youth aged 15 to 24. It is worth noting that acquiring informal jobs is a way for young individuals to enter the labor market, gain experience, and eventually move into formal employment, as informality decreases quickly with age. The effect of age on informality is generally lower in magnitude for private sector workers. This is expected since the pri- vate sector in MENA remains largely informal. To better illustrate this phenomenon, figure 2.16 shows informality rates by age for men and women for Syria and Egypt. Results in the left-hand panel show that informality rates decrease rapidly as age increases up to ages 40 to 45 and increases again thereafter as individuals retire. Results also indicate that early retirement is quite common, especially among women. In the private sector (shown in the right-hand panel of figure 2.16), informality rates by age are rather flat and high, especially for women, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 2.16 Informality Rates by Age Group (Syrian Arab Republic and the Arab Republic of Egypt) all Syrian Arab Republic private 100 100 women women men 90 90 80 80 men 70 70 % informal % informal 60 60 50 50 40 40 30 30 20 20 10 10 0 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 age Egypt, Arab Rep. age 100 100 women 90 90 80 women 80 men 70 70 % informal % informal 60 men 60 50 50 40 40 30 30 20 20 10 10 0 0 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 15–19 20–24 25–29 30–34 35–39 40–44 45–49 50–54 55–59 60–64 age age Source: Angel-Urdinola and Tanabe 2011. 107   108 Profile and Micro-Determinants of Informality suggesting that the negative slope between age and informality is driven by workers entering public sector employment as they reach prime working age. • Highest educational level attained: Controlling for other factors, more education is associated with a lower probability of being employed in the informal sector. The negative relationship between attaining higher education and having a lower probability of being employed informally is much lower for private sector workers (confirming the results pre- sented in the profile of informality). Controlling for other characteris- tics, attaining middle school (high school) is associated with a 5 to 18 (12 to 37) percent lower probability of working in the informal sector compared with otherwise similar workers who attained at most p ­ rimary school. Attaining tertiary education is associated with up to a 47 per- cent lower probability of being employed informally compared with otherwise similar workers who attained at most primary school. In some countries (for example, Iraq, Syria, and the Republic of Yemen), having completed high school decreases the probability of workers being employed informally as much as (or even more than) having attained tertiary education. Generally, one would expect the opposite result (as in Egypt, Jordan, and Lebanon), namely, a lower probability of being employed informally for workers with tertiary education. Finally, to control for heterogeneity in skills beyond educational attain- ment, quintile dummies (omitting the lowest quintile) representing the results of a cognitive nonverbal test (measuring workers’ logical and analytical skills) in Lebanon and Syria are included. Interestingly, this factor was not a significant determinant of informality (results are avail- able upon request).3 • Sector of employment: Controlling for other factors, the association between informality and sector of employment varies across coun- tries. In countries where the tertiary sector is more developed toward high value-added services such as financial services, transport, tour- ism, and communications (such as in Egypt and Lebanon), workers in the tertiary sector are associated with a 4 to 13 percent lower probability of working informally compared with workers in the sec- ­ ondary sector (manufacturing and construction). On the other hand, in countries where the tertiary sector is mainly geared toward low value-added personal services and wholesale or retail (such as in the Republic of Yemen and Syria) workers in the tertiary sector are Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 109 ­ ssociated with a 4 to 25 percent higher probability of working infor- a mally compared with workers in the secondary sector. Workers in public administration and social services are associated with 6 to 14 percent lower informality rates than otherwise similar workers in the secondary sector. • Public sector employment: Controlling for other factors, this variable is perhaps the most important determinant of informality. In all countries where information about firm ownership is available, workers in the public sector are associated with a 30 to 85 percent higher probability of working formally compared with otherwise similar workers in the private sectors. Indeed, as illustrated in f­igure 2.9, the public sector hosts a significant share of all formal employment. As such, changes in the size of the public sector relative to the private sector will likely be important determinants of informality dynamics (box 2.2). • Firm size: Data on firm size were only available for a few countries in the region (Egypt, Iraq, Jordan, and Morocco). For these countries, firm size dummy variables are included in the regression analysis (small-size firm dummy, fewer than 10 workers; medium-size firm dummy, 10 to 50 workers; and large-size firm dummy, more than 50 workers). Regression coefficients are shown in figure 2.17; the results indicate an important association between informality and firm size. Workers in medium-size (large-size) firms are 16 to 21 (17 to 53) percent less likely to work in the informal sector compared with ­ workers in small-size firms. Conclusions Socioeconomic conditions in the region are quite heterogeneous, which has important implications for informality. Each country included in the analysis has important economic and demographic ­ factors that are likely to affect the level and characteristics of informal employment, such as the size of the agricultural sector compared with the secondary and ter- tiary sectors (higher levels of agricultural employment are associated with higher labor informality), the size of the public sector (a larger public sector is associated with lower levels of informality), the educa- tional level of the labor force (a more educated labor force is associated with lower levels of labor informality), and the age composition of the labor force (countries with younger populations are associated with Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 110 Profile and Micro-Determinants of Informality Box 2.2 Informality Trends and the Size of the Public Sector (Egypt, 1998–2006) Informality dynamics: Informality in Egypt, as proxied by the share of all employment not contributing to social security according to data from the Egypt Labor Market Panel Survey, increased rapidly from 49.0 percent in 1998 to 58.3 percent in 2006. This result holds true using other proxies for informality such as the share of workers without a contract, share of all employment in small firms, share of unpaid to total employment, and share of self-employed to total employment. These dynamics are consistent with Assaad (2009), who finds that informality, as proxied by the share of workers without a contract, increased from 57 percent in 1998 to 61 percent in 2006. Informality Rates in the Arab Republic of Egypt, for 1998 and 2006 (urban and rural areas) 100 80 62.8 58.3 56.0 60 52.6 percent 49.0 40.0 40 20 12.6 10.5 11.3 8.1 0 no contribution self- no contract rm size (<5 unpaid to social employment workers) work security 1998 2006 rate of change Source: Angel-Urdinola and Tanabe 2011. As suggested by the table below, the increase in informality is largely explained by the fact that the public sector contracted as a share of total employment (from 47 per- cent in 1998 to 39 percent in 2006). Indeed, informality within the public and private sectors displayed only a slight increase during the period of study. These results are also consistent with previous work by Radwan (2007), who argues that one of main box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 111 Box 2.2  Informality Trends and the Size of the Public Sector (Egypt, 1998–2006) ­ (continued) reasons for the rise of informal employment in MENA has been the decline in public sector employment as a share of total employment. Informality Rates and Employment Shares by Sector (Egypt, Arab Rep., 1998 and 2006) 1998 2006 Pop. share % informal Pop. share % informal Private 53.3 60.8 60.9 67.3 Public 46.8 2.5 39.1 4.3 Source: Angel-Urdinola and Semlali 2010; Angel-Urdinola and Tanabe 2011. Note: Inf. = informal; Pop. = population. Figure 2.17 Marginal Increase in the Probability of Being “Informal” according to Firm Size (Omitted Category, Small Firms, 2–9 Workers) percent –21.0 Morocco (Urban) –52.6 Egypt, Arab Rep. –15.9 –32.4 Jordan –18.8 –24.6 –18.3 Iraq –17.2 medium rm dummy (10−50 workers) large rm dummy (50+ workers) Source: Angel-Urdinola and Tanabe 2011. higher levels of labor informality), among others. There are important variations across countries: Some of them are still very rural, and agricul- tural employment constitutes almost half of overall employment (the Republic of Yemen and Morocco); some countries have more educated workers (Egypt, Jordan, and Lebanon); and in some, the public sector still represents a significant share of overall employment (Egypt, the Republic of Yemen, Iraq, Jordan, and Syria). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 112 Profile and Micro-Determinants of Informality Several factors make informality in the MENA region a persistent phenomenon that may continue to rise in the years to come. The current demographic transition, the reduced importance of public employment, and the increase in private low-productivity employment are all likely to contribute to an increase in informality in the near future. Declining fertility and mortality rates, coupled with an increasing share of young people who attain tertiary education (notably women), are important factors contributing to the expansion of the informal sector. Informal employment is increasingly becoming (even for some educated individu- als) a permanent state of employment associated with low pay, poor working conditions, and limited mobility to the formal sector. One of the main reasons for the increase of informal employment in MENA is the decline in public sector employment as a share of total employment. The creation of formal private sector jobs has not been sufficient to off- set the downsizing of the public sector in many countries. Formal busi- nesses in many MENA countries face important challenges that restrain their capacity to grow, such as dealing with complex bureaucratic proce- dures; access to poor infrastructure, credit, and technologies; and high labor taxes. Overall, the formal private sector is very small in most MENA countries, and some might argue that this sector might also unfairly bear the brunt of paying all taxes. Informality is a more pronounced phenomenon among the poor. Informality rates among workers generally decrease as their household wealth increases. Nevertheless, in some countries informality remains sig- nificant even among the wealthier segments of the population. In the Republic of Yemen and Morocco, for instance, more than two-thirds of all workers who belong to the richest households work in the informal sector. This result has important implications. As indicated in chapter 1, informal employment in non-GCC countries, albeit large, produces little output relative to its employment share, suggesting very low levels of worker productivity. This generally occurs when informal workers are constrained in access to credit, services (such as utilities), and/or technology. Contrary to other developing regions, patterns of urban and rural infor- mality in MENA are somewhat similar, because of the presence of the public sector in rural areas. Urban workers are only 5 to 12 percent less likely to be employed in the informal sector than otherwise similar work- ers in rural areas. Controlling for age and education, in countries such as Iraq, Morocco, and the Republic of Yemen, the probability of being employed in the informal sector does not vary across urban and rural workers. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 113 The size of the public sector and the size of the agricultural sector are perhaps the main determinants of informality in the MENA region. Countries in the MENA region are quite heterogeneous in terms of size, economic development, and demographic structure. Countries where agricultural employment still constitutes a large share of overall employment (such as Morocco and the Republic of Yemen) are associ- ated with higher levels of overall informality. On the other hand, coun- tries with larger public sectors and more urbanization such as Egypt, Syria, Jordan, and Lebanon are associated with lower levels of overall informality. The existence of a large public sector, which is still associ- ated with generous benefits and better employment quality, creates an important segmentation between public and private employment in many MENA countries. At the same time, the private sector in many MENA countries is p ­ rimarily composed of small and medium enter- prises, which account for about 95 percent of all registered enterprises. In all countries where information about firm ownership is available, workers in the public sector are associated with a 45 to 80 percent lower probability of working informally as compared with otherwise similar workers in the ­ private sector. Self-employed and agricultural workers in most MENA countries are associated with a 10 to 20 percent higher likelihood of being employed informally compared with other- wise similar workers employed as wage earners. Age, gender, firm size, and education also constitute important deter- minants of informality. Informality rates are generally highest among young people between ages 15 and 24, a group that in most countries in the sample accounts for 24 to 35 percent of total employment. Above age 25, informality rates decrease rapidly in most countries. Attaining second- ary technical and tertiary education is associated with a 40 percent higher probability of being employed formally compared with otherwise similar workers who attained at most a primary education. The relationship between gender and informality varies across countries. In Egypt and Morocco, controlling for other factors, being a male worker is associated with a 4 to 12 percent lower probability of being employed informally, as women generally queue for formal sector jobs. On the other hand, in Iraq, Jordan, and Lebanon, being a male is actually associated with a higher probability of working in the informal sector as is generally the case in many developing countries. Finally, results indicate an important associa- tion between informality and firm size. Workers in medium and large firms are 16 to 53 percent less likely to work in the informal sector com- pared with workers in small firms. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 114 Annex Annex Table 2A.1  Description of the Data Used for the Micro-Analysis Country Survey Year Description Egypt, Arab Labor Market Panel 2006 ELMPS was conducted by the Population Council and the Central Agency for Public Mobilization and Rep. Survey (ELMPS) Statistics (CAPMAS) with support of USAID Egypt and the Ford Foundation. ELMPS 06 is designed as a panel survey in the sense that it follows the same households and individuals that were interviewed in the Egypt Labor Market Survey of 1998 and reinterviews them. Individuals who split from the original 1998 households in the intervening period are also tracked and interviewed together with their entire household. Of the 8,371 households interviewed in 2006, 3,701 were households that were interviewed in ELMS 1998, 2,167 were splits from the original households, and 2,498 were part of an entirely new refresher sample. Sample size is 17,364 individuals and 8,371 households. The surveys contain rich information about individuals’ education, employment status, occupation, economic activity, firm size, wage, pension contribution, and so on. Egypt, Arab Household Income, 2005–2008 HIECSs have been conducted every five years since 1995 by CAPMAS, and they have been the main Rep. Expenditures and (and the only official) source for poverty and inequality data in Egypt. In late 2007, faced with Consumption multiple policy demands arising from social tensions, the authorities decided to make the data Surveys (HIECS) collection more frequent, and they decided to revisit in 2008 the households interviewed in the February during HIECS 2004–2005. The new survey was conducted during 2008–2009. As part of these efforts, CAPMAS revisited in 2008 the households from one month of 12-month HIECS 2004– 2005, applying the same questionnaire. In February 2009 CAPMAS repeated the panel, revisiting again the full set of addresses of the 2004–2005 February sample. The data used in this report come from the last survey conducted in April 2008–March 2009. The sample provides information on informality, earnings, and poverty for 9,228 individuals of working age 16–64 and 3,456 households. table continues next page Annex Table 2A.1  Description of the Data Used for the Micro-Analysis (continued) Country Survey Year Description Iraq Household 2006–2007 In an effort to reduce poverty and promote social development, the Ministry of Planning and Socioeconomic Survey Development Corporation and the Central Organization for Statistics and Information Technology (IHSES) had undertaken IHSES with the support of the World Bank. Providing essential data for understanding the nature and causes of poverty among Iraqi households, IHSES is the largest household social and economic survey ever conducted in Iraq and reached a total of 18,144 households. IHSES provides information about education, labor, health, income, and expenditure in Iraq. IHEES has a sample size of 17,822 households and 127,189 individuals. Jordan Labor Market Survey 2010 The Jordan Labor Market Panel Survey (JLMPS 2010) was carried out by the Economic Research (JLMPS) Forum in cooperation with the National Center for Human Resource Development and the Jordanian Department of Statistics. For the first time in Jordan, detailed information about labor market experiences and behaviors is available in JLMPS. JLMPS has a sample size of 25,969 individuals, containing rich information about individuals’ education, employment status, occupation, economic activity, firm size, wage, pension contribution, and decision making of labor force participation. JLMPS allows for a much richer linking of individual characteristics with labor market outcomes. Lebanon National Survey of 2004 The NHS is the multipurpose survey conducted by the Ministry of Social Affairs, the Central Household Living Administration for Statistics, and the UNDP in 2004. The purpose of this survey is to assess the Conditions (NHS) economic and social conditions of the households in Lebanon. It provides varied data and social indicators on Lebanese households’ demographic status, educational conditions, employment and unemployment, health insurance, chronic diseases, disability, and leisure activities. The survey also provides data concerning the characteristics of residences and their available appliances. Out of the sample size of 14,948, 13,003 households—consisting of 56,513 individuals—completed the data in the questionnaire. The response rate reached 87% of the households sampled. table continues next page 115   116 Annex Table 2A.1  Description of the Data Used for the Micro-Analysis (continued) Country Survey Year Description Lebanon Lebanon Employer- 2011 The Lebanon Employer-Employee Survey was conducted by the World Bank under the Lebanon Employee Survey MILES program. It is a nationally representative household-based survey covering a sample of 1,841 households. The survey collects basic information such as age, education, employment for the entire household, detailed information on employment (current and history), skills and training, wages, work benefits for each individual in the household who is over 15 years of age and is either unemployed, self-employed, or a salaried employee, and the level of cognitive and noncognitive skills of workers. Morocco Household 2000–2001 HCES was conducted in 2000–2001 by the Ministry of Economic Forecasts and Planning. The Consumption and purpose of this survey was to provide information concerning the living condition of households Expenditure Survey and the structure of their consumption and expenditure. HCES provides demographic (HCES) characteristics of the family members, education, employment, household consumption and expenditures, and income. HCES has 14,243 household and 85,509 individual observations. Morocco Morocco Household 2009–2010 MHYS 2009–2010 was administered from December 2009 through March 2010 and collected and Youth Survey information from a nationally representative sample of 2,000 households across the country (1,216 (MHYS) households were urban and 784 were rural) on their demographic and educational characteristics, economic activities, migration, and social program participation. Data on household asset ownership were used to construct a household wealth index and classify households into welfare deciles. In addition to the household module, which collected information on all members, a separate youth module focused on young people aged 15 to 29 in the 2,000 surveyed households. Consequently, information related to youth economic inclusion, community participation, and use of key public services was collected from 2,883 young people. The survey thus gathered information on understudied issues related to youth, such as labor force participation and intermediation, career choices and perceived employment opportunities, use of free time, and use of youth-oriented recreational and educational services that complement formal education. table continues next page Annex Table 2A.1  Description of the Data Used for the Micro-Analysis (continued) Country Survey Year Description Syrian Arab Household Income 2003–2004 HIES 2003–2004 was conducted by the Central Bureau for Statistics, Syria’s official statistical agency. Republic and Expenditure HIES provides information on education, employment, household expenditure, and housing Survey (HIES) conditions. HIEA has a sample size of 29,790 households and 173,330 individuals. Syrian Arab Syria Employer- 2010–2011 EES 2010–2011 was conducted by the World Bank in the context of the Syria MILES program. EES Republic Employee Survey (EES) collects matched employer-employee data from firms and workers in registered firms in the manufacturing and services sector in Syria. A representative sample was drawn from firms interviewed in the 2009 World Bank Investment Climate Assessment. A random sample of workers in each firm was drawn systematically. EES has a sample of 116 firms and 961 individuals. The survey collects information on employment (current and history), skills and training, wages, and work benefits. Yemen, Rep. Household Budget 2005–2006 HBS 2005–2006 was conducted by the Central Statistical Organization of  Yemen. The HBS data Survey (HBS) contain information on household roster, economic activities, dwelling conditions, health, education, anthropometrics, income, durable goods, and consumption. One of the main objectives of the HBS 2005–2006 is producing aggregates of the statistical indicators at the level of the urban and rural communities of each governorate to serve the purposes of economic and social development planning on the central and local levels. HBS 2005 consists of 13,136 household and 98,941 individual observations. 117   118 Annex Table 2A.2a Informality Profile (All Workers) Syrian Arab Iraq Egypt, Arab Rep. Lebanon Morocco Republic Yemen, Rep. Jordan % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. All workers National level 100.0 66.9 100.0 58.3 100.0 56.2 100.0 81.9 100.0 71.0 100.0 91.4 100.0 44.2 Urban 72.9 63.5 43.0 42.7 50.1 48.3 56.5 72.7 50.4 65.9 28.2 84.7 82.9 47.5 Rural 27.1 76.1 57.0 70.0 49.9 66.7 43.5 93.8 49.6 76.3 71.8 94.0 17.1 28.3 Gender Male 89.8 70.5 76.9 57.9 76.2 62.5 84.0 83.5 81.8 72.8 44.0 82.1 83.0 47.8 Female 10.2 35.3 23.1 59.3 23.9 36.1 16.0 73.2 18.2 63.0 56.0 98.7 17.0 26.7 Marital status Single 30.8 73.7 29.4 76.4 41.4 62.3 40.3 86.4 64.4 95.4 64.4 95.4 35.7 43.3 Married 69.2 63.85 70.7 50.7 58.6 53.4 59.7 78.8 57.9 63.3 35.6 89.2 64.4 44.7 Age group 15–24 22.3 84.7 19.5 87.1 16.4 69.1 17.1 90.9 29.2 89.0 38.0 97.3 18.5 49.9 25–34 35.6 61.9 31.5 61.4 30.8 55.8 28.7 85.2 26.5 69.1 25.5 85.8 35.8 36.9 35–54 36.5 62.1 40.6 43.4 43.3 52.3 43.1 76.8 37.5 58.0 30.2 87.8 40.7 45.6 55–64 5.64 58.8 8.4 51.2 9.5 53.3 11.1 79.2 6.9 73.2 6.2 95.2 5.0 63.9 Education Primary or below 45.9 82.2 42.4 77.8 38.8 75.7 64.2 90.9 64.6 86.2 80.2 97.4 10.8 71.9 Preparatory/ secondary general 24.1 67.6 6.4 63.7 36.7 56.6 27.8 74.1 19.6 60.6 13.7 83.9 46.7 48.1 Secondary vocational 17.2 35.2 34.4 50.0 7.4 45.1 — — 8.4 19.1 0.7 56.1 16.7 34.2 table continues next page Annex Table 2A.2a Informality Profile (All Workers) (continued) Syrian Arab Iraq Egypt, Arab Rep. Lebanon Morocco Republic Yemen, Rep. Jordan % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. Tertiary education 12.8 25.4 16.8 23.7 17.2 28.4 8.0 36.5 7.4 24.3 5.4 40.7 25.8 22.0 Employment status Wage worker 68.9 51.7 64.3 42.6 65.4 43.1 51.4 67.1 55.6 52.2 23.9 63.7 81.7 32.4 Employer — — 13.2 78.4 4.8 84.0 3.9 98.4 6.7 89.4 — — 7.2 94.6 Self-employed 20.5 100.0 9.9 82.1 27.2 80.1 29.6 98.1 24.0 93.4 20.3 100.0 9.9 98.5 Unpaid worker 10.6 100.0 12.6 98.3 2.6 86.0 15.2 99.0 13.8 99.2 55.9 100.0 1.1 100.0 Industrya Primary sector 4.6 52.2 25.5 94.1 6.6 94.4 28.4 94.1 31.6 92.4 18.4 95.6 4.4 82.3 Secondary sector 30.7 78.7 21.3 65.8 24.5 75.3 12.6 78.9 6.1 75.5 3.4 86.0 20.6 61.8 Tertiary sector 27.5 71.8 28.0 62.7 68.9 46.0 54.5 82.2 37.7 87.2 40.0 93.3 31.9 71.3 Public administration and social services 37.2 12.2 25.3 11.0 — — 4.5 18.5 24.6 17.6 38.2 11.7 43.0 6.17 Ownership (only for wage workers) Public enterprises 36.9 11.8 30.0 5.3 13.5 9.5 19.2 24.4 26.8 10.8 38.6 7.8 35.6 1.1 Private enterprises 63.1 99.1 70.0 81.0 86.5 63.6 80.8 77.3 73.2 93.1 61.4 97.9 64.4 68.0 Note: Inf. = informality; Pop. = population; — = not available. a. Industry = primary sector (agriculture); secondary sector (manufacturing and construction); tertiary sector (wholesale, transport, services); public administration; and social services (including education and health). 119   120 Annex Table 2A.2b Informality Profile (Private Sector Workers) Syrian Arab Iraq Egypt, Arab Rep. Lebanon Morocco Republic Yemen, Rep. Jordan % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. Private sector National level 100.0 99.1 100.0 81.0 100.0 63.5 100.0 88.7 100.0 93.1 100.0 99.7 100.0 68.0 Urban 68.7 99.0 37.4 67.3 58.1 53.8 53.5 83.0 47.7 90.4 26.0 99.0 89.7 67.7 Rural 31.3 99.4 62.6 89.2 41.9 76.9 46.5 95.2 52.3 95.6 74.0 99.9 10.4 70.8 Gender Men 94.6 99.2 79.7 78.2 76.4 70.7 84.5 89.8 84.0 92.7 39.4 99.2 85.1 71.9 Women 5.4 98.9 20.3 91.9 23.6 40.1 15.5 82.5 16.0 95.1 60.7 100.0 14.9 45.7 Marital status Single 34.1 99.3 35.2 88.2 43.0 67.6 42.0 90.2 37.1 99.8 37.1 99.8 37.1 63.6 Married 65.9 99.1 64.8 77.1 57.0 62.2 58.0 87.6 53.9 91.1 62.9 99.6 62.9 70.6 Age group 15–24 28.3 99.5 25.7 92.4 17.8 72.6 18.4 92.2 33.9 96.9 40.5 99.9 20.3 69.6 25–34 32.7 98.9 33.6 78.9 31.2 61.9 29.8 89.3 26.5 92.1 23.9 99.6 32.3 62.8 35–54 33.9 99.2 33.0 74.9 42.2 60.6 41.1 87.1 32.3 90.1 29.1 99.4 41.0 69.8 55–64 5.1 98.5 7.8 78.0 8.9 64.7 10.7 87.4 7.4 91.9 6.5 99.9 6.4 77.5 Education Primary or below 60.0 99.4 52.9 88.3 41.8 80.4 67.7 93.5 77.3 95.5 84.9 99.9 14.1 88.2 Preparatory/ secondary general 25.8 99.1 7.2 80.1 36.0 66.0 26.4 84.8 17.0 89.8 12.4 99.3 48.0 74.3 Secondary vocational 9.2 98.4 30.4 77.8 7.3 51.5 — — 2.3 82.8 0.4 97.1 16.3 55.6 table continues next page Annex Table 2A.2b Informality Profile (Private Sector Workers) (continued) Syrian Arab Iraq Egypt, Arab Rep. Lebanon Morocco Republic Yemen, Rep. Jordan % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. % Pop. % Inf. Tertiary education 4.9 97.5 9.5 50.5 14.9 34.3 5.9 50.7 3.5 61.9 2.3 94.9 21.5 40.2 Employment status Wage worker 50.6 98.3 49.0 75.3 60.2 51.8 46.0 77.3 39.6 90.6 16.3 97.9 71.7 56.5 Employer — — 18.8 78.4 5.6 84.1 4.3 98.4 9.2 89.6 — — 11.2 94.6 Self-employed 32.6 100.0 14.1 82.1 31.3 80.2 32.8 98.1 32.6 93.5 22.3 100.0 15.4 98.5 Unpaid worker 16.8 100.0 18.1 98.3 3.0 85.7 16.9 99.0 18.7 99.3 61.4 100.0 1.7 100.0 Industrya Primary sector 4.3 98.6 35.1 97.3 7.7 94.4 30.8 95.2 41.1 96.7 29.6 98.0 6.4 92.6 Secondary sector 50.8 98.7 25.7 76.5 27.6 76.7 13.3 81.6 7.4 83.9 5.1 94.1 32.7 63.1 Tertiary sector 41.2 98.6 35.3 70.4 64.7 54.5 55.3 87.4 47.5 93.9 61.6 98.7 49.0 75.1 Public administration and social services 3.7 88.9 4.0 61.5 — — 0.6 29.3 4.0 63.8 3.7 87.9 12.0 30.7 Note: Inf. = informality; Pop. = population; — = not available. a. Industry = primary sector (agriculture); secondary sector (manufacturing and construction); tertiary sector (wholesale, transport, services); public administration; and social services (including education and health). 121   122 Profile and Micro-Determinants of Informality Notes 1. Because of data restrictions, the analysis in the remainder of this chapter as well as in chapters 3 and 4 is restricted to non-GCC MENA countries. Tunisia is excluded from the analysis because of a different and not comparable definition of informality. An overview of informality trends in Tunisia is given ­ in box 2.1. 2. Results including agricultural employment are very similar to those presented in the chapter and are available upon request. 3. The cognitive test used “Raven’s Progressive Matrices,” a nonverbal test in which individuals have to identify the missing piece of a particular pattern among multiple choices. Respondents are given five minutes to answer as many of the 12 matrices that are included in the test as possible. The matrices become progressively more difficult and require greater cognitive capacity. This test is independent of language, reading, or writing skills and focuses on measuring observation skills, analytical ability, and intellectual capacity. A score (out of 12) for each respondent is calculated based on the number of correct answers completed in the allocated time frame. See chapter 4 for a detailed description. References Angel-Urdinola, D., S. Brodmann, and A. Hilger. 2011. “Labor Markets in Tunisia: Recent Trends.” Mimeo. Washington, DC: World Bank. Angel-Urdinola, D., and A. Semlali. 2010. “Labor Markets and School-to-Work Transition in Egypt: Diagnostics, Constraints, and Policy Framework.” MPRA Paper 27674, University Library of Munich, Germany. Angel-Urdinola, D., and K. Tanabe. 2011. “Micro-Determinants of Informal Employment in the Middle East and North Africa Region.” Mimeo. Washington, DC: World Bank. Assaad, R. 2009. “Labor Supply, Employment and Unemployment in the Egyptian Economy, 1988–2006.” In The Egyptian Labor Market Revisited, ed. R. Assaad, 1–52. Cairo: American University in Cairo Press. Assaad, R., C. Binzel, and M. Gadallah. 2010. “Transitions to Employment and Marriage among Young Men in Egypt.” Middle East Development Journal 2 (1): 39–88. Elbadawi, I., and N. Loayza. 2008. “Informality, Employment and Economic Development in the Arab World.” Journal of Development and Economic Policies 10 (2): 25–75. Klasen, S. 1999. “Does Gender Inequality Reduce Growth and Development? Evidence from Cross-country Regressions.” Policy Research Report Working Paper 7. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Profile and Micro-Determinants of Informality 123 Perry, G., W. Maloney, O. Arias, P. Fajnzylber, A. Mason, and J. Saavedra-Chanduvi. 2007. Informality: Exit and Exclusion. Washington, DC: World Bank. Radwan, S. 2007. “Good Jobs, Bad Jobs, and Economic Performance: A View from the Middle East and North Africa Region.” In Employment and Shared Growth, ed. P. Pace and P. Semeels, 37–52. Washington, DC: World Bank. World Bank. 1999. “Kingdom of Morocco: Private Sector Assessment Update.” Report 19975-MOR. Washington, DC: World Bank. ———. 2003. “Unlocking the Employment Potential in the Middle East and North Africa toward a New Social Contract.” MENA Development Report. Washington, DC: World Bank. ———. 2009. “Morocco: Skills Development and Social Protection within an Integrated Strategy for Employment Creation.” Mimeo. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Cha p t e r 3 Informality and the Firm SUMMARY: Analyzing firms’ incentives to be informal is an important ­ complement to understanding informal employment. This chapter ­ characterizes informality among firms and their workers using data from (1) Investment Climate Assessment (ICA) surveys on registered firms, (2) ICA survey data on micro- and informal firms, including matched employer-employee ­ information, and (3) firm-level data on formal and informal small and medium enterprises. Important regularities emerge. First, informality is preva- lent among firms in the Middle East and North Africa (MENA). Many firms never formalize, and even those that eventually register still operate informally for a significant amount of time. Micro- and small enterprises in the region, which account for the lion’s share of enterprises and private sector jobs, are mostly unregistered and employ workers informally. Among currently formal firms, MENA has the world’s highest share of firms that start out as informal (one-fourth) and the longest operating period before formalization (four years). Registered firms in the region do not report up to one-fifth of their sales and workers to the tax and social security agencies, respectively. Second, business regulations (taxes, entry regulation, and tax and labor regulation) are ­ an important determinant of informality among firms. High taxation bur- den is the most significant constraint to formalization. Entry regulations seem to be strict, with longer and more ­ cumbersome procedures than in comparator ­ ­ countries. Tax and labor regulations are relatively rigid and are accompanied by strong enforcement but also corruption. Third, among regis- tered firms, worker underreporting for social security purposes is ­ associated with some characteristics of firms, such as size, manager’s education, and labor productivity. Fourth, among informal small and micro-firms in MENA coun- tries, the quality and determinants of employment are diverse, offering satisfac- tory employment for entrepreneurs (often previously unemployed) but lower Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 125   126 Informality and the Firm quality jobs for informal salaried workers. The correlates of informal employ- ment and labor market outcomes in micro-firms suggest that informal employ- ment might in itself be a way to cope with vulnerability and the need for flexible work arrangements. Introduction This chapter characterizes informality among firms and their workers. Analyzing informality among firms is an important step toward under- standing the institutional determinants and policy levers that might affect formalization in MENA, including implications for job quality and exclusion of workers from formal risk-sharing mechanisms. The evidence shows that micro- or small firms that do not register officially with the relevant government agency also typically do not register their workers with the social security agency. In addition, even firms that register often underreport a part of their workforce for social security purposes, which indicates that there is a continuum of informality. Different data sources are available to characterize informality from the firm’s perspective in MENA, including (1) Investment Climate Assessment (ICA) survey data, which are based on a sample of registered firms; (2) ICA survey data on samples of micro- and informal firms in Egypt and Morocco; and (3) firm-level data collected by the Economic Research Forum on formal and informal small and medium enterprises in the Arab Republic of Egypt, Lebanon, and Morocco (MEAS). Data from these various sources provide useful information on the extent of incomplete reporting for tax purposes and social security contributions, as well as on perceived constraints of complying with formalization requirements. This chapter also focuses on the determinants of informal salaried work, gaining insights from matched firm and worker data. The first section of this chapter discusses the definition and extent of informality among firms. The next section describes the costs and bene- fits of informality from a firm’s perspective and discusses the main styl- ized facts on informality among firms. Finally, the last two sections discuss informality among firms in more detail, outlining profiles of informality for both registered firms and micro–informal firms. Description of Informality among Firms Informality among firms is best understood as a continuum along different margins. As discussed earlier in the report, informality can be defined as an Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 127 activity that is unregulated by the formal institutions that govern e ­ conomic activities, such as registration, labor laws, and taxation (ILO 1993, 2002; De Soto 1989). This definition covers two main dimensions: (1) the firm’s perspective, which focuses on the legal documentation of a firm and the extent to which its activities are reported to public agencies (for example, the social security or tax authority), and (2) the workers’ perspective, which focuses on employment conditions such as contractual ties and provision of social security and health insurance coverage. Typically, a firm that is not registered with the relevant government agency is considered to operate informally. Among small or micro-firms, the concept of a fully informal firm is broader, including firms that fail either to be registered or licensed or to keep financial accounts. This usually includes small-scale production units with no legal separation from their owners, such as family-based businesses in which one or more family members participate, and micro-enterprises with at most five ­ employees.1 Nevertheless, infor- mality is not unique to fully informal firms. In fact, a significant share of the activities of registered firms is often unreported, including sales for tax purposes and the wage bill for social security contributions. Among micro- and small firms, there is an overlap between the lack of firm registration or tax identifier and a firm’s affiliation with a social security agency. Using a new set of data for micro- and small enterprises in Morocco (Oviedo 2008), Silva and others (2010) found that a firm’s lack of registration or tax identifier and its affiliation with social security were not mutually exclusive. Although only 40 percent of firms were informal with respect to all three margins simultaneously, 90 ­ percent were informal in at least one of them (figure 3.1). The most frequent margin of informality was the lack of affiliation with a social security agency (89 percent of firms), followed by a lack of registration ­ (55 ­percent of firms). Using responses from the workers of each firm, Silva and others (2010) found that 55 percent of informal workers were employed by ­formal firms. Using the Micro- and Small Enterprises Survey (MEAS) of the Economic Research Forum (ERF) for Lebanon, a similar picture emerges: 41 percent of firms were unregistered in all three regulatory agencies simultaneously, and 78 percent of firms lacked affiliation with a social security agency, again the most common margin of informality. Informality among firms is prevalent by all definitions in MENA. The region has the highest share of firms that start out as informal and has the longest operating period without formalizing in the developing world: On average, a quarter of firms with more than 20 workers start out as informal and operate for about four years without ­ registration.2 This is Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 128 Informality and the Firm Figure 3.1 Margins of Informality among Micro- and Small Firms in Morocco and Lebanon a. Morocco N = 219 rm not a liated with social security 75 (89%) 34% 30 1 14% 88 0% 40% 0 2 0% 1% 1 0% rm has no tax number rm not registered (41%) (55%) 22 fully formal rms (10% of total) Source: Calculations using informal ICA survey (2008) for Morocco. b. Lebanon N = 1,341 rm not a liated with social insurance scheme 325 (78%) 24% 86 89 6% 552 7% 41% 23 20 2% 12 1% 1% rm has no tax card rm not registered (50%) (50%) 234 fully formal rms (17% of total) Source: Calculations using ERF MEAS data (2004). followed by the East Asia and Pacific region (16 percent of firms and 0.9 year on average) and the Latin America and Caribbean region (7 per- cent of firms and six months on average) ­(figure 3.2). In MENA, micro- and small enterprises represent the largest share of all firms and of total nonagricultural private jobs (Galal 2005). For ­ ­ example, according to Elbadawi and Loayza (2008), in Egypt, micro- and small enterprises Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 129 Figure 3.2  Unregistered Firms, by Region 30 5 25 24.4 4.06 4 number of years 20 16.13 3 percent 15 2 10 7.46 0.9 0.95 5 1 0.54 3.45 0 0 Middle East and East Asia and Latin America Eastern Europe North Africa Pacific and the Caribbean and Central Asia % of rms that started operations without formal registration number of years rms operated without formal registration Source: World Development Indicators 2011. Note: These data are collected from currently formal firms with more than 20 workers. account for 97 percent of the enterprises, of which 81 percent are informal. Similarly, in Morocco and Lebanon, more than 50 percent of ­ small and micro-firms are unregistered, employ workers informally (for example, the probability of being a formal worker if employed by a firm with fewer than five workers is about 20 percent in Lebanon and Egypt and below 5 ­ percent in Morocco; table 3.1), and have lower labor pro- ductivity (evidence from Morocco is described in box 3.1). Among regis- tered firms, an estimated one-fifth of both sales and workers are not reported for tax and social security purposes. This share is lower than in Sub-Saharan Africa, comparable to that in Latin America, and higher than in South Asia and Europe and Central Asia (figure 3.3). In MENA firm informality is significantly associated with informal employment. Among small and micro-firms, the share of workers who do not have social security coverage is higher in informal than in formal firms: Although the share of informal employment relative to total micro- and small enterprise employment ranges from 47 percent for Egypt to between 66 and 70 percent in Morocco and Lebanon, most of these workers are hired by informal firms (figure 3.4). Similarly, in Egypt, when formal and informal firms were asked “What percentage of total employment would you estimate the typical establishment in your sector Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 130 Table 3.1  Firm Size and Worker Informality Lebanon (2011) Morocco (2010) Egypt, Arab Rep. (2006) Jordan (2010)   % of all Probability of % of all Probability of % of all Probability of % of all Probability of   informal being formal informal being formal informal being formal informal being formal <5 workers 42 16 56 4 73 20 60 9 5–9 workers 16 48 15 16 14 13 17 32 10–19 workers 15 65 8 32 16 65 20–49 workers 13 75 20 49 50–99 workers 5 79 3 84 100–149 workers 3 85 5 91 9 77 5 90 150 or more workers 6 82 Total 100 60 100 17 100 43 100 50 Source: Calculations using Morocco Household and Youth Survey (2010), Lebanon Employer-Employee Survey (2011), Egypt Labor Market Panel Surveys (1998–2006), and Jordan Labor Market Survey (2010). Note: See annex table 2A.1 in chapter 2 for a detailed description. Informality and the Firm 131 Box 3.1  Job Creation and Productivity: Small versus Large Firms in Morocco Data from the Longitudinal Census of Manufacturing Firms for Morocco (1995–2006) show the following: • Labor productivity, measured as sales per worker, of the smallest firms (with up to 10 workers) is about 2.6 times lower than labor productivity of the largest firms (with 100+ workers). Productivity and rm size, 2006 140 120 sales per worker, US$ 100 80 60 40 20 0 0–9 10–49 50–99 100+ workers Source: Calculations using Morocco industrial longitudinal census. Firm exit and birth annual rate by rm size, 1995–2006 (average) 100+ 50–99 workers 10–49 0–9 –10 –8 –5 –3 0 3 5 8 10 13 annual average annual average rm exitrate (%) rm entry rate (%) Source: Calculations using Morocco industrial longitudinal census. box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 132 Informality and the Firm Box 3.1  Job Creation and Productivity: Small versus Large Firms in Morocco (continued) • Smaller firms have higher rates of entry and exit than larger firms. • Small firms in Morocco appear to have employment growth rates lower than larger firms and have a smaller share of total job creation. Annual employment growth by rm size, 1995–2006 (average) 100+ 50–99 workers 10–49 0–9 –10 –5 0 5 10 percent Source: Calculations using Morocco industrial longitudinal census. Annual share of new jobs created by existing rms by rms size, 1995–2006 (average) 75% 10% 12% 2% 0–9 workers 50–99 workers 10–49 workers 100+ workers Source: Calculations using Morocco industrial longitudinal census. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 133 Figure 3.3 Informality among Registered Firms, by Region 45 40 35 30 percent 25 20 15 10 5 n/a 0 Sub-Saharan East Asia and Middle East Latin America South Asia Europe and Africa Paci c and and the Central Asia North Africa Caribbean share of workers underreported to social security share of sales underreported for tax purposes Source: Calculations using individual country ICA surveys for MENA and South Asia; 2005–10 consolidated data for Sub-Saharan Africa, Latin America and the Caribbean, Europe and Central Asia; and 2002–05 standardized data for East Asia and Pacific. Note: This figure was computed using firms’ responses in the enterprise surveys to the questions: “What percentage of total sales would you estimate the typical establishment in your sector reports for tax purposes?” and “What percentage of the work force would you estimate the typical establishment in your sector reports to social security?” The MENA economies included are Syria (2009), the West Bank and Gaza economy (2006), Egypt (2008), Jordan (2006), Lebanon (2009), and the Republic of Yemen (2010). n/a = not available. Figure 3.4 Informal Workers in Formal and Informal Micro- and Small Firms 80 70 60 50 percent 40 30 20 10 0 Egypt, Arab Rep. Lebanon Morocco hired by informal rms hired by formal rms total Source: Calculations using data from the MEAS survey of the ERF on 4,958 firms in Egypt (2003), 2,948 firms in Lebanon (2004), and 5,210 firms in Morocco (2002). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 134 Informality and the Firm reports for social security purposes?” the estimated share of underreport- ing by informal firms was higher than that reported by formal firms (annex table 3A.1). Even among micro-firms, the share of informal work- ers is significantly higher among informal firms (figure 3.5). Most informal workers work in firms with fewer than 10 workers. Data from household surveys for Morocco (2010), Lebanon (2011), Egypt (2006), and Jordan (2010) indicate that up to 71 percent, 87 ­ percent, workers work in 58 percent, and 77 percent, respectively, of all informal ­ firms with fewer than 10 workers (figure 3.6). Those employed in a firm Figure 3.5  Average Share of Informal Salaried Workers in Formal and Informal ­Micro-Firms in Morocco by Margins of Informality 100 95 percent 90 85 80 75 n n r r m ion r be be be tio tio r nu rat um um um be ra ra st st ax ist xn xn xn gi gi g re re no re ta ta ta no no d no an t n tio d an ra st gi re Source: Calculations using ICA survey sample of micro- and informal firms in Morocco, 2007. Note: Shares of informal salaried workers in each firm’s workforce were calculated. Bars represent the average of these shares in each firm category. Figure 3.6  Where Do the Informal Workers Work? 100 80 60 percent 40 20 0 Morocco Egypt, Arab Lebanon Jordan (2010) Rep. (2006) (2011) (2010) <5 workers 5−9 workers 10 or more workers Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 135 with fewer than five workers in Lebanon, Morocco, Egypt, and Jordan have a 16 percent, 4 percent, 20 percent, and 9 ­ percent probability, respectively, of having social security coverage through their job, five times smaller than the probability of having this benefit if employed in a large firm (table 3.1). Main Costs and Benefits of Informality for Firms Three main approaches in the literature explain the full or partial under- reporting of the workforce (and sales) to authorities. The first line of thought follows De Soto (1989), who characterizes informality as a forced choice of firms to hide from a predatory state. In this case, bureau- cratic burdens and rent seeking from public officials make underreporting the most rational choice for firms, especially those that are too small or unproductive to afford compliance.3 Research by Johnson and others (2000) and Djankov and others (2006) supports this idea. Levenson and Maloney (1998) propose an alternative interpretation in the case of Latin America; they represent informality as the normal state of the nascent entrepreneurial sector. Their model presents informality as a continuum, where fully informal and fully formal behaviors are just two corner out- comes. Firms are modeled as choosing how much to comply with various institutions based on their desired degree of institutionalization, and on the costs and benefits associated with that degree. For instance, firms may register or become licensed but still not find it advantageous to report their entire workforce or all of their sales for taxation purposes. Finally, Friedman and others (2000) depart from the classical explanation of cost- benefit analysis at the firm level and focus on the quality of the institu- tional environment as a determinant of overall noncompliance.4 Numerous barriers might preclude firms from formalizing. In the ­literature5 the costs and benefits of informality are categorized into three main groups: formalization costs (monetary, time, and information); com- pliance costs, which depend on the nature and degree of enforcement of the regulatory framework (the impact of stringency of regulation on firm costs, limits imposed on adjustment margins of employment and wages6 and legal consequences of not registering);7 and opportunity costs of operating informally8 (table 3.2). As in the case of workers, informality can be the result of a profit-maximizing choice that weighs the costs and benefits of formality. People or firms locate in the informal sector because of their belief that informality’s benefits outweigh its costs (for example, Ishengoma and Kappel 2006; Loayza and others 2004). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 136 Informality and the Firm Table 3.2  Benefits and Costs of Formality Benefits Costs Avoidance of government penalties and expansion •  Initial registration without fear of government intervention, which is Monetary costs •  particularly important if enforced Administrative costs and •  Ability to issue formal receipts (needed to expand •  opportunity cost of time and customer base to large firms and multinationals) effort Ability to create legally enforceable agreements with •  Ongoing compliance suppliers and customers; more negotiating power, Taxes and labor and other •  resulting in lower input prices contributions (such as Access to exporting and business with multinationals •  environmental or health taxes) Access to new and lower cost sources of financing (and •  Administrative costs and •  government programs) opportunity costs of time and Spillovers to personal life of entrepreneur (for example, •  effort access to personal loans often requires proof of income) Source: Based on Bruhn 2011. According to a survey of micro- and small firms, high taxation ­ urden is the most significant constraint to formalization. More than b 50 percent of micro- and small firms in Morocco and 40 percent in Egypt indicate that the level of taxes is the major reason for not regis- tering (figure 3.7). Similar results were found in Mexico, Brazil, and Bolivia (World Bank 2007). A high share of firms also report feeling constrained by both the minimum capital requirement and the level of administrative charges (about 30 percent for Morocco and 35 percent for Egypt). Although the share of Moroccan firms identifying taxes as an obstacle is significantly higher than those attributed to the other constraints, this difference is much less pronounced in Egypt, where corporate taxes were significantly reduced in 2005. Moreover, firms in Egypt did not identify the cost of registration and time necessary to register as major constraints to registration. Objective measures of tax rates for Morocco corroborate firms’ per- ceptions. Morocco’s corporate tax rate is one of the highest among devel- oping countries: In 2007, it was second only to Pakistan and remained significantly above the average for developing countries in 2008 (figure 3.8). Morocco’s profit taxes are also high relative to countries ­ with similar income levels (World Bank 2008).9 In contrast, Egypt’s cor- porate tax rate appears to be below average. As noted, Egypt reformed its corporate tax in 2005, reducing it from 42 percent then to 20 percent in 2007. Egypt also eliminated exemptions and tax holidays, clarified tax rules, replaced universal audits with self-assessment with sample audits, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 137 Figure 3.7 Obstacles to Formalization strict labor law time necessary lack of information cost of registration level of other charges level of adminstrative charges minimum capital level of taxes 0 10 20 30 40 50 60 % rms identifying constraint as major or very severe Egypt, Arab Rep. Morocco Source: Calculations using ICA surveys on samples of micro- and informal firms in Morocco 2007 and Egypt, Arab Rep., 2008. Figure 3.8 Corporate Tax Rates in Developing Economies 40 Corporate tax rate 2008 30 average developing countries 26% percent 20 10 0 pu an In Tun nd Cz n, C ysia Hu Ch ic m s Chazil h Jo ina Ar oc an Br ia oc In ina So nes a A u Ta M a, R ico pu nd Ro nga le M a ry Ar r c y Bu riti a ia ut P ia lg us ai 08 Ko M frica a la . Re la . Ph genco a Co ppi na Th 20 a iw a ep ak PoRep lo ne t, Tu bli ab ke i au ni h er bl co di b do is ar i Re rd or st la re e x ili ti h m M aki P or ec yp ov M Eg Sl Source: World Bank 2008. Note: a. Data from 2007. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 138 Informality and the Firm and s ­implified forms. As a result, compliance became less onerous, and corporate tax revenues rose from LE 22 billion in fiscal year 2004 to LE ­ 39 billion in fiscal year 2005.10 A pioneering World Bank study in Brazil is now testing the hypothesis that firm owners in countries that reform tax laws do not register because they mistakenly believe that tax rates are higher than they actually are (see chapter 5, part 1). Firms’ characteristics are significantly associated with costs and bene- fits of formalization. According to the World Bank (2007), small firms may face a lower risk of being caught by inspectors and may find it more difficult to amortize the fixed costs of registration. In addition, recently created firms may not know how profitable their business will be; they may wait until there is enough evidence that they will stay in business, especially if the monetary and administrative costs of registration are high. As emphasized by McKenzie and Sakho (2010), owner characteris- tics and family wealth may influence the firm’s ability to access credit to cover the minimum capital needed to register. These angles will be ana- lyzed in detail for MENA using the ICA surveys, the largest comparable datasets available for MENA countries, providing information on the relationship between informality and firms’ characteristics. The regulatory context matters. International evidence shows that external obstacles that limit firms’ and workers’ access to the formal market are key constraints to formality. Djankov and others (2002) suggest that the main policy action to enhance formalization should be ­ aimed toward lowering regulatory barriers. They find evidence that countries with more complex registration processes (that is, those that require more processes and/or days) have larger informal sectors. Within the profit-maximization framework, firms choose informality mainly because of lower costs and higher flexibility derived from avoid- ing tax, labor, and other types of regulations. Although some of these costs are fixed (for example, registration and licensing), the cost of compliance with labor market regulations and taxation obligations ­ varies with the number of workers and revenues/sales. Noncompliance exposes firms to penalties and sanctions if detected by the ­ corresponding regulatory entities. Firms must also outweigh other disadvantages of informality such as reduced access to financing and to legal e ­ nforcement options, and reduced business interaction with the government and other markets. Labor regulations (particularly firing regulations) in some MENA countries are relatively rigid and labor taxes high compared with coun- tries with similar levels of income. Hence, firms often identify labor Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 139 laws as a major obstacle for business growth. Firing regulations in MENA are quite strict, and firing costs are high (figure 3.9). The cost of advance notice requirements and severance payments and penalties due when t ­ erminating redundant workers, expressed in weeks of salary, is rather low, accounting for 11 weeks of salary on average in MENA versus 15 weeks in Latin America and the Caribbean, but only 7 weeks in OECD countries. Labor taxes and mandatory contributions paid by businesses as a percentage of commercial profit are 25 percent in North Africa versus an average of 10 percent and 13 percent in a typical coun- try in Latin America and East Asia, respectively. (See discussion in chapter 5, part 1, for more details.) Stringency of labor regulations is widely perceived as a major constraint by firms (figure 3.10), but this varies significantly across countries: In Syria and Lebanon, 50 percent and 36 percent of firms consider labor regulations as a major or severe obstacle for business, respectively, and in the West Bank and Gaza economy and the Republic of Yemen, it is 12 percent and 11 percent, respectively. In the following sections, it will be argued that these per- ceptions are strongly associated with informality. Evidence for Mexico shows that reforms that ease firms’ registration are associated with an increased number of registered firms, but this is due more to creation Figure 3.9  Difficulty of Redundancy Index MENA (Non-GCC) 43.3 SAR 41.3 AFR 39.8 ECA 25.9 LAC 24.1 EAP 19.6 GCC countries 4.0 0 20 40 60 di culty of redundancy index (0–100) Source: Doing Business indicators 2011. Note: The redundancy index ranges from 0 to 100. AFR = Sub-Saharan Africa; EAP = East Asia and Pacific; ECA = Europe and Central Asia; GCC = Gulf Cooperation Council; LAC = Latin America and the Caribbean; SAR = South Asia; LAC = Latin America and the Caribbean. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 140 Informality and the Firm Figure 3.10 Percent of Firms Indicating Labor Regulations as a Major Obstacle to Business Development 100 80 60 percent 40 50% 36% 20 26% 15% 12% 11% 0 Syrian Arab Lebanon Egypt, Arab Jordan West Bank Yemen, Rep. Republic Rep. and Gaza economy major or severe obstacle moderate obstacle no or minor obstacle Source: Calculations using individual country ICA surveys. of new businesses by former wage earners than to increased registration of existing informal businesses (Bruhn 2011). Entry regulations in MENA countries seem strict (figure 3.11). In MENA the average number of procedures required to start a business is second only to Latin America and the Caribbean (LAC). Within the region, Algeria, Djibouti, and the West Bank and Gaza economy stand out as economies where this process is particularly cumbersome, requiring more than 10 procedures and an average of between 20 to 80 days to start a business. Besides the fact that the average time is long, the playing field might be unequal between different types of firms, with privileged firms likely benefiting from faster registration processes. Tax regulations are accompanied by both strong enforcement and cor- ruption (figure 3.12). Firms in MENA report an average of four tax inspections per year, the highest regional average in the world. Strict enforcement also appears to be accompanied by widespread corruption, as, according to firm-level surveys, informal payments were requested in 17 percent of inspections, which is significantly above the 7 percent reported in the LAC and ECA regions. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 141 Figure 3.11 Entry Regulation across Regions and in Selected MENA Countries 16 90 14 80 70 12 number of procedures 60 10 50 days 8 40 6 30 4 20 2 10 0 0 y Ga Dj eria on i q sia A R n am an pu . ic Ye oro . en o ba . n iA n a Re ep M ep Le ep C Ar ECA ec ut om bi AF m cc no ud a Ira EN bl LA ria Isl Jord Sa Om za ibo ni ab ic R R ,R ra g M Tu ab Al t, yp Ar Sy n, Eg Ira nd ka an tB es W start-up procedures to register business days to start a business Source: World Development Indicators 2010. Note: The dotted bars show values for regions. AFR = Sub-Saharan Africa; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa. Figure 3.12 Enforcement of Tax Regulations and Bribes during Tax Inspections 4.5 25 percent of inspections where informal 4.0 number of inspections over 3.5 20 payment was requested the last 12 months 3.0 15 2.5 2.0 10 1.5 1.0 5 0.5 0.0 0 MENA AFR ECA LAC SAR EAP tax inspections bribe incidence Source: Latest available ICA survey data for individual MENA countries and standardized 2006–2010 ICA survey data for other regions. Note: AFR = Sub-Saharan Africa; EAP = East Asia and Pacific; ECA = Europe and Central Asia; LAC = Latin America and the Caribbean; MENA = Middle East and North Africa; SAR = South Asia. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 142 Informality and the Firm Stylized Facts on Informality among Firms in MENA Informality among Registered Firms: Firm-Level Correlates Underreporting employment (to social security) and sales (for tax ­ purposes) is common among registered firms but is heterogeneous across MENA countries with relatively similar levels of income. For example, firms in Syria and the Republic of Yemen do not pay social security con- tributions for about 40 percent of their workers, and in Egypt the share of underreported workers is about 12 percent ( ­figure 3.13). In the case of Syria, this information was compared with data collected in a 2011 ­ workers’ survey that revisited the firms involved in the enterprise survey but focused instead on workers. As seen in table 3.3, the estimates of Figure 3.13 Percentage of Underreported Sales and Workers, by Economy Jordan Egypt, Arab Rep. Lebanon West Bank and Gaza economy Yemen, Rep. Syrian Arab Republic 0 20 40 60 percent sales workers Source: Calculations using the World Bank Enterprise Survey database. Table 3.3 Informality among Formal Firms: Matched Employer- Employee Data, Syrian Arab Republic, 2011 Among workers Among sampled firms at sampled firms Share of workers underreported 43% 53% Firm size (employees) 10–49 49 52 50–99 46 30 100+ 32 20 Source: Calculations using firm-data from ICA Syria 2009 and Syria m ­ atched-employer employee survey 2011. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 143 informality based on firms’ responses are a lower bound, but roughly in line with those based on workers’ responses to enrollment in social security. Regardless of the high cross-country variation, some firm characteris- tics are systematically associated with higher degrees of informality among registered firms. First, size matters: Smaller firms tend to under- report more than larger firms. The fraction of workers and sales that firms hide from the government is about twice as large in small firms than in large firms (with the exception of Lebanon; figure 3.14). In the Republic of Yemen and Jordan, for instance, about 50 percent and 13 percent, respectively, of workers go underreported among small firms, compared with about 20 percent and 3 percent, respectively, among firms with 100 workers or more. In Syria, small and medium firms are similar with regard to informality but differ considerably from larger firms where informality is less prevalent. In Egypt, size is also strongly correlated with worker underreporting, and the relationship is almost linear. In the West Bank and Gaza economy, the smallest and largest firms differ considerably with regard to the level of informality. In Jordan, worker underreporting seems to be prevalent mainly among the smallest firms. Sales underreporting exhibits different patterns: (1) although it tends to decrease with firm size in the Republic of Yemen, Syria, and Egypt, this is not the case in Lebanon, the West Bank and Gaza economy, and Jordan; and (2) sales underreporting is significantly higher than workers’ underreporting among larger firms, with the exception of the West Bank and Gaza economy. The differences on underreporting between sales and workers might have important implications for productivity comparisons that use numbers on sales and employment from social security and tax adminis- trative registries. Doubling firm size is associated with a reduction in worker underreporting of more than 5 percentage points in Jordan and the ­ Republic of Yemen and more than 2 percentage points in Syria and Egypt ­ (figure 3.15). The negative relationship between underreporting and employment size in the case of Jordan, the Republic of Yemen, Syria, and Egypt is maintained even after controlling for the firm’s length of time in business, location, and sector. This result is confirmed using a three-wave panel dataset for Egypt, suggesting that as a firm’s employ- ment grows, it is more likely to “hide” a smaller part of its production and workers (see columns 6–8 of annex table 3A.2).11 The relationship between underreporting and size is found to be broadly linear across firms’ sizes; that is, relative to small firms, both large and medium firms Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 144 Figure 3.14  Worker and Sales Underreporting by Firm Size Workers underreported by rm size 70 Sales underreported by rm size 70 60 60 50 50 40 40 percent percent 30 30 20 20 10 10 0 0 ic b . n p. an . n p. an om d om d ic b ep ep bl ra no no on an on an bl ra Re Re rd rd y y ,R ,R pu n A pu n A ba ba ec nk ec nk Jo Jo ab ab en en Le Le Re yria Re ria za Ba za Ba Ar Ar m m Sy Ga est Ga st Ye Ye t, t, S yp yp e W W Eg Eg 10–14 15–49 50–99 100–249 250+ 10–14 15–49 50–99 100–249 250+ Source: Calculations using ICA survey data for MENA economies from the World Bank Enterprise Survey database. Informality and the Firm 145 Figure 3.15  Association between Doubling Employment Size and Worker ­Underreporting Jordan Yemen, Rep. Syrian Arab Republic Egypt, Arab Rep. −7 −6 −5 −4 −3 −2 −1 0 workers underreported change (percentage points) Source: Calculations using ICA survey data for MENA economies from the World Bank Enterprise Survey database. underreport less, but the difference is greater for larger firms (see ­ columns 1–2 of annex table 3A.2). Notably, underreporting of sales and underreporting of workers are highly correlated. In addition to reducing the revenue for taxation, sales underreporting may also have an effect on informal employment; that is, to pay its undeclared labor force, a firm may be forced to keep some of its sales off the books. For this reason, the incentive to sell informally may be correlated with an incentive to underreport workers. A simultaneous estimation of the determinants of sales and worker underreporting shows that similar determinants do affect underreporting of sales and workers (see columns 3–4 of annex table 3A.2). More productive firms appear to report a larger share of their workers to social security. The association between sales per worker (a rough estimate of productivity) and informality appears to be negative and statistically significant, even after controlling for other firm-level corre- lates. This finding is robust to the different measures of productivity associated with both sales and worker underreporting and is also robust to using longitudinal data (annex table 3A.2). Though causality cannot be inferred in this setup, figure 3.16 depicts the change in worker under- reporting that is associated with a doubling of productivity in selected countries in the region. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 146 Informality and the Firm Figure 3.16  Association between Doubling Labor Productivity and Worker Underreporting Syrian Arab Republic Jordan West Bank and Gaza economy −3 −2.5 −2 −1.5 −1 −0.5 0 workers underreported change (percentage points) Source: Calculations using ICA survey data for MENA economies from the World Bank Enterprise Survey database. Figure 3.17  Worker Underreporting by Manager’s Education Level Lebanon Yemen, Rep. Egypt, Arab Rep. 0 20 40 60 percent postsecondary secondary school less than secondary Source: Calculations using ICA survey data for MENA economies from the World Bank Enterprise Survey database. Firms managed by highly educated individuals consistently report a higher share of the workforce to social security, controlling for other firm characteristics, including education of its workforce and the manager’s prior experience. Firms whose top managers’ highest level of ­ education are below secondary school exhibit rates of worker underre- porting of 50 percent in the Republic of Yemen and 20 percent in Egypt, and firms whose managers completed postsecondary education exhibit underreporting rates of 35 percent in the Republic of Yemen and 11 percent in Egypt (figure 3.17).12 In Lebanon and Egypt,13 a hypothetical increase in the education level of the top manager from less than Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 147 Figure 3.18 Changes in Worker Underreporting Associated with a Change in ­Manager’s ­Education Level Lebanon Egypt, Arab Rep. −10 −8 −6 −4 −2 0 workers underreported change (percentage points) from “less than secondary” to “secondary” Source: Calculations using ICA survey data for MENA economies from the World Bank Enterprise Survey database. secondary to secondary is associated with 9 percent and 3 percent decreases in worker underreporting, respectively (figure 3.18).14 Cross- country regressions also suggest a strong association between the man- ager’s education level and underreporting of workers as well as of sales. An analysis of panel data for Egypt (using both Tobit random effects and OLS fixed effects models) confirms this result (annex table 3A.3). Note that fixed effects regressions control for time-invariant firm characteris- tics and eliminate regression bias from omitted firm-specific variables. The results show that, all else equal, the arrival of a manager with a higher education level than the previous manager is associated with lower levels of underreporting. The relationship between a manager’s education and informality is open to several interpretations. De Paula and Scheinkman’s (2007) model shows that a manager’s education is an imperfect measure of the person’s ability, and that more able managers tend to be less likely to work in informal firms (that is, they are found more often in productive or larger firms). However, the results here sug- gest that the relationship between a manager’s e ­ ducation and worker underreporting holds within the same firm (using firm fixed effects models and restricting the analysis to cases where the manager did not change), making this hypothesis less likely. A second hypothesis is that more educated managers may better understand the benefits of paying taxes or of hiring workers ­formally. Alternatively, more educated manag- ers may have a lower ­ tolerance for the risks associated with using infor- mal hiring as a tool to avoid regulations. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 148 Informality and the Firm Firms with a more educated workforce tend to exhibit lower levels of underreporting. In Syria and Lebanon, the level of compliance increases with workforce education (figure 3.19). For example, in Lebanon, in firms whose workforce consists mainly of workers with less than second- ary education, underreporting is 27 percent; in firms where most of the workforce have a secondary or higher than secondary education, under- reporting is 11 percent and 5 percent, respectively. Skill upgrading of the existing workforce could also be associated with increasing levels of formality. For a given firm size, age, location, and sector, a change in the most prevailing level of education of the firm’s workforce from less than secondary to secondary is associated with a decrease in underreporting of almost 10 percentage points in Lebanon. Likewise, a change in the most prevailing level of education from secondary to postsecondary is associated with a decrease in underreporting of 7 percentage points in Lebanon and 8 percentage points in Syria (figure 3.20). Note that a more educated workforce is associated with higher productivity, and therefore, a higher opportunity cost of being informal. Hence the relationship between workforce education and informality might be open to several interpretations. Informal firms are somewhat less connected and less likely to belong to a business network. As indicated by the World Bank (2010), the lack of network links might be an indication of “duality” in private Figure 3.19  Worker Underreporting and the Workforce Education Level Lebanon Syrian Arab Republic Egypt, Arab Rep. 0 10 20 30 40 50 percent postsecondary secondary less than secondary Source: Calculations using ICA survey data for MENA economies from the World Bank Enterprise Survey database. Note: Workforce education level is reported according to the most representative education level within the firm. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 149 Figure 3.20 Change in Worker Underreporting When Workforce Education Increases Syrian Arab Republic Lebanon –15 –10 –5 0 change in the rate of workers underreported from “less than secondary” to “secondary” from “secondary” to “above secondary” Source: Calculations using ICA survey data for MENA economies from the World Bank Enterprise Survey database. Note: Workforce education level is reported according to the most representative education level within the firm. Table 3.4 Informality, Connectivity, and Networks among Firms in MENA ­Economies Worker underreporting E-mail Web site Association Egypt, Arab Rep. −0.0883* −0.0918* Jordan −0.1600* −0.1335* Syrian Arab Republic −0.0941* −0.0984* −0.029 Yemen, Rep. −0.2542* −0.3099* −0.2335* Lebanon −0.1232* −0.0666* −0.0402* West Bank and Gaza economy −0.2221* ­ −0.2265* Source: Calculations using ICA survey data for MENA economies from the World Bank Enterprise Survey database. Note: Correlation coefficients reported. *Statistical significance at the 5% level. sector dynamics. A relevant finding common to all countries is a signifi- cant negative association between the share of worker underreporting and the use of e-mail, existence of a Web site (as a proxy for connectiv- ity), and belonging to a business association (as a proxy for networks) (table 3.4). The following chapter finds similar results for workers. More informal firms are also less likely to provide training to their workers. Across countries and controlling for firm size, age, and sector, firms that are more informal are found to be less likely to provide training to their workers (figure 3.21). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 150 Informality and the Firm Figure 3.21 Relative Provision of Worker Training in MENA Economies and ­Informality 0.8 Slovakia Slovenia Estonia Ecuador2006 0.6 Czech Rep. Croatia Peru2006 Bolivia2006 Lebanon Malawi2005 % firms provide training Lithuania LatviaChile2006 Colombia2006 Belarus ElSalvador2006 Swaziland2006 Poland Paraguay2006 Argentina2006 Serbia Bosnia and Montenegro Albania Kyrgyzstan Honduras2006 Namibia2006 Ukraine Cameroon2006 Panama2006 Hungary CapeVerde2006 Guatemala2006 SyrianBurkinaFaso2006 Arab 0.4 Tanzania2006 Uruguay2006 FYROM Republic Botswana2006 ArmeniaRussia Moldova Bulgaria Mexico2006Yemen, Rep. Uganda2006 Nicaragua2006 Romania Niger2005 Tajikistan Kazakhstan West Bank and GazaRwanda2006 Turkey Mauritania2006 Egypt, Arab Rep. Georgia Guinea2006 Gambia2006 Bangladesh2007 Burundi2006 0.2 Angola2006 Jordan Uzbekistan Azerbaijan India2005 GuineaBissau2006 Pakistan2007 DRC2006 0 0 20 40 60 80 % of workers underreporting Source: Calculations using available enterprise surveys. Firms and the Regulatory Environment Firms with higher rates of underreporting are more likely to identify stringent labor and tax regulations as a major constraint. The literature suggests that more rigid de jure labor regulations can increase firms’ incentives to hide part of their workforce (Botero and others 2004). A recent review of labor regulations in the region shows that MENA has fairly flexible hiring procedures but rigid dismissal rules that, according to employers, curb the employment potential (Angel-Urdinola and Kuddo 2010). Pierre and Scarpetta (2006) analyze an array of countries for which employment protection regulations have been evaluated in terms of stringency, and they show that employers’ perceptions about labor regulations are closely related to the actual stringency of labor laws. In this analysis, the correlation between the perceived stringency of de jure labor regulations and a greater propensity to hire informally is ­ analyzed (table 3.5). Following Pierre and Scarpetta (2006), employers’ perceptions of labor and tax regulations are used as a proxy for regulatory stringency. In Egypt, the Republic of Yemen, and Lebanon, worker underreporting and perceptions of the stringency of labor regulations are ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 151 Table 3.5 Correlation between Worker Underreporting and Perceptions of Labor Regulations and Tax Rates Economy Labor regulation Tax rate Egypt, Arab Rep. 0.1295* 0.2270* Jordan 0.0216 −0.0286 Lebanon 0.0682* 0.0471* Syrian Arab Republic −0.0109 0.0473* West Bank and Gaza economy −0.0504* −0.0388* Yemen, Rep. 0.1646* −0.0203 Source: Calculations using ICA survey data. Note: Correlation coefficients reported. *Statistical significance at the 5% level. positively and significantly associated. In contrast, in the West Bank and Gaza economy, this correlation is smaller and negative. Firms that reported labor regulations to be a severe or very severe obstacle to their business are consistently more likely to underreport their workforce and their sales. This result is found in cross-country regressions and confirmed using panel data (annex table 3A.4). Consistent with this result is the fact that firms that reported taxes as being an important obstacle to their business also tended to hide more sales. The study finds that over time and for the same firm, the more labor obstacles are perceived as constraining, the lower the volume of sales reported. Enforcement matters too. More frequent inspections are associated with less informality. The number of labor inspections that firms receive in a year is used to assess whether enforcement of regulations affect informality. In line with the results of Almeida and Carneiro (2009) for Brazil, more frequent labor inspections are found to be associated with less informal employment. Annex table 3A.4 reports the regression results for a pool of seven countries and for the panel of Egyptian firms. Tax inspections seem to be less effective than labor inspections in curbing sales evasion: Although labor inspections are significantly associated with less sales underreporting, the relationship between tax inspections and sales underreporting is not statistically significant. When corruption is associated with inspections, underreporting is higher. The effectiveness of enforcement depends on the extent to which inspections are associated with corruption. To control for this effect, a binary variable is added for cases where firms reported that labor inspectors expected an informal payment with their visit. As expected, corruption in Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 152 Informality and the Firm labor inspections is significantly associated with greater labor underreport- associated with ing. Similarly, bribing of tax officials is also significantly ­ underreporting of sales, which may explain the previous result that tax inspections, unlike labor inspections, are not associated with lower infor- mality. As was found for bribing, the share of informality among nonre- spondents is higher for labor and taxes. Importantly, the more binding the regulations, the less inspections are associated with lower informality (annex tables 3A.4 and 3A.5). In a survey of small firms, those that indicated they had transitioned from informality to formality reported that this was mainly driven by the fact that it became mandatory rather than beneficial. Data from the Micro- and Small Enterprises Survey (MEAS) of the Economic Research Forum (ERF) indicate that the reason for acquiring a business license or registering with the tax department after having started operations is ­ primarily because it is mandatory rather than because it became advantageous or simpler to do. In Egypt, Lebanon, and Morocco, manda- ­ tory registration was identified by 69 to 86 percent of firms, and either advantageous or simpler registrations were identified by 4 to 21 percent and 3 to 13 percent of firms, respectively ­ (figure 3.22). The only excep- tion for registering to obey the law versus registering because it became advantageous was in Morocco, where 51 percent of the firms expressed perceived advantages to business registration as the main motivation. The requirement to register followed in second place (32 percent of firms). Consistent with this finding, a World Bank study on firm informality in Bolivia (World Bank 2007) shows that the main benefit as perceived by firms of having a tax n ­ umber is compliance with the law (47 percent). Note, however, that not all small firms might be on the margin of formal- izing. In particular, and in line with McKenzie and Sakho (2010), registra- tion might lower profits for smaller firms because they might be too small to immediately benefit from formalization. Registering for taxes immediately involves more costs in the form of tax payments. If the firm is too small to benefit from an increased customer base or better access to credit, then immediate costs might outweigh benefits. Informality among Micro- and Small Informal Firms Some observers argue that the informal sector, and in particular informal small and micro-firms, is a vital employer of young and unskilled workers, a countercyclical employer for workers in transition, and an essential source of affordable goods and services for the poor. In this section, the informal ICA survey data on samples of micro- and informal firms in Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 153 Figure 3.22  Firms’ Reported Reasons for Registering or Acquiring a License or Tax Card Egypt, Arab Rep. registering acquiring business licence acquiring tax card registering Lebanon acquiring business licence acquiring tax card registering Morocco acquiring business licence acquiring tax card 0 20 40 60 80 100 percent became mandatory became advantageous procedures simpli ed other Source: Calculations using MEAS data (Egypt 2003, Lebanon 2004, and Morocco 2002). Egypt and Morocco are used to analyze informality among micro- and small firms and understand the extent to which these hypotheses hold true in MENA. Informality among micro- and small firms: A profile of employment. Among micro- and small firms in Egypt, entrepreneurs of informal firms are less qualified, on average, than entrepreneurs of formal firms. Using data from ICA survey (formal and informal firms) in Egypt, the World Bank (2010) indicates that managers of micro- and small informal firms have a lower level of education than those of formal firms. The share of managers with a primary education or less is 43 percent in informal firms and 7 percent in small formal firms. Similarly, the share of managers of informal firms with postsecondary education is significantly lower than percent, respectively) ­ that of formal firms (18 percent and 68 ­ (figure 3.23). In addition, 90 percent of informal firms indicated that the firm’s manager did not have previous work experience in the formal sector. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 154 Informality and the Firm Figure 3.23  Highest Education of Managers for Formal and Informal ­Manufacturing Firms in Egypt 100 90 80 70 60 percent 50 40 30 20 10 0 informal formal formal formal small medium large did not complete primary/did not enter school completed primary preparatory/incomplete secondary completed secondary (including vocational) some university degree or higher Source: World Bank 2010. Similarly, workers in informal firms seem to be less qualified, have a higher job turnover, and have a greater likelihood of being relatives or members of the household of the firm owner. Among micro- and small firms in Egypt, the workforce of informal firms is less educated than that of formal firms. The largest difference is in the share of workers with no formal education (33 percent in informal firms versus 18 percent in formal firms) and the share that completed secondary education (29 percent in informal firms versus 39 percent in formal firms; figure 3.24). Notably, 27 percent of micro- and small firms identify skills ­ and education of available workers as a major or very severe problem for their operations and growth. Nine percent of workers of these firms are part time, and 13 percent are relatives or members of the household of the firm owner. Among micro-firms in Morocco, only 7 percent of all workers are covered by social security; the patterns and associations observed among formal, registered firms between underreporting of workers and Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 155 Figure 3.24  Highest Education of Male Workers in Formal and Informal Manufacturing Firms in Egypt 100 90 80 70 60 percent 50 40 30 20 10 0 informal formal formal formal small medium large did not complete primary/did not enter school completed primary preparatory/incomplete secondary completed secondary (including vocational) some university degree or higher Source: World Bank 2010. firms’ characteristics also hold true among micro–informal firms. In ­ the Morocco ICA survey on micro-firms, each worker was surveyed. Among other questions, each worker indicated whether he or she w ­ ­ as  contributing to social security. Results are depicted in figure 3.25. The largest part of the workforce of these firms consisted of informal ­ salaried workers (58 percent), followed by informal entrepreneurs (34 percent). As in larger registered firms, informal employment among micro-firms spans beyond informal firms, with as many as 55 percent of all informal workers working in registered firms. This is not surpris- ing considering that informal workers represent up to 84 percent of the labor force of firms with a business registration, and 68 percent of firms with tax registration (recall figure 3.5). Finally, among informal small and micro-firms in MENA countries, the strongest predictors of infor- mality, measured as a firm’s lack of registration, are the owner’s level of education, having started the business following a period of unemploy- ment, absence of business with larger firms, and concerns about tax increases and labor law enforcement. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 156 Informality and the Firm Figure 3.25  Worker Categories in Micro-Firms in Morocco formal entrepreneurs 2% informal salaried workers informal 58% entrepreneurs 34% formal salaried workers 5% Source: Moroccan Enterprise Survey of micro-firms 2007. See description in the annex in chapter 2. Note: Pie chart shows share of total work force in data set. Entrepreneurs are both self-employed and employers. In Morocco, informal workers of micro-firms are a heterogeneous group, with informal salaried workers and entrepreneurs differing consid- erably with regard to age, gender, salary, and hours worked. The median informal salaried worker was 28 years old, male, had seven years of schooling, was not married and had no children, was paid weekly and received DH 6.9 ($0.83) per hour of work, worked 54 hours per week, and was most likely to be employed in manufacturing. The median infor- mal entrepreneur was older (36 years old), male, had eight years of schooling, was married, was head of the household and had one child, received DH 7.9 ($1.00) per hour, worked 63 hours per week, and was most likely to work in the service sector. The median informal employer was 40 years old, male, had nine years of schooling, was married with two children, received DH 15.6 ($1.90) per hour, and worked 57 hours a week (table 3.6). Looking beyond averages, variability in labor earnings among informal workers of micro-firms in Morocco is high, reflecting significant ­ heterogeneity. Only 8 percent of all salaried workers are in the highest hourly wage quintile, and the rest are uniformly distributed among the other quintiles (figure 3.26). The self-employed category is likely to cover a variety of workers; most are in the two lowest and highest quintiles, and Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 157 Table 3.6 Informal Employment in Micro-Firms in Morocco Salaried worker Self-employed Employer Mean Median Mean Median Mean Median Age 29.8 28 39.5 36 42.0 40 Female 23.9% 12.0% 16.9% Education (years) 7.3 7 8.6 8 8.2 9 Married 26.3% 64.0% 76.9% Number of children 0.6 0 1.6 1 2.5 2 Head of household 34.5% 68.0% 72.9% Paid daily 13.5% 50.0% 61.5% Paid weekly 54.6% 5.0% 13.5% Paid every two weeks 1.7% 0.0% 0.0% Paid every month 30.1% 45.0% 25.0% Hourly wage 8.1 6.9 9.1 7.9 19.3 15.6 Weekly hours worked 55.5 54 62.1 63 57.4 57 Manufacturing 40.9% 16.0% 35.6% Construction 12.0% 4.0% 13.6% Trade 14.9% 32.0% 16.1% Services 32.2% 48.0% 34.7% Source: Silva and others 2010. Figure 3.26  Distribution of Workers by Hourly Wage Quintile in Micro-Firms in Morocco 60 percent of workers in each group 50 40 30 20 10 0 informal salaried self-employed informal workers employer quintile 1 2 3 4 5 Source: Silva and others 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 158 Informality and the Firm about 10 percent are in the middle quintile. This may indicate an even ­ elf-employed distribution of successful (higher earning) and less successful s workers. Employers are highly concentrated in the higher earnings quin- tiles; more than 50 percent of all employers earn an hourly wage in the top quintile of the hourly earnings distribution. The distribution of hourly earnings of formal salaried workers has two peaks, with a small group of formal salaried workers earning high wages. The distribution of earnings of informal entrepreneurs (self-employed and employers) is more dispersed and skewed to the right than that of both formal and informal salaried workers (figure 3.27). Informal salaried workers differ considerably from formal salaried workers: They are significantly younger (mostly between 25 and 29 years old), less educated, have lower hourly wages, and are significantly more likely to be paid weekly. Figure 3.28 shows the distribution of workers’ age by job status among informal workers in micro-firms. Salaried workers are highly concentrated between the ages of 25 and 29 years old. Although the distribution of employers is more dispersed than those of salaried workers, they tend to be older (concentrated around the age of 40). Similarly the largest age group of the self-employed is between 30 and 39 years old, with a second peak around age 60+, mainly following early retirement. Beyond differing in age, informal Figure 3.27  Distribution of Hourly Earnings of Workers in Micro-Firms in Morocco 0.8 0.6 kernel density 0.4 0.2 0 0 2 4 6 log hourly wage informal salaried workers formal salaried workers informal entrepreneurs Source: Silva and others 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 159 Figure 3.28  Distribution of Workers’ Age, by Type of Informal Employment 0.05 0.04 kernel desity 0.03 0.02 0.01 0 0 20 40 60 80 age informal salaried workers self-employed employer of micro- rm Source: Silva and others 2010. salaried workers differ considerably from formal salaried workers in other dimensions: A significantly lower share have postsecondary edu- cation and lower hourly wages and are significantly more likely to be paid weekly (table 3.7). The evidence from micro firms in Morocco suggests that labor infor- mality might in itself be a way to cope with vulnerability and the need for flexible work arrangements. Silva and others (2010) show that in micro- firms in Morocco, the characteristics of workers that most strongly corre- late with informal employment are household size and being a married woman. Gender is the sole systematic determinant of hours worked, with women working fewer hours than men. Besides being systematically related to age and education, wages of informal workers rise with the number of children and household size, and tend to be higher for the head of the household. Firms’ characteristics play an important role in wage determination, particularly labor productivity and, to a lesser extent, size. Finally, in micro-firms, a significant wage (or hours worked) premium does not seem to exist between formal and informal salaried workers. Among informal workers, entrepreneurs command higher e ­ arnings. Holding gender, owner’s age, marital status, and education constant, owners of registered firms in Morocco are 8.3 to 20 percent more likely Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 160 Informality and the Firm Table 3.7  Differences between Informal and Formal Salaried Workers in Micro-Firms in Morocco Salaried worker Informal Formal Mean Mean Age 29.8*** 34.5 Female 23.9% 31.8% Married 26.3% 31.8% Number of children 0.6 0.9 Head of household 34.5% 45.5% Education (years) 7.3 8.8 Primary education or less 47.5% 36.4% Secondary education 39.3% 31.8% Postsecondary education 13.2%** 31.8% Hourly wage 8.1*** 14.7 Weekly hours worked 55.5 60.0  Paid daily 13.5% 9.1%  weekly 54.6%*** 4.5%  every two weeks 1.7% 0.0%  every month 30.1%*** 86.4% Source: Silva and others 2010. Note: Asterisks indicate statistical significance of mean differences. *p<.10, **p<.05, ***p<.01. to indicate that they earn a good income from their activity than their peers who own ­ unregistered firms of the same sector, region, age, and size (see figure 3.29 and Silva and others 2010).15 Informality as countercyclical employment for those out of work. Among micro-firms in Morocco, one-fourth of firm owners were unemployed before starting their business. This suggests that this type of entrepreneur- ship might often be a choice of last resort. Micro-firms in Morocco operate with relatively low product margins. Among micro-firms there, the median sales margin on products is about 15 percent. Figure 3.30 shows the average cost share in micro-firms’ expenditures. Human resources and production materials constitute the largest shares with 42.1 percent and 29.4 percent, respectively. The expenditure share on the rent of machines is much smaller (12.2 ­ percent), but distinctively larger than the energy and communication expenditure shares. For labor and production materials, informal firms might have a cost advantage over formal firms, because informal labor and acquisitions without a receipt are usually not subject to taxes, leading to lower pro- duction costs. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 161 Figure 3.29 Likelihood of Earning “Satisfactory Income” in Registered versus Unregistered Micro-Firms in Morocco two-steps estimation propensity score matching probit estimation 0 5 10 15 20 25 percentage points Sources: Silva and others 2010 and calculations using Morocco Enterprise Survey of micro-firms 2007. Figure 3.30 Cost Categories among Micro-Firms in Morocco rent for machines and land, 12.2% water, 1.6% gas, 6.1% communication, human 2.9% resources, 42.1% electricity, 4.7% production materials, 29.4% Source: ICA survey on sample of micro- and informal firms in Morocco 2007. Note: Pie chart shows average cost shares of firms in data set. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 162 Informality and the Firm On the revenue side, Moroccan micro-firms’ sales mainly target local markets and are not subject to taxes. Over 80 percent of their sales are directed to the local market; of the remaining sales, 12.5 percent goes to the regional market, nearly 5 percent goes to the national market, and less than 1 percent goes to foreign countries. Moroccan consumers make up 87 percent of their client base; 9 percent of clients are other Moroccan micro-firms, and almost 2 percent are large Moroccan firms (annex Three-fourths of micro-firms report that they do not pro- table 3A.6.). ­ vide receipts to clients, and only 49 percent report that they accept checks. A significant share of firms reported innovating and improving their products’ quality in the last 12 months in response to increasing competition: 30 percent and 19 percent of firms reported introducing new products and production processes, respectively, and 40 percent indicated they had performed quality upgrades. Notably, as many as 77 percent reported being recently forced to either reduce price or sales to respond to increasing competition (annex table 3A.6). Among micro–informal firms in Morocco, financing is also mainly infor- mal. Financing of micro-firms in Morocco depends heavily on internal financing (such as savings before starting the business) and amounts to an average of 84 percent of all investment (figure 3.31). On average, each firm mobilizes as little as 1.4 percent of its total resources from bank credit; fewer than 13 percent of all firms have ever demanded a bank credit. Figure 3.31 Sources of Investment among Micro-Firms in Morocco credit from micro-credit, supplier, credit from 1.2% 3.7% clients, 0.2% bank loan, 1.3% loan from relatives/ friends, 9.5% internal nancing, 84.0% Source: Moroccan Enterprise Survey of micro-firms 2007. Note: Pie chart illustrates average shares of firms’ sources of investment. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 163 Figure 3.32 Obstacles to Credit Requests among Micro-Firms in Morocco demanded guarantees too high 41.2 interest rate too high 36.8 procedure too complicated 26.4 insu cient o er 7.1 percent Source: Moroccan Enterprise Survey of micro-firms 2007. Note: Limited to firms not requesting credit; bars show the share of firms identifying a constraint. This is not surprising given that 79 percent of owners indicated they do not have a business bank account, and 62 percent do not have separate household and firm expenses. Firms indicated that high guarantees were the main obstacle for requesting credit (41 percent of firms), followed by the high level of interest rate (37 percent) and complicated procedures (26 percent) (figure 3.32). Conclusions Chapter 3 characterizes informality among firms and their workers. The main findings are fourfold. First, informality is prevalent among firms in MENA. Many firms never formalize, and even those that eventually reg- ister still operate informally for a significant amount of time. MENA’s micro- and small firms are mostly not registered and typically do not pay social security contributions for their employees. Larger, registered firms, although operating within the broad regulatory framework, do not report up to one-fifth of their sales and workforce to the tax and social security agencies, respectively. The region has the world’s highest share of firms that start out as informal (one-fourth) and has the longest operating period before formalization (four years) among firms with more than 20 workers. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 164 Informality and the Firm Firms in MENA report that business regulations (taxes, entry regula- tion, and tax and labor regulation) are an important determinant of infor- mality. Informal firms identify compliance costs and, in particular, the fiscal burden of taxes as the most significant constraint to formalization and, vice versa, the fact that registration became mandatory (as opposed to advantageous) as the main reason for formalizing. In the region and, in particular in economies such as Algeria, Djibouti, Iraq, and Tunisia and the West Bank and Gaza economy, a large opportunity exists for simpli- fication of entry regulations. International evidence indicates that this may enhance employment creation through establishment of new formal firms by previously salaried workers. In contrast, a response based on increasing enforcement might have important limitations, because MENA already has the world’s highest average number of inspections reported by registered firms over the last 12 months, and in more than 15 percent of cases these are reported to be accompanied by requests for informal payments. Firms that reported labor regulations to be a severe or very severe obstacle to their business are consistently more likely to underreport their workforce and their sales. Enforcement matters too: More frequent inspection is associated with less informality. However, corruption, also prevalent, reduces the effectiveness of inspections in curbing informality. Among registered firms, worker underreporting is heterogeneous and appears to be a reflection of firm characteristics, in particular size, man- ager’s education, and labor productivity. These variables are all strong correlates of worker underreporting. In particular, all else equal, an increase in firm size is associated with lower underreporting, even among similar firms (that is, with the same sector, location, or age). Specifically, among similar firms, doubling firm size is associated with a reduction in worker underreporting of more than 5 percentage points in Jordan and the Republic of Yemen, and roughly half of those effects in Syria and Egypt. In addition, at the firm level, the manager’s and workers’ educa- tion are strong predictors of a firm’s level of registration, particularly in Lebanon, Egypt, and Syria. Furthermore, low firm productivity and high worker informality are strongly correlated among registered firms in MENA. For example, doubling labor productivity is associated with a 2 to 3 percentage point reduction in firms’ underreporting in economies such as Jordan and Syria and the West Bank and Gaza economy.16 Panel data confirm statistical significance of the effects of size, productivity, and manager’s education. More informal firms are also somewhat less connected (for example, a lower share have e-mail and a Web site) and ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 165 less likely to belong to a business network. Importantly, they are also less likely to offer training to workers than their formal counterparts. Among informal small and micro-firms in MENA countries, the qual- ity and determinants of employment are heterogeneous. Small informal firms appear to offer satisfactory opportunities for entrepreneurs but lower quality jobs to informal salaried workers, who are young, likely to work long hours, and earn a relatively low hourly wage. These workers are employed in jobs with high turnover rates and are often related to the entrepreneur. In Moroccan micro-firms, only 8 percent of all salaried workers are in the highest hourly wage quintile, and the rest are uni- formly distributed among the other wage quintiles. Moreover, the two characteristics of workers that most strongly correlate with informal employment are household size and being a married woman. Gender is the sole systematic determinant of number of hours worked, with women working fewer hours. Wages of informal workers rise with the number of children and household size and tend to be higher for the head of the household. These results suggest that labor informality might in itself be a way to cope with vulnerability and the need for flexible work arrangements. In summary, in MENA, although many firms stay informal, those that eventually register still operate informally for a significant amount of time. Compliance costs and entry regulations seem to be at the root of the problem, as is corruption. Worker underreporting is heterogeneous among registered firms and appears to be correlated with some character- istics of MENA firms, in particular size, manager’s education, and labor productivity. Among informal small and micro-firms, only a minority of all workers are covered by social security. Micro- and small firms provide employment to young and unskilled workers, and opportunities to unem- ployed or informally employed salaried workers who might later decide to become entrepreneurs themselves. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Annex Annex Table 3A.1 Informality Profile among Firms All Egypt, Arab Rep. (2008) Jordan (2006) % workers % sales % workers % sales % workers % sales % of under­ under­ % of under­ under­ % of under­ under­ Variable sample reporting reporting sample reporting reporting sample reporting reporting Workers underreporting 19 12 6 Sales underreporting 20 14 13 Firm size (employees)   0–10 23   11–14 32 24.8 24.4 25 18.8 18.9 26 12.8 13.5   15–49 33 18.6 19.0 33 12.3 13.4 34 5.3 7.9   50–99 12 19.0 20.2 11 13.0 16.3 15 2.6 8.9   100–249 12 15.3 17.2 13 9.4 11.3 14 3.2 16.0   250+ 12 7.3 12.9 18 5.2 8.3 12 0.0 24.6 Firm’s age   0–5 15 19.9 20.1 10 14.7 13.8 31 7.3 15.8   6–10 19 20.5 20.0 17 14.0 14.3 21 10.3 10.9   11+ 66 18.0 19.7 73 11.6 13.8 48 3.1 11.4 Sector   Manufacturing 65 18.5 20.0 69 11.7 13.7 70 4.8 13.7  Service 24 15.9 16.0 28 12.0 12.4 23 7.5 9.0   Construction 5 21.0 23.6 3 31.0 34.6 7 11.0 13.2  Trade 7 33.9 33.0 Sales per worker (log, PPP)   1st tertile 33 20.7 20.9 38 13.4 13.4 10 12.5 22.7   2nd tertile 33 18.1 19.6 34 12.3 14.1 44 6.8 11.5   3rd tertile 33 17.5 19.1 28 10.7 14.2 46 4.0 12.1 Loan 19 11 30 Manager’s education   Less than secondary 11 34.3 43.6 6 20.4 24.6  Secondary 14 24.0 23.9 11 16.3 12.6  Postsecondary 75 13.0 15.4 83 11.2 13.3 Manager’s experience   0–3 16 11.6 13.1 28 11.2 11.3 9 9.8 28.6   4–10 28 17.8 20.6 33 14.4 17.3 30 5.0 12.5   11+ 56 21.8 21.7 39 11.2 12.7 61 5.7 10.1 Workforce education   Less than secondary 38 21.0 21.0 30 12.3 10.6  Secondary 41 16.9 16.9 49 11.8 16.1  Postsecondary 21 17.4 17.4 21 12.8 14.9 Labor regulatory obstacle 25 26 15 Labor inspections 3 5 Bribing _ labor inspections 12 7 Nonresp. _ labor inspections 36 6 Nonresp. _ bribing labor inspections 38 12 Tax obstacle 46 46 52 Tax inspections 3 3 2 Bribing _ tax inspections 14 5 40 Nonresp. _ tax inspections 24 10 18 Nonresp. _ bribing tax inspections 16 10 2 Source: Calculations using individual country ICA survey data for MENA economies that included the questions: “What percentage of the workforce would you estimate the typical establishment in your sector reports to social security” and “What percentage of total sales would you estimate the typical establishment in your sector reports for tax purposes? 166 West Bank and Gaza Syrian Arab Republic (2009) Yemen, Rep. (2010) Lebanon (2009) economy (2006) % workers % sales % workers % sales % workers % sales % workers % sales % of under­ under­ % of under­ under­ % of under­ under­ % of under­ under­ sample reporting reporting sample reporting reporting sample reporting reporting sample reporting reporting 43 43 14 14 34 52 18 13 26 46.4 31.6 52 50.7 59.2 36 11.5 16.1 46 17.6 13.7 30 48.0 39.3 25 49.1 51.5 36 16.7 21.2 38 11.5 12.0 18 47.2 34.6 11 19.1 34.5 13 11.8 13.6 9 14.1 15.6 16 38.0 31.8 7 33.6 44.0 9 11.9 10.4 5 12.5 10.8 10 24.0 26.0 5 9.7 19.9 6 14.3 19.7 2 4.2 0.0 18 44.7 36.2 12 41.4 44.0 8 16.5 15.3 18 15.2 13.1 15 45.9 38.8 25 43.6 47.9 13 10.5 11.0 30 15.0 13.1 67 42.3 32.0 62 43.3 53.9 79 14.3 19.0 52 13.8 12.6 69 43.6 36.8 51 45.6 55.6 37 19.8 25.8 81 15.2 12.4 20 39.6 24.6 18 37.1 50.4 31 9.4 11.2 12 9.8 13.4 2 46.4 49.8 1 0.0 2.5 12 16.0 16.7 7 13.8 17.9 9 48.4 28.7 30 45.3 48.7 20 10.1 14.5 0 22 43.0 35.1 48 44.3 54.0 48 14.2 14.5 26 27.1 24.0 24 48.5 36.4 35 44.3 52.0 9 22.8 29.7 48 10.3 10.3 54 41.1 32.1 17 38.4 46.0 43 11.5 18.1 26 9.7 6.5 15 12 53 18 36 49.5 61.5 2 27.8 46.1 28 46.7 50.9 12 13.5 22.7 36 34.5 41.8 86 13.2 15.9 2 23.6 15.6 3 57.0 66.6 27 10.1 12.4 3 5.7 4.3 17 47.7 38.1 23 41.0 45.5 27 13.5 19.1 21 12.6 16.7 81 42.9 33.4 74 43.7 53.4 45 17.7 19.9 76 15.4 12.3 70 44.3 34.0 25 26.7 25.8 12 43.1 30.4 49 10.8 15.9 19 36.6 33.9 26 4.7 8.5 49 10 36 12 4 2 0 53 18 4 23 0 76 24 1 74 42 43 59 37 3 6 1 2 46 45 8 1 24 21 58 63 25 23 56 0 167   168 Informality and the Firm Annex table 3A.2  Underreporting, Firm Size, and Productivity All economies Tobit (censored), Seemingly unrelated marginal effects regression (1) (2) (3) (4) Variable Workers Sales Workers Sales Size: 15–49 (over 10–14) −2.076 −1.107 −3.862 −2.971 [0.000]*** [0.028]** [0.001]*** [0.023]** Size: 50–99 −3.178 −1.617 −6.549 −4.66 [0.000]*** [0.018]** [0.000]*** [0.008]*** Size: 100–249 −4.128 −2.385 −8.967 −6.218 [0.000]*** [0.001]*** [0.000]*** [0.000]*** Size: 250+ −7.532 −3.684 −14.514 −7.946 [0.000]*** [0.000]*** [0.000]*** [0.000]*** Log size Productivity −7.166 −3.007 −7.557 −1.882 (log sales/worker, PPP) [0.000]*** [0.000]*** [0.000]*** [0.274] Country dummies Y Y Y Y Year dummies N N N N Firm effects N N N N Cluster firm-level N N N N Governorate dummies N N N N Sector dummies Y Y Y Y Observations (N) 3,547 3,525 3,453 3,453 No. of firms 1,623 1,617 1,617 1,617 R2 or pseudo-R2 0.046 0.028 0.2133 0.1265 Source: Calculations using ICA survey data for MENA economies. Note: Columns 1 and 2 were estimated using Tobit (censored). Although in column 1 the dependent variable was the share of workers underreported to social security, in column 2 the dependent variable is the share of sales underreported for tax purposes. Marginal effects are reported throughout. Columns 3 and 4 were jointly estimated, using seemingly unrelated regressions. Column 5 uses pooled data for Egypt covering the three waves of data (2004, 2007, and 2008) and including year dummies. Columns 6, 7, and 8 use panel data for Egypt for 2004, 2007, and 2008. Columns 6 and 7 have the share of workers underreporting as a dependent variable. Results in column 6 were estimated using firm-fixed effects in a linear probability model, and results in column 7 were estimated using firm-random effects in a censored Tobit model. The dependent variable in column 8 was the share of sales underreported. Results were estimated using a censored Tobit model. Columns 1 to 4 include control for firm age. p values in brackets. *significant at 10 percent; **significant at 5 percent; ***significant at 1 percent. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 169 Egypt panel Pooled Panel (5) (6) (7) (8) Workers Workers Sales −2.781 −2.085 −1.895 −1.472 [0.000]*** [0.001]*** [0.000]*** [0.000]*** −0.775 −1.324 −0.524 −0.451 [0.002]*** [0.008]*** [0.002]*** [0.006]*** N N N N Y Y Y Y N FIXED RANDOM RANDOM N Y N N Y N Y Y Y N Y Y 3,022 3,028 3,028 3,022 1,623 1,623 1,623 1,617 0.018 0.028 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 170 Informality and the Firm Annex table 3A.3  Underreporting and Manager’s Education All economies Tobit (censored), Marginal Seemingly unrelated effects regression (1) (2) (3) (4) Variable Workers Sales Workers Sales Size: 15–49 (over 10–14) −1.949 −1.247 −4.133 −3.156 [0.012]** [0.089]* [0.004]*** [0.074]* Size: 50–99 −1.961 −0.983 −3.656 −2.311 [0.070]* [0.334] [0.069]* [0.345] Size: 100–249 −3.38 −3.03 −6.928 −7.194 [0.001]*** [0.003]*** [0.000]*** [0.002]*** Size: 250+ −5.878 −4.412 −10.41 −8.271 [0.000]*** [0.000]*** [0.000]*** [0.000]*** Log size Productivity −0.0512 0.133 −0.218 0.185 (log sales/worker, PPP) [0.595] [0.141] [0.205] [0.376] Secondary education −2.63 −3.765 −4.772 −13.29 (base: less than secondary) [0.065]* [0.005]*** [0.092]* [0.000]*** Postsecondary education −2.695 −3.392 −4.925 −9.992 [0.045]** [0.007]*** [0.052]* [0.001]*** Manager’s years 0.0236 0.0611 0.0523 0.117 of managerial experience [0.451] [0.037]** [0.350] [0.085]* Worker education: secondary −2.202 1.142 −5.138 5.044 (base: less than secondary) [0.115] [0.378] [0.045]** [0.106] Worker education: −6.216 −3.848 −9.545 −4.889 postsecondary [0.000]*** [0.014]** [0.001]*** [0.167] Country dummies Y Y Y Y Year dummies N N N N Firm effects N N N N Cluster firm-level N N N N Governorate dummies N N N N Sector dummies Y Y Y Y Observations (N) 1,801 1,790 1,782 1,782 No. of firms 1,801 1,790 1,782 1,782 R2 or pseudo-R2 0.012 0.010 0.046 0.036 Source: Calculations using ICA survey data for MENA economies. Note: Columns 1 and 2 were estimated using Tobit (censored). In column 1 the dependent variable was the share of workers underreported to social security, but in column 2 the dependent variable is the share of sales underreported for tax purposes. Marginal effects are reported throughout. Columns 3 and 4 were jointly estimated, using seemingly unrelated regressions. Columns 5, 6, 7, and 8 use panel data for Egypt for 2004, 2007, and 2008. The dependent variable (columns 5 and 7) is the share of workers underreported, whereas in columns 6 and 8 it is the share of sales underreported. Results in columns 5 and 6 were estimated using firm-fixed effects in a linear probability model, and results in column 7 and 8 were estimated using firm-random effects in a censored Tobit model. Columns 1 to 4 include control for firm age. p values in brackets; *significant at 10 percent; **significant at 5 percent; ***significant at 1 percent. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 171 Egypt panel Linear probability model Censored Tobit (5) (6) (7) (8) Workers Sales Workers Sales −1.535 −1.349 −1.628 −1.325 [0.021]** [0.099]* [0.000]*** [0.000]*** −1.172 −0.281 −0.444 −0.364 [0.022]** [0.606] [0.012]** [0.031]** −7.364 −11.005 −1.551 −1.713 [0.042]** [0.003]*** [0.100]* [0.062]* −9.591 −11.024 −2.834 −1.905 [0.003]*** [0.001]*** [0.002]*** [0.026]** −0.163 −0.092 −0.065 −0.029 [0.014]** [0.198] [0.004]*** [0.182] 4.424 10.989 0.551 2.38 [0.183] [0.002]*** [0.614] [0.024]** 5.84 1.382 −1.767 −4.009 [0.290] [0.823] [0.340] [0.026]** N N N N Y Y Y Y FIXED FIXED RANDOM RANDOM Y Y N N N N Y Y N N Y Y 2,999 2,993 2,999 2,993 1,613 1,607 1,613 1,607 0.039 0.020 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 172 Informality and the Firm Annex table 3A.4  Underreporting, Regulations, Enforcement, and Corruption All economies: pooled Workers Variable (1) (2) (3) (4) Size: 15–49 (over 10–14) −2.031 −2.031 −1.993 −1.995 [0.000]*** [0.000]*** [0.000]*** [0.000]*** Size: 50–99 −3.216 −3.198 −3.179 −3.221 [0.000]*** [0.000]*** [0.000]*** [0.000]*** Size: 100–249 −4.229 −4.25 −4.203 −4.212 [0.000]*** [0.000]*** [0.000]*** [0.000]*** Size: 250+ −7.21 −7.205 −7.203 −7.172 [0.000]*** [0.000]*** [0.000]*** [0.000]*** log size Productivity (log sales/worker, PPP) −0.105 −0.106 −0.104 −0.102 [0.074]* [0.072]* [0.077]* [0.083]* Regulation obstacle 2.253 2.56 2.212 0.566 [0.000]*** [0.000]*** [0.000]*** [0.462] Inspections (log) −0.833 −0.829 −1.222 −1.416 [0.027]** [0.028]** [0.005]*** [0.001]*** Bribe inspectors 2.632 3.269 0.451 2.458 [0.000]*** [0.000]*** [0.743] [0.001]*** Nonresponse bribe inspectors 5.514 5.541 5.023 5.257 [0.000]*** [0.000]*** [0.000]*** [0.000]*** Regulation obst_bribe −1.437 [0.231] Inspections*bribe 1.488 [0.065]* Regulation obst*insp. 1.565 [0.003]*** Country dummies Y Y Y Y Sector dummies Y Y Y Y Location dummies Y Y Y Y Firm random effects N N N N Observations (N) 3,537 3,537 3,537 3,537 Pseudo-R2 0.0508 0.0509 0.0510 0.0513 Source: Calculations using ICA survey data for Egypt 2004, 2007, and 2008. Note: Columns 1 to 8 were estimated using Tobit (censored) using pooled data for Egypt with year dummies. Columns 6 to 8 use the panel dimension of the data with firm-random effects. In columns 1 to 4 and 9 to 10, the dependent variable was the share of workers underreported to social security, but in columns 5 to 8 and 11 the dependent variable is the share of sales underreported for tax purposes. Marginal effects are reported throughout. Columns 1 to 4 include control for firm age. p values in brackets; *significant at 10 percent; **significant at 5 percent; ***significant at 1 percent. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 173 All economies: pooled Panel Sales Workers Sales (5) (6) (7) (8) (9) (10) (11) −0.992 −0.966 −1.003 −0.966 [0.052]* [0.060]* [0.050]* [0.060]* −1.620 −1.574 −1.636 −1.574 [0.020]** [0.024]** [0.019]** [0.024]** −2.17 −2.156 −2.204 −2.156 [0.003]*** [0.003]*** [0.002]*** [0.003]*** −3.826 −3.75 −3.87 −3.75 [0.000]*** [0.000]*** [0.000]*** [0.000]*** −1.651 −1.626 −1.542 [0.000]*** [0.000]*** [0.000]*** −0.0721 −0.0726 −0.0697 −0.0726 −0.551 −0.554 −0.396 [0.196] [0.193] [0.212] [0.193] [0.003]*** [0.003]*** [0.025]** 1.152 1.76 1.124 1.76 3.141 3.128 2.113 [0.007]*** [0.000]*** [0.008]*** [0.000]*** [0.000]*** [0.000]*** [0.037]** 0.546 0.537 0.913 0.537 −1.232 −0.779 0.018 [0.115] [0.122] [0.029]** [0.122] [0.001]*** [0.069]* [0.972] 2.716 4.554 4.282 4.554 1.714 5.612 −0.151 [0.000]*** [0.000]*** [0.000]*** [0.000]*** [0.018]** [0.002]*** [0.830] 0.145 0.116 0.386 0.116 −2.742 −2.787 1.023 [0.854] [0.884] [0.633] [0.884] [0.389] [0.380] [0.746] −3.256 [0.002]*** −1.104 [0.116] −3.256 −2.226 −0.609 [0.002]*** [0.016]** [0.321] Y Y Y Y N N N Y Y Y Y Y Y Y Y Y Y Y Y Y Y N N N N Y Y Y 3,483 3,483 3,483 3,483 2,696 2,696 2,713 0.0321 0.0327 0.0323 0.0327 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 174 Informality and the Firm Annex table 3A.5 The Effect of Firm Size and Productivity on the Association between Regulations and Underreporting X = Regulation X = Inpections MENA Egypt panel MENA Egypt panel countries: Linear probility Censored countries: Linear probility Censored pooled model Tobit pooled model Tobit Variable (1) (2) (3) (4) (5) (6) Regulatory obstacle −0.745 15.428 7.332 2.276 8.317 2.972 [0.568] [0.016]** [0.001]*** [0.000]*** [0.000]*** [0.000]*** Inspections (log) −0.833 −3.606 −1.335 −1.93 −3.778 −2.925 [0.028]** [0.002]*** [0.000]*** [0.007]*** [0.334] [0.018]** Bribes 2.549 3.88 2.444 2.441 3.933 2.492 [0.001]*** [0.100] [0.001]*** [0.001]*** [0.097]* [0.001]*** Nonresponse bribes 5.481 5.37 −0.875 5.212 5.513 −0.718 [0.000]*** [0.556] [0.749] [0.000]*** [0.548] [0.795] X_size15_49 0.224 −1.237 [0.852] [0.047]** X_size50_99 3.049 −0.474 [0.065]* [0.564] X_size100_249 1.521 −1.256 [0.363] [0.137] X_size250+ 0.492 0.367 [0.802] [0.681] X_size −1.624 −1.008 0.103 0.095 [0.128] [0.008]*** [0.875] [0.692] X_productivity 0.282 −0.268 −0.043 0.212 −0.054 0.33 [0.016]** [0.813] [0.905] [0.000]*** [0.947] [0.185] Country dummies Y N N Y N N Sector dummies Y N N Y N N Firm effects N Y-F Y-R N Y-F Y-R Observations (N) 3,537 2,696 2,696 3,537 2,696 2,696 Pseudo-R2 0.0514 0.052 0.052 0.051 Note: Columns 1, 4, 7, and 10 include control for firm size, age, sector, location dummies and year dummies; and columns 2–3, 5–6, 8–9, and 11–12 include control for firm size and age, and year dummies. p values in brackets; *significant at 10 percent; **significant at 5 percent; ***significant at 1 percent. Y-F stands for firm fixed effects and Y-R stands for firm random effects. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 175 X = Bribes X = Nonresponse labor bribe MENA Egypt panel MENA Egypt panel countries: Linear probility Censored countries: Linear probility Censored pooled model Tobit pooled model Tobit (7) (8) (9) (10) (11) (12) 2.233 8.229 2.92 2.269 8.352 2.982 [0.000]*** [0.000]*** [0.000]*** [0.000]*** [0.000]*** [0.000]*** −0.823 −3.592 −1.326 −0.762 −3.588 −1.329 [0.03]** [0.003]*** [0.000]*** [0.046]** [0.003]*** [0.000]*** −0.2 12.864 9.352 2.468 3.994 2.511 [0.907] [0.130] [0.002]*** [0.001]*** [0.092]* [0.001]*** 5.529 5.233 −0.776 7.044 4.358 −4.59 [0.000]*** [0.568] [0.778] [0.000]*** [0.902] [0.605] 0.57 1.574 [0.716] [0.169] 3.28 1.699 [0.09]* [0.289] 2.817 1.896 [0.161] [0.258] 2.78 1.893 [0.245] [0.348] −0.867 −0.564 4.878 1.699 [0.628] [0.289] [0.175] [0.311] 0.17 −1.613 −1.055 −0.376 −5.713 −0.861 [0.244] [0.236] [0.018]** [0.001]*** [0.362] [0.655] Y N N Y N N Y N N Y N Y N Y-F Y-R N Y-F Y-R 3,537 2,696 2,696 3,537 2,696 2,696 0.0512 0.052 0.0516 0.052 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 176 Annex table 3A.6  Description of Data Sources Economy Survey Description Egypt, Arab Rep.; Investment Climate Data from the World Bank Enterprise Survey repository, which assembles data from the ICA survey,17 are Jordan; Syrian Assessment (ICA) used. The countries included were those where the ICA surveys asked questions about the percentage of Arab Republic; Surveys sales and the percentage of workers that a typical firm of a specific sector would report for tax or social Yemen, Rep.; security purposes.18 Information provided through this type of question has been widely used in the Lebanon; West literature with consistent results. Informality is measured using the percentage of workers that firms do not Bank and Gaza report for social security purposes. Although the percentage of sales reported for tax purposes are also economy considered, the main focus in this report is on worker underreporting. Morocco (2007) Investment Climate Survey of micro-firms with up to five workers in Morocco in 2007. These data include matched employer- Assessment Surveys employee information for 219 firms, 264 salaried workers, 127 employers, and 26 self-employed workers. of micro- and small All workers in each firm were interviewed in the context of this survey. Data cover the cities of Casablanca, firms Rabat, Salé, Témara, and Fès, which are located in urban and rural areas. Four sectors were chosen for the survey: manufacturing, construction, trade, and services. The survey covers all the key sectors of informal employment identified by the 2004 Morocco national survey on the nonagricultural informal sector (ENSINA): trade, services, construction, textile, clothes, and shoes. ENSINA indicates that these sectors employ 37 percent, 20 percent, 7 percent, and 50 percent of their workforce informally, respectively (Direction de la Statistique 2004; Alami 2008). It also indicates that entrepreneurship19 forms an important part of the national informal sector by representing 69 percent of the total informal employment. These data consist of 219 firms and 417 individuals (264 salaried workers, 26 self-employed,20 and 127 employers21). Summary statistics on firms, firm owners, and employees in the sample are presented below. Egypt, Arab Rep. Investment Climate Survey of micro- and small firms in Egypt 2008. The survey covers 500 firms selected from a sample frame of (2008) Assessment Surveys 25,000 firms of the Center for Social Research at the American University of Cairo. Firms have up to 50 workers. of micro- and small This sample frame includes firms that are either unregistered or unlicensed or do not keep formal accounts. It firms covers eight governorates: Cairo, Alexandria, Damittra, Gharbia, Giza, Fayoum, Assuit, and Souhag. About 89 percent of firms in the sample have fewer than five employees; 115 of the sample have between 6 and 50 employees; 134 are manufacturing firms; and 366 are in services, including retail. Summary statistics on the sample are presented below. Lebanon (2004); Firm-level data Data from the ERF on surveys in Lebanon, Egypt and Morocco of small and medium enterprises with up to Egypt, Arab collected by the 50 workers. The sample for Egypt includes 4,958 firms, for Lebanon 2,948 firms, and for Morocco 5,210 Rep. (2003); Economic Research firms. For a more detailed description of the data see Elbadawi and Loayza (2008). Morocco (2002) Forum (ERF) (MEAS) Annex table 3A.7  Descriptive Statistics on Moroccan Micro-Firms, Their Workers, and Their Owners Firms Mean Median Mean Median General Manufacturing 31% Infrastructure Owns computer 20% and innovation Construction 7% Has telephone service 31% Trade 21% Connected to public electricty 86% Services 41% Uses Internet 21% Urban 66% Improved quality within last year 40% Suburban 33% Introduced new product within last year 29% Rural 2% Introduced new materials within last year 16% Location rented by firm 43% Firm characteristics Age of firm (years) 10.19 7 Location owned by firm 50% Permanent workers 1.92 2 Location is mobile 7% Profit (DH) 5,756 1,850 Productivity 5,453 3,625 Market Local market (% of sales) 81% orientation Sales margin (%) 17.77 15 Regional market (% of sales) 13% National market (% of sales) 5% Invoices clients 25% Export market (% of sales) 1% Accepts checks 49% Moroccan consumers (% of client base) 87% Finance Moroccan micro-firms (% of client base) 9% Has business bank account 21% Large Moroccan firms (% client base) 2% Requested bank credit in 13% Quality problems with clients 7% the past % of firm resources from 136% bank credit table continues next page 177   178 Annex table 3A.7  Descriptive Statistics on Moroccan Micro-Firms, Their Workers, and Their Owners (continued) Firms Mean Median Mean Median Competition Government Bribe as % of market to receive public 32% 41% market Intense competition: 33% Application of law is foreseeable and 48% informal producers consistent Intense competition: small 36% Bribe often expect by inspectors 43% and medium producers Intense competition: large 31% Firm knows level of bribe expected in 31% producers advance Intense competition: 19% % of sales devoted to bribes 11% 19% imports Impact of competition: 25% Firm subject to inspections 46% forced to reduce price Impact of competition: 18% forced to reduce sales Impact of competition: 24% forced to reduce both Impact of competiton: 33% none Informality and the Firm 179 Firms’ Owners Mean Median Female owner 17% Education of owner (years) 8.63 9 Age of owner (years) 42.03 39 Owner is married 74% Number of children 2.45 2 Owner was previously unemployed 26% Household income for owner (DH) 4,109.62 3,500 Workers Mean Median Age (years) 30.17 28 Education (years) 7.46 7 Paid daily 13% Paid weekly 50% Paid every two weeks 2% Paid every month 35% Male 75% Hourly wage (DH) 8.71 7 Weekly hours of work 55.89 54 Married 27% Number of children 0.64 0 Employer himself 4% No relationship to employer 73% Married to employer 1% Child of employer 8% Sibling of employer 4% Parent of employer 2% Other blood relation to employer 8% Household income (DH) 3,141.26 3,000 Age of head of household 48.89 48 Size of household 8.00 5 Annex table 3A.8  Descriptive Statistics on Egyptian Micro-Firms, Their Workers, and Their Owners Small and micro-firms Mean Median Micro- ≤ 5 89% Small >5 11% No. of employees 4 3 Industry sector 26.60% Trade sector 48.60% table continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 180 Informality and the Firm Annex table 3A.8  Descriptive Statistics on Egyptian Micro-Firms, Their Workers, and Their Owners (continued) Small and micro-firms Mean Median Service sector 24.80% Urban Egypt, Arab Rep. 76.60% Lower Egypt, Arab Rep. 12.40% Upper Egypt, Arab Rep. 11.00% Age of firm 17 12 Sales (LE) 27,384 8,000 Productivity 7,835 2,498 Sales growth −15.26% −15.00% Percent of sales not paid in 2004 4.85% Has checking account (0/1) 5.92% Number of visits by tax inspector 1.60 Informal payment was expected by tax inspector (0/1) 4.38% Are skills and education of workers a constraint to growth? 27% In establishment, business activity varies by season 67% Percent of bank loan to total financing 2.95% Workers (%) Mean Share of workers with postsecondary education 16.89 Share of workers with secondary education 39.16 Share of workers with incomplete secondary education 11.49 Share of workers with complete primary education 6.88 Share of workers with incomplete primary education 25.58 Share of female employees 5.56 Share of relatives among employees 13.63 Lost production days due to strikes or absenteeism 8.40 Share of workers who left firm in the last 12 months 10.44 Firm owners (%) Mean Female owner 15 Education of owner/top manager   Postsecondary education 29.40   Completed secondary education 35.20   Incomplete secondary education 5.20   Complete primary education 7.00   Incomplete primary education; no education 23.20 Age of the owner (years)   Less than 21 0.40   21–30 11.42   31–40 29.46   41–50 28.66   51+ 30.06 Manager had previous experience in formal firm (0/1) 9.60 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 181 Notes 1. See Oviedo (2008). 2. See La Porta and Shleifer (2008) for a recent review of this strand of the ­literature. 3. See Perry and others (2007) for a detailed description of these three ­viewpoints. 4. Source: World Development Indicators 2011 and calculation using ICSs. 5. See, for example, Asiedu and Freeman (2008), Chong and Gradstein (2006), Elbadawi and Loayza (2008), McKenzie and Sakho (2010), and World Bank (2009). 6. Informality may be chosen to avoid burdensome government regulations such as hiring and firing costs, government standards for products and production processes, and strict working hours and wages. 7. For example, impossibility of providing receipts to clients and risk of being caught. 8. For example, limited access to markets, formal financing, courts, or other forms of contract enforcement and government services. 9. For a more detailed analysis of the tax wedge see chapter 5. Note that for firms at the cusp of formalizing, factors such as how progressive the taxation system is (including income tax for individuals), and whether there are exemptions from taxes below a certain threshold, should be considered. 10. For more details, see Ramalho (2008). 11. Methodologically, in the regression analysis of worker underreporting, the dependent variable measures the share of the workforce that is not reported to social security; for sales underreporting, the dependent variable measures the share of sales that are not reported to the tax office. Following Johnson and others (2000), the presented results were estimated using censored estimation models (Tobit) to take into account the fact that sales and worker underreporting in MENA are left censored. Each estimation provides Tobit marginal effects at the censoring point, representing the marginal effect of each explanatory variable on the percent of sales or workers underreported, conditional on the probability that a firm is in the uncensored group (that is, an informal firm). The figures in the main text present the results of the country by country analysis. This analysis was complemented with cross-country and panel data analysis. The latter over- comes the issue of firm unobserved characteristics that might drive both shares of tax/social security reporting and other outcomes (for example, productivity). Sales and worker underreporting for tax purposes and social security have been widely used as proxy for informality (see, for example, Gatti and Honorati 2008; Johnson and others 2000; and Ingram and others Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 182 Informality and the Firm 2007; Perry and others 2007). They are computed using the information from two questions: (1) “What percentage of total sales would you estimate the typical establishment in your sector reports for tax purposes?” and (2) “What do you think is the percentage of total workforce that is reported for purposes of labor regulation and social insurance in a typical establish- ment in your sector?” 12. Note that in MENA, the only countries where manager education and worker underreporting variables were collected in ICS surveys covering underreport- ing were Egypt, the Republic of Yemen, and Lebanon. 13. In contrast, in the Republic of Yemen, after controlling for firm size, changes in the manager’s level of education are not significant correlated with worker underreporting. In fact, in the Republic of Yemen, firm size predicts almost perfectly the manager’s level of education. For example, in small firms, 81 percent of managers have less than postsecondary education, whereas in large firms, fewer than 13 percent of managers are in this ­situation. 14. Lebanon’s controls do not include location because of the unavailability of corresponding information in the survey data. 15. The two estimates are obtained from 2SLS and probit estimation techniques, respectively. 16. Note that this chapter is not arguing that being informal is causing firms to be less productive or to offer workers less protection. Formal firms might be more productive as a result of formalization, or just because they have more capital, use more technology, and have better skilled owners. Although some of these factors are controlled for and the findings tested for robustness using panel data, the results should not be interpreted as causal. 17. In this instance, a word of caution is necessary given that the collected infor- mation can indeed describe the incidence of informality on firms, but does not allow aggregates of informality over the entire population due to the lack of weighted data. Although the results might not be descriptive at the general level, they still remain indicative for the policy standpoint. 18. These questions are: “What percentage of total sales would you estimate the typical establishment in your sector reports for tax purposes?” and “What do you think is the percentage of total workforce that is reported for purposes of labor regulation and social insurance in a typical establishment in your sector?” 19. Defined as the job status of either self-employed or employer. 20. Firm owners who do not employ other workers are defined as self-employed. 21. Firm owners who employ other workers are defined as employers. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 183 References Alami, R. M. 2008. “Comment définir l’informel?” La revue economia 2 (Feb.): 80. Almeida, R., and P. Carneiro. 2009. “Enforcement of Regulation, Informal Employment, Firm Size and Firm Performance.” Journal of Comparative Economics 37 (1): 28–46. Angel-Urdinola, D., and A. Kuddo. 2010. “Key Characteristics of Employment Regulation in the Middle East and North Africa.” SP Discussion Paper 1006. Washington, DC: World Bank. Asiedu, E., and J. Freeman. 2008. “The Effect of Corruption on Investment Growth: Evidence from Firms in Latin America, Sub-Saharan Africa and Transition Countries.” Working Papers in Theoretical and Applied Economics 2008/02. Lawrence: University of Kansas, Department of Economics. Botero, J. C., S. Djankov, R. La Porta, F. Lopez-de-Silanes, and A. Shleifer. 2004. “The Regulation of Labor.” Quarterly Journal of Economics 119 (4): 1339–1382. Bruhn, M. 2011. “License to Sell: The Effect of Business Registration Reform on Entrepreneurial Activity in Mexico.” Review of Economics and Statistics 93 (1): 382–386. Chong, A., and M. Gradstein. 2006. “Inequality and Informality.” Discussion Paper 5545. London: CEPR. De Paula, A., and J. A. Scheinkman. 2007. “The Informal Sector.” Working paper 13486. Cambridge, MA: NBER. De Soto, H. 1989. The Other Path: The Invisible Revolution in the Third World. New York: HarperCollins. Direction de la Statistique. 2004. “L’enquête nationale sur le secteur informel non agricole.” Rabat, Morocco: Direction de la Statistique. Djankov, S., R. F. Lopez-de-Silanes, and A. Shleifer. 2002. “The Regulation of Entry.” Quarterly Journal of Economics 117 (1): 1–37. Djankov, S., C. McLiesh, and R. Ramalho. 2006. “Regulation and Growth.” Economics Letters 92 (3): 395–401. Elbadawi, I., and N. Loayza. 2008. “Informality, Employment and Economic Development in the Arab World.” Journal of Development and Economic Policies 10 (2): 25–75. Friedman, E., S. Johnson, D. Kaufmann, and P. Zoido-Lobaton. 2000. “Dodging the Grabbing Hand: The Determinants of Unofficial Activity in 69 Countries.” Journal of Public Economics 76 (3): 459–493. Galal, A. 2005. “The Economics of Formalization: Potential Winners and Losers from Formalization in Egypt.” In Investment Climate, Growth, and Poverty, ed. G. Kochendorfer-Lucius and B. Pleskovic. Washington DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 184 Informality and the Firm Gatti, R., and M. Honorati. 2008. “Informality among Formal Firms: Firm-Level, Cross-Country Evidence on Tax Compliance and Access to Credit.” Policy Research Working Paper 4476. Washington, DC: World Bank. ILO (International Labour Organization). 1993. “Resolutions Concerning International Classification of Status in Employment.” Adopted by the 15th International Conference of Labour Statisticians. Geneva: ILO. ———. 2002. “Decent Work and the Informal Economy: Sixth Item on the Agenda.” International Labour Conference 90th: 2002. Report VII (2A) Geneva: ILO. http://www.ilo.org/public/english/standards/relm/ilc/ilc90​ /­pdf/rep-vi.pdf. Ingram, M., V. Ramachandran, and V. Desai. 2007. “Why Do Firms Choose to be Informal? Evidence from the Africa Investment Climate Surveys.” Washington, DC: World Bank. Ishengoma, E., and R. Kappel. 2006. “Economic Growth and Poverty: Does Formalization of Informal Enterprises Matter?” Working Paper Series 20. Hamburg: GIGA (German Institute of Global and Area Studies). Johnson, S, D. Kaufmann, J. McMillan, and C. Woodruff. 2000. “Why Do Firms Hide? Bribes and Unofficial Activity after Communism.” Journal of Public Economics 76 (3): 495–520. La Porta, R., and A. Shleifer. 2008. “The Unofficial Economy and Economic Development.” Brookings Papers on Economic Activity 2: 275–364. Levenson, A., and W. Maloney. 1998. “The Informal Sector, Firm Dynamics, and Institutional Participation.” Policy Research Working Paper WPS1988. Washington, DC: World Bank. Loayza, N., A. M. Oviedo, and L. Serven. 2004. “Regulation and Macroeconomic Performance.” Washington, DC: World Bank. McKenzie, D., and Y. S. Sakho. 2010. “Does It Pay Firms to Register for Taxes? The Impact of Formality on Firm Profitability.” Journal of Development Economics 91 (1): 15–24. Oviedo, A. 2008. “Economic Informality: Causes, Costs, and Policies: A Critical Survey of the International Literature.” Washington, DC: World Bank. Perry, G., W. Maloney, O. Arias, P. Fajnzylber, A. Mason, and J. Saavedra-Chanduvi. 2007. Informality: Exit and Exclusion. Washington, DC: World Bank. Pierre, G., and S. Scarpetta. 2006. “How Labor Markets Can Combine Workers’ Protection with Job Creation: A Partial Review of Some Key Issues and Policy Options.” Social Protection Discussion Paper 0716. Washington, DC: World Bank. Ramalho, R. 2007. “Adding a Million Taxpayers.” Doing Business 2008. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality and the Firm 185 Silva, J., M. Benyagoub, and G. Wallner. 2010. “Informality, Firm Performance and Labor Market Outcomes: Evidence from Matched Employer-Employee Data for Morocco.” Mimeo. Washington, DC: World Bank. World Bank. 2007. “Republic of Bolivia: Policies for Increasing Firms’ Formality and Productivity.” Report 40057-BO. Washington, DC: World Bank. ———. 2008. “Climat de l’Investissement au Maroc: Créer les conditions du changement structurel.” Washington, DC: World Bank. ———. 2009. “From Privilege to Competition: Unlocking Private-Led Growth in Middle East and North Africa.” Washington, DC: World Bank. ———. 2010. “Egypt Investment Climate Assessment 2009: Accelerating Private Enterprise-Led Growth.” Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 C H A PTER 4 Informality: Choice or Exclusion? SUMMARY: Building on the analysis of earning differentials, job mobility patterns, self-reported attitudes toward moving jobs, and self-rated satisfac- tion, this chapter provides a characterization of the quality of informal jobs. The existing evidence corroborates the view that segmentations exist across formal and informal jobs and that much of the observed informality is likely due to exclusion factors. Returns to human capital differ sharply, with esti- mated formality premiums ranging up to 50 percent for salaried workers. Informal salaried workers also report lower access to nonpension benefits, including annual, maternity, and official leave, and to training and skills upgrading. Moreover, informal jobs in MENA countries have longer durations than in comparator countries, thus dispelling the notion that informal jobs are mainly an entry point to the labor market to help employers and new entrants (youth) overcome asymmetric information constraints. In some countries, notably Egypt, a prominent margin of segmentation is along the public/private employment margin, as indicated, for example, by the extremely long duration of public jobs. The lack of information on earnings for the self-employed for most countries limits a full analysis of self-employed work. However, where these data are available, as in Lebanon, the evidence suggests that the self- employed have higher earnings than informal salaried workers, but overall lower earnings than formal salaried workers, while a small percentage of them earns more than both. The large majority of self-employed workers, who tend to be older and have lower education than both informal and formal salaried workers, do not have access to social security. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 187   188 Informality: Choice or Exclusion? Introduction The previous chapters, in particular chapter 2, discussed the profile and characteristics of informal workers, pointing at common trends but also important heterogeneities. Overall, a large informal sector and the result- ing coverage gap represent a suboptimal equilibrium that calls for policy action: Welfare can be improved by extending instruments to better manage risk. However, the ultimate drivers of informality are critical for ­ identifying the appropriate policies. This chapter investigates the attri- butes of formal and informal jobs, including earnings, job mobility, access to leave and other benefits, and investment in training, as well as workers’ subjective perceptions. In addition to providing a full characterization of informal work, this evidence can help document empirically the extent to which informality is the result of workers’ voluntary choice or exclu- sion. This chapter first uses a variety of techniques, from simple compari- sons of wage distributions to multivariate regression analysis, including the use of interesting and innovative controls for measured ability, to estimate whether a formality premium exists. Although persistent for- mality premiums are prima facie evidence that returns to skills are not equalized across formal and informal jobs for individuals of observed similar skills, wage differences cannot be taken alone as evidence that segmentations prevent arbitrage between the two sectors. Many factors can confound comparisons between earnings from different jobs but are unmeasured (for example, flexibility and independence). To gain a better understanding, this chapter complements the information on wage dif- ferentials with evidence on job mobility. High mobility is an indicator of integrated labor markets, where workers and enterprises, following mar- ket signals, search for matches between skills and vacancies. In countries with integrated labor markets, informal jobs help reduce asymmetric information for new labor market entrants, who, once they have proven their skills, would move rapidly to formal employment or self-employment.­ Thus workers that are observed working informally might indeed choose to do so, because in labor markets with easy transitions they could move out of informality if they so wished. In line with evidence from around the world, formal jobs have a longer duration than informal jobs. However, high persistence in informal, lower-paying jobs (for workers of skills comparable to those in the formal sector) is likely to indicate that many barriers, including labor regulations, prevent informal workers from moving to better jobs. In addition to analyzing data on actual job transi- tions and associated wage differentials, this chapter discusses newly Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 189 ­ ollected data on self-reported attitudes toward job search, as well as the c role of networks. Finally, the chapter discusses access to benefits (beyond old-age pensions) and training, and overall working conditions. Informality and Wages On average, wages in the formal sector are higher than in the informal sector. Wage data from surveys in Egypt (2006, and youth, 2009), Lebanon (2010), Morocco (available for youth only, 2010), Syria (2010), and the Republic of Yemen (2006) are used to quantify whether a premium to formal work exists beyond the returns to specific measured skills and experience. To address the high degree of heterogeneity among informal workers, the analysis is restricted to those individuals living in urban areas who are wage earners and working full time (between 30 and 60 hours a week). Data from Egypt, Morocco, and the Republic of Yemen include public and private sector workers, but data for Lebanon and Syria include only the private sector.1 Whenever possible, a distinction is made between private and public sector formal wages. Table 4.1 provides information about average wages by gender of formal and informal workers across dif- ferent education levels and sectors. Informal sector wages are uniformly lower than formal sector wages, especially if compared with the formal private sector. Women systematically command lower wages than men. In Egypt, public sector wages are higher than private formal sector wages for low-to-medium-skill workers, thus providing incentives for a large share of the population to queue for public sector jobs. Kernel density plots provide a useful representation of differences in wages among formal and informal workers because they plot (smoothed) densities at all wage levels. The heavy solid line in figure 4.1 shows informal salaried workers’ ­ wages. In Egypt and Lebanon (to a larger extent), the distribution of wages for workers in the informal sector is such that wages in the informal sector are always “‘within’” the distribution of public and formal sector wages; that is, they are stochastically dominated by them, confirming the exis- tence of a formality wage premium throughout the wage distribution and lending prima facie support to the view that informality is an exclusion phenomenon. In other words, workers would prefer to move from the low-paid informal sector to the high-paid formal sector if possible. In the Republic of Yemen, however, this is not the case. Although in that country average formal worker earns slightly more than the average informal the ­ worker, a small pool of workers at the high end of the wage distribution is Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 190 Table 4.1  Average Hourly Wages in Local Currency Yemen, Rep., 2006 Egypt, Arab Rep., 2006 Lebanon, 2010 Syrian Arab Republic, 2010 Private Private Private Private Private Private Private Private   Public formal informal Public formal informal formal informal ­ formal informal Educational attainment   Primary and below 104.48 111.12 148.35 3.31 2.95 2.32 5.09 3.13 78.08 60.88   Preparatory/secondary 140.25 229.12 149.51 3.33 2.87 2.26 5.38 3.60 93.58 75.65 General   Secondary vocational 129.22 142.14 114.34 3.52 3.20 2.19 5.74 4.07 113.76 110.34   University and above 198.46 296.52 222.62 5.22 8.27 3.13 8.21 6.45 168.55 132.10 Gender  Female 124.01 207.55 69.93 2.83 2.68 1.37 6.85 4.00 91.76 72.61  Male 158.42 219.80 164.45 3.03 3.25 2.08 6.85 4.87 110.75 92.45 Age group   15–24 87.70 74.32 124.06 2.76 2.68 1.78 5.89 4.05 76.21 72.51   25–34 136.04 182.49 161.16 3.54 5.27 2.66 6.80 4.88 97.34 83.43   35–54 184.26 248.48 196.94 4.32 5.55 2.85 7.57 3.82 114.62 94.58   55–64 163.63 367.91 181.51 5.65 8.37 2.69 5.47 — 156.91 149.94 Sector of employment  Primary 138.33 185.93 157.53 3.67 5.13 2.09 6.50 3.38 — —  Secondary 182.95 174.68 105.85 4.60 5.15 2.42 6.17 3.79 — —  Tertiary 161.47 237.40 170.24 5.08 5.88 2.41 6.98 4.73 — — Total average 154.48 218.87 158.03 4.21 5.29 2.39 6.85 4.55 107.43 87.31 Sources: Angel-Urdinola and Tanabe, 2011; Alloush and others, 2011. Note: — = not available. Informality: Choice or Exclusion? 191 Figure 4.1  Kernel Density Plots of Hourly Wages by Employment Sector Egypt, Arab Rep., 2006 0.8 0.6 density 0.4 0.2 0 –4 –2 0 2 4 6 log hourly wages public private informal private formal Egypt, Arab Rep. (youth), 2009 0.8 0.6 density 0.4 0.2 0 –2 0 2 4 6 log hourly wages public private informal private formal Yemen, Rep., 2006 0.8 0.6 density 0.4 0.2 0 0 2 4 6 8 log hourly wages public private informal private formal figure continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 192 Informality: Choice or Exclusion? Figure 4.1  Kernel Density Plots of Hourly Wages by Employment Sector (continued) Morocco (youth), 2010 0.8 0.6 density 0.4 0.2 0 1 2 3 4 5 6 log hourly wage formal salaried informal salaried Lebanon, 2010 1 0.8 0.6 density 0.4 0.2 0 0 1 2 3 4 log hourly wage formal salaried informal salaried self-employed Syrian Arab Republic, 2010 0.8 0.6 density 0.4 0.2 0 2 3 4 5 6 7 log hourly wage formal informal Sources: Angel-Urdinola and Tanabe 2011; Alloush and others 2011. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 193 able to earn higher wages in the private informal sector than would be earned in the public sector (seen at the right tail of the wage distribution). For this small pool of workers, working in the informal sector is probably a profitable choice.2 A wage gap in the earnings of formal and informal workers persists, especially for women, even after controlling for a number of observable factors. Although kernel density plots are informative, they represent unconditional distributions; that is, they compare wages in the formal and informal sectors without controlling for observed characteristics that might affect the distribution (such as lower education and/or age). To overcome this limitation, standard Mincerian regressions were used. In their basic form, Mincerian regressions estimate wages as a function of a number of key observable characteristics, such as gender, age (nonlinear), and years of education. A binary variable for working in the formal sector is then added to this standard specification. The coefficient on this vari- able captures an estimate of the formality premium.3 This estimate varies significantly from country to country, highlighting the heterogeneity of the informality phenomenon in the region (figure 4.2). In Lebanon, the Figure 4.2 Estimated Formality Premiums by Gender in Different Countries 80 70 60 50 percent 40 30 20 10 0 Egypt, Arab Morocco Lebanon, 2010 Syrian Arab Rep., 2006 (youth), 2010 Republic, 2010 total males females Sources: Angel-Urdinola and Tanabe, 2011; Alloush and others, 2011. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 194 Informality: Choice or Exclusion? average estimated formality premium is 29 percent, in Egypt, 32 percent, and among Moroccan youth, as high as 53 percent.4 The formality pre- mium is higher among female workers in all cases. In Egypt this phenom- enon is particularly notable, where, on average, a female worker in the formal sector earns 1.5 times as much than an otherwise similar woman in the informal sector. Very low wages paid to female workers in the informal sector often explains why many women decide not to partici- pate in the labor market at all, especially if the opportunities to find formal jobs are limited. This finding is quite relevant in a country such as Egypt, where female participation rates remain low by international stan- dards.5 In both Lebanon and Syria, the formality premium is higher for workers with a low level of education, which suggests that unskilled informal workers might have lower bargaining power in wage setting (Alloush and others 2011). The differences in wages between the formal and the informal sector are explained by differences in the characteristics of formal and informal workers, as well as by differences in how the labor market rewards those characteristics. It is interesting to assess whether the observed differences in wages of formal and informal workers are mainly due to differences in the characteristics (endowments) of workers in the two groups, or to ­ differences in the extent to which formal and informal jobs reward those endowments. Returns to observed skills can differ for many reasons between formal and informal jobs, depending on firm scale, complemen- tarities with physical capital investment, and incentives to acquire experi- ence and upgrade skills. Although many observed and unobserved variables affect workers’ distribution into formal and informal occupa- tions, a simple Oaxaca decomposition provides an alternative insight into the likely drivers of observed wage differences.6 Figure 4.3 plots the result of this decomposition: The dark gray bars illustrate the share of the dif- ference in wages explained by differences in observable characteristics between formal and informal workers, as attributed to characteristics (endowments), returns (how formal and informal jobs reward these char- acteristics), and an interaction effect. For example, part of the difference in wages can be explained by the fact that, on average, formal workers are older and more educated than informal workers (Angel-Urdinola and Tanabe 2011). These differences in “endowments” explain about one- third of the wage differential in Lebanon and about half in Egypt. One could argue that this component of the wage differential is “fair,” because the market should positively reward higher education attainment and experience. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 195 Figure 4.3 Oaxaca Decomposition of the Differences in Wages 80 72% 70 60 49% 50 40 percent 33% 30 26% 24% 20 10 –5% 0 –10 Egypt, Arab Rep. 2006 Lebanon 2010 di erence in endowments di erence in rewards interaction Source: Based on Angel-Urdinola and Tanabe 2011. Note: For Egypt these are wages for urban workers in private and public sectors. For Lebanon, the wages are for private sector employees with no distinction between rural and urban areas. The light gray bars in figure 4.3 illustrate the gap in returns to endowment due to being an informal rather than a formal worker, or the extent to which equal endowment/work is not rewarded by equal pay across the two sectors (for example, having a university education is rewarded less if the degree holder works as a taxi driver than if he works in the financial sector). These differences in rewards explain a large share of wage differences in Lebanon but a smaller share in Egypt. The higher the share of the wage differential that is explained by dif- ferences in “rewards,” the higher the chance that informality is an exclusion phenomenon (because workers would prefer to be rewarded more for a given observable characteristic, such as educational attain- ment, if they could). Last, the black bars represent the share of the difference in wages that arises because of the interaction between how the market rewards characteristics of formal versus informal workers and their differences in endowments. Looking beyond individual characteristics, where workers are employed (small firms or large firms) affects the formality premium. However, the formality premium still persists after controlling for size. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 196 Informality: Choice or Exclusion? Where people work can potentially make an important difference in the earnings that they command. In particular, it is expected that informal wage workers in micro-enterprises (two to five employees) will earn significantly less than workers in larger firms, because smaller firms usu- ally have lower productivity. In turn, firm productivity could account for much, if not all, of the formality premium across workers. Omitting this variable from an analysis of wage differentials could lead one to errone- ously attribute the formality premium to informal employment when it might instead reflect firm characteristics. Once firm size in included in the regressions, the earnings gap between formal and informal workers still persists. The resulting formality premium is now estimated to be between 10 and 50 percent (figure 4.2). However, ample heterogeneity is observed in how firm size affects the estimated formality premium. In Syria, working in micro-firms does not change the formality premium significantly. This could be because of the nature of the survey sample, since survey participants all work in registered firms. Small micro- “for- mal” firms might share some of the key productivity attributes of larger firms so that controlling for firm size might not make a significant dif- ference. In Egypt the formality premium continues to be significant but increases. In Lebanon, up to one-third of the informality premium is attributable to firm characteristics, with workers in micro-firms earning 23 percent less than workers in medium size firms (10 to 49 employees). Unlike in Egypt, informal workers in Lebanon are not primarily concen- trated in micro-enterprises, because a large share of them work in larger firms. In this sense, returns to informal workers’ skills might differ in important respects because of the widely different productivity levels of the firms where they work (ranging from micro-informal to large for- mal). If uncaptured, this difference would be spuriously attributed to workers’ status (formal versus informal) rather than to firms’ character- istics and productivity. Note that, when direct measures of individual ability are explicitly accounted for, the formality premium is virtually unchanged. Quality of education and skills often differ starkly across social strata and are cor- related with the likelihood of working in the informal sector. Assignment to the informal sector is not random, and so it is important to control in some way for individual ability/skills to ensure that the attribution of the formality premium does not instead reflect other factors. In particular, unmeasured individual ability, which is likely negatively correlated with selection into the informal sector, could be spuriously driving these results, because years of education are unlikely to capture well Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 197 actual quality of education and skills. To overcome this limitation in the interpretation of the estimation, the results of direct cognitive and noncognitive tests are included in the regression. These tests were administered in conjunction with the Lebanon labor force survey and the Syria matched employer-employee survey. Both surveys fielded a battery of ability tests to all participants as well as a questionnaire to self-assess personality and relational skills (box 4.1 describes the meth- odology and main features of these tests). The results of these tests are translated into scores and added as additional controls in the wage regressions. Although concerns have been expressed related to possible Box 4.1 Cognitive and Noncognitive Tests Traditionally, studies have focused on measuring skills through educational attain - ment and experience, without measuring ability directly. Cognitive and noncogni- tive tests can be good proxies for individual ability. In fact, an individual’s skill set can be defined as a combination of his or her cognitive, noncognitive, and technical skills (see table below). In turn, given that human capital is multidimensional, both cogni- tive and noncognitive skills can serve as good predictors of economic outcomes such as income and wealth (Bowles and others 2001; Cawley and others 2001; Hansen and others 2004; Heckman and others 2006).Two innovative labor market surveys con- ducted in the MENA region (in Syria and Lebanon) have measured cognitive and noncognitive skills of labor force participants and have produced interesting results. Relationship between technical, cognitive, and noncognitive skills Type of skill What it measures Measurable concepts Cognitive Ability Verbal, math, logic reasoning Noncognitive Ability, interaction Personality, communication Technical Knowledge, experience Breadth, depth of field of knowledge The cognitive test used “Raven’s Progressive Matrices,” a nonverbal test in which individuals have to identify the missing piece of a particular pattern among multiple choices. Respondents are given five minutes to answer as many as they can of the 12 matrices included in the test, which become progressively more difficult and require greater ­cognitive capacity. This test is independent of language, reading, or writing skills and focuses on measuring observation skills, analytical ability, and intellectual capacity. A score (out of 12) is calculated for each respondent based on the number of correct answers completed in the allocated time. box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 198 Informality: Choice or Exclusion? Box 4.1  Cognitive and Noncognitive Tests (continued) The noncognitive test consists of a list of statements describing personal behaviors and characteristics corresponding to the five core dimensions of personality used in psychol- ogy literature to classify human personality (referred to as the “Big Five”): • Conscientiousness: tendency to be organized, responsible, and hardworking • Emotional stability: tendency to be predictable and consistent in emotional ­reactions • Agreeableness: tendency to act in a cooperative and unselfish manner • Extraversion: tendency to orient one’s energies toward the outer world of people and things • Openness to experience: tendency to be open to new cultural or intellectual experiences. ­ These traits are considered to be relatively stable throughout one’s lifetime. The ­ oncognitive test used in the Syria and Lebanon surveys, an adaption of the Goldberg n test, asks respondents to rank the applicability of a total of 15 traits/behaviors to them- selves (three of each corresponding to one of the “Big Five”) from a scale of 1 to 7. Accordingly, a score for each of the “Big Five” traits is calculated. Main observed patterns: Cognitive skills are positively correlated with educational attainment, but more strongly so in Syria than in Lebanon. A positive difference in the average cognitive score between formal and informal workers is observed, which is significant in Lebanon but not in Syria. Cognitive Score Means and Correlations Syria Arab Republic Lebanon Unconditional correlation Unconditional correlation between years of education between years of education Mean and cognitive score Mean and cognitive score Total 4.56 0.41 4.93 0.29 Salaried formal 4.44 0.37 5.54 0.26 Salaried informal 4.86 0.51 4.85 0.23 Self-employed — — 4.53 0.26 Note: In Lebanon, 9.4 percent of the total population surveyed did not take the test. The highest nonresponse rate was for the illiterate (50 percent). Note, however, that only 10 workers in Lebanon reported themselves as illiterate. The results in Lebanon are in line with the “Flynn effect,” which describes an intergenerational increase in scores. box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 199 Box 4.1  Cognitive and Noncognitive Tests (continued) Noncognitive Score Means in the Syrian Arab Republic and Lebanon Syrian Arab Republic Total Informal Formal Conscientiousness 0.89 0.88 0.89 Emotional stability 0.16 0.17 0.16 Agreeableness 0.75 0.80 0.74 Extraversion 0.31 0.29 0.31 Openness to experience 0.52 0.52 0.52 Lebanon Total Informal Formal Conscientiousness 0.89 0.9 0.89 Emotional stability 0.22 0.17 0.23 Agreeableness 0.5 0.55 0.51 Extraversion 0.39 0.35 0.39 Openness to experience 0.49 0.54 0.46 Source: Alloush and others 2011. endogeneity of cognitive and noncognitive score measures, the inclusion of these scores is likely a good proxy for individual ability. Note that when cognitive and noncognitive scores are included in the estimation, they do not affect the estimate of the informality coefficient. Even after taking ability and personal characteristics into account, formality still has a 20 percent premium in Lebanon, and a 12 percent premium in Syria (tables 4.2 and 4.3).7 Unobserved firm characteristics can also affect earnings differentials between formal and informal occupations. In most household surveys, information on the types of establishments where individuals work is not available, with the exception of information on firm size. For example, characteristics such as whether an individual firm is registered with tax authorities or the average level of education of its workforce are not known. These characteristics might matter and drive the ability of firms to reward skills. If unmeasured, they can also confound the size and attri- bution of estimated wage gaps between formal and informal workers. The Syrian data, which allow for these controls, are collected as part of a matched employees-employer survey, so that workers can unequivocally Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 200 Informality: Choice or Exclusion? Table 4.2  Formality Premium, Lebanon: Ordinary Least-Squares (OLS) Estimates for Private Sector Formal and Informal Workers OLS OLS OLS OLS   (1) (2) (3) (4) Dependent Log hourly Log hourly wage Log hourly wage Log hourly wage variable wage Informal (dummy) −0.2920*** −0.1915*** −0.2024*** −0.1953*** Controls Gender, age, Gender, age, age Gender, age, age Gender, age, age age squared, squared, squared, squared, Years of Years of education, Years of education, Years of education, ­ education firm size dummies firm size dummies, firm size dummies, sector sector cognitive score cognitive score, noncognitive score Source: Annex tables 4.1 and 4.2 from Alloush and others 2011. ***Significant at 1 percent. be mapped to their firms and firm-specific effects controlled for. Even after controlling for individual cognitive and noncognitive ability and for firm fixed effects, the formality premium is unaffected; workers who are hired as informal command a lower wage. The wage analysis developed in this section suggests that informal workers command uniformly lower pay than formal workers with similar characteristics. One could argue that, given the overall worse terms of informal jobs (that is, lower duration and limited or no access to leave and pension benefits), informal jobs would have to pay higher wages to attract workers. However, in economies such as those in the MENA region where job creation is rationed, lower pay and worse conditions in the informal sector might be seen by workers as a preferable alternative to unemployment, especially for workers with low skills and education. Job (Im)Mobility Although wage comparisons can be informative, differences in earnings alone cannot unequivocally prove segmentations in labor markets. Earnings differences between the formal and informal sector might reflect a range of unobservable job characteristics, including flexible hours, value of training, and value (or lack) of social security benefits, all of which can affect the salary wedge between observationally similar jobs (Maloney 1999, 2004). Moreover, selection into formal or informal jobs Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Table 4.3  Formality Premium, Syrian Arab Republic: Ordinary Least-Squares (OLS) Estimates for Private Sector Formal and Informal Workers OLS OLS OLS OLS (1) (2) (3) (4) Firm fixed effects Dependent variable Log hourly wage Log hourly wage Log hourly wage Log hourly wage Log hourly wage Informal (dummy) −0.109** −0.123*** −0.123*** −0.119*** −0.127*** Controls Gender, age, age Gender, age, age Gender, age, age Gender, age, age Gender, age, age squared, squared, squared, squared, squared, Years of education Years of education, Years of education, Years of education, Years of education, firm size dummies firm size dummies, firm size dummies, cognitive score, cognitive score cognitive score, noncognitive score noncognitive score Source: Annex tables 4.1 and 4.2 from Alloush and others 2011. **Significant at 5 percent; ***significant at 1 percent. 201   202 Informality: Choice or Exclusion? might be driven by unobservable workers’ skills and ability, which might in turn affect earnings. The concern that this might lead to estimation bias is mitigated in wage regressions for Lebanon and Syria, due to inclusion of direct test score measures of cognitive and noncognitive ability. These controls did not affect the size of the formality premium, hence suggest- ing that “informality” status did not, to a large extent, spuriously reflect ability. However, other unobservables might still persist. To complete and complement this analysis, patterns of labor mobility in and out of infor- mality and wage differentials for individuals moving from informal to formal jobs and vice versa were examined using actual transitions from the Egypt panels as well as data on self-reported jobs search. Workers’ mobility patterns in MENA provide evidence that segmenta- tions exist in the region’s labor markets. Mobility patterns can provide important insights on the extent to which labor markets are segmented or integrated. A good level of mobility and integration in labor markets indicates that workers and employers have the chance to find productive matches between labor demand and supply of skills. In this section, lon- gitudinal data from two separate sources for Egypt are used to study mobility patterns between the formal and informal sectors: the ELMPS, which surveyed individuals in 1998 and 2006, and the HIECS panel, which surveyed individuals in 2008 and 2009. Over the eight-year span of the two ELMPS surveys, persistence in formal or informal status was high (figure 4.4): 46 percent of informal salaried workers in 1998 contin- ued to be informal salaried in 2006. By 2006, only 19 percent had tran- sitioned to formal employment, either in the private or in the public sector. In the second panel survey, 65 percent of informal salaried workers in 2008 remained so in 2009. Upward mobility was minimal, with only 9 percent of informal salaried workers transitioning toward jobs with social security coverage (public or private). Only a minority of all informal workers in 2008 became unemployed in 2009 despite the financial crisis. It is interesting to note that when the financial crisis struck Egypt in 2008, formal private sector employ- ment shrank substantively (35 percent) with a large share of workers transitioning to informal salaried jobs in the private sector (about 28 percent). In a crisis year, many of these transitions are likely involun- tary and might have been the result of termination of temporary ­ contracts (initially with benefits) and their subsequent conversion into informal jobs, as well as of firm closures after which workers rap- idly found informal subsistence employment. It is notable that transi- tions into unemployment from formal jobs are almost nonexistent. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 203 Lacking effective unemployment insurance, ­ informal work (which com- mands average lower wages) represents a safety net type of employ- ment.8 Table 4.4 depicts the 1998–2006 transition matrix, and table 4.5 depicts the 2008–2009 transition matrix. The yearly transition probabil- ities in 2008 translate into an estimated job duration of about three years for informal salaried workers,9 above that observed in comparator countries in other regions, such as Mexico (where movement in and out of salaried informal jobs is high and average job duration is estimated at about two years).10 Bosch and Maloney (2010) offer a sophisticated approach to reading transition matrices, which adjusts for the extent to which vacancies open up and workers leave in different sectors (for example, formal private, public, or private informal).11 Applying their Figure 4.4  Distribution of Labor Market Transitions in Egypt transitions among those originally informal transitions among those originally salaried private sector, male informal salaried private sector, female 100 100 90 90 80 80 70 70 60 60 percent percent 50 50 40 40 30 30 20 20 10 10 0 0 1998–2006 2008–2009 1998–2006 2008–2009 unpaid self-employed self-employed public public informal salaried informal formal salaried private salaried private private formal salaried unemployed private unemployed out of labor force out of labor force box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 204 Informality: Choice or Exclusion? Figure 4.4  Distribution of Labor Market Transitions in Egypt (continued) transitions among those originally out of transitions among those originally out of the the labor force, male labor force, female 100 100 90 90 80 80 70 70 60 60 percent percent 50 50 40 40 30 30 20 20 10 10 0 0 1998 2006 2008 2009 1998 2006 2008 2009 unpaid self-employed unpaid self-employed public informal salaried public informal salaried private private formal salaried unemployed formal salaried unemployed private private out of labor force out of labor force Source: Silva and others 2010. methodology to the Egypt context confirms that the public sector continued to be an engine of employment growth in the crisis, second ­ only to informal salaried work, absorbing 28 percent of formal private sector workers. Overall, persistence in informal status (including, in addition to informal wage workers, the self-employed, many employers, and unpaid workers) is extremely high, at about 90 percent. Panel data also allow for observing earning changes along transitions. On average, when a worker moves from an informal to a formal job, his or her net earnings increase by about 24 percent and even more if bene- fits are accounted for (table 4.6). Note that by tracking earnings changes for the same individual across the informal-formal transition, the concern that higher earnings in formal jobs might reflect unobserved workers’ characteristics is eliminated. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Table 4.4 Labor Market Transitions in Egypt (1998–2006) Formal Informal Out of labor Unemployed salaried salaried private Public Self-employed Unpaid force 2006 2006 private 2006 2006 2006 2006 work 2006 Total Out of labor force 1998 45% 10% 4% 21% 7% 6% 8% 100% No. of observations 1,283 279 121 602 207 168 216 2,876 Unemployed 1998 8 9 9 37 11 23 3 100 No. of observations 26 30 31 122 37 77 10 333 Formal salaried private 1998 5 1 55 14 10 15 0 100 No. of observations 13 4 149 37 28 41 1 273 Informal salaried private 1998 6 2 11 46 8 25 1 100 No. of observations 53 20 98 414 76 225 9 895 Public 1998 11 0 2 2 81 3 0 100 No. of observations 150 6 29 27 1057 40 1 1310 Self-employed 1998 5 2 3 7 6 78 1 100 No. of observations 30 10 18 44 39 511 6 658 Unpaid work 1998 7 1 5 19 10 32 27 100 No. of observations 16 2 11 45 23 78 66 241 Total 24 5 7 20 22 17 5 100 No. of observations 1,571 351 457 1,291 1,467 1,140 309 6,586 Source: Silva and others 2011. Note: Percentages are reported in bold. The shaded area highlights transition among employed workers. Sample: Male workers, 15–64 years old in 2006. 205   206 Table 4.5 Labor Market Transitions in Egypt (2008–2009) Formal Informal Out of labor Unemployed salaried salaried Self-employed Unpaid work force 2009 2009 private 2009 private 2009 Public 2009 2009 2009 Total Out of labor force 2008 73% 5% 2% 10% 2% 3% 5% 100% No. of observations 760 48 16 109 24 32 47 1,036 Unemployed 2008 16 31 3 29 13 3 5 100 No. of observations 26 49 5 45 20 4 9 158 Formal salaried private 2 2 32 28 28 7 0 100 2008 No. of observations 8 7 105 92 94 25 0 332 Informal salaried private 4 3 4 65 5 14 4 100 2008 No. of observations 41 28 40 607 51 132 40 940 Public 2008 2 0 2 3 90 2 0 100 No. of observations 21 2 25 25 886 22 1 982 Self-employed 2008 3 0 2 14 3 76 2 100 No. of observations 29 3 20 129 30 713 17 941 Unpaid work 2008 14 1 1 23 6 7 49 100 No. of observations 39 3 2 62 16 19 135 276 Total 20 3 5 23 24 20 5 100 No. of observations 924 140 214 1,071 1,121 948 249 4,666 Source: Silva and others 2011. Note: Percentages are reported in bold. The shaded area highlights transition among employed workers. Sample: male workers, 15–64 years old. Informality: Choice or Exclusion? 207 Table 4.6  Average Level of Earnings Increase When Moving from Formality to Informality percent Mover formal Mover Stayer Mover Stayer salaried private Stayer formal I-F I-I F-I FF to public salaried private Earnings (mean % change) (1) 24 2 −9 2 7 2 Source: Silva and others 2010 using Egypt HIECS 2008/09. Note: Line (1) presents the mean percentage change in real earnings between 2008 and 2009. For this calcula- tion, nonadjusted earnings (take-home pay), which exclude social security benefits, were used. Sample: Male, 15–64 years old, employed in 2008 and 2009. Patterns of social security contribution densities from administra- tive data confirm the limited mobility between formal and informal jobs. Social security contributions offer an alternative data source to understand job duration patterns. Contribution density is defined as the share of time an individual contributes to the pension system dur- ing his or her active life period. Individual contribution densities are the longitudinal companion of the coverage rate, which measures the share of employed who are contributing to social security at a certain moment in time. Figure 4.5 illustrates the age-specific contribution density patterns toward the mandatory social insurance scheme administered by the Social Security Corporation (SSC) in Jordan. The SSC dataset used for the analysis includes everyone who contributed to or received benefits from the SSC between 1980 and 2010. The pattern indicates that in Jordan, individuals tend to complete either long contributory spells or long noncontributory spells (that is, they stay in or out of the system for long periods of time). As illustrated by the figure, adult individuals between ages 25 and 55 are expected to stay in the system for about 30 years (which is generally the time needed for full retirement). Across all individuals and cohorts, contri- bution densities are low for youth, but they increase rapidly as indi- viduals reach adulthood, leave school, and establish themselves in formal employment. As expected, contribution rates drop significantly when individuals are in their late fifties, because early and statutory retirement rules drive down contributions (a pattern that is observed in most countries).12 Contribution density patterns show the coexistence of two different groups of workers: some with an established contributory status, and Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 208 Informality: Choice or Exclusion? Figure 4.5 Contribution Density and the Distribution of Contributory and ­Noncontributory Spells in Jordan average densities of contribution: total 40 30 density 20 10 0 20 30 40 50 60 70 age Source: Calculations using administrative datasets from the SSC of Jordan. some who hardly ever contribute. The average length of a continuous contributory spell in Jordan is longer than four years, with little differ- ence between women and men (49 and 53 months, respectively), and with 40 percent of all contributors contributing for at least three years (figure 4.6). Only 16 percent of all contributors (mainly new entrants to the labor market) have a contributory period shorter than six months. The distribution of the length of noncontributory spells signals the exis- tence of a group consistently out of coverage. For example, once workers who are enrolled in a contributory job (that is, registered with SSC) at some point in their career start a noncontributory spell, 69 percent of them stay out of coverage for a minimum of three years and an average of approximately six years (94 months). Changes in contributory status in Jordan are significantly less fre- quent than in a comparator country such as Chile. The third and fourth panels of figure 4.6 illustrate the transition patterns in and out of con- tributory status based on monthly administrative data on social insur- ance contributions from Chile. In the case of Chile, the most frequent length category for both being in or out of social insurance con­tributory status is less than six months: 42 percent of those contributing to the social insurance system in Chile are likely to discontinue contributions Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 209 within six months, and 37 percent of those registered with social security will restart contributions within six months. In contrast, the ­ Jordanian contributory pattern is extremely skewed: 40 percent of contributors are likely to complete a spell longer than three years, and Figure 4.6  Distribution of the Length of Social Insurance Contributory and ­Noncontributory Spells in Jordan and Chile Jordan—Contributory 70 60 50 40 percent 30 20 10 – …<6 6 < … < 12 12 < … < 24 24 < … < 36 36 < … months Jordan—Noncontributory 70 60 50 40 percent 30 20 10 – …<6 6 < … < 12 12 < … < 24 24 < … < 36 36 < … months box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 210 Informality: Choice or Exclusion? Figure 4.6  Distribution of the Length of Social Insurance Contributory and ­Noncontributory Spells in Jordan and Chile (continued) 70 Chile—Contributory 60 50 percent 40 30 20 10 – …<6 6 < … < 12 12 < … < 24 24 < … < 36 36 < … months Chile—Noncontributory 70 60 50 40 percent 30 20 10 – …<6 6 < … < 12 12 < … < 24 24 < … < 36 36 < … months Source: Administrative datasets from the SSC of Jordan and the Chilean Pension Supervisory Agency. 69 percent of those once registered as contributors, but currently not contributing, will be outside of the system longer than three years. These results confirm the existence of an insider/outsider structure in MENA labor markets that is notably different from other parts of the world.13 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 211 What Are Informal Jobs Like? This section aims to provide as assessment about the quality of informal employment. To begin with, available evidence indicates higher levels of job dissatisfaction among informal workers.14 Table 4.7 reports responses to the question “Would you like to change your job?” for Lebanon, Egypt, and Syria. In Lebanon and Egypt, informal workers were signifi- cantly more likely to want to change jobs or were searching for a job. A higher level of job dissatisfaction persists in Egypt and Lebanon even after many individual characteristics are accounted for, including age, education, and pay (with the expectation that workers with higher pay are less likely to want to change jobs). In Syria, where differences between formal and informal workers are less pronounced in the sur- veyed sample, formal workers actually appear to be marginally more willing to change jobs than informal workers, but this difference is insig- nificant when individual characteristics are accounted for (Alloush and others 2011). In general, informal jobs entail worse working conditions. Table 4.8 presents descriptive statistics related to self-perception of employment quality for formal and informal workers in Lebanon and Syria. Results indicate that, in many dimensions, formal workers experience better job quality, as measured by higher earnings, having a written contract, longer employment tenure, and higher chances of choosing when to work over- time (and be paid for such work). A noticeable feature is that informal jobs have a significantly shorter tenure than formal private sector jobs, which suggest that informal jobs provide firms with the needed flexibility in hiring and firing (especially relevant in countries where labor regula- tions are particularly restrictive).15 Little difference is seen in the amount of work time (hours per day and days per week) between formal and informal workers in Lebanon and Syria. However, access to annual or other leave is much lower for informal workers (as discussed below), so Table 4.7  Individuals Who Want to Change Jobs percent Egypt, Arab Rep., youth Lebanon Syrian Republic Arab Formal workers 11.6 21.8 37.1 Informal workers 36.9 45.5 34.4 N 3,180 757 961 Sources: Egypt youth survey 2010, Lebanon Labor Force Survey 2011, Syria Employee-Employer matched module 2011. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 212 Informality: Choice or Exclusion? Table 4.8  Differences in Job Quality between Formal and Informal Jobs Lebanon Syrian Arab Republic Informal Formal Informal Formal Are you currently looking for another job? (% yes) 20.6% 8.1% 22.7% 20.2% Hours (weekly) 48.42 46.96 50.80 48.73 Days of work per week 5.78 5.66 5.96 5.83 Earnings per hour ( pounds) 4.55 6.85 72.36 89.48 Earnings per hour (bonuses taken into account) 4.57 6.9 89.92 110.77 Length at job (years) 5.12 7.81 4.35 8.03 Implicit annual growth rate of earnings per hour 9.8% 8.4% 9.3% 10.8% Can request overtime 27.2% 38.4% — — Can refuse overtime 33.4% 47.4% — — Paid overtime 36.9% 49.3% — — Written contract 18.4% 64.1% 21.7% 68.2% Would like to be formal (registered in NSSF) 67.9% — 68.6% — Prefer to be self-employed 63.6% 57.1% 63.5% 65.2% Confidence in stability of job 83.4% 65.8% 77.6% 84.0% % full time 84.2% 95.2% — — Secondary job 5.6% 5.9% 7.7% 5.6% Secondary source of income 4.7% 4.2% 16.9% 16.5% Financial need to work 76.0% 73.5% — — Personal satisfaction 24.0% 26.5% — — Source: Alloush and others 2011. — = Not available. that they may end up working a significantly greater number of days per year than formal workers. As expected, informal workers enjoy fewer benefits than formal work- ers. In Lebanon, 56 percent of informal workers do not have any annual leave versus 13 percent of formal workers. Similarly, informal workers have significantly less access to benefits of family, sick, or maternity leave (figure 4.7). Although no significant difference is seen between formal and informal workers in terms of their access to benefits, such as provi- sion of a firm nursery or a lunch allowance, large differences emerge for benefits that are fairly common in the formal sector. The evidence from Egypt suggests that informal jobs require, on average, shorter commutes, indicating that they are more likely to be “local.” Still, transport expenses (which are not factored in the reported wage) can be substantial. For example, in Lebanon, daily transportation costs are anecdotally reported to be somewhat higher (at about 0.8–1.2 times the average hourly wage). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 4.7 Lack of Access to Job Benefits between Informal and Formal Sectors (% of Workers without Access) Lebanon Syrian Arab Republic annual leave 56% 35% annual leave official leave 29% 10% official leave emergency 46% leave/childcare 40% emergency leave/childcare maternity leave 50% 2% maternity leave sick leave 31% 36% sick leave transport allowances 70% 52% transport allowances firm nursery 87% 8% firm nursery flexible working hours 50% 44% flexible working hours 84% 76% lunch allowances lunch allowances 84% 38% family allowances family allowances informal formal Source: Alloush and others 2011. 213   214 Informality: Choice or Exclusion? Furthermore, informal workers are more likely to be working in unsafe conditions. Data from the Morocco youth survey provide additional information on job safety and dignity at the job place for formal and informal workers (table 4.9). Results indicate that informal workers are systematically worse off: more likely to be physically or verbally abused, more likely work longer period of time for less pay, more likely to be harassed at the workplace, and more likely to feel they are exposed to exhausting workloads. Finally, informal workers are much more likely to feel that their job is not challenging and/or not interesting, which ­ supports the view that informal jobs might entail waste of talents. Informal workers appear to be less likely to receive training than formal workers, even after controlling for individual characteristics. Evidence in chapter 3 showed that firms with a higher share of infor- mal workers are less likely to provide training. In the Syria sample, and after controlling for individual characteristics such as education and cognitive and noncognitive scores, informal workers are found to be significantly less likely to receive training. However, this association loses significance once firm fixed effects are included, indicating that the correlation identified cross-sectionally might all be due to differ- ences in firms’ types. Among Moroccan workers who did not report benefiting from technical training, informal workers were significantly more likely to report that no training is accessible to them, and that they have little information on what training to take. Moreover, they were almost twice as likely as formal workers to report that they could not afford the fees. Table 4.9  Job Conditions among Moroccan Youth percent Total Formal Informal Verbal abuse from clients or colleagues 18 14.3 18.9 Physical abuse 2 0.0 2.5 Long hours many days in a row 35 25.6 36.5 Too little pay 51 33.6 54.1 Exhausting work load 39 21.5 42.7 Harassment from clients or colleagues 3 2.7 3.5 Hazardous workplace 14 8.7 14.5 Harassment during commute 4 3.9 4.1 Unchallenging/very boring 29 12.0 32.2 Working without definite information about pay 8 5.9 8.8 Source: Morocco Household and Youth Survey 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 215 Self-Employment: A Choice? For some self-employed, informality may be a matter of choice. As men- tioned in chapter 2, self-employed workers are somewhat different from other informal workers (mainly wage earners). Results from Egypt and Lebanon indicate that self-employed workers are mainly male (72 percent in Egypt and 85 percent in Lebanon), belong predominantly to the age group 35–54 (about 50 percent in both countries), and have attained at most a secondary education (93 percent in Egypt and 87 percent in Lebanon). On average, self-employed workers are older and less educated than informal wage earners (see chapter 2). Nevertheless, as illustrated in figure 4.8, self-employed workers in Lebanon generally earn, on average, 32 percent higher wages than informal wage earners (although they still earn lower wages than formal wage earners in net terms). This holds true all along the wage distribution. Indeed, a small Figure 4.8 Net Wages by Employment Status (Lebanon, 2010) average wage rate (L£) 8 6.85 7 6.01 6 5 4.55 4 3 2 1 0 formal wage informal wage self-employed earners earners Lebanon, 2010 1.0 minimum wage 0.8 0.6 0.4 0.2 0 0 1 2 3 4 log hourly wage formal salaried informal salaried self-employed Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 216 Informality: Choice or Exclusion? pool of self-employed is seen at the top end of the wage distribution (see figure 4.8) earning even higher wages than formal wage earners. These results would be consistent with self-employed workers prefer- ring their employment status versus that of informal wage earners and even versus formal wage earners (that is, those at the very high end of the wage ­ distribution). Self-employed workers enjoy being independent, although they would like to be covered by a social security scheme. Table 4.10 provides infor- mation on self-stated preferences and perceptions of being self-employed versus being a wage earner. Results indicate that 70 to 75 percent of all workers would like to be wage earners mainly for job security reasons (that is, having social security and greater employment security). Note that only 7 to 15 percent of all workers interviewed claimed that better opportunities and/or greater job satisfaction would be a reason for want- ing to be a wage earner. At the same time, the majority of workers inter- viewed claimed that greater independence and higher earnings are the main two reasons for wanting to be self-employed. Although enjoying the benefits of independence and higher earnings, the majority of self- employed workers (53 percent) would like to be registered in a social security scheme. Table 4.10 Preferences for Wage Earning versus Self-Employment (Lebanon, 2010) percent Informal wage Formal wage earners earners Self-employed Reasons for wanting to be a wage earner Greater job security 64.5 57.7 52.3 Access to social security 10.3 13.8 24.4 Better prospects 4.7 9.0 4.7 Greater professional satisfaction 4.7 5.8 2.3 Reasons for wanting to be self-employed Greater independence 57.1 53.6 63.2 Higher income level 19.9 21.7 20.4 More flexible hours 5.8 4.9 6.5 Better prospects 6.3 8.0 4.2 Greater professional satisfaction 6.8 8.4 2.3 Would like to be formal (registered in NSSF) (% yes) 67.9 53.5 Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 217 How Much Are Workers Willing to Contribute to Social Security? By and large, informal workers would like to have access to social secu- rity benefits and are willing to pay for them. In Lebanon and Syria, 70 percent of informal workers indicate they would like to have social security benefits and are willing to give up on average 5.8 percent of their salary. This compares to a contribution rate of 2 percent for NSSF, which provides health insurance coverage with an additional 9 percent paid by the employer. Of the 30 percent of informal salaried workers who are not interested in being registered in NSSF, 37 percent say that having private health insurance from a different source is the main rea- son for their lack of interest (table 4.11). Although only limited infer- ences can be made because of the small sample size (these questions were asked only of informal workers who did not want to contribute to social security), it should be noted that both the poor perceived quality of the public service and lack of information are reported as leading reasons for opting out of the system. Similarly in Morocco, lack of infor- mation and how the system works are the top reasons for not wanting social security (table 4.11). Although no strong correlations emerge, generally males and the better educated are more likely to know how social security works. It is interesting that a high share of noncontribut- ing workers in Syria report that they are not concerned about retirement. Table 4.11 Reasons Why Informal Workers Do Not Want to Contribute to Social Security percent Syrian Arab Morocco   Lebanon Republic (youth) Informal workers who do not want to con- 28 30 tribute Reasons for not wanting to contribute   Private health insurance 37 — 0.9   Reduction of earnings too much 22 39 15.3   Low quality of service 39 42 6.7   Do not know how it works 19 55 50.9   Not concerned about retirement — 56 —   Employer does not want to pay — — 26.9   Does not think current job gives right to   — — 66.7   social security Source: Salaried workers (Lebanon Labor Force Survey 2011, Syria 2010, Morocco Household and Youth Survey 2010). Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 218 Informality: Choice or Exclusion? It is not clear whether this is attributable to myopia or to the existence of strong informal social safety nets that would substitute for retirement savings. Also in Syria, half of the noncontributing respondents report that “not knowing how it works” is an important reason for not contrib- uting. This suggests that improving financial literacy and communication about pension systems’ roles and rules might be an important compo- nent of effective risk protection. Networks and Intergenerational Persistence Networks constitute an important mechanism for workers to find formal and informal employment. The large majority of workers in the MENA region report having found jobs through personal connections. This is particularly the case for informal workers. For example, 85 percent and 74 percent of informal workers in Lebanon and Syria, respectively, report finding jobs through personal contacts (table 4.12). Overall, this evidence points at a very limited role for private and public institutional mecha- nisms to support broad matching of skills. Employers seem to fulfill their labor demand through mechanisms that bypass market signals of skills or quality of education, either because these signals might not be informative or because other factors, such as trust, are dominant and unlikely to be picked up by market matching mechanisms. To the extent that personal contacts and networks develop within, rather than across, specific social Table 4.12  How People Find Jobs percent Lebanon Formal Informal Personal contacts 77.3 84.5 Other 10.4 6.3 Advertisement 7.3 6.6 Online job search 4.2 1.7 Recruitment agency 0.66 0.99 National employment agency 0.22 0.0 Syrian Arab Republic Personal connections 62.3 74.1 Direct contact with company 21.5 16.3 Advertisement (newspaper) 6.1 6.6 Other 4.6 2.2 National employment agency 3.8 0.2 Private employment agency 1 0.5 Online advertisement 0.6 0.16 Source: Lebanon Labor Force Survey 2010 and Syria 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 219 and economic strata, these job search methods are likely to perpetuate segmentations within labor markets, particularly across formal and infor- mal jobs, thus preventing the best matches and highest returns to skills from being realized. Evidence from Lebanon and Syria stresses the importance of family relationships for finding informal employment. Table 4.13 presents the self-reported relationship to the firm owner for employed individuals in Lebanon and Syria. Results indicate that informal workers tend to be related to the firm owner (that is, be family or friend) more often than are formal workers. Among informal workers, this phenomenon is par- ticularly present among family-owned companies and almost absent in large enterprises. The opposite is observed among formal workers, for whom the importance of networks does not change substantially across firm size. Consistent with these findings, in Lebanon and Egypt, infor- mal jobs are less likely to require significant commutes because net- works are generally built within the community of residence (for example, in Egypt, youth are less likely to have a lengthy commute if working informally). Networks seem to be an important determinant for finding “formal” employment. Regression analysis indicates that in Morocco, having a father with a formal job significantly increases one’s chances of having a formal job, even after controlling for own and parental education, as well as for whether one’s father speaks French (used here as a proxy for social status; table 4.14). Moreover, the probability of being informal decreases significantly if another member of the household holds a formal job. In Morocco, the formality definition captures both pension and health insurance coverage, because the benefits are bundled. The literature on informality often makes the assumption that informality is higher because, with bundled and family coverage, incentives to participate in Table 4.13 Relationship to Owner percent Lebanon Syrian Arab Republic Relationship to firm owner Formal Informal Formal Informal Member of household 3.6 7.3 1.4 2.8 Relative outside household 1.8 5.9 2.1 6.3 Friend/neighbor 7.8 16.3 8.9 13.4 From the same town 4.4 10.6 0.8 0.3 Unrelated 82.4 59.8 86.7 77.3 Source: Lebanon Labor force survey 2010 and Syria 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 220 Table 4.14  Determinants of Informality: Networks in Morocco OLS OLS OLS OLS   (1) (2) (3) (4) Dependent variable Informal Informal Informal Informal Father has formal job −0.124*** −0.118*** −0.109*** −0.108*** Controls Gender, age, age squared, Gender, age, age squared, Gender, age, age squared, Gender, age, age squared, Education, province, urban Education, province, urban, Education, province, urban, Education, province, urban, father education, father father education, father father education, father French speaking French speaking, wealth French speaking, wealth, cell phone ownership Source: Gatti and others 2011. ***Significant at 1 percent. Informality: Choice or Exclusion? 221 the system for family members of a formal worker are very limited. However, the evidence for Morocco points at the opposite: Formal jobs appear as a privilege of the few, and family connections are successfully leveraged to access them. Conclusions Results indicate that the average worker in the formal sector earns higher wages compared to the average worker in the informal sector. Controlling for other factors such as education, experience, and gender, workers in the formal sector earn higher wages than otherwise similar workers in the informal sector. This “formality premium” is generally high, ranging from 10 percent in Syria to up to 53 percent among youth in Morocco. The formality premium is higher among female workers in all cases. On aver- age, a female worker in the formal sector earns 1.5 times as much as an otherwise similar worker in the informal sector. Very low wages paid to female workers in the informal sector often explain why many women decide not to participate in the labor market at all, especially if the oppor- tunities to find formal jobs are limited. The observed differences in wages between the informal and formal can be attributed to differences in the characteristics of formal and infor- mal workers (endowments), as well as to differences in how the labor market rewards those characteristics (returns). About one-third of the wage differential in countries such as Egypt and Lebanon can be explained by the fact that, on average, formal workers are older and more educated than informal workers. In Lebanon, the wage differential is explained mostly by the labor market treating similar skills of formal and informal workers differently. Where workers are employed (small versus large firms) and their skills and ability can both potentially affect the observed formality premium, but the formality premium is found to persist even when these additional factors are accounted for. A priori there might be important differences in workers’ wages across small and large firms because of differences in firms’ productivity. Omitting this factor could be problematic in estimating the formality premium, especially when informal workers for the most part find employment in micro-firms. When firm size is controlled for, workers in micro-firms in Lebanon earn 23 percent less than workers in medium size firms (10 to 49 employees) and about one-third of the formality premium is attrib- utable to firm characteristics. However, a formality premium of about Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 222 Informality: Choice or Exclusion? 20 percent still persists. In Egypt and Syria, the formality premium is virtually unaffected by inclusion of controls for where workers are employed. Although firm size might be an imperfect measure of firm productivity, even using firm fixed-effects estimation (Syria) confirms that informal workers earn less than formal workers, independent of where they work. Individual ability does not seem to be an additional important deter- minant of the formality premium. Quality of education and skills often differ starkly across social strata and are correlated with the likelihood of working in the informal sector. Assignment to the informal sector is not random, and so it may be important to control for individual ability and skills to ensure that the attribution of the formality premium does not instead reflect other factors. Data from Lebanon and Syria allow proxying for individual ability through results of direct cognitive and noncognitive tests. When cognitive and noncognitive scores are included in the estima- tion (and controlling for other factors), the observed informality pre- mium remains largely unchanged. These results suggest that informal workers are generally worse-off, because individual ability is better rewarded by formal jobs than by informal jobs. Limited mobility between formal and informal employment points at the existence of segmentations in the region’s labor markets. The fact that measured rewards (wages) to skills are consistently different across formal and informal workers cannot be taken alone as an indicator of labor market segmentation, because many other features of jobs are usually unmeasured. Mobility patterns provide important insights into the extent to which labor markets are segmented or integrated. In this chapter longitudinal data from two separate sources for Egypt are used to study mobility patterns between the formal and informal sectors. Results indicate that between 1998 and 2006, persistence in formal or informal employment status was very high and upward mobility was minimal, with only 9 percent of informal salaried workers transitioning into formal jobs. Transition from informal to formal jobs was associated with an average increase in earnings of 24 percent, confirming the hypothesis that formal workers are generally better off. Beyond earnings, workers in the informal sector are generally disadvan- taged in many other dimensions. Informal workers are more likely to be unsatisfied with their job relative to formal workers. This is explained by the fact that informal jobs entail worse working conditions. Results in this chapter indicate that, in many dimensions, formal workers display better job quality indicators, such as higher earnings, having a written contract, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 223 longer employment tenure, and higher chances to choose when to work overtime (and be paid for such work). Furthermore, informal workers enjoy fewer benefits than formal workers, such as paid leave and mater- nity leave. On the positive side, evidence from Egypt suggests that infor- mal jobs require, on average, shorter commutes, which indicates that they are more likely to be “local.” Finally, results in this chapter indicate that individual networks are important to explain the likelihood that individuals find formal and infor- mal employment. The large majority of workers in the MENA region report having found jobs through personal connections. This is particularly the case for informal workers. For example, 85 percent and 74 percent of informal workers in Lebanon and Syria, respectively, report finding jobs through personal contacts. Informal workers tend to be somewhat more related to the firm owner (that is, a family or friend) than formal workers. Networks are also important for individuals to find formal employment. Evidence from Morocco indicates that, controlling for other factors, hav- ing a father with a formal job significantly increases one’s chances of hav- ing a formal job. In Morocco, the evidence also suggests that, everything else equal, workers are more likely to find a formal jobs if someone else in the family already has a formal job. In countries where pension and health insurance are bundled and where spouses and dependents can rely on the worker’s coverage, this evidence suggests that formality is likely driven by the use of network to secure the “good jobs.” Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 224 Annex Annex table 4A.1 Estimating the Formality Premium in Lebanon (1) (3) (4) (5) Variables ln(earnings/hr) ln(earnings/hr) ln(earnings/hr) ln(earnings/hr) Informal −0.2920*** −0.1915*** −0.2024*** −0.1953*** Years of education 0.0325*** 0.0264*** 0.0247*** 0.0260*** Male 0.0982*** 0.0926** 0.0859** 0.0645 Age 0.0133 0.0117 0.0147 0.0161 Age2 −0.0001 −0.0001 −0.0001 −0.0001 Firm size <5 (base 10–49) −0.2312*** −0.2195*** −0.2157*** Firm size 5–9 −0.0374 −0.0629 −0.0668 Firm size >50 0.0165 −0.0065 −0.0117 Cognitive score quintile 2 0.0523 0.0549 Cognitive score quintile 3 0.1325** 0.1268** Cognitive score quintile 4 0.1446** 0.1615*** Cognitive score quintile 5 0.1347** 0.1514** Open to experience −0.0451 Conscientious −0.0072 Extravert 0.0044 Agreeable −0.1303*** Emotionally stable −0.0607 Constant 1.1544*** 0.9748*** 0.8845*** 0.9322*** Observations 582 582 533 519 R2 0.2526 0.3379 0.3430 0.3769 Source : Alloush and others 2011. Note: Estimation with OLS, robust standard errors. ***p<.01, **p<.05, *p<.1. Annex table 4A.2 Estimating the Formality Premium in Syria (1) (2) (3) (4) Firm fixed effects Variables Log hourly wage Log hourly wage Log hourly wage Log hourly wage Log hourly wage Informal −0.109** −0.123*** −0.123*** −0.119*** −0.127*** Years of education 0.0436*** 0.0420*** 0.0349*** 0.0334*** 0.0302*** Male 0.133*** 0.136*** 0.132*** 0.122*** 0.156*** Age 0.0459*** 0.0458*** 0.0472*** 0.0465*** 0.0415*** Age2 −0.000477*** −0.000465*** −0.000468*** −0.000470*** −0.000384** Firm size 1–5 0.0470 0.0314 0.000480 Firm size 6–10 −0.0740 −0.0555 −0.0460 Firm size >50 −0.0738 −0.0722 −0.0683 Cognitive score quintile 0.00837 −0.00731 −0.00632 2 Cognitive score quintile 0.0668 0.0397 0.0197 3 Cognitive score quintile 0.112* 0.0903 0.0522 4 Cognitive score quintile 0.258*** 0.229*** 0.126 5 Open to experience 0.0882*** 0.0991*** Conscientious −0.00841 0.0113 Extravert −0.0324 −0.00719 Agreeable −0.168*** −0.135*** Emotionally stable −0.0361 0.0152 Source: Alloush and others 2011. Note: Estimation with OLS, robust standard errors. ***p<.01, **p<.05, *p<.1. 225   226 Annex table 4A.3 Estimating the Formality Premium in Egypt Informality defined by not having Informality defined by not having access to Dependent variable = log access to social security social security and not having a work contract hourly wage Total Male Female Total Male Female Informal dummy −0.262*** −0.192*** −0.583*** −0.145*** −0.116*** −0.515*** [0.031] [0.033] [0.065] [0.032] [0.033] [0.083] Middle school 0.196*** 0.163*** 0.413*** 0.208*** 0.170*** 0.402*** [0.042] [0.045] [0.127] [0.042] [0.045] [0.123] High school 0.324*** 0.311*** 0.478*** 0.342*** 0.322*** 0.490*** [0.026] [0.027] [0.084] [0.026] [0.027] [0.087] Tertiary 0.714*** 0.743*** 0.826*** 0.740*** 0.759*** 0.843*** [0.033] [0.037] [0.088] [0.033] [0.037] [0.093] Experience 0.037*** 0.031*** 0.040*** 0.039*** 0.031*** 0.046*** [0.003] [0.003] [0.005] [0.003] [0.003] [0.005] exp2 −0.000*** −0.000*** −0.000*** −0.000*** −0.000*** −0.000*** [0.000] [0.000] [0.000] [0.000] [0.000] [0.000] Alexandria −0.068** −0.018 −0.189*** −0.060* −0.015 −0.147** [0.034] [0.041] [0.059] [0.034] [0.041] [0.059] Urban upper −0.107*** −0.124*** −0.157*** −0.113*** −0.131*** −0.142*** [0.030] [0.037] [0.051] [0.031] [0.037] [0.052] Urban lower −0.140*** −0.122*** −0.260*** −0.143*** −0.125*** −0.245*** [0.032] [0.039] [0.054] [0.032] [0.039] [0.054] Rural upper −0.112*** −0.151*** −0.163* −0.126*** −0.163*** −0.172* [0.033] [0.037] [0.097] [0.033] [0.037] [0.104] Rural lower −0.191*** −0.192*** −0.300*** −0.196*** −0.196*** −0.300*** [0.030] [0.035] [0.063] [0.030] [0.035] [0.064] Secondary sector 0.267*** 0.222*** 0.167** 0.267*** 0.221*** 0.174** [0.025] [0.026] [0.073] [0.025] [0.026] [0.072] Tertiary sector 0.244*** 0.222*** 0.211*** 0.253*** 0.228*** 0.239*** [0.027] [0.030] [0.062] [0.028] [0.030] [0.064] table continues next page Annex table 4A.3  Estimating the Formality Premium in Egypt (continued) Informality defined by not having Informality defined by not having access to Dependent variable = log access to social security social security and not having a work contract hourly wage Total Male Female Total Male Female Public sector dummy −0.113*** −0.088** −0.007 −0.047 −0.045 0.056 [0.032] [0.035] [0.066] [0.032] [0.035] [0.069] Married dummy 0.090*** 0.079*** 0.084* 0.103*** 0.091*** 0.100** [0.025] [0.030] [0.044] [0.025] [0.030] [0.045] Constant 0.151*** 0.263*** −0.139 0.024 0.188*** −0.331*** [0.055] [0.062] [0.123] [0.054] [0.061] [0.123] Observations 5,144 3,960 1,184 5,144 3,960 1,184 R2 0.249 0.234 0.423 0.24 0.228 0.404 Formality premium 23.0 17.5 44.2 13.5 11.0 40.2 Source: Angel-Urdinola and Tanabe 2011, using the 2006 Egypt ELMPs. Note: Omitted variables: Education: Primary; Region: Cairo; Sector of employment: Primary sector; Primary sector (agriculture); Secondary sector (manufacturing and construction); Tertiary sector (wholesale, transport, services); Public administration and social services (including education and health). 227   228 Informality: Choice or Exclusion? Notes 1. To appropriately control for hours worked, the analysis is based on hourly wages in current local currency (that is, wage rates). 2. Formal workers are required to pay taxes, and so some high-earning workers have an incentive to evade taxes and earn higher net wages, especially if they have access to health insurance through a family member and/or to a pension. 3. Because of the likely presence of omitted variables, the estimate of this coef- ficient is likely biased. In particular, if (unobservable) ability is positively cor- related with formality, this coefficient is likely biased upward. However, when direct measures of individuals’ cognitive and noncognitive ability are intro- duced for Lebanon and Syria, the size of the coefficient remains unchanged. 4. Herrera and Badr (2011) estimate a formality premium for Egypt of 13.2 percent using a different definition of informality (workers not having a contract and not contributing to social security). Using the same definition, very similar results are found in this analysis and are available upon request. 5. Controls are year of education, age, age squared, and a gender dummy. 6. See, for example, Arias and Khamis (2007) and Pianto and others (2009) for approaches that correct for selectivity bias. 7. Two main sources of concern are found when it comes to endogeneity. First, ability influences education outcomes, which in turn affect measured ability. This would tend to overestimate the coefficients on skills and underestimate that of schooling. Second, reverse causality may exist between labor outcomes and socioemotional skills. Personal characteristics can influence labor market outcomes, but labor market outcomes can also influence the way one feels about oneself. For example, having a low paying job can lower one’s self- esteem and emotional stability (Heckman and others 2006). 8. Note that the two rounds of surveys are likely to capture the early crisis impact because they were fielded in April 2008 and March 2009. Unfortunately, because of the lack of information on type of contracts (temporary vs. perma- nent), it is not possible to further disentangle mobility across formal/informal states. 9. Following Maloney (1999), job duration can be calculated as 1/(1−Pii) where Pii is the individual probability of not leaving sector/state i. See Silva and others (2011) for a detailed discussion. ­ 10. See Maloney (1999). 11. Bosch and Maloney (2010) compute an adjusted propensity to transit from sector i to sector j (C-stats). This statistic weighs the raw transition probabilities by the rates of leaving the origin sector and the relative opening in the receiving sector compared to the economy as a whole. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Informality: Choice or Exclusion? 229 12. Note that contribution density here, in contrast to the coverage rate, is not defined in relation to the labor force, but rather to the population registered with the social insurance administration. This population includes (at least temporarily) inactive people who exited the labor force after a period of past contributions; therefore the contribution density defined in this manner is naturally smaller than the coverage rate for the identical population. 13. The Chile and Jordan datasets differ somewhat, which may confound the interpretation of these results. For example, the underlying Jordanian admin- istrative data cover a period of 363 months, whereas the Chilean data cover only 288 months; that is, it might be more likely to observe long (in excess of 36 months) contributory and noncontributory spells in Jordan than in Chile because of data differences. However, it is unlikely that the radical ­ difference in the shape of the distributions can be explained solely by this sample difference. 14. In this chapter “job quality” is referred to in a way that is akin to the “decent work” definition. Decent work involves opportunities for work that is pro- ductive and delivers a fair income, security in the workplace and social pro- tection for families, better prospects for personal development and social integration, freedom for people to express their concerns, organize and par- ticipate in the decisions that affect their lives, and equality of opportunity and treatment for all women and men. The private sector and innovation literature usually refers to job quality as “high value-added jobs.” The discus- sion of which policy interventions can promote the development of high value-added production (for example, moving from textiles to electronics) is complementary to the treatment of informality herein. However, it is beyond the scope of this report. 15. An interesting exception is in Lebanon, where informal workers report having higher confidence in the stability of their jobs than do formal workers. References Alloush, M., C. Chartouni, R. Gatti, and J. Silva. 2011. “Informality and Exclusion: Evidence from Lebanon and Syria.” Mimeo. Washington, DC: World Bank. Angel-Urdinola, D., and K. Tanabe. 2011. “Micro-determinants of Informal Employment in the Middle East and North Africa Region.” Mimeo. Washington, DC: World Bank. Arias, O., and M. Khamis. 2007. “Comparative Advantage and Informal Employment.” Mimeo. Washington, DC: World Bank. Bosch, M., and W. Maloney. 2010. “Comparative Analysis of Labor Market Dynamics Using Markov Processes: An Application to Informality.” Labour Economics 17: 621–631. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 230 Informality: Choice or Exclusion? Bowles, S., H. Gintis, and M. Osborne. 2001. “The Determinants of Earnings: A Behavioral Approach.” Journal of Economic Literature 39 (4): 1137–1176. Cawley, J., J .J. Heckman, and E. J. Vytlacil. 2001. “Three Observations on Wages and Measured Cognitive Ability. Labour Economics 8 (4): 419–442. Gatti, R., M. Ivanic, and J. Silva. 2011. “It’s Who You Know: Networks and the Chance of Getting a Formal Job.” Mimeo. Washington, DC: World Bank. Hansen, K. T., J. J. Heckman, and K. J. Mullen. 2004. “The Effect of Schooling and Ability on Achievement Test Scores.” Journal of Econometrics 121 (1–2): 39–98. Heckman, J. J., J. Stixrud, and S. Urzua. 2006. “The Effects of Cognitive and Non- cognitive Abilities on Labor Market Outcomes and Social Behavior.” Working Paper 12006. Cambridge, MA: National Bureau of Economic Research. Herrera, S., and K. Badr. 2011. “Why Does the Productivity of Education Vary across Individuals in Egypt? Firm Size, Gender and Access to Technology as Sources of Heterogeneity in Returns to Education.” Mimeo. Washington, DC: Word Bank. Maloney, W. F. 1999. “Does Informality Imply Segmentation in Urban Labor Markets? Evidence from Sectoral Transitions in Mexico.” World Bank Economic Review 13 (2): 275–302. ———. 2004. “Informality Revisited.” World Development 32 (7): 1159–1178. Pianto, D. M., M. Tannuri-Pianto, and O. Arias. 2009. “Informal Employment in Bolivia: A Lost Proposition?” Mimeo. Washington, DC: World Bank. Silva, J., M. Benyagoub, and G. Wallner. 2010. “Informality, Firm Performance and Labor Market Outcomes: Evidence from Matched Employer-Employee Data for Morocco.” Mimeo. Washington, DC: World Bank. Silva, J., Y. Dukhan, and D. Marotta. 2011. “Determinants and Earnings Consequences of Mobility In and Out of Informality in Egypt.” Mimeo. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 c ha p t e r 5 Barriers to Coverage and Policy Options SUMMARY: Widespread informality implies that large portions of the workforce are not protected against old age, disability, and, often, health risks. ­ Although the need to address these vulnerabilities provides a clear rationale for government intervention, the actual drivers of informality should inform policy levers and choices. Many institutional constraints determine the labor market segmentation that underpins informality in MENA, such as the design of pension systems, business and labor regulation, incentives and pay in the public sector, and the design of interventions to improve skills upgrad- ing. This chapter analyzes these institutional constraints and presents a related set of policy interventions to effectively expand coverage and promote the creation of better quality jobs. First, the chapter discusses the importance of a healthy business environment that fosters competition and facilitates firm entry. Easing certain provisions of the labor legislation and keeping the cost of labor at a realistic level can help employment creation and reduce informality, ­especially if coupled with reforms geared toward protection of workers’ transitions. Realigning the pay and benefit package that is offered by public sector employment can reduce important distortions. Moreover, the low productivity dimension of informality, a phenomenon that is particularly relevant among the poorest countries in the region, calls for productivity- enhancing interventions, including those that aim to improve access and realign training and skill-upgrading programs to the needs of the informal sector. Second, the c­ hapter discusses how reforms in the design of the social insurance system in MENA are critical for addressing informality. Limited legal coverage, the short minimum vesting requirements, generous early retire- ment provisions, and the use of an average wage measure from the final years Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 231   232 Barriers to Coverage and Policy Options before retirement for pension benefit calculation all ­ contribute to limited coverage. Addressing the coverage gap will require ­ governments to look beyond reforming the existing social insurance system to seek special coverage extension schemes targeted to the informal sector and to those with limited savings capacity. This chapter is organized in two parts and addresses five distinct and complementary policy angles, each linked to the drivers of informality in the MENA region. Part 1 discusses institutional barriers to formality in MENA, including (1) the business environment, (2) labor market regula- tions, and (3) public sector employment bias, and provides relevant policy options. Part 1 also discusses the need for (4) addressing the productivity trap faced by informal workers, particularly through training and skills- upgrading interventions. Part 2 of this chapter focuses on (5) the social insurance system and presents barriers to coverage linked to the incentive design of pension systems in MENA. It presents policy options and recommendations for coverage extension through both improving the ­ design of existing social insurance systems and introducing new special schemes targeting informal workers. Part 1. Addressing Institutional Barriers and the Low Productivity Trap Introduction to Part 1 Although growth materialized in the last decade in MENA, not enough good jobs were created. Instead, informal employment grew, character- ized by lower pay, lack of access to training, worse working conditions, and lower job tenure than formal employment. These conditions affect the vast majority of workers, whereas relatively few “insiders,” including those employed in the public sector, benefit from privileged circum- stances. Reforms aimed at decreasing informality can potentially affect two aspects of this process: (1) by generating direct productivity gains and increasing job creation (increasing the “size of the pie”) and (2) by improving how equitably rents and benefits from the development process are redistributed. In this sense, many of the policy options dis- ­ cussed here can have an impact both on employment creation as well as on formalization and moving toward fulfilling work for all. Because of the multidimensionality of informality, it is important to acknowledge that a complex set of policy interventions might be needed to effectively overcome barriers to formality in a sustainable manner and help the Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 233 growth process to become progressively more inclusive. Here, too, the mechanism through which formalization is achieved matters greatly for ­ its effects on employment, efficiency, and growth. If formalization is based purely on enforcement, it will likely lead to unemployment and low growth. If, on the other hand, it is based on improvements in both the regulatory framework and the quality and availability of public ser- vices, it is likely to bring about more efficient use of resources and higher growth. Certain features of the regulatory framework in MENA, encom- passing the business environment, set of labor regulations, and nature of employment in the public sector, pose barriers to formality. Informality appears to be strictly intertwined with the development process and can be largely explained as a suboptimal private sector response to restrictive regulation. The first part of this chapter describes institutional barriers to formality and provides relevant pol- icy options that extend beyond mere enforcement. In particular, this section addresses the need for a conducive business environment, the importance of less restrictive labor regulations, and, at the same time, more effective protection of workers’ income d ­uring employment transitions. Moreover, the first part of this chapter argues that the preference for public sector jobs in many MENA ­ countries affects informality outcomes, requiring a realignment of incentives to limit queuing for these jobs. Finally, recognizing that many informal workers face a low-­ productivity trap because of their limited access to relevant skill acquisition, numerous skills-upgrading interventions are explored for specific MENA context. Improving the Business Climate Having to obey more stringent regulations may imply lower flexibility in employment and production decisions, and therefore, lower profits firms’ ­ and productivity (Almeida and Carneiro 2005). Excessive entry regulations and high taxes matter for informality and growth. Across countries, a significant correlation is found between the size of the informal sector and the ease of doing business (figure 5.1). Barriers to firm entry give discretion to public officials to exclude or advantage specific investors (World Bank 2009) and thus continue to perpetuate a dual model of development in which a few “protected” firms thrive and share rents while many small firms strive to survive. These barriers work as an impediment to growth, especially for the most productive among these outsiders, who might be excluded from Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 234 Barriers to Coverage and Policy Options Figure 5.1 Correlation across Countries between the Size of the Informal ­Sector and the Ease of Doing Business correlation: –0.69*** 60 CMR N = 74 % of informality in labor force IDN MDG DOMCOL VNM BOL PRY ETH 40 BGD PAKHND ZMB PER PHL NIC GTM JAM IRN VEN ECU SYR THA LKA MNG SLV DZA PANGRC BLZ MEX EGY BRA MAR KOR NAMURY CHL CRI ARG TUN FJI ITAPRT 20 TTO CYPMYS TWN MUS ZAF NZL ESP IRL CAN MLT BRB BEL CHE SGP ISL GBRAUS ISR FRA OMN JPNNLD DEUAUT HKG DNK USA NOR LSO 0 3 4 5 6 7 8 business regulatory freedom (index: higher, less regulated) Source: Loayza and others 2005. Note: *** = 1% significance level. important opportunities (or alternatively, who might have to divert resources from productive activities to rent-seeking activities). In paral- lel, corporate taxes are also systematically identified as a constraint to business and formalization. These constraints are particularly powerful in MENA. As highlighted in chapter 3, many firms in MENA never formalize, and even those that eventually register still operate infor- mally for a significant amount of time. The region has the developing world’s highest share of firms that start out as informal (one-fourth) and the longest operating period before formalization (four years) among registered forms. The following discussion focuses on two main margins along which business environment reform can promote a more inclusive and dynamic private sector: (1) regulation of entry and (2) corporate tax reform. Could reforms in the regulation of entry improve firms’ incentives to formalize? Regulation is understood to be an important determinant of formalization. First, monetary and administrative registration costs (for example, a high number of procedures requiring extensive time Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 235 and effort) increase formalization costs. After LAC, MENA is the region with the highest average number of procedures to start a business. Economies such as Algeria and Djibouti and the West Bank and Gaza economy stand out as places where the registration process is particu- larly cumbersome, requiring more than 10 procedures and an average of between 20 to 80 days to start a business (see chapter 3). Second, discretion in the application of these procedures (for example, with connected firms potentially benefiting from less strict enforcement) increases barriers to formalization for those firms that, quality-wise, would have the potential to compete on a larger scale and fully within the regulatory framework. Thus, uneven enforcement sustains an equi- librium where few connected firms thrive, a large number of firms operate informally, and new entrants face the choice of either investing in rent seeking to secure the benevolent eye of the bureaucracy or being virtuous and thus bearing the disproportionate brunt of taxation and regulation. Reforms to the regulation of entry have been shown to have had positive, albeit moderate, effects on formalization. Interventions include (1) reducing the costs of registration, number of procedures, and minimum capital; (2) providing information and training to entre- preneurs (such as filling out forms); and (3) facilitating registration by establishing one-stop shops for registration. No experimental evidence exists in MENA on the likely impact of these reforms. In Mexico experiments were conducted and the impact evaluated (Bruhn 2011; Kaplan and ­ others 2011). Broadly, these studies found that simplifying the process of business registration had a moderate impact on formal- ization of existing firms. In contrast, they indicate that the interven- tion increased f ­ormalization because of more creation of new businesses by former wage earners (Bruhn 2008). Another potential reason for the limited impacts on formalization of existing firms is that results might take time because of uncertainty about reform reversal, which is not captured by these analyses. Finally, the effect of these policies tends to depend on how many firms are at the margin of for- malizing. Given the differences between MENA and LAC in this domain, the effects of this type of intervention could be significantly larger in the MENA region. High taxation burden was the constraint to formalization most widely identified by micro- and small firms in Egypt and Morocco (figure 5.2). High taxes imply high costs of regulatory compliance if a formal business. Morocco’s corporate tax rate is one of the highest among developing Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 236 Barriers to Coverage and Policy Options Figure 5.2 Obstacles to Formalization in Egypt and Morocco strict labor law time necessary lack of information cost of registration level of other charges level of adminstrative charges minimum capital level of taxes 0 10 20 30 40 50 60 % firms identifying constraint as major or very severe Egypt, Arab Rep. Morocco Source: Calculations using informal ICA surveys from Morocco 2008 and Egypt 2010. countries: In 2007, it was second only to Pakistan and remained signifi- cantly above the average for developing countries in 2008. It is interesting that in Egypt (where corporate tax rates are below the developing countries average), a significantly smaller share of firms than in Morocco ­ identified tax rates as being a major obstacle to formalization. The ­ country’s profit taxes are also high relative to countries with similar income levels.1 Similar results were also found in countries such as Mexico, Brazil, and Bolivia. Lowering the corporate tax rate can affect tax revenue through three main channels: (1) existing formal firms may invest more and earn more income on which they pay taxes, (2) existing informal firms may be induced to ­ formalize and start paying some taxes, and (3) new firms might be induced to operate formally. Evidence from other regions suggests that the net effect is likely to depend on whether a reduction in the tax rate is accompanied by additional enforcement and a reduction in exceptions. The short-run impact may also be negative; for example, Turkey lowered its corporate tax rate from 30 to 20 percent in 2007 and experienced a drop in overall tax revenues (Otonglo and Trumbic 2008).2 However, the converse happened in Egypt: When its corporate tax rate was lowered from 42 to 20 percent, it was accompanied by a significant increase in overall tax revenues. An experimental study in Brazil evaluated the impact of a reform that combines business tax Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 237 ­ eduction (of up to 8 percent among eligible firms) and regulation simpli- r fication and found interesting results. The emerging evidence indicates that the reform led to a significant increase in formality along several dimensions (Fajnzylber and Reyes 2010). In particular, the reform con- sisted of implementing a new simplified tax system for micro- and small firms, referred to as “SIMPLES.” The new national system consolidated several federal taxes and social security contributions into one single monthly payment, varying from 3 to 5 percent of gross revenues for micro-enterprises, and from 5.4 to 7 percent of revenues for small firms. Program eligibility excluded some sectors. This intervention suggests that this type of program can increase levels of registration and government revenues. Enforcement matters too. Overall, and not surprisingly, more frequent inspections are associated with lower underreporting of workers and sales (Almeida and Carneiro 2009). In MENA, firms report an ­ average of four tax inspections per year (the highest regional average in the world). Such strict enforcement appears to be accompanied by wide- spread corruption, because informal payments were requested in 17 ­percent of inspections, significantly above the 7 percent reported in the LAC and ECA regions. In the MENA region, investment climate reforms have accelerated in many countries in recent years. The evidence suggests that the imple- mentation of reform matters greatly to private sector development. The recent MENA Development Report From Privilege to Competition (World Bank 2009) estimates that in response to previous reforms, private investment in the MENA region increased by only 2 percent of GDP, compared with between 5 and 10 percent in Asia, Eastern Europe, and Latin America. The same report estimates that the number of registered businesses per 1,000 people in MENA is less than one- third that in Eastern Europe and Central Asia, and with less entry and average business is 10 years older than in East Asia or exits of firms, the ­ Eastern Europe. Close to 60 percent of business managers surveyed did not think that the rules and regulations were applied consistently and predictably, whereas policy uncertainty, unfair competition, and cor- ruption were identified as major concerns for investors. Discretionary enforcement of regulation is a strong deterrent to small entrepreneurs who start their businesses informally but are then forced to stay small to escape controls. Staying small may, in turn, make it prohibitively costly to formalize over time.3 Overall, the paucity of existing data on firms’ dynamics has not yet allowed identification of which margins matter most to promote formalization in the ­ context of MENA. This is Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 238 Barriers to Coverage and Policy Options evidence can effectively inform an important area where experimental ­ policy making. Addressing Constraints in Labor Market Regulation Labor policies can affect informality through three main channels. First, excessive labor costs, whether due to labor regulation (such as high minimum wages, severance costs, or labor taxes) or strong worker bar- gaining power, can depress labor demand in the formal sector. Second, legislation can create incentives for workers to voluntarily work infor- mally if perceived contributions exceed the benefits. Third, labor market institutions can impact productivity growth. Productivity gains arising from adoption of new technologies and production processes account for half of the differences in levels of economic development (the most important determinants of informality), not to mention long-run worker productivity and welfare more generally (World Bank 2007). Yet excessive restrictions on job reallocation or layoffs for economic reasons, or state- or union-induced inflexibilities, may reduce the adop- tion of such innovations. Across countries, there is debate whether overly rigid employment protection legislation (EPL) is an important determinant of the labor ­ market segmentations underlying informality. EPL is the set of rules governing the hiring and firing process that is provided through both labor legislation and collective bargaining agreements (box 5.1). Strong evidence suggests that overly rigid EPL tends to not only discourage hiring and firing but also may slow down adjustment to shocks, impede ­ the reallocation of labor, and promote informality (OECD 2010). Recently, Fialová and Schneider (2010) established a statistically signifi- cant effect of EPL on the size of the informal sector in European Union (EU) countries. In the countries that have the most rigid EPL, the share of the informal sector is estimated to be about 3.5 percentage points higher than in countries that have the most flexible EPL. This result ­ supports the view that unduly strict EPL leads some employers to hire workers informally to avoid costs imposed by the EPL. Specifically, strict EPL typically makes it harder for certain groups, including youth, women, and displaced older workers, to enter or reenter the labor market, at least on an open-ended contract. ­ Perceptions of employers. The extent to which labor regulations are perceived by employers as a constraint to expanding their formal ­ employment varies among MENA countries but in general is higher Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 239 Box 5.1 Employment Protection Legislation (EPL) EPL, a key state intervention in the labor market, aims to protect workers from arbitrary, unfair, or discriminatory actions by their employers, while addressing ­ potential market failures stemming from insufficient information and inadequate insurance against risk. As such, EPL governs the individual employment contract, including flexibility of hiring through part-time and fixed-term contracts, and conditions of employment, including maximum number of hours in a work week, ­ premiums for overtime work, paid annual leave, and minimum wage. It also governs flexibility of firing, including grounds for dismissal, notification rules for dismissal, priority rules for dismissal, and severance pay. The literature on EPL describes positive and negative effects on labor market performance. Among the former, it highlights the benefits of long-term employee- ­ employer contracts, including greater willingness to invest in on-the-job training. Among the latter is the concern that workers hired on regular contracts may enjoy a high degree of employment security to the detriment of other workers hired on tem- porary contracts or without formal contract and coverage. In addition, employment protection may ­ diminish firms’ ability to cope with a rapidly changing environment driven by globalization, technological change, and the derived organizational inno- vation (OECD 2004). Source: Angel-Urdinola and Kuddo 2010. Table 5.1 Employment Regulations and Skills and Education of Labor as a Major or Severe Obstacle for Expanding Employment percent of firms Region Employment regulations Skills and education of labor Eastern Europe and Central Asia 18.7 43.2 Africa 17.3 32.6 East Asia and Pacific 14.5 33.8 Latin America and the Caribbean 29.5 42.5 South Asia 19.0 24.8 Middle East North Africa 36.2 54.4 Source: BEEPS 2008 and Enterprise Surveys. than in other regions. According to ICA surveys, labor regulations and mandatory contributions are considered by firms as a factor that con- strain many ­enterprises from expanding formal employment. Table 5.1 shows the percent of firms indicating employment regulations and skills and education of labor are a major or severe obstacle. The MENA Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 240 Barriers to Coverage and Policy Options region has the highest share of employers dissatisfied with the existing labor regulations, although it should be noted that the skills and educa- tion of the labor force are a more significant problem for all regions than ­ employment regulations. In Egypt, Lebanon, and Syria, labor regulations are perceived by firms as a major constraint to expanding formal employment, although this is true to a lesser extent in Algeria, Jordan, Morocco, and the West Bank and Gaza economy (figure 5.3). Manufacturing firms, service firms, and hotels in Egypt report that they would hire a net of 21 percent, 9 percent, and 15 percent more work- ers, respectively, if there were fewer restrictions on hiring and firing workers (Angel-Urdinola and Kuddo 2010). Similarly, according to enterprise surveys, firms in Lebanon would be willing to hire more workers (by an average of more than one-third of the workforce) in the absence of existing restrictions on labor regulation. The results of the enterprise survey analysis support those of previous studies (Pierre and Scarpetta 2006), showing that firms in countries with more stringent employment regulations are more likely to report labor regulations as a major or very severe obstacle, even after controlling for other factors such as GDP and unemployment. Overall, EPL as an obstacle to busi- ness growth tends to be more pronounced in countries that are more Figure 5.3 Share of Firms Identifying Labor Regulations as a Major Constraint in Doing Business in MENA (%) Syrian Arab Republic 2009 50 Lebanon 2009 36 Egypt, Arab Rep. 2008 27 Morocco 2007 16 Algeria 2007 14 Jordan 2006 14 West Bank and Gaza 12 economy 2006 Yemen, Rep. 2010 9 0 10 20 30 40 50 60 Source: World Bank: www.enterprisesurveys.org. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 241 likely to enforce it (including through court challenges) and less of a problem in countries where the capacity of labor market institutions is weaker. In most MENA countries, labor regulation is a key mechanism for protecting workers’ rights, because collective bargaining is not wide- spread. Trade unions in MENA rarely represent many workers effectively (although exceptions are found, as in Tunisia, where unions are influen- tial social partners). Moreover, workers have limited ways of challenging private employers; for instance, in many countries in the region, strikes remain illegal. Thus, labor regulations have an important role to play in protecting workers (Angel-Urdinola and Kuddo 2010). Below, key aspects of EPL that can affect informality are explored, including (1) hiring regulations and contract types, (2) minimum wage, (3) firing regulations, and (4) tax wedges. Hiring regulations and contract types. Hiring regulations in MENA are generally aligned with international standards, and MENA countries are joining the international trend in increasing the prevalence of fixed-term work contracts. In general, MENA countries do not have strict hiring regulations compared with international standards. In reforming labor legislation in the region, most attention is paid to relevant arrangements associated with fixed-term employment, which in the past was deemed to be an exceptional form of employment, conditioned by the nature of work or other objective conditions. In recent years, fixed-term work has been increasing not only in EU15 countries4 but also in MENA (table 5.2).5 Fixed-term contracts are heavily concentrated among young people and other new labor market entrants, such as the formerly unem- ployed and those with lower education levels, that is, among people who have weaker bargaining power. For these workers, fixed-term work can provide a bridge to formal employment and an opportunity to gain ­ experience and skills. Among MENA countries, Morocco has the most restrictive laws: Fixed-term contracts are prohibited for permanent tasks, the duration is limited to 12 months, and renewal is prohibited. At the other end of the spectrum, no restrictions or limits are placed on the use of fixed-term contracts in Bahrain, Egypt, Jordan, Kuwait, Oman, and the Republic of Yemen. Fixed-term employment contributes to more flexible labor markets. It provides a buffer for cyclical fluctuations in demand, allowing compa- nies to adjust employment levels without incurring high firing costs. Fixed-term work also allows companies to reap market opportunities by Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 242 Barriers to Coverage and Policy Options Table 5.2  Arrangements for Fixed-Term Contracts around the World in 2010 number of countries Fixed-term Maximum cumulative duration of a contracts are fixed-term employment relationship prohibited for including all renewals (months) permanent Region Total tasks No limit 12 24 36 48 60+ East Asia and Pacific 24 5 20 1 1 1 0 1 Europe and Central Asia 25 15 10 1 2 4 0 8 Latin America and the ­Caribbean 32 16 22 2 6 0 0 2 Middle East and North Africa 18 4 12 1 3 0 1 1 High income: OECD 30 9 13 1 10 1 1 4 South Asia 8 3 6 1 1 0 0 0 Sub-Saharan Africa 46 20 23 3 8 2 7 3 Total 183 72 106 10 31 8 9 19 Source: Doing Business 2011 database. e ­ ngaging in projects of short duration without bearing disproportionate personnel costs. This is especially important in labor markets where per- manent employment is protected by strict regulations and high firing costs. To counterbalance the latter, some countries have established a minimum service length of one to three years with the same employer for a worker to be eligible to claim severance pay, making short fixed- term contracts less attractive. Although temporary jobs can be useful for promoting employment opportunities, they can also lead to undesirable labor market outcomes and informality. From the firm’s perspective, temporary jobs can pro- vide a “screening” device, allowing the firm to evaluate workers’ ability or adequacy for the job. They can also act as a buffer, facilitating a firm’s adjustment to temporary demand shocks, thereby avoiding costly adjustments to its “core” labor force (European Commission 2010). Conversely, temporary contracts can simply be a convenient way for firms to reduce labor costs. From the workers’ perspective, fixed-term jobs are subject to higher turnover and pay lower wages on average. Estimates show that in the EU, temporary workers earn on average 14 percent less than workers on open-ended contracts after controlling for a number of personal characteristics. Temporary work- ers also tend to have reduced access to ­ training provided or subsidized by firms. Labor market reforms are met with resistance by the s ­ egments Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 243 of the society benefiting from the status quo, and so it is likely that fixed-term contracts will become more and more common. From the perspective of improving coverage, more ­ temporary work contracts are desirable if they provide access to basic social risk management tools to workers. Minimum wage and wage rigidities. Minimum wages affect informal- ity through at least two channels. First, if the minimum wage is set higher than what employers are willing to pay for an unskilled worker, the ­latter’s employment is likely to be undeclared. Second, minimum wage policy can reduce tax evasion where underreporting is a problem (World Bank 2008) (see box 5.2 for a brief overview of minimum wages). Although the evidence of the impact of statutory minimum wages on informality is limited, a large body of empirical literature, albeit inconclusive, exists on the impact of statutory minimum wages on worker flows, particularly in the United States. In the United States, although early studies tended to find a negative impact of minimum wages on job retention for individuals at, or close to, the minimum wage, more recent studies have generally found no significant impact (Abowd and others 2005; Zavodny 2000). Draca and others (2008) found that the introduction of a minimum wage in the United Kingdom in 1999 led to insignificant changes in firm entry and exit patterns (OECD 2010). Evidence from other countries is limited. Abowd and others (2005) found no impact of real minimum wages on entry into employment in France, but a strong positive impact on exit from employment. By contrast, Portugal and Cardoso (2006), exploring a specific Portuguese reform that in 1987 dramatically lifted minimum wages for very young workers, found that raising minimum wages had a significant negative effect on both separations and hirings. The effect of introducing a higher minimum wage appears to be large and negative in Colombia and small or negligible in Costa Rica and Mexico. In Brazil, evidence was found of a positive effect of an increase of minimum wage on employment; however, this was mainly the result of changes in the composition between hours worked and number of jobs (Maloney and Medez 2003). Less is known about the impact of minimum wages on informal employment, but some findings show that a rise in the minimum wage has a positive impact on wages in the informal sector, through what is known as the “lighthouse” effect—as workers in the informal sector use the Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 244 Barriers to Coverage and Policy Options Box 5.2 Minimum Wages Setting minimum wages is a common practice in many countries: Both OECD and developing countries set minimum wages with the intended objectives to promote a fair wage structure, to provide minimum acceptable standards of living for low-paid workers, and to eventually alleviate poverty. Minimum wage legislation exists in more than 90 percent of countries, with “universal” minimum wage for both public and pri- vate sectors the most dominant, although not the only, practice.a Minimum wage regulations have many dimensions: (1) the level set, (2) coverage, (3) differentiation in the level (such as by age, sector, or region), (4) methods of adjusting levels to reflect inflation, (5) how the level is set (for example, by the gov- ernment or by the social partners), (6) whether the level applies to the private and/or public sector, and (7) sanctions for noncompliance. International evidence demonstrates that if the minimum wage is set at a moderate level, then it is not likely to entail substantial employment losses. At the same time, minimum wages tend to have only a limited and often transitory impact on earnings of low-wage workers. Overall, this suggests that statutory minimum wages have at best second-order impacts on labor reallocation (OECD 2010). If minimum wages (in relation to the average wage) are set too high, they can be counterproductive. Higher minimum wages can have a non-negligible adverse impact on employment in low- wage sectors. Minimum wage as a policy tool to improve the living standards of low-paid workers has clear advantages. First, depending on its level, the minimum wage could ­ be a less distortionary policy tool compared with alternative mechanisms that include changes in tax policy (reduction in income taxes for low-skilled workers, non- distortive negative income tax, or implicit subsidies). Second, a minimum wage may be easier to implement and enforce than a change in taxes (Smits 2008). Neverthe - less, minimum wage should be viewed as only one option in a menu of policy instru- ments available to governments to affect income distribution, poverty, and employ- ment levels of low-income earners. Source: Angel-Urdinola and Kuddo 2010. a. An overview of the minimum wages in about 100 countries can be found in the ILO online database: http:// www.ilo.org/travaildatabase/servlet/minimumwages. ­ inimum wage as a reference for their own wages. Although minimum m wages are not legally binding in the informal sector, they still seem to influ- ence informal sector wage distribution. From the labor supply side, the minimum may be a benchmark for “fair” wages. On the demand side, employers may pay a wage comparable to the formal sector market wage Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 245 for a particular occupation so that employees will not leave for a similar job in the formal sector, or employers may not be willing to provide all legislated labor benefits but at least will pay the minimum wage. In par- ticular, Lemos (2004) found adverse effects of higher minimum wages on employment in both the formal and informal sectors in Brazil. Based on data from Costa Rica, a country with a complex minimum wage policy, Terrel and Gindling (2002) found that employers responded to a mini- mum wage increase by increasing the hours of full-time workers and decreasing them for part-time workers, who, in turn, switched to self- employed work in the informal sector. The subsequent increase in supply of workers in the informal sector then placed downward pressure on wages in the informal sector. About half of all MENA countries do not have a legal minimum wage; those that do set them with considerable variation. Djibouti and the West Bank and Gaza economy are examples of MENA members that have no minimum wage in practice (Angel-Urdinola and Kuddo 2010). In countries with minimum wages, settings vary, complicating cross-country comparisons. For example, minimum wages are set at a monthly rate in Egypt, Jordan, Lebanon, and Tunisia, whereas Morocco has a minimum hourly wage. Figure 5.4 presents the ratio of minimum wages to average value added per worker in selected countries. In countries that have mini- mum wages in some form, the ratio of minimum wage to average value added per worker varies from 1.80 in Zimbabwe and 1.17 in Mozambique to 0.05 in Burundi and Gabon. In the reviewed MENA countries, the ratio varies from 0.11 in Egypt to 0.72 in Morocco. A high minimum wage can be damaging in some low-paid sectors and regions with below-average wages; it is also typically more damaging for small and medium enterprises (SMEs) because these enterprises tend to be more labor intensive and financially weaker. This likely contributes to keeping many SMEs smaller than they might otherwise be and gives them an incentive to remain infor- mal. In Egypt, despite the relatively low minimum wage, a significant number of workers earn below the minimum wage, which suggests low levels of enforcement. However, the wage distribution for both formal and informal workers in Egypt seems quite centered (and compressed) close to the minimum wage; this suggests that the minimum wage may serve as a benchmark wage for new entrants in the labor market (figure 5.5). Overall, minimum wages do not appear to be binding in most MENA countries. In most MENA countries, minimum wages are rather low, and sanctions for noncompliance with minimum wage rules are weakly enforced. Independent of how high or low the minimum wage is relative Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 246 Barriers to Coverage and Policy Options Figure 5.4 Ratio of Minimum Wages to the Average Value Added per Worker in ­Selected Countries in 2010 Morocco 0.72 Jordan 0.4 Lebanon 0.32 Tunisia 0.27 Egypt, Arab Rep. 0.11 Italy 0.36 United Kingdom 0.35 Greece 0.29 Spain 0.27 Portugal 0.26 Australia 0.24 United States 0.21 Germany 0.21 France 0.14 South Africa 0.7 Turkey 0.47 Argentina 0.45 China 0.38 Brazil 0.28 India 0.16 Russian Federation 0.14 Mexico 0.11 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 Source: Doing Business 2011 database. Note: Because of a lack of consistent cross-country data on average earnings, the average gross national income ­ opulation per capita is used as a proxy for average earnings. This ratio is adjusted to represent the percentage of p of working age as a share of the total population. to average wages, the extent to which minimum wage policy affects employment outcomes and the wage distribution depends on its enforce- ment. Although most MENA countries with defined legal minimum wage have regulations on enforcement, enforcement is rather weak, inspections are rare due to a lack of resources, and fines are rarely imposed. A fairly significant mass of workers (those to the left of the vertical bar in figure 5.5) report wages below the minimum wage, which indicates that the minimum wage is unlikely to be a significant constraint to formal employment in most MENA countries. Centralized wage setting can, however, be an important determinant of informality. In ­ Tunisia, in parallel to the general minimum wage, employer and employee representatives negotiate a pay scale based on professional levels in each Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 5.5  Hourly Wage Distribution and Minimum Wages in Egypt, the Republic of Yemen, and Morocco (kernel density plots of monthly wages) Egypt, Arab Rep., 2006 Yemen, Rep., 2006 Morocco (youth), 2010 0.8 1 0.8 0.6 0.8 0.6 density density density 0.6 0.4 0.4 0.4 0.2 0.2 0.2 0 0 0 2 4 6 8 10 12 4 6 8 10 12 1 2 3 4 5 log monthly wages log monthly wages log hourly wage (Morocco 2010) kernel = Epanechnikov, bandwidth = 0.0974 formal informal Source: Calculations using Egypt’s 2006 Labor Market Panel Survey. Note: The vertical line illustrates the level at which the minimum wage is set; only wages in urban areas are c ­ onsidered. 247   248 Barriers to Coverage and Policy Options sector; the differences are significant in some sectors. In reality, a relatively high minimum wage for university graduates is ­ institutionalized, which likely contributes to the high graduate unemployment. The bargaining process is such that on the employee side, wages are monopolistically negotiated by unions whose members are all employed, thus possibly resulting in artificially high wages. This especially affects first-time job seekers. Those who cannot afford to wait for a formal job at or above the minimum wage will tend to accept lower wages in informal jobs.6 Firing arrangements. The strictness of firing regulations and the associ- ated cost can affect the incentives of firms to keep workers informal. In general, the procedures for dismissal often require notification or even approval by unions, workers’ councils, the public employment service, a labor inspector, or a judge (table 5.3). Some countries also mandate retraining and reassignment to another job and establish priority rules for dismissal or reemployment of redundant workers. In Tunisia, companies must notify the labor inspector of planned dismissals in writing one month ahead, indicating the reasons and the workers affected. The inspector may propose alternatives to layoffs. If these proposals are not accepted by the employer, the case goes to the regional tripartite commit- tee comprising the labor inspector, the employers’ organization, and the labor union. The committee decides by a majority vote: If the inspector Table 5.3  Firing Regulations around the World, 2010 number of countries Where employers must Before dismissal of one redundant worker Before collective dismissal Notify or Obtain prior Notify or Obtain prior consult approval of consult approval of Region Total third party third party third party third party East Asia and Pacific 22 10 3 13 4 Europe and Central Asia 29 6 0 9 0 Latin America and the ­Caribbean 29 10 6 12 6 Middle East and North Africa 19 10 5 12 6 High income: OECD 24 9 1 13 2 South Asia 8 6 3 6 4 Sub-Saharan Africa 46 36 14 44 18 Total 177 87 32 109 40 Source: Doing Business 2011 database. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 249 and union reject the proposal, no dismissal is possible. The committee may also suggest retraining, reduced hours, or early retirement. Only 14 percent of dismissals end up being accepted. As a result, annual layoffs occur in less than 1 percent of the workforce, compared with more than 10 percent in the average OECD country. In Egypt, the employer has the right to close down or downsize the establishment, but it is a cumber- some process. Currently the employer may pay terminated workers one month of salary for each of the first five years of service, and one and a half month’s salary for each year after that, one of the most generous severance payments in the world. Eliminating or limiting some or all of the associated firing restrictions would give employers greater flexibility in responding to market fluctuations. Employers must have reasonable freedom to dismiss employees or they will be reluctant to hire and more inclined to operate in the informal sector. Most countries mandate severance pay with layoffs but differ in important details including extent of coverage, eligibility conditions, cause of dismissals, generosity of benefits, and level of benefits associated with seniority. Some countries require a minimum number of years worked before a worker is entitled to severance pay. In MENA severance pay for redundancy dismissal (for workers with 10 years of job tenure) is the highest in Egypt with 27 weeks of salary paid, followed by the Islamic Republic of Iran, the West Bank and Gaza economy, and the Republic of Yemen, at 23 weeks each (figure 5.6). In general, firing costs in poor countries are 50 percent higher than in rich countries. Some argue that this is justified because governments in poor countries do not have enough resources to provide unemployment insurance, so the cost should be borne by businesses. However, heavy regulation of dismissal is also associated with more unemployment, so those who want to work in poor countries frequently get neither a job nor unemployment insurance (World Bank 2004). In labor markets with less rigid and less costly firing regulations, appro- priately designed unemployment insurance (UI) schemes can provide adequate protection to workers. This allows firms to discontinue unpro- ductive employee–employer relationships, while maintaining adequate income protection through UI, a powerful support for creating higher productivity, good quality (formal) jobs. According to many (including Auer 2007; Auer and others 2004; Grazier 2007), legislative focus should be shifted from protection of jobs to protection of transitions, so that the individual risk of unemployment and income loss is reduced, while the potentially negative effects of job protection are avoided. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 250 Barriers to Coverage and Policy Options Workers ­ themselves feel better protected by a support system for unem- ployment than by EPL (European Commission 2006). This is particularly important in a world characterized by the gradual disappearance of life- long jobs and an increasing need for job mobility. Only a few countries in the region have UI systems, namely, Algeria, Bahrain, Egypt, the Islamic Republic of Iran, and Kuwait (table 5.4). Even in countries with UI sys- tems in place, such as Egypt, systems are underutilized because of a lack of public awareness about UI benefits among plan members, restrictive eligibility conditions, the difficulty of and the stigma attached to docu- menting a “just-cause” firing decision, and low overall layoff risks among covered open-ended contract employees (Angel-Urdinola and Kuddo 2010). The shift from rigid firing rules to less restrictive regulation Figure 5.6 Severance Pay for Redundancy Dismissal (Average for Workers with 1, 5, and 10 Years of Tenure, in Salary Weeks) by region South Asia 19.41 Sub-Saharan Africa 14.88 Latin America & 14.55 Caribbean East Asia & Pacific 12.64 Middle East & North Africa 10.86 (Non-GCC) GCC 10.73 Eastern Europe & 10.66 Central Asia OECD 6.76 0 5 10 15 20 25 weeks figure continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 251 Figure 5.6  Severance Pay for Redundancy Dismissal (Average for Workers with 1, 5 and 10 Years of Tenure, in Salary Weeks) (continued) by MENA country Egypt, Arab Rep. 26.7 West Bank and Gaza 23.1 Iran 23.1 Yemen 23.1 United Arab Emirates 18.1 Qatar 16 Saudi Arabia 15.2 Kuwait 15.1 Morocco 13.5 Algeria 13 Tunisia 7.8 Syrian Arab Republic 0 Djibouti 0 Lebanon 0 Jordan 0 Oman 0 Iraq 0 0 5 10 15 20 25 30 weeks Table 5.4 Existence of Unemployment Protection Legislation around the World Number of economies with unemployment scheme/total Economies and territories with number of economies in region unemployment scheme East Asia and Pacific 9/24 China; Hong Kong SAR, China; Lao PDR; Mongolia; Papua New Guinea; Solomon ­ Islands; Taiwan, China; Thailand; Vietnam Europe and Central Asia 23/25 All countries except Georgia and Kosovo Latin America and the 8/32 Argentina, Brazil, Chile, Ecuador, Puerto ­Caribbean ­ ruguay, R.B. Ven- Rico, St. Kitts and Nevis, U ezuela Middle East and North Africa 5/18 Algeria; Bahrain; Egypt, Arab Rep.; Iran, Islamic Rep.; Kuwait High income: OECD 30/30 All countries South Asia 1/8 India Sub-Saharan Africa 4/46 Mauritius, Seychelles, South Africa, Tanzania ­ Total 80/183 Source: Doing Business 2011 database. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 252 Barriers to Coverage and Policy Options accompanied by unemployment insurance creates the precondition for a more efficient allocation of resources. In simple terms, the easier it is for firms to discontinue a formal employment contract tomorrow, the more likely the firm will create that job today. This is especially true in sectors exposed to a volatile product market. Of course, the argument for economic efficiency should not justify reducing worker protection to ­ inadequate levels, but rather shifting the form of protection from protect- ing jobs to protecting income for workers in transition through UI. By contributing to UI, employees and employers share the social costs of unemployment, but not in a manner that forces firms to maintain unpro- ductive employment in downturns or to limit their incentives to open up vacancies when demand is stronger (World Bank 2008). Tax wedge. Labor taxes create a wedge between the labor cost to the employer and the worker’s take-home pay. Studies suggest that a higher tax wedge reduces both employment and economic growth (World Bank 2007). For example, a 10 percent reduction in the tax wedge (the difference between the cost of labor and take-home pay) could increase employment between 1 and 5 percent (Kugler and Kugler 2003; Rutkowski 2003). The literature for developing countries and emerging markets economies is limited, but a World Bank study on Turkey con- cluded that labor tax cuts would not have a major impact on formal employment (Betcherman and Pages 2007). An across-the-board reduc- tion of 5 percent in pension contributions paid by employers would bring about a 0.8 percent increase in employment overall and would reduce the unemployment rate by about 0.2 to 0.3 percent. The effect could be stronger (an increase in employment of almost 1.5 percent) if the reduction in pension contributions was targeted at workers younger than 30 years old, who have less bargaining power to capture most of the tax reduction in higher wages. Labor taxes in countries such as Morocco (where they account for about 39 percent of total labor cost) and Egypt (37 percent) are as high as the average for OECD countries. In Egypt, Jordan, and Lebanon, social security contributions are the dominant component of labor taxes (table 5.5). Social security contributions in these countries (paid by both the employer and the employee) account for the bulk of the tax wedge. In all reviewed MENA countries, social security contributions are paid largely by the employer; employees pay only a minor part. Table 5.5 shows the calculation of the tax wedge in some MENA countries, and figure 5.7 shows the tax wedge in various countries. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 253 Table 5.5 Contribution Rates for Social Security Programs, 2009/2010, and Tax Wedge on Average Wages in Private Sector All social security programs Personal Tax Country Year Insured person Employer Total income tax (%)a wedge (%)b Egypt, Arab Rep. 2009 14 26 40 10 37.3 Jordan 2010 6 12.75 18.75 7 22.2 Lebanon 2010 2 21.5 23.5 7 24.9 Morocco 2009 6.29 18 24.29 24 39.4 Yemen, Rep. 2010 6 13 19 25 30.4 Source: Calculations based on SSA and ISSA (2008, 2009, and 2011). a. Data refer to effective rates on average wage. b. The tax wedge is calculated as a sum of social security contributions paid by the employer and the employee, and the personal income tax is expressed as a percentage of total labor cost. Total labor cost is gross wage plus employers’ social security contributions. Gross wage is net wage plus employees’ social security contributions and the personal income tax. Figure 5.7 Percent Tax Wedge in Various Countries Jordan 22.2 Lebanon 24.9 Yemen, Rep. 30.4 Egypt, Arab Rep. 37.3 Morocco 39.4 Mexico 15.3 Korea, Rep. 19.7 Australia 26.7 Japan 29.2 United States 29.4 United Kingdom 32.5 Portugal 37.2 Turkey 37.5 Spain 38.2 Greece 41.5 Italy 46.5 France 49.2 Germany 50.9 0 10 20 30 40 50 60 Sources: OECD 2010. Policy implications for MENA. MENA countries could ease certain provisions of labor legislation to achieve more compliance and improved employment outcomes, while shifting from protection of particular jobs to protection of transitions. Overall, even though cer- tain provisions in labor legislation in some MENA countries might be rigid de jure, de facto they are widely evaded. It is unlikely that merely improving enforcement would result in reducing informality. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 254 Barriers to Coverage and Policy Options As ­discussed in this section, the rigidity of certain labor regulations in some countries contributes to the prevalence of informality (for example, hiring arrangements in Morocco, and firing arrangements in Tunisia and Egypt). A shift toward a richer and more flexible set of labor contracts (including more fixed-term contracts and fewer open- ended contracts), despite their drawbacks, would provide opportuni- ties for young workers and new entrants to join the formal sector through flexible working arrangements with social insurance coverage. Such policy reforms that ease regulations and make them more realistic to comply with should be supported by a reform of social ­ protection systems to better protect the income position of workers and their employment transitions.7 For example, recent experience shows that moderately strict EPL, when combined with a well- designed system of unemployment benefits and a strong emphasis on active labor market programs, can help create a dynamic labor market while also providing adequate employment security to workers (OECD 2008). Adequate safety nets could also play an important role in protecting workers from sudden job loss, help them transition between jobs, and prevent more people from slipping into poverty. The newly legislated unemployment insurance schemes in Jordan and Egypt provide an example to be considered by other ­ countries in MENA. Keeping the cost of labor at a realistic level via affordable social ­ security contributions in MENA and relaxing wage rigidities are likely to reduce informality. In general, institutionalized minimum wages in MENA are neither high (with the exception of Morocco) nor binding. Yet centralized wage setting mechanisms, such as those discussed in the case of Tunisia, contribute to informality by artificially setting high wage floors for certain occupations and skill levels. In addition to keeping minimum wages at low levels that can be realistically enforced, wage-setting mech- anisms in MENA should benefit from some kind of quantitative anchor to provide an objective baseline measure on changes in productivity. In countries where minimum wages are high (whether economy wide or sector specific), and where it is not politically feasible to reduce them, governments could consider reducing the minimum wage for youth to at least improve transitions of new entrants into formal employment, while maintaining the higher-level minimum wages that protect well- established workers. This section has also shown that the cost of labor attributable to labor taxes is not very binding in MENA, with the exception of Morocco and Egypt. In general, tax wedges could be reduced through Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 255 social insurance reforms that reduce the social security contribution rates (as already legislated in Egypt) or by shifting a portion of the labor taxes toward other general revenue taxes such as consumption taxes. Engaging in a more inclusive social dialogue is key to sustaining these types of reforms. Important political economy aspects to labor market reform are found. In particular, the traditional tripartite structure that convenes government actors, trade unions, and employer representatives is likely to show a bias for the status quo of protective regulation for employed, unionized workers. Including representation from the outsid- ers, informal workers, youth, and the unemployed would likely rebalance the dialogue toward facilitating entry in labor markets, improving mobility, and promoting a more equitable redistribution of returns across ­ different strata of the population. ­ Addressing the Preference for Public Sector Employment The widespread preference for public sector jobs in many MENA countries has important implications for informality. The presence of ­ large public sectors has been explained as the consequence of an implicit social contract in the Arab region that promised well-­ compensated public sector jobs to those reaching higher levels of ­ educational attainment (Yousef 2004). Recently hiring has slowed, and in practice, public sector jobs are offered only to those who are suffi- ciently patient to queue for them. Yet the preference for public employment is very widespread. For example, according to the 2010 youth survey in Egypt, 70 percent of youth say it is best to work for the public sector. This preference for p ­ ublic sector jobs is grounded in a rational evaluation of costs and benefits. On average, public sector jobs (1) are better compensated than s ­imilar private sector jobs; (2) offer full job security, good fringe benefits, and solid social status; and (3) tend not to require as much effort as p ­ rivate sector jobs. Jordan, Syria, and Egypt have the highest proportion of the workforce in the public sector in the region (30 percent in each of Jordan and Syria and 39 ­percent in Egypt). In Syria, the average public sector wage is 32 percent higher than the average private sector wage (representing both formal and informal workers). Even in Egypt, where public sector pay is considered low, the average wage in the public sector is 6 percent higher than the average wage in the formal private sector.8 The public employment bias and the queuing phenomena also exist in Jordan, where public sector wages are on average 20 percent higher than pri- vate sector wages.9 Bodor and others (2008) show that in Morocco Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 256 Barriers to Coverage and Policy Options public s ­ector employees, regardless of the level of education, have ­ better career paths, in terms of wages and pension benefits, than those working in the private, formal nonagricultural sector and the informal agricultural sector. The existence in some countries of a distinct single registry for those seeking public sector jobs indirectly contributes to the prevalence of informality. In Jordan and Syria, those who wish to work in the public sector must register with an agency. This registry is the institution of the “queue”; once registered, an individual does not need to exert any effort to procure a public sector job, he or she just needs to wait for his or her turn in the queue. In Syria, the total number of registered jobseekers was over 1.7 million persons in 2009. When the Syrian government attempted to use this registry to offer training, with the opportunity for private ­ sector placement upon conclusion, it found that those in the registry were typically unwilling to accept the prospect of formal private sector employment. They believed that presence in the social insurance agency’s administrative records would lead to deletion from the public sector job queue. Therefore, only a quarter of those offered formal jobs took them.10 In contrast, informal employment was acceptable to the same individuals; many of them were already engaged in informal employment.11 In the short run, MENA governments should consider eliminating institutionalized public sector employment queues. In the medium and long run, a reform of civil service by realigning incentives is needed. Modernized public employment and placement services should require active job search effort from applicants. Placement services should place workers based on their interests, training, and skills, instead of an expressed preference for public or private employment. Workers cur- rently in formal private sector jobs should not be disqualified from moving into public sector jobs later in their careers. In the long run, civil service reforms should establish stronger performance evaluation mea- sures, linking worker compensation to performance. Further, public sector wage scales should be rationalized so they no longer constrain flow of talent into the private sector. Enhancing the Productivity of Informal Workers through Training and Skills Upgrading The productivity dimension of informality is especially predominant in the poorer countries, in rural areas where low-skilled workers are engaged in micro-entrepreneurship and low-yield agricultural work. Programs aiming at increasing productivity in the informal sector are potentially Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 257 important interventions to promote inclusive growth and avert a productivity trap. However, effectively tackling productivity improve- ­ ments in the informal sector, particular in rural areas, is a complex agenda, which involves not only effectively upgrading skills, but also creating opportunities that would allow for returns to training to materi- alize. As such, complementary investments in infrastructure and access to markets are needed. An exhaustive analysis of policies to increase pro- ductivity is outside the scope of this report; this section focuses on policy and program options that improve access to training opportunities and realign training programs to the needs of informal workers. Improving access to training. Informal workers have limited opportu- nities to benefit from training provided by governments, employers, and private training providers. Active labor market programs (ALMPs), which include training and skills-upgrading programs, are ­ interventions provided by the government or by nongovernmental organizations that aim at increasing workers’ employability. In MENA, government provision of such programs is mainly directed to the unemployed. This may be because of the belief that the unemployed are more vulnerable and worse off than the employed, regardless of job quality. However, the working poor could actually be worse off than certain groups of the unemployed in some countries. Workers from higher income households might be able to afford to be unemployed and queue for better quality jobs, whereas those from poor households are forced to ­ take available low-quality and low-wage informal sector jobs. Figure 5.8 shows that unemployment is lowest among those with the lowest skill levels. This would suggest that those who can afford to stay in the education system could also be those who can afford to stay unem- ployed and wait for b ­ etter jobs. In addition to government provision, the ­private sector plays a large role in providing fee-based training ­ programs in MENA. However, an evaluation of privately provided ALMPs targeting young people in MENA revealed that the informal sector rarely has access to such programs: 80 percent of beneficiaries are educated males from middle- or high-income groups in urban areas (Angel-Urdinola, Semlali, and Brodmann 2010). It is also worth mentioning that only 5 percent of all surveyed training programs tar- get rural areas, where a large share of informal workers reside. Finally, as shown in chapter 3, informal salaried workers working in informal firms or small firms are less likely to receive on-the-job training where they work. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 258 Barriers to Coverage and Policy Options Figure 5.8 The Relationship between Educational Attainment and the Unemployment Rate in Urban Egypt and the Republic of Yemen ­ 18 unemployment rate 16 14 12 10 8 6 4 2 0 ow ne ry l y na ar ge nda l io rti el ra rb at te co oc yo se yv y/ ar ar or im nd at pr ar co ep se pr educational attainment Egypt, Arab Rep. Yemen, Rep. Sources: Egypt 2006 Labor Market Panel Survey, the Republic of Yemen 2006 Household Budget Survey. To address low access to training, as well as to provide incentives for firms and workers themselves to pursue training, governments may extend their provision of ALMPs to informal sector workers through ­ ouchers). direct targeting (for example, through training cooperatives or v Training vouchers can be used to empower recipients to buy training in the open market and thereby promote competition between public and private providers of training and the efficient delivery of training ser- vices. The Jua Kali program in Kenya, which offered training vouchers to those working in the informal sector in the mid-1990s, provides an interesting perspective on the response of public training institutions to the demand for skills created by the vouchers. The Jua Kali vouchers produced a positive supply response to the demand created for skills, but mainly from master craftspersons in the informal sector and non- governmental organizations (NGOs). Programs were tailored to the needs of voucher recipients and offered in off-hours to fit work sched- ules. The Jua Kali voucher program was successful in its pilot stage in expanding the supply of training to workers in the informal sector. Evidence was also noted of its positive impact on the earnings of par- ticipants as well as strengthened capacity of local Jua Kali associations Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 259 responsible for distribution of the vouchers (Steel and Snodgrass 2008). A commonly used policy intervention to address the underprovision of training by firms is the establishment of training funds, financed through general revenue or some payroll taxes. However, financing training through general revenues is regarded as less distorting to employment outcomes than through payroll taxes (and thus preferable). Training funds can be used to target firms with low levels of training, such as SMEs. For example, in the Republic of Korea, all firms pay a training levy, but the government provides reimbursement to employers who offer training. When the government found that mostly larger firms were actually providing training and benefiting from the reimburse- ment, they provided a ­ dditional incentives to SMEs to establish a train- ing consortium through which they could collectively mobilize resources for training while benefiting from a higher reimbursement rate and a subsidy to hire financial m­ anagers (Lee 2009). Tailoring training to informal workers’ needs. Traditional training programs generally require a minimum level of literacy and proficiency ­ and do not adequately address the need for a general and flexible set of skills that is typical of informal employment, especially in rural areas. General literacy skills are a barrier to productivity as well as a barrier for informal workers to access and benefit from training programs, especially in rural areas. Although the expansion of the basic education system in MENA has been remarkably successful in initially enrolling almost all children in rural and urban areas, the dropout rates are still high, and education quality often lags behind in rural areas, leaving rural agricul- tural workers with low levels of literacy.12 This is especially important given the evidence that literacy is associated with increased productivity. An often cited example is that literacy improves the use of fertilizers when workers can adequately read and comply with directions written on the labels. General education reforms addressing the low literacy and quality of education, particularly among rural workers, are necessary con- ditions for improved productivity in the medium term. However, in the shorter run, rural workers could benefit from training programs that are made accessible to them, and for which literacy is not a necessary precondition. Such training should also be associated with support for ­ micro-entrepreneurship and accessing markets, so as to broaden the set of available opportunities. How training is delivered matters too. Training programs for the informal poor need to offer clear, concrete, and immediate reasons to ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 260 Barriers to Coverage and Policy Options motivate enrollment and ensure that individuals participate and benefit from the program. Many informal workers are too poor to take time off from work to participate in daytime training. Thus, programs should ­ provide opportunities to combine earning and learning as well as flexible schedules. If alternative schedules are offered, such as evenings and week- ends, beneficiaries can then continue to contribute to household income and/or take care of their children during regular hours, thereby increasing beneficiary retention (Singh 2005). Moreover, experience shows that nonformal training programs should be adapted to the work context sur- rounding the beneficiaries, that the teaching methods should be partici- patory, and that some of the program instructors should ideally come from the neighborhood itself, because they bring with them insights on community needs. Unless the training is provided in the rural villages or is hands-on at the homes of the informal workers, it can be difficult to attend for several reasons: lack of transportation, insufficient ­infrastructure, and lack of lights along roads. Women face additional challenges because they may not be allowed to travel without male company. Although school dropouts constitute a majority of informal workers in many MENA countries, very few “second-chance” programs are aimed at providing learning in a nonformal manner. Second-chance programs can provide enhancement of an individual’s literacy, work skills, equivalency education, and life skills training, crucial characteristics that facilitate integration into society. Education equivalency programs are designed for those who have missed opportunities for early and traditional education and are unlikely to return to a formal learning environment. People who have dropped out of school at an early age are generally poor and very vulnerable. Second-chance programs are usually provided in a nonformal manner (often via accelerated learning) because this increases the likeli- hood of reaching informal and vulnerable workers. Life skills include social and coping skills, and improving relations with family, community members, and authority figures, while increasing the beneficiary’s own self-confidence. They can also include counseling and mentoring and components related to risky behaviors. Participants are more likely to benefit from work skills training once life skills and coping mechanisms are included in the general training (Angel-Urdinola and Semlali 2010). The traditional approach to training programs in the MENA region is not well suited for the informal sector. Training appears to be associated with a positive impact on labor market outcomes when offered as part of a comprehensive package. According to the survey of privately provided ALMPs in MENA, about 70 percent focus solely on hard skills and are Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 261 provided in classrooms, less than 20 percent provide some type of practi- cal experience and/or apprenticeships, less than 35 percent focus on soft skills, and only 14 percent provide some type of employment services and/or labor market intermediation (Angel-Urdinola, Semlali, and Brodmann 2010). Country-specific evidence confirms that similar approaches also prevail among publicly provided ALMPs. Many countries, particularly in the OECD and Latin America, have moved toward a comprehensive training model that includes provision of classroom and ­ workplace t ­raining, monitoring, job search and placement assistance, and soft skills training. Evaluations of “comprehensive” youth programs from Latin America i ndicate that programs can have a significant positive ­ impact on employment and earnings of program participants, especially for women, if they are organized with flexible schedules, based on public- private partnerships (that is, demand driven), combined with internships and practical experience (in addition to in-class training), provide a com- bination of soft and hard skills, and are monitored and assessed for impacts. In many Latin American economies, youth unemployment rates soared during the late 1990s. To address this, the Chilean government designed what is known as the “Chile Joven” program, which offered com- prehensive “demand-driven” training programs to unemployed youth. The program was so ­ successful that similar models were customized in Argentina, Colombia, the Dominican Republic, Panama, Paraguay, Peru, and República Bolivariana de Venezuela. Depending on the specific needs identified, these programs can be targeted to either the general unem- ployed youth population or to specific marginalized groups. If improved and combined with theoretical knowledge taught by Technical and Vocational Education and Training (TVET), traditional apprenticeships13 could contribute to more productive employment within the informal sector in MENA. Traditional apprenticeships are dis- tinct from formal apprenticeships, which are registered with a ­government agency and administered by employers. The flexibility of traditional apprenticeships in combining hands-on training, work and learning, their affordability and self-financing, their connection with future employment, and their generally low entry standards make them attractive to disadvan- taged informal workers. However, master c ­raftspersons rarely provide theoretical knowledge alongside practical experience and often teach outdated technologies, and there are few market standards available for judging the quality of the training provided. Traditional apprenticeships suffer from the low education of those being trained, and the choice of trades tends to follow gender biases (Johanson and Adams 2004). If public Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 262 Barriers to Coverage and Policy Options financing for TVET institutions shifted to focus on outcomes (such as suc- cess in serving target populations of master craftspersons and apprentices) rather than ­ financing inputs (such as classrooms, courses offered, or instructors hired), both apprenticeships and TVET could be made more relevant. Performance-based budgeting for public institutions could pro- vide incentives to upgrade technical skills for master craftspersons and improve their pedagogy (Ziderman 2003). More attention and account- ability could be given to these institutions in partnership with apprentice- ships for addressing the low levels of basic education that handicap training of apprentices and master craftspersons and for providing the complementary theoretical training needed to accompany the practical training of apprenticeships (Van Adams 2008). Training programs targeted toward the low-skilled, self-employed, and micro-entrepreneurs among informal workers, or those individuals with the inclination to become self-employed, could help informal workers transi- tion into higher productivity and higher value-added s ­ elf-employment. The great majority of participants in informal self- or household-based enter- prises have had little formal education or training when entering self- employment. Some, especially woman entrepreneurs, may have had none at all. The knowledge and skills used in their businesses have likely been acquired from parents and other relatives in f ­amily enterprises. The self- employed can benefit from both technical training specific to their indus- try, such as on the use of modern production techniques in agriculture, as well as general entrepreneurship and business skills training. The latter could include bookkeeping, financial literacy, marketing, communication, life skills, and simple risk management. International experience has shown that a comprehensive package offering a set of services that both includes training and facilitates access to credit can be successful in improving entre- preneurial ventures. Moreover, micro-franchising programs are an emerging approach that entails helping individuals replicate an existing business rather than starting an original one. Box 5.3 presents two case studies of productivity-­ enhancing programs from Egypt and Jordan. Finally, moving toward more integrated and innovative social safety net systems that link income support to the poor with strategies to foster productivity should be considered in the context of MENA. The “Chile Solidario” Conditional Cash Transfer Project provided the poor with the means to attend training while linking them to employment opportunities. This well-targeted social protection project assisted extremely poor families, mostly in Chile’s rural areas. A social worker worked with ­ anyone in the family in need. Each individual’s needs Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 263 Box 5.3 MENA Case Studies of Productivity-Enhancing ­Programs in ­ rban Areas U The Cairo Earnings and Learning Project of the Children and Youth Garbage Collectors provides an example of an urban community-based nonformal project. By working ­ directly with the garbage-collecting community itself, this project, headed by a grass- roots NGO, has managed to reach several hundred children. The beneficiaries asked for training that would not alienate the children from their trade and families and for an education that would respect and build on the children’s existing skills of sorting and recycling waste. A curriculum was developed within the context of recycling: Participants were taught basic math and literacy, marketable skills within the weaving industry, waste recycling, health skills, family planning, personal hygiene, and sanitation. Children were removed from their hazardous environment and given opportunities to earn money in a clean and safe environment while also learning. The children would most likely not have been able to participate to the same extent had they not been able to work. The woven products they produced were sold through a cooperative system for young artisans and became popular among Egyptians and tourists alike, leading to a source of pride for the young weavers. Sales earnings were divided between the weavers. Results show that after six years, some 500 young women have graduated and achieved functional literacy. Moreover, 64 percent reported that they were practicing family planning, 56 percent said they would not circumcise their daughters, and 70 percent of single girls said they would not circumcise their daughters when they got married (Madhu 2005). Jordan’s “Questscope Program” provides urban street children, school dropouts, and young workers in the informal sector with a nonformal equivalency education, jobs, and life skills. In 2007 the Ministry of Education made the program an official alternative to tenth grade certification. The program was a response to the Department of Labor’s initiative to ensure that working children under the age of 16 (the legal age for work) are withdrawn from the labor force and reinserted into nonformal education (accelerated learning) or formal education. In addition to equivalency ­ education, job training, and life skills, the program provides income support to the beneficiaries as they attend evening classes in public schools. Certified teachers help beneficiaries earn a proficiency certificate (equivalent to tenth grade level). Vocational graduates also receive business-management training, thereby enhancing future employability and livelihood options. Other program elements include coaching in life and social/coping skills aimed at facilitating the integration of youth into society. The program is provided by a grassroots NGO with strong linkages to the community. The classes are flexible to ensure that attendance and learning are centered on the realities of the participants. The cost per beneficiary is around $350 per year ­ (Semlali 2008). Source: Semlali 2008. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 264 Barriers to Coverage and Policy Options were evaluated and the social worker then linked the person to the appropriate services, including literacy classes, soft skills training, help with job searches, links to internship opportunities/subsidized employ- ment opportunities, or self-employment assistance programs. The connection to the social worker was important for understanding the ­ needs of the client and for accurately informing the client of available opportunities, because many poor are not aware of the services avail- able to them. Conditional Cash Transfers in the form of a stipend were provided conditional upon training participation. Without this stipend, the poorest would not have been able to participate (Angel-Urdinola and Semlali 2010). Conclusions to Part 1 Numerous barriers to formality exist in MENA, requiring a complex set of policy interventions. Restrictive business and labor market regulation, the prominent role of the public sector as employer in a number of countries, and the productivity gap facing informal workers are all important barriers to inclusive growth and formalization. Although dif- ferent and complementary policy interventions that relax these barriers can be effective toward this goal, the process of formalization matters, and policy interventions to address informality should extend beyond mere enforcement and should aim to reduce informality in a sustainable manner while helping the growth process become progressively more inclusive. A healthy business environment and labor regulations that foster more mobility in labor markets, while protecting workers during job transitions, are important. The evidence suggests a negative correlation between the ease of doing business and the size of the informal sector. In MENA, barriers to entry, high taxes, and discretionary enforcement of regulation all collude to promote informality. Simplifying entry regulation, reducing compliance costs, and moving toward a fairer implementation of regulation are all necessary, and emerging evidence from other countries suggests that these interventions can be effective to move beyond informality. A large portion of employers in MENA perceive labor regulations as a major obstacle to business development and more employment growth. In some MENA countries, certain labor regulation provisions are rigid, including hiring arrangements in Morocco and firing arrangements in Tunisia and Egypt. Rigid EPL pro- motes informality, because firms can respond to rigid labor regulations by reducing overall employment or shifting employment into ­conditions Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 265 of informality. Easing certain provisions of labor legislation to achieve more compliance, supported by a reform of social protection systems to better protect the income position of workers and their employment transitions (for instance, through the introduction of unemployment benefits and a strong emphasis on active labor market programs), can decrease informality and promote employment creation. In parallel, it is also important to keep the cost of labor at a realistic level, including through affordable social security contributions. The generosity of ­ public sector employment conditions (including pay, benefits, and job security) in some countries also contributes to higher informality and important segmentations, making the need for a reform of civil service even more pressing. Many informal workers face a productivity trap. Especially in the poorer countries and in rural areas, the low productivity of jobs is the predominant aspect of informal employment. Low productivity is a result of different factors, including poor skills and limited opportuni- ties, particularly in rural areas. The evidence shows that informal work- ers consistently have limited access to training and ­ skills-upgrading opportunities. Although complementary investments in infrastructure and access to capital and markets will be necessary to increase the returns associated with skills upgrading, targeting such programs to informal workers can be one effective way to address the productivity trap. Providing incentives to firms (such as through training coopera- tives for small firms) and workers (such as with vouchers) to engage in training will address some of the determinants of underprovision of these programs. To make these interventions more effective, reorient- ing and tailoring the delivery and design of training toward the particu- lar needs of informal workers is necessary. Second-chance programs, traditional apprenticeship, and training specifically designed for the self-employed and micro-entrepreneurs are examples of interventions that are likely to be effective in the context of MENA. Part 2. Extending Social Insurance Coverage Introduction to Part 2 Lack of social insurance coverage exposes workers and their families to important risks and vulnerabilities. In MENA, these vulnerabilities loom large, with about 67 percent of the labor force not protected against a plethora of social risks, of which loss of income in old age may be the most pressing. Government interventions that effectively Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 266 Barriers to Coverage and Policy Options expand access to risk management instruments can be beneficial both from the ­ perspective of individual/household welfare and for society as a whole, because evidence exists that private markets are likely to underinsure social risks, such as loss of income-generating capacity due to old age, disability, health conditions, or layoffs. This is particularly true for the poor, who might engage in suboptimal strategies, such as selling productive assets or withdrawing children from school, to respond to shocks. From a societal point of view, important negative externalities can be found from having too much uninsured risk, including an adverse impact on productivity. If the rationale for govern- ment intervention is clear, assessing what drives lack of coverage is key to informing which policies are most likely to succeed in expanding it. For example, if most workers are observed to voluntarily opt out of social security systems because of a rational cost-benefit analysis, then improving the perceived quality of public service, including outreach and communication about benefits, would be needed. Specific design features of pension systems, including vesting periods, early retirement, and legal coverage provisions, might also provide ­ different incentives for workers to contribute to social insurance. Furthermore, a significant portion of the population might not possess sufficient saving and con- tributory capacity to pay for the true cost of the socially optimal degree of protection against these risks, which would justify government inter- vention to improve coverage with some level of subsidy beyond ­traditional contributory mandatory social insurance schemes. A ­complex set of reforms is likely to be needed, which includes moving beyond a sole focus on enforcement of mandatory social i ­nsurance rules (espe- cially under existing designs and lack of financial sustainability) toward a coverage extension strategy that acknowledges the realities of the informal economy in MENA. Part 2 of this chapter introduces a conceptual framework for social insurance coverage extension policies. In this context, it discusses design features of pensions systems in MENA, with special attention to the need for improving the design and incentive structure of existing formal sector pension schemes as a precondition for feasible and successful coverage extension efforts. Following that, a structured description of alternative coverage extension strategies is presented, including a set of guiding prin- ciples to select such strategies for specific vulnerable population groups. Finally, it concludes with a section on how the often disregarded evidence from the emerging field of behavioral economics could support better designed social insurance coverage extension policies. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 267 A Framework for Social Insurance Coverage Extension Policies: Why Does It Matter? Although the costs and benefits to informality have been discussed, the vulnerability associated with informal employment is a social concern requiring government action. The main objectives of this section are to describe (1) how the design of social insurance systems affects incentives to be informal and (2) the policies governments should consider to alleviate the social problems caused by informality. In line with the ­ emphasis of this report on informal employment, policies aimed at extending social insurance coverage are the focus of the following ­ discussion, because they provide a means for individuals to reduce vulnerability and excessive exposure to social and economic shocks, and ­ to ultimately improve social welfare. The ultimate objective of coverage extension policies is to improve social welfare by providing individuals with access to the risk management tools of social insurance systems. A framework for social insurance/coverage extension policies is illustrated in figure 5.9. Social insurance systems pro- vide protection against certain social and economic risks, such as the loss or sudden reduction in income (generating capability) due to old age, disabil- ity, work-related injury, death of an income-generating family member, sickness, loss of income due to maternity, or loss of a job.14 The ­realization Figure 5.9  A Framework for Social Insurance/Coverage Extension Policies: ­Objectives, Social and Economic Risks, Policies, Instruments, and Behavioral Impacts Framework for social insurance/coverage extension policies Objectives Consumption Protection from Preservation of smoothing falling into poverty human capital Risks Loss of income Old age Unemployment (generating capacity) due Sickness Work injury to Expense shocks Disability Maternity Policy instruments Risk pooling (Mandatory) saving Transfers (redistribution) Behavioral impacts Individual behavior Household behavior Firm behavior Source: Adaptation of the broader framework presented in Robalino and others 2010. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 268 Barriers to Coverage and Policy Options of any of these risks, at least temporarily, u ­ndermines an individual’s employment income-generating capacity, and in some cases a medical care- related expense shock is incurred as well (for example, with sickness, dis- ability, and work injury). The social insurance system is designed to provide income support and intervene in such cases to at least temporarily ­ cover a significant portion of the shock’s expenses. These interventions are in line with the underlying objectives of (1) consumption smoothing (lim- iting the extreme fluctuation of individual and household consumption and welfare), (2) poverty alleviation (preventing social and economic shocks from forcing individuals or households into poverty), and (3) pre- serving investment in human capital (in the cases of health insurance, dis- ability, and work injury benefits, there is a direct objective to prevent the—further—deterioration of health). In a broader social insurance con- text, temporary income support measures are justified as an attempt to prevent households from reducing their investment in human capital (such as spending on education) when social and economic shocks occur. Social insurance programs typically entail a combination of risk pooling, (mandatory) saving, and explicit redistributionary transfer mech- ­ anisms (Holzmann and Koettl 2010). The optimal choice of a policy instrument depends on the nature of the risk. In general, low-probability shocks with potentially devastating impact are best protected against through risk pooling. In contrast, the loss of income-generating capacity in old age is a high-probability (thus predictable) event; therefore (man- datory) saving-type pension schemes are more appropriate. Saving and risk pooling are, in theory, contributory mechanisms ensuring the finan- cial sustainability of the social insurance systems and ensuring that any redistribution among plan members happens based only on the ex ante unknown realization of risks (that is, in a random manner). The discussion to follow defines a necessary role for redistributionary transfers, especially for ensuring protection to low saving capacity individuals (and house- holds) who cannot afford to pay the true cost of a universally guaranteed social insurance package. Such transfers play an especially critical role in coverage extension policies. Understanding behavioral impacts is especially important for the design of coverage extension policies given the need to induce voluntary compliance and enrollment. Given the nature of informality, relying only on enforcement of mandates is not sufficient and can be counterproduc- tive. Economic actors respond to incentives. For example, self-employed individuals decide to register and enroll in social insurance schemes based on whether they find their participation beneficial given the information Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 269 they have and the level of effort required. Firms make employment offers with or without associated enrollment in the mandatory pension system dependent on factors such as the related additional cost of labor. Employees and employers may negotiate over a worker’s enrollment as part of a total compensation package. Households may optimize social insurance participation of one household member based on the degree of protection offered to the entire household. A focus on expected behav- iors is necessary for understanding the true effect of policy interventions. Later in this chapter it will be shown that the lessons learned from behav- ioral economics about bounded rationality can significantly inform the design of coverage extension policies. Improving the Design of the Existing Social Insurance System The link between pension design and coverage: A review of key options. This section describes an array of pension system designs and their features as related to coverage. The designs of old-age pensions vary greatly in achieving wide coverage, and few systems manage to provide access specifically to informal sector workers. In their simplest form, old-age pension systems are mandatory saving schemes that prevent myopic undersavings for old age and curb intentional undersavings and abuse of often generous social safety nets that redistribute income to those in need. Earnings-related pension schemes (see panel 1 of figure 5.10) provide old-age pension benefits dependent on individual contributions; they work, in effect, as saving mechanisms from the perspective of the indi- vidual and can be designed as defined benefit (DB) or defined contribu- tion (DC).15 In MENA, pension benefits are determined through a set of parameters taking into account the individual’s contribution performance over the active life cycle (such as number of contributory periods, and some average wage measure reflecting earnings before retirement or over the entire contributory life span). These are known as DB schemes. The expected value of pensions may be quite different than the value of life- long contributions. If the pension benefits are systematically higher than the value of contributions, contingent government liabilities emerge, giv- ing rise to implicit pension debt. Mandatory pension schemes in MENA tend to be pay-as-you-go (PAYG), as opposed to “funded,” in their under- lying financing mechanism, with pension benefits financed by the contri- butions of active age plan members.16 Pure earnings-related mandatory pension schemes are rare and lack the minimum old-age income feature that most modern pension schemes possess in one way or another. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 270 Barriers to Coverage and Policy Options Figure 5.10 Old-Age Pension Design and Coverage: Options Ranging from Pure Earnings-Related Pensions to Pure Noncontributory Flat Pensions pension pension pension income income income (1) Earnings-related pension (2) Minimum pension guarantee (3) Universal basic pension pension pension income income (4) Universal basic pension (5) Means targeted matching defined contribution income level–dependent subsidies income level–dependent contributions By ­definition, mandatory earnings-related pension schemes in general, regardless of their design or financing mechanism, do not offer protection to informal sector workers; these schemes typically require that employ- ers register their employees with the social insurance authority, report their earnings. and pay wage-proportional employee and employer con- tributions, while earnings of informal sector workers are rarely observed. Later in this section, the specific design features of PAYG pensions sys- tems in MENA that pose challenges to participation are explored. In contrast to earnings-related pension schemes, noncontributory basic (flat) pension schemes do not take into account an individual’s income position before retirement to determine his or her benefit amount. Such schemes assume a social responsibility for providing everyone with a minimum standard of living at old age regardless of their saving capacity before retirement (see panel 4 of figure 5.10). As such, they have limited scope for controlling myopic behaviors, a disadvantage over earnings- related pension schemes. Although earnings-related pension schemes attempt to smooth consumption patterns between an individual’s active and retired life, flat noncontributory pensions do not necessarily prevent Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 271 a radical drop in consumption in retirement but offer wider coverage. By design, noncontributory pensions rely on general revenue financing and cover formal and informal sector workers, as well as those outside the labor force. Noncontributory pension schemes can provide 100 percent coverage in exchange for government spending as low as 1 to 2 percent of GDP.17 The most commonly cited universal flat pension scheme is in New Zealand, where no other pension program is offered by the government.18 Establishing a minimum pension floor in an earnings-related pension scheme improves income protection compared with a pure earnings-related design but does not improve the potential for coverage. Additionally, it can undermine contribution incentives. The minimum pension is a redistribu- tory feature of earnings-related pension schemes, which ensures an income floor in old age to anyone with an acceptable lifetime contribution effort (for example. with a sufficiently large number of contributory years), regardless of what the earnings-related pension benefit determination would yield (see panel 2 of figure 5.10). The cost of a minimum pension (that is, the “top-up” over the earnings-related pension) is financed from general revenues or redistribution from other plan members. Its potential to increase coverage is limited because it does not extend benefits beyond the Bismarkian (earnings-related) scheme for the formal sector. Moreover, it may create incentives for individuals to hide income, move certain employ- ment activities into informality, or save outside the system, so that lower levels of observed income and contributions will create eligibility for the top-up. A recent innovation that addresses the flaws of the minimum pension combines an earnings-related pension scheme with a universal (flat) basic pension. One policy option is to combine a pure earnings-related scheme with a universal basic (flat) pension without any connection between the two components. This solution provides smooth income patterns across active and retired life for formal sector workers and establishes an income floor in old age for everybody. The criticism of this solution is its overall high cost, involving uniform general ­revenue-financed universal pension benefits to those with already high earnings-related pensions. An innovation over this simple unlinked combination is the so-called “universal basic pension with claw-back,” which attempts to combine the high coverage of the universal basic pension with the incen- tive design of a pure earnings-related pension, by gradually phasing out the universal basic pension component as old-age income from the earnings-related pension increases (see panel 3 of figure 5.10).19 This mechanism provides complete old-age income protection coverage Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 272 Barriers to Coverage and Policy Options while avoiding the contribution incentive trap of the minimum pension. Under the latter, for a class of low-income workers, any contribution that will not eventually elevate earnings-related pensions above the mini- mum pension is a sunk cost. In contrast, the claw-back feature reduces the marginal benefit from contributions among low-income contributors but does not eliminate it. Universal basic pensions with a claw-back are always cheaper than an otherwise identical flat universal benefit scheme because of the built-in targeting mechanism and benefit reduction. However, a universal pension with a claw-back is more expensive than an otherwise identical earnings-related pension scheme with a minimum pension because it covers more individuals using general revenue fund- ing and the top-up is larger, as apparent in figure 5.10. This innovation was recently enacted (but has yet to be implemented) in Egypt and already exists in Chile and South Africa. The most recent innovation for incentive-compatible pension cover- age extension is the matching defined contribution (MDC) design. MDCs are voluntary defined contribution saving mechanisms, offering old-age or other social insurance benefits, where the government or employer provides incentives to enroll by matching individual contribu- tions at a given rate and threshold. MDCs increase the incentive to save through matching contributions, effectively increasing the return on sav- ings. Many employer-sponsored pension programs, such as the 401(k) scheme with employer matching in the United States, already have this feature. In panel 5 of figure 5.10, the dark gray columns show the income level–dependent subsidies, and the light gray columns show the income level–dependent contributions under a general MDC design. The figure reflects a further extension to MDCs, combining the matching mecha- nism with a targeting mechanism (such as means testing or proxy means testing) to induce participation by those with limited saving capacity, without regressively offering higher matches to individuals with higher income or savings capacity. This feature makes MDCs suitable for cover- age extension. It is conceivable that a significant number of households desire social insurance services, but their income position does not allow them to pay for the true cost of their social insurance coverage in the form of contributions (see box 5.4 for an example of a targeted MDC in the case of health insurance in Lebanon). The MDC innovation demon- strates that targeting mechanisms often used to reach the poorest and most disadvantaged households can also be used to target those above the extreme poverty level, but below the income level that would allow them to enroll in the contributory social risk management mechanisms. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 273 In fact, the relative income status of households can be used to determine the amount of matching subsidies for which households are eligible if they are willing to enroll in the related social insurance mechanism. The MDC scheme is likely more affordable than universal basic pensions due to the targeting mechanism and the consequently reduced inclusion error. However, MDCs do not ensure complete coverage as well as universal schemes do; their potential to increase coverage can only be met if (1) the matching design adequately motivates contributions at all income levels, (2) the transaction cost of enrollment is small, and (3) lessons learned from behavioral economics on irrational behavior are used in the program design (discussed later in this chapter.)20 The key issues with increasing coverage through noncontributory (or heavily subsidized) special schemes are the potential unintended Box 5.4 Proposal for Expansion of Health Insurance in Lebanon through a ­Targeted MDC The government of Lebanon has embarked on reforms to expand health insurance coverage to individuals who are not covered by the National Social Security Fund ­ (NSSF) and other contributory schemes. The NSSF, which covers about 40 percent of the population, provides health insurance to formal sector workers in the private sector and to civil servants through earnings-related contributory schemes. For those not covered by the NSSF, the Ministry of Public Health acts as the insurer of last resort. These are mostly poor or near poor, low-income self-employed, and informal wage earners. The recent reforms (Health Card Initiative) launched by the government entail subsidizing the low-income population who cannot afford the premium to be covered. Figure B5.4.1 is an illustration of how the consolidation of the health plans can be achieved if the ­government chooses to use a targeting strat- egy offering three plans (a basic plan, an intermediate plan, and the more compre - hensive NSSF plan) and allocates subsidies based on the level of earnings. As shown in the figure, a universal subsidy is provided to all individuals regardless of income that covers externalities and public goods. Additional subsidies are allocated based on the income quintile to which beneficiaries belong. The proxy means testing sys- tem currently being developed under the National Poverty ­ Targeting Program could be used to indicate beneficiaries’ income brackets. As income increases, subsidies decline, and the required individual contributions rise. Thus, individuals with some savings capacity would contribute to the cost of the plan to receive the subsidy. Individuals who can afford to pay additional contributions can finance more “gener- ous” health packages. box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 274 Barriers to Coverage and Policy Options Box 5.4  Proposal for Expansion of Health Insurance in Lebanon through a ­Targeted MDC (continued) Figure B5.4.1 Subsidized and Individual Contributions by Income Quintile subsidized and individual contributions (monetary units) 35 NSSF plan 30 intermediate plan 25 basic plan 20 15 10 5 0 1 2 3 4 5 income quintile additional contributions individual contributions to BP redistribution externalities Source: World Bank 2010. Note: BP = Basic Plan ; NSSF = National Social Security Fund. ­onsequences in the form of behavioral responses; such schemes may c reduce willingness to participate in the contributory mechanism among those who are able to and could ease the path for firms to offer jobs without social insurance coverage. Clearly, the easiest way to extend coverage is through universal general revenue–funded noncontributory ­ transfers such as the universal basic (flat) pension. In addition to the higher costs of a universal program, the availability of noncontributory benefits may create incentives for current contributors (and some uncov- ered who otherwise would exert effort to contribute) to transition to the noncontributory scheme. Even if the services of the noncontributory system are of lower quality, this argument still holds as long as the services are valuable and include a higher subsidy component than the ­contributory Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 275 benefit alternative. Levy (2006) provides evidence in this regard, indicat- programs for uncov- ing that the availability of subsidized social insurance ­ ered workers may unintentionally allow firms to shift employment into contractual relationships not covered under the formal social insurance scheme or into self-employment. The macroeconomic result of such a shift may be that aggregate labor is overemployed in the informal sector and underemployed in the formal sector. This increases labor costs in the formal sector, and thus these programs can have an economy-wide nega- tive effect on productivity competitiveness, and potentially on growth. Alterido and others (2009) have shown that expansion of a noncontribu- tory health insurance program in Mexico (Seguro Popular)21 has a moder- ate downward influence on participation in the formal contributory social insurance scheme covering health insurance as well as pension and unem- ployment benefits. The authors show that 65 percent of Seguro Popular beneficiaries are nonpoor. Overall, the availability of noncontributory special schemes is likely to decrease incentives to participate in contribu- tory social insurance mechanisms. In addition, access to noncontributory schemes allows employees and firms to collude by not incurring the cost of enrolling the employee in the contributory social insurance system to share the emerging surplus. In cases of weak bargaining power on the side of employees, the availability of noncontributory schemes can give rise to firm behaviors that shift the creation of new jobs to informality, justifying this by the access of workers to the noncontributory benefits. Theoretically, the MDC design would not suffer from this problem if the matching subsidy were truly aligned with the income position of the contributor. In ­ particular, the better the targeting in assessing saving capacity, the less likely that undesirable behavioral responses would emerge. Design of pension systems in MENA and determinants of coverage. Universal social insurance coverage is a relatively modern policy goal; current coverage patterns in MENA reflect a history of social security ­ provision initially limited to the civil service, without the ambition of universality. Most social insurance systems in MENA were born in the late 1960s or early 1970s through the establishment of schemes specific to civil service employees. In fact, worldwide, public sector employees were usually the first to be covered by pensions and other social insurance schemes sponsored by governments in their role as employers.22 Gradual expansion to formal private sector employees and other groups took place at later stages. In MENA, though expansion of pension schemes happened beyond the civil service has been achieved in all countries, it has ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 276 Barriers to Coverage and Policy Options in a fragmented manner. As shown in figure 5.11, the share of MENA countries that have already integrated civil service schemes within larger schemes that cover workers in other sectors (such as the private sector) is lower than in the ECA and LAC regions, comparable to EAP, and greater than in SSA and SAR.23 Currently, 11 of the 19 MENA countries still operate a separate or only partially integrated scheme covering civil servants and private sector workers. Concerns regularly surface regarding the legacy costs of integrated civil service schemes in countries where the integration has already been implemented (such as in Jordan and Syria). None of the countries operate one single national scheme that covers all types of workers. The fragmentation of the system poses various prob- lems, especially as MENA economies seek efficient and productive resource allocation across sectors. First, if benefits are not portable across schemes, labor mobility across sectors with different schemes is con- strained (box 5.5). Moreover, generous sector specific schemes (such as civil service schemes) may crowd out labor supply in other sectors (such as the private sector) and create inequities (Robalino 2005). The follow- ing sections discuss three features that are common to many MENA pension systems and that are problematic for informality: legal coverage, early retirement, and calculation of benefits. Outside the public sector, pension system coverage in MENA is still limited, and legal provisions for covering the self-employed and agricul- tural workers are sporadic. MENA social insurance schemes reflect the traditional view that coverage is associated with a formal employment relationship, and they rarely accommodate special employment status categories. In several MENA countries, certain types of workers are legally excluded from coverage. Thus, in some cases, coverage extension may need to start with expansion of “legal coverage,” as was done recently for workers of agricultural cooperatives in Morocco. Traditional social insur- ance coverage implies that the self-employed are excluded from coverage by default. Only half of the MENA countries offer mandatory or volun- tary coverage for the self-employed (table 5.6; see Robalino 2005; SSA 2009, 2010). However, integrating the coverage of the self-employed under the current DB pension schemes in MENA is challenging because it would put a significant burden on the often low-productivity self- employed. Moreover, the existing pension schemes in the majority of MENA countries exclude a large portion of rural workers; at best, work- ers in large agricultural organizations such as cooperatives can enroll. As for unpaid family workers, coverage relies exclusively on household members accessing social insurance benefits (such as survival pensions or Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 5.11 Path Dependency in the Evolution of Pension Systems: Degree of Integration of Civil Service Pension Schemes in MENA and Other Regions Pension system integration in MENA 100 90 Separate or only partially integrated pension schemes for the civil service: 80 Bahrain; Djibouti; Iran, Islamic Rep.; Iraq; Lebanon; Morocco; 70 Oman; Saudi Arabia; Tunisia; Yemen, Rep. [x2]; West Bank and 60 Gaza economy 50 Fully integrated pension schemes for civil service: 40 Algeria; Egypt, Arab Rep.; Jordan; Kuwait; Libya [x2]; 30 Syrian Arab Republic; United Arab Emirates [x2] 20 Distinct pension scheme for the military: 10 Algeria; Djibouti; Lebanon; Libya [x2]; Oman; Saudi Arabia; 0 Tunisia; United Arab Emirates [x2]; Yemen, Rep. [x2] ECA LAC MENA EAP SSA SA separate or only partially integrated (%) fully integrated (%) Source: World Bank pension systems database. Note: Integration refers to integration of civil service pension schemes within schemes covering workers in other sectors, including the formal private sector. ECA = Europe and Central Asia; EAP = East Asia and Pacific; LAC = Latin America and the Caribbean; SA = South Asia; SSA = Sub-Saharan Africa. 277   278 Barriers to Coverage and Policy Options Box 5.5 Integration and Portability of Social Insurance Schemes The integration of distinct public and private sector social insurance schemes is driven by the need for reduced distortionary forces in the labor market. Programs covering distinct groups in the labor force without explicit portability arrangements of vested social insurance rights limit labor mobility and therefore cause economic and labor market distortions. For example, public sector workers who are vested in the public sector pension scheme for many years, but not long enough to draw pen- sion benefits, may be reluctant to accept otherwise desirable job offers from the private sector, because their past contributions may become a sunk cost, and future private sector vestment may not be long or secure enough to yield pension benefits either. To diminish this distortion, and to facilitate labor mobility, governments either integrate public and private sector social insurance programs or establish arrange - ments for their portability. With integration, the same set of rules applies for all enrolled plan members regard- less of their employment sector. Portability arrangements are often thought of as second best solutions” for labor mobility; distinct social insurance contribution “­ requirements and benefits still apply to different groups in the labor force, and the dif- ference in generosity (often favoring those working in the MENA public sector) may induce reluctance to switch jobs. Segmentation of social insurance ­ systems also limits international labor mobility, further motivating the current trend toward integration of sector specific schemes domestically, and establishing cross-border portability among integrated national social insurance schemes. health insurance) through covered family members. Temporary and casual workers are excluded from coverage in the Republic of Yemen and Syria. Another example of legal exclusion from coverage is in Syria, where employees in firms with fewer than five employees receive coverage limited to work injury only. In Lebanon, all private sector ­ employees are not eligible for pensions, but rather receive a lump-sum end-of-service indemnity. Further contributing to the fragmentation of pension systems and labor markets in MENA, various countries have established specific occupational schemes that are usually not subject to government regula- ­ tion. Occupational schemes exist in the Islamic Republic of Iran, Morocco, Egypt, and Tunisia. Tunisia’s experience in expanding coverage to the self- employed is based on the development of an income estimation system for different occupational groups. Based on the estimated lowest income bracket assigned to each occupational group, the required personal Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 279 Table 5.6 Legal Coverage for Self-Employed Workers and Agricultural Workers in ­Non-GCC MENA Economies Coverage for self-employed Coverage for mandatory (M) or agricultural Economy voluntary (V) workers Remark Egypt, Arab Mandatory coverage Covered In agriculture, casual workers covered Rep. Morocco No coverage Covered Agricultural workers in cooperatives covered Tunisia Mandatory coverage Covered Agricultural workers in cooperatives and agricultural wage earners covered Syrian Arab No coverage Covered Agricultural workers with formal jobs (a Republic marginal segment) covered under general social insurance scheme Algeria Voluntary coverage No coverage Special scheme for self-employed Jordan Voluntary coverage Lebanon No coverage No coverage Iran, Islamic Mandatory coverage No coverage Coverage of self-employed within Rep. scheme for private sector employees Yemen, Mandatory coverage No coverage Coverage of self-employed within Rep. scheme for private sector employees Djibouti No coverage Covered Agricultural workers in cooperatives covered under general social insurance system West Bank No coverage No coverage Excludes self-employed and Gaza economy Libya Mandatory coverage No coverage Coverage of self-employed within general social insurance scheme Iraq No coverage No coverage Excludes self-employed Sources: Robalino 2005; SSA 2009, 2010. ­ ontributions are identified. Workers can choose to contribute more, or c they can contribute less if they prove that their real income is lower than the bracket assigned to them. This kind of approach suffers from chal- lenges of financial sustainability and inaccuracy of estimated income. Although specific mandatory occupational schemes have been used as a gradual approach to expanding coverage, and have been sought in coun- tries such as South Africa, Korea, and Costa Rica (Von Ginneken 2009), their potential as an approach to increase coverage (especially compared with more wide-scale extension of coverage strategy) is very much affected by whether the groups are large and sufficiently homogenous (Olivier 2009; Robalino and Palacios 2009). Arguably, such a coverage Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 280 Barriers to Coverage and Policy Options extension approach increases the segmentation of the social insurance system and could ultimately undermine integrated coverage extension efforts that attempt to attract voluntary participation regardless of the nature of workers’ employment status. In terms of design, mandatory earnings-related pension schemes in MENA all follow the DB pension system design. Figure 5.12 illustrates the relative share of each of four mandatory pension design types (DB, DC, notional defined contribution [NDC], and provident fund [PF]) that exist around the world. In MENA, DB systems dominate. The widespread design weaknesses of these schemes in MENA contribute to the observed high degree of informality. Modern versions of DB pension design are capable of creating favorable participation incentives similar to those of DC schemes. In fact, it has been proven that DB and DC design can be identical if the DB design (1) adjusts the benefit level in an actuarially fair manner for early or delayed retirement before or after the statutory retirement age, (2) has an automatic mechanism to change the statutory Figure 5.12 Mandatory Pension Design Types across Regions of the World 100 90 80 70 60 percent 50 40 30 20 10 0 MENA SSA LAC ECA EAP SA DB DC NDC PF Source: Pallares-Miralles and others forthcoming. Note: ECA = Europe and Central Asia; EAP = East Asia and Pacific; LAC = Latin America and the Caribbean; SA = South Asia; SSA = Sub-Saharan Africa; DB = defined benefit; DC = defined contribution; NDC = notional defined contribution; PF = provident fund. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 281 retirement age in line with changing life expectancy, and (3) uses an average wage measure in the benefit formula, which incorporates the ­ wage level of all contributory periods in a revalorized manner.24 However, the mandatory DB schemes in MENA do not meet these criteria. In addi- tion to the problem of financial sustainability, MENA’s DB schemes have flaws in many of the critical design features, such as the minimum vesting period, early retirement provision, and average wage measure in the benefit formula, that provide disincentives to participate in the system. ­ Minimum vesting period: Minimum vesting periods can promote informality. Table 5.7 summarizes some of the design parameters of ­ mandatory pension schemes in MENA. With the exception of the ­ recently modernized Iraqi pension system, none of the featured MENA countries require individuals to contribute for longer than half of a potentially 35- to 45-year active life cycle (that is, between a labor market entry age of 15 to 25 years and retirement at the age of 60) to ­ ­ ensions. Primarily, a short vesting period causes either qualify for full p unsustainability or inadequate income protection. In many cases, the length of the minimum vesting period is close to or less than the length of time for receiving pension benefits, even without taking advantage of early retirement provisions. Theoretically, this can be sustainable only at prohibitively high contribution requirements or very low benefits. Low benefits would not protect against poverty in old age, so that the protec- tion offered is not adequate. Another consequence of a short vesting period is that individuals provide only what is required from them in terms of system participation and exert a significant portion of their income-generating effort in some form of informality hidden from the social insurance system. The best strategy for workers is to limit the length of contributions to the minimum necessary to draw full benefits (see further below). Gaming the pension system with strategically short- ened life-cycle contributions not only undermines sustainability but also increases informality. Conversely, a long minimum vesting requirement would exclude from pension eligibility those who cannot contribute suf- ficiently long in the formal sector. Under DC schemes, no need for a required minimum vesting period is required, and thus such schemes can be considered more “coverage friendly” than DB schemes. With funded DC or NDC25 pension designs, the saving effort during the active life cycle is measured only by the actual amount of the saving. Furthermore, the earnings-related DC mechanism could be augmented with a minimum old-age income guarantee feature (see the summary in the previous section), and the eligibility of the Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 282 Table 5.7 Pension Design Parameters of Mandatory Formal Sector Schemes in MENA Retirement Age, ​ Minimum Vesting Period, and Early Retirement Provisions Life Statutory Minimum expectancy retirement age vesting period Early retirement age Vesting for early retirement age at birth Female Male Female Male Female Male Female Male Algeria 73.5 55 60 10 15 any age/45 any age/50 32/14 32/20 Djibouti 58.5 55 55 — — any age any age 25 25 Egypt, Arab Rep. 73.5 60 60 10 10 any age any age 20 20 Iran, Islamic Rep. 73.3 55 60 19 19 any age/45 any age/50 35/30 35/30 Iraq 70.2 60 60 25 25 55/50 55/50 25/30 25/30 Jordan 73.6 55 60 15 15 50 50 22 25 Libya 75.1 60 65 20 20 — — — — Morocco 72.5 60 60 14 14 55 55 if employer pays if employer pays Syrian Arab Republic 76.1 55 60 15 15 any age/50 any age/55 25/20 25/20 Tunisia 74.4 60 60 10 10 50 50 30 30 Sources: Pallares-Miralles and others forthcoming; SSA 2009, 2010. Note: — = not available. Barriers to Coverage and Policy Options 283 ­ ensioner for a subsidy could be determined based on the annuity level p he or she receives given his or her saving effort and age. Such a scheme would avoid gaming behaviors that are prevalent under the vesting period requirement of DB design. The minimum vesting period requirement also puts upward pressure on informality through another channel. Jobseekers with access to temporary formal employment who are u ­ ncertain whether they can maintain formal employment status for the minimum required DB vesting period may decide to stay informal since any contributions could eventually become a sunk cost to them.26 Conversely, under DC schemes, no sunk cost risk is found for engaging in formal sector employ- ment for those who are uncertain about the stability of their formal employment status. Overall, DC design can be considered as the desirable “coverage friendly” system benchmark (as long as it is augmented with some minimum old-age income guarantee) under which there is no need to use proxies for contribution performance that are prone to be abused. The path dependency reality suggests that DB will continue to be the dominant design in MENA in the foreseeable future. However, it is pos- sible to incorporate features into a DB design pension to approximate the advantages of a DC scheme, although doing so is more difficult than tak- ing advantage of the DC design’s simplicity. For example, under a still uncommon but theoretically interesting DB design solution, the flaws associated with a minimum vesting period could be overcome. The requirements for drawing full pensions or accessing the minimum p ­ ension top-up could be conditioned on age and the true value of past contribu- tory effort (see box 5.6 on recent Jordanian pension reform). Early retirement provisions: Generous early retirement provisions cause problems similar to those of the short minimum vesting period by radically undermining financial sustainability and encouraging people to reduce the length of their formal sector employment. ­ Providing a degree of flexibility in determining the actual individual age of retirement is a logical feature of pension systems. In good practices of pension design, such flexibility is achieved through an actuarially fair adjustment of benefits. In particular, those who decide to retire before the statutory retirement age should compensate the pension system, in the form of reduced pension benefits, for their shorter contributory period and their longer term of benefit receipt. An upward adjustment should also be allowed for those who delay retirement beyond the statu- tory age and therefore contribute longer and receive benefits for a shorter period. MENA pension schemes are lax in defining the age at which individuals can start drawing benefits, and it is the exception rather than Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 284 Barriers to Coverage and Policy Options Box 5.6 Social Insurance Reform in Jordan In 2010 the Jordanian parliament passed a complex social insurance reform law. First, the reform implements a parametric readjustment of the mandatory DB pension pro- gram of the Social Security Corporation (SSC) to improve sustainability and incentives in line with the recommendations of this report, including the reduction of benefits in cases of early retirement. This is, in effect, a degree of approximation toward a DC design. Second, it extends the scope of social risk management by introducing mater- nity benefits and an unemployment insurance saving account mechanism, while also declaring health insurance as the next phase of the scope extension. Third, the reform extends full social insurance coverage to workers of small firms, in particular, firms with fewer than five employees; previously these workers were not covered due to the coverage assignment in accordance with firm size. This coverage extension effort is being supported by public communication activities and implementation efforts that make it easy for small firms to register and do business with the SSC, which has so far been used to collaborate with firms possessing their own human resource depart- ments to serve as counterparts in administering social insurance contributions. Source: Razzaz 2011. the rule that the actual benefit amount is less than that defined by the benefit formula.27 In many MENA countries, the actual age of retirement is not constrained at all. For example, in Algeria, males must complete 32 contributory years; in Djibouti, 25; in Egypt, 20;28 in the Islamic Republic of Iran, 35; and in Syria, 25. An individual who starts a formal sector job at the age of 20 could retire with a full pension at the age of 40 in Egypt or 45 in Syria, in other words, at ages still suitable for economic activity. Many of these individuals do decide to start drawing pension at this age, stop contributing to the pension system, and seek other informal employ- ment income while officially retired. These “noncontributing active age pensioners” are then counted as uncovered in widely used coverage ­ definitions. Even when the pension system actually imposes an age floor, the set ages are still considered suitable for economic activity; for exam- ple, in Iraq, Jordan, and Tunisia, the floor is age 50 for men. The flip side of this practice is that no incentive is given for delayed retirement. To the contrary, mandatory retirement-age rules are set at or close to the statu- tory retirement age, especially in the public sector. It is likely that this practice is related to the bias toward public sector employment. Without mandatory retirement, many would “overstay” in their relatively Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 285 ­ ell-compensated and low-effort public sector jobs. Allowing early w retirement for the public sector–dominated formal employees may be viewed as making room for the many young graduates who queue for public sector jobs, a situation exacerbated by the ever lengthening queue associated with the youth bulge and expansion of the higher education system. High informality rates among those 55 and older have been observed in the data used herein. If these are early retirees, they are already covered by pensions, and the vulnerability argument does not apply to them. Average wage measure in the DB pension formula: Another design flaw of MENA pension schemes is that the wage measure used in the DB formula does not represent full career wages; this feature provides incen- tives to report only high income in the years before retirement. If the average wage measure used in the DB pension formula represents all contributions throughout the active life cycle where past wages are revalorized to the changes in the average wage levels over time, then DB systems could have desirable incentive effects similar to those of DC systems. Table 5.8 summarizes the average wage measures applied in the national DB schemes of MENA; they tend to use an unrevalorized aver- age wage measure of the last few years before retirement. As a result, individuals tend to avoid reporting low earnings in these critical years and artificially inflate wages during the same period through side agreements with the employer. The average wage measure effects of the DB MENA pension schemes are stronger from a sustainability or a general equity perspective but can also contribute to gaming of earnings for reporting contribution purposes. What are optimal life-cycle pension system participation strategies? Pension systems in MENA are extremely generous and have a high internal rate of return for contributors. Given the existing design, indi- viduals often have the incentive to contribute for the minimum period required and still draw generous pensions once retired. MENA pension schemes are excessively generous and, not surprisingly, financially unsustainable. For anyone able to find formal employment and regis- tered with the pension administration, it is worthwhile to enroll in the pension system, although the optimal strategy is not necessarily to contribute all the time. Figure 5.13 shows that in Egypt and Jordan ­ (before the recent reforms) and in Syria, contributing to the pension system yields real returns in the ranges of 6 to 10.5 percent, 7 to 17 percent, and 7 to 14 percent, respectively, much higher than the risk- less 2 to 4 percent real return one could achieve through saving outside Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 286 Barriers to Coverage and Policy Options Table 5.8  Average Wage Measures in the Pension Benefit Formula in Selected MENA Countries Country Average wage measure Algeria Maximum of the average of the wages of five last years and the average of five best years without revalorization Djibouti average of the wages of the last 10 years/last month’s wage Egypt, Arab Rep. convoluted combination of unrevalorized average basic wage of the last two years and unrevalorized full career average variable wages Iran, Islamic Rep. average of unrevalorized wages in the last two years Iraq gradually expanding to revalorized full career average wages in public ­ sector/unrevalorized average of last three years of wages in private sector Libya average of unrevalorized wages in the last three years Syrian Arab average of last year’s unrevalorized wages with wage increase not to exceed ­ Republic 15 and 30 percent in the last two and five year periods, respectively Tunisia average revalorized wages in the last 10 years in main scheme; different rules in four other national schemes Yemen, Rep. last month’s salary Sources: Pallares-Miralles and others forthcoming; SSA 2009, 2010. the pension system.29 Figure 5.13 also illustrates that returns and con- tribution length are inversely related under the flawed design of MENA DB pension schemes. The highest return is associated with just completing the minimum vesting period necessary to draw full pen- sions, with contributions completed either early in the career (for example, a “frontloading strategy,” taking advantage of the generous early retirement provisions) or very late in the career, to raise the aver- age wage measure with the highest wages reached during a career (for example, a “backloading strategy”). Note that limiting the number of contributory years, if they are nearly evenly distributed between the years of labor market entry and the statutory retirement age, does not increase the return on contributions; the constant contribution density strategy (that is, a “homogeneous life-cycle strategy”) is suboptimal from the perspective of the individual. If individuals are in a position to move income between the formal and informal sectors, or if they strategically plan their life-cycle formal employment participation behavior, they further undermine the sustainability of the pension schemes and increase observed informality. The patterns of the derived optimal contribution strategies can explain some of the coverage pat- terns observed at the aggregate level. Reforms toward DC pension design (regardless of the financing mechanism) or improvements in DB scheme design would address the Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 5.13 Internal Rate of Return under Various Pension System Participation Strategies in Egypt, Jordan, and Syria 11 Egypt, Arab Rep. 19 Jordan 10 17 9 15 8 13 percent percent 7 11 6 9 5 7 4 5 3 3 20 22 24 26 28 30 32 34 36 38 40 42 44 18 20 22 24 26 28 30 32 34 36 38 40 42 44 contribution length in years contribution length in years frontloading contribution history backloading contribution history homogeneous contribution history 15 Syrian Arab Republic 13 11 percent 9 7 5 15 16 17 18 19 20 21 22 23 24 contribution length in years homogeneous contribution history backloading contribution history Source: di Filippo and Bodor forthcoming. 287   288 Barriers to Coverage and Policy Options adverse incentives of the current MENA pension schemes. Figure 5.14 illustrates the changing internal of rate of return profile under the recently legislated NDC of a funded DC pension reform in Egypt and a hypothetical DB reform in Syria. The Egyptian legislature passed a complex pension reform in 2010, which changed the PAYG pension scheme from DB to NDC design, and introduced a smaller funded DC mandatory pension component, as well as a universal basic pension for all, while reducing the tax wedge imposed by the pension system. One of the goals of this reform was to reinstate the pension system’s finan- cial sustainability, so it is not a surprise that the current system’s extreme generosity is reduced (as evidenced by the lower internal rate of return). It is also notable that the rate of return differential across various contribution lengths and strategies of “frontloading,” “backload- ing,” or keeping contribution density nearly constant is diminished; the pension system no longer provides incentives to work informally dur- ing the work career. Similar results can be achieved under a well- designed DB architecture as well. The second panel of figure 5.14 shows that the payoff for short career contribution (and system par- ticipation) strategies disappears if an actuarially fair benefit reduction for early retirement is introduced and is combined with an average wage measure that represents all career contributions in a revalorized manner. (These reform features are combined with reduced accrual rate,30 increased retirement age to improve sustainability, and a reduced contribution rate/tax wedge to reduce the system’s burden on the economy.) By design, pure DC pension schemes promote desirable incentives and are financially sustainable, yet on their own they cannot provide adequate income protection in old age for those who could not save enough during their active life period. As the name “defined contribution” suggests, ben- efits are solely dependent on the contributions/savings. The implications of this design are especially problematic among those who save too little, not because of a lack of effort, but rather because of labor market condi- tions and the marketability of their skills. That is why DC pension schemes are almost always augmented with some feature guaranteeing a minimum income level in old age. In this case, the minimum guarantee could be conditional on contributory effort rather than the value of con- tributions. Moreover, the income guarantee could be a top-up subsidy to augment the saving in the DC scheme, a broader noncontributory social pension, or a combination in the form of a social pension claw-back based on the DC saving. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 5.14 Internal Rate of Return Patterns of the Egyptian Pension System before and after Recently Legislated Pension Reform and of Hypothetical DB Reform in Syria Egypt, Arab Rep. Syrian Arab Republic 9 15 8 13 7 11 percent percent 6 9 5 7 4 3 5 30 32 34 36 38 40 42 44 15 16 17 18 19 20 21 22 23 24 contribution length in years contribution length in years frontloading contribution history homogeneous contribution history backloading contribution history backloading contribution history homogeneous contribution history backloading contribution history (reform scenario) frontloading contribution history (reform scenario) homogeneous contribution history (reform scenario) backloading contribution history (reform scenario) homogeneous contribution history (reform scenario) Source: di Filippo and Bodor forthcoming. 289   290 Barriers to Coverage and Policy Options The discussion in the previous subsections on the various distortions created by the design features of MENA DB pensions implies that the observed informality in the form of the coverage rate is likely to ­ overestimate vulnerability, at least when it comes to access to old-age benefits. Box 5.7 discusses how the coverage rate overestimates the informality-related vulnerability problem when it comes to accessing social insurance benefits that are based on life-cycle contributions, even without strategic behavioral responses to the rules of the social insurance (pension) systems. The flawed DB design of many of the mandatory MENA pension schemes further contributes to reducing Box 5.7 Coverage Rates and Vulnerability Due to Informality The social insurance coverage rate (that is, the share of the labor force contributing to the mandatory social insurance scheme at a point of time) adequately signals vulnerability when it comes to benefits conditional on current enrollment (such as health insurance) but overestimates the vulnerability for social insurance benefits based on lifetime contributions, such as pensions. The typical DB pension scheme in MENA sets minimum contributory length requirements for eligibility to draw full pensions (called the minimum vesting period). ­ Consider a hypothetical case in which the minimum vesting period is 20 years, equivalent to roughly half of a 40-year active life period. An observed 50 percent coverage rate (that is, half of the labor force is contributing at any point in time) could represent two extreme sce - narios: (a) half of the labor force contributes 100 percent of the time and the other half never contributes and (b) 100 percent of the labor force contributes half of the time. Clearly, the underlying implications for vulnerability are different in the two scenarios. Under the first scenario, half of the labor force has access to pension ben- efits, and the coverage rate is accurate and informative about access to benefits and about vulnerability. Under the second scenario, the entire labor force has access to pensions, but this is not reflected in the estimate of coverage. (Note that with a minimum vesting eligibility period just a bit longer than 20 years, none of the labor force would have access to full pensions.) The true vulnerability lies in between these extreme scenarios for social insurance benefits, which are based on life-cycle contribution (saving) performance. For other benefits, current enrollment matters for eligibility; in these cases, the coverage rate is a good direct indicator of vulnerability. Note that this illustration explains how the observed snapshot coverage rate overestimates the true vulnerability associated with informality regardless of strategic (endogenous) behavioral responses to the rules of social insurance provision. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 291 the observed coverage rate and increasing the gap between the informality measure provided by the coverage indicator and the actual ­ access to old-age income benefits. It should also be stressed that this usually comes at the cost of financial sustainability and, as such, has implications for intergenerational equity. Beyond pensions. In the previous section, the primary focus was on how pension system rules affect informality outcomes in detail, but a variety of other social insurance risk management mechanisms may have similar effects. With the notable exception of Lebanon, mandatory national pension schemes in MENA offer old-age pension benefits in the form of ­ pension payments until death. The old-age pension benefits are almost always augmented with survivor and disability pension benefits. Eligibility for survivor benefits in MENA tends to be generous; significant anecdotal evidence suggests household-level strategic planning around fully utilizing survivor pensions. Work injury coverage, almost always bundled with old- age pensions, can be more extensive than that of old age. For example, in Syria, registered workers of firms with fewer than five employees are covered only for this risk. Work injury coverage for private sector ­ workers in Morocco is mandatory (a unique case), but employers need to purchase coverage for their employees from competing private insur- ance providers. A survey from Syria31 suggests that employees value this social insurance component more than others, such as family or survivor- ship benefits, but less than old-age benefits and perceived job security due to coverage. In any case, it is unlikely to significantly drive informal- ity. Workers’ behaviors are more likely to be affected by health insurance coverage, but in MENA, broad access to health insurance is ensured through public provision of heavily subsidized but often low-quality health services, and privately or publicly provided health insurance benefits typically augment this or offer access to higher quality care (for ­ example, as in Iraq for private sector scheme members). Last, UI schemes in MENA rarely meet the international standards of UI provision, are hardly effective in ensuring the unemployment risk, and are unlikely to drive social insurance system participation decisions. However, recent UI reforms in Jordan and Egypt, which not only offer services to those laid off but also return the savings not used during periods of unemployment to the individual upon retirement, have the potential to make this a valued social insurance component. Bundling or debundling coverage for the various social insurance components may influence informality outcomes. The bundling of ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 292 Barriers to Coverage and Policy Options ­ enefits under a single social insurance package may exert downward b pressure on coverage through mandating contributions for components widely viewed as less valuable. The Syria employer-employee survey showed that old-age coverage is significantly more valued than any bundled components, although health insurance is not provided through social insurance. On the other hand, in Lebanon, where old-age coverage in the traditional sense of pensions does not exist (only a lump-sum end- of-service indemnity), the health coverage component is viewed as the most valuable. The “price elasticity” of coverage is dependent on what services are provided under the coverage umbrella. If employees and employers are forced to pay for social insurance components that are less valued because of the bundled access to social insurance services, par- ticipation may be reduced under the formal sector social insurance scheme. At the same time, bundling of important social insurance com- ponents that are not valued by some workers because of myopia and the lack of immediate benefits (such as old-age income security compared with health insurance whose benefits could be immediately experienced) may improve coverage of such insurance types. Often benefits that are not associated with social risks are also financed through the social insur- ance mechanism, which increases the cost of participation without underlying theoretical justification. For example, in Syria, heating subsi- dies and family allowances are financed through the social insurance contributions of employees and employers. Such social benefits may be legitimate social benefits categories in a particular country context, but no reason exists to mix their financing with the financing mechanisms of protection against the realization of social risks such as old age, disability, or death of income-generating family members. Social support toward families is not an insurance service per se; this would be better matched with general revenue financing. In summary, the set of risks covered by the social insurance mecha- nism should be linked to the underlying exposure to social risks and productivity of formal sector employment, so that a bundle of protection against the truly relevant risk is provided at an appropriate cost. Such bundling decisions should be revisited as MENA governments are now considering introducing unemployment insurance to protect against lay- off risk in formal sector employment. This was less of a relevant risk in the past, but one that is getting increasingly more relevant given the private sector–driven growth path. It should be noted, however, that the limited existing evidence does not support the hypothesis that bundling significantly affects the participation decision in Morocco. There the Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 293 chance of having a formal job was found to increase if another member of the family had a formal job. With health insurance extended to spouse and dependents, the “bundling” hypothesis would generate a negative correlation of formal jobs within the household. Instead, an explanation relying on networks and privileged access to formal jobs seems more plausible. The coverage of social insurance (pension) systems in MENA is historically low. Coverage extension is required to provide tools to indi- ­ viduals and households to protect against various critical social risks including, but not limited to, health and unemployment risks and dimin- ishing income-generating capacity in old age. Well-designed social risk management mechanisms increase welfare, but it is crucial that their design ensures adequate, affordable, and sustainable benefits in a manner that offers coverage to a large and increasing share of the population, without causing massive economic and labor market distortions. The first step in coverage extension is bridging the legal coverage gaps: coverage is rarely offered, for example, to the self-employed or agricultural workers. Beyond that, rethinking the design of existing MENA pension schemes should be a precondition for further coverage extension efforts. Vesting requirements, early retirement provisions, and average wage measures used for pension benefit calculation all contribute to reduced observed coverage. In general, the MENA pension systems provide high returns to those who have the chance to enroll, but paradoxically, this return can be made even greater by reducing the contribution length to a minimum and by taking advantage of the generous early retirement or by gaming representation of career wage by wages immediately before retirement. In addition to improving sustainability of the pension system and protecting against a wider array of social risks, social insurance reform is critical for addressing informality in MENA. Important political economy aspects to coverage extension can be identified. Given the generosity of existing systems and their financial unsustainability, current contributors have the incentive to extend coverage now to stave off reform. Without reestab- lishing financial sustainability first, extending coverage under the same conditions of existing pension systems is likely to magnify considerably the financial sustainability problem (box 5.8). Nevertheless, even if the existing formal sector pension schemes were reformed to match best practices, full, or even near full, coverage would still be an issue in MENA. Because many of informal workers have limited ability to save, it is unlikely that coverage extension can be achieved effectively without special subsidized social insurance features. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 294 Barriers to Coverage and Policy Options Box 5.8 Political Economy Considerations for Coverage Extension The political economy of coverage extension and its interaction with pension reform matters significantly. For example, the true intent behind certain coverage extension efforts may be motivated by the need to temporarily improve the financial sustain- ability of otherwise unsustainable PAYG pension schemes. Particularly, DB pension schemes with PAYG financing are often financially unsustainable due to the political economy of pensions: The future generations who inherit the burden of the “pension Ponzi scheme” do not participate in the political system. When the demographic transition in developing countries accelerates the aging process of the society and the benefits of those already in or close to retirement become harder to finance even on a PAYG basis, then the search for alternative funding sources is intensified in the political arena, at least in the short term. Paradoxically, the mass enrollment of young workers who used to be excluded from participating in the formal sector pension scheme offers such a temporary financing solution. New young contributors pay contributions to the PAYG scheme to cover pension expenditures in the near future. Of course, improving the current financial position of a pension scheme through cov- erage extension without reinstating the overall sustainability of the design is hardly even a temporary correction measure; in fact, such a move accelerates the accumula- tion of unfunded government liabilities. In the future, when coverage extension is no longer available as a temporary measure, the government will be forced to default on the pension promises; that is, it will reduce the value of benefits promised in exchange for enrolling in the system. Ultimately, chances are that coverage extension without realigned sustainability will treat those formerly informal in an unfair manner, as it shifts forward the burden of the financial sustainability of the system. Because of this, covered (formal) workers are likely to support coverage extension, because it ensures system sustainability in the short run and eases the pressure for pension reform (which would likely reduce the generosity of their benefits). In this sense, reestablish- ing the sustainability of pension (and broader social insurance) schemes is a precon- dition to functionally adequate coverage extension efforts. Extending Social Insurance Beyond the Existing Mandatory System: Guiding Tree with Application to MENA This section addresses how governments can develop coverage extension program strategies for their country-specific needs, assuming that the incentive design issues in existing earnings-related schemes are addressed in line with the discussion above. The following discussion acknowledges that informality is a persistent reality, and that access to innovative social Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 295 risk management tools needs to be developed and provided to the informal sector, as well as to those outside of the labor force. ­ Governments that opt to extend social insurance coverage beyond the existing mandatory social insurance system by targeting groups of informal workers are faced with a set of decisions related to strategy, design, and implementation of new schemes. Although modernization of social insurance systems is expected to contribute to improved coverage, certain groups are likely to remain vulnerable/excluded. For ­ ­ example, consider the case of workers with nontraditional jobs that have no easily identifiable employment relationship defined by a single employer, duration, frequency, and wage, and who experience high mobility between jobs and recurrent unemployment spells. It is diffi- cult, if not impossible, to mandate such workers to enroll under the formal earnings-related social security system. Moreover, some groups of informal workers are likely not to be able to afford any contributions required by the social insurance system and the transaction costs associ- ated with enrolling. Extending coverage to such groups of workers requires the introduction of one or more types of special coverage extension schemes. This section discusses the various decisions facing governments when introducing such schemes (figure 5.15). For exam- ple, schemes could be based on either a universal strategy, where all individuals in the broad population group are eligible to benefit from a ­ certain scheme, or on a targeted strategy, where only a portion of individuals within the broad population group is eligible, based on some ­ kind of means test or categorical differentiation. Additionally, the scheme might or might not require contributions on the part of the beneficiaries. Examples of contributory schemes include MDC schemes in which the matching for the employee’s contribution may be financed by the government or by employers. Noncontributory schemes include social pensions and free health insurance (figure 5.10 illustrates these types of schemes in the context of pensions). Governments may even consider introducing different schemes serving distinct homogenous social groups and/or different objectives. This section also presents guiding indicators that can be used to assess the relevance of different strategies in light of existing country conditions. The indicators are ­ chosen based on evidence from international literature and experience on different social insurance schemes. Based on the decision framework presented, this section also provides policy implications specific to MENA countries regarding the relevance of different types of coverage extension strategies. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 296 Barriers to Coverage and Policy Options Figure 5.15  Guiding Tree for Introducing Coverage Extension Strategies determine need for coverage extension to improve welfare of certain social groups assessment of existing SI system (gaps in systems and processes) extend coverage beyond existing mandatory improve existing contributory SI SI system system (incentives, enforcement, and legal coverage gap) targeted strategy universal strategy (means, proxy-means, categorical) noncontributory contributory noncontributory contributory universal targeted universal SI matching targeted SI matching bene ts (social subsidy bene ts (social subsidy pensions/HI) (MDC; pensions/HI) (MDC; government or government employer nanced) nanced) program implementation Note: SI = social insurance. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 297 A Framework to Assess the Existing Social Insurance System in Light of Coverage Extension Most countries are not starting with an empty slate when it comes to social insurance. The first step of decision making requires an assessment of the existing system with the objective of identifying possible reasons for low coverage outcomes resulting from implementation shortcomings. Palacios (forthcoming) presents a framework that can be used to identify gaps in social protection programs based on five main implementation “tasks,” presented as different layers of the pyramid in figure 5.16: (1) identification, (2) targeting, (3) enrollment, (4) benefit definition, and (5) management information systems. This framework suggests that a rough estimation can be made of the share of the eligible population that did not receive a certain social protection benefit due to shortcomings in each of the five main tasks. This then provides guidance on improving the implementation of existing programs as well as on the design and ­ implementation of new programs. Figure 5.16 also presents examples of indicators associated with each of the five tasks in the framework. A robust and integrated national identification system is an important prerequisite. In MENA, although a national identification (ID) system exists in all countries and national IDs are mandatory for adults, the robustness and integration of these systems across various public agencies and service providers vary among countries but remain under- ­ developed in most. Moreover, exclusion from the ID system is heavily dependent on income level. Morocco, the Republic of Yemen, and Djibouti have recently taken steps to develop a multiuse biometric national ID system. The national ID in Egypt is used for receiving various public services as well as for employment and some private transactions. In other countries, the use of national IDs is limited to election purposes. Overall, modernization of national ID systems is underway in the region, although completion may take time, and program-specific IDs can be used, with an outlook toward future development of an integrated unique ID. With respect to the second layer of the pyramid, targeting/ eligibility, the key question for governments is whether some people are being excluded because they cannot prove their eligibility or because of the quality of the targeting system. It is also important to determine if ineligible people are erroneously included as well. The upper levels of the pyramid consider the need to assess the procedures through which potential beneficiaries enroll in the system and evaluate the effectiveness of the benefit delivery processes and the accuracy and maturity of the management information systems. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 298 Figure 5.16  A Framework for Implementing Social Protection Programs Example of indicator ID National Identification System Robustness MIS National Identification System Integration Benefit T existence and quality of (proxy) means testing delivery (BD) National Identification System Integration E bureaucratic transaction costs Enrollment (E) legal constraints to enrollment strength of enforcement Targeting/eligiblity (T) Internet/mobile penetration BD access/availability of providers (such as Identification (ID) hospitals, local offices, banks) Internet/mobile penetration Source: Based on Palacios forthcoming. Barriers to Coverage and Policy Options 299 Extending Coverage beyond an Existing Mandatory Social Insurance System Existing country conditions can serve as a good guide for governments to assist in making design decisions when introducing coverage exten- sion strategies. Current country-specific conditions (such as the level of income of the target group, existing coverage patterns, government effectiveness, and effectiveness of existing programs and their support- ing infrastructure) can be used to assess the relevance of each type of scheme in a specific country context. Various guiding indicators are collected in the score card presented in table 5.9. It summarizes considerations and examples of relevant indicators that can be used to ­ evaluate existing country-specific conditions for guiding decisions on coverage extension strategies. The conditions are described in turn below. It is worth mentioning that although current country conditions will ultimately guide the decisions to be made when introducing new Table 5.9 Existing Country Conditions Guiding Indicators (Score Card) Universal vs. Targeted strategy Universal Targeting 1. Level of income and poverty in the country A. GDP per capita + B. Share of rural population + C. Share of “no” and “low” savings capacity population + D. Incidence of poverty among elderly + Effectiveness of existing social assistance programs 2.  and systems  xistence and effectiveness of social assistance E. E + programs  xistence and quality of means or proxy means testing F. E + 3. Government’s ability to reach different population groups G. Mobile/Internet penetration + H. Local government presence/strength + Noncontributory vs. Contributory Scheme Noncontributory Contributory Level of income and poverty in the country/target 4.  population A. GDP per capita + B. Coverage rate + C. Share of “no” to “low” savings capacity population + 5. Transaction costs (bureaucratic transaction costs) + 6. Trust in government (government accountability) + 7. Financial literacy + Note: GDP = gross domestic product. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 300 Barriers to Coverage and Policy Options schemes, governments’ future objectives and the projected enabling environment of a country, as envisioned by various reforms, should be taken into account. For example, although current country conditions may motivate the introduction of temporary programs or temporary design features, a different form of the system may be anticipated under more developed conditions. The key is that design features applied under current conditions should not serve as constraints toward future modernization. Deciding between a universal or a targeted strategy:      Level of income and poverty: The richer the country, the more likely (1)  that a universal strategy will result in higher inclusion errors; that is, more middle- and high-income individuals with no need for govern- ment support are likely to benefit from the scheme. This group would otherwise be excluded from a means-tested targeted scheme. Thus, a universal scheme would be both more regressive in nature and more costly in richer countries. In addition, because income is more likely to be observable in richer countries, both inclusion and exclusion errors will be less significant under a targeted scheme. Indicators to ­ consider under this category include GDP per capita, the share of rural population, and the savings capacity of the population. Finally, with respect to old-age benefits in particular, a higher share of elderly among the poor strengthens the argument for universal schemes.32 Effectiveness of existing social assistance programs and systems: In (2)  countries with well-established, effective, and high-coverage social as- sistance programs, a universal scheme implies that individuals are likely to receive multiple benefits. Moreover, effective social safety net programs often rely on well-developed targeting systems, on which governments can depend when introducing a targeted coverage exten- sion strategy. Relevant indicators to be considered under this category include scope, coverage, and impact of social assistance programs and the existence/quality of means or proxy means testing systems in the country (see box 5.9 for a discussion of targeting methods). Government’s ability to reach different population groups: In less (3)  urbanized countries with little government representation in ­ rural and remote areas, a targeting strategy may be very costly. ­ Examples of indicators that can be considered when assessing the government’s ability to reach different population groups include local g­ overnment p resence/strength and mobile or Internet p ­ ­ enetration rates. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 301 Box 5.9 Targeting Methods The ideal mechanism to target programs to low-income households is determining eligibility to a given program based on the household’s level of income and available assets.a Means testing implies that benefits could go directly to those most finan- cially  in need. It requires the collection and verification of household income information, a costly exercise whose credibility could be threatened by the underesti- ­ mation or misreporting of income by households, especially by the self-employed. Another mechanism of targeting based on means that is increasingly being used in the developing world is proxy means testing (PMT). Under proxy means testing, individuals and/or households are assigned a score based on a number of household indicators that serve as a proxy for income and assets. The method requires the development of a regression formula that relates easily observed household ­ characteristics to income (or expenditures). Common variables include number of household members and chil- dren, location, quality of housing, ownership of assets, and utility bills. Once individuals apply to a certain program, such information is usually collected through household visits, although in more developed targeting environments, a detailed national data- base such as the census could be used to ­ create a proxy means score independently of applications to a specific program. This method can be less expensive than means testing and subject to less underreporting. However, the effectiveness of the overall ­ mechanism depends on a strong ­ correlation between income and easily observed household and individual ­ characteristics. In addition to targeting based on means or proxy means, some programs deter- mine eligibility on the basis of specified geographic, socioeconomic, and demo - graphic characteristics that are correlated with poverty (for example, age, number of children, disabled status, single parenthood, unemployed status, and geographic location). Such methods are called categorical and geographic targeting and are associated with relatively low administrative and economic costs, although they are more likely to include the nonpoor. Moreover, the success of geographic targeting depends heavily on an accurate and detailed poverty map of the country. Source: Silva, Levin, and Morgandi 2012. a. For a discussion of concepts, methods, results, and implementation of targeting strategies, refer to chapter 4 in Grosh and others (2008). ­ lthough a higher share of rural population may make it more A expensive and ­ ­ difficult to use a targeting strategy, the cost may be offset in varying degrees if the majority of people can be reached through mobile phones or the Internet and if the government has a ­ resence. strong local p Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 302 Barriers to Coverage and Policy Options Besides country conditions, the fiscal implications of each scheme are also important. The average cost of a noncontributory pension scheme is estimated at about 1 percent of GDP, varying between universal and targeted schemes. For example, a means-tested scheme in Bangladesh costs 0.03 percent of GDP, whereas a universal scheme in Mauritius costs 2 percent of GDP. For the MENA region, Palacios and Sluchynsky (2006) estimate that the cost of a universal social pension that provides a benefit equal to 15 percent of GDP per capita at a retirement age of 65 would reach 1 percent in 2025 and 1.5 percent in 2040, depending on demo- graphic evolutions. The cost of a universal scheme may be considered unaffordable for countries with a low tax base and revenues and an aging population. Moreover, capping the number of eligible beneficiaries, restricting benefits to one individual per household (as in Bangladesh), and disqualifying persons who receive other forms of assistance could also contain costs. Countries with means tested social pensions include Bangladesh, rural Brazil, Chile, Costa Rica, India, Moldova, Nepal, South Africa, and Thailand. Deciding between a contributory and a noncontributory scheme: The choice between a contributory and a noncontributory scheme is, to a large extent, dependent on the ability and willingness of target social groups to contribute to a particular scheme and the extent to which their incentives to do so can be influenced. It is in this context that the follow- ing factors are evaluated (recall table 5.9). Level of income and poverty in country/target population: Indicators (1)  including GDP per capita, coverage rate of existing contributory schemes, and the ratio of “no” to “low” savings capacity population can be considered. If a large portion of the target population has no sav- ings capacity, the take-up rate of a contributory scheme is likely to remain very moderate even when the government or employers offer to subsidize a portion of contributions. Moreover, if coverage of exist- ing mandatory schemes is low (for example, below 20 to 25 percent), incentives are likely to play a less significant role compared with situ- ations where coverage is higher (for example, above 50 percent). GDP per capita and coverage rates are correlated, and so this is in line with the argument presented earlier. Transaction costs: The schemes under consideration are by nature vol- (2)  untary and require beneficiaries to make a conscious choice to enroll, knowing that they will have to incur some transaction cost related to simply registering or also making contributions.Thus, it is very ­important Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 303 understand how to consider the level of transaction costs involved and ­ they will influence people’s incentives to enroll. If high transaction costs cannot be overcome or mitigated, noncontributory schemes may achieve higher coverage rates than contributory ones. Trust in government: People’s trust in the government and its institu- (3)  tions can influence their decision to enroll and contribute to a scheme. Financial literacy: Contributory schemes require some financial un- (4)  derstanding or computation from the individual to assess whether it would be cost effective to contribute. In this context, if financial lit- eracy among the target population is low, it may be difficult to achieve high enrollment rates in contributory schemes.33 Examples of different schemes: Core and supplementary schemes are the two main types of noncontributory schemes that have emerged interna- tionally. In the case of social pensions, core schemes can be found in developing countries such as South Africa, Bolivia, Botswana, Mauritius, Kosovo, and rural Brazil. In this context, social pensions are used as the main tool for addressing the coverage gap, in some cases covering up to 100 percent of the elderly. Larger core schemes are more likely to be found in countries with limited coverage of mandated contributory schemes or no such schemes at all. On the other hand, supplementary social pension schemes are intended to assist the elderly poor who are excluded from the formal system; these are more likely to exist in coun- tries with moderate to large contributory schemes, such as Algeria, Egypt, Colombia, Turkey, Costa Rica, and Uruguay. MDC schemes are emerging as a potentially viable option for expanding coverage to informal workers. Although MDCs exist in some developed countries such as the United States and Germany, laws estab- lishing MDC schemes have been only recently passed in some middle- and low-income countries, including the Dominican Republic, Korea, China, India, Indonesia, Mexico, Vietnam, and Thailand. Yet only a few countries have recently launched implementation (see box 5.10 for a description of MDC schemes in India, Mexico, and China); thus little systematic evidence is available on their performance and impact. Robalino and Palacios (2009) argue that MDC schemes have numerous advantages for informal workers, namely, their portability across differ- ent jobs and their ability to accommodate flexible levels and frequency of contributions. Moreover, MDCs may be a more relevant tool to encourage enrollment among informal workers in middle- and low- income countries compared to other monetary incentives, such as tax Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 304 Barriers to Coverage and Policy Options Box 5.10 The Design of MDC Schemes in India, Mexico, and China MDC schemes in India, Mexico, and China exist as universal voluntary pension schemes. In the three countries, they are not stand-alone schemes but are instead plugged into existing umbrella schemes that leverage their infrastructure. The Indian Swavalamban a scheme is one of the most recent initiatives to motivate workers to enroll in pension systems. The scheme was rolled out nationwide in late 2010 and targets eligible informal sector workers who joined the national New Pension Scheme (NPS) in 2010 and 2011. The NPS is the country’s mandatory DC pension scheme for central government employees and is voluntary for all others. Under Swavalamban, the government provides a yearly matching of Re 1,000 ($25) for three years to every new account, conditional upon members individu- ally contributing between Re 1,000 and Re 12,000 yearly. Workers not employed by the government and not covered by any of the occupational provident funds in the country are eligible for the match. It is expected that the scheme will become permanent beyond the anticipated three years. The government has out- sourced the management and administration of the scheme to private intermedi- aries, including microfinance institutions, NGOs, and ­ others. The government plays a regulatory role and finances the match. It provides incentives to interme - diaries based on performance, that is, the number of members they enroll. The Mexican “Cuota Social” is an older scheme, introduced as a component of the Mexican Institute of Social Security reform package in 1997. A matching contri- bution is paid by the government at the rate of 1 peso per day, adjusted for inflation, disbursed on each day that an individual makes a contribution. Although the Cuota Social is not targeted, it is redistributive in nature because it is a flat rate. While the Indian and Mexican examples entail governments subsidizing con- tributions of new members, the widely recognized New Rural Pension Scheme introduced in China in 2009 subsidizes pension benefits. The government pro - vides a basic monthly contribution of Y 55 ($8) to retirees conditional upon having at least 15 years of contributions and a minimal annual contribution equivalent to 4–8 percent of the country’s average personal income during the previous year. Additional contributions are matched by the government. Although the scheme is quite specific to the context of rural China, one interesting design feature rele - vant to coverage extension is that the scheme still offers the match to retirees who have not accumulated enough savings, on the condition that all family members above the age of 16 enroll in the new scheme and contribute on their behalf. Moreover, although the scheme is universal, in the sense that all rural populations of China are eligible, the design entails an implicit geographical targeting, because box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 305 Box 5.10  The Design of MDC Schemes in India, Mexico, ​and China (continued) local governments in more affluent regions are encouraged to pay higher benefits and could require higher contributions on the part of their beneficiaries.b Source: Robalino and Palacios 2009. a. More information on the India New Pension Scheme and the Swavalamban scheme can be found at http:// www.pfrda.org.in/. b. For more on the Chinese New Rural Pension Scheme, see Shen and Williamson (2010) and Zhu (2009). deductions or exemptions. First, lower-income informal workers are less likely to pay taxes in lower-income countries, and, second, tax incen- universally, may be regressive. When MDCs are struc- tives, if applied ­ tured as a flat co-­ contribution by governments, they cease to be regressive. Furthermore, evidence on the magnitude by which tax incen- tives affect enrollment and savings levels is mixed, and a debate exists around whether savings induced by tax incentives crowd out existing savings (Antolin and Ponton 2007). The level of the matching contribution and the elasticities of the take- up and savings rate to the match are critical in determining the success of an MDC scheme in expanding coverage. The take-up rate is affected by both the level of the matching contribution and its magnitude relative to the individual’s income level. Robalino and Palacios (2009) present a method to determine the parameters and costs of an MDC scheme. However, ex ante evidence on the take-up/match elasticity, a critical fac- tor affecting the take-up of MDCs, is fairly limited, and the only evidence comes from U.S. experience in employer-matched 401(k) and individual retirement account (IRA) schemes. For example, a randomized field experiment offered matching incentives for IRA contributions in low- and middle-income neighborhoods in St. Louis, Missouri, at two levels (20 and 50 percent), while a control group was not offered any match. The higher match significantly affected both participation and contribu- tions to the scheme. Take-up rates among the 20 percent match group (8 percent) were almost three times that of the control group (3 percent) and reached 20 percent in the 50 percent match group. Savings were also four and seven times higher in the 20 and 50 match groups, respectively, compared with the control group (Duflo and others 2005). Not all stud- ies are conclusive about the power of matching with respect to increasing participation. Engelhart and Kumar (2006) used administrative data on 401(k)34 participation and estimated that the elasticity of 401(k) Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 306 Barriers to Coverage and Policy Options ­ articipation with respect to the match rate ranged between 0.02 and p 0.07; a 25 cent match per dollar of employee contribution resulted in an increase of 3.75 to 6 percent. The savings/match elasticity was also esti- mated to be modest. Under contributory schemes, some design features can further improve incentives of target groups to enroll. Allowing flexible contribu- tions and withdrawals are two potentially useful features to encourage take-up rates (Hu and Stewart 2009; Robalino and Palacios 2009). For types of work with volatile income (as in the agricultural sector or other seasonal sectors), it may be important to relax the rules on the regularity contributions. In Chile, participation of seasonal and temporary work- of ­ ers is encouraged by allowing irregular frequencies of contributions. The Central Provident Fund in Singapore, a mandatory social insurance scheme originally designed to provide old-age income security, eventu- ally allowed members to use their account funds for paying housing mortgages. As withdrawals to fund home ownership increased, concerns about the scheme’s retirement objectives increased (Hu and Stewart 2009). Since liquidity of savings above a certain threshold may be a relevant issue for low- and middle-income countries, schemes could be designed to restrict withdrawals from specific portions of the account and for ­certain acceptable reasons. Workers’ preferences and perceptions: Preferences and perceptions of workers in Syria and Lebanon suggest a genuine demand for social insurance and a general willingness to pay for its services.35 In Syria, about one-third of the surveyed workers are informal, 64 percent of which are interested in social insurance coverage. Similarly, 68 percent of informal workers in Lebanon are interested in being covered by social insurance. Figure 5.17 presents results from the survey. The majority of Syrian workers rank old-age benefits as the most important, whereas health insurance is the most important benefit for Lebanese workers, because the social insurance package there does not offer pen- sion coverage. In both countries, preferences for different benefits are similar among formal workers and informal workers demanding cover- age. The majority of formal workers in both countries consider the cost of social insurance to be reasonable, which indicates a general willing- ness to pay for coverage.36 Further confirming a willingness to pay, cost is not identified as the most important reason for lack of interest in social insurance among groups of informal workers. In Syria, individu- als’ myopia (56 percent) and lack of knowledge about the system (55 ­percent) seem to be more important reasons than cost ­considerations. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Figure 5.17 Preferences and Perceptions of Informality among Workers in Lebanon and the Syrian Arab Republic value of different SI benefits for workers in value of different SI benefits in the Syrian Arab Republic Lebanon old age 59 64 54 health job security 16 insurance 48 14 work injury 9 8 6 40 disability 6 end of service 45 survivorship 6 4 family 3 3 family 6 allowance 8 heating 2 2 0 10 20 30 40 50 60 70 0 10 20 30 40 50 60 % of workers ranking benefit as most important % of workers ranking benefit as most important cost of SI among formal workers in benefit in Lebanon and the Syrian Arab Republic is considered … Lebanon, 86 reasonable Syrian Arab Republic, 71 Lebanon, 6 high Syrian Arab Republic, 16 Lebanon, 9 low Syrian Arab Republic, 10 do not know how much Syrian Arab Republic, 3 0 20 40 60 80 100 % of formal workers informal (interested in SI) formal figure continues next page 307   308 Figure 5.17  Preferences and Perceptions of Informality among Workers in Lebanon and the Syrian Arab Republic (continued) most important reasons for not being interested in SI in the Syrian Arab Republic not worried about 56 retirement do not know how the 55 system works most important reasons for not being interested in SI in Lebanon have private health low quality of services 42 36 insurance no trust in the system 40 low quality of service 21 costly, too much costly, too much 40 16 deduction from salary deduction from salary 0 10 20 30 40 50 60 0 10 20 30 40 of informal workers (not interested in SI) ranking of informal workers (not interested in SI) ranking reasons as important or very important reasons as important or very important Sources: Syria Employee-Employer Survey (2011) and Lebanon Employee-Employer Survey (2011). Note: SI = social insurance. Barriers to Coverage and Policy Options 309 Cost considerations are also not the main reason in Lebanon either, where reliance on private health insurance as an alternative is most frequently cited. Implications for MENA countries. Among MENA countries, the rele- vance of contributory and noncontributory schemes (whether targeted or universal) as coverage expansion strategies varies. In Lebanon and Jordan, the existing coverage rates and the development of proxy means testing systems indicate the viability of a targeted-contributory scheme. The countries’ GDP rates, rural population rates, and poverty levels indicate a potential affordability of contributions, for at least some population groups (annex table 5A.1). Governments may consider fully subsidizing contributions of groups with no savings capacity, while matching contributions of groups with limited saving capacity, as deter- mined by the proxy means test. The willingness to pay for social ­ insurance, as implied by the survey results presented earlier, further confirms the potential relevance of a contributory scheme in Lebanon. In Syria, general country conditions (including lower levels of coverage and higher rural population rates) would guide policy choices away from introduction of a contributory scheme. However, the survey results indicate a certain level of demand and willingness to pay for social insurance among informal w ­ orkers in registered firms. In this case, the government might consider introducing a contributory scheme targeting such workers. ­ Country conditions (especially high poverty rates) in Djibouti and the Republic of Yemen would point toward universal noncontributory schemes, while acknowledging that limited fiscal space is an important barrier. In Djibouti, although the share of the rural population is low, mobile and Internet penetration are also quite low, poverty is high among the elderly and the rest of the population, and no proxy means testing (PMT) system is in place (see box 5.9 for targeting methods). Although a PMT system exists in the Republic of Yemen, the high share of rural population, low mobile and Internet penetration rates, and high poverty levels favor the relevance of a universal strategy at this time. In Tunisia, coverage patterns and level of income make a contributory scheme viable for at least part of the population, but with no PMT in place, a contributory scheme in the short run may need to be based on a universal flat matching provided by the ­government. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 310 Barriers to Coverage and Policy Options Program implementation. Implementation arrangements and capacity differ widely in MENA. Some of the key elements for consideration include whether some of the tasks outlined in the pyramid in figure 5.16 can be contracted out to the private sector, while others remain under public management. Another important set of decisions involves how to best leverage existing infrastructure and institutions to implement new schemes. Box 5.11 describes the recent implementation of a new noncontributory health insurance scheme in India. Its implementation ­ arrangements are continually being improved as the scheme evolves, a critical factor of its success so far. Box 5.11 Implementation of the Indian Rashtriya Swasthya Bima Yojana In 2008 the government of India introduced a noncontributory health insurance scheme. This is the first such scheme to cover the informal sector (95 percent of ­ population) after earlier attempts to extend health insurance by the Ministry of Health failed. The new scheme, called “Rashtriya Swasthya Bima Yojana” (RSBY), was launched by the Ministry of Labor in April 2008 and targeted the 300 million indi- viduals below the poverty line. RSBY provides hospitalization coverage of up to Re 30,000 for up to five members of each family. Individuals pay as little as Re 30 in registration fees yearly. The scheme recently expanded to households above the poverty line who are willing and able to pay a premium for the services provided. In only a few months, over 150,000 families enrolled. The government is now also con- sidering gradual expansion of coverage to different categories of workers. The arrangements through which the program is being implemented provide numerous lessons learned for other countries. A thorough investigation of the target group, poor informal workers, led to the realization that the scheme needed to be free, paperless, portable, independent of any identifiable employee-employer rela- tionship, and impose minimal transaction cost on target groups. The government designed implementation arrangements in the shape of a pure business model for a social service where the scheme is funded publicly but operated privately. Adminis- tration of the scheme is contracted entirely to a third party insurer (TPI) selected on a competitive basis in each state. The TPI is responsible for enrolling persons and is paid a premium by the government for every person enrolled, creating an incentive for TPIs to expand coverage. Moreover, hospitals are paid based on each beneficiary treated. The government provides the list of eligible households to the insurer and a schedule for enrollment. The insurer is responsible for announcing the scheme and the ­enrollment procedure in the state and managing mobile ­ enrollment stations box continues next page Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 311 Box 5.11  Implementation of the Indian Rashtriya Swasthya Bima Yojana (continued) frequently visited locations. The insurer also involves local grassroots and stations at ­ entities including NGOs and micro-finance institutions. Members were provided with a unique “smart card,” used for all transactions, on the spot when enrolling. In parallel, a unique national identification system in India is in the process of being developed. Take-up of RSBY was initially quite slow, but today over 45 million people across 29 Indian states are covered by the scheme. Stakeholder ownership and extensive information campaigns were instrumental in increasing take-up rates. Another factor considered critical for the success of the program was having the right information technology ­ systems to support the scheme in terms of operation and monitoring, as well as for fraud minimization. Today RSBY faces numerous challenges as well as promising opportunities. Increasing public awareness and information about the ser- vice and its benefits remains a huge challenge. Building capacity of all stakeholders and service providers is also a requirement. Another important challenge is improv- ing the quality of services offered through this scheme. To address that, the govern- ment is attempting to create incentives for improved quality of services by rating the quality of hospitals. Corruption is another challenge: Fraud and collusion behavior between beneficiaries and hospitals have been detected in more than 50 hospitals. Sources: Rashtriya Swasthya Bima Yojna, 2009; personal communication with Anil Swarup, Director G ­ eneral for Labour Welfare at the Ministry of Labor in India. It is also worth mentioning that financial education and generally rais- ing awareness about the value of social insurance can be a useful tool for supporting the implementation of social insurance programs and con- tributing to higher take-up rates. Evidence from developing countries shows that general awareness about social insurance can be improved. A survey in India shows that about 80 percent of the informal sector did not even know what pensions were. In Chile, the understanding of pen- sions among citizens is still very limited (rated at 2 on a scale of 2 to 7). Using data from Chile, Fajnzylber and Reyes (2010) show that an improvement in information provided by pension administrators increased the probability of making voluntary contributions by older workers but less so by younger workers. Women responded to the infor- mation significantly more than men. Even in developed countries such as the United Kingdom, research has shown that over 25 percent of pension credits remain unclaimed because people do not know they are eligible. Various initiatives can improve awareness and financial education, including training office staff in social insurance agencies ­ Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 312 Barriers to Coverage and Policy Options (as in China) and community outreach and awareness through house- hold visits to potential plan members (as in the United Kingdom). The recent Chilean pension reforms set up a “Pension Education Fund” to fund awareness-raising initiatives (Stewart 2006). The evidence from Lebanon and Syria also indicates low awareness about pension benefits and ­pension rights. Lessons from behavioral economics. Findings from behavioral eco- nomics can complement rational economic models in understanding individuals’ savings choices and behavior.37 Behavioral research has shown that individuals procrastinate when making choices, especially important ones. This is known as “inertia” or “status quo” bias. When the choice involves selection among several options and the need to make complex computations, individuals are even more likely to behave passively. Such behaviors are incompatible with traditional models of economics based on the rationality of economic agents. Another finding is the concept of “loss aversion”; people are more active in avoiding losing something they already have rather than in acquiring new gains (Kahneman and others 1991). Furthermore, behavioral research has shown that altering the description of a certain problem or question (also called “framing”) may result in different choices and judgments (Tversky and Kahneman 1981). To some extent, these findings explain certain savings behaviors, including, for example, the low take-up of voluntary retirement or other social insurance schemes. Figure 5.18 presents a notional framework for ­ addressing individual saving behavior. In the context of coverage expansion, auto-enrollment and default options in retirement savings schemes (or other types of social insur- ance schemes) can positively influence savings behavior. Under auto-enrollment, individuals are by default enrolled into a social insur- ­ ance scheme unless (or until) they chose to opt out. In addition, saving rates are set at a default (usually conservative) rate with a choice to enroll in different saving options if desired. The idea is motivated by the philosophy of libertarian paternalism under which the government can influence an individual’s choices without exerting coercion (Thaler and ­ Sustein 2003). Auto-enrollment in social insurance schemes has been used in various developed countries, including Italy, New Zealand, the United Kingdom, and the United States. Some evidence from 401(k) plan experiments shows that auto-enrollment and default options are effective in increasing take-up of voluntary savings schemes and can Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 313 Figure 5.18 Implications of Rational and Behavioral Economics Theories ­Explaining an Individual’s Savings Behavior theory policy/design says… implications rational economics behavioral economics optimal set optimal savings subsidy inertia auto- loss enrollment liquidity impose aversion and default constraints mandatory options time- savings to framing inconsistent counteract preferences myopia even be designed to increase savings. Using data from a large U.S. firm, Madrian and Shea (2001) found that automatic enrollment had a sig- nificant impact on enrollment and retirement savings behavior; an enrollment rate of 86 percent was achieved compared with 36 percent without auto-enrollment. Simply framing the question as “Check this box if you would not like to have 3 percent of your paycheck put into a 401(k)” instead of “Check this box if you would like to participate in a 401(k) and indicate how much you’d like to contribute” led to this difference. An interesting auto-enrollment default option scheme that does achieve increased savings is the experimental “Save More Tomorrow” scheme (Thaler and Benartzi 2004). Under this scheme, employees are invited to choose a target level of savings according to which they will automatically start contributing in one year at a low savings rate, with a gradually increasing rate until they reach their target. This feature is usually referred to as auto-escalation or auto- ­ acceleration of savings. Note that it makes a difference if one needs to commit current earnings to saving in the present or commit to saving yet unearned future income.38 Employees have the choice to opt out at any time. Evidence from four years of implementation shows a take-up rate of almost 80 percent; moreover, 80 percent of participants stayed through to the fourth year. Saving rates rose from 3.5 to 13.6 percent over the course of the program. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 314 Barriers to Coverage and Policy Options Conclusions to Part 2 Overall, coverage extension through the social insurance system is an important yet complex agenda in MENA. This part of the chapter first reviewed the link between the design of social insurance schemes, par- ticularly pensions, and coverage outcomes. With the exception of the recently approved social insurance reform in Egypt, all MENA countries have DB pension systems with design features not in line with best inter- national practices. Many of the existing substandard design characteris- tics of MENA DB pension schemes contribute to the low observed coverage rates beyond the already well-known concerns over financial sustainability. These flawed design features include limited legal cover- age, short minimum vesting periods that promote gaming of the system, generous early retirement provisions that distort participation incentives, and the use of average wages in the last few years of service as a basis of benefit computation. It is crucial that these incentive design issues are addressed through reforms that ensure adequate, affordable, and sustain- able benefits, while offering coverage to a large and increasing share of the population. It is unlikely that in MENA coverage extension will be achieved effectively without special subsidized social insurance schemes targeted to informal workers, or even to those outside the labor force. Although this notion might be considered controversial, social insurance coverage, at least for certain social groups, should be separated from the formal employment relationship without compromising the unified overall logic (integration platform) of the social insurance system most easily established based on DC design features. Even if this is achieved, no homogeneous recipe for success can be offered; country conditions mat- ter significantly. A decision tree approach that MENA governments could follow when deciding on the type of social insurance coverage extension programs was suggested. Once the incentive design of the existing formal sector social insurance scheme in MENA is improved, the key decisions with regard to coverage extension strategies are (1) whether contributory or noncontributory programs are more appro- priate and (2) whether universal or targeted subsidies should be used. To support these critical decisions, a stylized set of indicators based on country conditions was developed to help governments assess the rele- vance of different schemes. Among targeted contributory interventions, piloting a MDC scheme with subsidies conditional to low saving capac- ity emerges as a viable option to expand coverage in some MENA countries, including Lebanon, Syria, and Jordan. In the short run, Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 315 ­ oncontributory schemes will still play an important supplementary n role in these countries, particularly for those individuals who have already completed their working life and those with no savings capacity. Means-tested noncontributory strategies may have to be used as the main coverage extension strategy in poorer countries such as Djibouti and the Republic of Yemen where contributory schemes are not a strong option given current country conditions. In the case of new con- tributory schemes, design features and implementation arrangements geared at inducing voluntary participation that include auto-enrollment and auto-escalation of savings could be considered, in addition to com- plementary financial literacy and general awareness interventions. It was also emphasized that the ex ante predictions of traditional, agent rationality–based, economic models are not sufficient for determin- ing the subsidies needed to induce voluntary participation, critical in the informality context. The lessons learned from behavioral economics about observed behaviors being irreconcilable with rational agent assumptions should also be taken into account. Such lessons suggest the use of auto-enrollment arrangements and program features responding to loss aversion (as opposed to just risk aversion). The recent innovation of MDC program design for social insurance coverage expansion, especially if combined with a targeting mechanism that assigns subsidies condi- tioned on the income position of the individual or household, received special attention. Although little empirical evidence exists on the poten- tial impact of such innovations (and virtually no evidence from MENA, other than inferences based on perception surveys), one could confi- dently argue that successful coverage extension programs would combine (1) a program design predicting subsidy needs based on ex ante predictor economic models of agent rationality with (2) lessons from observed behaviors reflecting bounded rationality put into an implementation environment, which (3) ensures low transactions costs from the perspec- tive of the individual for enrollment and benefit delivery and (4) relies on investments in financial literacy and broader awareness building within the target group of the coverage extension effort. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 316 Annex Annex Table 5A.1 Country Conditions in MENA Countries Syrian Arab Egypt, Arab Iran, Islamic   Djibouti Yemen, Rep. Iraq Morocco Republic Jordan Rep. Algeria Tunisia Rep. Lebanon Rural population 12.3 68.78 33.5 43.64 45.44 21.54 57.24 34.14 33.1 31.02 12.92 (% of population)a GDP per capita, PPP 2,319 2,470 3,548 4,494 4,730 5,597 5,673 8,172 8,273 11,558 13,070 ($ current)a Coverage rate 9 33 18 29 42 37 45 44 (% employed with social security coverage)b Poverty gap (elderly/­ 43.52/45.09 34.57/41.75 13.2/18.98 9.81/14.18 4.09/8.62 — — — — nonelderly)c Mobile cellular subscription 13 16 57 72 34 91 51 93 83 60 34 (per 100 people)d Internet users (per 100 2 2 1 33 17 27 17 12 27 32 23 people)d ­ Corruption Perception Indexe 3.2 2.2 1.5 3.4 2.5 4.7 3.1 2.9 4.3 2.2 2.5 Existence of PMT No Yes No Yes In prepara- In prepara- Yes No In prepara- tion tion tion Sources: a. World Development Indicators, World Bank 2009. b. Specific surveys for the Republic of Yemen, Iraq, Morocco , the Syrian Arab Republic, Egypt, and Lebanon as presented in chapter 2; coverage rates for the other countries are based on 2008 World Development Indicators, pension contributors as percent of labor force. c. Robalino and others 2009. d. World Development Indicators, World Bank 2008. e. The Corruption Perception Index ranges from 0 to 10, 10 being for countries with no perceptions of corruption. From Transparency International 2010. Note: GDP = gross domestic product; PPP = purchasing power parity; PMT = proxy means testing; — = not available. Barriers to Coverage and Policy Options 317 Notes 1. Note that other taxes also contribute to the tax wedge. For detailed discus- sion, see below. 2. Note that as pointed out by Otonglo and Trumbic (2008), even in Turkey, the medium-run effect can be an increase in revenues, as the total amount of income declared increased, attributed to less understating of incomes. Importantly, evidence on the short-run effects of the Turkey case study should be interpreted with caution because this study is based on a before-and-after comparison, and many other changes might have occurred in the economy during the same period that also affected overall tax collection rates, making it hard to draw causal interpretations. 3. The fact that Doing Business rankings capture only a small part of the set of complementary reforms that are needed to effectively promote private sector growth can explain the apparent disconnect between the significant improve- ment in rankings recorded by some of the MENA countries and the contin- ued perceived challenges to an inclusive growth process. 4. The EU15 was the number of member countries in the European Union prior to the accession of ten candidate countries on 1 May 2004 compromising Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands, Portugal, Spain, Sweden, and the United Kingdom. 5. The term “fixed-term worker” means a person having an employment contract or relationship entered into directly with an employer where the end ­ of the employment contract or relationship is determined by objective condi- tions, such as reaching a specific date, completing a specific task, or the ­ occurrence of a specific event. 6. Wage data for Tunisia were not available to produce kernels similar to those depicted in figure 5.5. 7. The ILO, European Union, and OECD have embraced the concept of “flexi- curity,” combining flexible regulation, safety nets (such as unemployment insurance), and active social policies. One component of flexicurity policies is flexible and reliable contractual arrangements (from the perspective of the employer and the employee, of “insiders” and “outsiders”) through modern labor laws, collective agreements, and work organization. 8. This is only an indicative data point that does not take into account, for example, the underlying educational attainment composition in the public and private employment sectors. The point is that public sector pay is rela- tively high even in Arab region countries, where public sector salaries are generally perceived to be low such as in Egypt. 9. Annual Employment and Unemployment Survey, Department of Statistics, 2008. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 318 Barriers to Coverage and Policy Options 10. Data were collected from the Public Commission for Employment and Enterprise Development in Syria. 11. The Syrian example highlights distortions introduced by certain labor market institutions; the unintended contribution to the informality phenomenon is just one particular aspect. The public sector job queue consists of a self- selected group of job seekers with a preference for public sector employment; this preference drives their acceptance of informal and refusal of formal pri- vate jobs, at least temporarily. The government uses this queue to attempt to identify candidates for its training and private sector placement programs because the alternative registries are of little value; neither employers nor job seekers show up in other government registries because neither party believes in effective matches through such channels. 12. See MENA Education Flagship Report (World Bank 2008). 13. An apprenticeship is an arrangement that allows individuals to receive firm- specific training through a working relationship with an employer. Workers usually receive low wages during the apprenticeship period such that the training costs are shared among firms and workers during the period. 14. The expressions “social insurance system” and “pension system” are used somewhat interchangeably in this chapter. Pension systems are often thought of as a set of social insurance mechanisms, including old-age pension, disabil- ity pension, survivor benefit, and work injury benefit components. These are the social insurance components typically introduced at earlier stages of development. Pension systems in this sense are present in all the MENA countries, albeit with often limited coverage. A social insurance system is often thought of as one that incorporates some of the social insurance com- ponents typically introduced at higher levels of development, such as unem- ployment insurance, sickness and health insurance, and maternity benefits. These “second generation” social insurance benefits are typically introduced on top of the already existing pension system components. 15. See Holzmann and Hinz 2005; Holzmann and Palmer 2006. 16. In a funded DC scheme, savings are typically accumulated through invest- ment in actual financial saving instruments. 17. See the discussion on the range of the cost of universal basic pensions later in this chapter based on Palacios and Sluchynsky (2006). A cost benchmark at 2 percent of GDP is what the World Bank projections suggest on universal flat pension expenditures in the case of Egypt. 18. Currently, $NZ318.12 (net, for single persons) is paid to those eligible, as determined by age (65) and residency. Benefits are adjusted annually based on the consumer price index and average wages. The main task of the pension administration is not really administering this simple payment mechanism, but rather building tools that improve financial literacy and communicating Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 319 to the public that the quality of life in retirement is heavily dependent on individual saving behavior. 19. For example, a claw-back with a basic pension reduction rate of 25 percent yields no reduction in the universal basic pension if no contributions were made during an individual’s active life cycle (for example, this is the benefit for those who spent their entire career in the informal sector), whereas the top-up is completely phased out at an earnings-related pension level equiva- lent to four times the basic flat pension. The mathematical formula to calculate the pension benefit under the universal basic pension with a claw- ­ back design is p = ERP (y) – max (0, a – b × ERP (y)), where p is the amount of the pension benefit, ERP (y) is the pure earnings-related pension compo- nent benefit at life-cycle income measure y, a is the amount of the universal flat pension for those with no earnings-related pension, and b is a parameter reflecting the share of the earnings-related pension reducing the universal pension benefit. 20. For a detailed discussion on MDCs, see Robalino and Palacios (2009). 21. Notionally, participation requires a participation fee based on income, but as income is rarely observable outside of the formal sector, almost no one pays such a fee. 22. See Palacios and Whitehouse (2006) for more information on the evolution of civil service pension schemes worldwide. 23. See Pallares-Miralles and others (2011) for cross-country data on mandatory publicly and privately managed pension systems around the world. 24. Revalorization of past wages in the pension benefit formula means that past wages are taken into account with an adjustment equivalent to pro- portional changes in the general wage level between the time the pension contribution was paid and when the formula was applied to determine pension levels. 25. NDC schemes are DC schemes where contributions are not funded (that is, invested in financial instruments) and where individual earnings are based on individual contributions (PAYG). See Holzmann and Palmer (2006). 26. The lump-sum payments offered by MENA pension schemes to those who do not fulfill the minimum vesting requirement do not compare to the mag- nitude of pension payments received until death; contributions for these individuals would remain a sunk cost, for the most part, even with the lump- sum payment. 27. A simplified mathematical representation of the core of a defined benefit –, where T is the number of contributory years, pension formula is p = a × T × w – w is some measure of the wage based on which contributions were made dur- ing at least a part of the active life cycle, and a is a parameter called the replacement rate referring to what share of earnings one year of contributions Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 320 Barriers to Coverage and Policy Options replaced in old age by the DB pension system. Such a formula is calibrated (if at all) for a particular retirement age. Clearly a should be conditional on the actual age of retirement if sustainability matters. 28. In the case of Egypt, this features refers to the pension system before the recent legislated reform. 29. The internal rate of return (IRR) is a concept that creates an intuitive rela- tionship between life-cycle contributions and the pension benefit flow. One can think about the internal rate of return as a real interest rate that, if applied to all career employee and employer contributions, would yield a savings amount by retirement equivalent to the (expected) value of the pen- sion benefit flow until death. The IRR concept makes it easy to assess whether one should save inside or outside a pension system (provided that is a choice). 30. The accrual rate is a critical DB pension parameter. It is the rate by which pension benefits accumulate each year of contribution expressed as a percent- age of the final salary. 31. World Bank employee-employer survey on workforce development (2001) conducted under the World Bank MILES program in Syria. 32. For a discussion of universal and targeted social pensions, see Holzmann and others (2009) and Palacios and Sluchynsky (2006). 33. OECD work on financial education and pensions can be found at http://www.oecd.org/document/37/0,3343,en_2649_15251491_25698341​ _1_1_1_1,00.html. 34. 401(k) plans are widely used retirement savings plans among American ­ workers, where employers often choose to match contributions that workers make into their individual accounts. IRA schemes are provided by an employer and are similar to a 401(k) but offer simpler and less costly admin- istration rules. 401(k) and IRA schemes are DC schemes in design. 35. The firm-based survey in Syria is representative of workers in registered ­ private sector firms in manufacturing and services sectors. The household- based survey in Lebanon is representative of private sector workers more generally. 36. It is worth mentioning that the Syrian Social Insurance system is an unsus- tainable one because contributions do not cover the cost of services. It is unknown whether workers’ perceptions on the cost of social insurance would change if the contributions to be paid reflected the true cost of the services offered. 37. The field of behavioral economics brings together insights from psychology and economic theory in understanding individuals’ economic choices and Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 321 decisions, with a focus on bounded rationality of economic actors. For recent experiments and writings on this topic, see Ariely (2008, 2010). 38. This behavior is consistent with the theory of time-inconsistent preferences, a formal and promising modeling attempt to explain certain behaviors in the rational agent environment. References Abowd, J., F. Kramarz, D. Margolis, and T. Phillippon. 2005. “Minimum Wages and Employment in France and the United States.” Paris: CREST-INSEE. Almeida, R., and P. Carneiro. 2005. “Enforcement of Labor Regulation, Informal Labor, and Firm Performance.” Policy Research Working Paper 3756. Washington, DC: World Bank. ———. 2009. “Enforcement of Regulation, Informal Employment, Firm Size and Firm Performance.” Journal of Comparative Economics 37 (1): 28–46. Alterido, R., M. Hallward-Driemeier, and C. Pages. 2009. “Big Constraints to Small Firms’ Growth? Business Environment and Employment Growth across Firms.” Policy Research Working Paper 5032. Washington, DC: World Bank. Angel-Urdinola, D., and A. Kuddo. 2010. “Key Characteristics of Employment Regulation in the Middle East and North Africa.” World Bank SP Discussion Paper 1006. Washington, DC: World Bank. Angel-Urdinola, D., and A. Semlali. 2010. “Labor Markets and School-to-Work Transition in Egypt: Diagnostics, Constraints, and Policy Framework.” MPRA Paper 27674. Munich: University Library of Munich. Angel-Urdinola, Diego F., Amina Semlali, and Stefanie Brodmann. 2010. “Non- Public Provision of Active Labor Market Programs in Arab-Mediterranean Countries: An Inventory of Youth Programs.” Social Protection Discussion Papers No. 55673, World Bank, Washington, DC. Antolin, P., and E. L. Ponton. 2007. “The Impact of Tax Incentives on Retirement Savings: A Literature Review.” Conference Proceedings, OECD/IOPS Global Private Pensions Conference. Paris: OECD. Ariely, D. 2008. Predictably Irrational: The Hidden Forces That Shape Our Decisions. London: HarperCollins. ———. 2010. The Upside of Irrationality: The Unexpected Benefits of Defying Logic at Work and Home. London: HarperCollins. Auer, P. 2007. “In Search of Optimal Labor Market Institutions.” In Flexicurity and Beyond: Finding a New Agenda for the European Social Market, ed. H ​ . Jorgenson and P. K. Madsen, 67–98. Copenhagen: DJOF. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 322 Barriers to Coverage and Policy Options Auer, P., J. Berg, and I. Coulibaly. 2004. “Is a Stable Workforce Good for the Company? Insights into the Tenure-Productivity-Employment Relationship.” Employment Strategy Papers 2004-15. Geneva: ILO. Bartelsman, E., J. Haltiwanger, and S. Scarpetta. 2004. “Microeconomic Evidence of Creative Destruction in Industrial and Developing Countries.” Tinbergen Institute Discussion Papers 04-114/3. Amsterdam: Tinbergen Institute. Betcherman, G., and C. Pages. 2007. “Estimating the Impacts of Labor Taxes on Employment and Balances of the Social Insurance Funds in Turkey.” Washington, DC: World Bank. Bodor, A., D. Robalino, and M. Rutkowski. 2008. “How Mandatory Pensions Affect Labor Supply Decisions and Human Capital Accumulations? Options to Bridge the Gap between Economic Theory and Policy Analysis.” Washington, DC: World Bank. Bruhn, M. 2008. “License to Sell: The Effect License to Sell: The Effect of Business Registration Reform on Entrepreneurial Activity in Mexico.” World Bank Policy Research Working Paper WP4538. Washington, DC: World Bank. ———. 2011. “License to Sell: The Effect of Business Registration Reform on Entrepreneurial Activity in Mexico.” Review of Economics and Statistics 93 (1): 382–86. di Filippo, M., and A. Bodor. Forthcoming. “Old-Age Pension System Induced Incentives as They Affect Participation in Informal Sector Employment in the Middle East and North Africa Region.” Draca, M., S. Machin, and J. Van Reenen. 2008. “Minimum Wages and Firm Profitability.” Working Paper 13996. Cambridge, MA: National Bureau of Economic Research. Duflo, E., W. Gale, J. Liebman, P. Orszag, and E. Saez. 2005. “Saving Incentives for Low- and Middle Income Families: Evidence from a Field Experiment with H&R Block.” CEPR Discussion Paper 5332. Social Science Research Network. Engelhardt, G. V., and A. Kumar. 2006. “Employer Matching and and 401 K Savings: Evidence from the Health and Retirement Study.” Working Paper 12447. Cambridge, MA: National Bureau of Economic Research. European Commission. 2006. Employment in Europe 2006. Brussels: European Commission. ———. 2010. “Employment in Europe 2010.” Brussels: European Commission. Fajnzylber, E., and G. Reyes. 2010. “How Much Saving Is Enough? Evaluation of a Large Scale Intervention on Retirement Saving Decisions.” http://insr​ .­wharton.upenn.edu/documents/seminars/F10-Fajnzylber.pdf. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 323 Fialová, K., and O. Schneider. Forthcoming. “Labor Institutions and Their Impact on Shadow Economies in Europe.” Background paper for “In from the Shadow: Integrating Europe’s Informal Labor.” Washington, DC: World Bank. Grazier, B. 2007. “Making Transitions Pay: The Transitional Labor Market Approaches to Flexicurity.” In Flexicurity and Beyond: Finding a New Agenda for the European Social Model, ed. H. Jorgenson and P. K Madsen, 99–130. Copenhagen: DJOF. Grosh, M., C. del Ninno, E. Tesliuc, and A. Ouerghy. 2008. “For Protection and Promotion: The Design and Implementation of Effective Safety Nets.” Washington, DC: World Bank. Holzmann, R., and R. Hinz. 2005. “Old Age Income Support in the 21st Century.” Washington, DC: World Bank. Holzmann, R., and J. Koettl. 2010. “Portability of Pension, Health and Other Social Benefits.” Washington, DC: World Bank. Holzmann, R., and E. Palmer. 2006. “Pension Reform: Issues and Prospects for Non-Financial Defined Contributions (NDC) Schemes.” Washington, DC: World Bank. Holzmann, R., D. Robalino, and N. Takayama. 2009. “Closing the Coverage Gap: The Role of Social Pensions and Other Retirement Income Transfers.” Washington, DC: World Bank. Hu, Y., and F. Stewart. 2009. “Pension Coverage and Informal Sector Workers: International Experience.” OECD Working Papers on Insurance and Private Pensions 31. Paris: OECD. Johanson, R., and A. V. Adams. 2004. “Skills Development in Sub-Saharan Africa. Regional and Sectoral Studies.” Washington, DC: World Bank. Kahneman, D., J. Knetch, and R. Thaler. 1991. “Anomalities: The Endowment Effect, Loss Aversion and Status Quo Bias.” Journal of Economic Perspectives 5 (1): 193–206. Kaplan, D., E. Piedra, and E. Seira. 2007. “Entry Regulation and Business Start- Ups: Evidence from Mexico.” Policy Research Working Paper 4322. Washington, DC: World Bank. Kenyon, T., and E. Kapaz. 2005. “The Informality Trap.” Viewpoint Series, Note 301. Private Sector Development Vice-Presidency. Washington, DC: World Bank. Kugler, A., and M. Kugler. 2003. “The Labor Market Effects of Payroll Taxes in Middle Income Countries: Evidence from Columbia.” Discussion Paper 0920. Bonn: IZA. Lee, K. W. 2009. “Productivity Increases in SMEs: With Special Emphasis on In-Service Training of Workers in Korea.” Social Protection Discussion Paper 0917. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 324 Barriers to Coverage and Policy Options Lemos, S. 2004. “The Effect of Minimum Wage on Prices.” Bonn: IZA. Levenson, A., and W. Maloney. 1998. “The Informal Sector, Firm Dynamics, and Institutional Participation.” World Bank Policy Research Working Paper WPS1988. Washington, DC: World Bank. Levy, S. 2006. “Social Security Reforms in Mexico: For Whom?” World Bank Conference on “Equity and Competitiveness in Mexico.” Washington, DC: World Bank. Loayza, N., A. M. Oviedo, and L. Serven. 2005. “The Impact of Regulation on Growth and Informality Cross-Country Evidence.” Policy Research Working Paper 3623. Washington, DC: World Bank. Madhu Singh, ed. 2005. “Meeting Basic Learning Needs in the Informal Sector.” Technical and Vocational Education and Training, Vol. 2. Geneva: UNESCO- UNEVOC. Madrian, B. C., and D. F. Shea. 2001. “The Power of Suggestion: Inertia in 401(K) Participation and Saving Behavior.” Quarterly Journal of Economics 116 (4): 1149–87. Maloney, W., and J. Medez. 2003. “Measuring the Impact of Minimum Wages: Evidence from Latin America.” Working Paper 9800. Cambridge, MA: National Bureau of Economic Research. McKenzie, D., and Y. S. Sakho. 2010. “Does It Pay Firms to Register for Taxes? The Impact of Formality on Firm Profitability.” Journal of Development Economics 91 (1): 15–24. OECD (Organisation for Economic Co-operation and Development). 2004. Employment Outlook. Paris: OECD. ———. 2008. Employment Outlook. Paris: OECD. ———. 2010. Employment Outlook. Paris: OECD. Olivier, M. 2009. “ Informality, Employment Contracts and Extension of Social Insurance Coverage.” Geneva: International Social Security Association. Otonglo, C., and T. Trumbic. 2008. “Doing Business. Case Study of Turkey.” http:// www.doingbusiness.org/~/media/FPDKM/Doing%20Business/Documents​ /Reforms/Case-Studies/2008/DB08-CS-Turkey.pdf. Palacios, R. Forthcoming. “Framework for Implementing Social Programs (FISP).” Palacios, R., and O. Sluchynsky. 2006. “Social Pensions Part I: Their Role in the Overall Pension System.” Social Protection Discussion Paper 0601. Washington, DC: World Bank. Palacios, R., and E. Whitehouse. 2006. “Civil-Service Pension Schemes around the World.” Social Protection Discussion Paper 0602. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 325 Pallares-Miralles, M., C. Romero, and E. Whitehouse. Forthcoming. “International Patterns of Pension Provision II: A Worldwide Overview of Facts and Figures.” Pierre, G., and S. Scarpetta. 2006. “How Labor Markets Can Combine Workers’ Protection with Job Creation: A Partial Review of Some Key Issues and Policy Options.” Social Protection Discussion Paper 0716. Washington, DC: World Bank. Portugal, P., and A. Cardoso. 2006. “Disentangling the Minimum Wage Puzzle: An Analysis of Worker Accessions and Separations.” Journal of the European Economic Association 4 (5): 988–1013. Rashtriya Swasthya Bima Yojna. 2009. http://www.rsby.gov.in. Razzaz, O. 2011. “The Political Economy of Social Security Reform: Lessons from the Jordan Experience.” Mimeo. World Bank, Washington, DC. Robalino, D. A. 2005. “Pensions in the Middle East and North Africa: Time for Change.” Washington, DC: World Bank. Robalino, D., V. Gudivada, and S. Oleksiy. 2009. “Preventing Poverty amongst the Elderly in MENA Countries: Role and Optimal Design of Old-Age Subsidies.” Primer Pensions Series. Washington, DC: World Bank. Robalino, D. A., and R. Palacios. 2009. “Matching Defined Contributions: A Way to Increase Pension Coverage.” In Closing the Coverage Gap: The Role of Social ​ . Takayama, Pensions and Other Retirement Income Transfers, ed. D. A. Robalino, N and R. Holzmann, 187–202. Washington, DC: World Bank. Robalino, D. A., H. Ribe, and I. Walker. 2010. “Achieving Effective Social Protection for All in Latin America and the Caribbean: From Right to Reality.” Washington, DC: World Bank. Rutkowski, J. 2003. “The Minimum Wage: Curse of Cure?” Washington, DC: World Bank. Semlali, A. 2008. “Youth Employability and Job Matching Challenges and Opportunities in the Middle East and North Africa Region.” Washington DC: World Bank. Shen, C., and J .B. Williamson. 2010. “China’s New Rural Pension Scheme: Can It Be Improved?” International Journal of Sociology and Social Policy 30: 239–250. Silva, J., V. Levin, and M. Morgandi. 2012. “Inclusion and Resilience: The Way Forward for Safety Nets in the Middle East and North Africa.” World Bank, Washington, DC. Singh, M., ed. 2005. “Meeting Basic Learning Needs in the Informal Sector— Integrating Education and Training for Decent Work, Empowerment and Citizenship.” Hamburg: Institute for Education, UNESCO. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 326 Barriers to Coverage and Policy Options Smits, K. 2008. “Economic Aspects of the Minimum Wage Policy.” Washington, DC: World Bank. SSA. 2009. “Social Security Programs throughout the World: Africa, 2009.” Washington, DC: Social Security Administration (SSA) and International Social Security Association (ISSA). ———. 2010. “Social Security Programs throughout the World: Asia and the Pacific, 2009.” Washington, DC: Social Security Administration (SSA) and International Social Security Association (ISSA). Steel, F. W., and D. Snodgrass. 2008. “Raising Productivity and Reducing Risks of Household Enterprises—Diagnostic Methodology Framework.” WIEGO Network. Washington, DC: World Bank. Stewart, F. 2006. “Financial Education and Pensions.” Presentation at G8 International Conference on Improving Financial Literacy. Moscow: OECD. Terrel, K., and T. H. Gindling. 2002. “The Effects of Minimum Wages on the Formal and Informal Sector: Evidence from Costa Rica.” IZA Discussion Paper No. 1018. Thaler, R. H., and S. Benartzi. 2004. “Save More Tomorrow: Using Behavioral Economics to Increase Employee Saving.” Journal of Political Economy 112 (1): S164–S187. Thaler, R. H., and C. R. Sustein. 2003. “Libertarian Paternalisms.” American Economy Review 93 (2): 175–79. Transparency International 2010 Corruption Perception Index (CPI). http:// www.transparency.org/policy_research/surveys_indices/cpi/2010/results. Tversky, A., and D. Kahneman. 1981. “The Framing of Decisions and the Psychology of Choice.” Science 211 (4481): 453–58. Van Adams, A. 2008. “A Framework for the Study of Skills Development in the Informal Sector of Sub-Saharan Africa.” Washington, DC: World Bank. Von Ginneken, W. 2009. Extending Social Security Coverage: Concepts, Approaches and Knowledge Gaps. Geneva: International Social Security Association. World Bank. 2004. Doing Business 2005. Washington, DC: World Bank. ———. 2007. “Republic of Bolivia: Policies for Increasing Firms’ Formality and Productivity.” Report 40057-BO. Washington, DC: World Bank. ———. 2008. “Reducing Undeclared Employment in Hungary: Synthesis Report of the World Bank Study.” Washington, DC: World Bank. ———. 2009. “From Privilege to Competition: Unlocking Private-Led Growth in Middle East and North Africa.” Washington, DC: World Bank. ———. 2010. “Health Insurance in Lebanon: Improving Risk Pooling.” Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Barriers to Coverage and Policy Options 327 Yousef, T. 2004. “Development, Growth and Policy Reforms in Middle East and North Africa since 1950.” Journal of Economic Perspectives 18 (3): 91–116. Zavodny, M. 2000. “The Effect of the Minimum Wage on Employment and Hours.” Labour Economics 7 (6): 729–50. Zhu, Y. 2009. A Case Study on Social Security Coverage Extension in China. Geneva: International Social Security Association. Ziderman, A. 2003. “Financing Vocational Training in Sub-Saharan Africa.” Africa Region Human Development Series. Washington, DC: World Bank. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 Environmental Benefits Statement The World Bank Group is committed to reducing its environmental footprint. In support of this commitment, the Publishing and Knowledge Division lever- ages electronic publishing options and print-on-demand technology, which is located in regional hubs worldwide. Together, these initiatives enable print runs to be lowered and shipping distances decreased, resulting in reduced paper consumption, chemical use, greenhouse gas emissions, and waste. ­ The Publishing and Knowledge Division follows the recommended standards for paper use set by the Green Press Initiative. Whenever possible, books are printed on 50 percent to 100 percent postconsumer recycled paper, and at least 50 percent of the fiber in our book paper is either unbleached or bleached using Totally Chlorine Free (TCF), Processed Chlorine Free (PCF), or Enhanced Elemental Chlorine Free (EECF) processes. More information about the Bank’s environmental philosophy can be found at http://crinfo.worldbank.org/wbcrinfo/node/4. Striving for Better Jobs  •  http://dx.doi.org/10.1596/978-0-8213-9535-6 While many countries in the Middle East and North Africa region have experienced sustained economic growth for a number of years, this has not resulted in the creation of a sufficient number of jobs, especially in ­ egistering the highest the formal sector. In fact, informality in the region is widespread, with some countries r levels of informality in the world. The informal sector provides jobs that are vulnerable and low-paying; have low productivity; are poorly arranged and managed; and lack coverage against health, unemployment, and old-age risks. Unsurprisingly, these jobs generate low levels of satisfaction among workers, offering them scarce prospects for a real career or for a chance to move into formal jobs in the private or public sector. Striving for Better Jobs: The Challenge of Informality in the Middle East and North Africa assesses job informality from a human development angle, proposing five strategic directives to promote long-term inclusive growth and formality: foster competition; realign incentives, pay, and benefit packages in the public sector; move toward labor regulations that promote labor mobility and provide support to workers in periods of transition; enhance the productivity of informal workers through training and skills building; and reform existing social insurance systems and introduce new instruments to attain broader coverage. Removing the obstacles and distortions that currently impede an easier transition from informality to formality in employment would provide real opportunities for improving living standards and ­ socioeconomic equity. Achieving these results, however, will require an inclusive dialogue open to a broader set of social groups, particularly those whose voices have not been heard: youth, women, unemployed individuals, and informal workers. ISBN 978-0-8213-9535-6 SKU 19535