JOBS SERIES Issue No. 20 DIAGNOSTIC PAKISTAN Yo o n y o un g C h o an d Z ain e b M a j o k a Promoting Access to Quality Jobs for All DIAGNOSTIC PAKISTAN Yo o n y o un g C h o an d Z ain e b M a j o k a Promoting Access to Quality Jobs for All © 2020 International Bank for Reconstruction and Development / The World Bank. 1818 H Street NW, Washington, DC 20433, USA. Telephone: 202–473–1000; Internet: www.worldbank.org. Some rights reserved 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: Yoonyoung Cho and Zaineb Majoka. 2020. “Pakistan Jobs Diagnostic: Promoting Access to Quality Jobs for All.” World Bank, Washington, DC. License: Creative Commons Attribution CC BY 3.0 IGO. Translations — If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations — If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third‑party content — The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third‑party‑owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re‑use a component of the work, it is your responsibility to determine whether permission is needed for that re‑use and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202–522–2625; e‑mail: pubrights@worldbank.org. Images: © World Bank. Further permission required for reuse. ACKNOWLEDGMENTS The lead authors of the report are Yoonyoung Cho and Zaineb Majoka from the Social Protection and Jobs (SPJ) Global Practice of the World Bank. The authors would like to acknowledge the efforts of former task team leaders Laurent Bossavie and Victoria Strokova. Significant contributions were received from many World Bank colleagues at various points in the research process. Michael Weber, Joerg Langbein, Saleha Waqar, Upasana Khadka, Vincent Palmade, Soonhwa Yi, Ali Qureshi, and Andrew Beath provided critical support and background papers at various stages of the writing process. Nina Rosas Raffo and Priyanka Kanth made significant contributions to the report, especially regarding the topics of informality and skills. Dino Merotto helped use the Jobs Diagnostic Macro Tool to hash out overall macrolevel labor market trends in Pakistan. Valuable insights were shared by Maria Beatriz Orlando, Rohini Pande, Tanya Lynn D’Lima, Uzma Quresh, Saman Amir, Ghazala Mansuri, Rafay Khan, and Gonzalo Varela. Bhavya Paliwal provided excellent research and analytical support, and Fayre Makeig provided editorial support. Special thanks go to Aline Coudouel and Enrique Blanco Armas for providing advice and guidance on strengthening key messages and policy implications. The authors would also like to thank Syud Amer Ahmed, Cristina Isabel Panasco Santos, Gabi George Alam, Shabih Ali Mohib, Silvia Redaelli, and Thomas Farole for their review and valuable comments. The report benefited from discussions with and suggestions from policy makers, government officials, and researchers as well as Bank colleagues through multiple consultation workshops held in Islamabad, Lahore, and Karachi. The authors would like to acknowledge the collaboration of the Pakistan Bureau of Statistics in providing access to the microdata used in this report. This work was carried out under the leadership of Michal Rutkowski (Global Director of SPJ), Lynne D. Sherburne‑Benz (Regional Director of Human Development in South Asia), and Illango  Patchumuthu (Country Director of Pakistan), and was guided by Stefano Paternostro (Practice Manager of SPJ). The report was made possible through a grant from the Department for International Development’s Private Sector Development Trust Fund. i ABBREVIATIONS ADB.................. Asian Development Bank ASER.................Annual Status of Education Report BEOE.................Bureau of Emigration and Overseas Employment BISP...................Benazir Income Support Programme BPO...................Business Processing Outsourcing CAGR................Compound Annual Growth Rate FDI. ....................Foreign Direct Investment FLFP. ..................Female Labor Force Participation Rate FGD...................Focus Group Discussion FY. .....................Fiscal Year GCC..................Gulf Cooperation Council GDP. ..................Gross Domestic Product HEs....................Household Enterprises HIES. ..................Household Income and Expenditure Survey ICT....................Information and Communication Technology ILO....................International Labour Organization IT. .....................Information Technology KNOMAD..........Global Knowledge Partnership on Migration and Development KP(K).................Khyber Pakhtunkhwa LFP....................Labor Force Participation LFS....................Labor Force Survey LSS....................Labor Skills Survey MOPHRD...........Ministry of Overseas Pakistanis and the Ministry of Human Resources NEET. .................Not in Employment, Education, or Training NFNA. ................Non-Farm, Non-Agricultural NGO..................Non-Governmental Organization NTAVTTC. ..........National Vocational and Technical Training Commission OEC. ..................Overseas Employment Corporation OEP...................Overseas Employment Promoters OPs. ...................Overseas Pakistanis OPF...................Overseas Pakistanis Foundation PBS.................... Pakistan Bureau of Statistics POE...................Protectorate of Emigrants POEPA...............Pakistan Overseas Employment Promoters Association PSLM.................Pakistan Social and Living Standard Measurement SAR................... South Asia Region SMEs.................Small and Medium Enterprises TFR....................Total Fertility Rate TTCs.................. Trade Testing Centers TVET. .................Technical and Vocational Education and Training UNODC.............United Nations Office On Drugs And Crime WDIs................. World Development Indicators ii CONTENTS ABBREVIATIONS II FOREWORD 1 EXECUTIVE SUMMARY 3 Policies to strengthen the fundamentals of job creation.......................................................4 Policies to boost productivity ..................................................................................................5 Policies to promote the inclusiveness of the labor market ...................................................5 INTRODUCTION 7 1. MACRO DRIVERS, TRANSFORMATIONS, AND JOBS 11 1.1. Growth, poverty reduction, and job creation in Pakistan.............................................12 1.2. External factors affecting growth and job outcomes....................................................14 2. KEY LABOR MARKET OUTCOMES 19 2.1. Labor market trends over time.......................................................................................19 2.2. Labor market snapshot in 2017.......................................................................................21 2.3. Access to quality jobs.......................................................................................................26 3. PAKISTAN’S WORKFORCE 31 3.1. Education and skills..........................................................................................................31 3.2. Education, cognitive and noncognitive skills, and labor market returns ....................34 4. PAKISTAN’S ENTERPRISES 37 4.1. Business environment for enterprises.............................................................................38 4.2. Nonfarm, nonagricultural household enterprises..........................................................39 4.3. Digital jobs in Pakistan....................................................................................................45 5. URBANIZATION AND INTERNAL MIGRATION FOR JOBS 49 5.1. Urbanization in Pakistan.................................................................................................49 5.2. Internal migration and labor market implications........................................................50 6. OVERSEAS LABOR MARKET  55 6.1. Trends in outmigration from Pakistan............................................................................55 6.2. A migrant’s journey..........................................................................................................57 7. CONCLUSIONS AND POLICY DIRECTION 63 7.1. Brief summary of main findings......................................................................................63 7.2. Initial policy direction......................................................................................................65 REFERENCES 71 APPENDICES 77 Appendix A: Summary of background papers and data sources used................................77 Appendix B: Regression analysis of employment outcomes................................................78 Appendix C: Policy and institutional analysis of international migration in Pakistan..........79 iv FOREWORD Pakistan has made remarkable progress in alleviating poverty over the past 15 years. The share of the population living in poverty decreased from 64.3 to 24.3  percent between 2001 and 2015, as more than 30 million individuals rose above the poverty line. For Pakistan to sustain this progress and to realize the ambition of becoming an upper-middle-income country by 2047, it is critical that the country’s most abundant asset— labor—be utilized in productive income-generating activities. This is particularly so since the country’s workforce is young, providing great potential to reap demographic dividends. However, data indicate that the country’s human resources are underutilized, with low productivity levels on average. Labor market data from 2017 suggest that a total population of 208 million relies on the incomes of only 46 million earners. This is partly because of the country’s demographic structure, whereby 47 percent of the population is either over 65 or under 15. But even among those of working age, approximately half are out of the labor force, in large part due to a low rate of female participation in the labor force. Moreover, a significant share of workers in the labor market is in nonpaid or low-paying jobs. Amid the current demographic transition—characterized by declining fertility rates and a growing share of working-age individuals in the total population—policy efforts are urgently required to address the underutilization and low productivity of Pakistan’s human resources. This multisectoral Jobs Diagnostic provides a comprehensive overview of Pakistan’s job market, with the aim of informing key policy areas. It investigates a host of factors associated with labor outcomes, such as the general macroeconomic environment and job-specific conditions, workers’ education and skill levels, labor market segmentation and disparities, structural transformation, and workers’ mobility. It highlights issues related to job creation, quality, and access by looking at past and present trends in the labor market. It also provides a global and regional comparison of key dimensions of the job market, in an effort to assess Pakistan’s position. The Diagnostic highlights a prioritized set of interventions to address key priorities on the jobs agenda. These include policies and programs that aim to invest in the human capital of the future workforce, help the current stock of workers improve their productivity, promote women’s economic empowerment and thus unlock their potential contribution to the economy, support a business environment conducive to entrepreneurial activities and global competition, and provide stable macroeconomic conditions. Labor mobility, both domestic and international, and the technological advances that are pushing the frontier of emerging jobs should also be considered when setting labor policy. Jobs are at the center of the country’s sustained growth and shared prosperity. Given the complex and daunting nature of Pakistan’s job-related challenges, the Diagnostic calls for concerted policy efforts on multiple levels. Along with the World Bank study “Pakistan@100: Shaping the Future,” we hope that the Diagnostic will help stimulate debates on the policy priorities and instruments that will do the most to improve job prospects for the Pakistani people—benefiting their lives and the country’s future. Lynne D. Sherburne-Benz llangovan Patchamuthu South Asia Regional Director Country Director Human Development Vice Presidency Pakistan 1 EXECUTIVE SUMMARY Pakistan is undergoing a demographic transition as the number of working-age individuals continues to grow faster than the overall population. This could help create an environment conducive to economic growth and demographic dividends. An economy can realize demographic dividends when total fertility rates decline and the share of the working-age population grows, leading to a decline in dependency ratios and allowing for more savings and greater investment in human and physical capital. However, a demographic transition alone does not automatically lead to dividends. Three key aspects of the job market—the quantity, quality, and inclusiveness of jobs—must be simultaneously addressed to reap the potential benefits of a young labor market. Quantity of jobs Pakistan needs more workers to engage in productive activities and support the economy’s growth. If the market is to absorb more of the people currently out of the labor force, more job opportunities need to be created. In 2017, the share of the employed (57 million) out of the total population in the country (208 million) was only 27 percent. This means that one breadwinner provides for 3.6 individuals in Pakistan—a significantly higher ratio than in other countries (e.g., 2.7 people per breadwinner in Bangladesh and 1.9 in Vietnam). This is in part because almost half of the working-age population is out of the labor force, mainly due to a very low female labor force participation (FLFP) rate and a high percentage of youth who are not in education, employment, or training (NEET). Hence, more jobs are needed to induce potential workers who are currently out of the labor force to engage in it. Quality of jobs The quality of jobs,1 reflecting labor productivity, remains a significant challenge. This is especially true given the limited degree of structural transformation in the labor market to date.2 Transitions out of agriculture into industry and services have been slow, and the growth of productivity within sectors has been negligible. Slow structural transformation is in part due to the economy’s frequent boom-bust cycles: the share of industry and services has rapidly increased in good times, but then this progress has inevitably been reversed during an ensuing bust. If the economy had grown steadily without much fluctuation, the overall change in the quality of jobs could have been greater. Informality is prevalent in Pakistan, and is often considered as a source of low-quality jobs.3 Agricultural jobs, most of them informal, provide employment opportunities to a large share of workers: in 2017, over 40 percent of workers in Pakistan were employed in agriculture. In addition, a significant share of households (28 percent in urban areas and 17 percent in rural areas) reported being self-employed in nonfarm, nonagricultural household enterprises, most of which were also likely to be informal.4 Even among wage employees, the share of formal 1 The quality of jobs is assessed by the employment sector and status as well as earnings. Given that earnings information is available only for wage workers, and labor productivity is measured mostly in aggregate, the sector and status of employment are criteria commonly used for measuring the quality of jobs. 2 See Sharma, Nayyar, and Kim (2019) for more discussion of structural transformation in Pakistan. 3 Definitions and measures of informality vary depending on whether the focus is on enterprises or workers. 4 Close to 60 percent of nonfarm, nonagricultural household enterprises report having no workers but the owner, and 25 percent are family businesses run by the owner and nonpaid family workers. Among respondents to the 2014 Labor and Skills Survey (LSS), 77 percent of household enterprises were informal, with only 23 percent registered. 3 workers (with a written contract) was small (only about 26 percent in 2017), while their earnings were more than double those of informal wage workers. Many of the informal workers in the labor market engage in income-generating activities only to find that their earnings are not sufficient to lift them and their families out of poverty. The”working poor” status is associated with workers’ characteristics, such as gender, education, location, and status and sector of employment. People can gain access to better-quality jobs by moving. In Pakistan, spatial mobility is characterized by rural- to-urban and international migration. The large majority of within-province movement has been to urban destinations. In particular, there was a significant increase in the population of each capital district between the 1998 and 2017 censuses. There is suggestive evidence that internal migrants have better labor market outcomes (e.g., more wage employment, less of it in agriculture, and higher earnings) than those who stay put. Similarly, a significant number of individuals from Pakistan pursue better-quality jobs overseas. The number of labor migrants leaving the country reached a peak of 900,000 in 2015, although this volume had dropped to 500,000 in 2017. Overseas employment provides higher earnings than domestic, and remittances—equivalent to 7 percent of gross domestic product in 2017—are a critical source of funds, as foreign direct investment is limited and foreign reserves are low. Inclusiveness of jobs There is huge heterogeneity in access to quality jobs across workers’ characteristics and locations. Differences across the male/female and urban/rural divides are among the most notable. The FLFP rate has remained stubbornly low: it was gradually improving up until 2014, and then fell significantly between 2014 and 2017 (from 32.9 percent to 28.3 percent). Meanwhile, the urban FLFP rate is strikingly low and stagnant—it remained at about 11 percent during the period 2001–17. Frequently mentioned constraints include women’s lack of mobility, household responsibilities, social and individual perceptions of women’s work, and road and workplace safety issues.5 In addition, the quality and variety of jobs available to women is rather limited: the majority of employed women are almost exclusively in agriculture (78 percent) and manufacturing (19 percent), with a large share of working women being unpaid and underemployed. Other than gender, educational attainment levels and age are important determinants of access to quality jobs. The share of youth (ages 15–24) who are in the NEET category is high, showing large variations across the male/ female and urban/rural divides. Moreover, as the employment structure differs widely between urban and rural areas, so does the quality of jobs. The heterogeneity of labor market outcomes across different types of workers is further compounded by large variations across provinces. POLICIES TO STRENGTHEN THE FUNDAMENTALS OF JOB CREATION While the private sector creates jobs, the government can adopt measures to make the economic environment more conducive to quality job creation in line with strategies for growth. ∫ Ensure macroeconomic stability and move away from long-standing economic fluctuations. Measures such as modifying tax policies to improve revenue, removing distortionary subsidies, reforming public utilities and energy sectors to rationalize public spending, and strengthening exports through a robust exchange rate policy can potentially lead to stability. ∫ Improve the business environment by making it easier for firms to operate without disruption. This can be done primarily by reforming the processes and regulations that facilitate the entry of new business and accumulation of physical capital (e.g., construction permits, property registration, enrollment in tax networks, regular tax payments, etc.). 5 See Amir et al. (2018) and Amir and Pande (2019). 4 ∫ Foster exports’ competitiveness through initiatives that help reduce production and information costs. For this purpose, the government can promote a gradual reduction in tariffs on both intermediate and final goods. In addition to creating an enabling environment, efforts are needed to increase and upgrade inputs for production—especially physical and human capital—to strengthen the fundamentals. ∫ Mobilize resources to increase investment in physical capital. Reforming tax policy or facilitating the influx of foreign direct investment and remittances by making processes more straightforward and accountable can make resources available to firms for investment. ∫ Increase public financing in human capital investment, as it can potentially boost innovation and entrepreneurship in the long run. Human capital investments can be enhanced with measures that promote informed decisions on parenthood; early childhood health, nutrition, and development; school readiness; universal primary education; and the successful transition to secondary school with a particular focus on the poor and girls. POLICIES TO BOOST PRODUCTIVITY Given the heterogeneity in the productivity of informal firms and workers, there is room for interventions that give firms access to the advantages of operating in the formal sector and facilitate their pursuit of growth. ∫ Provide business development services for local microenterprises to address their growth challenges. Providing access to credit and opportunities for managerial capacity building, and promoting noncognitive skill building and a pro-growth mindset, can improve the efficiency of firms while allowing them to expand, grow, or undertake risks. ∫ Support informal workers in gaining access to the benefits of formality. This can also reduce the cost of firms. Practices such as adopting a standard contract do not incur any additional cost to employers even as they provide invaluable protection to workers. Similarly, where access to formal social insurance is lacking, the government can design and implement mechanisms through which informal workers can access contributory pensions or savings. There is a strong need to help poor and unskilled workers overcome key challenges to higher earnings in an environment where private sector jobs are already limited. ∫ Provide skill development opportunities. These could include basic literacy and numeracy, entrepreneurial skills, and noncognitive skills and social empowerment. ∫ Provide productive inclusion (also known as graduation) packages to enhance the labor market prospects of the ultra-poor. Based on global evidence, productive inclusion programs are successful when they adopt a holistic approach that (i) promotes basic protection (e.g., health, nutrition, consumption), (ii) supports livelihood activities, (iii) deepens financial inclusion, and (iv) strengthens social empowerment. Given the returns to domestic and international migration, policy measures can help workers take more advantage of such opportunities. These should focus on mitigating the risks associated with moving to a new place. ∫ Provide resources (financial and nonfinancial) to pay for the cost of migration and mitigate the risks. ∫ Adapt safety nets and labor programs to the urban context to help vulnerable workers (many of whom are migrants from rural areas) cope with risks and enhance their resilience. ∫ Manage international migration (and remittances) more systematically for safe, inclusive, and productive migration. POLICIES TO PROMOTE THE INCLUSIVENESS OF THE LABOR MARKET First and foremost, the highest priority on the policy agenda for labor market inclusiveness is to increase women’s economic empowerment. The Global Gender Gap Index 2018 produced by the World Economic Forum ranked Pakistan 148th out of 149 countries (WEF 2018). Deep-rooted social norms that affect women’s human capital accumulation and labor market status, as well as civic and political activities, are often pointed to as the culprit. 5 ∫ Address gender norms, lower procedural costs for women’s economic activities, and promote girls’ education to promote women’s overall economic empowerment. Awareness-raising activities and social messaging can change the perceptions of women’s work and mobility as well as shared household responsibilities. ∫ Promote income diversification and secondary activities in rural areas. Promoting secondary activities in agriculture (e.g., cash crops, livestock, petty trades, and food processing) can improve women’s productivity. ∫ Upgrade agricultural practices along the value chains for women . Improving female agricultural workers’ skills, knowledge, and linkages to value chains can significantly enhance agricultural productivity. ∫ Adopt measures that require or incentivize employers to provide female-friendly work environments (e.g., separate workspaces, transportation, childcare). Incentives (e.g., tax credits/exemptions or subsidies; gender equity certifications) combined with regulations (such as for workplace safety) can promote female- friendly workplaces that help mitigate employers’ costs while meeting the needs of women in the workplace. ∫ Harness information and communication technology (ICT) and social sector jobs (such as caregivers, health workers, social workers, etc.) for relatively well-educated women in urban areas. Promoting ICT jobs among women, for instance, can directly address mobility constraints while also allowing them to work flexible hours. The disadvantages that young people face in the labor market should not be overlooked. Despite young people’s higher educational attainment (compared with their older counterparts), the agricultural share of employment was about 40 percent among youth—a level similar to that of older workers. The following measures are proposed to address several of the challenges associated with at-risk youth: ∫ Strengthen youth training and entrepreneurship initiatives . Training programs must include supplementary services such as job search assistance, job placement, and on-the-job training. Instead of focusing only on financing, Pakistan’s entrepreneurship programs require a more holistic approach that combines technical support and coaching/mentoring. ∫ Make agriculture attractive for youth. This can be done by supporting the adoption of new technologies, the use of ICT, green fertilizers and pesticides, commercial crops, and nonfarm rural businesses. ∫ Provide job search assistance (and intermediation). Job fairs, the provision of information on the availability and requirements of jobs, and clinics supporting job seekers in updating of their resumes are among options to consider. ∫ Provide opportunities to build workplace skills and experience. These may include on-the-job training, internships/apprenticeships, and wage subsidies. ∫ Enhance the focus on soft skills. All the labor market programs mentioned above could integrate the development of soft skills (e.g., effective communication, creative thinking, etc.). Among the many policy options discussed, a prioritized set of interventions to address key challenges merits further emphasis. Priorities should be given to those efforts that are likely to have large spillover impacts on overall growth and the economy including on job markets (e.g., reform of tax policies, managed international migration). Also, those whose promised benefits are proven based on global research (e.g., early childhood development, business regulation reforms, promotion of soft skills); and those that can build on existing initiatives and development (e.g., youth programs, efforts to boost ICT) require further investment. 6 INTRODUCTION Pakistan’s economy has seen impressive poverty reduction despite its short-term volatility and vulnerable macroeconomic conditions. The Pakistan economy has been experiencing repeated boom-bust cycles, with large gaps in the growth rates between good and bad times. In spite of that, poverty reduction has made steady progress: the share of population living in poverty decreased from 64.3 percent in 2001 to 24.3 percent in 2015.6 The income growth of poorer households, boosted by remittances from international migrants, and flourishing informal economic activities associated with urbanization and household enterprises, are likely to have contributed to the substantial poverty reduction—beyond what would have been suggested by aggregated income growth. On this positive note, Pakistan is aspiring to become an upper-middle-income country by its centenary in 2047 (World Bank 2019a). In sustaining this progress and realizing ambitious aspirations, Pakistan’s large and growing population can be its main asset if the workforce is absorbed in productive activities through quality jobs. Pakistan’s population, about 208 million in 2017, continues to grow. While this exerts huge pressure on limited resources (physical and financial) in the country, it provides an abundant labor supply. Specifically, the country is experiencing a youth bulge, as the number of individuals entering the labor market expands at a faster rate than the total population. This can be turned into demographic dividends if the economy is able to absorb these young workers entering the labor force into quality jobs, and ensure that everyone—whether man or woman—has the opportunities and skills to carry out productive activities. Jobs—more, better, and inclusive—are therefore at the center of development in Pakistan. Leveraging the country’s critical asset (that is, its people) to generate incomes, whether through wage employment, or a range of household-based enterprises and other informal activities, or through overseas employment, is the most sustainable way out of poverty for individuals and households while boosting growth. Jobs provide benefits critical to development, such as upgraded workforce skills, enhanced productivity and competitiveness across the economy, women’s empowerment, and social cohesion through productive engagement of youth, especially in conflict-stricken contexts (World Development Report 2013: Jobs [world Bank 2012a]). How effectively the economy utilizes the accumulated human capital will also determine individuals’ and households’ investment in the human capital of the future generation. This report presents a comprehensive diagnostic of Pakistan’s jobs with the goal of better understanding the gains made to date as well as constraints to more, better, and inclusive jobs in the country. Job outcomes depend heavily on the country’s macroeconomic conditions as well as on the regional and global environment for trade. Most jobs are and should be created by the private sector, and understanding demand-side constraints to job creation helps identify policy measures to boost job creation. At the same time, the supply-side capability of workers and their skills (human capital) determine the productivity and quality of jobs. This Jobs Diagnostic aims to investigate a range of areas that influence job outcomes, and to provide an evidence base for policy development to promote more, better, and inclusive jobs in Pakistan. This Jobs Diagnostic takes a multisectoral approach that goes beyond traditional analytical tools to comprehensively analyze issues related to labor demand and supply, the market as well as policy environment, and the bigger picture along with its determinants. Standard labor analysis tends to focus on specific aspects of job outcomes (e.g., skills) without considering other dimensions influencing job creation, quality, and access. On the other hand, a broad growth diagnostic exercise tends to have no direct links with job outcomes. 6 Based on a new poverty line revised by the Pakistan Bureau of Statistics in 2016. See Redaelli (2019) for further discussion on poverty estimates. 7 In this context, a multisectoral Jobs Diagnostic aims to assess multiple dimensions and factors, and extend the examination of past trends into discussions of future jobs. The report is based on a number of background papers (see Appendix A for a summary of the main background papers, descriptions, and data sources). The base reference period is 2001–17, but based on the availability of data, other years may be covered and discussed.7 Box 1 provides definitions of key terms used throughout the report. These are based on the official definitions set by the International Labour Organization before its adjustment in 2013.8 BOX 1: DEFINITIONS OF EMPLOYMENT STATUS AND TYPE USED IN THE REPORT The employment-related terms and concepts used throughout the report are based on the official employment definitions of the International Labour Organization (ILO). The working-age population is classified into two groups: active and inactive. The active population is further divided into employed and unemployed. An individual is considered employed if he or she worked at least one hour over the previous seven days, even if the economic activity was unpaid. An individual is considered unemployed if he or she worked less than one hour in the past seven days, is actively looking for work, and is available to work. In this report, we further subdivide the employed population by type of employment. Definitions of the key terms used in this report are provided below. Labor Force Working-age population Individuals ages 15 to 64 Labor force (active population) Individuals who were employed or unemployed during the week prior to the survey Labor force participation rate Share of working-age population that is in the labor force Individuals who performed any type of economic activity (paid employment, Employed self‑employment, or unpaid work) for at least one hour a week prior to the survey, or who had a permanent job but were not working during the week prior to the survey Employment ratio (employment to Share of the working-age population that is employed working-age population ratio) Individuals who do not work (are not in employment) but are actively looking for Unemployed a job and are available to work Unemployment rate Share of the labor force that is unemployed Individuals not in the labor force; this includes all individuals who were not Inactive (out of the labor force) employed or unemployed during the week prior to the survey, including because they were in school, were retired, or were ill or disabled NEET (not in employment, education, Individuals of working age who are not employed and not attending school or training or training) Types of Employment Employed individuals who have declared being employed by an employer outside Wage employment the household and are paid (either with or without a contract, in cash or in kind); this includes regular wage employees as well as day labor/casual workers Employed individuals who work on their own account in agricultural (including Self-employment farming activities) and nonagricultural sectors to generate income Unpaid (family) worker Employed individuals who are not receiving remuneration 7 For the macro analysis using WDI and Jobs Structure Tool, the reference period is 2000–17. When the labor force survey is the main source of data, the reference period begins in 2001. 8 The norms for employment statistics were adjusted following the 19th Conference of Labor Statisticians held in 2013, but comparable employment indicators cannot be built retrospectively using the new indicators. To ensure comparability, this report builds comparable employment statistics based on the definitions used before the adjustment. 8 Formality of Wage Employment Formal employment Employed individuals who have declared being wage workers with a written contract Formal wage employees in the civil service, a para-public enterprise, or an Formal public international organization Formal private Formal wage employees in private firms Employed individuals who have declared being wage workers without a written Informal private wage employment contract Regularity of Wage Employment Regular employment Workers who have fixed, monthly wages with a regular job Nonregular employment Workers who serve as casual labor with piecemeal and output-based payment Self-employment Employer Self-employed individuals who employ at least one paid employee Self-employed individuals who do not employ any paid employee, but may employ Own-account worker unpaid (family) workers Source: Authors’ compilation. Key data sources used in the report are as follows (Table 1). While the macro analysis draws primarily on internationally comparable data from sources such as the World Bank’s World Development Indicators (WDI) database, the primary sources for the micro analysis are microdata produced by the Pakistan Bureau of Statistics (PBS), most notably the Labor Force Survey (LFS) and Household Income and Expenditure Survey (HIES). Table 1 Summary of Key Data Sources Used in the Jobs Diagnostic Main data source(s) Years covered in the analysis Source World Development Indicators (WDIs) Mostly 2000–17 World Bank Macro analysis National accounts FY  b 2000–17 Pakistan Bureau of Statistics (PBS) Labor Force Survey (LFS) FY2000–2018, various years PBS Labor Skills Survey (LSS) FY2014–16 World Bank Supply side Household Income and Expenditure Survey (HIES) and Pakistan Social and FY2000–16, various years PBS Living Standard Measurement (PSLM) HIES household enterprise modules FY2007, FY2016 PBS Demand side Doing Business Indicators Various years World Bank KNOMAD Migration Cost Survey  a 2015 World Bank Source: Authors’ compilation. a Conducted by the World Bank’s Global Knowledge Partnership on Migration and Development (KNOMAD). b The Government of Pakistan’s fiscal year (FY) runs from July 1 through June 30; so, for example, fiscal year 2017 runs from July 1, 2016, through June 30, 2017. 9 The remainder of the report is organized into several topical chapters as follows: ∫ Chapter 1. Macro drivers, transformations, and jobs: The first chapter of the report sets the context by providing an analysis of the evolution of outcomes in terms of growth, poverty reduction, and external factors associated with growth and jobs over the period 2000–17. ∫ Chapter 2. Key labor market outcomes: This chapter reviews the trends of key labor market indicators between 2001 and 2017, and highlights major labor market challenges using the latest labor statistics. Using stylized facts about labor force participation, unemployment, type and status of employment, and job quality, this section investigates the quantity, quality, and inclusiveness of jobs in Pakistan’s labor market. ∫ Chapter 3. Pakistan’s workforce: This chapter investigates the education and skill level of the workforce, and their labor market returns. The educational attainment and cognitive and noncognitive skills, and the heterogeneous association with earnings, by gender and type of employment are examined. ∫ Chapter 4. Pakistan’s enterprises: This chapter assesses the channels for job creation. Given that microdata on firms and enterprises that are sources of jobs for wage employees are lacking in Pakistan, the analysis focuses on two specific sources of jobs: nonfarm, nonagricultural household enterprises and digital jobs. ∫ Chapter 5. Urbanization and internal migration for jobs: This chapter reviews urbanization patterns and labor mobility within the country, and examines the labor market returns to migration. Labor movements between rural and urban areas within and across provinces highlight the dynamic forces in the labor market masked in the static aggregate numbers. ∫ Chapter 6. Overseas labor market: This chapter investigates the trends and patterns of international labor migration out of Pakistan. Huge opportunities as well as risks associated with international migration are discussed. ∫ Chapter 7. Conclusions and policy direction: The final chapter of the report reviews the main conclusions from the rest of the report and outlines a broad framework for considering policies to address the job-related challenges identified. 10 1. MACRO DRIVERS, TRANSFORMATIONS, AND JOBS Pakistan’s most abundant asset is labor, and human capital comprises a significant share of the country’s wealth. As the world’s fifth-most-populous country, with a population of almost 208 million, labor is an indispensable asset for Pakistan. A recently released report (Lange, Wodon, and Carey 2018) estimates the changing wealth of 141 nations over 20 years, from 1995 to 2014,9 and suggests that human capital wealth10 is by far the largest component, accounting globally for two-thirds of the total wealth. Pakistan is not an exception: human capital’s share of total wealth was estimated to be around 61 percent in 2014, growing from 46 percent in 1995. If this abundant asset is well utilized and invested, it could bring large returns. Moreover, Pakistan is experiencing a favorable demographic transition—a youth bulge, as the share of the young, working-age population grows at a faster rate than the dependent population (Figure 1 and Figure 2). This provides an opportune time to benefit from demographic dividends. The share of young dependents is falling amid declines in fertility rates and that of elderly dependents remains small. Pakistan’s youth bulge can generate huge benefits if the economy is able to absorb the substantial number of young workers entering the labor force by creating a sufficient number of good-quality jobs and to ensure that individuals of working age—both men and women—participate in the labor force (Ahmed, Cho, and Fasih, 2019). However, a demographic transition alone does not automatically lead to dividends. Demographic dividends can be reaped only if the economy creates sufficient jobs to absorb the large number of working-age people, workers engage in more productive activities, and more individuals participate in the labor force. The review of Pakistan’s macroeconomic environment in the following section suggests several macroeconomic constraints to the creation of more, better, and inclusive jobs. Figure 1 Figure 2 Population by five-year age group in 2017 Population by five-year age group in 2050 (projection) 2017 2050 100 100 90 90 80 80 FIVE-YEAR AGE COHORT FIVE-YEAR AGE COHORT 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 20 15 10 5 0 5 10 15 20 20 15 10 5 0 5 10 15 20 POPULATION (MILLIONS) POPULATION (MILLIONS) MALES FEMALES MALES FEMALES Source: UN World Population Prospects 2017 Revision, as cited in Ahmed, Cho, and Fasih, (2019). 9 National wealth is defined as the sum of produced capital such as factories and infrastructure, 19 types of natural capital (such as oil, minerals, land, and forests), human capital, and net foreign assets. 10 Defined as the discounted value of the future earnings of a country’s labor force. 11 Figure 3 Figure 4 Real annual GDP growth, 2000–17 Value added and annual growth of three key sectors, 2000–17 REAL ANNUAL GDP GROWTH: 2000–2017 VALUE ADDED BY MAJOR SECTOR, PAKISTAN 10 300,000 9 9 REAL ANNUAL GDP GROWTH (%) 8 250,000 MILLION CONSTANT 2010 USD 8 ANNUAL GROWTH RATE (%) 7 7 200,000 6 6 5 5 150,000 4 4 3 100,000 3 2 2 1 50,000 1 0 0 0 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 0 20 1 20 2 20 3 20 4 20 5 20 6 20 7 20 8 20 9 20 0 20 1 20 2 20 3 20 4 20 5 20 6 17 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 20 PAKISTAN LOWER MIDDLE INCOME BANGLADESH SOUTH ASIA (EXCLUDING PAKISTAN) AGRICULTURE INDUSTRY SERVICES VALUE-ADDED GROWTH (RHS) Source: World Development Indicators and Jobs Structure Tool. 1.1. GROWTH, POVERTY REDUCTION, AND JOB CREATION IN PAKISTAN Pakistan’s growth trajectory has been highly volatile with repeated boom-bust cycles. Between 2000 and 2017, the period that this report focuses on, the volatile patterns continued and economic performance can be categorized in three phases: (i) from 2000 to 2005, accelerating growth (“boom”); (ii) from 2006 to 2011, downward growth (“bust”); and (iii) from 2012 to 2017, slow upward growth (“recovery”) (Figure 3). The gap between the boom-bust growth rates (close to 6 percentage points) appears large and the country’s recession was deeper than that of peer countries. During this period, the service sector was the largest contributor to value added (Figure 4). The share of industry in value added increased from a very low base, and as of 2017, agriculture and industry each composed a similar share of GDP. An estimation of jobs’ elasticity to growth in multiple countries suggests that Pakistan’s ability to create jobs, given its growth performance, has been quite strong (Figure 5). In Pakistan, GDP growth of 1.0 percent is associated with 0.65 percent growth in the number of jobs. This elasticity is significantly larger than in other countries (Figure 6) and in part explains the country’s low unemployment rate despite a large and increasing number of labor market entrants.11 This means that the challenges facing Pakistan’s job market so far are more about the quality of jobs than their quantity. Despite fluctuating growth, maybe due to steady job creation, the country’s poverty rate fell from 64.3 percent in 2001 to 24.3 percent in 2015.12 This indicates that over 32 million individuals were lifted from poverty during this period. Poverty reduction was significantly more prominent in urban than rural areas: the poverty headcount fell from 50 to 12.5 percent in urban areas whereas it fell from 70.2 to 30.2 percent in rural areas. This coincided with a reduction in the share of rural population from 71 to 65 percent amid rapid urbanization.13 There are regional and distributional disparities in the pace of poverty reduction, however, and the gaps seem to be widening. According to the World Bank’s Global Database of Shared Prosperity, the trends of shared prosperity (measured as the average consumption of the poorest 40 percent of the population) suggest that the consumption growth among the bottom 40 percent was slower than that for the average of the population between 2001 and 2015. 11 This is in line with the estimation in Farole and Cho (2017) with a different reference period. 12 Actual coverage periods in the HIES and LFS are based on fiscal years. The surveys here referred to as being for 2001 and 2015 cover fiscal years 2002 and 2016. 13 According to the Population Census of 1998, “All localities which were metropolitan corporations, municipal corporations, municipal committees, town committees or cantonment at the time of the Census were treated as Urban.” Then any population, housing, or territory not included within an urban area can be considered as rural. 12 Figure 5 Figure 6 Growth rates of GDP and jobs, 2000–17 Jobs’ elasticity to growth, 2000–17 GDP AND JOBS GROWTH RATES: 2000–2017 JOBS ELASTICITY TO GROWTH 10 VIETNAM 0.29 8 ANNUAL JOBS GROWTH RATE (%) INDONESIA 0.32 6 SRI LANKA 0.11 4 PAKISTAN 2 INDIA 0.20 ID BD IN LK 0 BANGLADESH 0.37 -2 PAKISTAN 0.65 -2 0 2 4 6 8 10 12 ANNUAL GDP GROWTH RATE (%) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Sources: Jobs’ growth rates are based on LFS in Pakistan and ILO modeled data for other countries; GDP growth rates are from WDI. Note: BD = Bangladesh; ID = Indonesia; IN = India; LK = Sri Lanka. A closer analysis of the sources of growth in per capita value added and labor productivity show Pakistan is lagging behind its peers. During the period 2000–17, the value added per capita increased by 2.0 percentage points, of which 65 percent is explained by labor productivity growth and the rest by demographic change and employment and participation rates. Both the growth rate and share of productivity in explaining the growth are far lower in Pakistan than other countries (Figure 7). Overall productivity increases when workers shift from lower- to higher-productivity sectors, or when productivity within each sector improves, or both at the same time. Compared with its peers during the same period, Pakistan’s productivity gains have been very limited within both agriculture and industry, and relatively few jobs have shifted to the service sector (Figure 8). The pace of the workforce’s transition from agriculture to industry and services—a measure of structural transformation—varied by the macroeconomic performance of the economy. During the high-growth period of 2000–05, the share of agriculture in employment fell significantly, while productivity slightly increased. However, during the subsequent recession (2006–11) and recovery (2012–17) periods, the employment share and Figure 7 Figure 8 Cross-country comparison of decomposition of growth in per capita value Cross-country comparison of decomposition of labor productivity change, added, 2000–17 2000–17 DECOMPOSITION OF GROWTH IN PER CAPITA VALUE ADDED, BY COUNTRY CONTRIBUTION OF CHANGE IN PRODUCTIVITY (VALUE ADDED PER WORKER), BY COUNTRY AND MAJOR SECTOR VIETNAM 2000–17 VIETNAM 2000–17 TOTAL=5.3% TOTAL=4.5% INDONESIA 2000–17 INDONESIA 2000–17 TOTAL=3.9% TOTAL=3.7% SRI LANKA 2000–17 SRI LANKA 2000–17 TOTAL=4.5% TOTAL=4.6% INDIA 2000–17 INDIA 2000–17 TOTAL=5.7% TOTAL=5.7% BANGLADESH 2000–17 BANGLADESH 2000–17 TOTAL=4.6% TOTAL=4.0% PAKISTAN 2000–17 PAKISTAN 2000–17 TOTAL=2.2% TOTAL=0.9% -2.0 0 2.0 4.0 6.0 8.0 -2.0 0 2.0 4.0 6.0 8.0 ANNUAL CHANGE (PERCENTAGE POINTS) ANNUAL CHANGE (PERCENTAGE POINTS) WITHIN AGRICULTURE WITHIN INDUSTRY WITHIN SERVICES PRODUCTIVITY PARTICIPATION RATE INTER-SECTORAL INTER-SECTORAL INTER-SECTORAL EMPLOYMENT RATE DEMOGRAPHIC CHANGE AGRICULTURE INDUSTRY SERVICES Source: Jobs Structure Tool based on LFS for labor statistics in Pakistan and WDI. 13 Figure 9 Figure 10 Changes in productivity and employment share of three key sectors, 2000–17 Agricultural output in Pakistan and selected countries, 2005–15 CORRELATION BETWEEN CHANGE IN SECTORAL PRODUCTIVITY AGRICULTURAL OUTPUT (2005=100) AND EMPLOYMENT SHARES, BY PERIOD 150 8.0 ANNUAL CHANGE IN PRODUCTIVITY, 7.0 140 6.0 SERV 5.0 130 4.0 PERCENT 3.0 120 2.0 SERV 1.0 AGRIC AGRIC 110 0 SERV IND IND -1.0 IND 100 -2.0 AGRIC -3.0 90 -6.0 -4.0 -2.0 0 2.0 4.0 6.0 8.0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 ANNUAL CHANGE IN EMPLOYMENT SHARE, PERCENT PAKISTAN CHINA VIETNAM 2000–2005 2006–2011 2012–2017 BANGLADESH INDIA Source: Jobs Structure Tool based on LFS for labor statistics in Pakistan and WDI. Source: International Agricultural Productivity database: https://www.ers. usda.gov/data-products/international-agricultural-productivity. Note: The size of circles reflects the size of employment in the end of each period. productivity of agriculture remained stagnant (Figure 9). Meanwhile, the service and industry sectors experienced expansions in employment during the high-growth period, as they absorbed workers shifting away from agriculture. Then, in subsequent periods, industry continued to grow as discussed above, but the service sector was unable to absorb workers while experiencing a productivity boost. During the recent recovery period, the productivity of the service sector improved by 6 percentage points, but the sector shed jobs and was unable to create more. Its employment share declined by 2.3 percentage points, offsetting the gains of the high-growth period. Overall, there is a negative correlation between sectoral productivity and employment shares throughout all periods. While a large number of individuals still remain in agriculture, the sector’s productivity growth has been slow. Pakistan’s agriculture output index, using the output of 2005 as a reference point of 100, shows that growth in the sector has been slow compared with the country’s peers: it grew by 28 percent over the decade 2005–15 whereas that of Bangladesh, China, Vietnam, and India, for instance, grew by 36 to 42 percent over the same period (Figure 10). Studies suggest that a limited crop mix and overdependence on resources such as water and lack of machinization are likely associated with productivity growth (or lack thereof) in the sector. 1.2. EXTERNAL FACTORS AFFECTING GROWTH AND JOB OUTCOMES Global market integration, remittances, investment, and domestic demand are all important factors in growth and job creation. Pakistan’s market integration—reflected by the goods and services imported and exported as shares of GDP—shows that the country’s utilization of the global market is low and diminishing. Imports as a share of GDP remained as high as 30 percent in the region whereas in Pakistan, they hovered around 20 percent, until they started diminishing in 2012 (Figure 11). Similarly, exports also show that Pakistan is lagging behind in its sales to the global market (Figure 12). The share of exports has been decreasing since 2003 and currently is significantly lower than among Pakistan’s peers. Pakistan’s low level of integration into the global market is in large part due to its limited integration in the regional market. One major constraint is that Pakistan still maintains high tariff barriers, particularly for its neighbors. In 2017, Pakistan’s tariffs were three times as high as those in Southeast Asia (World Bank 2019a), and have resulted in Pakistan’s trade share with its neighbors being almost negligible. Moreover, the tense geopolitical rivalry between India and Pakistan has contributed to South Asia being the least integrated region 14 Figure 11 Figure 12 Imports as share of GDP, 2000–17 Exports as share of GDP, 2000–17 IMPORTS AS SHARE OF GDP EXPORTS AS SHARE OF GDP 35 35 30 30 25 25 SHARE (%) SHARE (%) 20 20 15 15 10 10 5 5 0 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 PAKISTAN INDIA SOUTH ASIA PAKISTAN INDIA SOUTH ASIA BANGLADESH LOWER MIDDLE INCOME BANGLADESH LOWER MIDDLE INCOME Source: World Development Indicators. globally.14 The World Bank estimates that if Pakistan starts trading with its neighbors, this could cause Pakistan’s economy to grow as much as 30 percent by 2047. Another factor that lowers external trade growth is Pakistan’s weak trade competitiveness: the country was ranked 107 out of 140 in 2018 based on 12 indicators, including an enabling environment, markets, human capital, and an innovation ecosystem.15 This ranking is behind its peers—Bangladesh at 103, India at 58, Sri Lanka at 85, and Vietnam at 77. Pakistan’s export structure has not changed much over time. In 2017, Pakistan’s total exports were valued at US$24.8 billion, which made it the 68th largest exporter in the world. This is significant growth from US$8.1 billion in 2000, although the share as a percentage of GDP declined during this time. Main export items in 2017 were from the textile sector: house linens (13 percent), nonknit men’s suits (8 percent), and nonretail pure cotton yarn (5 percent). There was little change in the composition of leading exports since 2000 (Figure 13). To the contrary, over the same period, Vietnam’s volume of exports increased from US$16.1 billion to US$220 billion with a significant change in export structure (Figure 14). Investment levels in the economy, measured by gross fixed capital formation and foreign direct investment (FDI) as a percent of GDP, are lower in Pakistan than in its peers (Figure 15 and Figure 16). On average, Pakistan’s investment-to-GDP ratio (about 16 percent) was 8 percentage points lower than that of its peers in 2001. The gap widened after some fluctuations, and in 2017, Pakistan’s investment-to-GDP ratio still remained around 15 percent, more than 10 percentage points lower than the regional average of 27.2 percent. With respect to FDI flows, Pakistan was in an upward trend and performed as well or better than its peers until 2007. FDI inflows were highly concentrated in three sectors: oil and gas exploration, communications, and finance. With robust economic growth, FDI grew significantly in 2001–08. However, it declined more rapidly than in other countries between 2008 and 2012. While FDI has been staggering, remittances have become a crucial source of foreign exchange inflow for Pakistan’s economy in the past decade. In 2015, the volume of out-migrating Pakistanis seeking overseas employment rose to a high of 950,000 compared with 200,000 in 2006 (Figure 17). In 2016–17, Pakistan received approximately US$20 billion in remittances, or close to 6.5 percent of GDP, placing it sixth on the list of top remittance-receiving countries (Figure 18). 14 See Deng, Illangovan, and Blanco Armas (2019) for more on regional connectivity under the Pakistan@100 research initiative. 15 An enabling environment includes macroeconomic conditions, institutions, infrastructure, and ICT adoption; market conditions include labor and product markets as well as financial systems and market size; human capital conditions include health and skills; and an innovation ecosystem considers business dynamism and innovation capability. 15 Figure 13 Figure 14 Export structure by sector in Pakistan, 2000 and 2017 Export structure by sector in Vietnam, 2000 and 2017 PAKISTAN EXPORTS, 2000 VIETNAM EXPORTS, 2000 TOTAL: $8.1B TOTAL: $16.1B NON-RETAIL PURE HEAVY PURE NON- KNOTTED KNIT LEATHER RICE CRUDE PETROLEUM LEATHER RUBBER CRUSTA- ELECTRICAL COTTON YARN WOVEN COTTON KNIT CARPETS MEN'S APPAREL FOOTWEAR FOOT- CEANS CONTROL 4.1% 21% 14% 4.5% MEN'S 2.7% SHIRTS 5.2% WEAR BOARDS SHIRTS 6.2% 4.8% 2.7% 3.9% 1.2% 3.4% KNIT SWEATERS 4.5% TEXTILE FOOTWEAR 4.3% OTHER LIGHT PURE SPORTS FURNITURE HOUSE LINENS WOVEN COTTON EQUIPMENT RICE 1.7% 2.1% NON-KNIT 10% 4.1% 4.2% MEN'S COATS 2.2% LIGHT SYNTHETIC COTTON FABRICS NON-KNIT COFFEE 3.9% MEN'S SUITS 3.6% 1.8% PAKISTAN EXPORTS, 2017 VIETNAM EXPORTS, 2017 TOTAL: $24.8B TOTAL: $220B HOUSE LINENS NON- NON- HEAVY KNIT RICE LEATHER BROADCASTING BROAD- OFFICE KNIT OTHER COFFEE 13% RETAIL KNIT PURE SWEAT- 6.6% APPAREL EQUIPMENT CASTING MA- SWEAT- FURNITURE 1.4% PURE WOMEN'S WOVEN ERS 1.8% ACCES- CHINE ERS 14% 2.5% COTTON SUITS COTTON 2.8% SORIES PARTS 1.6% YARN 4.2% 4.0% 2.3% 2.3% 5.1% LIGHT PURE WOVEN ALCOHOL COTTON >80% ABV TELEPHONES 2.8% 1.6% 6.8% NON-KNIT TEXTILE MEN'S SUITS FOOTWEAR 7.7% 4.3% INTEGRATED CIRCUITS 6.6% LEATHER FOOTWEAR 2.8% Source: MIT Observatory of Economic Complexity: http://atlas.media.mit.edu/en/. Figure 15 Figure 16 Gross fixed investment as share of GDP, 2000–17 Foreign direct investment as share of GDP, 2000–17 GROSS FIXED INVESTMENT AS SHARE OF GDP FOREIGN DIRECT INVESTMENT (NET INFLOWS) AS SHARE OF GDP 40 4 35 3 SHARE OF GDP (%) SHARE OF GDP (%) 30 25 2 20 1 15 10 0 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 PAKISTAN INDIA SOUTH ASIA PAKISTAN INDIA SOUTH ASIA BANGLADESH LOWER MIDDLE INCOME BANGLADESH LOWER MIDDLE INCOME Source: World Development Indicators. 16 Figure 17 Figure 18 Total number of outmigrants from Pakistan, 2006¬–17 Remittances as share of GDP, 2000–17 TOTAL NUMBER OF OUTMIGRANTS FROM PAKISTAN REMITTANCES AS SHARE OF GDP 1,000,000 12 900,000 10 800,000 SHARE OF GDP (%) 8 700,000 600,000 6 500,000 4 400,000 2 300,000 200,000 0 00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 100,000 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 0 PAKISTAN INDIA SOUTH ASIA 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 BANGLADESH LOWER MIDDLE INCOME Source: Bureau of Emigration and Overseas Employment (migrant volume) and WDI (remittances). A series of simulations suggest that in order for Pakistan to achieve upper-middle-income country status by its centenary in 2047, significant increases in investment and productivity as well as decreases in population growth are required (World Bank 2019a). Additional simulations also confirm that increases in the utilization of human resources by boosting female labor force participation, continued structural transformation, and productivity increases through human capital investment are necessary. 17 2. KEY LABOR MARKET OUTCOMES 2.1. LABOR MARKET TRENDS OVER TIME Both labor force participation (LFP) and employment ratios have experienced only modest changes over time on average (Figure 19 and Figure 20). The overall LFP rate and employment ratio were about 50 percent in 2017, a similar level as the previous two decades. The gap between the LFP rates of rural and urban men, albeit relatively small, had declined a little but remained visible. While female labor force participation (FLFP) in rural areas had been gradually improving before 2017, it then reverted to a level seen a decade back. This is mostly driven by a huge drop in FLFP in rural Balochistan as well as in Sindh, although the decreases in rural FLFP are observed in all provinces.16 The urban FLFP rate has been strikingly low and stagnant—it has remained stubbornly at around 10 percent since 2005. Also, urban FLFP did not decrease in 2017, unlike the rural rate. With these trends, the significant gender gap in labor supply has not been reduced. While masked in the average, each sector’s ability to create jobs varies across time and also across the urban-rural divide. Net increases in employment (Figure 21) show that the service sector added the largest number of jobs (about 4.0 million) during the period between 2000 and 2005, with agriculture and industry adding about 0.8 million and 3.2 million net jobs, respectively. However, the pattern significantly changed during the economic slowdown of 2006–11. Net increases in employment in both industry and services declined significantly, while that in agriculture increased, indicating that agriculture likely provided labor market buffers during the economic downturn. During the recovery period of 2012–17, the industry sector turned around, adding 3.4 million jobs. However, the service sector did not rebound during this period, and the number of jobs fell by 760,000. This is reflected in changes in the sectoral share of employment (Figure 22). During the Figure 19 Figure 20 Labor force participation rates of men and women in rural and urban areas, Employment ratio of men and women in rural and urban areas, 2001–17 2001–17 LFP OVER TIME EMPLOYMENT RATIO OVER TIME 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0 0 2001 2005 2007 2008 2010 2012 2013 2014 2017 2001 2005 2007 2008 2010 2012 2013 2014 2017 RURAL MEN URBAN MEN OVERALL RURAL MEN URBAN MEN OVERALL RURAL WOMEN URBAN WOMEN RURAL WOMEN URBAN WOMEN Source: Bureau of Emigration and Overseas Employment (migrant volume) and WDI (remittances). 16 FLFP in rural Balochistan declined from 21 percent to 8 percent between 2014 and 2017. This may have been due to increased conflicts and violence in Balochistan in 2016 despite the province’s expectation of an economic boon coming from the China-Pakistan Economic Corridor (Ricks 2017). 19 Figure 21 Figure 22 Net increase in employment by sector across three time periods, Changes in sectoral share of employment across three time periods, 2000–17 2000–17 NET INCREASE IN EMPLOYMENT BY SECTOR NET CHANGE IN EMPLOYMENT SHARE BY SECTOR 9,000 8.0% 8,000 6.0% 7,000 4.0% 6,000 5,000 2.0% 4,000 0 3,000 -2.0% 2,000 -4.0% 1,000 0 -6.0% -1,000 -8.0% 2000–2005 2006–2011 2012–2017 2000–2005 2006–2011 2012–2017 (BOOM) (BUST) (RECOVERY) (BOOM) (BUST) (RECOVERY) AGRICULTURE INDUSTRY SERVICES AGRICULTURE INDUSTRY SERVICES Source: Jobs Structure Tool based on LFS. boom, the industry and service sectors created more jobs than did agriculture, progressing toward a structural transformation. During the subsequent periods, the employment share of services declined. With the varying pace of structural transformation, a major shift of the workforce from agriculture to industry and services has not taken place over time. The share of workers in agriculture have remained at the same low level in urban areas (Figure 23) and decreased only slightly in rural areas (Figure 24). While industry’s share of employment has increased, that of the service sector declined, in both urban and rural areas. The expansion of the industry sector is encouraging; however, it seems to have come at the expense of services rather than agriculture. This is of concern given that the economy relies on the service sector for a large share of value added. An overall transition to wage employment has been on the rise, but mostly for men.17 The increase in overall wage employment has been most pronounced among rural males, but this is mainly due to increases in nonregular wage employment that includes day labor and piecemeal contract work (Figure 25). The share of Figure 23 Figure 24 Three key sectors’ employment share in urban areas, 2001, 2008, 2014, Three key sectors’ employment share in rural areas, 2001, 2008, 2014, and 2017 and 2017 EMPLOYMENT SECTOR OVER TIME: URBAN AREAS EMPLOYMENT SECTOR OVER TIME: RURAL AREAS 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0 0 2001 2008 2014 2017 2001 2008 2014 2017 URBAN RURAL AGRICULTURE INDUSTRY SERVICES AGRICULTURE INDUSTRY SERVICES Source: LFS various years. 17 Recall that wage employment is classified as regular or nonregular. Regular wage employment is based on an extended employment tenure in a certain occupation under a continuous payment scheme, whereas nonregular employment is based on piecemeal and task- based tenure and payment. 20 Figure 25 Figure 26 Men’s employment status in urban vs. rural areas, 2001, 2008, 2014, and Women’s employment status in urban vs. rural areas, 2001, 2008, 2014, 2017 and 2017 EMPLOYMENT STATUS OVER TIME (MALES): BY REGION EMPLOYMENT STATUS OVER TIME (FEMALES): BY REGION 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0 0 2001 2008 2014 2017 2001 2008 2014 2017 2001 2008 2014 2017 2001 2008 2014 2017 MALES, RURAL MALES, URBAN FEMALE, RURAL FEMALE, URBAN REGULAR WAGE EMPLOYEE SELF-EMPLOYED, AGRICULTURE REGULAR WAGE EMPLOYEE SELF-EMPLOYED, AGRICULTURE NON-REGULAR WAGE EMPLOYEE SELF-EMPLOYED, NON-AGRICULTURE NON-REGULAR WAGE EMPLOYEE SELF-EMPLOYED, NON-AGRICULTURE UNPAID FAMILY WORKER UNPAID FAMILY WORKER Source: Jobs Structure Tool based on LFS various years. agricultural self-employment and nonpaid family work has declined significantly over time among rural males, indicating progress in rural areas. Urban males also experienced a similar pattern as overall wage employment (mainly nonregular) increased, but to a lesser extent than their rural counterparts. For women, the trends are too volatile to make a conclusive observation (Figure 26). However, it is clear that women’s share of wage employment gradually increased and the share of nonpaid family work declined in both rural and urban areas, at least during the period 2008–17. 2.2. LABOR MARKET SNAPSHOT IN 2017 The current status of the labor market in Pakistan shows that about 208 million people rely on only about 46 million earners (Figure 27). This is in part due to the population structure: 45 percent (or about 92 million) are either over 65 or under 15 years old. Apart from this, the heavy underutilization of human resources in a country where 47 percent of the working-age population remains out of the labor force explains the dependency on a small share of income-generating workers. This is largely due to a very low FLFP—less than one quarter of working-age women participated in the labor force in 2017. Moreover, a large majority of inactive working-age individuals (40 million out of 53 million) were not in school, which translates into a high share of people not in employment, education, or training (NEET).18 Besides, not everyone who is working receives remuneration for their work—46 million out of 60 million workers were engaged in income-generating activities, whereas 14 million workers did not earn incomes. In general, the overall LFP rate (53 percent) is low and characterized by a huge gender gap and variations across provinces (Figure 28 and Figure 29). The likelihood of participating in the labor force is higher in rural areas, probably due to the availability of more opportunities in agriculture and unpaid work in rural areas. The urban-rural gap in LFP is more prominent for women than men. The FLFP rate in 2017 was barely over 10 percent in urban areas, whereas it was close to 30 percent in rural areas. The relatively high FLFP in rural areas was largely driven by Punjab and by agriculture. To the contrary, the urban-rural divide in LFP was not so striking among males, while overall LFP among men was over 80 percent except in Khyber Pakhtunkhwa (KPK) province. This suggests that there is little room for men to further increase LFP, whereas there is a huge potential for women to do so. 18 In discussing FLFP rates reported in LFS, it should be noted that some studies, such as those of Saqib and Arif (2012) and Arshad (2008), have highlighted that economic activities performed by women, including paid work conducted at home, as well as their unpaid work, tend to be underreported. This may be due to individual and social perceptions of what constitute work and jobs. In particular, some activities typically performed by females may not be considered as work by survey respondents, despite the reference to family help as work in the LFS questionnaire. 21 Figure 27 A snapshot of Pakistan’s labor market in 2017 TOTAL POPULATION: 208 MILLION WORKING-AGE POPULATION: 116 MILLION NON-WORKING AGE POPULATION: 92 MILLION IN SCHOOL: LABOR FORCE: 59 MILLION OUT OF SCHOOL: 44 MILLION 13 MILL. EMPLOYED (FEMALE): 12 MILLION EMPLOYED (MALE): 45 MILLION UNEMPLOYED: 1 MILLION PAID (MALE): 40 MILLION PAID (FEMALE): 6 MILLION 0 50 100 150 200 MILLIONS Source: LFS 2017, adjusted for the population recorded in 2017 census. Figure 28 Figure 29 Rural and urban women’s labor force participation rates by province, 2017 Rural and urban men’s labor force participation rates by province, 2017 FEMALE LFP: BY PROVINCE MALE LFP: BY PROVINCE 40% 100% 35% 90% 80% 30% 70% 25% 60% 20% 50% 15% 40% 30% 10% 20% 5% 10% 0% 0% BALOCHISTAN KHYBER PUNJAB SINDH BALOCHISTAN KHYBER PUNJAB SINDH PAKHTUNKHWA PAKHTUNKHWA FEMALE RURAL FEMALE URBAN MALE RURAL MALE URBAN Source: LFS 2017. Unemployment rates are measured based on the number of individuals who are not employed even if they are willing and available for work, and are actively searching for jobs. Some individuals may be unemployed because the available jobs are not desired due to low earnings, poor working conditions, long commutes, or other reasons. Others may be unemployed because no job opportunities are available. When unemployment rates are disaggregated by gender and education (Figure 30), it is clear that unemployment is a more serious issue for relatively well-educated individuals, especially for females. It is also more of an issue for youth than older adults (nonyouth) (Figure 31). Some youth, especially educated ones, may let go of currently available job opportunities in the anticipation of finding better ones in the future. Also, the fact that youth from higher-income quintiles who can afford to not earn incomes are more likely to be unemployed appears to corroborate the argument that Pakistan’s unemployment on average is rather by choice and due to mismatches between available jobs and job seekers’ expectations. High unemployment rates, especially among educated females in urban areas, highlight the labor market challenges faced by women. The FLFP rate is very low, particularly in urban areas. That relatively well-educated women are often unemployed and typically leave the labor market after entry signifies the underutilization of human 22 Figure 30 Figure 31 Unemployment rates by gender, education, and urban/rural location, 2017 Unemployment rates by age group and income quintile, 2015 14% 16% 15 12% 14% 10% 12% 8% 10% 8% 8 8 6% 7 6% 5 4% 4% 2% 2 2 2 2 2 2% 0 MALE, URBAN MALE, RURAL FEMALE, URBAN FEMALE, RURAL 0 1ST 2ND 3RD 4TH 5TH NO EDUCATION LOWER SECONDARY POST-SECONDARY QUINTILE QUINTILE QUINTILE QUINTILE QUINTILE PRIMARY UPPER SECONDARY YOUTH NON-YOUTH Source: LFS 2017. Source: from Bossavie, Khadka, and Strokova (2018) based on HIES 2015. capital in the country. Among unemployed youth, almost half (49 percent) expressed a desire to be employed full time in the government sector, indicating that unemployment is in part explained by a gap between actual labor market conditions and job seekers’ expectations (only less than 5 percent of jobs are in the public sector). The share of NEET further underscores the issue of low FLFP rates (Figure 32). For both urban and rural men, the NEET share is already quite low from age 15 but further diminishes as the individual ages. To the contrary, there is a quite different pattern among urban and rural women as they age. At age 15, there are more rural girls in NEET than urban girls, indicating higher and earlier school dropout rates among rural girls. Urban women aged 15–29 are less likely to transition to the labor market and hence have a higher share in NEET as compared to rural women in the same age group, who are more likely to engage in unpaid work, mostly agriculture.19 A comparison with peer countries also shows that the share of women in the NEET group is higher in Pakistan than in peer countries, on average (Figure 33). Disaggregated by gender, however, it is observed that the NEET share in Pakistan is lower than in other countries for males, but it is far higher for females, raising the overall average. Figure 32 Figure 33 NEET share by age, gender, and urban vs. rural location, 2017 Cross-country comparison: NEET share among youth (ages 15–24), various years SHARE OF INDIVIDUALS IN NEET NEET SHARE: CROSS-COUNTRY COMPARISON 80% 50 70% 45 40 60% 35 50% PERCENT (%) 30 40% 25 30% 20 20% 15 10 10% 5 0 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 0 PAKISTAN BANGLADESH INDONESIA TURKEY VIETNAM (2017) (2016) (2015) (2015) (2013) MALES, URBAN FEMALES, URBAN MALES, RURAL FEMALES, RURAL MALE FEMALE TOTAL Source: LFS 2017 (Pakistan); LFS various years (other countries). Note: NEET = not in employment, education, or training. 19 Likewise, the LFP rates are similar for urban and rural women when unpaid work is excluded. 23 Figure 34 Figure 35 Beliefs regarding women’s work outside the home by urban vs. rural location Labor force participation rates by income quintile and gender AGREEMENT WITH GENDER NORMS REGARDING 100% 97 96 96 95 96 WOMEN’S WORK OUTSIDE THE HOME 90% WOMEN URBAN 80% MEN 70% 60 60% 56 WOMEN 51 51 50 RURAL 50% MEN 40% 30% WOMEN 24 ALL 20% 16 MEN 9 10 10% 7 0 20% 40% 60% 80 100% 0 1ST 2ND 3RD 4TH 5TH MARRIED WOMEN SHOULD BE ALLOWED TO WORK OUTSIDE THE HOME QUINTILE QUINTILE QUINTILE QUINTILE QUINTILE IF WIFE WORKS, HUSBAND SHOULD HELP WITH HOUSEWORK ALL MALES FEMALES Source: LSS 2015. Note: Household wealth is measured by a wealth index computed using dwelling characteristics and household durables using a principal component analysis. Gender norms constitute a great impediment to female labor market activities. A large share of people— 71 percent of women and 51 percent of men—agree that women should be allowed to work outside the home (Figure 34). Despite a significant gender difference in such perceptions, attitudes toward married women’s labor market activities are generally favorable. The positive attitudes are particularly prevalent in urban areas, but urban FLFP remains significantly lower than rural FLFP. Bursztyn, Gonzalez, and Yanagizawa-Drott (2018) show that in Saudi Arabia, correcting the beliefs of husbands about other men’s attitudes toward their wives’ work by informing them that there is widespread rejection of traditional social norms around women’s work, significantly increased FLFP rates among the wives of informed husbands. Focus group discussions in Pakistan corroborate that beliefs regarding the perceptions of other men often explains men’s opposition to women’s work for pay.20 When the LFP rates are disaggregated by income quintiles, it is revealed that females are more likely to work when in poverty (Figure 35). The FLFP rates in the 1st and 2nd quintiles of household wealth—at 24 and 16 percent, respectively—are significantly higher than their non-poor counterparts. This suggests that women tend to work out of necessity, which can signal a household’s financial deficiency. The association of women’s work and the households’ financial needs can induce significant underreporting of women’s work. Associated with gender norms, women face significant mobility constraints, which are a great impediment to FLFP (Figure 36). The share of respondents who believe that women can visit a local market or health facility alone is only about 30 percent. Such restrictions may be associated with safety concerns regarding traveling alone: about half of respondents reported that they did not feel safe walking alone even during the day and would never walk alone (Figure 37). There have been some studies that have attempted to identify the impact of addressing mobility constraints on female labor market outcomes (Box 2). A qualitative survey among educated, urban women in four major cities21 in Pakistan also provides additional insights regarding barriers to women’s labor market activities (Amir and Pande 2019). Many women prefer home-based work over office-based employment mainly for two reasons: (i) work and life balance and (ii) workplace harassment. Due to rampant workplace, gender-based harassment, gender-segregated offices are also preferred in an office setting. Family disapproval of women’s work poses another challenge. For the minority of women who do enter the labor force, their choice of occupation is severely limited to a socially accepted, gender-appropriate set of jobs (e.g., teachers, health-care providers). 20 Box 2 from Amir and Pande (2019). 21 Focus group discussions took place in the capital city of each province: Lahore (Punjab), Karachi (Sindh), Quetta (Balochistan), and Peshawar (KPK). 24 Figure 36 Figure 37 Women’s mobility: General perceptions of what is appropriate Women’s own perceptions of their personal safety while walking in public CAN WOMEN GO TO [PLACE X] ALONE? DO YOU FEEL SAFE WALKING ALONE OUTSIDE YOUR COMMUNITY/NEIGHBORHOOD? NEARBY MASJID 26 37 37 OR SHRINE NO, I NEVER WALK ALONE HOMES OF FRIENDS 52 35 13 IN NEIGHBORHOOD YES, DURING THE DAY ONLY LOCAL HEALTH FACILITY 31 56 13 OR DOCTOR YES, ANYTIME LOCAL MARKET 31 49 20 0 20% 40% 60% 80% 100% 0 10% 20% 30% 40% 50% 60% ALONE NOT ALONE NEVER Source: Amir et al. (2018) based on LSS 2015. BOX 2: TRANSPORT AND WOMEN’S LABOR MARKET ACTIVITIES Access to transport is expected to have a significant impact on women’s probability of employment. Ejaz (2011) finds that women in Pakistan who own a private vehicle are more likely to participate in the labor force, all else considered. This positive association is not limited to private car ownership but extends to access to transport in general. For example, in San Francisco and Boston (United States), low-skilled workers without private vehicles were more likely to be employed if they had access to public transport (Kawabata 2003). A recent study in Lima, Peru, explores this relationship, particularly for female workers living in urban areas. The analysis was conducted in the metropolitan area of Lima, where access to transportation remains a challenge. The city opened a bus rapid transit system in 2010 and an elevated rail system (called Line 1) in 2011. These systems connected peripheral areas of the city to major employment centers and offered much safer options than other public transit modes. Martinez et al. (2018) find that women were more likely to use these systems as compared to men, which in turn improved their access to jobs. They also find higher employment gains for women as a result of these two public transit interventions. According to the 2018 Economic and Social Wellbeing Survey, conducted by the Punjab Commission on the Status of Women, less than half (48 percent) of the total surveyed women in Pakistan’s Punjab province have access to public transport, and the share is much lower in rural areas (38 percent) than urban areas (64 percent) (Sajjad et al. 2017). Such constraints are expected to have a direct impact on women’s mobility, which has implications for their labor force participation and educational attainment. In a recent study of 243 villages in Pakistan, it was found that women’s participation in training programs was substantially increased by addressing their mobility constraints (Cheema et al. 2019). A research project is underway to examine the role of a new metro bus system in Lahore. However, lack of access to transportation is not the only factor limiting women’s mobility. Other than social and cultural norms, safety considerations are an important determinant. Women often get harassed on public transport. For example, a small study in Karachi found that 85 percent of women who travel regularly for work had been harassed on public transport (ADB 2014). This could be a reason why women are more likely to participate in training sessions when group transportation as opposed to individual transportation is arranged (Cheema et al. 2019). In Pakistan, access to safe public transportation for women is almost nonexistent, with only small-scale initiatives such as the women-only Pink Rickshaw or Pink Bus routes in Lahore in place. Providing access to safe public transport can alleviate some of the mobility constraints that women face especially when it is unacceptable for them to ride bicycles or motorbikes to commute. This in turn is likely to have a positive impact on their employment outcomes. Source: Based on Sajjad et al. 2017; Cheema et al. 2019; and Martinez et al. 2018. 25 Demand-side constraints are also likely to affect women’s decisions regarding labor market participation. Some survey respondents suggested that employers expressed reservations about hiring women. Frequently cited reasons include that there were not enough female applicants and that women’s labor market attachment is weak, since they leave the workforce once they are married (Malkani 2017). This is in line with reports that a large share of employers cite family commitments and the limited availability of women with expected skills as reasons for not hiring women (Amir et al. 2018). In addition to this are potential implications for workplace dynamics, additional expenses for providing separate workplace facilities, and regulations such as maternity leave. Bias against women’s skills and competencies poses additional challenges to women’s work. 2.3. ACCESS TO QUALITY JOBS While job quality is a multifaceted concept that is not easily measurable, it is typically assessed based on earning level as well as the sector, status, and formality. There is a strong association between per capita consumption expenditure and the household head’s employment type (Figure 38). Employers have the highest level of consumption expenditure, followed by paid employees and the self-employed. The service and manufacturing sectors have positive association with higher consumption expenditure. Also, the gap between urban and rural areas is quite large. When we look into employment type by welfare quintiles, the top quintile is dominated by formal wage employees22 and the nonagricultural self-employed. On the contrary, nonpaid and agricultural workers compose a large share of the bottom quintile (Figure 39). The distribution of workers across welfare quintiles, and the sector and status of employment suggest a large variation in individual workers’ ability to access quality jobs. Almost one quarter (24 percent) of workers can be classified as the working poor—those who work but live in households that are below the poverty line. This is indicative of the low quality of existing jobs, which do not provide the level of earning required to lift workers out of poverty. More than half of the working poor are engaged in agriculture, whereas only one third of the working non-poor are in agriculture. Female workers are more likely to be in this category (21.8 percent for men, 29.1 percent for women) corroborating both poor job quality for women and labor force participation among women from poor households. Figure 38 Figure 39 Monthly per capita expenditure by household head’s employment type Employment type by welfare quintiles 14,000 12,000 PER CAPITA CONSUMPTION EXPENDITURE BOTTOM QUINTILE 10,000 PAKISTAN RUPEES 8,000 6,000 TOP QUINTILE 4,000 2,000 ALL 0 QUINTILES EMPLOYER SELF-EMPLOYED FAMILY WORKER PAID EMPLOYEE AGRICULTURE MANUFACTURING CONSTRUCTION SERVICES URBAN RURAL MEAN—OVERALL 0 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% WAGE, FORMAL UNPAID SELF, NON-AGRICULTURE EMPLOYMENT STATUS INDUSTRY LOCATION WAGE, INFORMAL SELF, AGRICULTURE Source: HIES 2015. 22 A formal wage employee is defined as a wage employee with a written contract. 26 Figure 40 Distribution of employment sector by gender, 2017 A. MALES B. FEMALES 3.8% 0.2% 0.2% 1.4% GOVERNMENT SERVICES TRANSPORT AND COMMUNICATIONS FINANCE, INSURANCE, AND REAL ESTATE FINANCE, INSURANCE, 0.4% 2.0% AND REAL ESTATE 1.3% GOVERNMENT SERVICES WHOLESALE AND RELATED TRADE OTHER PRIVATE SERVICES 0.1% 9.0% 0.3% OTHER PRIVATE SERVICES TRANSPORT AND OTHER INDUSTRY COMMUNICATIONS 19.3% MANUFACTURING 31.0% AGRICULTURE 23.3% WHOLESALE AND RELATED TRADE 17.9% 77.5% 12.3% MANUFACTURING AGRICULTURE OTHER INDUSTRY Source: LFS 2017. With respect to the employment sector, a striking gender difference is noted (Figure 40). The employment sectors of male workers are distributed widely across agriculture (31 percent), industry (18 percent manufacturing and 12 percent other industry), and services (23 percent wholesale and related trades and 16 percent in other services). To the contrary, over two-thirds of female workers are in agriculture and 19 percent in manufacturing, with the remaining few working in other services. This distribution suggests that women in Pakistan are particularly deprived of opportunities in services.23 Thus, there is much scope for public intervention to expand the opportunities for female-friendly occupations, such as teachers, health service providers, care providers, and retailers, given their huge underrepresentation. The distribution of employment status across gender also shows a similar picture (Figure 41). The gender difference is relatively muted in urban areas where a similar share of male and female workers engages Figure 41 Distribution of employment status by gender and urban vs. rural location, 2017 A. EMPLOYMENT STATUS: URBAN B. EMPLOYMENT STATUS: RURAL 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0 0 MALE FEMALE MALE FEMALE WAGE, FORMAL WAGE, INFORMAL UNPAID SELF-EMPLOYED, AGRICULTURE SELF-EMPLOYED, NON-AGRICULTURE Source: LFS 2017. 23 In Bangladesh in 2016, the share of women in agriculture, manufacturing, and other services was 36 percent, 28 percent, and 32 percent, respectively; in Turkey in 2015, the share of women in agriculture, manufacturing, and other services was 33 percent, 15 percent, and 41 percent, respectively; and in Indonesia in 2015, the share of women in agriculture, manufacturing, and other services was 31 percent, 15 percent, and 30 percent, respectively. See Bossavie, Khadka, and Strokova (2018) for further discussion. 27 Figure 42 Figure 43 Prevalence of formality by employment status and urban vs. rural location, 2017 Earning by level of formality among wage employees, 2017 PREVALENCE OF WRITTEN CONTRACTS AMONG REGULAR AND NON-REGULAR WAGE EMPLOYEES 35,000 NON-REGULAR WAGE 30,000 RURAL AVERAGE MONTHLY EARNING, REGULAR WAGE 25,000 PAKISTAN RUPEES 20,000 NON-REGULAR WAGE URBAN 15,000 REGULAR WAGE 10,000 5,000 NON-REGULAR WAGE OVERALL 0 REGULAR WAGE WAGE EMPLOYMENT WAGE EMPLOYMENT (REGULAR) (NON-REGULAR) 0 20% 40% 60% 80% 100% WITH A WRITTEN CONTRACT MINIMUM WAGE (2017) WITHOUT A CONTRACT WITH A CONTRACT WITHOUT A WRITTEN CONTRACT Source: LFS 2017. in formal and informal wage employment. In rural areas, however, there is a substantial difference between male and female workers in employment type and quality of jobs: most females work in unpaid family work with a negligible share in formal wage employment, whereas male workers tend to work in informal wage employment and agricultural self-employment. Nonagricultural self-employment, typically nonfarm business activities, is also less common for women than men, which underscores the pertinent challenge of the quality of jobs among women. The micro-determinants of the status and sector of employment suggest uneven access to quality jobs by gender and region, as well as across age groups. The results of multinomial logit regressions on the determinants of working in wage employment (compared with agricultural self-employment) and agriculture (compared with services) are as follows (see Appendix B for a detailed discussion). Education has a positive (negative) effect on the probability of being a wage employee (agricultural worker). The association between education level and employment type (employment status and sector) is generally stronger for women than for men. This suggests that labor market segmentation by educational attainment is stronger for women. The results also show that living in an urban area is positively associated with wage employment and negatively with agricultural self-employment. Even within regular wage employment, there is great heterogeneity in job quality depending on the formality of contractual arrangements. The concept and measurement of formality is not always straightforward (see Box 3 for further discussion). The formality of employment is defined in this Diagnostic based on whether the worker has a written contract or not. Among all wage employees, only slightly over one quarter have a written contract (Figure 42). The prevalence of formality is higher in urban than rural areas. There is a large difference between the earnings of formal employees (with a written contract) and those of informal employees (without a written contract) (Figure 43). Formal employees earn 2.2 times more than informal employees. The difference by formality is a lot larger among regular wage employees than nonregular employees. There are large variations in the share of formal wage workers with a written contract by employment sector (Figure 44). Almost everyone in government services in the public sector is a formal employee. To the contrary, almost no agricultural workers have established employment relations based on a written contract. In high-value services, such as finance, insurance, and real estate, about 72 percent of the employees have written contracts. A comparison with Bangladesh shows that jobs in manufacturing and wholesale and trade (equivalent to commerce) in Pakistan are far less formal than in Bangladesh (Figure 45). The large difference in 28 Figure 44 Figure 45 Share of formal wage workers by sector in Pakistan, 2017 Share of formal wage workers by sector in Bangladesh, 2016 100% 97 100% 90% 90% 83 80% 80% 72 71 70% 70% 60% 60% 55 50% 50% 46 40% 40% 30 30% 26 30% 24 22 20% 18 20% 16 11 8 10% 8 10% 5 4 0 0 . TA E, IEN Y A IC S- IC LY- RE NG RY TE D ES ES M AN ININ URE NG N CE CO IEN AND ON D S RS ES NC LA N TIE OR NIT ERV NES TIO I N TE M IC IC TE ND ES ES U RV I D ST HE AC ) TE B S SE AM ER RE LE A RI AT A G RI LT CO M ULT ILI AL RA RV RV DU TU UC TU M OT S I IC RT CU D US UT D F SE SE RE SU SA M AC ND IC UN PO IN D TR RI CO AN IC (A GR D , IN . E LE VT UF NS ER FIN MN ANS AG UF AT BL HO A O TH AN CE AN RT CO IV PU G CO TR W PR O OR AL O AN M M SP U CI ER FIN M AN AN TH M TR O Source: LFS 2017. Source: LFS 2016. the formality observed in manufacturing is likely because Bangladesh’s manufacturing is led by large firms in the export-oriented ready-made garment sector, with the foreign direct investment of global brands and retailers. Large, global manufacturing firms are less prevalent in Pakistan. Typically, the working hours are shorter in rural than urban areas, and the urban-rural gap in working hours is similar for both genders. About 56 percent of female workers report having worked less than 35 hours a week, whereas the equivalent figure for male workers is about 9 percent (Figure 47). The share of female workers with less than 35 hours a week varies widely across provinces with the highest shares observed in Khyber Pakhtunkhwa and Punjab. To the contrary, the share of male workers with less than 35 hours a week is low (around 7 percent) and more consistent across provinces with the exception of Khyber Pakhtunkhwa (with a higher share at 17 percent). Figure 46 Figure 47 Average weekly hours worked by gender and urban vs. rural Share of those working less than 35 hours per week by gender and location, 2017 province, 2017 60 70% 65 53 61 50 60% 57 50 48 SHARE OF WORKERS THAT WORKED NUMBER OF HOURS WORKED 50% LESS THAN 35 HOURS 40 38 IN THE PAST WEEK 42 33 34 40% 35 30 30% 20 20% 19 9 9 10 10% 8 10 0 0 PUNJAB SINDH BALOCHISTAN KHYBER ALL URBAN RURAL ALL PAKHTUNKHWA MALE FEMALE MALE FEMALE Source: LFS 2017. 29 BOX 3: MEASUREMENT AND PREVALENCE OF INFORMALITY IN PAKISTAN Informal employment refers to the hiring of labor that is not subject to labor regulations, including those regarding statutory minimum wages, mandated social contributions and benefits, or the termination of contracts. Informal employment is a large component of labor markets in many developing countries and even of some member countries of the Organisation for Economic Co-operation and Development. Although the concept of informal employment is clear, measuring it in practice is not straightforward. In that regard, the International Labor Organization (ILO 2013) makes a distinction between the informal sector, which refers to the formality of production units, and informal jobs, which refer to workers’ type of employment. While the formal sector, made up of firms that are registered with authorities, mostly consists of formal employment, not all workers employed in the formal sector are necessarily employed in formal jobs. (In)formal sector in Pakistan: With limited data of production units, such as enterprise surveys or economic census, it is difficult to measure the size or prevalence of the informal sector. In a sole proprietorship (single-member company), the business is considered legitimate and formal as long as the owner has a National Tax Number under the Federal Board of Revenue and pays the income taxes due for individuals. A limited liability company (either private or public) has two or more shareholders (owners) of the firm. These firms are required to register with the Securities and Exchange Commission of Pakistan and are governed by the Companies Act 2017. They are by definition formal firms and are required to fulfill their legal duties including paying taxes and benefits for workers.24 (In)formal jobs in Pakistan: Throughout the report, a formal wage employee is defined as a wage worker who holds a written contract in their current employment, as in many other jobs studies.25 A rationale for the use of that proxy is that a written contract constitutes the legal basis for the enforcement and compliance with labor laws that characterize formal employment. Among workers who are not wage employees, unpaid contributing family workers are systematically classified as informal as per the ILO guidelines, as their employment is not subject to labor legislation, social security regulations, collective agreements, etc. For own-account workers and employers, the criteria is the formality status of the production unit they own, which is classified as formal if the entity is registered. How prevalent are informal jobs in Pakistan? Based on LFS 2017, the share of formal workers (i.e., wage workers with a written contract) out of the total number of wage workers is about 26.6 percent. If all nonwage workers are assumed to have informal jobs, the share of formal jobs out of all workers is about 11.6 percent. Source: Based on LFS 2017. 24 Detailed business process for companies, see World Bank (2018)—Doing Business 2019. 25 See Farole and Cho (2017); Christiaensen and Premand (2017); and Cunningham and Munoz (2018), among others. 30 3. PAKISTAN’S WORKFORCE 3.1. EDUCATION AND SKILLS Educational attainment has been slowly improving in Pakistan. The expansion of secondary schooling and reduction in the share of population with no educational attainment over time is noticeable (Figure 48). However, in 2017 over 30 percent of working-age men and over 50 percent of working-age women still remain without any formal education, and the share of those with a postsecondary education is less than 10 percent for both genders. Also, despite significant progress, mean years of schooling for women of the 1995 cohort in 2017 is only about 6.4, (which is just about primary school completion), which is lower than that of male counterparts by 1.2 years (Figure 49). With educational progress, literacy rates are on the rise among younger generations. Among younger cohorts, the basic ability to read and write, which has been shown to provide important returns in the labor market (de Baldini Rocha and Ponczek 2011; Hanushek and Woessmann 2008), has improved significantly. Close to 80 percent of children aged 10–14 are able to read, compared with 40 percent of individuals aged 55–64. However, some groups are considerably left behind, particularly the working-age population in rural areas and specifically rural women, for whom the literacy rate is only 35 percent. Despite overall progress, levels of schooling and literacy among the workforce remain low. Forty‑three percent of the working-age population in Pakistan has no formal education (Figure 50). There are large disparities in levels of schooling between rural and urban areas. More than half of the workers in rural areas have no formal education, compared with one quarter in urban areas. When benchmarked against comparator countries, Pakistan ranks poorly in terms of literacy (Figure 51). The low level of literacy among the workforce remains a key bottleneck; literacy-related skills are a prerequisite for productive employment to emerge in the first place, and furthermore because they usually cannot be acquired on the job. Without numeracy and literacy skills, the prospects of improving employment opportunities and earnings, whether in agriculture or urban settings, are poor (World Development Report 2013: Jobs [World Bank 2012a]). Figure 48 Figure 49 Distribution of educational attainment of working age population Mean years of schooling by birth cohort and gender by gender, 2001–17 100% 8 90% 7 80% MEAN YEARS OF EDUCATION EDUCATION TRENDS 70% 6 60% 5 50% 40% 4 30% 3 20% 2 10% 0 1 2001 2005 2008 2012 2014 2017 2001 2005 2008 2012 2014 2017 MALE FEMALE 0 1960 1965 1970 1975 1980 1985 1990 1995 NO EDUCATION SECONDARY BIRTH COHORT PRIMARY POST-SECONDARY MALE FEMALE Source: LFS various years. Source: LFS 2017. 31 Figure 50 Figure 51 Educational distribution of working-age population by urban vs. rural Cross-country comparison: Literacy rates location, 2017 100% 100% 92 90% 90% 80% 80% 76 70% 73 69 70% 60% 61 60 LITERACY RATES 50% 60% 40% 50% 30% 40% 20% 30% 10% 20% 0 OVERALL URBAN RURAL 10% NO EDUCATION LOWER SECONDARY POST-SECONDARY 0 PAKISTAN INDIA BANGLADESH SRI NEPAL LOWER MIDDLE PRIMARY UPPER SECONDARY LANKA INCOME Source: Pakistan LFS 2017 (Figure 50); WDI (Figure 51). All data are from 2015 except Pakistan (2017), Sri Lanka (2010), and Nepal (2011). Skill acquisition and human capital formation largely happen at school, which leaves behind the high number of out-of-school children in Pakistan. The HIES 2015 suggests that Pakistan has around 21.5 million out-of-school children and youth (aged 4 to 18), of which more than half are of primary school age. This results in low human capital formation, particularly among the poor, girls, and those in rural areas. The World Bank’s recently released Human Capital Index shows that a child in Pakistan is expected to complete only 4.8 years of learning-adjusted schooling; 5.1 years for boys and 4.4 years for girls,26 indicating both low rates of school enrollment and poor learning outcomes. Net enrollment rates in Pakistan are low at all levels of schooling, and the transition from primary to secondary school is particularly weak (Figure 52).27 Enrollment rates for children aged 5–9 years are 62 percent nationally with a significant gender difference. The enrollment rates drop during the transition period between primary to lower secondary school (ages around 10), and continuously decline during the upper secondary period (ages 11 and onward). The gender difference in school enrollment rates widens with age. Children living in poverty are in a more dismal situation (Figure 53).28 The poor and non-poor gap is far larger for girls (about 21 percentage points) than boys (about 16 percentage points). At the same time, the gender difference, particularly among children aged 13 to 16, is far greater for poor children. This indicates that in the midst of overall struggles, girls living in poverty are significantly deprived of the opportunity for education. The low enrollment rate is in part explained by high dropout rates. The majority of dropouts occur at the end of primary school (grade 5), when many fail to progress to lower secondary education, and middle school (grade 8), when even larger numbers fail to progress to upper secondary education (Figure 54). The dropout rates are, as expected, highest among children living in poverty. Close to 30 percent of children attending grade 5 drop out after this grade, and this increases to 42 percent among the poor. Similarly, at the end of grade 8, 36 percent of students overall and 44 percent of the poor drop out. For older children, outside work options and earnings are likely significant factors in the decision to drop out, especially among boys. Figure 55 shows that a significant share of boys, especially those in poverty, report having worked in the last month. The transition to labor market activities tends to accelerate from age 13 (around grade 8) for boys. See Box 4 for more on child labor in Pakistan. 26 See http://www.worldbank.org/en/publication/human-capital for country-level Human Capital Index (HCI) scores and respective breakdowns by gender. 27 The education system in Pakistan consists of compulsory schooling of 12 years from 5–16 years: that is, primary school from 5 to 9 years, until grade 5; middle school from 10 to 12 years, until the end of grade 8; and secondary school from 13 to 16 years, until the end of grade 10. 28 Poverty is defined based on equivalent household expenditure per adult, adjusted for spatial price differences using the Paasche Index under the poverty line of PRs 3,250.28. These estimates constitute approximately 30 percent of the population. 32 Figure 52 Figure 53 Net enrollment rates of children aged 5–16 by gender, 2015 Net enrollment rates of children aged 5–16 by gender and poverty status, 2015 90% 90% 80% 80% 70% 70% ENROLLMENT RATE ENROLLMENT RATE 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0 0 5 6 7 8 9 10 11 12 13 14 15 16 5 6 7 8 9 10 11 12 13 14 15 16 AGE AGE BOYS (OVERALL) GIRLS (OVERALL) BOYS (POOR) GIRLS (POOR) BOYS (OVERALL) GIRLS (OVERALL) Source: HIES 2015. Poor-quality education not only discourages children from attending schools, but also affects skill building among school-attending children. The World Bank’s Learning Poverty Indicator highlights that about 27 percent of primary-school-aged children are not enrolled in school (thus excluded from learning in school) and 65 percent of students attending school do not attain the minimum proficiency level in reading (World Bank 2019b). According to studies conducted for the Annual Status of Education Report (ASER 2015a, 2015b), in urban areas only three-fifths of grade 3 students could correctly perform a subtraction like 54–25 in 2015; in rural areas this share was limited to only two-fifths of the students. ASER (2016) shows similar findings: in rural areas, 48 percent of grade 5 students and 83 percent of grade 3 students could not read a grade 2 story in Urdu, Sindhi, or Pashto. Such competency levels vary widely by household income level. According to Andrabi et al. (2008), learning outcomes may also be influenced by whether a child attends public or private school. The quality difference is so large that it would take 1.5 to 2.5 years of additional schooling for students in public schools to catch up with their counterparts in private schools in grade 3. Learning gaps between good and bad schools are also prominent as well as those between rich and poor households. Figure 54 Figure 55 Dropout rates by grade level and poverty status, 2015 Percentage of children aged 10–18 who are working, by gender and poverty status, 2015 50 90% 45 80% PERCENTAGE OF CHILDREN WORKING DROPOUT RATES BY GRADE (%) 40 70% 35 60% 30 50% 25 40% 20 30% 15 10 20% 5 10% 0 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 GRADE AGE OVERALL POOR BOY (OVERALL) GIRL (OVERALL) BOY (POOR) GIRL (POOR) Source: HIES 2015. Note: The dropout rates were the fraction of people who did not go on to the next grade using survival analysis methods for cross-sectional data, based on the self-reported “highest grade attended.” Survey questions capture information on children’s employment only for those who are 10 years of age or older. 33 BOX 4: CHILD LABOR IN PAKISTAN Statistics Over 12.5 million children in Pakistan are estimated to be working, according to the Pakistan Child Rights Movement national secretariat. The largest share is engaged in unpaid family work in the agriculture sector. Pakistani children who work are almost exclusively the poorest. The likelihood that children aged 10–14 living in the poorest households will participate in a learning activity is somewhere between 30 percent (for girls) and 55 percent (for boys). Legal framework The role of provincial governments has been increasing. In 2016, Punjab Province passed the Punjab Restriction on Employment of Children Act, which establishes 15 as the minimum age for employment and 18 as the minimum age for employment in hazardous work. The province also passed the Punjab Prohibition of Child Labor at Brick Kilns Act, which prohibits the employment of children under age 14 in brick kilns. Sindh passed laws prohibiting children under age 15 from working in factories and commercial establishments. Sindh also passed the Sindh Bonded Labor System (Abolition) Act. In 2017, Sindh Province passed the Sindh Prohibition of Employment of Children, which establishes 15 as the minimum age for employment and 19 as the minimum age for employment in hazardous work. Balochistan and Khyber Pakhtunkhwa tend to lag behind in these efforts. Federal and provincial laws, however, are not completely in compliance with international standards on child labor, including the worst forms of child labor. They tend to exclude informal work or bonded labor. Policy efforts International evidence suggests that cash transfer programs (both unconditional and conditional) have shown to be effective in reducing child labor (Skoufias et al. 2001; Bourguignon, Ferreira, and Leite 2003; de Janvry, Finan, and Sadoulet 2004; Attanasio et al. 2010; Edmonds and Schady 2012). However, as Bhalotra and Heady (2003) suggest, children in land-rich households tend to work more and be less likely to attend school than children in land-poor households. If income is not the sole reason for child labor, then income support for households should be combined with other efforts to reduce it. Source: Ahmed, Cho, and Fasih. 2019. 3.2. EDUCATION, COGNITIVE AND NONCOGNITIVE SKILLS, AND LABOR MARKET RETURNS Earnings are positively correlated with education level, with the largest effect at the postsecondary level (Figure 56). In 2017, men with a postsecondary education earned about 8 times more than those with a primary education; for women, the difference was over 13 times more. Returns to education are far greater for women at all levels, and in particular, the relative returns to secondary education are significantly larger for women. These high returns, especially for secondary and postsecondary education, further highlight that there is significant room to promote female education. The returns to education (at each level of education or for average years of schooling in the Mincerian results) have remained at more or less similar levels with the exception of year 2008. In 2008, when the economy experienced a significant downturn due to the global recession, the returns to education plummeted for both men and women. The average monthly earnings of men and women between 2001 and 2017, in real terms, highlight significant and persistent gender gaps as well as high returns at post-secondary-level education (Figure 57). While the earnings for male workers experienced annual growth rates of 2.1 to 2.6 percent in period 2001–17, the growth rates of female workers’ monthly earnings over time were far lower, especially at the primary level (0.7 percent per year). Returns to education are significantly larger for wage workers than for workers running their own small enterprises (Figure 58). Based on LSS data, the average return to an additional year of schooling is estimated to be a 7.5 percent increase in earnings among wage workers, whereas the equivalent figure for 34 Figure 56 Figure 57 Returns to education over time by gender and education level Monthly earnings by gender and education level, 2001 and 2017 RETURNS TO EDUCATION MONTHLY EARNINGS (PKR) 160% 10% 30,000 140% 8% 120% 25,000 100% 6% 20,000 80% PERCENT 4% 60% 2% 15,000 40% 20% 0 10,000 0 -2% -20% 5,000 -40% -4% 1999 2008 2014 2017 1999 2008 2014 2017 0 MALES FEMALES MALE FEMALE MALE FEMALE 2001 2017 NOTE: THE BASE IS WHEN A PERSON HAS NO EDUCATION. PRIMARY SECONDARY POST-SECONDARY MINCERIAN (RIGHT AXIS) NO EDUCATION PRIMARY SECONDARY POST-SECONDARY Source: LFS various years. Note: Wages and earnings are reported only for wage employees and adjusted with the Consumer Price Index at 2017 value. the self-employed is about 5 percent.29 This difference widens dramatically at higher levels of schooling.30 Completing lower secondary schooling yields higher returns for the self-employed—a 38 percent increase in labor earnings—as opposed to 24 percent of wage workers; in contrast, completing upper secondary school or beyond is associated with large returns for wage workers: earnings increase by 92 percent among wage workers as opposed to only 53 percent for the self-employed. For both wage workers and the self-employed, returns to schooling are larger in urban areas than in rural areas. Returns to schooling remain robust after controlling for cognitive and noncognitive skills, suggesting that schooling produces skills or signals the level of skills beyond cognitive and noncognitive skills captured by standard measures. The role of cognitive and non-cognitive skills in earnings varies by the type of employment (Figure 59). Cognitive skills are positively associated with earnings, but the association is statistically significant only for wage workers. Meanwhile, openness and extraversion (known as beta factor noncognitive skills) are strongly associated with the earnings of the self-employed. Stark differences in cognitive and noncognitive skill levels, and heterogeneous association between these skills and earnings, between wage workers and the self-employed, may explain the differential returns to education. Differences in cognitive test scores between the two groups are also large: cognitive scores among wage workers are about 0.3 standard deviations higher than among the self-employed. This suggests significant selection into wage work, as workers with higher levels of schooling and cognitive ability likely have greater access to wage work. The utilization of skills at work also varies by skill level, which may further explain the returns to education. Higher levels of cognitive and noncognitive skills are associated with a greater use of these skills at work. Evidence from the LSS indicates that individuals with higher socioemotional and cognitive skills tend to be employed in jobs where those skills are used more intensively. In particular, workers with higher scores in openness and extraversion also report being in jobs where they have frequent contact with people other than their colleagues, such as customers.31 Similarly, workers who scored high on cognitive ability tests are more likely 29 The positive association between earnings and years of schooling is statistically significant at the 1 percent level. The results are in line with the findings from the analysis of the LFS. 30 This finding is robust to accounting for selection in wage and nonwage work using the Heckman selection procedure. 31 The Big Five indicators were aggregated using an approach derived from Digman (1997) and subsequently implemented by Cunningham, Torrado, and Sarzosa (2016). The approach combines the individual elements of the Big Five into two noncognitive aggregates: an “alpha” and a “beta” trait. The alpha trait encompasses agreeableness, conscientiousness, and neuroticism (emotional stability) and can be defined as “stability,” meaning that the individual is consistent in motivation, mood, and social interactions (van der Linden, te Nijenhuis, and Bakker 2010). The “beta”‘ factor encompasses extraversion and openness to experience and is interpreted as “striving toward personal growth,” or “plasticity,” meaning the extent to which a person actively searches for new and rewarding intellectual and social experiences (Digman 1997). 35 Figure 58 Figure 59 Returns to education by employment status (wage vs. self-employed), Returns to skills by employment status (wage vs. self-employed) and 2015 skill type, 2015 RETURNS TO EDUCATION—WAGE VS SELF-EMPLOYED RETURNS TO COGNITIVE AND NONCOGNITIVE SKILLS 40% 100% 8% 90% 7% 30% 80% 6% 20% 70% 60% 5% 10% 50% 4% 40% 0 3% 30% 2% -10% 20% 10% 1% -20% WAGE SELF-EMPLOYED 0 0 WAGE SELF-EMPLOYED COGNITIVE SKILLS NON-COGNITIVE SKILLS—BETA FACTOR PRIMARY SECONDARY POST-SECONDARY MINCERIAN (RIGHT AXIS) NON-COGNITIVE SKILLS—ALPHA FACTOR Source: LSS 2015. to be engaged in tasks that are cognitively challenging as opposed to routine tasks. For example, 44 percent of workers in the top quintile of the cognitive test score distribution reported being engaged in work where the cognitive challenge is of medium intensity, as opposed to 28 percent of workers in the bottom quintile. While cognitive skills strongly correlate with years of schooling, the association between schooling and noncognitive skills is much weaker. This indicates that while formal education carries a lot of information on the cognitive skills of workers, some dimensions of skills that matter for labor productivity are not well captured by years of formal education, especially among self-employed workers. Given the strong association between some dimension of socioeconomic skills and labor earnings among the self-employed, this could partly explain why estimated returns to schooling are relatively low for self-employed workers in Pakistan. While formal schooling plays a strong role in labor earnings among wage employees and socioemotional skills do not, socioemotional skills play a more prominent role among the self-employed while years of schooling are more weakly associated with earnings. Returns to experience, which could be interpreted as on-the-job learning, are also considerably larger in wage employment than in self-employment. This is especially true for younger workers, for whom earnings increase much faster with experience than they do for young self-employed workers. This indicates limited on-the-job learning among workers running their own small enterprises and can reflect the prevalence of subsistence self-employment with limited business growth prospects. Overall, the returns of years of work experience are larger for workers in urban areas than in rural areas, particularly at a younger age, a finding that is consistent with evidence from other developing countries. Finally, there exists a positive association between technology use and profits among the self-employed. Beyond cognitive and socioemotional skills, the use of simple technology—which includes a telephone, mobile phone, pager, or other communication device—is strongly and positively associated with profits among the self-employed.32 This suggests that technology use tends to boost productivity among the self-employed. In contrast, the association of technology use with labor earnings among wage workers is very weak and statistically insignificant. 32 The results are robust to the use of technology either at work or outside work. Respondents are asked to answer yes or no to the following questions: (i) As part of this work do you (did you) regularly use a telephone, mobile phone, pager, or other communication device? (ii) On a normal day in your life [outside of work as (occupation)], do you use a telephone, mobile phone, pager, or other communication device? 36 4. PAKISTAN’S ENTERPRISES Enterprise start-ups, growth, and transformation toward more sophisticated and higher-productivity activities are key sources of more and better-quality jobs. These are also critical corollaries to the structural transformation process. In a dynamic and healthy economy, many new firms are created and, at the same time, many firms are destroyed when not competitive. Competitive firms not only persist but also grow, creating more jobs. Different production activities, across sectors, may utilize different intensities of capital and labor. Thus, their contributions to productivity, value added, and job creation may vary. Widespread research on firm dynamics specifically focuses on how frequently firms are created and destroyed, and how such patterns differ by firm characteristics (e.g., size, sector, capital and labor intensity, and managerial capacity, among others).33 Answers to questions about the dynamics of firms and how these relate to the overall economy often come from well-structured data. Analysis of the overall landscape relies on economic censuses: they investigate all enterprises registered at a fixed location (see Box 5 for various types of establishments under labor regulations in Pakistan). In Pakistan, data that enable research on the landscape and dynamics of firms are very limited. The latest economic census in Pakistan was conducted from 2001 to 2003, and released in 2005, by the Pakistan Bureau of Statistics (PBS). The latest Enterprise Survey in Pakistan was conducted in 2013–16 for nonagricultural enterprises with five or more employees. However, the survey suffered low response rates, undermining the credibility of the results. The PBS conducted a Census of Manufacturing Industries (CMI) in 2015–16 but the data have not been released yet. Under these circumstances, this chapter discusses an overall business environment that impacts enterprise dynamics and the two important sources of domestic job creation. With respect to the business environment, the World Bank’s Doing Business results in Pakistan are briefly summarized. The first source of domestic job creation to be discussed is nonfarm, nonagricultural household enterprises where a significant number of nonwage workers outside agriculture are engaged, using household surveys including the Household Income and Expenditure Survey (HIES) and Labor Skills Survey. The second is digital jobs, which constitute a small, but increasingly important, part of Pakistan’s labor market. BOX 5: DEFINITIONS OF BUSINESS ESTABLISHMENTS UNDER LABOR REGULATIONS IN PAKISTAN Typically, enterprise surveys or economic censuses aim to collect key information on economic establishments. There are different types of business establishments in Pakistan, and the definitions vary slightly by province. Shops: Any premises used wholly or in part for the wholesale or retail sale of commodities or articles, either for cash or on credit, or where services are rendered to customers. These may include an office, a showroom, a “godown” (that is, storage below ground level), a warehouse, or another place of work mainly used in connection with such trade or business. 33 See Haltiwanger (2006); Haltiwanger, Jarmin, and Miranda (2010); Haltiwanger et al. (2013, 2016); and Acs and Muller (2008) for further discussion. 37 Commercial establishments: Any premises in which the business of advertising, commissioning, or forwarding is conducted; a commercial agency, including a clerical department of a factory; the office establishment of a person who for the purpose of fulfilling a contract with the owner of any other commercial establishment or industrial establishment employs workers; a unit of a joint stock company; an insurance company; a banking company or a bank; a broker’s office or stock exchange; a club; a hotel; a restaurant or an eating house; a cinema or theater; societies registered under the Societies Registration Act; and any other establishment designated as “commercial” via government notification. In Sindh, this definition also includes schools, colleges, private educational institutions, hospitals, private health centers, clinical laboratories, and private security agencies. In Khyber Pakhtunkhwa, the definition is similar to Sindh’s but excludes all education-related establishments. Industrial establishments: A factory; railway, tramway, or motor omnibus service; dock, wharf, or jetty; mine, quarry, or oilfield; a workshop or other establishment in which articles are produced, adapted, or manufactured, with a view to their use, transport, or sale; establishment of a contractor who, directly or indirectly, employs workers to do any skilled or unskilled, manual, or clerical labor for hire or reward in connection with the execution of a contract to which he is a party. Road transport establishments: Any road transport service in which workers are employed, and where the business of such service is not directly managed by an owner. Source: Based on shop/establishment ordinances and acts, factory acts, and industrial and commercial employment standing orders of Punjab, Sindh, and Khyber Pakhtunkhwa; Sindh Terms of Employment Act; and Road Transport Workers Ordinance. 4.1. BUSINESS ENVIRONMENT FOR ENTERPRISES The entry and performance of enterprises depends largely on business environment. In particular, there is a large body of literature that has recognized the importance of key regulations governing business activities such as procedures to start a firm, access to finance, access to infrastructure such as electricity, and tax policies.34 Firm activities and entrepreneurial resources can flourish when rules governing their operation are clear, efficient, and coherent—as opposed to arbitrary, cumbersome, and contradictory. In this spirit, the Doing Business captures about 45 key indicators in 11 dimensions35 of a business environment that have implications on business dynamics and performance, and provides a cross country comparison (see World Bank 2018, 2019c, for example). The Doing Business reports both the ease of doing business score for each indicator and its ranking in countries where data were collected. Pakistan has experienced a major downward trend in its Doing Business ranking from 2007 to its lowest in 2018, but due to greater policy attention and reform efforts, it has made progress in 2019 and 2020. The Doing Business exercise ranked Pakistan 74th in 2007, deteriorating to 147th in 2018. With recent business climate reform drives, the ranking improved to 136th in 2019 and 108th out of 190 economies in the latest 2020 exercise. Major progress came from the ease of getting electricity, where the score increased by 20.9 percentage points from 43.1 in 2019 to 64.0 in 2020. Also, the scores related to construction permits increased by 14.6 percentage points from 51.9 in 2019 to 66.5 in 2020. Although the Doing Business data are collected only in the two largest cities in Pakistan—Karachi and Lahore—the ranking and scores highlight some sobering needs for policy reforms. In particular, the processes and regulations that influence the accumulation of firms’ physical capital (e.g., dealing with construction permits, registering property, paying taxes) and areas that are critical to firms’ operations (e.g., getting electricity and enforcing contracts) lagged despite some notable progress. The electricity challenges with very high costs have been well recognized in Pakistan. The enterprise survey 2013 (World Bank 2012b) also pointed out that one of the greatest difficulties faced by enterprises is electricity. The recent improvement in construction permits has been focused on reducing procedures, time, and costs to complete all formalities 34 See Djankov (2016) for background and summary. 35 The 11 dimensions include: starting a business, dealing with construction permits, getting electricity, registering property, getting credit, protecting minority investors, paying taxes, trading across borders, enforcing contracts, resolving insolvency, and labor market regulation. 38 to build a warehouse and get the inspection regarding the building code and safety regulations done. Also, regulations related to taxes, which measure the time taken to prepare, file, and pay (or withhold) the corporate income taxes, value added or sales taxes, and labor taxes, including payroll taxes and social contributions (in hours per year), suggest inefficiency in tax collection. Along with a very low tax revenue–GDP ratio in Pakistan (about 11 percent), tax related Doing Business scores call for policy intervention in this area. Compared with the macroeconomic environment, which is affected by external factors and is not entirely under the government’s policy control, the business environment can be significantly improved by a strong government commitment. Policy reforms and prioritization, and the adoption of innovative technology, can make a huge difference in the business environment. Several countries have shown examples of the success driven by strong commitment and use of technology.36 Pakistan has renewed its efforts in this agenda, recognizing the importance of the business environment for investment and firm productivity. While the Doing Business exercise captures the environment for small and medium sized domestic enterprises, the overall enhancement in the business environment will likely increase entrepreneurial activities in the economy regardless of the size of the firms, promote foreign direct investment as well as trade competitiveness, thus potentially contributing to regional connectivity, and further catalyze structural transformation (see World Bank 2019a for further discussions). 4.2. NONFARM, NONAGRICULTURAL HOUSEHOLD ENTERPRISES (Informal) household enterprises are widespread in developing countries and have the potential to act as facilitators of a structural change from agricultural to nonagricultural jobs.37 Such enterprises are defined as any entrepreneurial, income-generating activity owned by a household member. They are often conducted on household premises and are automatically excluded from the purview of labor regulations (discussed in Box 5). Many studies (e.g., Fox and Sohnesen 2012; Filmer and Fox 2014; Lanjouw and Lanjouw 2001) argue that household enterprises matter for growth, jobs, and livelihoods in the context of Sub-Saharan Africa. The reasons are that the majority of new nonfarm jobs can be created through household enterprises; incomes generated from their activities are important to families’ livelihoods and for escaping poverty; and enterprises that survive the initial start-up phase tend to persist and contribute to economic productivity. Are household enterprises a driver of jobs, livelihoods, and growth in Pakistan? In the context of Pakistan and most South Asian economies, informal businesses and self-employment dominate the labor market, where formal businesses are the exception. Thus, household enterprises have absorbed large shares of new labor market entrants in the past, underlining that they constitute an important source for employment and livelihoods. In addition to integrating workers in the mostly informal labor market, nonfarm, nonagricultural (NFNA) household enterprises are believed to be a stepping-stone to better-paying industry and service sector jobs, thereby stimulating a much-needed structural change in Pakistan. The role of household enterprises in growth may depend on their productivity, regarding which little consensus exists given its large heterogeneity. This section focuses on the profiles and productivity of NFNA household enterprises captured in the HIES in Pakistan. NFNA household enterprises, either in urban or rural areas (Figure 60), operate mostly in an informal setting without a business registration under the country’s company act. These are primarily self-employed individuals who are running unincorporated businesses, sometimes with family or casual workers, with or without pay. Depending on the status of their owner, NFNA household enterprises are thus further disaggregated into four categories: (i) employer with paid employees; (ii) own-account worker with casual laborers and helpers who incur labor costs but are not employees; (iii) own-account worker with nonpaid family employees; and (iv) own-account worker without any employees.38 36 For case studies that include several success stories, see: https://www.doingbusiness.org/en/reports/case-studies/view-all. 37 This section builds on the Pakistan Jobs Diagnostic’s background paper (Weber and Langbein 2018). 38 In HIES 2015–16, there is no information regarding the exact number of employees or casual laborers in each household enterprise. Instead, the data report the total labor costs used for wages for employees for only those enterprises with fewer than 10 paid employees. 39 Figure 60 Classification of household enterprises HOUSEHOLDS: DOES ANY HOUSEHOLD MEMBER ENGAGE IN AN ENTREPRENEURIAL, INCOME-GENERATING ACTIVITY? NO: NO HOUSEHOLD ENTERPRISE YES: HOUSEHOLD ENTERPRISE IN FARMING OR AGRICULTURE? YES: FARM ACTIVITIES (FAMILY FARM) NO: NONFARM, NONAGRICULTURAL (NFNA) ENTERPRISE URBAN RURAL • EMPLOYER WITH PAID EMPLOYEES • OWN-ACCOUNT WORKER WITH CASUAL LABORERS • OWN-ACCOUNT WORKER WITH NONPAID (FAMILY) WORKER • OWN-ACCOUNT WORKER WITHOUT ANY EMPLOYEE Source: Based on the household enterprise module of HIES (2015). The HIES 2015 shows that about a fifth of Pakistan’s households operate NFNA enterprises. This is equivalent to almost 6.9 million households—a significant number. The prevalence of NFNA household enterprises varies across provinces and also between rural and urban areas (Figure 61 and Figure 62). Punjab has the highest share; here, about one out of four households operate NFNA household enterprises. As expected, NFNA household enterprises are more common in urban areas (27.8 percent of households) than rural ones (17.1 percent of households). But the urban rate is notably lower than that of several countries in Sub-Saharan Africa: the share of urban households that operate NFNA enterprises is 56 percent in Ethiopia, 35 percent in Malawi, 68 percent in Niger, 71 percent in Nigeria, 60 percent in Tanzania, and 54 percent in Uganda.39 NFNA household enterprises consist of mostly own-account workers without any employees (Figure 63). Close to 60 percent of NFNA household enterprises report having no workers but the owner, and Figure 61 Figure 62 Share of nonfarm, nonagricultural household enterprises by province, Share of nonfarm, nonagricultural household enterprises by rural vs. urban 2015 location, 2015 30% 30% 25% NONAGRICULTURAL ENTERPRISES AS 25% A SHARE OF ALL HOUSEHOLDS 20% 20% 15% 15% 10% 10% 5% 5% 0 0 KHYBER PUNJAB SINDH BALOCHISTAN RURAL URBAN PAKHTUNKHWA Source: HIES 2015. 39 See Nagler and Naudé (2017) for the period 2005–13. 40 Figure 63 Figure 64 Employment structures of nonfarm, nonagricultural household enterprises Shares of households running nonfarm, nonagricultural enterprises in rural in urban vs. rural locations, 2015 vs. urban locations, 2007 and 2015 TYPE OF NFNA HOUSEHOLD ENTERPRISES BY REGION SHARE OF NFNA HOUSEHOLD ENTERPRISES OVER TIME 30.0% 27.8% 26.4% URBAN 25.0% 19.5% 20.0% 17.1% 15.0% RURAL 10.0% 0 20% 40% 60% 80% 100% 5.0% OWN ACCOUNT WORKER WITHOUT ANY EMPLOYEE OWN ACCOUNT WORKER WITH NON-PAID (FAMILY) WORKERS 0 RURAL URBAN OWN ACCOUNT WORKERS WITH CASUAL LABORERS EMPLOYER WITH PAID EMPLOYEES (LESS THAN 10) 2007 2015 Source: HIES 2015. Note: NFNA = nonfarm, nonagricultural. 25 percent are family businesses run by the owner and nonpaid family workers. The shares of own-account workers with casual laborers, and employers with paid employees are higher in urban areas (19 and 7 percent, respectively) than rural areas (7 and 3  percent, respectively). This suggests that the ability to create wage employment in these enterprises is quite limited in rural areas. However, given the prevalence of NFNA household enterprises in urban areas and that about 26 percent of them hire some workers, urban enterprises appear to serve as a significant source of job creation. Between 2007 and 2015, the overall share of households that operated NFNA household enterprise activities remained at a similar level, with a marginal decline in rural areas (from 19.5 to 17.1 percent) and increase in urban areas (from 26.4 to 27.8 percent) (Figure 64). This is equivalent to a 4.5 percent and 3.2 percent a year growth in the number of household enterprises in urban and rural areas, respectively. These are far faster than the average growth of the number of workers in the economy, suggesting that NFNA household enterprises are increasingly playing an important role in absorbing workers. About 86 percent of both rural and urban NFNA household enterprises are engaged in the service sector (Figure 65 and Figure 66). Wholesale and retail trade are the most common activities, followed by manufacturing and transportation and storage. The Labor Skills Survey provides some specific examples of enterprises such as corner shops selling general goods, grocery stores, transport services, courier services, flour mills, tailoring and garment shops, auto mechanic garages, and so on. The prevalence of retail and services among NFNA household enterprises mirrors the pattern in other regions (Fox and Sohnesen 2012; Vijverberg and Haughton 2004). Activities in the transportation and storage sectors are more common in rural than urban areas probably due to the need for storing agricultural goods such as grain and dairy products. Based on HIES 2015, an estimated 9.9 million individuals engage in NFNA household enterprises.40 Given that about 10.8 million workers are reported to be self-employed in the nonagricultural sector according to the LFS 2017, this number appears to be reasonable. This indicates that each NFNA household enterprise has about 1.4–1.6 workers on average, including the owner. When the total number of individuals engaged in NFNA household enterprises is disaggregated by province, about 70 percent are from Punjab (Figure 67). The large majority are own-account workers without any laborers or employees (Figure 68). Urban representation is significantly noticeable among employers and own account workers with casual laborers, corroborating the role of urban NFNA household enterprises in creating jobs. To the contrary, rural NFNA household enterprises are much more likely to receive help from nonpaid family workers. 40 The HIES does not provide an exact number of workers for household enterprises with paid employees or casual laborers. However, it does report labor costs in case the enterprises incur payments to workers, and also the number of nonpaid family workers. The number of paid employees and laborers is estimated based on the minimum wage, following Weber and Langbein (2018). 41 Figure 65 Figure 66 Shares of rural nonfarm, nonagricultural household enterprises by sector Shares of urban nonfarm, nonagricultural household enterprises by sector or activity, 2015 or activity, 2015 NFNA HOUSEHOLD ENTERPRISES ACROSS INDUSTRY, RURAL NFNA HOUSEHOLD ENTERPRISES ACROSS INDUSTRY, URBAN WHOLESALE AND RETAIL TRADE WHOLESALE AND RETAIL TRADE TRANSPORTATION AND STORAGE MANUFACTURING MANUFACTURING TRANSPORTATION AND STORAGE RETAIL TRADE OUTSIDE STORES RETAIL TRADE OUTSIDE STORES OTHER PERSONAL SERVICE ACTIVITIES OTHER PERSONAL SERVICE ACTIVITIES REPAIR OF MOTOR VEHICLES, REPAIR OF MOTOR VEHICLES, MOTORCYCLES, AND ELECTRONICS MOTORCYCLES, AND ELECTRONICS RESTAURANTS AND MOBILE FOOD SERVICES EDUCATION; HUMAN HEALTH AND SOCIAL WORK EDUCATION; HUMAN HEALTH AND SOCIAL WORK PROFESSIONAL, SCIENTIFIC, AND TECHNICAL ACTIVITIES PROFESSIONAL, SCIENTIFIC, AND TECHNICAL ACTIVITIES OTHER ACTIVITIES OTHER ACTIVITIES ACCOMMODATION AND FOOD SERVICE ACCOMMODATION AND FOOD SERVICE RESTAURANTS AND MOBILE FOOD SERVICES CONSTRUCTION CONSTRUCTION MINING AND PUBLIC UTILITIES 0 10% 20% 30% 40% 50% 0 10% 20% 30% 40% 50% Source: HIES 2015. Note: NFNA = nonfarm, nonagricultural. Figure 67 Figure 68 Number of individuals working in nonfarm, nonagricultural household Number of individuals working in nonfarm, nonagricultural household enterprises by province, 2015 enterprises by employment structure and by rural vs. urban location, 2015 6,966,550 EMPLOYER WITH PAID EMPLOYEES (LESS THAN 10) OWN ACCOUNT WORKERS WITH CASUAL LABORERS OWN ACCOUNT WORKER WITH NON-PAID (FAMILY) WORKERS 1,677,751 990,971 OWN ACCOUNT WORKER WITHOUT ANY EMPLOYEE 224,894 0 1,000,000 2,000,000 3,000,000 4,000,000 KHYBER PUNJAB SINDH BALOCHISTAN PAKHTUNKHWA URBAN RURAL Source: HIES 2015. The majority of household enterprises are run by male household heads, which points at gender barriers in pursuing entrepreneurial activities. In contrast to the higher engagement of women in entrepreneurial activities in Sub-Saharan Africa (Ackah 2013) and other countries in Asia (e.g., Myanmar), in Pakistan only 4 percent of the households running an NFNA enterprise are female headed.41 There is no significant difference in the share of female-headed NFNA enterprises between urban and rural areas. Apart from the many barriers to female labor force participation discussed earlier, there are additional obstacles to women seeking to start and operate a business. These include lack of access to capital, inadequate exposure to markets and business networks, and limited property rights. Weak decision-making power within a household further limits women’s access to time or resources for business purposes (Minniti and Naudé 2010; Giné, Mansuri, and Picón 2011). 41 Nationally, almost 11 percent of households are female headed (HIES 2016). 42 Figure 69 Figure 70 Educational attainment (in years) of household head running a nonfarm, Share of household heads running a nonfarm, nonagricultural household nonagricultural household enterprise by employment structure and rural enterprise who own property by employment structure and type of vs. urban location, 2015 property, 2015 SCHOOLING OF HOUSEHOLD HEAD OF NFNA HOUSEHOLD ENTERPRISES: 12.5% NUMBER OF YEARS 9 9 9.4% 8.8% 7.9% 7 7 6 6.3% 6 5.5% 4.7% 4.4% 5 4.0% 4 3.1% 2.6% 1.8% OWN ACCOUNT OWN ACCOUNT OWN ACCOUNT EMPLOYER WITH WORKER WORKER WITH WORKER WITH PAID EMPLOYEES WITHOUT ANY NON-PAID (FAMILY) CASUAL LABORERS (LESS THAN 10) OWN ACCOUNT OWN ACCOUNT OWN ACCOUNT EMPLOYER WITH EMPLOYEE WORKERS WORKER WORKER WITH WORKER WITH PAID EMPLOYEES WITHOUT ANY NON-PAID (FAMILY) CASUAL LABORERS (LESS THAN 10) SHARE OF HOUSEHOLDS OWNING NON-AGRICULTURAL LAND EMPLOYEE WORKERS SHARE OF HOUSEHOLDS OWNING AGRICULTURAL LAND RURAL URBAN SHARE OF HOUSEHOLDS OWNING COMMERICAL PROPERTY Source: HIES 2015. Note: NFNA HEs = nonfarm, nonagricultural household enterprises. The educational attainment of the household heads running NFNA enterprises suggests a clear difference between those who hire paid workers and those who work without them. The average years of schooling of employers and own-account workers hiring laborers is about 8.5 (9.2 years in urban areas and 6.7 years in rural areas), which is significantly higher than the average for working-age males (8.2 in urban areas and 5.6 in rural areas) (Figure 69). By contrast, the average years of schooling of own-account workers without paid workers is about 5 years (5.9 years in urban areas and 4.5 years in rural areas), which is significantly lower than that of average working-age males. Such large variations in educational attainment suggest a difference between subsistence versus growth-oriented household enterprises. Better-educated, growth-oriented entrepreneurs seem to start and operate businesses that create job opportunities for other people, whereas less-educated, subsistence-type entrepreneurs may fail to grow and fail to generate employment. However, the difference in the asset ownership of enterprises with paid workers versus those without is not as stark as for educational attainment (Figure 70). Profits of NFNA household enterprises are calculated by deducting enterprise input costs from enterprise revenues. For employers and own-account workers with casual laborers, the labor expenses are also deducted. A small share (about 1.3 percent of NFNA enterprises) reports a negative profit; these are included here as zero-profit enterprises. Figure 71 shows that except for own-account workers with casual laborers, the profit levels of urban NFNA household enterprises are significantly greater than those in rural areas. Also, it is observed that service sector profits are far greater than those of manufacturing in both urban and rural areas (Figure 72). When households with NFNA enterprises are compared with those without enterprises, a few significant differences are observed (Figure 73). The share of female-headed households is significantly lower among households with NFNA enterprises than the average households. Land ownership—whether of agricultural or nonagricultural land—in general does not appear to be a strong predictor of an NFNA household enterprise. Residential and commercial property ownership, on the other hand, is strongly and positively correlated with NFNA enterprises. This may indicate that having access to a dwelling is important to operating an enterprise in both rural and urban settings. Commercial property ownership is the strongest predictor of nonagricultural household enterprises: the relative probability of engagement in NFNA enterprises is over two times higher for commercial property owners. 43 Figure 71 Figure 72 Yearly profits of nonfarm, nonagricultural household enterprises by Yearly profits of nonfarm, nonagricultural household enterprises in two key employment structure and urban vs. rural location, 2015 sectors by urban vs. rural location, 2015 2,500,000 1,693,992 2,000,000 ANNUAL PROFITS (PKR) 1,500,000 1,134,735 1,000,000 500,000 496,678 279,619 0 OWN ACCOUNT OWN ACCOUNT OWN ACCOUNT EMPLOYER WORKER WORKER WITH WORKER WITH WITH PAID WITHOUT ANY NON-PAID (FAMILY) CASUAL EMPLOYEES EMPLOYEE WORKERS LABORERS (LESS THAN 10) MANUFACTURING SERVICES URBAN RURAL URBAN RURAL Source: HIES 2015. Note: Top 5 percent outliers were dropped in these graphs. Figure 73 A comparison of key characteristics across households that run a nonfarm, nonagricultural enterprise vs. those that do not, 2015 NFNA VS. NO HE 100 90 80 70 60 50 40 30 20 10 0 SHARE OF SHARE OF SHARE OF OWN OWN OWN OWN COMPUTER INTERNET FEMALE-HEADED HOUSEHOLDS IN LITERATE NON-AGRICULTURAL AGRICULTURAL RESIDENTIAL COMMERCIAL USAGE IN USAGE IN HOUSEHOLDS URBAN AREAS HOUSEHOLD LAND LAND PROPERTY PROPERTY LAST YEAR LAST YEAR MEMBERS NFNA HE NO HE Source: HIES 2015. Note: NFNA = nonfarm, nonagricultural; HE = household enterprise. The poverty rate is lower among households with NFNA enterprises compared with the national average (Figure 74). The overall poverty rate based on HIES 2015 is 24.3 percent. Poverty levels are lowest among employers (4 percent) followed by own-account workers with casual laborers (7 percent), significantly different from those of own-account workers (around 21 percent). Compared with the significantly higher level of profits in the service sector relative to manufacturing, as discussed above, there is no significant difference in poverty rates among household enterprises. Overall, the poverty rates of households with NFNA enterprises are lower than the national average, largely driven by low poverty rates among those in urban areas (Figure 75). This underlines the capacity of nonagricultural household enterprises to generate a sustainable livelihood. 44 Figure 74 Figure 75 Poverty rates of nonfarm, nonagricultural household enterprises by Poverty rates of nonfarm, nonagricultural household enterprises in general employment structure and sector, 2015 and by rural vs. urban location, 2015 NATIONAL POVERTY RATE 24.3 25.0 NATIONAL POVERTY RATE 24.3 24.9 21.1 21.5 18.7 20.0 17.6 18.5 15.0 11.4 7.0 10.0 4.0 5.0 OWN OWN ACCOUNT OWN EMPLOYER MANU- SERVICES ACCOUNT WORKER WITH ACCOUNT WITH PAID FACTURING WORKER NON-PAID WORKERS EMPLOYEES WITHOUT ANY (FAMILY) WITH CASUAL (LESS THAN 0 EMPLOYEE WORKERS LABORERS 10) NFNA HOUSEHOLD NFNA HOUSEHOLD NFNA HOUSEHOLD TYPE SECTOR ENTERPRISES ENTERPRISES RURAL ENTERPRISES URBAN Source: HIES 2015. Note: NFNA HEs = nonfarm, nonagricultural household enterprises. These businesses tend to continue traditional ways of operation without pursuing much innovation, risk taking, or firm growth. While firms do seem well established in their local markets, many do not have a systematic mechanism to learn about prices and expressed a need for more market information to properly set prices (based on LSS data on the self-employed). Since firms are primarily set up using individual capital, household income, or asset sales (not through loan finance), many owners do not express plans for expansion, even if they are making profits and investments. Studies note two main constraints to firm growth—lack of access to credit and business management capacity—and Pakistan is no exception. Financial inclusion rates are rather low in Pakistan: in 2017, 47 percent of adults indicated that they had borrowed money in the past year and only 3 percent had borrowed money for a farm or business.42 In regard to business management, Giné and Mansuri (2018) found that offering an eight-day business training to entrepreneurs led to improved business and household outcomes. However, these results were significant only among men. The gender difference is consistent with findings from other countries (e.g., Sri Lanka and Ghana). In addition to these two major constraints, some emerging studies emphasize the importance of fostering a growth mindset among the self-employed. For instance, Campos et al. (2017) found that the provision of training focused on a self-starting, future-oriented, and persistently proactive mindset significantly improved business outcomes. 4.3. DIGITAL JOBS IN PAKISTAN While traditional forms of economic activities in household enterprises and agricultural activities seem to dominate the labor market of Pakistan, information and communication technology (ICT) is changing the world of job markets. Computing, the Internet, and mobile phones create opportunities for firms and individuals to increase their productivity, reduce transaction costs, and overcome entry barriers in various markets. At the same time, these technologies could displace human effort through automation, or shift the location of work through a mix of globalization and outsourcing. Pakistan is also affected by ICT development and has seen digital jobs changing labor market opportunities. Digital jobs can be created through three main channels: the technology sector, online gig economy (freelancing and sharing economy), and digital society. 42 Global Findex Database. 45 Technology sector This sector includes those businesses where production and distribution of ICT-based goods and services are the main outputs. Such activities include software development, computer products, and services related to telecommunications and information technology (IT). From a nascent stage in the late 1980s, Pakistan’s technology sector has grown significantly, generating around US$3.5 billion in annual revenues as of 2017, equivalent to just under 1 percent of nominal GDP. There are over 1,500 registered IT firms in the country, and over 15 IT parks (with IT-enabled infrastructure) have been established. The growth in part is driven by the growth in IT exports, with an estimated IT export revenue of US$5 billion by 2020 (Pakistan Software Export Board). It is hard to estimate the exact number of workers in the sector but according to a survey conducted by the Pakistan Software Houses Association in 2015, 120,000 jobs have been created. Freelancing (online gig) economy and sharing economy Technological development has given workers more access to consumers and investors locally as well as globally and has exponentially expanded freelancing opportunities for them. Well-known freelancing platforms include Freelancer.com, Upwork.com, Guru.com, and Fiverr.com (Iftikhar 2017). These are the platforms where individuals or agencies can bid for contracts posted by employers from around the world, and then complete the work online. These tasks can be done virtually e.g. graphic design, software programming, and legal and accounting services. More than 200,000 Pakistanis are estimated to be working on various online work platforms as independent contractors. Meanwhile, platforms for the sharing economy (e.g., ride-sharing service Uber) provide earning opportunities to a large number of affiliated workers. In Pakistan, the global platform Uber and regional leader Careem, established in Dubai, have been active in large cities like Karachi and Lahore since 2016. Compared with traditional digital jobs in the technology sector, freelancing and the sharing economy offer several advantages in creating better and more inclusive jobs. Since the online gig economy allows flexible work hours and remotely completed tasks, it can provide job opportunities to those who have mobility and/or time constraints—women, students, and people with disabilities, for instance. Also, online freelancing links the local supply of talent to the global pool of demand with a large set of potential markets. Estimating the size of the workforce in the freelancing and sharing economy and its value added through activities is a challenge due to the lack of a consistent definition or measurement methods. Given the flexibility and diversity of platform activities, individuals can have multiple jobs without any employment-employee relations—a setup that may or may not be captured by traditional household surveys. The Online Labour Index43 ranks Pakistan third after India and Bangladesh, with about 13 percent of world’s total supply of workers in online freelancing (Figure 76) (Kässi and Lehdonvirta 2016). This is not directly translated into market shares of the revenue generated from freelancing work. The global average hourly rate is US$21 and on average people work for 36 hours per week according to a Payoneer Freelancer Income Survey in 2015. In Pakistan, freelancers generally work for 34 hours per week and types of jobs include IT and programming; design and multimedia; writing and translation; and sales and marketing. The hourly rates for these jobs are between US$17 and US$23.44 The digital society in the real economy The use of digital technology is expanding on many fronts, transforming the ways companies do business and transactions, and providing them with new digital assets, means of communication, and models of customer engagements. Many traditional firms adopt digital technologies in their business processes and heavily invest in online transactions, for example, business processing outsourcing (BPO), which involves the contracting of the operations and responsibilities of a specific business process to a third-party service provider. Although the sector in Pakistan has been repeatedly identified for rapid growth potential, 43 It is published by the Oxford Internet Institute. 44 http://www.technologyreview.pk/working_hard_for_money/ 46 Figure 76 Selected countries’ market share of freelancing work, by type of work, 2017 INDIA BANGLADESH PAKISTAN UNITED STATES PHILIPPINES UNITED KINGDOM UKRAINE CANADA EGYPT KENYA LEBANON SRI LANKA GERMANY NIGERIA ITALY 0 2.5% 5% 7.5% 10% 12.5% 15% 17.5% 20% 22.5% 25% 27.5% SHARE SOFTWARE DEVELOPMENT AND TECHNOLOGY CREATIVE AND MULTIMEDIA SALES AND MARKETING SUPPORT WRITING AND TRANSLATION PROFESSIONAL SERVICES CLERICAL AND DATA ENTRY Source: Online Labour Index 2017. the reality remains that its BPO sector generally lags behind in this area:45 Pakistan ranked 28th in terms of attractiveness among 55 global BPO destinations.46 Despite the low labor costs in Pakistan, workers’ inadequate skills and a weak business environment (e.g., prompting concerns associated with the protection of intellectual property rights, infrastructure, and security) appear to be the main factors behind Pakistan’s low ranking. 45 However, its subsector—call centers, where helpdesks, reservations, and customer services are handled—saw the most substantial growth in Pakistan with projected growth of 3 percent annually between 2015 and 2010 (Ahmed and Javed 2015). 46 2016 A.T. Kearney Global Services Location Index. 47 5. URBANIZATION AND INTERNAL MIGRATION FOR JOBS 5.1. URBANIZATION IN PAKISTAN The spatial landscape of Pakistan’s population and employment growth has been changing with a varying degree across provinces. All four provinces witnessed a significant increase in population between 1998 and 2017 based on the population census of the respective year. The overall population rose from 132 million to 208 million, which is equivalent to 2.4 percent growth per year.47 About half of this increase came from Punjab (Figure 77). Meanwhile, between 1999 and 2017 based on the labor force surveys of the respective year, the overall number of workers in the labor market increased by 26.6 million, of which 55.5 percent were from Punjab (Figure 78). Jobs grew at an overall rate of 3.1 percent per year, almost 0.7 percentage points higher than the general population growth. This highlights a favorable demographic and labor market transition, with a more rapid increase in workers than the general population. The faster jobs growth compared to population growth is particularly significant in Punjab and Sindh. The fact that more jobs were created in certain parts of the country than in others may be related to internal migration and inflows of the working-age population into urban areas. In 1999, only 29 percent of workers, equivalent to 10 million people, resided in urban areas (Figure 79). It increased steadily and by 2017 about 34 percent of workers (21.1 million) were in urban areas. The pace and level of urbanization varies by province (Figure 80) where Sindh is by far the most urbanized province, with over 47 percent of workers Figure 77 Figure 78 Increases in population by province, 1998 and 2017 Increases in number and share of workers by province, 1999 and 2017 POPULATION GROWTH BY PROVINCE EMPLOYMENT GROWTH 120 60% 40 60% 35 100 50% 50% 30 80 40% 40% 25 MILLIONS MILLIONS 60 30% 20 30% 15 40 20% 20% 10 20 10% 10% 5 0 0 0 0 KHYBER PUNJAB SINDH BALOCHISTAN KHYBER PUNJAB SINDH BALOCHISTAN PAKHTUNKHWA PAKHTUNKHWA 1998 2017 SHARE OF INCREASE 1999 2017 SHARE OF INCREASE Source: Population census (left); LFS 1999, 2017 (right). 47 This growth has implications for many aspects of the economy. The most pressing priorities would be to improve investment in the human capital of the incoming generation and at the same time increase the utilization of the current stock of human capital (see Ahmed, Cho, and Fasih [2019] for further discussion). 49 Figure 79 Figure 80 Distribution of workers over time in rural vs. urban areas by number and Changes in the urban share of workers over time by province share of total URBANIZATION OF EMPLOYED POPULATION URBANIZATION OF EMPLOYED PEOPLE BY PROVINCE 40 36% 60% 35 35% 34.9% 50% 48.4% 34% 45.3% 44.6% 45.6% 30 33% 25 40% 32% 33.5% MILLIONS 20 29.6% 29.2% 31.7% 30% 27.7% 31.5% 31% 25.0% 15 22.2% 23.5% 30% 29.8% 20% 16.3% 10 29% 18.1% 18.6% 15.6% 16.4% 5 28% 10% 1999 2008 2014 2017 0 27% 1999 2008 2014 2017 KHYBER PAKHTUNKHWA PUNJAB RURAL URBAN URBAN SHARE OF WORKERS (RHS) SINDH BALOCHISTAN Source: LFS various years. employed in urban areas in 2017. Balochistan’s urbanization started from a low base (at 15 percent) with a high growth rate, and about 25 percent of workers were in urban areas by 2017. This is in line with a global trend of urbanization. Urbanization has been often listed as one of the global megatrends that affect the economic lives of people, along with others such as demographic transition, use of technology, and climate change. In 2014, high levels of urbanization at or above 80 percent characterized Latin America and the Caribbean region. Africa and Asia remain mostly rural, with 40 percent and 48 percent of their respective populations living in urban areas, but are expected to urbanize faster than other regions (Gentilini 2015). The increase in urbanization and growth in urban population come from three sources: natural population growth with births, migration, and reclassification of urban areas. With relatively high fertility rates in Pakistan, natural growth is an important factor explaining population growth in urban areas. However, the Pakistan Demographic and Health Survey (DHS) 2017–18 shows that the total fertility rate in rural areas at 3.9 is significantly higher than that in urban areas at 2.9.48 Academically, the classification of urban areas considers accessibility, density, and population threshold. However, the urban classification in Pakistan census followed predefined historical criteria. This raised the issue of undercounting the urban population and some called for reclassification of urban areas.49 Domestic migration appears to be a significant source of growth in urban population in Pakistan. 5.2. INTERNAL MIGRATION AND LABOR MARKET IMPLICATIONS The LFS 2017 reports individuals’ migration experiences in the past 10 years. About 6.7 percent of the working-age population reports coming back from abroad during that time period (Table 2). Almost one-quarter of those who migrated into KPK came back from abroad, indicating a significant number of workers from KPK seeking temporary overseas employment and also returning. Excluding those who came back from abroad, about 70 percent of the internal migration is within the province; the other 30 percent is across provinces. About 29 percent of all internal migration was from rural to urban areas. Interestingly, rural-to-rural migration was equally common, at 29 percent, followed by urban-to-urban migration (24 percent). Urban-to-rural migration was also not uncommon, at about 18 percent of all internal migration. 48 See NIPS (2019) for further discussion. 49 See Bari (2012) for further discussion. 50 Table 2 Prevalence of migration into the four provinces in Pakistan within the past 10 years Pattern of In-Migration by Province 2017 (%) Migration from Migration to Overall KPK Punjab Sindh Balochistan KPK 44.9% 12.1% 7.3% 3.9% 17.5% Punjab 16.7% 77.9% 19.6% 4.1% 57.5% Sindh 11.9% 6.0% 71.3% 22.1% 16.4% Balochistan 1.7% 1.0% 1.6% 69.9% 1.9% Abroad 24.9% 3.0% 0.3% 0.0% 6.7% Total 100.0% 100.0% 100.0% 100.0% Overall 18.9% 66.1% 14.0% 1.0% 100% Source: LFS 2017. Note: KPK = Khyber Pakhtunkhwa. Among within province migrants, a significant share of workers moves into urban areas (Figure 81). This is clear not only in Sindh, where urbanization is the highest in the country, but also in less-urbanized provinces such as Balochistan. There is a significant increase in the population of each capital district between the 1998 and 2017 censuses (Figure 82). The annual growth rate in these districts far exceeded overall population growth (2.4 percent per year) in the period. Significant growth rates are noticeable in Quetta, Islamabad, and Peshawar, which all started from a relatively low base. Figure 81 Figure 82 Increase in the prevalence of internal migration by destination province, Population growth in capital districts by number and percentage, between 1999–2017 1998 and 2017 SHARE OF MOVES TO URBAN AREAS POPULATION GROWTH IN CAPITAL DISTRICTS AMONG WITHIN-PROVINCE MIGRANTS 12 7.0% 90% 5.8% 80% 10 6.0% POPULATION (IN MILLIONS) 4.9% 70% 5.0% 8 4.0% 60% 4.0% 6 3.0% 50% 2.4% 3.0% 40% 4 2.0% 30% 2 1.0% 20% 10% 0 0 ISLAMABAD LAHORE PESHAWAR QUETTA KARACHI 0 KHYBER PUNJAB SINDH BALOCHISTAN PAKHTUNKHWA 1998 2017 CAGR (1998-2017) RHS Source: LFS 1999, 2017. Source: Population census 1998, 2017. 51 Figure 83 Figure 84 Prevalence of internal migration by education level, 2017 Average years of schooling by migrant status and pattern of migration, 2017 PREVALENCE OF INTERNAL MIGRATION BY EDUCATION LEVEL 10 9.3 10% 9.8% 9 9% AVERAGE YEARS OF SCHOOLING 8 7.6 8% 7 6.7% 6.9% 6.1 7% 6.3% 6 6% 5.8% 5.3 5 5% 4 3.9 4% 3.4 3 3% 2 2% 1 1% 0 0 STAYEE STAYEE MIGRANT MIGRANT MIGRANT MIGRANT NO EDUCATION PRIMARY LOWER UPPER POST- RURAL URBAN RURAL- RURAL- URBAN- URBAN- SECONDARY SECONDARY SECONDARY RURAL URBAN URBAN RURAL Source: LFS 2017. Education is positively associated with the likelihood of internal migration. Migration is generally more common among those with a postsecondary education (Figure 83). Migrants who move from rural to urban areas have significantly more years of schooling (6.1 years) than those who stay in rural areas (3.9 years) or those who move from rural to rural areas (3.4 years) (Figure 84). Relatedly, internal migrants tend to have better labor market outcomes, as reflected in their status and sector of employment as well as earnings. For instance, compared with those who stay in rural areas, migrants from rural to urban areas are more likely to work in wage employment and nonagricultural self-employment, and less likely to be nonpaid workers (Figure 85). Even those migrants who move from one rural area to another are better off in that they are significantly less likely to work in agriculture and as nonpaid workers—although many find themselves in informal wage employment. In terms of the sector of employment, compared with those who stay in rural areas, people who move to other rural areas or to urban areas are much more likely to be in the nonagricultural sector (Figure 86). Migrating to urban areas is positively associated with formal employment and better earnings among wage workers. While only 20 percent of people who stay behind Figure 85 Figure 86 Employment type by migrant status and pattern of migration, 2017 Sector of employment by migrant status and pattern of migration, 2017 100% 100% 90% 90% 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0 0 STAYEE STAYEE MIGRANT MIGRANT MIGRANT MIGRANT STAYEE STAYEE MIGRANT MIGRANT MIGRANT MIGRANT RURAL URBAN RURAL- RURAL- URBAN- URBAN- RURAL URBAN RURAL- RURAL- URBAN- URBAN- RURAL URBAN URBAN RURAL RURAL URBAN URBAN RURAL WAGE, FORMAL UNPAID FAMILY WORKER SELF, NON-AGRICULTURAL WAGE, INFORMAL SELF, AGRICULTURAL AGRICULTURE INDUSTRY SERVICES Source: LFS 2017. 52 in rural areas have a written contract for formal employment, about 25 percent of rural-to-urban migrants do. Moreover, urban-to-urban migrants receive higher wages by over 30 percent than people who stay in the same urban center. Similarly, rural-to-urban migrants’s wages are higher by 6 percent than those of workers who stay in the same rural areas. These substantial wage differentials suggest potentially large returns to urban migration. Urban migration, by attracting better educated people and promoting non-agricultural employment, can be a catalyst for structural transformation and a growing middle class. Productivity increases through the agglomeration of people and enterprises in closer locations to one another and spurring job creation in manufacturing and services are considered as key trait of urbanization (see Ellis and Roberts 2016). Political economy of development also suggests that a growing income share among the middle class can facilitate a higher level of income and growth as well as public goods (see Easterly, 2000). Despite the generally positive effects in terms of labor market opportunities, there are a few risks associated with urbanization that could potentially reduce the benefits of urban migration. Some of these potential risks include expansion of the informal economy; poor-quality jobs among migrants; higher cost of living; increased burden on already weak public services such as housing, water, sanitation, electricity, and social services (e.g., health and education); and congestion leading to increased traffic and average commute time (SPDC 2016). The unmet housing demand is significant where Pakistani cities are characterized by a large share of informal settlements (known as katchi abadis; see Box 6). The overcrowding of poor-quality housing in marginal areas perpetuates the marginalization of the poor, and residents of certain neighborhoods are prone to climate hazards and general discrimination (Collective for Social Science Research 2017). BOX 6: KATCHI ABADIS AND SLUMS IN PAKISTAN Urban slums are characterized by poverty and challenging living conditions. Many residents are recent migrants. Literature in Pakistan uses the terms “slum,” “katchi abadi,” and “squatter settlement” interchangeably, although the definitions of katchi abadis and slums, as recognized by the Government of Pakistan, are slightly different. Both terms are used to describe unserviced or underserviced settlements. Katchi abadi is a popular term used to denote low-income as well as deprived settlements that may or may not be legal. Approved katchi abadis are earmarked for regularization through a 99-year lease and local government infrastructure development against a payment to the state. Historically, katchi abadis are populated by the working poor and are an element of urban sprawl. Recent urbanization has kept them alive. Typical katchi abadis are characterized by insecure tenure, irregular or informal settlements, lack of basic services, substandard housing or illegal and inadequate building structures, overcrowding and high population density, unhealthy living conditions and hazardous locations, and poverty and social exclusion of minorities. Source: Based on UN-HABITAT (2003). 53 6. OVERSEAS LABOR MARKET  50 6.1. TRENDS IN OUTMIGRATION FROM PAKISTAN The number of Pakistani labor migrants seeking earnings opportunities abroad is significant, making remittances sent home crucial to the economy. The number of Pakistanis legally outmigrating each year51 has risen significantly in the past decade, peaking at 900,000 in 2015 (Figure 87). Pakistan receives a large amount of remittances, which is a significant part of the economy. In 2015–16, Pakistan registered nearly US$20 billion (equivalent to 7 percent of GDP). Studies find evidence that remittances make a positive contribution to economic growth and poverty reduction. In Pakistan, positive links have been established between remittances and social outcomes such as children’s increased enrollment in school and reductions in child labor, particularly of girls.52 At the province level, in Khyber Pakhtunkhwa (KPK), the origin of many migrants, remittances have played an important role in providing food and income security as well as rebuilding destroyed homes and basic infrastructure affected by conflict (Gioli, Khan, and Scheffran 2013). The strongest incentive for workers to seek overseas opportunities must be the higher level of earnings. International migrants earn much higher wages than they did prior to migration. A migration survey Figure 87 Number of outmigrants and remittances as share of GDP, 2006–17 1,000,000 8 900,000 7 800,000 6 700,000 5 600,000 % OF GDP 500,000 4 400,000 3 300,000 2 200,000 1 100,000 0 0 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 TOTAL MIGRANTS REMITTANCE AS SHARE OF GDP Sources: BEOE (migrant volume) and WDI (remittances). 50 This chapter is based on Cho et al. (2018). 51 The United Nations Office on Drugs and Crime (UNODC 2012) estimates that up to 300,000 people have been leaving Pakistan each year though irregular channels. 52 See Mansuri (2006); Javid, Arif, and Qayyum (2012); and Arif (2010). These studies complement findings from around the world, including Indonesia (Adams and Cuecuecha 2010), Philippines (Yang and Choi 2007), Latin America and the Caribbean (Acosta et al. 2006), and Nepal (Lokshin, Bontch-Osmolovski, and Glinskaya 2010) on the poverty-reducing impact of remittances, and studies in the Philippines (Yang 2004) and Ecuador (Ponce, Olivié, and Onofa 2011), which have seen positive impacts on education and health outcomes. 55 Figure 88 Differences in monthly earnings (US$), 2015 700 623 600 500 482 437 448 395 400 370 US$ 300 239 200 157 156 139 151 130 100 0 NO EDUCATION PRIMARY SECONDARY SECONDARY POST-SECONDARY ALL LEVELS INCOMPLETE COMPLETE PRIOR TO DEPARTURE AT DESTINATION Source: KNOMAD Migration Cost Survey 2015. in 201553 shows that Pakistani migrants earn on average three times as much as what they earned at home, irrespective of the level of their educational attainment (Figure 88). Reflecting these benefits, a significant share of workers desire to seek overseas employment. The LSS 2015 reports that about 13 percent of respondents had intended to migrate abroad. It is notable that those with a relatively high level of educational attainment tended to be more likely to report their desire to pursue overseas employment (Figure 89). This suggests that the perceived returns to migration may be larger for educated workers. In addition, those with social networks (friends and relatives) already working abroad are also more likely to report the intention to migrate. The intention to migrate was strongest among workers from urban areas, especially in the KPK province (Figure 90).54 Figure 89 Figure 90 Share of those intending to migrate abroad across various characteristics Likelihood of reporting an intention to migrate abroad across various characteristics MALES, AGE 15–64 KHYBER RURAL PAKHTUNKHWA AGE*10 URBAN IS MARRIED RURAL SINDH URBAN PRIMARY EDUCATION RURAL LOWER SECONDARY PUNJAB URBAN UPPER SECONDARY AND ABOVE SOCIAL YES FRIENDS/RELATIVES ABROAD LIVES IN AN URBAN AREA NETWORK NO FRIENDS/RELATIVES ABROAD KHYBER PAKHTUNKHWA PROVINCE UPPER SECONDARY AND ABOVE SINDH PROVINCE LOWER SECONDARY EDUCATION PRIMARY KHYBER PAKHTUNKHWA*URBAN NO EDUCATION HAS FRIENDS/RELATIVES ABROAD 0 10% 20% 30% 40% –.3 –.2 –.1 0 .1 .2 Source: LSS 2015. 53 The survey was conducted by the Global Knowledge Partnership on Migration and Development (KNOMAD) and referred to as the KNOMAD Migration Cost Survey. 54 Another specification was run that also controlled for the household wealth quintile, where wealth was proxied by an asset index, but wealth quintile dummies had no significant impact on the desire to migrate abroad. 56 Figure 91 Figure 92 Share of outmigrants by destination country, 2005–17 Share of common occupations of Pakistani migrants, 2010–17 100% 100% 90% 90% SHARE OF MIGRANTS BY DESTINATION COUNTRY 80% 80% 70% 70% 60% 60% 50% 50% 40% 40% 30% 30% 20% 20% 10% 10% 0 0 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2010 2011 2012 2013 2014 2015 2016 2017 UNITED ARAB EMIRATES OMAN EAST ASIAN COUNTRIES LABORER DRIVER MASON TECHNICIAN CARPENTER SAUDI ARABIA OTHER GULF COUNTRIES OTHERS ELECTRICIAN SALESMAN AGRICULTURE MECHANIC OTHERS Source: BEOE. The opportunities for international migration, however, are not evenly distributed among workers. Punjab and KPK together account for 92 percent of the flow of outmigrants in 2016, while their share of Pakistan’s total working-age population (aged 15–64) is 72  percent. In particular, the share of migrants from KPK, at 35 percent, far exceeds the province’s share of the employed population, at 11 percent. This indicates that more than 5 percent of KPK’s active labor force leaves the province each year to work abroad. Large emigration waves and consequential remittances may have contributed to significant poverty reduction in KPK. Also, women form a negligible share of labor migrants from Pakistan, representing less than 1 percent of total outmigrants. This is extremely low in absolute value terms as well as when compared to other migrant-sending countries. For instance, Bangladesh experienced a sharp increase in the share of female migrants after a bilateral agreement with Saudi Arabia on female domestic work, and the lifting of a ban on the work of women abroad. As of 2016, the female share of outmigrants in Bangladesh was about 16 percent. One obvious driver for the low female share of migrants in Pakistan is the fact that female labor participation is very low in the first place. The Pakistan Emigration Rules also restrict women’s migration to those who are 35 and above (with a 5-year age exception granted in some cases). While overseas incomes of migrant workers are a great resource, overconcentration in specific destination countries and occupational sectors poses a risk to Pakistan. Most Pakistani labor migrants have headed to either Saudi Arabia or the United Arab Emirates in the past decade, exposing them to macroeconomic shocks as seen in recent years (Figure 91). Most workers are employed in occupations classified as unskilled and less skilled, with over 60 percent working as laborers, drivers, or masons (Figure 92). Therefore, the occupational concentration of Pakistani migrants is also risky given that labor is replaceable without specialized skills, and hence, is largely exposed to competition with other low-skilled migrants. 6.2. A MIGRANT’S JOURNEY The journey of a migrant can be divided into five broad phases: pre-decision, recruitment, pre-departure, in-service (i.e., in overseas employment), and return (Figure 93). Migration is a complex process that involves a series of decision-making points at each stage, and various stakeholders come into play. Also, market failures leave migrants exposed to diverse types of risks and uncertainties, leading to varied outcomes in their migration experience. Understanding the choices they face at each stage of the process, their perceived risks and benefits, as well as potential market failures that influence their motivations and decision making, is necessary. 57 Figure 93 A migrant’s journey PRE-DECISION RECRUITMENT PROCESS PRE-DEPARTURE IN-SERVICE RETURN TAKEN BY MIGRANTS ACTIONS/DECISIONS DECISION TO MODE OF PRE-DEPARTURE PAPERWORK MEDICAL TESTS REMITTANCE REINTEGRATION MIGRATE RECRUITMENT/VISA TRAINING WORKING SKILLS RECRUITMENT COSTS CONDITIONS NETWORKS, SOCIAL NETWORKS, OEPs, EMPLOYER, LABOR INVOLVED ACTIONS OEPs, AGENTS, PROTECTORATE, MEDICAL CENTER, TRAINING PROTECTORATE, ATTACHE, DIASPORA RETURNEE AGENTS, CENTER, TTC, FINANCIAL INSTITUTIONS, MONEY OEPs, MRC NETWORK, FINANCIAL NETWORK MEDIA, MRC LENDERS INSTITUTIONS Source: Authors’ compilation. Note: MRC = migrant resource center; OEPs = overseas employment promoters; TTC = trade testing center. The migration institutions and governance architecture (in Appendix C) are designed to address various risks and vulnerabilities along the migration journey although implementation remains weak. Pakistani migrants are often unable to reap the benefits of migration given that they are constrained by costs, both monetary and nonmonetary (e.g., opportunity, information, and psychological costs), as well as by key market and government failures that include: ∫ Information asymmetry. Employers and particularly recruiters gather far more information about workers than workers can learn about potential employers, the nature of potential jobs, or the process of migration itself. ∫ Lack of regulations for intermediaries. Adequate oversight is not provided to regulate various intermediaries in the recruitment market, and unlicensed brokers often charge exorbitant rates and fees to migrant workers. ∫ Lack of access to credit. Without access to formal channels of credit, many migrants borrow informally at high interest rates to cover recruitment costs. ∫ Unequal bargaining power between employers and migrant workers (e.g., under the Kafala, or “sponsorship,” system used in several Gulf Cooperation Council [GCC] countries) leaves migrants vulnerable to physical abuse and contract infringement without adequate channels of dispute arbitration. It is critical to assess these vulnerabilities along the journey, and address gaps that limit the realization of the full benefits of international migration and their positive spillover impacts on the local economy. Pre-decision An aspirant migrant weighs the expected costs and benefits associated with migration and compares various alternative options, making information access a necessity at this stage. Focus group discussions (FGDs) with outgoing migrants in Pakistan mentioned various sources from which they acquired information on overseas opportunities and life abroad. These include: social networks both in the country of origin as well as in destination countries; Overseas Employment Promoters (OEPs) and subagents; newspapers and social media, and, finally, migrant resource centers. The KNOMAD Migration Cost Survey also suggests that employment information through relatives or friends is higher in Pakistan than other comparator countries except India (see Figure 94). However, social networks and intermediaries may not be the most reliable sources of information as such informal sharing can mislead migrants’ decisions. In an environment with high search and information 58 Figure 94 Main sources of information used to find a job in a foreign destination by migrants’ country of origin, 2015 2 2 0 2 2 100% 4 4 12 1 13 90% 19 32 80% 38 25 51 30 70% 66 29 60% 25 8 50% 40% 15 30% 62 56 49 49 20% 41 32 31 10% 0 PAKISTAN ETHIOPIA INDIA NEPAL PHILIPPINES VIETNAM TOTAL RELATIVE/FRIEND INDIVIDUAL RECRUITER/BROKER MANPOWER AGENCY OTHERS Source: KNOMAD Migration Cost Survey 2015. costs, OEPs and particularly unlicensed subagents can monopolize or withhold information in such a way as to support their profit motives and not necessarily the migrants’ best interests. The very reason they exist is because of information asymmetry: both workers and employers have very limited information about each other, thus leaving a space for intermediaries to intervene and engage in rent-seeking behavior. Recruitment Pakistani migrants emigrate through different channels and thus use different visas; information received at the pre-decision stage shapes the route used. Special attention should be given to a category of work arrangement referred to informally as the Azad or “freedom” visa, often arranged by social networks or subagents. The attractiveness of the Azad visa is that it is seemingly not tied to one employer; instead, the Kafeel (sponsor) provides the visa to a migrant for a certain monthly fee, connected with a broad set of employment options. In reality, the visa is grossly misunderstood and visa holders are still tied to a Kafeel. If employment is not mentioned in the Iqama card (work license) provided by the Kafeel, they are considered illegal and face imprisonment/deportation. In large part due to the dominance of unlicensed subagents and the popularity of the Azad visa, Pakistanis pay some of the highest recruitment costs in the region (Figure 95). The KNOMAD Migration Cost Survey shows that Pakistani workers pay more than US$3,500 on average to migrate to Gulf countries, of which an average of US$2,500 is covered by loans. Most of these costs are associated with visa fees, at US$2,900 on average. Such high costs are equivalent to over 10 months of wages in Saudi Arabia (Figure 96). This is far higher than the official cost of a visa to Saudi Arabia or the United Arab Emirates (below US$100). Apart from migration costs, the skills and productivity of migrant workers require greater policy attention. Skills certificates can help signal workers’ productivity to employers, and thus contribute to improving employment quality and earnings. Outflows of skilled workers can lead to higher remittances, which can further spur human capital accumulation and provide improved incentives for skill formation in the home country (Lucas 2004). Furthermore, the return of skilled migrants can facilitate the exchange of know-how and skills between the origin and destination countries (Alvarez and Barney 2014). 59 Figure 95 Figure 96 Average migration costs (US$) by sending country and funding source, 2015 Average migration costs in months of earnings by origin and destination, 2015 4,000 IN MONTHS OF EARNINGS BY DESTINATION 12 3,500 10 3,000 8 2,500 6 USD 2,000 4 1,500 2 1 MONTH OF EARNINGS AT DESTINATION 1,000 0 PAKISTAN ETHIOPIA INDIA PHILIPPINES NEPAL BANGLADESH INDIA SRI LANKA PAKISTAN BULGARIA ECUADOR MOROCCO INDONESIA THAILAND VIETNAM VIETNAM 500 0 PAKISTAN ETHIOPIA INDIA NEPAL PHILIPPINES VIETNAM SAUDI QATAR KUWAIT UAE SPAIN KOREA MALAYSIA AMOUNT BORROWED OWN FUNDS ARABIA Source: KNOMAD Migration Cost Survey 2015. Pre-departure Upon completion of required paperwork and prior to departure, a migrant is generally expected to go through mandatory pre-departure training (Section 15 of the Emigration Ordinance, 1979). While global evidence on the effectiveness of pre-departure training is weak (McKenzie and Yang 2015), it can serve to advance a migrant’s familiarity with labor and immigration laws and rights, as it is important for the migration system to respond to the evolving landscape. The mandatory training, however, is not as useful or accessible to migrant workers as it could be. FGDs suggest that it is not universally enforced, in part due to protectorates’ lack of capacity. Amid the limitations of the mandatory predeparture training, OEPs and even unlicensed intermediaries provide basic orientation to migrants on information regarding life abroad (e.g., rules, employment, resources available in times of difficulty, culture, finances). During migration (in service) Two elements of this phase—worker protection and remittances—are among policy priorities. Measures to protect workers are often difficult to implement unilaterally from the sending side, especially when migrants are in the jurisdiction of the receiving country. Key players at the destination that are providing worker protection services include embassies and community welfare attachés (formal) and diaspora networks (informal). However, FGDs revealed that many workers view embassies/attachés merely as providers of administrative support such as renewing passports rather than as reliable resources for worker protection. Common topics of dispute are working hours, wages, work-related injury or illness, and workers’ rights (KNOMAD 2015). Destination countries are also introducing modest policy innovations to protect workers, and these should be further advocated for. For instance, the Wage Protection System (Jureidini 2017), which is being implemented in all GCC countries except Bahrain mandates wage payments by bank transfer to provide an official record that can be monitored, with penalties for noncompliance. Such changes can positively impact migrant welfare. A 2011 rule that allowed workers to legally switch jobs without permission from an employer in the United Arab Emirates led to increased earnings among incumbent migrants (Naidu, Nyarko, and Wang 2014). Information on these options and programs could be further disseminated through predeparture training and other avenues of information sharing. Remittances are sent via multiple channels by migrants, and Pakistan has made good progress in formalizing remittances through the Pakistan Remittance Initiative (PRI). The initiative was launched to facilitate the faster, cheaper, more convenient, and more efficient flow of remittances, and to create investment opportunities in Pakistan for overseas Pakistanis. The PRI encouraged banks in Pakistan to enhance their outreach 60 Figure 97 Figure 98 Association between duration of stay (in months) and cost of migration Association between duration of stay (in months) and monthly earnings (in US$) (in US$) at destination among Pakistani migrants 50 1,000 DURATION OF STAY AT DESTINATION 40 800 MONTHLY EARNINGS (USD) 30 600 20 400 10 200 0 0 0 5,000 10,000 15,000 0 10 20 30 40 50 TOTAL MIGRATION COST DURATION OF STAY AT DESTINATION Source: KNOMAD Migration Cost Survey 2015. by entering into bilateral arrangements with overseas financial institutions (as a result of which 600 bilateral agreements were negotiated) (Pakistan Remittance Initiative 2015). Return Return migration represents an important step in the migration process as a large share of outmigration, especially to GCC countries, is typically governed by contracts of two to three years. Migrants come back to Pakistan either voluntarily (upon completion of contract) or involuntarily (deportation, layoff), and return temporarily (before remigrating) or permanently. As temporary migrants are not eligible to apply for citizenship in GCC countries, migrants eventually return, but the duration abroad varies. Migrants who have paid higher costs to migrate tend to stay longer in destination countries to recoup the costs (Figure 97). Also, the duration of stay is longer for those who earn more (Figure 98). The reintegration of former migrants into the local labor market is considered an important policy priority given their accumulated human capital. It is well recognized that there is generally a positive selection for migration, with those who have migrated and returned (former migrants) being more likely to be educated and to have received training, compared with those who have never migrated (nonmigrants) (Figure 99). As discussed above, migrants from KPK (compared with other provinces) and rural areas (compared with urban) are overrepresented in the composition of those who have migrated and returned. Employment outcomes show that, compared with nonmigrants, former migrants are significantly more likely to be self-employed, but less likely to be in informal wage employment (Figure 100). One likely reason for the high probability of self-employment upon return from overseas employment is an accumulation of savings and capital that allows returning migrants to start entrepreneurial activities (see Mesnard 2004; Dustmann and Kirchkamp 2002; McCormick and Wahba 2001). Dustmann and Kirchkamp (2002) emphasize that savings may be put into productive uses and that the entrepreneurial activities of returning migrants may contribute to wealth generation, and create jobs in economies where capital constraints are major bottlenecks for starting an enterprise. At the same time, however, former migrants are significantly more likely to be unemployed (Figure 101). Migrants who had returned within a year before the survey had a particularly high unemployment rate of close to 30 percent. The difference in unemployment rates was far greater in urban than in rural areas and also remains highly significant after controlling for workers’ observable characteristics such as experience and education. Policies for return migration have been increasingly important in Pakistan given the significant increase in the volume of return migrants recently due to mass layoffs and nationalization policies in popular destination countries. 61 Figure 99 Characteristics of nonmigrants vs. former migrants among working-age men DIFFERENT CHARACTERISTICS BETWEEN NON-MIGRANTS VS. FORMER MIGRANTS 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0 MARRIED RECEIVED NO PRIMARY LOWER UPPER POST- PUNJAB SINDH KHYBER BALOCHISTAN RURAL SKILLS EDUCATION EDUCATION SECONDARY SECONDARY SECONDARY PAKHTUNKHWA TRAINING EDUCATION EDUCATION EDUCATION THOSE WHO HAVE NEVER MIGRATED (NON-MIGRANTS) THOSE WHO HAVE MIGRATED AND RETURNED (FORMER MIGRANTS) Source: LFS 2014. Note: The differences are all statistically significant at least at a 5 percent level. Figure 100 Figure 101 Shares of rural and urban residents by prior migration and employment status Shares of unemployed rural and urban residents by migration status 100 35% 90 30% 80 70 25% 60 50 20% 40 15% 30 20 10% 10 5% 0 NON-MIGRANTS RETURN NON-MIGRANTS RETURN MIGRANTS MIGRANTS 0 RURAL AREAS URBAN AREAS RURAL AREAS URBAN AREAS ALL WAGE, FORMAL UNPAID FAMILY WORKER ALL RETURNING MIGRANTS RETURNED LESS THAN ONE YEAR AGO WAGE, INFORMAL SELF-EMPLOYED RETURNED MORE THAN A YEAR AGO NON-MIGRANTS Source: LFS 2014. 62 7. CONCLUSIONS AND POLICY DIRECTION 7.1. BRIEF SUMMARY OF MAIN FINDINGS Pakistan’s labor market is young, offering an opportunity to reap substantial demographic dividends. The working-age population is growing faster than are dependents, meaning that more breadwinners are potentially available for the economy. Among the working-age population, the share of youth (aged 15–24) is high, at almost 35 percent, and every year a large number of young individuals (around 2 million) enter the labor force. New entrants tend to be better educated than their predecessors, seek more modern types of employment in nonagricultural sectors, and are mobile within and across borders. With such a demographic boon, Pakistan’s labor market has the potential to be increasingly dynamic. Pakistan has been moving in the right direction in terms of educational achievement levels and a structural transformation process. Labor market transitions from agriculture to the service and industry sectors, from nonwage to wage employment, and from rural to urban economies, are all taking place although the pace of such transitions varies. Improved education in part explains these trends, given the strong association between education and labor market outcomes. The overall level of education has been increasing and so has the labor productivity. The gender gap in education has been narrowing, and women’s labor force participation was slowly improving until a recent slip in 2014–17. A digital job sector has started emerging, with good potential to provide an engine of growth. An increasing number of workers pursue overseas employment and their remittances have provided huge incomes flows to the economy. This progress has been made, albeit gradually, despite volatile macroeconomic conditions. However, many challenges remain. First and foremost is that Pakistan needs more workers to engage in productive activities and support the economy’s growth. In 2017, the share of the employed (57 million) out of the total population in the country (208 million) was only 27 percent. This means that one breadwinner provides for 3.6 individuals in Pakistan. This is significantly higher compared with other countries: 2.7 people in Bangladesh and 1.9 people in Vietnam per breadwinner. This is because almost half of the working-age population is out of the labor force, which is due to a very low FLFP rate. For the market to absorb more workers now out of the labor force, more job opportunities need to be created. During this period of 2000–17, the number of jobs managed to grow at pace with the number of workers in the labor force—mostly men. Indeed, the employment ratios and working hours of working-age men are very high, and unemployment rates have remained low, despite some fluctuations in an economic cycle of boom (2000–05), bust (2006–11), and recovery (2012–17).55 For instance, in 2017, the labor force participation rate of working-age men was over 80 percent, and weekly working hours exceeded 50 hours on average. The rate of employment growth with respect to economic growth56 is higher in Pakistan than in many other countries, which corroborates the finding that the economy has been absorbing the labor force as much as it can for those who cannot afford not to work. This indicates that there is little room for the labor supply of already working men to increase, and that more jobs are needed to induce potential workers (now out-of- the-labor-force) to engage in employment. 55 Pakistan has seen frequent boom-bust cycles without sustained, continued growth. For more discussion, see World Bank (2019a). 56 This is often referred to as employment intensity or the elasticity of growth. 63 The quality of jobs,57 reflecting labor productivity, remains a significant challenge amid limited structural transformation.58 Structural transformation is characterized by labor moving from lower to higher productivity activities, typically from agriculture to industry and services, and from nonwage to wage employment. At the aggregate level, compared with 2000, in 2017 the labor market was less agriculture based and featured more wage employment.59 However, the pace of this change has been very slow, and Pakistan’s labor productivity growth rate has been far lower than that of comparator countries (e.g., Bangladesh, India, Sri Lanka, Indonesia, and Vietnam) during the same period. This is in part due to the economy’s frequent boom-bust cycles, depending on which the pace of quality job creation has varied. The share of industry and services increased during the boom, leading to a significant decrease in the share of agricultural employment. During the economic bust that followed, the progress was reversed, as many workers relied on agriculture while industry and services added fewer employment opportunities. Then, as the economy slowly recovered, the industry sector gained momentum in adding jobs, while the share of jobs in the services sector fell further. Similarly, the growth of wage employment (vis-à-vis nonwage employment) was also procyclical. If the economy had grown steadily without much fluctuation, the overall change in the quality of jobs could have been larger. On average, over 40 percent of workers still find jobs in agriculture and 45 percent in nonwage employment. Informality is prevalent in Pakistan and is often considered as a source of low-quality jobs.60 Agricultural jobs; nonfarm, nonagricultural household enterprises (self-employment); and wage employment without a written contract all compose informal jobs. In total, roughly 80  percent of urban workers and more than 90 percent of rural workers are classified as having informal jobs in 2017. Many of these informal workers in the labor market engage in earnings-generating activities only to find that their incomes are not sufficient to lift them and their families out of poverty, and improving the productivity and earnings of these workers is considered to be the best way to reduce poverty. However, it should be noted that the majority of jobs of the working poor are informal, but not all informal sector jobs have low productivity. Indeed, the productivity of the nonfarm, nonagricultural enterprises varies: some better-educated owners in urban areas tend to hire workers and earn as much as formal wage workers (with a written contract), but others, especially in rural areas, remain as own-account workers without any paid employees or laborers and their earnings are significantly lower. People can gain access to better-quality jobs by migrating. In Pakistan, spatial mobility is characterized by rural-to-urban and international migration. With respect to urban migration, the movement of working-age people within and across provinces is common. Focusing on migration within the past 10 years and excluding those returning from abroad, about 70 percent of internal migration was within the same province and the other 30 percent was across provinces. Although not all internal migration is from rural to urban areas, the large majority of within-province movement was to urban destinations. In particular, there was a significant increase in the population of each capital district between the 1998 and 2017 censuses. There is suggestive evidence that internal migrants have better labor market outcomes (e.g., more wage employment, less of it in agriculture, and for higher earnings). Similarly, a significant number of individuals from Pakistan pursue overseas employment for better-quality jobs. Overseas employment provides higher earnings than domestic, and remittances—equivalent to 7 percent of GDP in 2017—are a critical source of funds as foreign direct investment is limited and foreign reserves are low. International migration is particularly common in the province of Khyber Pakhtunkhwa. There are huge variations in how workers access quality jobs, depending on their characteristics and location. Differences across the male/female and urban/rural divides are among the most notable. The FLFP rate 57 The quality of jobs is assessed by the employment sector and status as well as earnings. Given that earnings information is available only for wage workers and labor productivity is measured mostly in aggregate, the sector and status of employment are commonly used criteria for measuring the quality of jobs. 58 See Sharma, Nayyar, and Kim (2019) for more discussion of structural transformation in Pakistan. 59 The share of agriculture declined from 48 percent in 2000 to 43 percent in 2017; industry from 17 to 26 percent; and services from 34 to 31 percent. Meanwhile, the share of wage employment increased from 37 percent in 2000 to 44 percent in 2017. 60 Definitions and measures of informality vary depending on whether the focus is on enterprises or workers. See discussion in Appendix B. 64 has remained stubbornly low. In addition, the quality and variety of jobs available to women tend to be limited, and employed women are found almost exclusively in agriculture (78 percent) and manufacturing (19 percent), with a large share of working women being unpaid and underemployed. Along with gender, educational attainment levels and age are important determinants of access to quality jobs. The share of youth (ages 15–24) who are in the category of NEET is high, showing large variations across male/female and urban/rural divides. Moreover, as the employment structure differs widely between urban and rural areas, so does the quality of jobs. The heterogeneity of labor market outcomes across different types of workers is further compounded by large variations across provinces. To reap the benefits of a young labor force, Pakistan would have to simultaneously focus on quantity, quality, and inclusivity of jobs. Moving forward, the economy has to create sufficient jobs to absorb the large number of working-age people, engage more workers in productive activities, and make quality jobs accessible to everyone. 7.2. INITIAL POLICY DIRECTION To address the key challenges of quantity, quality, and accessibility of jobs, Pakistan can use targeted interventions that (i) strengthen the fundamentals of job creation, (ii) boost productivity, and (iii) promote inclusiveness of the labor market. Each requires a multisectoral approach and can go beyond the scope of traditional job policies. Policies to strengthen fundamentals of job creation While the private sector creates jobs, the government can adopt measures to make the economic environment more conductive to quality job creation in line with strategies for growth. ∫ Ensure macroeconomic stability and move away from long-standing economic fluctuations. Stable growth helps create quality jobs while busts easily reverse the progress made. Short-term stabilizers addressing the issues of current account imbalance, depleted foreign exchange reserves, and lack of confidence in the economy help avoid deeper recessions.61 At the same time, measures to ensure fundamental stability (e.g., tax policies to improve revenue and remove distortionary subsidies, reforms for public utilities and energy sectors to rationalize public spending, an exchange rate policy to strengthen the export sector) should continue to be upheld even after short-term crises are avoided. There is renewed emphasis on these policies in light of the recent, near-crisis macroeconomic conditions of 2019.62 ∫ Improve the business environment. While macroeconomic conditions affect the overall economic performance of a country, the business environment has a direct bearing on the private sector’s ability to start and operate business, and thus create jobs. The World Bank’s Doing Business ranked Pakistan 136th in 2019 and 108th out of 190 economies in the latest 2020 exercise, and highlighted areas for improvement. These include streamlining processes and regulations that influence the accumulation of firms’ physical capital (e.g., dealing with construction permits, registering property, paying taxes), as well as areas that are critical to firms’ operations (e.g., getting electricity and enforcing contracts). The business environment can be significantly improved by a strong government commitment to policy reforms and their prioritization, and to the adoption of innovative technology.63 ∫ Foster export competitiveness with initiatives that help reduce production and information cost. For this purpose, the government can promote a gradual reduction in tariffs on both intermediate and final goods. Reduced tariffs on intermediate goods helps reduce the production cost for firms downstream by allowing them to access cheaper, better quality, and more varied intermediate inputs. This makes their products more competitive in the global market. In the meantime, reducing tariffs on final products will bring 61 Short-term adjustment measures include increasing the interest rate, raising energy prices, allowing currency devaluation, tightening fiscal spending, and mobilizing resources from financial markets and international financial institutions (e.g., the International Monetary Fund). 62 The recent macroeconomic conditions led to a negotiation with the International Monetary Fund for a support program. See Ahmed (2018) and World Bank (2019a, 2019b) for further discussion of policy measures for macroeconomic stability. 63 For case studies that include several success stories, see: https://www.doingbusiness.org/en/reports/case-studies/view-all. 65 in more competition and will spur efficiency gains by helping more productive firms to grow. At the same time, exports can be promoted by reducing information costs for new exporters in finding new markets or introducing new products. In addition to creating an enabling environment, efforts are needed to increase and upgrade inputs for production—especially physical and human capital—to strengthen the fundamentals. ∫ Mobilize resources to increase investment in physical capital. Pakistan lags in capital investment—as measured by the rate of fixed capital formation and foreign direct investment (FDI). Weak financial markets and limited domestic savings in part explain the lack of availability of resources and difficulty of firms in getting access to credit. Fiscal space for public investment is also limited due to very limited tax revenues.64 Thus, resource mobilization through tax policies and drives for financial inclusion in the domestic market are required (see World Bank [2019a] and the National Financial Inclusion Strategy 2015). Also, facilitating FDI and remittances from the international market is critical, and a more systematic management of international migration and facilitation of remittances transactions would be needed.65 ∫ Increase public financing in human capital investment while getting the priorities right. It is worth highlighting that investing in human capital and building a strong foundation for labor productivity are critical. As the World Bank’s Human Capital Index and policy discussions around it have indicated, the highest priority goes to starting early in the life cycle for effective human capital accumulation (Ahmed, Cho, and Fasih 2019). Given that the human capital indicators of vulnerable individuals are significantly lagging behind, targeted support for low-income families and girls in particular should be prioritized. A recently announced policy framework, Ehsaas (meaning “compassion”), and corresponding provincial investments are a welcome approach.66 Policies to boost productivity Given the heterogeneity in the productivity of informal firms and workers, there is room for intervention to make operating in the formal sector and pursuing growth more attractive and easier. The obligations associated with formality—such as paying taxes, contributing to social insurance benefits, and adhering to labor regulations—are often recognized as deterrents to higher levels of formality, which leads to deregulation and lower tax wedges as prescriptions to increasing formality.67 At the same time, efforts to lower the barrier for informal enterprises to become formal, such as simplifying business registration, are recommended.68 In addition to such efforts as part of policies to improve the business environment, the following interventions hold promise to provide incentives for formality. ∫ Provide business development services for local microenterprises to address their challenges while granting pathways toward formality. Lack of access to credit and business management capacity are two main constraints to firm growth.69 The majority of nonfarm, nonagricultural household enterprises in Pakistan report being primarily set up using individual capital, household income, or asset sales (not through loan finance). They tend to be local without being connected to broader value chains. These businesses are less likely to pursue innovation or risk taking that can potentially lead to growth. Providing access to credit and opportunities for managerial capacity building, tied with some efforts to encourage owners to formalize their operations, could make a difference in enhancing firm productivity. 64 Pakistan collects less than 13 percent of GDP in tax revenue, with a very narrow tax base. 65 In 2017, remittances were equivalent to 7 percent of GDP, whereas foreign direct investment was around 1 percent of GDP. 66 http://www.pakistan.gov.pk/ehsaas-program.html. 67 See reviews from di Porto, Elia, and Tealdi (2017); Packard, Koettl, and Montenegro (2012); Feld and Schneider (2010); and Maloney (2004), among others. 68 Multiple studies have examined the impact of policy efforts to formalize informal firms on the productivity, performance, and formality of firms. See de Mel, McKenzie, and Woodruff (2013); Bruhn (2011); Fajnzylber, Maloney, and Montes-Rojas (2011); and Kaplan, Piedra, and Seira (2011), for example. 69 For instance, Giné and Mansuri (2018) found that offering an eight-day business training to entrepreneurs led to improved business and household outcomes in Pakistan. 66 ∫ Support informal workers in gaining access to the benefits of formality, which can reduce the costs to the firms. Informal workers typically do not have access to job security (through written contracts), social insurance (e.g., pension or health insurance typically provided through employment), worker protections (through labor regulation), and/or ability to invest in skills development (on-the-job training). This lack affects workers’ labor productivity. Countries have been experimenting with mechanisms through which workers who may have the resources to save, but are not in a formal employer–employee relationship, can access contributory pensions/savings (see Guven [2019] for discussion). Studies also suggest that workers’ valuation of access to such benefits varies, as does employers’ willingness to provide them.70 For instance, written contracts that provide job security are valued by workers, while employers would not mind providing them. A significant policy question is what the government can do for workers whose productivity and earnings are low. Alongside robust poverty reduction between 2002 and 2016, the likelihood that workers would find themselves in poverty even while working also declined across all types of workers. The pace of poverty reduction among workers in urban areas, in Khyber Pakhtunkhwa, and in the services sector is particularly noticeable. Despite progress, close to one quarter of workers were in poverty in 2016; workers’ characteristics, such as location, gender, and education as well as the status and sector of their employment, are important determinants of the working poor status. There is a strong need to help poor, unskilled workers overcome key challenges to higher earnings in an environment where private sector jobs are limited. Multiple barriers are faced by these workers. They include lack of skills, productive assets (credit, land, livestock, etc.), financial products and services (savings or borrowings), and confidence and agency. Several key steps would support such workers in gaining access to quality jobs: ∫ Provide skills development opportunities. In Pakistan, there are stark differences in the returns to cognitive and noncognitive skills between wage workers and the self-employed. Some dimensions of noncognitive skills are strongly associated with higher earnings, but only for the self-employed, conditional on years of schooling and a cognitive score. A growing body of research corroborates the impact of efforts to encourage a “growth mindset” for entrepreneurial outcomes, and proposes cost-effective measures.71 Interventions to foster the development of noncognitive skills as part of active labor market programs—such as mentoring, community-led volunteer or club activities, and peer networks—should be further investigated and expanded in the context of Pakistan. ∫ Provide productive inclusion (also known as graduation) packages, a globally promoted approach to enhance the labor market prospects of the ultra-poor. Such efforts are based on a holistic framework that (i) promotes basic protection (e.g., health, nutrition, consumption), (ii) builds livelihood activities, (iii) deepens financial inclusion, and (iv) strengthens social empowerment. Pakistan has implemented related interventions but these have tended to focus narrowly on livestock transfers. Given the returns to domestic and international migration, several questions arise: Why don’t more people take advantage of such opportunities? How can policies help them? The main mobility challenges boil down to (i) uncertainties and costs associated with moving and (ii) prospects (or lack thereof) of high earnings and quality of living in the destination markets. Workers do not have sufficient information on better jobs, and tend to receive information through social networks or middlemen (in the case of international migration). The costs associated with a failure to secure employment opportunities are often too large to bear, especially for the poor, who have limited access to credit and savings. The urban labor market, as noted in the literature, is characterized by income volatility, the expansion of informality and the working poor, congestion and other constraints that make commuting to jobs difficult, slums and risky housing conditions, low-quality and highly congested social services, and potential socioeconomic marginalization due to the lack of social or kinship-based networks. The following measures are proposed to address several of the challenges associated with workers on the move: 70 See Kumar et al. (2019) for further discussion. 71 Campos et al. (2017) found that the provision of training focused on a self-starting, future-oriented, and persistently proactive mindset significantly improved business outcomes. 67 ∫ Provide resources (financial and nonfinancial) to pay for the costs of migration and mitigate the consequences of risks. Incentives to migrate, information on urban jobs, and interventions to promote safe migration are among a few examples.72 ∫ Adapt safety nets and labor programs to the urban context to help vulnerable workers (migrants) cope with risks and enhance their resilience. It is well recognized that urban areas pose a different set of opportunities and challenges than do rural areas, and policy responses must be adapted to the environment (Gentilini 2015). For instance, public works could focus more on local community projects that improve urban public services (e.g., water and sanitation, waste management), labor programs could support green enterprises (e.g., recycling, eco-construction, and green energy), and safety net programs could be linked with specific urban measures such as housing subsidies. ∫ Manage international migration (and remittances) more systematically for safe, inclusive, and productive migration. Many areas can be strengthened along the process: pre-decision and pre-departure support, services in destination countries, and support of return migration. In the short run, immediate support for workers and better regulations for middlemen (i.e., OEPs) would be helpful. Further, regulations regarding OEPs and their utilization of subagents can be better monitored and enforced by introducing a robust grievance redressal mechanism. In the longer run and at the macro level, greater efforts can be made to establish bilateral agreements to promote diversified productivity, the sustainability of labor migration systems, and workers’ protection. Policies to promote the inclusiveness of the labor market First and foremost, the highest priority on the policy agenda for labor market inclusiveness is to increase women’s economic empowerment. The Global Gender Gap Index 2018 produced by the World Economic Forum ranked Pakistan 148th out of 149 countries (WEF 2018). Deep-rooted social norms that affect women’s human capital accumulation and labor market status, as well as civic and political activities, are often pointed to as the culprit. In this daunting environment, steady and long-term policy efforts are needed to influence social norms toward encouraging women’s empowerment. ∫ Address gender norms, lower procedural costs for women’s economic activities, and promote girls’ education for overall women’s economic empowerment. Awareness-enhancing activities and social messaging can change the perceptions towards women’s work and mobility as well as shared household responsibilities. Relatedly, there is a strategic need to promote girls’ secondary education given its significant benefits (e.g., delayed marriage, delayed childbearing, and significant returns in the labor market). Provinces such as Punjab and Sindh operate secondary girls’ stipend programs to incentivize their school attendance. Similar initiatives can be introduced in other provinces, and such programs can be further strengthened to ensure the inclusion of girls in rural areas and from poor households. In addition, policy reforms are needed to make procedures less costly and cumbersome for women’s ownership of land, formal financial accounts, and enterprises. Meanwhile, several targeted active labor market programs could be adopted to specifically address some challenges and make a difference in women’s economic engagement and labor market activities at the margin. ∫ Promote income diversification and secondary activities in rural areas. The data show that women from low-income families who need secondary incomes are more likely to work.73 However, women’s choice of occupations is quite limited; many women engage in agricultural activities in rural areas where mobility constraints are relatively muted and working hours are flexible. Thus, agriculture is important for women’s employment. To further women’s economic empowerment while increasing their productivity, secondary 72 For the benefits of providing incentives to migrate in Bangladesh, see Bryan, Chowdhury, and Mobarak (2014). For efforts to promote safe migration in Bangladesh, see BRAC and World Bank (2016). In rural Philippines, providing individuals with vouchers to attend a job fair encouraged job seekers to look for work in urban areas and increased their chances of getting a job in the formal sector (Beam 2016). Similarly, in India, spreading information in villages about city-based business process outsourcing jobs for women, and offering placement services, increased the likelihood of young women working outside the home (Jensen 2012). 73 The FLFP rates of women in the first and second quintiles of household wealth—at 24 and 16 percent, respectively—are significantly higher than those of their non-poor counterparts as per the Labor Skills Survey (2015). 68 activities (such as cash crops, livestock, petty trades, and food processing) can be promoted in addition to the main agricultural activities.74 In doing so, the female workforce can be further utilized for government programs such as extension services to encourage other women to participate in labor market activities. ∫ Upgrade agricultural practices along the value chains for women. As much as women need agriculture for their employment, agriculture needs the female labor force as the female share of agricultural workers increased from 23 percent in 2001 to 39 percent in 2017. Thus, upgrading female agricultural workers’ skills and knowledge and providing better linkages to value chains can significantly enhance agricultural productivity. Gender-smart investments and technology adoption in the agricultural value chain—inputs, production, postharvest processing and storage, and transportation, marketing, and sales—in Pakistan should be further investigated and strengthened.75 ∫ Adopt measures that require or incentivize employers to provide gender-friendly work environments (e.g., separate workspaces, transportation, and childcare). Such efforts can start from large enterprises in manufacturing. Focus group discussions, however, suggest that many women have reservations about working in offices (or factories) vis-à-vis home-based employment, mainly for two reasons: work-life balance and workplace harassment (see Amir et al. 2018). At the same time, employers are concerned with the costs associated with hiring women such as potential implications for workplace dynamics, additional expenses for providing separate workplace facilities, and regulations regarding maternity leave. Incentives (e.g., tax credits/exemptions or subsidies) combined with regulations (e.g., requirements for workplace safety) can help mitigate employers’ costs while meeting the needs of women in the workplace. ∫ Harness information and communication technology (ICT) and social sector employment (e.g., caregivers, health workers, social mobilizers, etc.) for relatively well-educated women in urban areas. The unemployment rate among urban women with postsecondary education was about 12 percent in 2017. This rate is significantly higher than for men or women at any education level in rural or urban areas. Many unemployed women hope for public sector jobs, but opportunities are few. Emerging ICT jobs that could provide new opportunities for women, can be promoted. In Pakistan, freelancers in ICT generally work 34 hours a week, with flexible hours, and a gender gap in earnings does not seem to exist; some records even suggest that female freelancers in Pakistan earn more than their male counterparts. Measures to promote women’s ICT jobs include creating workspaces for women with internet connections, networks, and mentors. Also, social sector jobs (e.g., health-care workers, teachers, and caregivers) that tend to be considered female friendly and socially acceptable jobs for women, are significantly lacking. The disadvantages that young people face in the labor market should not be overlooked. School-to- work transitions take place as early as age 12. Children from low-income families are more likely to start working at earlier ages, with limited educational attainment. Despite the higher educational attainment among youth (compared to their older counterparts), the agricultural share of employment was about 40 percent among youth—a similar level to that of older workers. Globally, efforts to tackle the challenges of youth employment have not been successful, and Pakistan is no exception. The consensus is that effective interventions take a holistic approach to tackling the multiple challenges faced by youth. The following measures are proposed to address several of the challenges associated with at-risk youth: ∫ Strengthen youth training and entrepreneurship initiatives. Pakistan’s youth programs have been focusing on skills development, particularly formal technical and vocational education and training, but the access, coverage, governance, and quality challenges are widely recognized (Nomura 2019). In addition to addressing these challenges, training programs can be strengthened by other supplementary services including job search assistance, job placement, and on the job training. ∫ Make agriculture attractive for youth. The share of agricultural employment is about the same for youth and older adults—at around 40 percent. Frequently proposed areas to utilize the young workforce to improve 74 See FAO (2015) for further discussion. 75 See IFC for discussion on investments in women along agribusiness value chains; FAO (2015) for women in agriculture; and World Bank (2019c) for an example of gendered value chain analysis in Nepal. 69 agricultural productivity include: technology adoption, use of ICT, green fertilizers and pesticides, commercial crops, and nonfarm rural business (see World Bank 2017; FAO 2014). ∫ Provide job search assistance (and intermediation). Surveyed Pakistani jobseekers mention that their main methods of such information are their personal networks (friends and relatives), indicating significant barriers for youth (particularly women) to obtain labor market information. Studies have suggested that job search assistance, core to public employment services, is a cost-effective way of reducing youth unemployment in developed countries (Card, Kluve, and Weber 2018; Marinescu 2017). Job fairs, provision of information on the availability and requirements of jobs, and clinics for resumes, can be considered ∫ Provide workplace skills and experience building opportunities. These can be done through on-the- job training, wage subsidies, and apprenticeship. On-the-job training typically refers to employer-facilitated opportunities for workers to participate in skills-building opportunities while on the job. Wage subsidies are given to either firms or workers on the condition of employment of a targeted group of workers (in many cases youth). Apprenticeship is also an employment arrangement between employers and apprentices, where the apprentices can work at lower wages during a certain period of time. The mechanisms of these instruments vary slightly, but all these can be considered for workplace skills and experience building. ∫ Combine soft skills development with programs supporting youth. All the labor market programs mentioned above can add soft skills development given the growing evidence and emphasis on the role of soft skills in labor market success, as discussed above. Among many policy options discussed, a prioritized set of interventions to address key challenges merits further emphasis. The prioritized set of interventions include: those likely to have large spillover impacts on overall growth and economy including job markets (e.g., reform of tax policies, managed international migration); those proven to have promises based on global research (e.g., early childhood development, business regulation reforms, promotion of soft skills); and those that can build on existing initiatives and development (e.g., initiatives for youth, ICT). 70 REFERENCES Ackah, C., 2013. “Nonfarm Employment and Incomes in Rural Ghana.” Journal of International Development 25 (3): 325–39. Acosta, P., C. Calderón, P. Fajnzylber, and H. López. 2006. “Remittances and Development in Latin America.” The World Economy 29 (7): 957–87. Acs, Z. and P. Muller. 2008. “Employment Effects of Business Dynamics: Mice, Gazelles, and Elephants.” Small Business Economics, 30 (1): 85–100. Adams Jr, R. H., and A. Cuecuecha. 2010. “The Economic Impact of International Remittances on Poverty and Household Consumption and Investment in Indonesia.” Policy Research Working Paper WPS5433, World Bank, Washington, DC. ADB (Asian Development Bank). 2014. Rapid Assessment of Sexual Harassment in Public Transport and Connected Spaces in Karachi. Manila: ADB. Ahmed, S. A., Y. Cho, and T. Fasih. 2019. “Pakistan at 100: Human Capital.” Policy note, World Bank, Washington, DC. Ahmed, M. 2018. “Why Does Pakistan Have Repeated Macroeconomic Crises?” CGD Note, Center for Global Development, July. Ahmed, R., and S. H. Javed. 2015. “Business Process Outsourcing: A Case Study on Pakistan’s Out-Bound Call Centers.” SSRN Electronic Journal. http://doi.org/10.2139/ssrn.2643315. Alvarez. A., and J. Barney. 2014. “Entrepreneurial Opportunities and Poverty Alleviation.” Entrepreneurship Theory and Practice 38 (1): 159–84. Amir, S., and R. Pande. 2019. “Labor Force Aspirations, Experiences and Challenges for Urban, Educated Pakistani Women: Discussions in Four Metro Cities.” Policy Note, Pakistan Gender and Social Inclusion Platform and Center of Gender and Policy Studies, World Bank, Washington, DC. Amir, S., A. Kotikula, R. Pande, L. Bossavie, and U. Khadka. 2018. “Female Labor Force Partipation in Pakistan: What Do We Know?” Policy Note, Pakistan Gender and Social Inclusion Platform and Social Protection and Jobs, World Bank, Washington, DC. Andrabi, T., J. Das, A. I. Khwaja, T. Vishwanath, and T. Zajonc. 2008. Pakistan – Learning and Educational Achievements in Punjab Schools (LEAPS): Insights to Inform the Education Policy Debate. World Bank, Washington, DC. Arif, G. M. 2010. “Economic and Social Impacts of Remittances on Households: The Case of Pakistani Migrants Working in Saudi Arabia.” Pakistan Institute of Development Economics, International Organization for Migration (IOM), Geneva. Arshad, S. 2008. The Economic Contribution of Pakistani Women through their Unpaid Labour. Islamabad, Pakistan: Society for Alternative Media and Research and HealthBridge. ASER. 2015a. Annual Status of Education Report Urban. Lahore, Pakistan: South Asian Forum for Education Development. ———. 2015b. Annual Status of Education Report National. Lahore, Pakistan: South Asian Forum for Education Development. ———. 2016. Annual Status of Education Report National. Lahore, Pakistan: South Asian Forum for Education Development. Attanasio, O., E. Fitzsimons, A. Gomez, M. I. Gutiérrez, C. Meghir, and A. Mesnard . 2010. “Children’s Schooling and Work in the Presence of a Conditional Cash Transfer Program in Rural Colombia.” Economic Development and Cultural Change 58 (2): 181–210. 71 Bari, F. 2012. “Rural-Urban Boundaries.” Pakistan Today, April 17. https://www.pakistantoday.com.pk/2012/04/16/rural-urban-boundaries/ Beam, E. A. 2016. “Do Job Fairs Matter? Experimental Evidence on the Impact of Job-Fair Attendance.” Journal of Development Economics 120: 32–40. Bhalotra, S., and C. Heady. 2003. “Child Farm Labor: The Wealth Paradox.” The World Bank Economic Review 17 (2): 197–27. Bossavie, L., U. Khadka, and V. Strokova. 2018. Pakistan—A Labour Market Overview. Draft Version. Washington, DC: World Bank Group. Bourguignon, F., F. H. Ferreira, and P. G. Leite. 2003. “Conditional Cash Transfers, Schooling, and Child Labor: Micro-Simulating Brazil’s Bolsa Escola Program.” The World Bank Economic Review 17 (2): 229–54. BRAC and World Bank. 2016. “Promoting Safe Migration through Innovation.” Mimeo, World Bank, Washington, DC. Bruhn, Miriam. 2011. “License to sell: the effect of business registration reform on entrepreneurial activity in Mexico.” Review of Economics and Statistics, 93(1): 382-86. Bryan, G., S. Chowdhury, and A. M. Mobarak. 2014. “Under-Investment in a Profitable Technology: The Case of Seasonal Migration in Bangladesh.” Econometrica 82 (5): 1671–748. Bursztyn, L., A. Gonzalez, and D. Yanagizawa-Drott. 2018. “Misperceived Social Norms: Female Labor Force Participation in Saudi Arabia.” Mimeo, University of Chicago, Chicago, IL. Campos, F., M. Frese, M. Goldstein, L. Iacovone, H. Johnson, D. McKenzie, and M. Mensmann. 2017. “Teaching Personal Initiative Beats Traditional Training in Boosting Small Business in West Africa.” Science 357 (6357): 1287–2190. Card, D., J. Kluve, and A. Weber. 2018. “What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations.” Journal of the European Economic Association 16 (3): 894–931. http://doi.org/10.1093/jeea/jvx028. Cheema, A., A. Khwaja, F. Naseer, and J. N. Shapiro. 2019. “Glass Walls: Experimental Evidence on Access Constraints Faced by Women.” Mimeo, Harvard University. Cho, Y., A. Denisova, S. Yi, and U. Khadka. 2018. “Migrant Worker Outcomes: Korean Employment Permit System Triumphs and Challenges.” World Bank, Washington, DC. Christiaensen, L., and P. Premand. 2017. Cote d’Ivoire Jobs Diagnostic: Employment, Productivity, and Inclusion for Poverty Reduction. Washington, DC: World Bank. Collective for Social Science Research. 2017. “Social Processes and Individual Behaviour: A Qualitative Study of Public Service Delivery in Informally Settled Areas of Karachi.” http://www.researchcollective.org/project_info.php?id=C_077. Cunningham, W., and R. Munoz. 2018. Myanmar’s Future Jobs: Embracing Modernity. Washington, DC: World Bank. Cunningham, W., M. P. Torrado, and M. Sarzosa. 2016. “Cognitive and Non-Cognitive Skills for the Peruvian Labor Market: Addressing Measurement Error through Latent Skills Estimations.” Policy Research Working Paper WPS 7550, World Bank, Washington, DC. De Baldini Rocha, M. S., and V. Ponczek. 2011. “The Effects of Adult Literacy on Earnings and Employment.” Economics of Education Review 30 (4): 755–64. De Janvry, A., F. Finan, and E. Sadoulet. 2004. “Can Conditional Cash Transfers Serve as Safety Nets to Keep Children at School and Out of the Labor Market?” Department of Agricultural and Resource Economics, UC Berkeley. https:// escholarship.org/uc/item/5fp0g5p2. De Mel, S., D. McKenzie, and C. Woodruff. 2013. “The Demand for, and Consequences of, Formalization among Informal Firms in Sri Lanka,” American Economic Journal: Applied Economics, 5 (2): 122–50. Deng, F., P. Illangovan, and E. Blanco Armas. 2019. Pakistan at 100: Regional Connectivity (English). Washington, DC: World Bank Group. Di Porto, E., L. Elia, and C. Tealdi. 2017. “Informal Work in a Flexible Labour Market.” Oxford Economic Papers 69 (1): 143–64. 72 Digman, J. M. 1997. “Higher-Order Factors of the Big Five.” Journal of Personality and Social Psychology 73 (6): 1246–56. Djankov, S. 2016. “The Doing Business Project: How It Started: Correspondence.” Journal of Economic Perspectives, 30 (1): 247-48. Dustmann, C., and O. Kirchkamp. 2002. “The Optimal Migration Duration and Activity Choice after Remigration.” Journal of Development Economics 67 (2): 351–72. Easterly, W. 2000. “The Middle Class Consensus and Economic Development.” Policy Research Working Paper WPS 2346, World Bank, Washington, DC. Edmonds, E. V., and N. Schady. 2012. “Poverty Alleviation and Child Labor.” American Economic Journal: Economic Policy 4 (4): 100–24. Ejaz, M. 2011. “The Determinants of Female Labor Force Participation in Pakistan: An Instrumental Variable Approach.” CREB Working Paper 01-11, Centre for Research in Economics and Business, Lahore School of Economics, Lahore, Pakistan. Ellis, P. and M. Roberts. 2016. Leveraging Urbanization in South Asia: Managing Spatial Transformation for Prosperity and Livability. World Bank, Washington, DC. Fajnzylber, P., W. Maloney, and G. Montes-Rojas. 2011. “Does Formality Improve Microfirm Performance? Evidence from the Brazilian SIMPLES Program.” Journal of Development Economics 94 (2): 262–76. FAO (Food and Agriculture Organization). 2014. Youth and Agriculture: Key Challenges and Concrete Solutions. Rome, Italy: FAO. ———. 2015. Women in Agriculture in Pakistan. Islamabad: FAO. Farole, T. and Y. Cho. 2017. “Bangladesh Jobs Diagnostic”, Jobs Series: No. 9. World Bank, Washington, DC. Feld, L. P., and F. Schneider 2010. “Survey on the Shadow Economy and Undeclared Earnings in OECD Countries.” German Economic Review 11 (2): 109–49. Filmer, D., and L. Fox. 2014. Youth Employment in Sub-Saharan Africa. Washington, DC: World Bank. Fox, L., and T. P. Sohnesen. 2012. “Household Enterprises in Sub-Saharan Africa: Why They Matter for Growth, Jobs, and Livelihoods.” Policy Research Working Paper WPS 6184, World Bank, Washington, DC. Gentilini, U. 2015. “Entering the City: Emerging Evidence and Practices with Safety Nets in Urban Areas.” Social Protection and Labor Discussion Paper 1504, World Bank, Washington, DC. Giné, X., and G. Mansuri. 2018. “Money or Management? A Field Experiment on Constraints to Entrepreneurship in Rural Pakistan.” Policy Research Working Paper WPS 6959, Impact Evaluation series, World Bank, Washington, DC. Giné, X., G. Mansuri, and M. Picón. 2011. “Does a Picture Paint a Thousand Words? Evidence from a Microcredit Marketing Experiment.” The World Bank Economic Review 25 (3): 508–42. Gioli, G., T. Khan, and J. Scheffran. 2013. “Remittances and Community Resilience to Conflict and Environment Hazards in North Western Pakistan.” In Remittance Flows to Post-Conflict States: Perspectives on Human Security and Development. Pardee Center Task Force Report. Boston, MA: Frederick S. Pardee Center for the Study of the Longer-Range Future, Boston University. Guven, M. 2019. “Extending Pension Coverage to the Informal Sector in Africa.” Social Protection and Labor Discussion Paper 1933, World Bank, Washington, DC. Haltiwanger, J. 2006. “Entrepreneurship and Job Growth.” Mimeo. University of Maryland. Haltiwanger, J., R. S. Jarmin, and J. Miranda. 2010. “Who Creates Jobs? Small vs. Large vs. Young.” Working Papers 10–17, Center for Economic Studies, U.S. Census Bureau. Haltiwanger, J., R. S. Jarmin, R. Kulick, and J. Miranda. 2016. “High Growth Young Firms: Contribution to Job, Output, and Productivity Growth.” In Measuring Entrepreneurial Businesses: Current Knowledge and Challenges, 11–62. Cambridge, MA: National Bureau of Economic Research. Haltiwanger, J., S. Davis, K. Handley, R. Jarmin, J. Lerner, and J. Miranda. 2013. “Private Equity, Jobs, and Productivity.” NBER Working Paper 19458, National Bureau of Economic Research, Cambridge, MA. 73 Hanushek, E. A., and L. Woessmann. 2008. “The Role of Cognitive Skills in Economic Development.” Journal of Economic Literature 46 (3): 607–68. Iftikhar, P. 2017. “Pakistan’s Freelancing Industry Is Thriving.” The Express Tribune, October 30. https://tribune.com.pk/ story/1544773/2-export-earnings-pakistans-freelancing-industry-thriving/. ILO (International Labour Organization). 2013. The Informal Economy and Decent Work: A Policy Resource Guide Supporting Transitions to Formality. Geneva: ILO. Javid, M., U. Arif, and A. Qayyum. 2012. “Impact of Remittances on Economic Growth and Poverty.” Academic Research International 2 (1): 433–47. Jensen, R. 2012. “Do Labor Market Opportunities Affect Young Women’s Work and Family Decisions? Experimental Evidence from India.” The Quarterly Journal of Economics 127 (2): 753–92. Jureidini, R. 2017. Wage Protection Systems and Programmes in the GCC. Gulf Labour Markets and Migration (GLMM) Research Report 01/2017, Migration Policy Center (MPC) and the Gulf Research Center (GRC). http://gulfmigration.eu/media/pubs/rp/GLMM_EN_2017_RR01.pdf. Kaplan, D., E. Piedra, and E. Seira. 2011. “Entry regulation and business start-ups: Evidence from Mexico” Journal of Public Economics 95(11-12): 1501-15. Kässi, O., and V. Lehdonvirta. 2016. “Online Labour Index: Measuring the Online Gig Economy for Policy and Research.” Paper presented at Internet, Politics & Policy 2016, September 22–23, Oxford, UK. http://ilabour.oii.ox.ac.uk/online-labour-index/. Kawabata, M. 2003. “Job Access and Employment among Low-Skilled Autoless Workers in US Metropolitan Areas.” Environment and Planning A: Economy and Space 35 (9): 1651–68. Kumar, K., M. Mahmud, S. Natarji, and Y. Cho. 2019. “Employer and Employee Preferences for Worker Benefits: Evidence from a Matched Survey on the Bangladesh Informal Sector.” IZA working paper 12064, IZA Institute of Labor Economics, Bonn, Germany. Lange, G.-M., Q. Wodon, and K. Carey. 2018. The Changing Wealth of Nations 2018: Building a Sustainable Future. Washington, DC: World Bank. Lanjouw, J. O., and P. Lanjouw. 2001. “The Rural Non-Farm Sector: Issues and Evidence from Developing Countries.” Agricultural Economics 26 (1):1–23. Lokshin, M., M. Bontch-Osmolovski, and E. Glinskaya. 2010. “Work-Related Migration and Poverty Reduction in Nepal.” Review of Development Economics 14 (2): 323–32. Lucas, R. 2004. International Migration Regimes and Economic Development. Draft report to the Expert Group on Development Issues, Swedish Ministry for Foreign Affairs. Malkani, A. 2017. “Employing Women.” Dawn, June 18. https://www.dawn.com/news/1340243. Maloney, W. 2004. “Informality Revisited.” World Development 32 (7): 1159–78. Mansuri, G. 2006. “Migration, School Attainment and Child Labor: Evidence from Rural Pakistan.” Policy Research Working Paper 3945, World Bank, Washington, DC. Marinescu, I. E. 2017. “Job Search Monitoring and Assistance for the Unemployed.” IZA World of Labor 2017: 380. Martinez, D., O. Mitnik, E. Salgado, L. Scholl, and P. Yanez-Pagans. 2018. “Connecting to Economic Opportunity? The Role of Public Transport in Promoting Women’s Employment in Lima.” IZA Discussion Paper 12020, IZA Institute of Labor Economics, Bonn, Germany. McCormick, B., and J. Wahba. 2001. “Overseas Work Experience, Savings and Entrepreneurship amongst Returnees to LDCs.” Scottish Journal of Political Economy 48 (2): 164–78. McKenzie, D., and D. Yang. 2015. “Evidence on Policies to Increase the Development Impacts of International Migration.” World Bank Research Observer 30 (2): 155–92. 74 Mesnard, A. 2004. “Temporary Migration and Capital Market Imperfections.” Oxford Economic Papers 56 (2): 242–62. Minniti, M., and W. Naudé. 2010. What Do We Know about the Patterns and Determinants of Female Entrepreneurship across Countries?. The European Journal of Development Research 22 (3): 277–93. Nagler, P., and W. Naudé. 2017. “Non-Farm Entrepreneurship in Rural Sub-Saharan Africa: New Empirical Evidence.” Food Policy 67: 175–91. Naidu, S., Y. Nyarko, and S. Wang. 2014. “Worker Mobility in a Global Labor Market: Evidence from United Arab Emirates.” NBER Working Paper 20388, National Bureau of Economic Research, Cambridge, MA. NIPS (National Institute of Population Studies). 2019. Pakistan Demographic and Health Survey 2017–18. Islamabad. Nomura, S. 2019. “Pakistan: Skills Assessment for Economic Growth.” Mimeo, World Bank, Washington, DC. Packard, T. G., J. Koettl, and C. Montenegro. 2012. In from the Shadow: Integrating Europe’s Informal Labor. Washington, DC: World Bank. Pakistan Remittance Initiative. 2015. “Pakistan Remittance Initiative: A Brief.” State Bank of Pakistan, Karachi, Pakistan. http://www.pri.gov.pk/wp-content/uploads/2015/06/PRI-Brief-2015.pdf. Ponce, J., I. Olivié, and M. Onofa. 2011. “The Role of International Remittances in Health Outcomes in Ecuador: Prevention and Response to Shocks.” International Migration Review 45 (3): 727–45. Redaelli, S. 2019. “Pakistan at 100: From Poverty to Equity.” Policy note, World Bank, Washington, DC. Ricks, T. 2017. “Balochistan Looks Ready to Blow, and When It Does, It Likely Will Spill Over.” https://foreignpolicy.com/2017/02/09/bolochistan-looks-ready-to-blow-and-when-it-does-it-likely-will-spill-over/. Sajjad, F., G. A. Anjum, E. Field, and K. Vyborny. 2017. “Overcoming Barriers to Women’s Mobility: Improving Women’s Access to Public Transport in Pakistan.” Policy Brief, Center for Economic Research in Pakistan. https://cerp.org.pk/filemanager/files/1/pdfs/5d397dd63e958.pdf. Saqib, N. U., and G. M. Arif. 2012. “Time Poverty, Work Status and Gender: The Case of Pakistan.” The Pakistan Development Review 51 (1): 23–46. Sharma, S., G. Nayyar, and K. Y. Kim. 2019. “Structural Transformation in Pakistan: A Note for Pakistan@100.” World Bank, Washington, DC. Skoufias, E., S. Parker, J. Behrman, and C. Pessino. 2001. “Conditional Cash Transfers and Their Impact on Child Work and Schooling: Evidence from the PROGRESA Program in Mexico,” Economia 2(1): 45-96 SPDC (Social Policy and Development Centre). 2016. “Social Development in Pakistan: Annual Review 2014–15.” SPDC, Karachi, Pakistan. UN-Habitat. 2003. The Challenge of Slums: Global Report on Human Settlements 2003. London: Earthscan. UNODC (United Nations Office on Drugs and Crime). 2012. Migrant Smuggling in Asia: A Thematic Review of Literature. Vienna: UNODC. van der Linden, D., J. te Nijenhuis, and A. B. Bakker. 2010. “The General Factor of Personality: A Meta-Analysis of Big Five Intercorrelations and a Criterion-Related Validity Study.” Journal of Research in Personality 44 (3): 315–27. Vijverberg, W., and J. Haughton. 2004. “Household Enterprises in Vietnam: Survival, Growth, and Living Standards.” In Economic Growth, Poverty, and Household Welfare in Vietnam, edited by P. Glewwe, N. Agrawal, and D. Dollar, 95–132. Washington, DC: World Bank. Weber, M., and J. Langbein. 2018. “Can Household Enterprises Help Create More and Better Jobs in Pakistan? An Overview of Productivity and Welfare Impacts of Household Enterprises in Pakistan.” Background paper, World Bank, Washington, DC. WEF (World Economic Forum). 2018. The Global Gender Gap Report 2018. Geneva: WEF. World Bank. 2012a. World Development Report 2013: Jobs. Washington, DC: World Bank. ———. 2012b. Enterprise Survey 2013: Pakistan Country Profile. Washington, DC: World Bank. 75 ———. 2017. “Access to the German Labor Market for Moroccan Youth.” Recognition of progress note, mimeo, World Bank, Washington, DC. ———. 2018. Doing Business 2019: Training for Reform. Washington, DC: World Bank. ———. 2019a. Pakistan @ 100: Shaping the Future. Washington, DC: World Bank Group. ———. 2019b. Ending Learning Poverty: What Will It Take? Washington, DC: World Bank. ———. 2019c. Doing Business 2020: Comparing Business Regulation. Washington, DC: World Bank. ———. N.d. “Human Capital Project.” http://www.worldbank.org/en/publication/human-capital. ———. Various years. Doing Business: Measuring Business Regulations. Washington, DC: World Bank. https://www. doingbusiness.org/en/reports/case-studies/view-all. Yang, D. 2004. “International Migration, Human Capital, and Entrepreneurship: Evidence from Philippine Migrants’ Exchange Rate Shocks.” Working Paper, University of Michigan. Yang, D., and H. Choi. 2007. “Are Remittances Insurance? Evidence from Rainfall Shocks in the Philippines.” World Bank Economic Review 21 (2): 219–48. 76 APPENDICES APPENDIX A: SUMMARY OF BACKGROUND PAPERS AND DATA SOURCES USED Table A.1 Summary of background papers and data sources used in the Jobs Diagnostic Background papers Description Main data source(s) This analysis examines aggregate 1. Demographics and Structural trends and prospects in employment, World Development Indicators (WDIs) Transformation and Jobs productivity, sectoral composition, and 2000–2017 structural transformation. This analysis provides the most up-to- Labor Force Surveys (various years “Core” 2. Workers and Jobs date information on the supply side of from 2001/02 to 2017/18); Labor Jobs Diagnostic the labor market in Pakistan. Skills Surveys (2014–16) This analysis looks into the characteristics of formal sector firms Doing Business Survey 2018; Limited 3. Firms and Jobs and employment in the manufacturing microdata sector. 4. Constraints to female LFP and This in-depth investigation focuses on Labor Force Surveys (various years) access to better jobs gender using the existing analysis. Labor Skills Surveys (2014–16) This analysis looks into the Household Income and Expenditure 5. Household enterprises characteristics of informal household Surveys (2015/16) enterprises This stock-taking exercise covers more than 70 distinct national- and 6. Stock-taking of government’s Compilation of various sources such province-level active labor market active labor market programs as provincial acts and provisions programs and their budgets for FY 2016–17. This review compiles and benchmarks all recently passed labor market 7. Review of labor market Compilation of various sources such regulations at the provincial level and, regulations as provincial acts and provisions Complementary hence, fills a critical gap in the analysis understanding of labor policies. This note provides a view into the potential for the increasing use of Compilation of various sources such 8. Digital jobs in Pakistan information and communication as the Online Labor Index technology to create more, better, and inclusive jobs in Pakistan. Labor Force Surveys (1999/00, 9. Constraints to youth This note investigates the skills gap in 2004/05, 2009/10, 2014/15); Labor employment the labor market. Skills Surveys (2014–16) Compilation of administrative data of This note investigates the trends and the Bureau of Emigration & Overseas 10. International migration determinants of overseas migration Employment (BEOE) and the Global from Pakistan. Knowledge Partnership on Migration and Development (KNOMAD) 77 APPENDIX B: REGRESSION ANALYSIS OF EMPLOYMENT OUTCOMES The micro-determinants of the status and sector of employment suggest uneven access to quality jobs by gender and region, as well as across age groups. Figure B.1 and Figure B.2 present the results of multinomial logit regressions on the determinants of working in wage employment (compared to agricultural self-employment) and agriculture (compared with services). As discussed, education has a positive (negative) effect on the probability of being a wage employee (agricultural worker). However, only an upper secondary education and above shows a strong association with wage employment, whereas any education, even at the primary level, has a positive association with nonagricultural employment. The association between education level and employment type (employment status and sector) is generally stronger for women than men. The results also show that living in an urban area is positively associated with wage employment and negatively with agricultural self-employment. Married women are less likely to work as wage employees, but more likely to work in agriculture, suggesting that, for females, marriage is associated with engaging in agricultural activities that are flexible and conducive to balancing work and family responsibilities. Figure B.1 Figure B.2 Likelihood of working in wage employment, 2017 Likelihood of working in agriculture, 2017 BALOCHISTAN BALOCHISTAN SINDH SINDH PUNJAB PUNJAB POST-SECONDARY POST-SECONDARY UPPER SECONDARY UPPER SECONDARY LOWER SECONDARY LOWER SECONDARY PRIMARY PRIMARY MARRIED MARRIED URBAN URBAN HOUSEHOLD SIZE HOUSEHOLD SIZE YOUTH YOUTH –0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 –0.5 –0.4 –0.3 –0.2 –0.1 0 0.1 0.2 MALE FEMALE MALE FEMALE Source: LFS 2017. Note: Marginal effects from multinomial logit regressions are reported with agricultural self-employment as a base category (left) and the service sector (right) as a base category; Khyber Pakhtunkhwa is a base category for provinces; and no education is a base category for education. 78 APPENDIX C: POLICY AND INSTITUTIONAL ANALYSIS OF INTERNATIONAL MIGRATION IN PAKISTAN The Emigration Ordinance 1979 provides a basic legal framework governing outmigration while the Emigration Rules 1979 stipulate more detailed measures that address various aspects of the full cycle of migration. The Emigration Ordinance 1979 mandates the state’s role and necessary intervention to promote overseas employment, control and regulate outflows, and protect migrant workers’ rights and welfare. The ordinance thus authorizes the government to make emigration rules (Clause 16) as necessary in line with the purposes of the ordinance, which stipulates more detailed measures that address various aspects of the full cycle of migration.76 Measures include: regulation of overseas employment promoters (OEPs) (including of licensing, security deposits, recruitment service fees, and documents to protect migrants); support mechanisms for accommodation and medical care abroad for migrants; implementation of the welfare fund for migrants and their families left behind; and a bond mechanism to help ensure the return of migrants and receive returning migrants. The bond is refunded to migrants only if they honor the terms of the employment contract, the visa, and other legal conditions specified in the bond agreement. In practice, however, enforcement falls short. Multiple examples show lapses in the enforcement of the measures discussed above. Major ones include measures related to migration costs, use of subagents, and wages that put migrants at significant risk. The Emigration Rules (Clause 15-A) prohibit OEPs from charging fees beyond the actual expenses incurred for air tickets, medical exams, work permits, levies, visas, and documentation along with a service fee capped at PRs 6,000. However, the Global Knowledge Partnership on Migration and Development’s (KNOMAD’s) Migration Cost Survey reveals that temporary labor migrants have been paying large recruitment service fees, and making informal payments to expedite the emigration process. Also, the Emigration Rules (Clauses 7, 11) prohibit OEPs from subcontracting work to unlicensed agents. In practice, however, OEPs and migrants both rely heavily on unlicensed subagents to process overseas employment; this is especially true for migrants from rural, remote areas where information penetration is likely to be low. The Emigration Rules (Clause 20-A) also mandate that the Bureau of Emigration & Overseas Employment (BEOE) assess whether the wages and other terms and conditions of overseas positions offered are “reasonable” prior to granting OEPs permission to process the demand. But it does not define a “reasonable” wage level. Multiple government efforts to formulate and implement overarching migration policies have been made. Various national development plans77 recognize the importance of remittances to the country’s economy, and, as part of employment generation strategies, outline ways the government might do more to harness overseas employment opportunities. Domestic labor policies at both the federal and provincial levels also aim to lower barriers to emigration.78 Unilateral domestic policies are further complemented by bilateral and multilateral initiatives aimed at promoting migrant welfare. Pakistan has signed memorandums of understanding (more flexible than a legally binding arrangement such as a bilateral labor agreement) with multiple countries over time.79 Moreover, Pakistan has also been participating in multilateral dialogues to promote Pakistanis’ overseas employment and enhance their protection in destination countries. It is a member of the Abu Dhabi Dialogue, the Colombo Process, and the more recent Global Forum on Migration and Development.80 76 In addition to the major migration legislation and rules discussed above, supplemental legislation related to international migration includes the Federal Investigation Agency Act 1974, which established the Federal Investigation Agency, mandated with responsibilities to investigate all specified offenses including those punishable under the Emigration Ordinance; Prevention and Control of Human Trafficking Ordinance 2002 under the Ministry of Interior; Passport Act and Passport Rules 1974, which regulates departure from and entry into Pakistan and visits to foreign countries undertaken by citizens of Pakistan; and Exit Control Ordinance 1981, which empowers the government to prevent the exit of any person from the country regardless of having valid documentation for the exit, without specifying the foundation for such prevention from exit. 77 These include the 11th Five-Year Plan 2013–18, and the government’s long-term development plan, Vision 2025. 78 Labor Policy 2010 at the federal level calls for simplifying and streamlining emigration procedures by revising the Emigration Rules, and recommends that OEPs be offered incentives to improve their performance. Also, a draft Punjab Labor Policy 2015 calls for building a collaborative mechanism with relevant federal government agencies to better tap overseas employment opportunities for the Punjab workforce. 79 Bahrain (2014), Italy (2009), Jordan (1978), the Republic of Korea (2006), Kuwait (1995, revised in 2013), Libya (2009), Malaysia (2003), Qatar (1987, revised in 2008), and the United Arab Emirates (2006). 80 The objectives of such international forums include: first, sending countries at different levels of maturity can offer lessons to one another that will enhance efforts to tackle common challenges; second, to leverage their bargaining power in relation to receiving countries, sending countries can come together to form coalitions; and third, sending and receiving countries can work together to maximize mutual gains from migration while promoting migrant welfare. 79 Several key institutions operate under the Ministry of Overseas Pakistanis and Human Resource Development (MOPHRD) and execute the legal and policy efforts mentioned above (see Box C.1).81 The MOPHRD, established in 2013 through a merger of two preexisting ministries (the Ministry of Overseas Pakistanis and the Ministry of Human Resources), develops measures to enhance the welfare of Pakistani emigrants and their family members left in Pakistan, within the ambit of overseeing broader human resource development in the country.82 Three of the ministry’s agencies—the BEOE, the Overseas Employment Corporation (OEC), and the Overseas Pakistanis Foundation (OPF)—plus the community welfare attachés it oversees implement migration‑related policies, rules, and regulations (Figure C.1). Figure C.1 Institutional arrangements of Pakistan’s migration system MINISTRY OF OVERSEAS PAKISTANIS AND HUMAN RESOURCE MINISTRY OF FOREIGN AFFAIRS (FOREIGN DEVELOPMENT (MOPHRD) MISSIONS) BUREAU OF EMIGRATION AND OVERSEAS OVERSEAS EMPLOYMENT OVERSEAS PAKISTANIS COMMUNITY WELFARE ATTACHES EMPLOYMENT (BEOE) CORPORATION (OEC) FOUNDATION (OPF) (CWA) OTHER PLAYERS PROTECTOR OF MINISTRY OF INTERIOR (DIRECTORATE GENERAL OF PASSPORTS AND NADRA) EMIGRATION (POE) (7 OFFICES IN PUNJAB, STATE BANK OF PAKISTAN, COMMERCIAL BANKS, MONEY TRANSFERS SINDH, KP, AND BALOCHISTAN) TVET, NTAVTTC, TRAINING CENTER, TRADE TESTING BOARDS/CENTERS PRIVATE SECTOR INGOs/NGOs/MEDIA/MEDICAL CENTER OVERSEAS EMPLOYMENT PROMOTERS MIGRANT RESOURCE CENTERS Source: Authors’ compilation. Note: INGO = international nongovernmental organization; KP = Khyber Pakhtunkhwa; NGO = nongovernmental organization; NTAVTTC = National Vocational and Technical Training Commission; TVET = technical and vocational education and training. 81 The Ministry of Foreign Affairs (MOFA) is tasked with managing Pakistan’s diplomatic and consular relations as well as its foreign policy. Thus, MOFA is responsible for maintaining Pakistani government offices abroad and supporting overseas Pakistanis. However, the report focuses on labor migrants, and thus highlights the role under the MOPHRD. 82 Its vision is to be “vibrant, efficient and transparent official business process based on integrated IT networking and to provide better services to the Overseas Pakistanis (OPs), rehabilitation of returning OPs, taking measures for protection of their investment, to secure emigration of right worker for right job, to promote workers welfare by adopting workers friendly policies and to ensure provision of social protection and decent working condition” as well as “the preparation of special incentives for the Pakistani Diaspora to proactively participate in the National Development and economic growth” as per the Ministry’s Yearbook 2013–14. 80 BOX C.1: KEY INSTITUTIONS UNDER THE MINISTRY OF OVERSEAS PAKISTANIS AND HUMAN RESOURCE DEVELOPMENT The Bureau of Emigration & Overseas Employment (BEOE) and the Protectorate of Emigrants (POE) under the BEOE play a key role in migration governance. The BEOE, under the Emigration Wing of the MOPHRD, is the main agency that implements rules related to emigration procedures. Key functions and responsibilities of the BEOE include facilitation of overseas employment, dissemination of information on the migration process and employment opportunities, and the provision of worker protection measures and regulatory oversight of private sector overseas employment promoters (OEPs). The BEOE operates primarily through its seven POE Offices—Lahore, Rawalpindi, and Multan in Punjab; Karachi in Sindh; Peshawar and Malakand in Khyber Pakhtunkhwa; and Quetta in Balochistan. POEs are responsible for managing private recruitment agencies and prospective migrants in the recruitment process. The Overseas Employment Corporation (OEC) is a public agency that is authorized to locate jobs abroad and recruit workers for these jobs, but its role has been limited. It maintains a data bank of resumes of jobseekers and sends these to foreign employers in response to requests for job matching. Unlike private OEPs, the OEC is concerned with specific occupations (e.g., doctors needed in Saudi Arabia and Oman) or arrangements. However, the role of the OEC as a recruiting agency has gradually declined over the years. Between 1971 and 2015, around 58 percent of all migrant workers going abroad formally went through OEPs, followed by 40 percent through their own efforts; only 2 percent used the OEC’s services. The Overseas Pakistanis Foundation (OPF), established under the Emigration Ordinance 1979, is a public entity that provides welfare services to overseas Pakistanis and their families left in Pakistan. The OPF also supports special cases such as the economic rehabilitation of returning disabled Pakistanis, transportation of corpses, and assistance to those affected by emergencies. The activities and services of the OPF are financed through the OPF Welfare Fund, in which each Pakistani going abroad to work is required to deposit PRs 2,500 (US$21) as insurance. The insurance can be claimed for up to PRs 1 million in case of unforeseen extreme events, and the insurance membership is valid during the stay abroad and up to five years within the worker’s return to Pakistan. Community welfare attachés, which are operating in 15 major destination countries, serve to provide protection for migrant workers abroad. These are expected to provide dispute resolution and intermediation services when migrants file complaints against their employers, to handle extreme cases such as death and criminal charges against migrant workers, and to keep in regular contact with migrant communities. OEPs are registered private recruitment companies that facilitate overseas employment. They need to acquire a license from the BEOE to operate and to establish contracts with foreign employers for Pakistani workers. In addition, the OEPs often represent overseas employers or intermediaries such as recruitment companies. OEPs play a major role, as the majority of migrants use their recruitment services. As of 2016, there were 2,195 licensed OEPs in the country, of which the largest share (35 percent) was in Rawalpindi, followed by Lahore (20 percent), and Karachi (16 percent) (Figure C.2). Only 20 percent of the licensed OEPs are located in KPK (e.g., Peshawar and Malakand) whereas the large majority are in Punjab (e.g., Rawalpindi, Lahore, Multan). The capacity of OEPs varies widely such that over 60 percent of the OEPs were able to mobilize fewer than 100 migrants per year, while only 1 percent of OEPs had the capacity to mobilize over a 1,000 migrants per year (Figure C.3). OEPs have formed the Pakistan Overseas Employment Promoters Association (POEPA),83 which collectively represents OEPs. Licenses and their renewal are the main mechanisms through which the BEOE manages OEPs. BEOE-issued licenses for OEPs can be revoked or their renewal can be denied in case of misconduct or violations such as worker abuse or falsified foreign employment. If an OEP has successfully facilitated 50 or more workers 83 POEPA is registered with the Securities and Exchange Commission of Pakistan under the Companies Ordinance (1984). Apart from regulations from the BEOE, the POEPA’s member OEPs follow the business rules developed under the Emigration Ordinance and the Code of Conduct for ethical recruitment. 81 Figure C.2 Figure C.3 Distribution of overseas employment promoters by location Distribution of overseas employment promoters by capacity QUETTA 1,000+ 0% MALAKAND 500–999 7% 1% 10% 300–399 MULTAN 8% RAWALPANDI 9% 35% 200–299 9% PESHAWAR 10% 100–199 14% 0–99 KARACHI 61% 16% LAHORE 20% Source: BEOE. in the last three years of the license, the license renewal is granted for another three years.84 In addition, OEPs with reliable performance are also recognized by the government through a performance certificate. As of 2016, 129 licenses had been canceled and 15 percent of OEPs had received at least one complaint in the past three years. To promote the transparent regulation of OEPs, the BEOE has been making key OEP-related information (e.g., contact information, complaints, OEPs with licenses revoked) publicly available through its website, though the information remains underutilized given the general population’s level of access to the Internet. A large majority of migrants, meanwhile, interact with unlicensed recruitment agents who may or may not be associated with OEPs. These are most often individual brokers who match aspirant migrants with OEPs and, sometimes, directly with a foreign-based recruiter. These brokers, unlicensed and unregistered, are outside the realm of regulatory policies governing OEPs and operate informally, often charging a large service fee. In fact, their presence is the major culprit for the markedly high migration costs borne by Pakistani workers. Since they are not registered with official channels, there is no credible estimate of the number of active labor migration brokers in Pakistan; however, the demand for their services is strong, among both OEPs and aspiring migrants. Most OEPs are based in large cities, and it may be financially unviable for them to have branch offices, although they are legally allowed to have up to two. Thus, many OEPs find it more cost-effective to rely on brokers to mobilize aspirant migrants from rural areas. Meanwhile, aspiring migrants are most likely to rely on brokers from their own communities to navigate the bureaucratically and administratively challenging migration process at a premium. Other than damage to their reputation, these brokers do not bear any costs or consequences for misconduct that may include price setting, excessive fees, as well as failure to provide migrants with key information on contracts. 84 Similarly, licenses are renewed for a period of two or one year(s) if an OEP mobilizes 10–49 workers and up to 10 workers, respectively. 82 83 84 Address: 17 76 G S t NW, Washington, DC 20006 Website: ht tp://w w w.worldbank .org/en/topic/jobsanddevelopment Twit ter : @W BG_ Jobs Blog: ht tp://blogs.worldbank .org/jobs