© 2023 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; internet: www.worldbank.org Some rights reserved 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 Exec- utive 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 immuni- ties 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:/ - /creative� commons.org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, dis� - tribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: World Bank. 2023. Pakistan Human Capital Review: Building Ca- pabilities throughout Life. Washington, DC: World Bank. 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 attri- bution: 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 attri- bution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adapta- tion 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 with- in the work. The World Bank therefore does not warrant that the use of any third-party-owned individual compo- nent 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 reuse a component of the work, it is your responsi- bility to determine whether permission is needed for that reuse 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; email: pubrights@worldbank.org. ISBN: 978-0-9635254-7-5 Photographs on pages 3, 15, 41, 42, and 130 by Insiya Sayed. Photograph on page 157 by SS 360/Shutterstock. Contents Foreword ix Acknowledgments xi Abbreviations xiv Policy highlightsxv Numerical highlights xvii Overview 1 1 Pakistan’s human capital: Opportunities amid crises 25 Summary 25 Introduction 26 Pakistan’s persistent human capital crisis 27 The importance of human capital for growth 30 Conclusion and recommendations 36 2 Unpacking Pakistan’s human capital crises 43 Summary 43 Introduction 44 The evolution of Pakistan’s human capital 44 Inequalities in human capital formation 53 Why does Pakistan underutilize its human capital? 58 Conclusion and recommendations 58 3 Fostering early childhood development 65 Summary 65 Contents | Pakistan Human Capital Review Introduction 66 Early childhood development outcomes are low 66 Multiple risk factors impede early childhood development and learning 68 Household-level stress and early childhood development in the era of COVID-19 77 High cost of inaction: early childhood education 80 The existing policy and programming environment is undermined by implementation shortfalls 82 Accelerating progress to improve early childhood development outcomes 90 Conclusion and recommendations 91 iii 4 Boosting nutrition 101 Summary 101 Introduction 102 Malnutrition situation and trends 105 Determinants of stunting 109 Nutrition financing 111 Nutrition stewardship and institutional arrangements 114 Addressing constraints to program effectiveness 121 Conclusion and recommendations 125 5 Providing education to all children 131 Summary 131 Introduction 131 Children out of school 132 Factors affecting school enrollment 138 Providing education to all children 138 Giving children a better education 143 Conclusion and recommendations 147 6 Improving labor market outcomes of the poor 157 Summary 157 Introduction 157 Low unemployment but poorer labor market outcomes for youth, women, and the less educated 158 Employment in Pakistan is overwhelmingly informal 163 Gender and household status in informal settings dampen the benefits of human capital investments 165 Improving labor market outcomes­ —­economic inclusion programs and beyond 170 Conclusion and recommendations 172 Annexes 178 About the authors and contributors 204 Contents | Pakistan Human Capital Review iv Boxes 1 Measuring human capital 4 1.1 Human capital drives development 32 2.1 Human capital impacts of the 2022 floods 51 3.1 Acting early to promote protective factors for Pakistani children: What the evidence tells us 69 4.1 Malnutrition: Key definitions 102 4.2 Stunting reduction in exemplar countries 103 4.3 Indicators used for assessing the adequacy of food, environmental health, and care 111 4.4 The Scaling Up Nutrition Movement 116 4.5 The biggest impediment to nutrition program effectiveness and impact is implementation 119 6.1 The COVID-19 pandemic has adversely affected Pakistan’s labor market through a sudden drop in labor demand and an increase in unemployment 159 Figures 1 Pakistan’s Human Capital Index value is among the world’s lowest and lower than expected given its level of economic development 1 2 Building, protecting, and deploying human capital 2 3 Pakistan’s Human Capital Index value falls dramatically when adjusted for utilization, particularly for women 6 4 Gender gaps on the basic utilization-adjusted Human Capital Index remain wide 6 5 Can Pakistan harness its two distinct demographic dividends? 7 6 Distribution of eight categories of adequacies in children younger than 24 months in Pakistan 10 7 Trends in prevalence of stunting, underweight, and wasting in children under 5 in Pakistan, 1990–2018 11 8 Parents report that many children age 3 and 4 are not developmentally on track 12 9 Gaps in enrollment in early childhood education between children age 3–5 in the lowest and highest income quintiles, in Pakistan and by province 13 10 More girls than boys are out of school 15 11 Four-fifths of wageworkers are informally employed 19 12 Almost 90 percent of wageworkers, formal and informal, are employed in microenterprises 19 13 The economic benefits of actions to build human capital are substantial 21 1.1 Pakistan has a young population, which is growing much faster relative to its peers 26 1.2 Pakistan’s Human Capital Index value is among the world’s lowest and lower than expected given its level of economic development 27 1.3 Pakistan’s Human Capital Index value is low even among the richest quintile, relative to the average for its peers 28 Contents | Pakistan Human Capital Review 1.4 Pakistan’s utilization-adjusted Human Capital Index value is lower than expected given its economic development 29 1.5 Labor force participation of women remains low, though it is growing for those with postsecondary education 30 1.6 Gender gaps on the basic utilization-adjusted Human Capital Index remain wide 30 1.7 Pakistan’s basic utilization-adjusted Human Capital Index value for women is lower than that in peer countries 31 1.8 Pakistan’s Human Capital Index value falls dramatically when adjusted for utilization, particularly for women 31 1.9 The Human Capital Index gains that Pakistan could make if it improves all components over the next 25 years 33 v 1.10 Improving the Human Capital Index: Component contributions 34 1.11 Can Pakistan harness its two distinct demographic dividends? 35 1.12 The fall in fertility that Pakistan needs to accelerate 35 1.13 Government education and health spending in Pakistan is below that in comparator countries 36 1.14 Real absolute investment in public education, health, and social protection has risen in Pakistan since 2000, but changed little since 2007 36 1.15 The economic benefits of building human capital will be substantial 37 2.1 Pakistan’s Human Capital Index value increased very slowly from 1990 to 2021 44 2.2 Pakistan’s Human Capital Index growth is similar to that of peer countries, but starting from a lower base 45 2.3 The period 1990–2012 saw the highest relative improvements in schooling, and under-5 survival increased the most in 2012–17 45 2.4 Expected years of schooling contributed the most to Pakistan’s Human Capital Index growth between the two periods, but at a declining rate 45 2.5 Pakistan underperforms on all Human Capital Index components except adult survival rate 46 2.6 Pakistan’s vaccination coverage for children under 5 fell from January to April 2020 49 2.7 The COVID-19 pandemic may have wiped out more than eight years of gains on Pakistan’s Human Capital Index 50 2.8 Girls’ outcomes are worse than boys in access to school, but not in other components of the Human Capital Index 54 2.9 Girls enroll in much lower numbers than boys 54 2.10 Components of the Human Capital Index by wealth quintiles 55 2.11 The Human Capital Index gap between the richest and poorest groups has persisted over the past three decades 56 2.12 Pakistan Human Capital Index values by gender and wealth 56 2.13 Human Capital Index trend in Pakistan, by province 57 2.14 Human Capital Index components in Pakistan’s provinces vary little on survival but widely on stunting 57 2.15 In three decades, rural areas have closed the Human Capital Index gap with urban areas a little 58 2.16 Human Capital Index components in Pakistan by economic development 59 2.17 Labor force participation of women continues to grow for women with post-secondary education 60 2.18 For women especially, having a postsecondary education means a higher chance of employment 60 3.1 Parents report that many children age 3 and 4 are not developmentally on track 67 3.2 Differing trajectories of human development as a function of exposure to protective and Contents | Pakistan Human Capital Review risk factors 69 3.3 Early childhood education enrollment of children age 3–5 falls far short of universal access, in Pakistan and by province, 2019–20 71 3.4 Gaps in enrollment in early childhood education between children age 3–5 in the lowest and highest income quintiles, in Pakistan and by province 71 3.5 Gaps in enrollment in early childhood education between children age 3–5 in rural areas and those in urban areas, in Pakistan and by province 72 3.6 Gaps in enrollment in early childhood education are large between children age 3–5 with a functional disability and those without one, in Pakistan and by province 72 3.7 Rates of early stimulation by caregivers in provinces in Pakistan are low, children age 2–5 73 vi 3.8 Rates of early stimulation by caregivers and access to playthings in the household are low across provinces in Pakistan, children age 3–5 73 3.9 Rates of early stimulation by caregivers and access to resources in the household are low across provinces in Pakistan, children age 0–2 74 3.10 Use of harsh discipline is common in Pakistan among children age 0–2 and 3–5 74 3.11 Even very young children often lack supervision 75 3.12 At age 3–5, girls and boys are equally likely to experience multiple risks to on-track developmental outcomes, Punjab province 75 3.13 Gender gaps in education attainment disadvantaging girls begin to emerge in early childhood education enrollment in Pakistan and across provinces, children age 3–5 76 3.14 Rates of multiple exposure to the most prevalent risk factors are high among children age 3 and 4 in Punjab and Sindh 76 3.15 The likelihood of being developmentally on track declines with the number of risk factors, children age 3 and 4 77 3.16 Across Pakistan, children age 3 and 4 who enroll in early childhood education are much more likely to be rated developmentally on track 78 3.17 Children under 3 in high-stimulation households have higher levels of child development than children in low-stimulation households in Pakistan 79 3.18 School readiness scores increase with access to early stimulation, maternal education and enrollment in early childhood education among children age 5 80 3.19 Parents who had lost income since 2020 report higher distress than parents who had not lost income 81 3.20 Not raising early childhood education enrollment rates is costly in the long run 82 3.21 Progress toward universal basic hygiene from 2015–20, among countries with less than 99 percent coverage in 2020 84 3.22 Inequalities by wealth quintile in water, sanitation, and hygiene coverage in Pakistan 84 3.23 Private providers of early childhood education serve large shares of children age 3–5 in Pakistan provinces, 2019–20 85 3.24 After rising between 2004 and 2018, preprimary enrollment began to decline in Pakistan among children age 3–8 85 3.25 Enrollment among children age 3–8 rises by age but is very low for children age 3–5 in Pakistan, 2019–20 86 3.26 More than half of 5-year-old children in Pakistan are out of school, 2019–20 86 3.27 Access to essential interventions among young children in Punjab 88 3.28 Access to essential interventions among young children in Sindh 89 3.29 Investing across sectors to promote early childhood development will accelerate progress in human capital development in Pakistan 91 Contents | Pakistan Human Capital Review 4.1 Top 10 risk drivers of deaths and disabilities in Pakistan, 2019 105 4.2 Key nutrition indicators for children under 5 in South Asian countries 106 4.3 Trends in prevalence of stunting, underweight, and wasting in children under 5 in Pakistan, 1990–2018 107 4.4 Stunting prevalence in Pakistan, rural versus urban areas 107 4.5 Stunting by wealth quintiles among children under 5 in Pakistan 108 4.6 Stunting by age in months in Pakistan 108 4.7 The Nutrition Causal Framework 110 4.8 Distribution of eight categories of adequacies in children younger than 24 months in Pakistan 112 4.9 Expenditure trends for nutrition-specific and nutrition-sensitive interventions 113 4.10 Federal nutrition spending by thematic sectors, FY17/18–19/20 113 vii 4.11 Nutrition budget absorption rate by nutrition category, federal level, FY17/18–19/20 114 4.12 Nutrition spending in Punjab and Sindh, FY17/18–19/20 115 4.13 Nutrition spending by thematic sector in Punjab and Sindh, FY17/18–19/20 115 4.14 Nutrition budget absorption capacity, Punjab and Sindh, FY17/18–19/20 116 5.1 More girls than boys are out of school 133 5.2 Rural children are more likely to be out of school 134 5.3 The distribution of out-of-school children by province changed between 1998 and 2017 135 5.4 The older the child, the less likely to be in school 136 5.5 Only 1 student in 3 entering the education system finishes secondary school 136 5.6 Many children have never been to school, and large numbers drop out 137 5.7 Why so many children do not go to school in Pakistan 139 5.8 Improving both access and quality reduces learning poverty faster 142 6.1 Pakistan’s unemployment before the COVID-19 pandemic was generally much lower than the regional average 160 6.2 Pakistan’s labor force participation rate shows wide gender gaps 160 6.3 While Pakistan’s overall labor force participation rate is similar to the South Asia average, the youth cohort is less so 160 6.4 Gender gaps in labor force participation and not in employment, education, or training rates are similar across provinces, but greatest for Balochistan 161 6.5 Unemployment rates are higher for better educated workers in Pakistan 162 6.6 In Pakistan’s youth age cohort, unemployment are higher in urban areas and climbed faster than in rural areas 163 6.7 Four-fifths of wageworkers are informally employed 163 6.8 Almost 90 percent of wageworkers, formal and informal, are employed in microenterprises 164 6.9 Informality rates are higher among youth wageworkers in urban areas than among urban wageworkers generally 165 6.10 Gaps in income by gender and formality status are wide in Pakistan 166 6.11 Formally wage-employed women in Pakistan are the best educated among the four groups 166 6.12 Socioemotional skills by education level in Pakistan 168 6.13 Economic inclusion programs boost not just livelihoods of the poor but also education, health, and nutrition outcomes 171 Tables 1.1 Pakistan’s Human Capital Index value compares poorly with those of other world regions 28 2.1 Disruptions to essential services mean fewer people receive services 49 4.1 Federal expenditure on nutrition interventions, FY19/20 112 4.2 Nutrition public spending in Punjab and Sindh (FY19/20) 114 Contents | Pakistan Human Capital Review 4.3 Key functions of the World Health Organization health system stewardship 117 5.1 Distribution of out-of-school children age 5–16 in Pakistan, by gender, 2019/20 132 5.2 Balochistan has the country’s highest concentration of out-of-school children, 2018/19 135 5.3 The sticker price of having all children in school under different scenarios—lower bounds 140 5.4 Matrix to guide sustainably reducing out-of-school children and learning poverty 146 6.1 Distribution of workers by employment type, 2001 and 2019 165 6.2 Mincerian regressions with log of monthly earnings as the dependent variable, Pakistan 167 6.3 Probit estimations of the probability of being in unpaid work or self-employment, Pakistan 167 6.4 Heckman selection model results to illustrate the relationship between monthly wages and various characteristics including personality traits and cognitive skills 169 A5.1 Most public schools in Pakistan are for primary education 192 viii Foreword The role of human capital in sustainable growth of their potential. Pakistan’s human capital is and development is well recognized, but the pro- low in both absolute and relative terms and cess of designing policies and building institu- has improved only marginally over the past tions that foster human capital accumulation can three decades. Inequalities in human capi- be complex. The benefits of investing in human tal outcomes have persisted or even widened capital often take a long time to materialize and between the rich and poor, men and women, are not always immediately visible, leading to and rural and urban areas and among the underinvestment and to disappointing outcomes. provinces. While Pakistan has made progress In 2018, the World Bank Group launched the in human development over the past sever- Human Capital Project to address this incentive al decades, the agenda remains unfinished: problem through advocacy, measurement, and more than 20 million school-age children are analytical work to raise awareness, increase de- not in school, 40 percent of children under 5 mand for human capital investments, and inform are stunted, and 65 percent of primary school policy action. The Project aims to support human children are failing to achieve minimum profi- capital development as a core element of a coun- ciency in reading. The COVID-19 pandemic has try’s overall strategies to increase productivity deepened these challenges and reversed a de- and growth, and to achieve rapid progress toward cade of hard-won gains in human capital. The a world in which all children can achieve their full adverse impacts of the 2022 floods on house- potential. holds and social services further increased malnutrition and learning poverty in the affect- As part of this effort and building on the lessons to ed districts. date, this Human Capital Review (HCR) assesses the challenges and opportunities to improve Pa- The HCR shows that Pakistan can realize sub- kistan’s human capital outcomes. Several priority stantial economic growth by bringing its popula- outcomes that require improvement­ —­including tion growth rate under control, investing more in early childhood development, child malnutrition, the supply and quality of health and education, out-of-school children, learning poverty, and labor and bringing women to the labor force. For ex- market outcomes of the poor­ —­are analyzed to ample, if Pakistan invests in human capital at the inform human capital development policies. To level of its aspirational peers, while ensuring its Foreword | Pakistan Human Capital Review provide a richer understanding of Pakistan’s de- full utilization over the next 25 years­—­to 2047, velopment trajectories, the HCR adopts a long the 100th anniversary of its founding­ —­its GDP time frame in its analysis. It identifies policies and per capita could grow eight times more than institutions that may prove crucial to improving under business as usual. To make this a reality, human capital outcomes, while underscoring the the HCR recommends that Pakistan exercise interdependence among the multiple sectors fun- effective stewardship over its human capital, damental for human capital accumulation over the declare emergencies over its health and educa- life cycle. tion crises, and ensure long-term commitment that goes beyond the tenure of any government The HCR estimates that children born in Paki- or political cycle. Pakistan should make fam- stan today would be able to achieve 41 percent ily planning a priority across all development ix initiatives. And it should invest more in its peo- remarkably low human capital with a fierce ple and create avenues for them to deploy their sense of urgency. The World Bank is commit- human capital more productively. ted to supporting Pakistan in this journey and to helping give its children the opportunity to We hope this report provides impetus for all achieve their full potential and thrive now and in stakeholders to pay more attention to Pakistan’s the economy of the future. Najy Benhassine Lynne Country Director, Sherburne-Benz Pakistan Regional Director, The World Bank Human Development South Asia Region The World Bank Foreword | Pakistan Human Capital Review x Acknowledgments The Pakistan Human Capital Review (HCR) is animations used to highlight the findings of the a part of the World Bank Group’s Human Cap- HCR and in their dissemination. ital Project (HCP), a global effort that supports countries through data, policies, and research The HCR is based on six background studies to accelerate more and better investments in that involved many contributors in and outside people for greater equity and economic growth. the Bank: Pakistan is one of the early adopters of the HCP, which ensured high-level leadership for the • Pakistan’s Human Capital: Opportunities human capital agenda. The HCR puts a spotlight amid Criseswas led by Koen Geven, Amer on the state of human capital in Pakistan, high- Hasan, and Lire Ersado with contributions lights Pakistan’s opportunities and challenges, from Yi Ning Wong (Consultant) and Tianshu and puts forth recommendations to take the Guo (Consultant). Ahmed Raza (Consultant) next steps in the country’s growth and develop- and Mahin Tariq (Consultant) provided excel- ment trajectory by investing in its most impor- lent research assistance and data analysis. tant resource: the Pakistani people. The team appreciates Moritz Meyer (Senior Economist) for useful comments and Joao The HCR was prepared by a core team led by Pedro Azevedo (Lead Economist) for work on Lire Ersado (Practice Leader) and comprising learning poverty estimates. Amer Hasan (Senior Economist), Ashi Kathu- ria (Senior Health Specialist), Juan Baron (Se- • Unpacking Pakistan’s Human Capital Crises nior Economist), Koen Geven (Senior Econ- was led by Koen Geven, Amer Hasan, and Lire  omist), May Bend (Consultant), and S. Amer Ersado with contributions from Yi Ning Wong Ahmed (Lead Economist, Program Leader). (Consultant), Tianshu Guo (Consultant), Izza It was prepared under the guidance of Najy Farrakh (Senior Education Specialist), Elena Benhassine (Country Director, Pakistan), Lynne Roseo (Senior Economist), Juan Baron, and Sherburne-Benz (Regional Director, Human De- May Bend. Ahmed Raza (Consultant) and Acknowledgments | Pakistan Human Capital Review velopment), Stefano Paternostro (Practice Man- Mahin Tariq (Consultant) provided excellent ager, Social Protection and Jobs), Feng Zhao research assistance and data analysis. The (Practice Manager, Health Population and Nutri- team appreciates Moritz Meyer (Senior Econ- tion), and Keiko Inoue (Practice Manager, Edu- omist) for useful comments and Joao Pedro cation). We would like to thank Pakistan Human Azevedo (Lead Economist) for work on learn- Development Administrative and Client Services ing poverty estimates. (ACS) colleagues: Zulfiqar Ali Raza (Senior Pro- gram Assistant), Syed Farrukh Ansar (Senior • Fostering Early Childhood Development w  as Program Assistant), Shahnaz Meraj (Program led by Amer Hasan with contributions from Assistant), and Ayesha Tabassum (Program As- Heather Tomlin­ son (Consultant), Mahreen sistant) for excellent assistance. Special appre- Tahir-­Chowdhry (Consultant), Elizabeth ciation to Zulfiqar for his editorial support and Hent­schel (Consultant), Aisha Yousaf­zai for the instrumental role he played in the cover (Consultant), Mina Za­mand (Education Spe- design and production of a series of videos and cialist), and Amna Ansari (Consultant). The xi authors gratefully acknowledge the  collab- Organization, Nutrition International, and the oration and support of the following indi- Bill and Melinda Gates Foundation­ —­and Ox- viduals: Sohail Saeed Abbasi (Senior Social ford Policy Management for sharing findings Protection Specialist), Farrah Ilyas (UNICEF from its nutrition financial tracking exercise Pakistan), Gul Najam Jamy (Consultant), in Sindh. Sania Khursheed (Consultant), Dr. Ghazala Ra­fique (Consultant), Soofia Yunus (Senior • Providing Education to All Children  was led by Health Specialist), Ziauddin Hyder (Senior Juan Baron and May Bend with contributions Health Specialist), and Mohammad Far­ han­ from Zohair Zaidi (Consultant). The team ullah Sami (Senior Water Supply and Sanita- would like to thank Freya Perry (Education tion Specialist). Advisor, UK Foreign, Commonwealth and De- • Boosting Nutrition  was led by Ashi Kathu- velopment Office) and Akiko Hanaya (Educa- ria with contributions from Manav Bhattarai tion Advisor, Japan International Cooperation (Senior Health Specialist), Jahanzaib Sohail Agency) for useful comments on earlier ver- (Health Economist), Akmal Minallah (Senior sions of the paper. Public Sector Specialist), Mamata Ghimire (Health Economist), Soofia Yunus (Senior • Improving Labor Market Outcomes of the Health Specialist), Ali Mirza (Health Special- Poor was led by S. Amer Ahmed with contri- ist), Aliya Kashif (Senior Health Specialist), butions from Thelma Choi (Consultant) and and Ziauddin Hyder (Senior Health Special- Jyotirmoy Saha (Consultant). Ali Qureshi (So- ist). John Lincoln Newman (Consultant), cial Protection Specialist) and Sakib Sherani Huma Qureshi (Consultant), and Wajiha (Consultant) provided useful comments and Ghias (Consultant) provided inputs based on guidance. Priyanka Kanth (Economist) pro- a review and analysis of programs, policies, vided valuable inputs on existing programs and institutional arrangements for nutrition in Pakistan, and Colin Andrews, Hugo Brous- interventions in Pakistan. The team is grate- set, Sundas Liaqat, and the Partnership for ful for the support of several ministries and Economic Inclusion provided helpful advice departments of the federal and provincial on economic inclusion programs. This paper governments, including the Federal Nutri- benefited from stakeholder consultations tion Wing at the Ministry of National Health with counterparts at the Planning Commis- Services, Regulations and Coordination for sion, the Pakistan Poverty Alleviation Fund, generously sharing useful insights, the Of- the National Vocational and Technical Train- fice of the Controller General of Accounts ing Commission, the Poverty Alleviation and Acknowledgments | Pakistan Human Capital Review for sharing data on nutrition expenditure, Social Safety Division, the Pakistani Micro­ and officials from provincial departments finance Investment Corporation, the Punjab for their thoughts and data on their respec- Social Protection Authority, the Rural Support tive nutrition programs. The team is thank- Program Network, the Punjab Planning and ful to the National Nutrition Survey team at Development, the Ministry of Finance–Khyber Aga Khan University, including Dr. Zulfiqar Pakhtunkhwa, and the Office of the Special Bhutta, Dr. Abdul Baseer Khan Achakzai, Dr. Advisor for Institutional Reform. The team is Khawaja Masuood Ahmed, and Mr. Shehzad grateful for comments from Mr. Haris Gazdar Hasan, for data and overall nutrition program of the Collective for Social Science Research. and policy inputs. The team sincerely appre- ciates the valuable support and engagement The team is also grateful to peer reviewers Colin of development partners­ —­including the Andrews (Program Manager, HSPGE), Eman- Scaling Up Nutrition Secretariat, the United uela Galasso (Senior Economist, DECPI), Luc Nations Children’s Fund, the World Health Laviolette (Practice Leader, HMNDR), Mohamed xii Ihsan Ajwad (Senior Economist, HMNSP), Rafa- Poverty­ —­for support and valuable inputs that el de Hoyos (Lead Economist, HECED), Willem strengthened the evidence base and presenta- Janssen (Lead Agriculture Economist, SSAA1), tion of the HCR. and Yasuhiko Matsuda (Country Manager, EACMA) for useful comments. The team ap- A team at Communications Development preciates comments from participants at the Incorporated­ —l­ed by Bruce Ross-Larson with Pakistan Institute of Development Economics, Joe Caponio, Meta de Coquereaumont, Mike the World Bank Consultative Workshop in Islam- Crumplar, Christopher Trott, and Elaine Wilson, abad, and the Technical Clinic on Human Capital and including Debra Naylor­ —d­ esigned, edited, Reviews in Washington, DC, where preliminary and laid out the report. findings were presented. The team would like to extend its appreciation to many stakehold- The HCR is a product of the staff of the World ers who shared their insights during consulta- Bank Group. The findings, interpretations, and tions and meetings and the many colleagues conclusions expressed in this report do not nec- from across the Global Practices­ particularly —­ essarily reflect the view of the Executive Direc- Education; Health, Nutrition, and Population; tors of the World Bank (or the governments they Social Protection and Jobs; Water; Governance; represent), or the Government of the Islamic Finance, Competitiveness, and Innovation; and Republic of Pakistan. Acknowledgments | Pakistan Human Capital Review xiii Abbreviations ADP Annual Development Plan NCF Nurturing Care Framework B40 Bottom 40 percent NNS National Nutrition Survey BaU Business as usual NSER National Socio-Economic Registry BCC Behavior change communication NSPS National Social Protection Strategy BISP Benazir Income Support Program OOP out of pocket BMI Body mass index PKR Pakistani rupee CCT Conditional cash transfer PMNS Pakistan Multisectoral Nutrition CREDI Caregiver-Reported Early Strategy Development Instrument PSLM Pakistan Social and Living Standards DHS Demographic and Health Survey Measurement Survey ECD Early childhood development RMNCAH&N Reproductive, maternal, newborn, ECE Early childhood education child, and adolescent health and FFP Food Fortification Programme nutrition FY Fiscal year SDG Sustainable Development Goal GDP Gross domestic product SUN Scaling Up Nutrition HCI Human Capital Index TIMSS Trends in International Mathematics HCP Human Capital Project and Science Study HCR Human Capital Review TMIS Tackling Malnutrition Induced HIES Household Integrated Economic Stunting in Pakistan Survey UHCI Utilization-adjusted Human Capital IFMIS Integrated financial management Index information system UN United Nations IQ Intelligence quotient UNICEF United Nations International LAYS Learning-adjusted years of schooling Children’s Emergency Fund LFS Labor Force Survey USI Universal salt iodization LHWP Lady Health Worker Program WASH Water, sanitation, and hygiene NC Nutrition Cluster WHO World Health Organization Abbreviations | Pakistan Human Capital Review xiv Policy highlights Pakistan has reached middle-income status and capital externalities and market failures pro- made significant progress in reducing poverty vide a strong rationale for the government to over the past two decades. But low human capi- invest in human capital. As in most develop- tal development could limit the realization of its ing countries, poor people in Pakistan stand ambition to become an upper-middle-income to lose the most when the public sector fails country by 2047. Pakistan needs a healthy, to make these investments. To boost its skilled, and resilient population to ensure high human capital, Pakistan needs to invest more economic growth that is both sustainable and in the supply of health and education through inclusive. With the right policies and invest- domestic resource mobilization, shifting re- ments, the growing working-age population can sources from costly energy subsidies and im- become healthier, more educated, more skilled, proving efficiency in the existing allocations and more productive­ —a­nd can earn more if to human development sectors. the economy generates more and better jobs. • Improve child development and learning out- This Human Capital Review recommends that comes by integrating high-quality, contextu- Pakistan: ally relevant parenting and behavioral change • Exercise effective stewardship over human programs into existing health, social protec- capital and declare emergencies over its tion, and education platforms. Early inter- health and education crises. This requires ventions, including access to early childhood long-term planning beyond the tenure of education, are vital to mitigate risks and pro- any government and political cycle. With the mote protective factors that shape healthy COVID-19 pandemic and devastating floods brain development, yielding gains for future deepening its already poor state of human health, learning, and productivity. capital, Pakistan needs to declare emergen- • Make nutrition a national priority by focusing cies and take bold actions to tackle its health on strengthening stakeholder momentum and and education challenges. coordination­ —­and on mobilizing financing • Make family planning a priority across all and tracking spending and progress. It should human development initiatives. Pakistan formulate a national nutrition policy, ad- should integrate population planning in ac- dress bottlenecks, and ensure convergence Policy highlights | Pakistan Human Capital Review ademic, religious, and national policies and of sectoral interventions. That policy should develop its labor market to accommodate address the immediate and underlying de- the growing youth population. Key actions terminants of nutrition and prioritize the include training more Lady Health Workers most vulnerable areas and populations with to educate women on family planning, cre- high-impact nutrition-specific interventions. ating more awareness about the use of mod- • Enhance enrollments and learning outcomes. ern contraceptives, increasing the provincial More schools, particularly in rural areas and health budget, and focusing on structural in- staffed with a higher proportion of female equalities in access to health and education. teachers, are needed to encourage girls’ ed- • Invest more and more smartly in people and ucation. In addition, safe transportation; ad- develop avenues for them to deploy their equate water, sanitation, and hygiene facili- human capital more productively. Human ties; conditional cash transfers; and remedial xv learning programs are imperative to reduce require clear identification of the segment of the high number of out-of-school children in the labor market that the program will target, Pakistan. the package of interventions to be bundled • Support the labor market and human capital together, and the use of pre-existing nation- outcomes of the poor through multipronged al systems for targeting such programs as interventions, including those anchored on asset transfers and financial support for cash transfer programs. The interventions entrepreneurship. Policy highlights | Pakistan Human Capital Review xvi Numerical highlights 0.41 Pakistan’s Human Capital Index (HCI) value of 0.41 is low in both absolute and relative terms. It is lower than the lower-middle-income countries. After the pan- demic and the recent floods are accounted for, learning poverty is estimated to have risen to South Asia average of 0.48, with Bangladesh at 79 percent. 0.46 and Nepal at 0.49. Pakistan’s human cap- ital outcomes are more comparable to those in Sub-­ Saharan Africa, which has an average HCI value of 0.40. 2% The main determinants of nutri- tion are food intake, environmental health, and care for children and women, and very few of Pakistan’s children under 2 have ad- 144% If Pakistan continues on its current trajectory in human capital development, its GDP per capita would equacy on all three. The proportion of children with multiple adequacies is low in Pakistan: more than 26 percent of children lack adequa- grow overall by a mere 18 percent through 2047, cy on any of the three determinants, and fewer the 100th anniversary of its founding. If Pakistan than 2 percent of children have adequacy on all can boost human capital investments and its HCI three dimensions. value to the level of its peers, per capita GDP could grow by 32 percent. But if Pakistan improves both its human capital and its use of human capital, bringing adults into employment outside farming, 0.37 The pandemic has likely erased nearly a decade of progress on human capital for both boys and girls. Simula- GDP per capita could rise by 144  percent, eight tions accounting for the pandemic show that times more than under business as usual. Pakistan’s HCI value would be reduced from 0.41 to 0.37, lower than its 2012 level. The main 40% Early childhood development outcomes in Pakistan are low rel- ative to those in middle-income countries. Only culprit is reduction in the quantity of schooling (due to dropouts) and in quality (due to learning loss), both of which have been confirmed by em- Numerical highlights | Pakistan Human Capital Review 40–59 percent of young children, depending on pirical studies. the province, are reported by their parents to be developmentally on track, whereas the average in peer countries is 75 percent. Some 40 percent of children under 5 are stunted, and 18 percent 0.08 Economic gains from human cap- ital can be realized only if people can utilize their skills and ingenuity in productive under 5 are wasted. And fewer than one in five activities, such as gainful employment in the labor children enrolls in early childhood education. market. After utilization in the labor market is adjusted for, Pakistan’s HCI value falls from 0.41 79% An estimated 20.3 million of Pa- kistan’s school-age children are out of school. In addition, Pakistan’s learning to 0.20. There is an enormous gender inequal- ity in utilization of human capital: the utilization-­ adjusted HCI value is 0.31 for men and only 0.08 —­­ poverty rate­­ the percentage of children unable for women. This underscores the low female labor to read and understand a short age-appropriate force participation, at just over 20  percent (and text by age 10­­ stood at 75 percent before the —­­ even lower for those age 15–24), in Pakistan. COVID-19 pandemic and the 2022 floods, more About 60 percent of working-age women are not xvii than 19 percentage points above the average for in employment, education, or training. Overview BUILDING HUMAN CAPITAL to human capital development at all stages of THROUGHOUT THE LIFE CYCLE the life cycle. Disruptions to health services and income losses due to the pandemic have forced Pakistan can realize major economic growth and many to skip regular health checkups and treat- development by investing in its people and their ments, with adverse consequences that are like- human capital. But the reality is that Pakistan’s ly to be felt for decades to come. And school-age human capital is low and has improved only children and youths have similarly experienced marginally over the past three decades. Inequal- huge disruptions in access to education and job ities in human capital outcomes have persisted opportunities. Making matters worse, the 2022 or widened over time between the rich and poor, floods further limited access to health and edu- men and women, and rural and urban areas and cation services and likely increased malnutrition among the provinces. Human capital outcomes and learning poverty.2 are low across the board, with even the most economically advantaged groups in Pakistan These low human capital investments will limit having lower human capital outcomes than less the realization of Pakistan’s ambition to become economically advantaged groups in peer coun- an upper-middle-income country by 2047. If tries. Pakistan’s Human Capital Index (HCI) Pakistan continues on its current trajectory in value of 0.41 is low in both absolute and relative human capital development, its GDP per capita terms (figure 1; see box 1 on the methodology for measuring human capital). It is lower than FIGURE 1  Pakistan’s Human Capital Index value is among the the South Asia average of 0.48, with Bangladesh world’s lowest and lower than expected given its level of economic at 0.46 and Nepal at 0.49. Pakistan’s human development capital outcomes are more comparable to those Human Capital Index value, 2020 in Sub-Saharan Africa, which has an average 1.0 HCI value of 0.40. Pakistan thus remains far from realizing its po- 0.8 tential and forgoes substantial growth and de- Vietnam Sri Lanka Türkiye velopment. Despite favorable demographic Overview | Pakistan Human Capital Review Indonesia 0.6 tailwinds, Pakistan is currently unable to reap Bangladesh Kazakhstan India Malaysia its rightful dividends due to inadequate invest- Nepal Bhutan Afghanistan ments in education, health, social protection, and 0.4 Punjab 0.42 Pakistan 0.41 the labor force. Its public investment of about Khyber Pakhtunkhwa Sindh 0.36 0.39 Balochistan 0.32 2.5 percent of GDP in education and 0.9 percent on health is much lower than the global average 0.2 and the average for similar economies. Pakistan 6 7 8 9 10 11 12 Log of GDP per capita (constant 2010 US$) spends about 0.6 percent of GDP on social safe- ty nets, compared with the global average of Source: Pakistan provincial calculations from 2017 and 2019 Demographic and Health 1.5 percent.1 The COVID-19 pandemic has deep- Surveys. Country estimates from the Human Capital Project. GDP from World Bank na- tional accounts data and Organisation for Economic Co-operation and Development na- ened the challenges, adding a significant threat tional accounts data. 1 FIGURE 2  Building, protecting, and deploying human capital HUMAN CAPITAL LIFE CYCLE Active School labor market infrastructure policies, skilling Teacher training Public insurance Public–private and safety nets Inclusive social Immunization partnerships Women’s pensions and nutrition Conditional cash empowerment A ordable programs transfers Reducing and accessible Early childhood education School feeding harassment healthcare Positive parenting programs Age-smart support infrastructure Prenatal and postnatal Water, sanitation, investments healthcare and hygiene services Micro-nutrient supplements Family planning IN-UTERO AND SCHOOLING YOUTH AND ADULTHOOD AGING AND ELDERLY EARLY CHILDHOOD (Early and lifelong (Healthy, safe, (Skilled, empowered, (Aging with dignity) advantage) and educated) productive, and resilient) would grow overall by a mere 18 percent through Pakistan needs a healthy, skilled, and resilient 2047, the 100th anniversary of its founding. If population to ensure high economic growth that Pakistan can boost human capital investments is both inclusive and sustainable. Bringing more and its HCI value to the level of its peers, per women into education and the labor force will Overview | Pakistan Human Capital Review capita GDP could grow by 32  percent. But if require increased focus on their safety, not just Pakistan improves both its human capital and at school and in the workplace but also in transit its use of human capital, bringing adults into to and from home, while creating more jobs ac- employment outside farming, GDP per capita cessible to them in all parts of the country. With could rise by 144 percent, eight times more than the right policies and investments, the growing under business as usual. working-age population can become health- ier, more educated, more skilled, and more To boost its human capital, Pakistan needs to productive­ —­and can earn more, if the econo- bring population growth under control, invest my generates more and better jobs. This calls more in the supply of health and education, for a deliberate effort from multiple stakehold- and enhance female labor force participation. ers and sectors to build on intrasectoral and 2 intersectoral externalities. For example, healthy and well-nourished children learn better and earn more as adults, and higher levels of educa- tion help improve health outcomes throughout the life cycle as well as improve access to em- ployment opportunities. To enhance its human capital, Pakistan should adopt a life cycle approach to building, protect- ing, and deploying human capital, starting be- fore birth, continuing through early childhood The COVID-19 pandemic and the 2022 development and schooling, culminating in in- floods have deepened these challenges, creasingly productive employment (figure 2). This calls for a long-term commitment, recog- adding a significant threat to human nition of the multidimensional and cumulative capital at all stages of the life cycle. nature of human capital investments, deliberate efforts from multiple stakeholders and sectors to build on intersectoral linkages, and a con- competition is intensifying, only countries that tinuity of policies across political parties and enable their citizens to consistently create and governments. Many countries previously at Pa- productively apply knowledge can thrive. And kistan’s level of development have managed to as the disruptive impacts of the COVID-19 pan- do precisely this, even with regional variations demic and the 2022 floods continue to unfold in and gaps just as large. Pakistan has the tools to Pakistan, strategic policy decisions backed by implement the recommendations in this report, evidence are more vital than ever to protect and provide stewardship for human capital invest- strengthen human capital outcomes. ments, and enhance economic growth over the long term. Pakistan’s handling of the COVID-19 Key issues pandemic has shown that the country can man- age complex challenges, despite its institutional Human capital makes up 61  percent of Paki- constraints.3 stan’s wealth, yet its levels of human capital are among the world’s lowest. About 7  percent of newborns in Pakistan do not live to their fifth BUILD HUMAN CAPITAL birthday. Around 40 percent of children under 5 are stunted, relegating them to a lifetime of Human capital is a central driver of sustainable physical and cognitive deficits. Prior to the growth and development, with more human COVID-19 pandemic and the 2022 floods, an es- Overview | Pakistan Human Capital Review capital associated with higher earnings for peo- timated 75 percent of Pakistani children were in ple and higher income for countries. Human learning poverty, unable to read and understand capital accounts for an estimated 64 percent of a simple age-­ appropriate story by age 10. Dis- global wealth­—­ among high-income countries, parities across socioeconomic and geographic 70  percent; among lower-middle-income coun- groups are large. Together, these human capital tries, 51 percent; among upper-middle-income shortfalls and equity gaps explain Pakistan’s low countries, 58 percent; and among low-income earnings, low economic growth, and low poverty countries, 41 percent. In a global environment reduction. where natural and human-made socioeconomic shocks are proliferating, global markets are shift- On most components of the HCI, Pakistan ing, technologies are ever-changing, and global performs the South Asia region and even under­ 3 BOX 1 Measuring human capital Imagine the trajectory from birth to adulthood of a child those in other countries, despite being in school for a born in Pakistan today. There is a risk that the child may similar amount of time. This measure is the same as not survive to her fifth birthday. Even if she does reach learning-adjusted years of schooling. The HCI defines school age, there is a further risk that she does not start a complete, quality education as 14 learning-adjusted school, let alone complete the full cycle of preprimary, pri- years of schooling, which serves as a benchmark for mary, and secondary education that is the norm in more comparing performance. advanced countries. The time she does spend in school may translate unevenly into learning. When she reaches • Health: This component uses two indicators for a age 18, she carries with her lasting effects of poor nutri- country’s overall health environment: the prevalence of tion and health in childhood that limit her physical and stunting among children under 5 and the adult surviv- cognitive abilities as an adult. al rate, defined as the proportion of 15-year-olds who will survive until age 60. The first indicator reflects the The Human Capital Index (HCI), developed as part of health environment experienced during prenatal, in- the World Bank’s Human Capital Project, measures the fant, and early childhood development; the second re- amount of human capital that a child born today can ex- flects the range of health outcomes that a child born pect to attain by her 18th birthday, given the risks of in- today may experience as an adult. complete education and poor health that prevail in her birth country. It highlights how investments that improve The dimensions of human capital are complementary education and health outcomes today will affect the pro- and start interacting with each other early in life. Each ductivity of the next generation of workers. The HCI mea- component is interpreted in terms of its contribution to sures current education and health outcomes since they worker productivity, relative to a benchmark correspond- can be influenced by current policy interventions to im- ing to complete education and full health. The resulting prove their quantity and quality. The HCI quantitatively il- overall index measures the productivity of a worker rela- lustrates the key stages in the trajectory described above tive to this benchmark. The HCI ranges from 0 to 1, and a and their consequences for the productivity of the next value of x means that a worker of the next generation will generation of workers, in a summary measure. The HCI be only x × 100 percent as productive as she would have consists of three components: been under the benchmark of complete education and full health. Equivalently, the gap between x and 1 measures • Survival: This component reflects the fact that chil- the shortfall in worker productivity due to gaps in educa- dren born today need to survive until human capital tion and health relative to the benchmark. Overview | Pakistan Human Capital Review accumulation through formal education can begin. Survival is measured using the under-5 mortality rate. The HCI captures the size of the income gains when to- day’s children become tomorrow’s workers. But this as- • Education: This component combines information on sumes that when today’s child becomes a future worker, the quantity of education a child can expect to obtain —­ she will be able to find a job­ which may not be the case by age 18 with a measure of quality: how much children in countries with low employment rates. Moreover, even if learn in school based on countries’ relative perfor- today’s child can find employment in the future, she may mance on international student achievement tests. By not be in a job where she can fully utilize her skills and adjusting for quality, this component reflects the real- ity that children in some countries learn far less than (continued) 4 BOX 1, continued cognitive abilities to maximize her productivity. To ad- for the possibility that when today’s child becomes a dress these and to complement the HCI, the report worker in the future, she may not be able to fully use her includes two utilization-adjusted Human Capital Indi- skills and cognitive abilities even if she finds a job. The ces (UHCIs) that adjust the HCI for underutilization of full UHCI also assigns a basic utilization value to workers human capital in the labor market. The basic UHCI ad- without any human capital because they contribute pro- justs a country’s HCI value for labor force participation ductively to the economy. The full UHCI may thus be a (the share of workers in each age bracket who are work- better measure of the productivity of human capital than ing or actively looking for work). The full UHCI adjusts the basic UHCI. the average for Sub-­Saharan Africa. On stunting Economic gains from human capital can be re- and the quality of education, it performs below alized only if people can use their skills and in- the average for Sub-­ Saharan Africa. Child sur- genuity in productive activities, such as gainful vival rates until age 5 are far below the South employment in the labor market. The basic Asia regional average and the same as the av- utilization-adjusted HCI (UHCI)5 shows that erage for Sub-­Saharan Africa. While children in the utilization of human capital is remarkably Pakistan stay in school slightly longer than their low, especially for women. Adjusting Pakistan’s peers in Sub-­ Saharan Africa, a 2019 regional HCI value for the utilization of human capital assessment ranked Pakistan second from the reduces it from 0.41 to 0.20 (figure 3). A child bottom globally in science and mathematics born in Pakistan today can expect to utilize only performance by its 4th graders. Against peer 20 percent of their human capital potential. This and regional comparators, the only indicator means that the benefits that Pakistan can real- on which Pakistan outperforms is adult survival ize from its human capital are further depressed rate. because it does not have an environment for citizens to put their human capital to work. Provincial inequalities are wide on the HCI and Further ringing the alarm bell, the UHCI value its components, notably for stunting and ex- is only 0.08 for women. In other words, a baby pected years of schooling. In schooling, Khyber girl born in Pakistan today can expect to utilize Pakhtunkhwa and Punjab are at the expected only 8  percent of her human capital potential level for Pakistan’s economic development, by the time she turns 18. Gender inequality in with the average child growing up with at least 9 the utilization of human capital has largely per- years of formal schooling­ —­though when adjust- sisted over the past three decades, with basic ed for quality, this is equivalent to only 5.1 years.4 UHCI values increasing by 3 percentage points Sindh and Balochistan are at much lower levels, for men and by 4 percentage points for women Overview | Pakistan Human Capital Review with the average child growing up with around (figure 4). 7.7 and 5.4 years of schooling, respectively. With only 89 percent of children surviving until their A second adjustment to the HCI, the full UHCI fifth birthday, Balochistan is at the very bottom­ accounts for the possibility that when today’s —­ globally. The trajectory of child development child becomes a worker in the future, she may and human capital accumulation for children not be able to fully utilize her skills and cogni- born in Balochistan and Sindh is further under- tive abilities to increase her productivity even if mined by the fact that almost half of the children she finds a job. The full UHCI value is designed under 5 in the two provinces are stunted. Even in to capture the productivity gains more closely Punjab, where stunting is the lowest among the and excludes agricultural employment. It is es- provinces, one-third of children are stunted. timated at 0.26 for Pakistan, slightly higher than 5 FIGURE 3  Pakistan’s Human Capital Index FIGURE 4  Gender gaps on the basic value falls dramatically when adjusted for utilization-adjusted Human Capital Index remain utilization, particularly for women wide Human Capital Index (HCI) value Utilization-adjusted Human Capital Index value Basic utilization-adjusted HCI value Full utilization-adjusted HCI value 0.4 0.5 Male 0.31 0.4 0.4 0.3 0.4 0.4 0.28 0.3 0.3 0.3 Overall 0.2 0.19 0.3 0.16 0.2 0.2 0.2 0.1 0.08 Female 0.1 0.04 0.1 0.0 0.0 1990 2012 2017 2019 Overall Male Female Source: World Bank calculations using data from Demographic Source: World Bank calculations using data from Demographic and Health Surveys, Household Income and Expenditure Surveys, and Health Surveys, Household Income and Expenditure Surveys, Pakistan Social and Living Standards Measurement Surveys, and Pakistan Social and Living Standards Measurement Surveys, and Labor Force Surveys. Labor Force Surveys (see Annex 1B). Note: Basic utilization-adjusted HCI represents all employment. Full utilization-adjusted HCI denotes nonagricultural employment. women are ill-equipped to obtain better-paying jobs. Marriage norms (with women expected the basic HCI value (see figure 3). The gap be- to withdraw from the labor for force after mar- tween the two measures means that the larger riage) imply that their parents would not reap problem facing Pakistan is underemployment the benefits of investing in their education. (particularly for women) relative to job opportu- nities in the labor market.6 This becomes appar- Pakistan is the world’s fifth most populous ent when breaking the measure down by gen- country, with about 230 million people. An “early der. For men, there is no difference between the demographic dividend” country with immense two UHCI values, whereas women, when they potential for accelerated economic growth from do enter the labor market, tend to secure more the changes in its population age structure, Pa- productive employment, attributable to the rel- kistan can reap dividends from two phases of atively large share of highly educated women the demographic transition (figure 5). The first joining the labor market. dividend can be captured as the demograph- Overview | Pakistan Human Capital Review ic transition speeds up and the population age The strongest driver of low human capital utili- structure becomes more concentrated among zation in Pakistan is low female labor force par- those of working age. The right policies and in- ticipation rate­—­23 percent in 2018. Having more vestments in human capital can lead to youth than three in every four women outside the cohorts that are healthier, more educated, more labor force underscores the need to make better skilled, and more productive and that can earn use of the human capital that Pakistan gener- higher incomes if the economy can create good ates now and in the future. Women’s low returns jobs for them. to education at the primary and secondary lev- els may reduce incentives for parents to invest The second phase comes along later in the de- in their daughters’ schooling. Poorly educated mographic transition if Pakistan can increase 6 FIGURE 5  Can Pakistan harness its two distinct demographic dividends? Total dependency ratio FIRST DIVIDEND SECOND DIVIDEND Pakistan • Permanent increase in • Accumulation capital–worker of human ratio and • More and physical output per capita disposable capital • More • More income to production • More people workers save in working age Demographic transition Source: World Bank elaboration. savings and investments as a result of fewer de- health and education crises. This requires pendent children, more disposable income, and long-term planning beyond the tenure of prospects for longer lives. As this generation any government and political cycle. With becomes older, and more children follow, the the COVID-19 pandemic and devastating country has both a tremendous opportunity and floods deepening its already poor state of responsibility to reinforce and solidify human human capital, the Government of Pakistan capital development. Seizing the second demo- needs to declare emergencies and take bold graphic dividend through human and physical actions to tackle its health and education capital accumulation and higher labor produc- challenges. tivity is going to be key to deal with the challeng- • Make family planning a priority across all es presented by an aging population. human development initiatives. Pakistan should integrate population planning in ac- However, Pakistan is currently m ­ issing the op- ademic, religious, and national policies and portunity to capitalize on its potential demo- develop its labor market to accommodate graphic dividend. If existing trends continue, it the growing youth population. Key actions will miss out on the associated economic devel- include training more Lady Health Work- opment possible from a potential demographic ers to educate women on family planning, Overview | Pakistan Human Capital Review dividend in the future as well. This is due to a creating more awareness about the use confluence of low human capital, low labor force of modern contraceptives, increasing the participation, and low productivity, as well as a provincial health budget, and focusing on pace of job creation that is slower than popula- structural inequalities in access to health tion growth. In short, Pakistan’s poor foundation and education. for human capital development is imperiling its • Invest more and more smartly in people and ability to capitalize on its demographic dividend. develop avenues for them to deploy their human capital more productively. Human High-level recommendations capital externalities and market failures • Exercise effective stewardship over human provide a strong rationale for the govern- capital and declare emergencies over the ment to invest in human capital. As in most 7 7% 40% 79% About 7 percent of Around 40 percent of In the aftermath of the newborns in Pakistan children under 5 are pandemic and 2022 floods, do not live to their stunted, relegating them an estimated 79 percent fifth birthday. to a lifetime of physical of Pakistani children are and cognitive deficits. in learning poverty, unable to read and understand a simple age-appropriate story by age 10. developing countries, poor people in Pakistan creates a cycle of unmet need for basic ser- stand to lose the most when the public sector vices in health and education and for jobs. fails to make these investments. To boost its Existing family planning programs are not ef- human capital, Pakistan needs to invest more fective, with key gaps in supply and demand. in the supply of health and education through Contraceptive usage has stagnated, and con- domestic resource mobilization, shifting re- traceptive prevalence in age 15–49 is low, at sources from costly energy subsidies and im- 34 percent, with only 26  percent using mod- proving efficiency in the existing allocations ern contraceptive methods. Pill usage among to human development sectors. married women age 15–49 is low, at 2 percent, versus 25 percent in Bangladesh, for example. About 17 percent of women age 15–49 report FOSTER EARLY CHILDHOOD unmet needs for family planning services,7 with DEVELOPMENT 10 percent wanting to space births and 8 per- cent wanting to limit births. ...by starting before birth Pregnant women need more support to pre- Key issues pare for healthy childbirth. Antenatal care has been reported to reach 86  percent, but 7 in The fertility rate in Pakistan is much higher 10 women still report difficulties accessing than in comparator countries and limits Pa- healthcare for themselves. Immediate postna- kistan’s development potential. High fertility tal practices could be improved dramatically Overview | Pakistan Human Capital Review with early support. However, only 11 percent of babies have skin contact with the mother after delivery, only 20 percent of children are breast- fed within 1 hour of birth, and only 48 percent are exclusively breastfed during the first six months of life. Micro­nutrients, such as iron and folic acid, need to be universalized during preg- nancy. In Pakistan, only 29  percent of preg- nant women take iron supplements, and about 43  percent of all women of reproductive age are anemic. 8 Malnutrition is one of Pakistan’s biggest challenges, with long-term implications for human capital development and economic growth. High-level recommendations change will be needed if Pakistani children are • Promote demand for family planning methods to attain their full growth and development toward men and women. Continue behavioral potential. change campaigns at the national level, com- plemented by community-level communica- Stunting has lifelong and even intergenerational tion for men and women. Target information consequences for individuals­ —­ and nations. As on birth-spacing, limiting family size, fertile an indicator of chronic malnutrition, it is most periods, and low cost modern contraceptive closely associated with brain development, methods and their side effects. Create high- physical growth, and human capital develop- trust information sources on the internet and ment. Childhood stunting leads to increased social media, making it easy to understand mortality, increased morbidity (in childhood for men and women in all local languages. and later in adulthood), decreased cognitive • Increase the supply of long-lasting modern ability, a delayed start and reduced attainment contraceptives. Make pills, intrauterine devic- at school, and far lower individual earnings and es, implants, and injectables universally avail- slower national economic growth. Stunted chil- able. Instill procedures for frontline providers dren are a third less likely than other children to to discuss family planning when women visit escape poverty as adults. Childhood stunting clinics. has intergenerational implications: malnour- • Focus more on pregnant women. Further ished mothers are more than twice as likely increase coverage of antenatal provision. as well-­ nourished mothers to have stunted Provide better information on pregnancy at children. delivery points. Train providers and monitor service delivery. Food intake, environmental health, and care Overview | Pakistan Human Capital Review for children and women are the main determi- …by boosting nutrition nants of malnutrition at the most immediate level.8 Factors such as income poverty, gender, Malnutrition is one of Pakistan’s biggest devel- and education underpin all three, alongside the opment challenges, with long-term implications larger political, economic, social, and cultural for human capital development and economic environments. But the adequacy of these three growth. It remains the top cause of death and categories of indicators has a strong associa- disabilities in the country. And while there has tion with how well-nourished children will be. been progress in reducing childhood malnutri- Children can receive adequate feeding, environ- tion, the pace has been slow, despite two de- mental health, and care­—­or none, or anything in cades of national and provincial efforts. Faster between. Very few of Pakistan’s children under 9 2 have adequacy in all three determinants, and Among adolescents, about 7  percent of boys the proportion of children with multiple adequa- and 5 percent of girls are obese.9 cies is low. More than 26 percent of children lack adequacy on any of the three determinants, and The state of malnutrition in Pakistan is worse fewer than 2 percent of children have adequacy than in all other countries in South Asia and on all three dimensions (figure 6). Overall, Paki- has improved little over the past half century. stan comes up short in providing adequate feed- More generally, malnutrition indicators among ing, environmental health, and care. The strong children under 5 (stunting, underweight, wast- association between nutrition outcomes and the ing, overweight) are worse in Pakistan than in adequacy of food intake, environmental health, other South Asian countries. Malnutrition is the and care suggests that it is critical to ensure number one risk factor for death and disabili- that all three factors reach every mother and ties in Pakistan across all ages. In particular, 7 child during the narrow window of opportunity. out of 100 children do not live past the age of stan compared with 4 out of 100 in the 5 in Paki­ Key issues region. Pakistan has a high burden of malnutrition in The national averages mask wide subnational all its forms. One-fifth of children are born with disparities in malnutrition. For example, stunt- low birthweight, and among children under 5, ing is worse among rural and poor households. 29 percent are underweight, 40 percent stunt- In Khyber Pakhtunkhwa, 48 percent of children ed, 17 percent wasted, 54 percent anemic, and under 5 are stunted, and in Sindh, 50 percent. 52 percent deficient in vitamin A (figure 7). The Even in the wealthiest income quintile, about prevalence of overweight among children under 23 percent of children under 5 suffer from stunt- 5 almost doubled between 2011 and 2018, from ing, so poverty is not the only factor. Making 5 percent to 9.5  percent. Malnutrition is also matters worse, the repeated waves of COVID- prevalent in women of reproductive age: more 19 since March 2020 and the floods in 2022 than 14  percent are underweight, 24  percent have reduced incomes, threatened livelihoods, overweight, 14  percent obese, 27  percent de- exacerbated food insecurity, and disrupted the ficient in vitamin A, and 43  percent anemic. provision and use of basic health and nutrition FIGURE 6  Distribution of eight categories of adequacies in children younger than 24 months in Pakistan Food intake, environmental health, and care 1.6 Food intake and environmental health 9.1 Overview | Pakistan Human Capital Review Environmental health and care 8.5 Food intake and care 0.5 Food intake only 5.6 Environmental health only 45.6 Care only 2.5 None 26.2 0 10 20 30 40 50 Percent Source: World Bank calculations using data from National Nutrition Survey 2018. 10 FIGURE 7  Trends in prevalence of stunting, services. And they have set back Pakistan’s ef- underweight, and wasting in children under 5 in forts to reduce malnutrition. Pakistan, 1990–2018 Percent Micronutrient deficiencies have substantial im- Stunting pacts on health and human capital. Iodine de- 60 ficiency, particularly in pregnancy, affects fetal Demographic and Health Surveys development and child intelligence quotient 50 (IQ). An average loss of 12.5–13.5 IQ points was 40 found in children born to iodine-deficient moth- 40 37 38 ers. Children with iodine deficiency lose 13 IQ National Nutrition Surveys points on average, making them less educable. Iron deficiency is a common cause of anemia, 20 which contributes to increased maternal mor- bidity and mortality and low birthweight. The de- ficiencies have impaired brain development and reduced productivity in adults. Vitamin A defi- 0 ciency increases the severity of measles, malar- 1990 1995 2000 2005 2010 2015 2018 ia, and diarrhea in children and increases mor- Underweight bidity, including preventable pediatric blindness, 50 and death. Vitamin A supplementation for chil- 40 dren has been linked to a 23 percent reduction 40 38 National Nutrition Surveys in child mortality. Zinc deficiency is associated with reduced immunity and with the increased 30 incidence, severity, and duration of diarrhea; 29 Demographic and Health Surveys and it has a negative effect on child growth.10 23 20 High-level recommendations 10 Prioritize cost-effective nutrition interventions informed by evidence 0 1990 1995 2000 2005 2010 2015 2018 • Raise the national profile of stunting, which is a major human capital catastrophe requir- Wasting 20 ing national and local efforts. Ensure political 17 championing for stunting reduction at the na- National Nutrition Surveys tional and highest provincial levels. Strength- en nutrition-focused development partners’ Overview | Pakistan Human Capital Review 15 13 platforms at the national and provincial levels to improve cross-sectoral nutrition gover- 10 nance and coordination bodies and oversee 9 Demographic and Health Surveys multisectoral nutrition actions. Systematize 7 monitoring and evaluation for debottleneck- 5 ing and course adjustments. • Remove existing constraints to program imple- 0 mentation, including long delays in approvals, 1990 1995 2000 2005 2010 2015 2018 frequent changes in leadership, procurement Source: Pakistan Demographic and Health Surveys 2001, 2013, delays, and funds flow issues. Ensure timely 2018; Pakistan National Nutrition Survey 2018. decisions and problem resolution. Provide 11 The eight key risk factors disrupting children’s healthy development 1 poverty • Identify priority districts to implement inter- 2 stunting ventions based on hands-on global knowledge 3 living in a rural area on convergence and sequencing of interven- tions. Map out interventions in water, sanita- 4 disability tion, and hygiene (WASH); social protection; 5 low maternal education education; nutrition; and maternal, neonatal, 6 inadequate early stimulation at home and child health that might be able to con- verge. Empower district health authorities, 7 harsh parenting and ensure that they have the capacity and 8 lack of early childhood education resources commensurate with their respon- sibilities to implement nutrition and early childhood development programs. powers commensurate with responsibility, es- …by helping parents support their pecially to implementation-level bodies, and children’s early development strengthen their stewardship capacity. Build technical and operational capacity across the Key issues sectors that contribute to nutrition. • Support “kangaroo” mother care to keep the Young children in Pakistan are not developing newborn skin to skin with the mother. Re- as well as their low- and middle-income country place iron and folic acid supplementation for peers. Fewer than two-thirds of preschool-age pregnant women with multiple micronutrient children are on track, according to their parents: supplementation. 59.4 percent in Punjab (2017–18), 47.5 percent in Sindh (2018–19), 54.6 percent in Khyber Pakh- Coordinate nutrition interventions and tunkhwa (2019), and 39.1 percent in Balochistan advocacy with those for early childhood development FIGURE 8  Parents report that many children • Address constraints affecting the effective- age 3 and 4 are not developmentally on track ness of existing strategies. Ensure timely ap- Children developmentally on track (percent) proval and updating of strategies, particularly 100 those that are cross-sectoral. Give more at- tention to the sustainability of interventions. 80 Assess and articulate financing needs and Low- and middle-income country average: 75 relative priorities. Improve communication. 60 • Integrate high-quality, contextually relevant 59.4 54.6 parenting and behavioral change programs Overview | Pakistan Human Capital Review 47.5 40 into health, social protection, and education 39.1 platforms. Early interventions, including ac- cess to early childhood education, are vital to 20 mitigate risks and promote protective factors that shape healthy brain development, yield- 0 ing gains for future health, learning, and pro- Punjab Sindh Khyber Balochistan (2017–18) (2018–19) Pakhtunkhwa (2019–20) ductivity. For example, healthy and well-nour- (2019) ished children learn better and earn more as Source: Multiple Indicator Cluster Surveys. adults, and higher levels of education help Note: Based on parent reports of whether children age 3–4 are developmentally on track in at least three of the following four improve both health outcomes and access to domains: literacy-numeracy, physical, social-emotional, and ap- better employment opportunities. proaches to learning. 12 (2019–20) (figure 8). Every province in Pakistan FIGURE 9  Gaps in enrollment in early childhood education shows a lower prevalence of children on track between children age 3–5 in the lowest and highest income quintiles, than the 75  percent average across low- and in Pakistan and by province middle-income countries. The eight key risk Bottom 20% Top 20% factors disrupting children’s healthy develop- Overall ment are living in poverty, being stunted, living 15.2 27.2 in a rural area, living with a disability, having a mother with low education attainment, receiv- Punjab 25.3 33.4 ing inadequate early stimulation at home, harsh parenting that includes psychological or phys- Sindh 5.1 27.6 ical aggression, and not participating in early childhood education. Each risk factor is associ- Khyber Pakhtunkhwa 12.4 17.4 ated independently with lower rates of being de- velopmentally on track, but multiple risks often Balochistan 2.7 9.0 co-occur. 0 10 20 30 40 50 Children receive very little stimulation at home, Percent of early childhood education enrollment and the prevalence of harsh punishment re- mains high. Early stimulation of cognitive Source: World Bank calculations using data from the Pakistan Social and Living Standards Measurement Survey 2019–20. and social-emotional skills, which takes place through reading books, telling stories, sing- ing songs, going outside together, playing, or enroll in early childhood education, further wid- naming or counting, is low across the board: ening developmental gaps. In addition, although 35.3  percent in Punjab, 45.0  percent in Sindh, girls and boys experience similar exposure to 35.0  percent in Khyber Pakhtunkhwa, and risks and similar early childhood development 49.3  percent in Balochistan. Lack of adequate scores at age 3 and 4, gender disparities start to social-emotional and cognitive support inter- emerge as early as preprimary enrollment, with sects with lack of proper nutrition and other more boys than girls enrolled in early childhood critical inputs to increase the likelihood of stunt- education in every province. ing and other long-term risks to development. Use of harsh parenting (psychological or phys- Policy and program environments to support ical aggression) in the previous month among early childhood development and nutrition do preschool-age children is also very high in Pa- not cover the majority of Pakistan’s children, kistan: 84.9  percent in Punjab, 81.5  percent in and implementation of those that do exist falls Sindh, 87.1 percent in Khyber Pakhtunkhwa, and short. Implementation challenges facing pro- 57.0 percent in Balochistan. gram effectiveness range from staffing to ser- Overview | Pakistan Human Capital Review vice delivery to the households. Resources cur- Inequities in access to preprimary school rently allocated by the education sector for early place some children at a deep disadvantage, childhood education are inadequate, despite particularly children living in poverty. Inequal- overwhelming evidence showing high returns on ity between the rich and the poor is striking: investment, and only 19 percent of children age 15.2 percent of the poorest quintile are enrolled 3–5 are enrolled. In 2020, national investment in in preprimary school relative to 27.2 percent of education represented just 2.5 percent of GDP, the wealthiest quintile (figure 9). Already behind down from 2.9 percent in 2017 and considerably developmentally, poor children’s developmen- lower than the global average of 4.5  percent. tal trajectory is further undermined by the fact Precise financing figures on preprimary educa- that they are less likely than other children to tion are unclear but estimated to be low. 13 High-level recommendations • Strengthen monitoring and evaluation sys- tems for early childhood development and Expand cash transfer support to young nutrition by incorporating children’s devel- families from pregnancy through preschool opmental milestones into health information • Supplement cash transfers, where they exist, systems while also strengthening education with programs such as the Lady Health Work- information systems to collect key health er Program, and develop training and aware- markers. ness programs to empower mothers at the • Regularly evaluate the reach and quality of community level to support each other. such multisectoral programs to ensure they • Use community-based mother support groups remain fit-for-purpose. to monitor child development outcomes and identify early markers for intervention. MAKE SURE CHILDREN ARE IN Improve the supply of clean water and SCHOOL AND LEARNING adequate sanitation with piped water delivery wherever feasible An estimated 20.3 million of Pakistan’s school- • Improve existing tariff structures with a view age children were out of school prior to the to enhancing sustainability, treatment of COVID-19 pandemic and the 2022 floods that in- human waste, water treatment, and water undated nearly one-third of the country.11 Due to quality tracking. school closures and the socioeconomic impact • Ensure all public facilities such as schools of the pandemic and the 2022 floods on families, and basic health units have adequate WASH an estimated 2–3.5  million additional children facilities. will drop out of school in Pakistan, erasing years of enrollment gains.12 At 10  percentage points, Encourage behavior change at the the gender gap in enrollment has narrowed household and community levels slightly, from 13 percentage points in 2007 (fig- • Develop and implement parenting programs ure 10). In addition, Pakistan’s learning poverty that teach caregivers how to engage in re- rate­—­the percentage of children unable to read sponsive interactions with their children and and understand a short age-appropriate text provide early stimulation. These programs by age 10­ —­at 75 percent before the pandemic would encourage breastfeeding, teach par- and the 2022 floods, is more than 16  percent- ents how to provide stimulation, and under- age points above the average for South Asia and score the benefits of proper nutrition and more than 19 percentage points above the aver- positive parenting. age for lower-middle-income countries. The high • Encourage reduction in the incidence of open number of primary and secondary school–age defecation and encourage good hygiene children who are not in school and the low qual- Overview | Pakistan Human Capital Review practices such as regular handwashing. ity of education­ —­about 65  percent of children perform below minimum proficiency in reading­ Ensure the convergence and geographic —­ explain Pakistan’s high learning poverty rate. co-location of sectoral interventions for So, simply bringing all children to school will not nutrition and early development at the be enough to end Pakistan’s learning poverty household level and build human capital. • Empower district health and education au- thorities, and ensure they have the capacity The COVID-19 pandemic has likely increased the and resources commensurate with their re- number of out-of-school children and worsened sponsibilities to implement early childhood learning poverty, increasing the cost of bring- development and nutrition programs. ing all children into school and improving the 14 Bringing all children into school while also improving learning outcomes and overall education system efficiency is challenging, but it can be done. FIGURE 10  More girls than boys are out of school Percent Male Female 50 47 40 43 40 39 37 36 36 34 30 30 27 27 25 25 25 20 10 Overview | Pakistan Human Capital Review 0 2004–05 2006–07 2008–09 2010–11 2012–13 2014–15 2018–19 Source: World Bank calculations using data from the Pakistan Social and Living Standards Measurement surveys 2004–19. quality of education. It is estimated that learn- young women. Among the top barriers to chil- ing-adjusted years of schooling will decline by dren’s enrollment and progression in school, one-third to two-thirds of a school year and that the most common are high cost, distance to learning poverty will rise to at least 79 percent.13 schools, perceived poor quality of education, Disruption of education during the pandem- and shortage of teachers. These problems are ic will disproportionately affect disadvantaged particularly relevant in rural schools and persist and hard-to-reach children, including girls and across education levels. 15 Pakistan needs to educate all its children to a world, behind India. The large share of children level that promises them a more productive fu- out of school contributes to nearly half of Paki- ture. An important first step is to recognize that stan’s high (75–79 percent) learning poverty. out-of-school children are not a uniform group. Policies need to be tailored to the characteris- The population of out-of-school children broadly tics of their different groups to increase impact. comprises three distinct groups, each with dif- Bringing out-of-school children into education or ferent needs, and policies need to be tailored training is a huge challenge. The longer children to maximize impact. The youngest out-of-school are out of school, the less likely they are to go to children (age 5–9), around half of the total school, especially children who have dropped (9.5 million), have the most straightforward out.14 Tackling dropout requires targeted, aggres- needs: access to schools, a caring teacher in sive, and innovative approaches. Bringing all chil- front of the classroom, and perhaps a nudge dren into school while also improving learning to parents from the community to enroll their outcomes and overall education system efficien- child in school. But children who are in school cy is challenging, but it can be done. and at risk of dropping out (typically age 8–16) often require additional learning support, to Key issues make sure that parents see the value of keeping their children in school. As the labor and mar- With an estimated 20.3  million children out of riage markets start pulling older children away school (age 5–16), Pakistan has one of the larg- from school, parents may need incentives that est number of children in the world not attend- reduce their cost of sending children to school, ing school. The 20.3 million figure corresponds such as stipends. Older children who have never to nearly a third (31.7 percent) of Pakistan’s chil- been to school require specific literacy interven- dren age 5–16, 82 percent of whom have never tions and short-term skills training to help them gone to school. Due to population increases, become literate and obtain job-related skills. this absolute number remains large, even as the share of out-of-school children has fallen in the Making sure that all children are in school and past few decades. In absolute terms, the out- learning will require a substantial financial in- of-school population is second highest in the vestment from Pakistan’s government. The total cost of providing all current children in Pakistan with a quality education would conservatively cost 4.8 percent of GDP, including 2.9  percent An estimated 20.3 million to increase school enrollments under current children (age 5–16) conditions and 1.9  percent to improve quality. out of school … This assumes that all children can be enrolled in public schools or low-fee private schools, at an Overview | Pakistan Human Capital Review average annual cost of US$240 for each child. 31.7% High-level recommendations 82% Expand the supply of safe schools so that every child has a guaranteed seat in school … equals nearly … 82% of • Prioritize public sector provision and public–­ one-third of whom have private partnerships. Strong controls on pro- Pakistan’s never viders to guarantee school outcomes, along children … attended with the safety and welfare of students and school. teachers, are paramount to get and keep 16 Making sure that all children are in school and learning will require a substantial financial investment from Pakistan’s government. children in school. Clean water and basic dedicated and free transport for secondary WASH facilities, including safe bathrooms, school girls. The transition to secondary must be present in all schools. school is a key attendance drop-off point for • Rehabilitate and build schools where they are girls and boys in Pakistan. Direct provision of needed. School and classroom construction monetary support and safe school transport is particularly important in areas where diffi- can ease families’ concerns, increasing enroll- cult or no access to school is a key reason for ment and reducing dropout. Conditional cash dropouts. This has been successful in parts of transfers are most effective when payments Balochistan, where abandoned public build- are directly linked to children’s enrollment ings have been adapted and made adequate and regular attendance at school. In the most for service provision of education. disadvantaged districts, adding a premium • Hire more qualified (especially female) teach- for food or food vouchers can offset families’ Overview | Pakistan Human Capital Review ers based on merit to increase enrollment and need for children to participate in household attendance among adolescent girls. Like girls, or paid work, freeing resources for the educa- female teachers may require safe and dedicat- tion required to build human capital. ed transport to reach schools, particularly in • Raise awareness about the importance of remote areas. Parents often see female teach- education, particularly for girls. Communi- ers as an indicator of security for young girls. ty approaches that bring parents, and other household and community members into Support households to increase enrollment advocating for and facilitating girls’ access to and reduce dropouts education and security from home to school • Expand the education conditional cash trans- can increase education demand. More re- fer program to both boys and girls and provide search is needed on measuring and shifting 17 common learning disorders (such as dyslex- ia), informing them of available support. • Introduce basic literacy and numeracy pro- grams, coupled with skills development pro- grams, to prepare older children who have never attended school or are longer-­ term dropouts for integration into the labor mar- ket. Partner with local community organiza- tions and potential employers to meet the needs in the locality. social norms around girls’ education. This IMPROVE LABOR MARKET work includes sharing information on returns OUTCOMES FOR THE POOR to education and the value of education for all for human capital development. Poorer households in Pakistan are trapped in a vicious cycle of low human capital and poor Prioritize literacy to increase retention, and labor market outcomes, the most serious of enable human capital accumulation which is low earnings. Boosting human capital • Develop literacy training materials for care- through means such as education attainment givers and teachers of young children, using would enhance earnings, but households must existing curriculum and materials. also increase their earnings to be able to have • Conduct a rapid assessment of all reading ini- the resources to invest in themselves. Poor- tiatives across Pakistan. er households not only lag in human capital • Draw on existing initiatives. Government, outcomes but also invest far less in absolute nongovernmental organizations, and devel- terms on education and health than wealthier opment partners have offered early reading households. In Pakistan, women, youth, and the programs throughout Pakistan for years that less-educated generally have far poorer labor can provide important resources and imple- market outcomes, and women and vulnerable mentation experiences. groups tend to suffer more from shocks, such as • Direct support to teachers. Structured peda- the COVID-19 pandemic and climate change. gogy and training in teaching at the right level have shown positive impacts on children’s lit- Breaking the cycle of low human capital invest- eracy levels. ment and poor labor market outcomes requires • Train teachers and school leaders to identify not just improvements in the supply of health children at risk of dropping out and provide and education services and measures to boost Overview | Pakistan Human Capital Review in-school remediation centered on foun- productivity but also financial support and other dational learning and numeracy. Low-per- complementary measures. Multipronged in- forming students are often the first to drop terventions, including those anchored on cash out of school. International evidence is clear transfer programs, can be powerful mecha- that teachers are key to student learning nisms to support the labor market (and human outcomes and strengthening their ability to capital) outcomes of the poor. But they must connect with at-risk students is crucial for clearly identify the segment of the labor market decreasing dropout rates. To better equip that the program will target and coordinate the teachers to support students, preservice and package of interventions to be bundled togeth- in-service training curricula should instruct er. And for their targeting, they should build on teachers how to identify at-risk children and the National Socio-Economic Registry (NSER) 18 and the Benazir Income Support Program’s In Pakistan, women’s labor force participation (BISP) poverty scorecard. rate is low, at just over 20  percent (and even lower for women age 15–24), while men’s is Key issues 80  percent. And 60  percent of working-age women are not in employment, education, or Informality predominates in Pakistan, with training, against only 6 percent of working-age 94.9 percent of wageworkers not having a for- men. Social norms, limited mobility, and wide- mal contract:15 83  percent of all wageworkers spread safety concerns are associated with are informally employed male workers, and in- women’s inability to be in the labor force. formally employed female workers account for another 12  percent (figure 11). Microenterpris- Overlapping crises imply more pronounced set- es (1–9 workers) account for 88 percent of all backs to human capital outcomes, particularly employment, formal and informal, for men and for the poor. Pakistan now is faced with the im- women, with small firms (10–99) accounting pacts of the war in Ukraine, increasing natural for another 12 percent (figure 12). Gaps are sub- disasters, and an uncertain economic outlook stantial in the earnings of women and informal marked by rising debt and inflation. These chal- sector workers, and informally wage-employed lenges come on top of the COVID-19 pandemic females have the lowest earnings. Formal- and the 2022 floods. To cope, families are often ly wage-employed females, informally wage- forced to reallocate resources away from health employed males, and formally wage-employed and education and to switch to lower-quality males earn 1.9, 2.3, and 2.6 times as much per food, resulting in less diversified food intake and month. nutrient deficiency. These coping mechanisms FIGURE 11  Four-fifths of wageworkers are FIGURE 12  Almost 90 percent of informally employed wageworkers, formal and informal, are Distribution of wage-employed workers (percent) employed in microenterprises Male Female Distribution of wage-employed workers (percent) Informal Formal Informal Formal 100 Micro Small Medium and large 100 80 80 60 60 40 40 Overview | Pakistan Human Capital Review 20 20 0 0 Micro Small Medium Large All rms Informal, Formal, Informal, Formal, All types Source: World Bank calculations using data from Pakistan Labour male male female female Force Survey 2018–19. Source: World Bank calculations using data from Pakistan Labour Note: Microenterprises are firms with fewer than 10 workers, Force Survey 2018–19. small firms are those with 10–99 workers, medium firms are Note: Microenterprises are firms with fewer than 10 workers, those with 100–250 workers, large firms are those with more small firms are those with 10–99 workers, medium firms are than 250 workers. See Annex 6A for the definition of formal wage those with 100–250 workers, large firms are those with more employment. than 250 workers. See Annex 6A for the definition of formal wage employment. 19 • Ensure that economic inclusion programs that combine safety nets with complementa- ry services fit the needs of different segments of the poor and vulnerable. Productive inclu- sion programs that go beyond cash can help improve household resilience to withstand shocks in the long term. For programs with a focus on women, interventions must ac- count for social norms, which have a major impact on women’s social interactions by Boosting human capital through means restricting access to services or certain oc- such as higher education attainment would cupations deemed not suitable for women. Possible interventions to influence norms enhance earnings, but households must also include strategic use of positive messaging increase their earnings to be able to have about strong female role models and engag- the resources to invest in themselves. ing to change norms surrounding women’s economic activities. Similarly, employing women in public leadership positions can have long-lasting impacts on health, education, support the acceptance of ambitions and and labor outcomes across the life cycle, with career aspirations among women. Youth, young children often the worst affected. women, and other marginalized groups may need to engage in self-employment but Even before these crises, Pakistan faced high may lack capital, technical skills, and expe- levels of learning poverty and malnutrition. Pro- rience running an enterprise. Depending on tecting human capital is thus crucial and re- the combination of needs they face, pro- quires flexible measures to respond quickly to grams for them could include a combination the most vulnerable. of classroom or even on-the-job learning (such as apprenticeships), classroom-based High-level recommendations training in business management, socio- • Build the human capital of the poor, especial- emotional skills development, and access ly against shocks like the COVID-19 pandem- to finance, as through collateral-­ free micro- ic and climate change, through interventions finance. Youth entrepreneurship programs over the life cycle to ensure long-term liveli- in other emerging economies show that hood gains. Cash transfers and other safety successful programs pair business manage- net programs protect the human capital of ment training with other services such as the poor by supporting their consumption, access to capital and counseling to enhance Overview | Pakistan Human Capital Review especially when faced with adverse shocks. noncognitive ability. When these programs are complemented • Use existing programs as a basis for econom- with additional services, often collectively ic inclusion interventions. An asset transfer– called economic inclusion programs, they based pilot in Sindh spanned assets, finance, can improve human capital investments, consumption, and time use­ —­ and its package such as children’s education, health, and of services significantly increased incomes nutrition, as well as household earnings. For and revenues. The National Poverty Grad- households not poor enough to receive cash uation Program, in addition to transfers for transfers, seed capital, microfinance, and consumption, packages coaching, business technical and business management training capital, financial service facilitation, market are more relevant. links, and skills training. As poor households 20 begin to expand or diversify their business policymakers may still face key challenges­ —­ under the program, they become eligible such as identifying the appropriate benefi- to apply for loans; asset beneficiaries and ciaries, the right mix of interventions for the loan recipients are eligible to receive basic target group, and the best mix of implement- training to use the assets and loans effec- ing agencies in their design of new programs­ tively. And the Kamyab Pakistan Program, —­leveraging these major strengths will stand launched in 2021, is using financial services them in good stead and provide a stable and for entrepreneurship­ —­ rather than asset high-quality starting point for new programs, transfers­ as the core intervention for bun- —­ whether implemented by federal or provincial dling other services. entities. • Leverage existing national systems for effi- cient and effective delivery of programs. Pa- kistan has a strong starting point to deliver BENEFITS OF BUILDING relevant and comprehensive services given MORE HUMAN CAPITAL its existing strategic policy initiatives, and its investments in systems. It has made notable Without solid investments in human capital, progress in building and strengthening the no country can sustain economic growth over social protection system through its flagship the long run. As Pakistan seeks to become an BISP. It has invested in systems such as the upper-middle-income country with a thriving NSER and the poverty scorecard, as well as economy and population by its centennial, ef- a robust national identification system (civil forts to accelerate progress in human capital registration, digital and biometric, or voter development will be essential. identification). A key factor in the poten- tial effectiveness of support programs­ —­ to Simulations of several realistic scenarios show both protect the human capital of the poor how much Pakistan’s economic productivity and improve their productivity­ —­is whether could increase by investing more in human cap- the services reach the intended segments ital over the next 25 years­—­ to 2047, the 100th of the poor, by location and gender. While anniversary of its founding (figure 13). FIGURE 13  The economic benefits of actions to build human capital are substantial Percentage increase in Pakistan’s Human Capital Index (HCI) value by 2047 From rise in HCI value From full use of human capital From higher HCI value + full use of human capital 160 140 144 120 Overview | Pakistan Human Capital Review 100 80 71 60 65 57 52 40 42 39 32 20 18 9 13 9 0 Growth at Growth at levels Growth under typical transition Growth under best transition current rates of today’s peer to upper-middle-income status to upper-middle-income status Source: World Bank simulations. Note: Baseline is Pakistan’s 2018 GDP per capita of US$1,483. 21 Business as usual: If Pakistan continues to im- rate as an average country that transitions from prove its HCI value at the current rate, its GDP lower-middle-income status to upper-middle-in- per capita is expected to grow by a mere 18 per- come status, which Pakistan aspires to, its GDP cent by 2047. per capita would rise by 39 percent. Matching its current peers: If it can boost its HCI Matching the best performing lower-middle-in- value to the level of its peers, its per capita GDP come country having advanced to upper-mid- could grow by 32 percent by 2047. dle-income status: If it can improve both its human capital and its use of human capital, bring- Matching the average lower-middle-income ing adults into productive work outside farming, country advancing to upper-middle-income sta- its GDP per capita would rise by 144 percent, eight tus: If it can improve its HCI value at the same times more than under business as usual. Overview | Pakistan Human Capital Review 22 NOTES 1. World Bank 2018. are also fewer opportunities for making better use of 2. Government of Pakistan 2022. that human capital in better employment. 3. The Government of Pakistan declared a national 7. Demographic and Health Survey 2017/18. emergency and rapidly implemented a vaccination 8. Environmental health refers to the availability of campaign, in collaboration with many of its stakehold- drinking water and sanitation facilities; care includes ers. Its schooling system implemented strict standard such factors as antenatal care, nutritional supple- operating procedures, allowing schools to open earlier mentation during pregnancy, and appropriate treat- than other countries in the region. And the shutdown ment of childhood diarrhea. of the economy was relatively short lived, giving busi- 9. National Nutrition Surveys. nesses the opportunity to get back to work quickly. 10. Shekar et al. 2016. 4. Provincial disaggregation of learning-adjusted years 11. NEMIS and AEPAM 2018. of schooling is not possible. 12. This estimation comes from phone surveys car- 5. “Utilization” is the HCI value adjusted for labor force ried out by the World Bank in January 2021 and participation. September–­October 2022. 6. The full utilization has a U-shaped relationship with 13. World Bank 2020b. GDP per capita. In low-income countries (as op- posed to middle-income countries), the full utiliza- 14. OECD 2020. tion rate is usually higher than the basic utilization 15. Pakistan’s high rates of informality are not uncom- rate. As discussed in World Bank (2020a, p. 91 ff.), mon in developing countries. Close to 90 percent of the intuition behind this is that in low-income coun- all employment in South Asia is informal (ILO 2018; tries, there is less human capital and hence there Loayza and Meza-Cuadra 2018). REFERENCES Government of Pakistan. 2022. “Pakistan Floods 2022: during School Closures and School Re-openings.” Paris: Post-Disaster Damage Assessment (PDNA).” Ministry OECD Publishing. https:/ /www​.oecd​.org​/coronavirus/ of Planning Development and Special Initiatives: Gov- policy​ - responses​ / the​ - impact​ - of​ - covid​ -19​ - on​ - student ernment of Pakistan. https:/ /www.pc.gov.pk/uploads​ -equity​-and​-inclusion​-supporting​-vulnerable​-students​-during​ /downloads/PDNA-2022.pdf. -school​-closures​-and​-school​-re​-openings​-d593b5c8/. ILO (International Labor Organization). 2018. Women and Shekar, M., J. Kakietek, J. D. Eberwein, and D. Walters. Men in the Informal Economy: A Statistical Picture. Gene- 2016. An Investment Framework for Nutrition: Reaching va: International Labor Office. the Global Targets for Stunting, Anemia, Breastfeeding, and Wasting. Directions in Development: Human Develop- Loayza, N., and C. Meza-Cuadra. 2018. “A Toolkit for In- Overview | Pakistan Human Capital Review ment. Washington, DC: World Bank. formality Scenario Analysis.” Washington, DC: World Bank. http://www.worldbank.org/informality. Accessed 3 World Bank. 2018. The State of Social Safety Nets 2018. March 2022. Washington, DC: World Bank NEMIS (National Education Management Information World Bank. 2020a. The Human Capital Index 2020 Up- System) and AEPAM (Academy of Educational Planning date: Human Capital in the Time of COVID-19. Washing- and Management). 2018. Pakistan Education Statistics ton, DC: World Bank. 2016–17. Islamabad. http://library.aepam.edu.pk/Books​ World Bank. 2020b. Learning Losses in Pakistan Due /Pakistan%20Education%20Statistics%202016–17.pdf. to COVID-19 School Closures: A Technical Note on Sim- OECD (Organisation for Economic Co-operation and De- ulation Results. Washington, DC: World Bank. https://​ velopment). 2020. “The Impact of COVID-19 on Student openknowledge.worldbank.org/handle/10986/34659. Equity and Inclusion: Supporting Vulnerable Students 23 Pakistan’s human capital: Opportunities amid crises SUMMARY many to skip regular health checkups and treat- 1 ments, with adverse consequences that are like- Pakistan can realize major economic growth and ly to be felt for decades to come. And school-age development by investing in its people and their children and youths have similarly experienced human capital. But the reality is that Pakistan’s huge disruptions in access to education and job human capital is low and has improved only opportunities. Making matters worse, the 2022 marginally over the past three decades. Inequal- floods further limited access to health and edu- ities in human capital outcomes have persisted cation services and likely increased malnutrition or widened over time between the rich and poor, and learning poverty.2 men and women, and rural and urban areas and among the provinces. Human capital outcomes These low human capital investments will limit are low across the board, with even the most the realization of Pakistan’s ambition to become economically advantaged groups in Pakistan an upper-middle-income country by 2047. If having lower human capital outcomes than less Pakistan continues on its current trajectory in economically advantaged groups in peer coun- human capital development, its GDP per capita tries. Pakistan’s Human Capital Index (HCI) would grow overall by a mere 18 percent through value of 0.41 is low in both absolute and relative 2047, the 100th anniversary of its founding. If terms. It is lower than the South Asia average of Pakistan can boost human capital investments 0.48, with Bangladesh at 0.46 and Nepal at 0.49. and its HCI value to the level of its peers, per Pakistan’s human capital outcomes are more capita GDP could grow by 32 percent. But if comparable to those in Sub-Saharan Africa, Pakistan improves both its human capital and which has an average HCI value of 0.40. its use of human capital, bringing adults into employment outside farming, GDP per capita Pakistan thus remains far from realizing its po- could rise by 144 percent, eight times more than tential and forgoes substantial growth and de- under business as usual. velopment. Despite favorable demographic tailwinds, Pakistan is currently unable to reap To boost its human capital, Pakistan needs to its rightful dividends due to inadequate invest- bring population growth under control, invest ments in education, health, social protection, and more in the supply of health and education, and Chapter 1 | Pakistan Human Capital Review the labor force. Its public investment of about enhance female labor force participation. Paki- 2.5 percent of GDP in education and 0.9 percent stan needs a healthy, skilled, and resilient pop- on health is much lower than the global average ulation to ensure high economic growth that is and the average for similar economies. Pakistan both inclusive and sustainable. Bringing more spends about 0.6 percent of GDP on social safe- women into education and the labor force will ty nets, compared with the global average of require increased focus on their safety, not just 1.5 percent.1 The COVID-19 pandemic has deep- at school and in the workplace but also in tran- ened the challenges, adding a significant threat sit to and from home, while creating more jobs to human capital development at all stages of accessible to them in all parts of the country. the life cycle. Disruptions to health services and With the right policies and investments, the income losses due to the pandemic have forced growing working-age population can become 25 healthier, more educated, more skilled, and Given that better education and health trans- more productive­ —a­nd can earn more, if the late into improved productivity, and that economy generates more and better jobs. This human capital is often the only asset for the calls for a deliberate effort from multiple stake- poor, sustainable poverty reduction is unlike- holders and sectors to build on intrasectoral ly to materialize without substantial invest- and intersectoral externalities. For example, ments in human capital. Despite widespread healthy and well-nourished children learn better agreement that Pakistan has a human capital and earn more as adults, and higher levels of ed- crisis, there is surprisingly little consensus ucation help improve health outcomes through- around the enormous opportunity of investing out the life cycle as well as improve access to in human capital. With a high fertility rate and a employment opportunities. growing and young population (see figure 1.1), successive governments in Pakistan have tried to keep pace and have continued to provide INTRODUCTION subpar health and education services to ever more people. Pakistan faces a human capital crisis. Its popu- lation, three-quarters of which is under the age Improving human capital takes a long-term of 35 (figure 1.1), has very recent memories of commitment. As a first order of business, Pa- failures in the schooling system and in health kistan needs to bring population growth under services. Federal and provincial governments, as control, to ensure that it can invest more in its well as Pakistan’s policymakers, often cite high existing population and reap a demographic stunting rates, high childhood mortality, high dividend, rather than continue to provide low- out-of-school populations, and low learning lev- quality services to the ever-growing population. els. Gender inequality is particularly stark in ed- It needs to invest more in the supply of health ucation and the labor market, and all outcomes and education services. Pakistan’s population remain unequal across wealth quintiles and geo- challenge can be turned into a demographic op- graphic regions. Pakistani leaders are also keen- portunity only by improving the quality of basic ly aware of public opinion polls showing voter services and improving human capital. And concerns about the lack of jobs, reflecting the Pakistan needs to enhance female labor force low utilization of human capital. participation, to ensure that families and the FIGURE 1.1  Pakistan has a young population, which is growing much faster relative to its peers a. Pakistan’s population pyramid b. Average number of children per adult woman (2018) 4 75 and older 70–74 65–69 3.5 Chapter 1 | Pakistan Human Capital Review 60–64 3 55–59 50–54 Male Female 45–49 2.4 40–44 2.3 2 2.2 35–39 2.1 2.0 1.9 30–34 25–29 20–24 1 15–19 10–14 5–9 0–4 0 8 4 0 4 8 an h a a n al co si di es Ir a ep oc st ne Percent In ad N ki or do gl Pa M an In 26 B Source: 2017 population census, World Bank Open Data. country as a whole can benefit from the talent PAKISTAN’S PERSISTENT available. The surest ways to reduce the popu- HUMAN CAPITAL CRISIS lation burden are to keep girls in school longer,3 improve health services for young women,4 and Pakistan’s human capital crisis has two facets. bring women into the labor force.5 With the right The first is the low absolute and relative level of policies and investments, the growing work- human capital itself. Pakistan has an HCI value ing-age population can become healthier, more of 0.41, which is lower than expected for its level educated, more skilled, and more productive­ of economic development (figure 1.2). There is —­and can earn more if the economy generates also substantial geographic variation, with Ba- more and better jobs. lochistan at the global bottom, and Punjab at the regional bottom. Pakistan’s human capital Pakistan’s handling of the COVID-19 pandemic outcomes are more comparable to Sub-­Saharan has shown that the country can manage com- Africa’s, which has an average HCI value of 0.40 plex human capital challenges despite its insti- (table 1.1). The South Asia average is 0.48, with tutional bottlenecks. While the pandemic has Bangladesh (0.46) and Nepal (0.49) having deepened the human capital crisis, infection higher HCI values than Pakistan. and death rates in Pakistan were much lower than those of its peers. The government rap- To put this in perspective, consider that a child idly implemented a national vaccination cam- born in Pakistan today can expect to be as pro- paign, in collaboration with many of its stake- ductive as a child born in Sub-­ Saharan Africa holders. Its education system implemented and only 72  percent as productive as a child strict standard operating procedures, allowing born in the Middle East and North Africa or schools to open earlier than in other countries 59 percent as productive as one born in Europe in the region. And the shutdown of the econo- and Central Asia (see table 1.1). Given that bet- my was relatively short, giving businesses the ter education and health translate to improved opportunity to get back to work fast. With sim- ilar commitment and urgency, Pakistan can FIGURE 1.2  Pakistan’s Human Capital Index value is among the turn a corner and enhance its human capital world’s lowest and lower than expected given its level of economic outcomes and economic growth in the medi- development um to long term. Human Capital Index value 1.0 This chapter reviews the benefits Pakistan can realize from investing in its people and their human capital. Based on newly devel- 0.8 oped measures of human capital, including the Vietnam Sri Lanka Türkiye World Bank’s HCI, it provides a comprehen- Chapter 1 | Pakistan Human Capital Review Indonesia 0.6 sive assessment of human capital indicators Bangladesh Kazakhstan India Malaysia in Pakistan and their state today. The chapter Nepal Bhutan Afghanistan analyzes the level and degree of utilization of 0.4 Punjab 0.42 Pakistan 0.41 human capital, considering job opportunities Khyber Pakhtunkhwa Sindh 0.36 0.39 Balochistan 0.32 available to people. It details important in- equalities among the provinces and between 0.2 the rich and poor and men and women. The 6 7 8 9 10 11 12 Log of GDP per capita (constant 2010 US$) key argument is that every group in Pakistan stands to gain from improved human capital in- Source: Pakistan provincial calculations from 2017 and 2019 Demographic and Health vestments, even those in the highest consump- Surveys. Country estimates from the Human Capital Project. GDP from World Bank na- tional accounts data and Organisation for Economic Co-operation and Development na- tion quintile. tional accounts data. 27 TABLE 1.1  Pakistan’s Human Capital Index value compares poorly with those of other world regions Middle Europe & Latin East & Sub-­ East Asia Central America & North North South Saharan Indicator & Pacific Asia Caribbean Africa America Asia Africa Pakistan HCI Component 1: Survival  Probability of survival to age 5 0.98 0.99 0.98 0.98 0.99 0.96 0.93 0.93 HCI Component 2: Education  Expected years of school 11.9 13.1 12.1 11.6 13.3 10.8 8.2 9.4  Harmonized test scores 432 479 405 407 523 374 374 339 HCI Component 3: Health  Survival rate, from age 15 to 60 0.86 0.90 0.86 0.91 0.92 0.84 0.74 0.85  Fraction of children under 5 not stunted 0.76 0.90 0.85 0.82 — 0.69 0.69 0.62 Human Capital Index (HCI) 0.59 0.69 0.56 0.57 0.75 0.48 0.40 0.41 Source: World Bank 2020. productivity, and that human capital is often the everyone. Even the population in the top quin- only asset the poor have, sustainable poverty tile of Pakistan’s wealth distribution experienc- reduction is unlikely to be achieved without sub- es high stunting rates, high childhood mortality, stantial investments in education, health, and and low learning outcomes. A child born today social protection. in a household in the top wealth quintile in Pa- kistan will have an HCI value of 0.50, that is, she All social groups in Pakistan suffer from will be only half as productive as she could be low human capital, creating a strong incen- (figure 1.3). In contrast, for low- and middle-in- tive for governments to improve services for come countries such as Lesotho and Vietnam, those in the highest quintile have HCI values FIGURE 1.3  Pakistan’s Human Capital Index value is low even ranging from 0.59 to 0.85.6 among the richest quintile, relative to the average for its peers Human Capital Index value The second facet of Pakistan’s human capital 1.0 crisis is its low utilization. Economic gains from human capital can be realized only if people 0.8 can use their skills and ingenuity in produc- Average: 0.67 tive activities, such as gainful employment in the labor market. The basic utilization-adjust- Chapter 1 | Pakistan Human Capital Review 0.6 ed HCI (UHCI) based on employment status is used to get an idea of how much human capital 0.4 is being utilized in the labor market.7 The basic UHCI measure simply multiplies the HCI value 0.2 by the employment to population ratio to see how much human capital translates into gainful 0.0 employment. m u e an o ye l a an va ga bw di th r na Pe do ki st st ne In so r ba Pakistan’s UHCI value is only 0.20 (figure 1.4), ki kh et ol Tü Se Le Pa Vi m M za Zi Ka much lower than expected given its economic Source: World Bank 2021b. development and virtually at the global bottom. 28 This means that the benefits that Pakistan can FIGURE 1.4  Pakistan’s utilization-adjusted Human Capital Index realize from its human capital are further de- value is lower than expected given its economic development pressed because it does not have an environ- Utilization-adjusted Human Capital Index value ment for citizens to put their human capital to 0.6 work. In fact, with their low UHCI values, the provinces of Balochistan (0.15) and Khyber Pa- 0.5 Vietnam khtunkhwa (0.16) are at the very bottom of the Kazakhstan global distribution of human capital utilization. 0.4 Nepal Indonesia Malaysia Bhutan 0.3 Gender is the main factor driving Pakistan’s low Sri Lanka Türkiye labor force participation. Fewer than one in four Punjab 0.22 Bangladesh India 0.2 Pakistan (23 percent) women are in the labor force, com- 0.20 Sindh 0.17 Afghanistan Balochistan Khyber pared with four in five men (81 percent) (figure 0.1 0.15 Pakhtunkhwa 0.16 1.5a). Although the female labor force participa- tion rate more than doubled between 1990 and 0.0 2018, Pakistan’s value is still lower than in peer 6 7 8 9 10 11 12 Log of GDP per capita (constant 2010 US$) countries. Female labor force participation has a U-shaped relation with human capital, with Source; Human Capital Project, World Bank national accounts data, and Organisation for Eco- mostly uneducated women and those with high- nomic Co-operation and Development national accounts data. Note: For provincial values of log GDP per capita (constant 2010 US$), country value was used. er education degrees participating in substantial numbers (figure 1.5b). But women with primary and secondary education are mostly outside of of human capital has largely persisted over the the labor market in larger numbers. past three decades, with basic UHCI values in- creasing by 3 percentage points for men and by Women’s low returns to education at the pri- 4 percentage points for women (figure 1.6). mary and secondary levels may trigger a cycle that reduces incentives for parents to invest The UHCI for women is also lower in Pakistan in their daughters’ schooling. Poorly educated than in other countries, even those with similar women are ill-equipped to obtain better-paying UHCI gaps between men and women (figure jobs. While high-paying occupations improve 1.7).​ Gender gaps in India are similar despite along the range of education levels for men, overall better UHCI values for both men and women’s labor force participation­ —­and thus women there. The gender gap is wider only in their wages­ —i­ncrease only beyond 10 years the Islamic Republic of Iran, with a similar UHCI of education. The key factors that limit female value for women and better outcomes for men. labor force participation are household work, marriage norms (with women expected to The HCI can also be adjusted for “better Chapter 1 | Pakistan Human Capital Review withdraw from the labor force after marriage), employment”­ —­a full utilization-adjusted Human female mobility norms (with women not being Capital Index (full UHCI). The idea behind the allowed to travel outside the household or vil- full UHCI is to measure the share of the working-­ lage), and safety concerns for women on the age population that utilizes its human capital job and in transit. more productively. This is a measure designed to more closely capture the productivity gains While Pakistan’s gender gap on both the basic from human capital by classifying agricultural and full UHCI (see below) is comparable to the employment as less productive employment regional average, South Asia has a higher UHCI (from a human capital perspective).9 The full among women (13 percent, against 8 percent in UHCI also corrects for a country’s potential for Pakistan).8 Gender inequality in the utilization realizing human capital gains, giving those of 29 FIGURE 1.5  Labor force participation of women remains low, FIGURE 1.6  Gender gaps on the basic though it is growing for those with postsecondary education utilization-adjusted Human Capital Index remain a. Labor force participation by gender, age 15–65 wide Percent Utilization-adjusted Human Capital Index value 100 0.4 85 Male 80 Male 0.31 81 0.3 0.28 60 Total Overall 0.2 52 0.19 49 40 0.16 Female 0.1 20 23 0.08 Female 0.04 11 0 0.0 1990 1995 2000 2005 2010 2015 2018 1990 2012 2017 2019 b. Female labor force participation by education, age 15–65 Source: World Bank calculations using data from Demographic Percent and Health Surveys, Household Income and Expenditure Surveys, Pakistan Social and Living Standards Measurement Surveys, and 50 Labor Force Surveys. 43 40 Postsecondary facing Pakistan is underemployment (partic- ularly for women) relative to job opportunities 30 in the labor market.10 This becomes apparent 31 No education 29 when breaking the measure down by gender. 20 For men, there is no difference between the 19 two UHCI values, whereas women, when they Primary and secondary 14 10 do enter the labor market, tend to secure more 10 productive employment, attributable to the rel- 0 atively large share of highly educated women 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 joining the labor market. Source: World Development Indicators and World Bank calculations using data from Labor Force Surveys for Pakistan. THE IMPORTANCE OF HUMAN working age with no human capital (and agricul- CAPITAL FOR GROWTH Chapter 1 | Pakistan Human Capital Review tural workers) a minimum value for productivity (considering this as raw labor). In low- and mid- The benefits of focusing on human capital in- dle-income countries, the full UHCI is typically vestments are well established in the econom- higher than the basic UHCI, because there is ic literature. Human capital investments have less potential to utilize human capital in better a dynamic and circular relation with economic jobs. growth, and both are linked through various pathways. Improvements in early health out- According to the full UHCI measure, human comes translate into better education out- capital utilization in Pakistan is 26 percent (fig- comes, leading to more opportunities for pro- ure 1.8). The gap between the basic and full ductive employment. Productive workers are UHCI values means that the larger problem better able to utilize technology, accelerating 30 FIGURE 1.7  Pakistan’s basic utilization-adjusted Human Capital Index value for women is lower than that in peer countries Female Male Pakistan 0.08 0.31 Punjab 0.12 0.33 Sindh 0.04 0.29 Khyber Pakhtunkhwa 0.04 0.30 Balochistan 0.27 0.03 Iran, Islamic Rep. 0.09 0.39 India 0.10 0.36 Bangladesh 0.17 0.37 Türkiye 0.22 0.45 Indonesia 0.29 0.42 Malaysia 0.33 0.46 Russian Federation 0.48 0.49 0.0 0.1 0.2 0.3 0.4 0.5 Basic utilization-adjusted Human Capital Index value Source: World Bank calculations using data from Demographic and Health Surveys, Household Income and Expenditure Surveys, Pa- kistan Social and Living Standards Measurement Surveys, and Labor Force Surveys for Pakistan and provincial numbers. Benchmark values from the Human Capital Project. the transition to technology-based produc- FIGURE 1.8  Pakistan’s Human Capital Index tion, which in turn rewards human capital in- value falls dramatically when adjusted for vestments. And as Pakistan reaps the benefits utilization, particularly for women of this growth, it will be able to invest more in Human Capital Index (HCI) value health and education services that strengthen Basic utilization-adjusted HCI value this virtuous cycle. Full utilization-adjusted HCI value 0.5 Human capital complements physical and nat- 0.4 0.4 0.4 ural capital in the production process and is key 0.4 to technological innovation. Societies with high- 0.3 er human capital and equality of opportunity 0.3 0.3 are more cohesive, more able to find solutions 0.3 to complex challenges, and ultimately more 0.2 0.2 0.2 prosperous (box 1.1). Human capital makes up 61 percent of Pakistan’s wealth.11 As the nature Chapter 1 | Pakistan Human Capital Review 0.1 of work evolves in response to technological 0.1 advancements, global integration, and other 0.0 changes, investing in human capital becomes Overall Male Female even more important. Labor markets will in- Source: World Bank calculations using data from Demographic creasingly demand workers with higher human and Health Surveys, Household Income and Expenditure Sur- capital, especially those with advanced cogni- veys, Pakistan Social and Living Standards Measurement Sur- veys, and Labor Force Surveys (see Annex 1B). tive and socio-behavioral skills.12 Note: Basic utilization-adjusted HCI represents all employ- ment. Full utilization-adjusted HCI denotes nonagricultural Of course, Pakistan’s economic difficulties employment. are caused by many other factors beyond low human capital. Its economy has many structural 31 BOX 1.1 Human capital drives development A country’s assets or wealth underpins national income A country’s level of economic development is strongly cor- and well-being. These assets include produced capital related with its wealth composition. Human capital, mea- (made by people), natural capital (offered by nature), sured as the value of earnings over a person’s lifetime, is human capital (embedded in people), and net foreign as- the most important component of wealth globally. In 2014, sets (box figure 1). Sustaining long-term growth requires human capital accounted for an estimated 64 percent of investing in and managing this broad portfolio of assets. global wealth. It is thus a much larger driver of econom- GDP indicates how much monetary income or output ic development than traditionally understood and rises a country creates in a year, while a country’s wealth in- above tangible physical assets as the most crucial in- dicates the value of the underlying national assets and gredient for economic success and poverty reduction. therefore the prospects for maintaining and increasing Natural capital is the largest component of wealth in low-­ that income over the long term. The Changing Wealth income countries (47 percent) and accounts for more than of Nations report series by the World Bank (2005, 2011, one-quarter of wealth in lower-middle-income countries. 2018a, and 2021a) draws on a large and unique global asset database to estimate the underlying value of a na- Pakistan has substantially less produced capital than the tion’s wealth, considering the four asset types. The 2018 typical lower-middle-income country, but it has corre- and 2021 reports presented the most comprehensive ac- spondingly more human capital than other countries in its counts to date of the wealth of nations. income group­ underscoring its untapped potential. —­ Box figure 1 Share of wealth by asset type and income level in 2014 Produced capital Natural capital Human capital Net foreign assets World Pakistan Low income Lower middle income Chapter 1 | Pakistan Human Capital Review Upper middle income High-income OECD 0 10 20 30 0 10 20 30 40 50 0 20 40 60 80 –4 –3 –2 –1 0 Percent Percent Percent Percent Source: World Bank 2018a. 32 distortions that limit growth and productivity.13 FIGURE 1.9  The Human Capital Index gains that Pakistan could These distortions include undertaxation of real make if it improves all components over the next 25 years estate, export substitution because of high im- Human Capital Index value port duties, and crop subsidies that limit the di- 0.8 Peer country versification of the agricultural sector. Pakistani maximum, 0.78 firms, even those that focus on exports, tend to Middle-income country be small, and they exhibit low growth. There are 0.6 average, 0.57 also several loss-making and so-called zombie­ - Pakistan, firms (11 percent of the top 500 firms have been 0.39 0.41 Peer country average, 0.53 0.48 0.4 loss-making for more than three years). That said, without solid investments in human 0.2 capital, no country can sustain economic growth over the long run. Pakistan has more to gain from human capital investments in economic terms 0.0 than most peer countries. As Pakistan seeks to 2017 2019 2047 become an upper-middle-income country with a Source: World Bank calculations using data from the Human Capital Project, 2020. thriving populace and economy, stronger efforts to accelerate progress in building human capital status: A third scenario projects what would are critical. Four growth scenarios for Pakistan to happen if Pakistan improved its human cap- invest more in human capital and realize better ital outcomes at the same pace as a typical outcomes for its citizens are presented below. country that transitions from lower-mid- Each of these scenarios is based on basic as- dle-income status to upper-middle-income sumptions of growth and considers a timeline of status, reaching an HCI value of 0.57. This is 25 years, to the year that Pakistan will turn 100: perhaps the most realistic scenario that Pa- • Business as usual: The first scenario as- kistan could aim for, given its ambitions to sumes that the long-run rate of change ob- transition to upper-middle-income status by served in the Pakistan data continues with- 2047. out any change. At this rate, Pakistan’s HCI • Matching the best performing lower-middle-­ value would still be only 0.48 in 2047 (figure income country advancing to upper-mid- 1.9). Thus, even 100 years after the country’s —­ dle-income status: Finally, a fourth scenario­ independence, a child born in Pakistan in —­ more ambitious but still realistic­ would have 2047 would be less than half as productive as Pakistan seek to reach the maximum of its she could be given the risks to complete edu- peers. This would result in substantially high- cation and full health that prevail. er gains on human capital, with a child born in • Matching its current peers: A second scenario Pakistan in such circumstances being more Chapter 1 | Pakistan Human Capital Review is where Pakistan seeks to reach the present-­ than 70 percent as productive as she could day average of its peers. This is also a con- be. In all scenarios, much of the progress on servative estimate, as Pakistan would still the HCI would come from improving learn- substantially lag its peers, which are likely to ing-adjusted years of schooling, suggesting continue to improve their human capital out- a strong focus on ensuring that children have comes in the same period. That would mean access to quality education (figure 1.10). reaching an HCI value of 0.53, so children born in Pakistan in 2047 would be just over Realizing the economic gains half as productive as they could be. • Matching the average lower-middle-income Pakistan is three-fourths of the way to the country advancing to upper-middle-income centennial of its founding in 2047, with a goal 33 FIGURE 1.10  Improving the Human Capital Index: Component Pakistan’s demographic dividend contributions Increase in Human Capital Index value Pakistan is the world’s fifth most populous Learning-adjusted years of schooling country, with about 230  million people. An Under-5 survival rate Not stunted Adult survival rate “early demographic dividend” country with 0.4 immense potential for accelerated economic growth from the changes in its population age 0.3 structure, Pakistan can reap dividends from two phases of the demographic transition (figure 1.11). The first dividend can be captured as the 0.2 demographic transition speeds up and the pop- ulation age structure becomes more concen- trated among those of working age. The right 0.1 policies and investments in human capital can lead to youth cohorts that are healthier, more 0.0 educated, more skilled, and more productive Pakistan Peers Middle-income Peers and that can earn higher incomes if the econo- (current (average) countries (maximum) trajectory) (average) my can create good jobs for them. Source: Human Capital Project. The second phase comes along later in the de- mographic transition if Pakistan can increase of becoming an upper-middle-income coun- savings and investments as a result of fewer try with a thriving economy. To reach this dependent children, more disposable income, ambitious goal, efforts to accelerate progress and prospects for longer lives. Seizing the sec- in human capital development are essential. ond demographic dividend through human and There is general consensus that countries physical capital accumulation and higher labor need to boost investments in human capital productivity is going to be key to deal with the to prepare their workforce for the more skilled challenges presented by an aging population. jobs of the future and compete effectively in the global economy. However, Pakistan is currently m ­ issing the op- portunity to capitalize on its potential demo- Critical inputs and care during all stages of life­ graphic dividend. If existing trends continue, it —­starting from conception to birth and through will miss out on the associated economic devel- infancy, childhood, adolescence, and adulthood­ opment possible from a potential demographic —­determine the development of human capital. dividend in the future as well. This is due to a Children’s early development requires nurtur- confluence of low human capital, low labor force Chapter 1 | Pakistan Human Capital Review ing care that includes health, nutrition, water participation, and low productivity, as well as a and sanitation, security and safety, responsive pace of job creation that is slower than popula- caregiving, and early learning, to attain their full tion growth. development potential. Since this window of op- portunity closes quickly, investing in the physi- In Pakistan, fertility rates are comparatively cal, cognitive, linguistic, and socio-emotional high, and government investment in human development of children is critical. When human capital is low, thus producing low human cap- capital investments begin in the early years, ital outcomes. Even though the total fertility they lay a strong foundation for the prosperity rate has been falling over the past six decades, and resilience of adults, and for the growth and from 6.6 children per woman in 1960 to 3.5 in competitiveness of nations. 2018 (figure 1.12), it is still much higher than in 34 FIGURE 1.11  Can Pakistan harness its two distinct demographic dividends? Total dependency ratio FIRST DIVIDEND SECOND DIVIDEND Pakistan • Permanent increase in • Accumulation capital–worker of human ratio and • More and physical output per capita disposable capital • More • More income to production • More people workers save in working age Demographic transition Source: World Bank elaboration. comparator countries. Fertility rates have been FIGURE 1.12  The fall in fertility that Pakistan needs to accelerate falling in all provinces, but in 2018, women in Total fertility rate (children per woman) Population (millions) Punjab and Sindh still had more than three 8 240 children on average, and women in Balochistan Total fertility rate and Khyber Pakhtunkhwa had four children on average. 6 180 Financing education and increasing the effi- 4 120 ciency of spending remain key challenges for guaranteeing that every child receives a quality education. Public investment in education and Population 2 60 health as a share of Pakistan’s GDP remains low relative to that in comparator countries (figure 1.13), even if the real absolute amount has in- 0 0 creased (panel a). Pakistan’s total investment 1960 1970 1980 1990 2000 2010 2018 in education is 2.5 percent of GDP, far below the Source: World Bank calculations using data from World Development Indicators. international median of 4.4 percent. Each year, Chapter 1 | Pakistan Human Capital Review the federal and provincial governments invest roughly 10 percent of their total budgets in ed- 3.2 percent of GDP in 2018, against the glob- ucation. The Third International Conference on al average of 6.5 percent and the average for Financing for Development (in Addis Ababa, July lower-middle-income countries of 4.1 percent. 2015) recommended that at least 4–6 percent While investment in health has been rising of GDP and 15–20 percent of total public expen- worldwide over the past two decades, Paki- diture go to education to achieve the Sustain- stan’s remained stagnant at around 3 percent able Development Goals. of GDP. Pakistan’s annual investment in educa- tion, health, and social protection has grown in Similarly, Pakistan’s total investment in absolute terms but not kept pace with popula- health is low in international comparison, at tion growth (figure 1.14). 35 FIGURE 1.13  Government education and health spending in CONCLUSION AND Pakistan is below that in comparator countries RECOMMENDATIONS Percent of GDP a. Education expenditure Despite favorable demographic tailwinds, Paki- 5 stan has not been able to reap its dividends due Lower-middle-income countries largely to inadequate investments in human 4 capital, which are essential to improve the pro- 3.8 South Asia ductivity and employability of the workforce. 3.0 3 2.9 Low levels of human capital investments in Pa- 2.6 Pakistan 2.5 kistan as elsewhere may be due partly to the 2 long time it takes to see the benefits, unlike 1.8 physical capital investments, which can have 1 immediate and visible results. Returns to in- vestments in education and health, by contrast, 0 may take years or even decades to materialize. 2000 2002 2004 2006 2008 2010 2012 2014 2016 2019 Some of the benefits go beyond the generation b. Health expenditure making those investments. And many individ- 5 uals fail to see the benefits that investments South Asia 4.1 in their human capital can have on others and 4 Lower-middle-income countries society. These considerations have often led to 3.7 3.4 3.5 underinvestment in human capital­ —­missing an 3.2 3 2.9 opportunity to create a virtuous cycle of invest- Pakistan ment and growth. In Pakistan, public expendi- 2 ture on the social sectors falls short of needs. While health expenditure has increased in the 1 past 10 years, it remains well below the aver- age in South Asia. Expenditure on education, 0 by contrast, has dropped in the past five years 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 alone, bringing it even further below the aver- Source: World Bank elaboration. age for the region. FIGURE 1.14  Real absolute investment in public education, health, and social protection has risen in Pakistan since 2000, but changed little since 2007 Expenditure, billions BKS (adjusted for inflation) Education Health Social 2,500 Chapter 1 | Pakistan Human Capital Review 2,000 1,500 1,000 500 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 36 Source: PRSP. There is ample global evidence that no country education and health services are available to all can sustain economic growth over the long run its citizens. without solid investments in human capital. As Pakistan seeks to become an upper-middle- Pakistan can realize major economic growth income country with a thriving economy and and development by investing in its people and population by its centennial, efforts to acceler- their human capital. Simulations of several real- ate progress in human capital development will istic scenarios show how much Pakistan’s eco- be essential. Given the low level of human capi- nomic productivity could increase by investing tal, the policy agenda for Pakistan is clear. It will more in human capital over the next 25 years­ need to focus on fundamentals: bring its popu- —­ to 2047, the 100th anniversary of its founding lation growth rate under control, invest more in (figure 1.15). If Pakistan continues to improve its the supply and quality of health and education, HCI value at the current rate, its GDP per capi- and bring women to the labor force. Bringing ta is expected to grow by a mere 18 percent by more women into education and the labor force 2047. If it can boost its HCI value to the level of will require a focus on their safety, not just at its peers, its per capita GDP growth could near- school and in the workplace, but in transit to and ly double to 32 percent by 2047. Raising its HCI from home, while creating more jobs accessible value to the maximum among its peers would to them in all parts of the country. allow GDP per capita to grow by 71  percent. Even greater economic gains can be realized by The policies, programs and processes of coun- improving the utilization of human capital­ —b ­y tries that invested heavily in human capital and bringing more adults into the labor force and reaped the resulting benefits have focused on into more productive employment outside agri- ensuring the continuity of government sup- culture, where workers can make greater use of port across political cycles, closely coordinat- their human capital. Investments in human cap- ing sectoral programs at different levels and ital, while ensuring its full utilization, will result in branches of government and building an evi- dence base to improve and update human cap- FIGURE 1.15  The economic benefits of building human capital ital strategies. Therefore, there is an important will be substantial role for the Government of Pakistan in break- Percent ing out of the low demand for human capital Improvement in Human Improvement Improvement in HCI Capital Index (HCI) value in full utilization value and full utilization investments. The country’s highly favorable 150 demographic conditions mean that such in- 144 vestments can yield huge benefits in the com- 120 ing decades, enhancing the income-generating potential of the workforce and Pakistan’s devel- 90 opment prospects. Chapter 1 | Pakistan Human Capital Review 71 Pakistan needs to invest more systematically 60 65 57 in its people and develop avenues for them to 52 42 39 deploy their human capital more productively. 30 32 Human capital externalities and market failures 18 13 provide a strong rationale for the government to 0 9 9 invest in human capital. As in most developing Pakistan’s Growth to the Growth from Growth to expected current level an average lower maximum of lower countries, poor people in Pakistan stand to lose growth of its peers middle income to middle income to upper middle upper middle the most when the public sector fails to make income transition income transition these investments. The government should thus Source: World Bank simulations. prioritize investments that ensure that quality Note: Baseline is Pakistan’s 2018 GDP per capita of US$1,483. 37 the largest gains in Pakistan’s economic growth develop its labor market to accommodate and development. These range from an increase the growing youth population. Key actions of 42 percent from the current GDP per capita include training more Lady Health Work- to 144  percent, eight times more than under ers to educate women on family planning, business as usual. creating more awareness about the use of modern contraceptives, increasing the In summary, the Government of Pakistan needs provincial health budget, and focusing on to prioritize investments that ensure quality ed- structural inequalities in access to health ucation and health services for all its citizens. and education. • Exercise effective stewardship over human • Invest more and more smartly in people and capital and declare emergencies over the develop avenues for them to deploy their health and education crises. This requires human capital more productively. Human long-term planning beyond the tenure of capital externalities and market failures pro- any government and political cycle. With vide a strong rationale for the government to the COVID-19 pandemic and devastating invest in human capital. As in most develop- floods deepening its already poor state of ing countries, poor people in Pakistan stand human capital, the Government of Paki­ stan to lose the most when the public sector fails needs to declare emergencies and take bold to make these investments. To boost its actions to tackle its health and education human capital, Pakistan needs to invest more challenges. in the supply of health and education through • Make family planning a priority across all domestic resource mobilization, shifting re- human development initiatives. Pakistan sources from costly energy subsidies and im- should integrate population planning in ac- proving efficiency in the existing allocations ademic, religious, and national policies and to human development sectors. Chapter 1 | Pakistan Human Capital Review 38 NOTES 1. World Bank 2018b. employed in small agricultural settings, where work 2. Government of Pakistan 2022. is primarily physical, and it is difficult to realize gains from higher human capital levels. 3. For a review, see Wodon et al. (2018). 10. The full utilization has a U-shaped relationship with 4. See, for instance, Bhuiya et al. (2004); Douthwaite GDP per capita. In low-income countries (as op- and Ward (2005); Hennink and Clements (2005). posed to middle-income countries), the full utiliza- 5. Bloom et al. 2009. tion rate is usually higher than the basic utilization 6. Quintile data are available for 15 low-income and 22 rate. As discussed in World Bank (2020, p. 91 ff.), the lower-middle-income countries. intuition behind this is that in low-income countries, there is less human capital and hence there are also 7. “Utilization” is the HCI value adjusted for labor force fewer opportunities for making better use of that participation. human capital in better employment. 8. Figure 31A, Pennings 2020. 11. World Bank 2018a. 9. The basic idea behind this is that in many devel- 12. World Bank 2019. oping countries, large shares of the population are 13. See World Bank (2022). REFERENCES Alderman, H., and J. Behrman. 2006. “Reducing the In- Bau, N., and J. Das. 2020. “Teacher Value Added in a cidence of Low Birth Weight in Low-income Countries Low-Income Country.” American Economic Journal: Eco- Has Substantial Economic Benefits.” The World Bank Re- nomic Policy 12 (1): 62–96. search Observer 21: 25–48. Bhuiya, I., U. Rob, A. H. Chowdhury, L. Rahman, N. Haque, Andrabi, T., B. Daniels, and J. Das. 2020. “Human Capital and S. Adamchak. 2004. “Improving Adolescent Repro- Accumulation and Disasters: Evidence from the Pakistan ductive Health in Bangladesh.” FRONTIERS Final Report. Earthquake of 2005.” Journal of Human Resources 57 (2). New York: Population Council. Aslam, M., G. Kingdon, and M. Söderbom. 2008. “Is Fe- Bloom, D. E., D. Canning, G. Fink, and J. E. Finlay. 2009. male Education a Pathway to Gender Equality in the “Fertility, Female Labor Force Participation, and the De- Labor Market? Some Evidence from Pakistan.” In M. mographic Dividend.” Journal of Economic Growth 14: Tembon (ed.), Girl’s Education in the 21st Century: Gen- 79–101. https:/ /doi.org/10.1007/s10887-009-9039-9. der Equality, Empowerment and Growth. Washington, DC: Chandir S., D. A. Siddiqi, H. Setayesh, and A. J. Khan. World Bank. 2020. “Impact of COVID-19 Lockdown on Routine Immu- Aslam, M., G. Kingdon, and M. Söderbom. 2012. “Returns nisation in Karachi, Pakistan.” Lancet Global Health 8 (9): to Schooling, Ability and Cognitive Skills in Pakistan.” Ed- e1118–e1120. Chapter 1 | Pakistan Human Capital Review ucation Economics 20 (2): 139–173. Collin, M., and D. N. Weil. 2020. “The Effect of Increasing Autor, D. H., and D. Dorn. 2013. “The Growth of Low-Skill Human Capital Investment on Economic Growth and Pov- Service Jobs and the Polarization of the US Labor Mar- erty: A Simulation Exercise.” Journal of Human Capital 14 ket.” American Economic Review 103 (5): 1553–97. (1): 43–83. Azevedo, J. P., A. Hasan, D. Goldemberg, K. Geven, and Douthwaite, M., and P. Ward. 2005. “Increasing Contra- S. A. Iqbal. 2022. “Simulating the Potential Impacts of ceptive Use in Rural Pakistan: An Evaluation of the Lady COVID-19 School Closures on Schooling and Learning Health Worker Programme.” Health Policy and Planning Outcomes: A Set of Global Estimates.” The World Bank 20 (2): 117–123. Research Observer 36 (1): 1–40. https://doi.org/10.1093​ Field, E., S. U. Junaid, H. Majid, A. Malik, A. Shahid, and /wbro/lkab003. K. Vyborny. 2020. Transport and Urban Labour Market 39 Integration: Evidence on Travel Time and Congestion from World Bank. 2005. Where is the Wealth of Nations? Mea- a Mass Transit Quasi-experimental Evaluation and Evi- suring Capital for the 21st Century. Washington, DC: dence on Firms from a Randomised Control Trial in Paki- World Bank. stan, 3ie Grantee Final Report. New Delhi: International World Bank. 2011. The Changing Wealth of Nations: Mea- Initiative for Impact Evaluation (3ie). suring Sustainable Development in the New Millennium. Filmer, D., H. Rogers, N. Angrist, and S. Sabarwal. 2018. Environment and Development. Washington, DC: World “Learning Adjusted Years of Schooling (LAYS): Defining Bank. a New Macro Measure of Education.” Policy Research World Bank. 2018a. The Changing Wealth of Nations Working Paper 8591, World Bank, Washington, DC. 2018: Building a Sustainable Future. Washington, DC: Global Financing Facility. https:/ /www​.global​financing​ World Bank. facility​ .org​ /sites​ /gff_new/files​ /documents/Pakistan​ World Bank. 2018b. The State of Social Safety Nets 2018. -Covid-Brief-GFF.pdf Washington, DC: World Bank. Government of Pakistan. 2022. “Pakistan Floods 2022: World Bank. 2019. Pakistan at 100: Shaping the Future. Post-Disaster Damage Assessment (PDNA).” Islamabad. Washington, DC: World Bank. Hennink, M., and S. Clements. 2005. “The Impact of Fran- World Bank. 2020. The Human Capital Index 2020 Up- chised Family Planning Clinics in Poor Urban Areas of Pa- date: Human Capital in the Time of COVID-19. Washing- kistan.” Studies in Family Planning 36 (1): 33–44. ton, DC: World Bank. Pennings, S. 2020. “The Utilization-Adjusted Human Cap- World Bank. 2021a. The Changing Wealth of Nations 2021: ital Index.” Policy Research Working Paper 9375, World Managing Assets for the Future. Washington, DC: World Bank, Washington, DC. Bank. UNESCO (United Nations Educational, Scientific and World Bank. 2021b. Insights from Disaggregating the Cultural Organization). 2017. Reducing Global Poverty Human Capital Index, 2021. Washington, DC: World Bank. through Universal Primary and Secondary Education. UIS fact sheet, Volume 44. Paris: UNESCO. World Bank. 2022. Swimming through Sand: Pakistan’s Country Economic Memorandum. Washington, DC: World WHO (World Health Organization). 2021. Second Round of Bank. the National Pulse Survey on Continuity of Essential Health Services during the COVID-19 Pandemic. Geneva: WHO. World Bank. n.d. World Development Indicators data- https://www.who.int/publications/i/item/WHO​ - 2019​ base. Washington, DC: World Bank. https://databank​ -nCoV​-EHS-continuity-survey-2021.1. .worldbank.org/metadataglossary/world-development​ -indicators/series. Wodon, Q., C. Montenegro, H. Nguyen, and A. Onagoruwa. 2018. Missed Opportunities: The High Cost of Not Educat- ing Girls. Washington, DC: World Bank. Chapter 1 | Pakistan Human Capital Review 40 Unpacking Pakistan’s human capital crises SUMMARY top districts in northern Punjab now have twice 2 as much human capital as children in the lowest In the past three decades, Pakistan’s progress districts in rural Sindh and Balochistan. on human capital has been uneven and slower than that of its peer countries. Its Human Cap- A deeper look at the components of the HCI ital Index (HCI) value improved from an esti- shows what is driving the changes in human mated 0.32 in 1990 to only 0.41 in 2020, which capital outcomes: 1990–2012 saw a 34 percent means that babies born today can expect to gain in the quantity of schooling, a 17 percentage reach only 41 percent of their human potential point reduction in the stunting rate, and a 2.6 per- by the time they reach age 18. When adjusted for centage point gain in the probability of survival to using human capital in the labor market, Paki- age 5. These changes result in considerable dif- stan’s HCI value falls dramatically­—­ to 0.20. For ferences in the extent to which the components females, it is further depressed to 0.08, meaning contribute to changes in HCI values over time. that a baby girl born today can expect to reach only 8 percent of her human capital potential by Children from the top income quintile spend twice the time she turns 18. as long in school as the poor. Pakistan has the sec- ond largest out-of-school population in the world, While Pakistan has slowly raised its HCI value, at 20.3  million children, with its low enrollment recent disruptions by the COVID-19 pandemic rate of 71 percent of all children age 6–16. This is a and the floods in 2022 quickly set back the mod- highly gendered problem, with 37 percent of girls est gains. Based on simulations for education age 5–16 in Pakistan out of school versus 25 per- and health, the pandemic has erased more than cent of boys. A large share of children (22 percent eight years of progress on human capital, reduc- of all children age 5–16) never enroll in school. The ing Pakistan’s HCI value to 0.37, which is lower reasons include low accessibility and availabili- than in 2012. The 2022 floods not only caused ty of schools, safety concerns on the way to and deaths and destroyed physical assets; they also from school and work, poverty, and social norms reduced children’s opportunities to accumulate preventing the enrollment of girls. human capital. The combined effects of the two Chapter 2 | Pakistan Human Capital Review shocks have likely worsened the already high While gender gaps run deep across Pakistan’s learning poverty and malnutrition and limited society, gender inequality on the HCI is not as the cognitive, socioemotional, and healthy de- large as one might expect. The main source of velopment of Pakistan’s next generation. gender inequality in human capital accumu- lation is expected years of school, with boys Inequalities in human capital outcomes be- spending 1.4 more years in school than girls. tween the rich and poor have also grown, imply- Gender gaps in the probability of survival to age ing that the limited gains in human capital are 5 and the fraction of children under 5 not stunt- attributable to improvements in the top income ed are small and are even in girls’ favor. This quintiles. And provincial inequalities have wors- means that almost all of the gender inequality ened, with Punjab’s HCI value improving faster in HCI values comes from the education compo- than the rest of the country’s. Children in the nent and not from the health component. 43 So, by investing in girls’ education and training to materialize. It also details inequalities among and enhancing their physical mobility and safety the provinces and between the rich and poor to and from school and work, Pakistan can re- and men and women. duce gender gaps in employment and increase the use of its human capital, especially by cre- ating platforms and institutions that match THE EVOLUTION OF PAKISTAN’S women job-seekers with firms. And workplace HUMAN CAPITAL environments providing childcare, flexible work hours, dedicated transport, and separate rest Pakistan has not improved its human capital areas for women can be more conducive to outcomes at a rate comparable to those in other women’s working. countries: between 1990 and 2020, its HCI im- proved from an estimated 0.32 to 0.41, or an av- erage of 0.03 HCI point per decade (figure 2.1). INTRODUCTION Compare that with the faster HCI growth in the Russian Federation (0.08 point per decade) and As the nature of work evolves in response to Peru (0.06) (figure 2.2). technological advancements, global integra- tion, and other changes, investing in human A deeper look at the components of the HCI capital has become ever more important. And shows what is driving the changes in Pakistan’s since countries with higher human capital adapt human capital outcomes. Figure 2.3 shows the better to technological changes, the returns relative improvement of the HCI caused by the to human capital investments are also likely to underlying components, and figure 2.4 the rela- increase. While the speed of technological dif- tive improvement of the components over time. fusion and subsequent demand for skills in The period 1990–2012 saw a 34  percent gain Pakistan are still emerging, there is little doubt in the quantity of schooling,2 a 17 percent re- that children entering school today will confront duction in the stunting rate, and a 2.6 percent labor market conditions very different from what their parents faced.1 Pakistan thus needs FIGURE 2.1  Pakistan’s Human Capital Index to prioritize the health and education of its pop- value increased very slowly from 1990 to 2021 ulation so that its citizens have more opportuni- Human Capital Index (HCI) value ties to stay healthy, to learn continually, and to 1.0 gain new skills to stay competitive through high- er human capital. In a global environment where 0.8 markets are shifting, technologies are proliferat- ing, and competition is growing, Pakistan needs 0.6 Chapter 2 | Pakistan Human Capital Review to empower its citizens to create knowledge and O cial HCI values embody it in new technologies and products. 0.40 0.41 HCI values from survey data 0.4 0.40 0.40 0.40 0.39 0.37 This chapter looks at the evolution of human 0.32 capital in Pakistan using the HCI­—­ which com- 0.2 bines indicators of health and education into a measure of the human capital that a child born 0.0 today can expect to obtain by her 18th birthday, 1990 2012 2017 2018 2020 2021 given the risks of poor education and health Source: World Bank calculations using data from Demographic in the country. It reviews the evolution of Paki- and Health Surveys, Household Income and Expenditure Surveys, stan’s human capital over the past three de- Pakistan Social and Living Standards Measurement Surveys, and Labor Force Surveys, 1990–2017, and Human Capital Project 2020. cades and assesses the potential that has yet 44 FIGURE 2.2  Pakistan’s Human Capital Index growth is similar to FIGURE 2.3  The period 1990–2012 saw the that of peer countries, but starting from a lower base highest relative improvements in schooling, and Human Capital Index growth under-5 survival increased the most in 2012–17 2010 2021 Share of contribution in human capital improvement (percent) Pakistan Expected years of schooling Stunting Under-5 survival rate Punjab Sindh 100 18.3 40.3 Khyber Pakhtunkhwa Balochistan Vietnam 80 9.8 Türkiye Russian Federation 71.9 Kazakhstan 60 Malaysia 23.6 Iran, Islamic Rep. Moldova Peru 40 Indonesia Morocco 36.1 Zimbabwe Senegal 20 Lesotho 0.0 0.2 0.4 0.6 0.8 1.0 0 Human Capital Index 1990–2012 2012–17 Source: World Bank calculations using data from Demographic and Health Surveys, 2012–19, Source: World Bank calculations using data from Demographic Pakistan Social and Living Standards Measurement Surveys, 2012–17, and Human Capital and Health Surveys, 1990–2017, and Human Capital Project 2020. Project. Note: Contribution is calculated by share of increase in Human Capital Index value from the preceding year. Baseline year is 1990. gain in the probability of survival to age 5. These FIGURE 2.4  Expected years of schooling contributed the most to changes result in considerable differences in the Pakistan’s Human Capital Index growth between the two periods, extent to which the components contribute to but at a declining rate changes in HCI values over time. Between 1990 Percent and 2012, the gain in HCI value was due large- Expected years of schooling Stunting Under-5 survival rate 50 ly to improvements in the quantity of schooling (72 percent), with smaller gains due to improve- ments in stunting (10 percent) and child mortal- 40 ity (18 percent). However, in the most recent pe- 33.8 riod, 2012–17, gains in the quantity of schooling 30 slowed to 36 percent due to a slowdown in en- rollment growth, while reductions in child mor- 20 tality (40 percent) and stunting (24  percent) 17.0 Chapter 2 | Pakistan Human Capital Review contributed more to HCI growth. This shows 15.1 13.1 10 that the relative contributions of different com- ponents depend on their starting values and im- 2.6 1.8 0 provements over time. 1990–2012 2012–17 Source: World Bank calculations using data from Demographic and Health Surveys, 1990–2017. Why do Pakistan’s human capital outcomes remain low? below the average for Sub-Saharan Africa. The probability of survival to age 5 is far below the On most components of the HCI, Pakistan un- South Asia average and the same as the aver- derperforms the South Asia region and even age for Sub-­Saharan Africa (figure 2.5, panel a). Saharan Africa. On stunt- the average for Sub-­ While children in Pakistan stay in school slightly ing and the quality of education, it performs longer than their peers in Sub-­ Saharan Africa, 45 FIGURE 2.5  Pakistan underperforms on all Human Capital Index components except adult survival rate a. Probability of survival to age 5 b. Children under 5 not stunted 1.0 South Asia average 1.0 Sub-Saharan Africa average 0.8 0.8 Sub-Saharan Africa and South Asia average 0.6 0.6 0.4 0.4 0.2 0.2 0.0 0.0 do ye ng am an ka Tü ia Vi sia Bh h a an za iye In nam an al ia P h lo jab n l n tu an a dh n Sr t a n M n ka Bh ia Vi a n Pu ia ng ab sh a Pa pa hw di hw s nd ta er gh t a ta s s s kh Ind ep i an de de st ut st kh ist ut nj Ba Sin ay ay ne ne rk K a ür k B a un In e n is yb Af kis is s a Si nk nk N N ki kh kh et et iL iL la la ch Pa a n ch al al do T Pa tu za M Sr lo In Ba Ba Ka Pa er yb c. Expected years of schooling d. Learning-adjusted years of schooling Kh Kh 15 15 12 12 South Asia average Sub-Saharan Africa average 9 9 South Asia average Sub-Saharan Africa average 6 6 3 3 0 0 e ye m an a n an ka do sia a Tü al a n dh ng t an sh Pa ak b tu an m an a ka a al ia an sh an an iy si di gh w si si ta ta P ja d ep ep na na ki an an de de kh ist st st ut st st Ba Sin ay rk Af nkh ay ne ne un In In B a hu is is r N N ki ni kh kh et et Bh Tü iL iL la la ch al al do P B Pa ha Vi Vi ng za za M M Sr Sr lo In In g Ba Ka Ka Af er yb e. Survival rate from age 15 to 60 Kh 1.0 South Asia average Sub-Saharan Africa average 0.8 0.6 Chapter 2 | Pakistan Human Capital Review 0.4 0.2 0.0 ye m ka ia sh al a an an a an an si di s ep na ki an de st st ut st ay ne In r N ki ni kh et Bh Tü iL la al do Pa ha Vi ng za M Sr In g Ba Ka Af Source: Pakistan provincial calculations from Demographic and Health Surveys 2017 and 2019. Country-level estimates from the Human Capital Project. GDP from World Bank national accounts data, and national accounts data from the Organisation for Economic Co-operation and Development. 46 Note: Provincial breakdown is not available for learning-adjusted years of schooling because the Trends in International Mathematics and Science Study is represen- tative only at the national level. results from the Trends in International Mathe- children. Data from the 2018 National Nutrition matics and Science Study (TIMSS) 2019 rank Survey show that mothers and children suffer Pakistan as the second from bottom globally in from similar deficiencies that contribute to high the science and mathematics performance of child and maternal mortality. In Pakistan, one in its 4th graders (figure 2.5, panel d). Against peer five children is born with low birthweight, 54 per- and regional comparators, the only indicator on cent of children are anemic, and 52 percent of which Pakistan outperforms is adult survival children lack vitamin A. Among women of repro- rate (figure 2.5, panel e). ductive age, more than 14  percent are under- weight, 43 percent are anemic, and 27 percent Provincial inequalities are wide on the HCI and are deficient in vitamin A. its components, notably for stunting and expect- ed years of schooling (figure 2.5, panels b and c). Biological and psychosocial risks, such as ma- In schooling, Khyber Pakhtunkhwa and Punjab ternal health and malnutrition, maternal depres- are at the expected level for Pakistan’s economic sion, violence, poverty, and inadequate care and development, with the average child growing up early-learning opportunities, all prevalent in Pa- with at least 9 years of formal schooling­ ­ hough —t kistan, expose the developing infant to stressors when adjusted for quality, this is equivalent to even if survival is not a concern (chapter 3). The only 5.1 years.3 Sindh and Balochistan are at onset, severity, and chronicity of these stressors much lower levels, with the average child grow- can influence early brain development, which ing up with around 7 years of schooling. With is most rapid during early childhood and has only 89 percent of children surviving until their greater plasticity. While genetics establish the fifth birthday, Balochistan is at the very bottom­ architecture for the developing brain, the nur- —­ globally. The trajectory of child development turing environment and experiences exert a sig- and human capital accumulation for children nificant influence. Given the widespread risks born in Balochistan and Sindh is further under- to poor early development in Pakistan and the mined by the fact that almost half of the children large potential gains in returns to investments under 5 in the two provinces are stunted. in early childhood, more attention to early child- hood development (ECD) is warranted. High child mortality High stunting rates With 93  percent of newborns surviving until age 5, Pakistan is below the South Asia average Pakistan has one of the highest rates of child (96  percent) and on a par with Sub-­ Saharan malnutrition in all its forms: malnutrition (stunt- Africa (93 percent).4 Across low-income coun- ing, underweight, and wasting; micronutrient de- tries, the probability of survival to age 5 aver- ficiencies; and obesity and overweight (chapter Chapter 2 | Pakistan Human Capital Review ages 93.8  percent. While childhood survival in 4). Malnutrition has serious long-term implica- Punjab (92 percent) is slightly lower than those tions for human capital and affects individual and in Sindh (93 percent) and Khyber Pakhtunkhwa national growth and productivity. While malnu- (94 percent), Balochistan is at the bottom. The trition manifests itself in various forms, the HCI probability of survival to age 5 in Balochistan is focuses on stunting. As an indicator of chronic one of the lowest in the world (89 percent). malnutrition, stunting is most closely associ- ated with brain development, physical growth, As discussed in detail in chapter 4, child malnu- and human capital development. Stunting var- trition is a major complication, caused by low ies widely across provinces, with 48 percent of food intake, poor environmental health (includ- children stunted in Balochistan and 50 percent ing inadequate clean drinking water and sani- in Sindh, compared with the national average of tation), and lack of quality care for women and 40 percent and Punjab’s 30 percent. 47 Low access and retention in which has a LAYS value of 12.8.5 Not only is schooling Pakistan’s value far short of the expected 9.4 years of schooling, but the value also suggests The expected years of schooling for children that almost half the time that students spend in age 5–18 in Pakistan (9.4 years) is just below the school is not spent in learning. Compared with expected value for a country with its GDP per its income and regional peers, Pakistan under- capita, but as with stunting rates, differences performs on the quality of education. across provinces are striking. Punjab performs above the expected value for a country with Pa- The low level of learning outcomes stems from kistan’s GDP per capita but slightly below the a raft of factors, including limited access to South Asia average of 10.8 years. Khyber Pakh- schools, low school readiness of children, poor tunkhwa slightly underperforms both, at 9.0 quality teaching, teacher absences, poor school years, while Sindh and Balochistan’s expected environment, and ineffective management prac- years of schooling are well below expectations tices. With 31.5  percent of three- to four-year- (7.5 and 6.9 years, respectively). Pakistan lags olds having low learning levels, Pakistan ranks all regional and lower-middle-income peers on second from the global bottom on the Early this dimension. Childhood Development Index. A recent survey in Punjab showed that 16  percent of teachers Pakistan has one of the largest out-of-school are not present in the classroom at a given time, populations in the world, with an estimated and 90 percent of those who are present do not 20.3  million school-age children not attending use effective teaching practices, such as provid- school. This is a highly gendered problem, with ing feedback to students or sequencing lessons 37 percent of girls age 5–16 in Pakistan out of appropriately. Teachers have few incentives to school versus 25 percent of boys. A large share perform in the public sector, and there is no re- of children (22 percent of all children age 5–16) lation between teacher quality and teacher ten- never enroll in school. The reasons include low ure.6 Schooling in Pakistan also suffers from po- accessibility and availability of schools, safety litical and social challenges around the language concerns on the way to and from school and of instruction, with only a minority of students work, poverty, and social norms preventing the attending a school in the language of the paren- enrollment of girls. While the private sector led tal home. the expansion in access to school in the 1990s and 2000s, private school enrollment has pla- Impact of the COVID-19 pandemic teaued at roughly 40 percent of all enrollments on human capital in the past decade. Pakistan has slowly raised its HCI value, but dis- Low quality of schooling Chapter 2 | Pakistan Human Capital Review ruptions can quickly set back the gains. Earlier crises in Pakistan have shown that the effects of The learning-adjusted years of schooling (LAYS) school closures can persist for many years. For combines the quantity (access) and quality example, an entire cohort of students age 3–15 (learning outcomes) of schooling in a single at the time of the 2005 earthquake in Pakistan easy-to-understand metric. The HCI defines a had lower academic scores four years later, de- complete, quality education as 14 LAYS, which spite substantial remedial efforts.7 Similar effects serves as a benchmark for comparing perfor- have been seen for other crises, such as the 2010 mance. The LAYS value in Pakistan is only 5.1 floods, which destroyed about 11,000 schools. meaning that the average child learns only the equivalent of 5.1 years of schooling versus a The COVID-19 pandemic is a much larger set- child growing up in countries like Singapore, back for Pakistan’s human capital than any 48 previous crisis. A strained health system cou- FIGURE 2.6  Pakistan’s vaccination coverage for children under 5 pled with people’s fear of infection led many fell from January to April 2020 to avoid treatment or regular health checkups, Percent thus risking Pakistan’s recent gains in health January February March April 100 outcomes (table 2.1). The disruptions in health services and COVID-19-related income losses have worsened child mortality and stunting, 80 with effects that will be felt for decades. School- age children and young people have similarly 60 experienced huge disruptions in access to edu- cation and job opportunities. 40 The COVID-19 lockdown’s impact on routine im- 20 munization has also exacerbated existing ineq- uities in access (figure 2.6). A geospatial analy- 0 sis in Karachi, Sindh, revealed that declines in BCG Penta 3 Measles immunization during the initial lockdown period Source: World Bank calculations using data from Demographic and Health Surveys, House- from April 1–9, 2020, were highest in slums and hold Income and Expenditure Surveys, Pakistan Social and Living Standards Measurement squatter settlements, which already had histor- Surveys, and Labor Force Surveys, 1990–2017, and Human Capital Project 2020. ically low vaccine coverage.8 More affluent city neighborhoods were less affected by lockdown data from a student cohort in Punjab show that restrictions. Between January and March 2021, students scored 10 percentage points lower in the share of services experiencing at least some grade 5 mathematics and 14 percentage points disruptions in Pakistan was slightly higher than lower in English after school closures from the global average of 38 percent.9 March 2020 to September 2021 alone (this pe- riod representing less than half of all school clo- The evidence so far shows significant adverse sures). These impacts are worse on the poor and impacts of the COVID-19 pandemic on educa- vulnerable, exacerbating poverty and inequality.10 tion. Gallup data show that between March 2019 and June 2020, Pakistan lost 12 percent of stu- Based on global simulations for education and dents age 8–15 from the school system, with health, the impact of the COVID-19 pandem- the impact particularly pronounced for adoles- ic could have erased more than eight years of cents. In Punjab, for example, 21 percent of ad- progress on human capital for both boys and olescent boys and 8 percent of adolescent girls girls (figure 2.7). The simulations suggest that had dropped out by September 2021. Further, Pakistan’s HCI would be reduced from 0.406 Chapter 2 | Pakistan Human Capital Review TABLE 2.1  Disruptions to essential services mean fewer people receive services Current coverage Coverage if services People losing services (percent) disrupted (percent) due to disruptions Oral antibiotics for pneumonia in children 79 40 5,424,900 DPT vaccinea for children 75 36 5,441,800 Facility-based delivery 66 34 980,400 Contraceptive prevalence rateb 37 23 4,021,800 Source: Global Financing Facility. a. Diphtheria, pertussis (whooping cough), and tetanus vaccinations.  b. Includes both modern and traditional methods and is calculated for married women only.  49 FIGURE 2.7  The COVID-19 pandemic may have wiped out more The floods took the lives of more than 1,700 peo- than eight years of gains on Pakistan’s Human Capital Index ple, one-third of them children. Rain-induced Human Capital Index floods, accelerated glacial melt, and the resulting 0.5 landslides devastated millions of homes and key infrastructure, submerging entire villages and destroying livelihoods. The scale of the disas- ter is unprecedented in Pakistan, exceeding the 0.42 0.4 0.41 0.40 0.40 damages of the 2010 floods. About 94 districts­ 0.39 0.39 0.38 —­ —­ more than half of all districts in Pakistan­ were 0.37 0.37 declared calamity hit. Of the 25 poorest districts, 19 were calamity affected. Initial estimates by 0.3 the World Bank suggest that, as a direct conse- quence of the floods, the national poverty rate has likely increased by 3.7–4.0 percentage points, 0.2 pushing 8.4–9.1 million people into poverty. The 2012 2020 2021 2012 2020 2021 2012 2020 2021 floods reveal Pakistan’s high vulnerability to cli- Total Male Female mate change despite accounting for less than Source: World Bank calculations using data from Demographic and Health Surveys, House- 1 percent to global greenhouse gas emissions. hold Income and Expenditure Surveys, Pakistan Social and Living Standards Measurement Surveys, and Labor Force Surveys, 1990–2017, and Human Capital Project 2020. The floods exacerbated the already dire levels of learning poverty and malnutrition. Standing water, food insecurity, malnutrition, and water-­ to 0.373, a reduction of 0.033, lower than its borne and vector-borne diseases are serious 2012 level. The main culprit is a reduction in the risks for human capital development. Natural quantity of schooling (due to dropouts) and in disasters not only cause death and destruction quality (due to learning loss). Between February to physical capital; they also have large and long- 2020 and August 2021, 67–84  percent of the term effects on children’s opportunities to accu- school system was closed, where ranges varied mulate human capital due to the adverse impacts depending on assumptions on partial or com- on nutrition, education, health, and income.13 In plete closures.11 Another factor is child survival, most of the flood-affected districts, 50  percent given that child mortality is expected to have or more of children under 5 were already stunt- increased sharply due to reduced household in- ed. Almost 24,000 schools have been damaged comes and health services during lockdown pe- or destroyed, most of them in Sindh. More than riods. No simulations are available for stunting 7,000 schools had to accommodate people dis- or adult survival, making it likely that the simu- placed by flooding, so they could not provide Chapter 2 | Pakistan Human Capital Review lations here present lower bounds of the full im- schooling. As a result, an estimated 3.5  million pact of the pandemic on human capital. children had their schooling disrupted.14 Impact of the 2022 floods on The floods made it even more difficult for chil- human capital dren to attend school and receive quality learning. Before the floods, about 20.3 million Between June and August 2022, torrential rains children were out of school, and 75 percent of and riverine, urban, and flash flooding put about 10-year-olds could not read an age-appropriate one-third of Pakistan under water.12 According text. With the COVID-19 pandemic and the to the National Disaster Management Authority, floods, learning poverty could increase to around 33  million people were affected by the 80  percent, according to simulations by the floods, nearly 8 million of whom were displaced. World Bank. The floods affected access to safe 50 drinking water, transport, and basic services. As for children and older family members. And with most disasters, the poorest families were health risks have increased due to water-borne hit the hardest. Food insecurity has become and associated diseases, including dengue and acute, affecting nutritional and health outcomes malaria (box 2.1). BOX 2.1 Human capital impacts of the 2022 floods To assess the human development impacts of the floods, Box figure 1 Almost one-third of families with children a national phone survey gathered information from age 3–17 reported some flooding in their area about 4,000 households with children age 3–17. Original- Completely ooded ly designed to track children’s return to school following 14% school closures due to the COVID-19 pandemic, the sur- vey helped document and understand the challenges fac- Partially ooded 6% ing families with children. Any ooding 26% 7% Somewhat ooded Impacts on families with children No Of households with children age 3–17, 26 percent report- ooding ed living in areas somewhat, partially, or completely flood- 74% ed, including 14 percent in completely flooded areas (box figure 1). These high percentages show the detrimental effects on the future human development of a substan- Source: World Bank phone survey. tial share of children. About 5 percent of families reported having married off one of their children to cope with the Box figure 2 Families with children reported direct impacts, and 15 percent reported that their children are impacts to their household and income-related likely to stay home to help with household chores and activities, and indirect impacts through limited reconstruction. Of parents surveyed, 28  percent envi- access to public goods sioned their child working instead of returning to school Damage to house due to the adverse economic impacts of the floods, with Loss of crops the rate of child labor expected to be the highest in Bal- Loss of livestock ochistan (41 percent against about 27 percent in Khyber Damage to roads Chapter 2 | Pakistan Human Capital Review Pakhtunkhwa and Sindh). Loss of house Electricity connection Of families with children in flooded areas, 39 percent re- Means of earning a living ported losing agricultural crops, 32  percent losing live- Damage to bridges stock, and 15–29 percent damage to roads, bridges, and Mobile phone outages electricity (box figure 2). There were also direct hits to Severe illness households, with 44  percent reporting damage to their Loss of life house and 26 percent reporting losing their house. One in 0 10 20 30 40 50 five households reported losing the means to earn a living, Percent casting doubt on their capacity to quickly recover. (continued) Source: World Bank phone survey. 51 BOX 2.1, continued Impact on children’s education Box figure 4 Engagement of children in learning Of the families with flooding in their area, just over half activities during school closures due to floods reported that the floods hurt their children’s education Children engaged in learning activities (box figure 3). More than a quarter reported that their children’s schools were disrupted and/or damaged, and almost a third reported that their children’s school was completely damaged. Damages to schools due to the Yes 28% floods occurred most often in Sindh and Balochistan. Approximately 17  percent of schools were also used as temporary shelters for displaced families, which further No 72% depleted school infrastructure. Box figure 3 More than half of families interviewed reported that the floods hurt their children’s education Studied at home No impact on education Teacher visits Received educational materials Physical damage to school Temporary learning centers Distance Sessions disrupted learning 0 10 20 30 40 50 0 10 20 30 40 50 Percent Percent Source: World Bank estimates based on phone survey. Source: World Bank phone survey. transport, roads, and other services. This aligns with find- About 72 percent of parents reported that their children ings that parents in Pakistan generally consider transpor- were not studying during flood-caused school closures (box tation services to be a key determinant of school atten- figure 4), while 28 percent of parents said their children were dance for girls and young women, even before the floods.1 Chapter 2 | Pakistan Human Capital Review continuing to learn. Among the latter group, 28 percent re- ported that teachers visited their children, and 20 percent Health impacts reported that their children received printed educational Survey respondents said health facilities were either materials. About 9  percent indicated that their children completely damaged (21  percent) or partially damaged continued their studies in temporary learning centers, (25 percent) in their area; 17 percent of respondents said and just 4 percent participated in distance learning. these facilities were somewhat damaged (box figure 5). In short, only one in three healthcare facilities in areas About 33 percent of parents agreed that it would be more covered by the survey seem to have survived the floods difficult for girls than boys to go back to school in the areas unscathed. The lack of health facilities can reduce school where roads had been destroyed. Parents were more attendance as areas underwater generate disease. concerned for girls than boys about the deterioration of (continued) 52 BOX 2.1, continued Box figure 5 Almost two-thirds of health facilities Importantly, in response to the phone survey, 28 percent were damaged by floods reported malaria and dengue outbreaks in the flooded areas. Unlike the unequal impact of the floods on differ- Completely damaged ent households with different education backgrounds, flood-related diseases show no preference­ all levels of —­ education attainment bear the burden equally. In addi- Partially damaged tion, 32 percent of households reported a lack of regu- lar access to clean drinking water, one of most common sources of infectious diseases. Somewhat damaged Note 1. Baron et al. forthcoming. Not damaged and functional Source: Bend and others 2022. 0 10 20 30 40 50 Percent Source: World Bank phone survey. INEQUALITIES IN HUMAN and even in girls’ favor (see figure 2.8, panels a CAPITAL FORMATION and b). This means that almost all the gender in- equality on the HCI comes from the education Understanding how Pakistan can improve its component, not from the health component. human capital outcomes requires an analysis that adopts a longer timeframe and identifies Girls in school have far higher learning out- the many interventions that can lead to positive comes, about 38 points ahead of boys on the changes.15 The persistent gaps between wealth 2019 TIMSS Science test, which is equivalent groups are only one of several dimensions of to more than a year of schooling in an average the inequality that continue to undermine Paki- country. This again shows the great potential for stan’s human capital potential. The challenges Pakistan to raise its overall HCI value by improv- of low human capital accumulation and utiliza- ing human capital accumulation (education) tion are compounded by significant inequalities and utilization (employment) for girls. between men and women and among provinces and urban–rural locations. Girls are disadvantaged mostly by lower initial Chapter 2 | Pakistan Human Capital Review enrollment (figure 2.9). The peak of the figure Gender inequalities shows the share of children who ever enroll in school at any age: 86  percent of boys and While gender gaps run deep across Pakistan’s 74 percent of girls ever enroll in school. During society, gender inequality on the HCI is not as the first five years of schooling, there are no vis- large as one might expect. The main source of ible gender differences in dropout patterns. But gender inequality in human capital accumu- there is a large dropout of girls after grade 5, the lation is the expected years of schooling, with end of primary school. About 23 percent of boys boys spending more time in school (9.7 years) and 28 percent of girls are lost from the school than girls (8.3 years) (figure 2.8). Gender gaps in system at this transition. After grade 8, boys and the probability of survival to age 5 and the frac- girls drop out at similar rates, and by grade 10, tion of children under 5 not stunted are small the inequality between boys and girls is virtually 53 FIGURE 2.8  Girls’ outcomes are worse than boys in access to school, but not in other components of the Human Capital Index Percent Percent a. Probability of survival to age 5 b. Children under 5 not stunted 100 Low-income country average 100 80 80 Low-income country average 60 60 40 40 20 20 0 0 Male Female Male Female c. Expected years of schooling d. Learning outcomes 12 400 Low international benchmark 10 Low-income country average 300 8 6 200 4 100 2 0 0 Male Female Male Female Source: Demographic and Health Survey 2017; Trends in International Mathematics and Science Study (TIMSS) Science test 2019. Note: Only one round of TIMSS has been conducted in Pakistan. FIGURE 2.9  Girls enroll in much lower numbers than boys Percent 100 80 Chapter 2 | Pakistan Human Capital Review 60 40 20 Boys Girls 0 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Age Source: Pakistan Social and Living Standards Measurement Survey 2018/19. 54 eliminated. The large gap between boys and probability of survival to age 5, and those girls at the end of primary school is due to inad- from the poorest, 92  percent. The difference equate access to schools beyond grade 5 (es- in stunting is also large, with fewer than half pecially in rural areas, where most schools offer (44  percent) of children in the poorest quin- only primary education) and to parental views tile not stunted, versus more than three-quar- that primary education is enough for girls. ters (78 percent) in the richest. The difference between rich and poor is most pronounced in Socioeconomic inequalities expected years of schooling, with the richest quintile spending twice as long in school (12.7 As discussed, even the richest groups in Pa- years) as the poorest (6.4 years). However, the kistan have relatively low human capital out- difference in learning outcomes is not as large comes in international comparisons. That said, as one might expect, with about 34 points in human capital outcomes vary widely by socio- the TIMSS Science score in favor of children economic background (figure 2.10). Children of highly educated parents, or about a year of from the richest two quintiles have a 95 percent learning in an average country. FIGURE 2.10  Components of the Human Capital Index by wealth quintiles Percent Percent a. Probability of survival to age 5 b. Children under 5 not stunted 100 100 Low income average, 93 Low income average, 66 80 80 60 60 40 40 20 20 0 0 Poorest Poor Middle Rich Richest Poorest Poor Middle Rich Richest c. Expected years of schooling d. Learning outcomes 15 400 Low international benchmark, 400 Low income average, 7.8 12 300 Chapter 2 | Pakistan Human Capital Review 9 200 6 100 3 0 0 Poorest Poor Middle Rich Richest Poorest Poor Middle Rich Richest Source: Demographic and Health Surveys 2017; Trends in International Mathematics and Science Study (TIMSS) Science test 2019. Note: Only one round of TIMSS has been conducted in Pakistan. 55 Inequalities in human capital outcomes be- the richest and poorest quintiles has increased tween the richest two quintiles, and the poorest from 0.18 to 0.19. And the gap has widened be- two quintiles appear to have grown in the past tween the second and fourth richest quintiles, 30 years (figure 2.11). The richest two quintiles from 0.04 to 0.09. This means that Pakistan have improved to be roughly around the lower- has raised its HCI value by investing more in the middle-income country average (HCI value of richest quintiles than in the poorest two. 0.48), whereas the poorest two quintiles remain below the Sub-­ Saharan African average (HCI The gender gap in human capital accumulation value of 0.40). The gap in HCI values between also varies with wealth (figure 2.12). Between 1990 and 2017, the gap widened for those in FIGURE 2.11  The Human Capital Index gap between the richest the poorest quintile but narrowed for those in and poorest groups has persisted over the past three decades all other wealth quintiles. For the poorest quin- Human Capital Index value tile, the HCI gender gap increased from 0.03 to 1.0 0.04, as the HCI value rose from 0.27 to 0.31 for females and from 0.31 to 0.35 for males. In the 0.8 second richest quintile, the female HCI value im- High-income countries in 2020 proved from 0.35 to 0.47, surpassing the male HCI value in 2017, which improved from 0.39 to 0.6 Lower-middle-income countries in 2020 0.46. The gap was minimal among those in the Richest Richer richest quintile over the period. 0.4 Sub-Saharan Poorer Africa in 2020 Poorest Geographic inequalities 0.2 Breaking down HCI values by province sug- 0.0 gests wide inequality of outcomes: the lowest- 1990 2012 2017 performing province in Pakistan­ —­ Balochistan (0.32)­—­is at the global bottom, at the same Source: World Bank calculations using data from Demographic and Health Surveys 1990– 2017. Official Human Capital Index values are from the Human Capital Project. level as Niger. Sindh, with an HCI value of 0.36, is Note: The middle quintile is excluded to allow better comparison against benchmark values. comparable to Nigeria and Sierra Leone (0.36). Khyber Pakhtunkhwa, with a HCI value of 0.39, FIGURE 2.12  Pakistan Human Capital Index values by gender and is comparable to Burundi and Tanzania. The wealth HCI value of the highest-performing province in Female Male Pakistan­—­Punjab (0.42)­—­is comparable to that 1990 of Senegal (0.42) and just below that of South Poorest 2012 2017 Africa (0.43). Chapter 2 | Pakistan Human Capital Review 1990 Poor 2012 North Punjab has the highest HCI value (about 2017 1990 0.50), while rural Sindh and Balochistan are Middle 2012 home to districts with some of Pakistan’s low- 2017 1990 est HCI values (about 0.25). In other words, Rich 2012 children born in the top-performing districts 2017 of Pakistan today can expect to have, on aver- 1990 Richest 2012 age, twice as much human capital (be twice as 2017 productive) by age 18 as children born in the 0.0 0.1 0.2 0.3 0.4 0.5 0.6 lowest-performing districts. Without urgent re- Human Capital Index value medial action, this will reinforce the intergener- Source: World Bank calculations using data from Demographic Health Surveys 1990–2017. ational cycle of inequality. 56 FIGURE 2.14  Human Capital Index components in Pakistan’s provinces vary little on survival but widely on stunting Changes in provincial inequalities Percent a. Probability of survival to age 5 Punjab has had the highest HCI value among Total Male Female 100 the provinces for the past three decades, ris- ing from about 0.34 in 1990 to about 0.42 80 in 2019 (figure 2.13). Khyber Pakhtunkhwa had the fastest rate of increase in HCI of any province, going from 0.31 in 1990 to 0.39 in 60 2012, but then levelling off. Sindh and Baloch- istan grew at the same rate over the 30 years 40 (0.05), although from different baselines. Bal- ochistan improved at a slower pace, with its 20 HCI value increasing from about 0.27 to 0.32, while Sindh’s HCI value improved from 0.31 to 0 0.36 over the same period. Punjab Sindh Khyber Balochistan Pakhtunkhwa Pakistan’s provinces do not show large varia- b. Children under 5 not stunted Total Male Female tions in the probability of survival to age 5 but 100 exhibit wide differences in under-5 stunting rates (figure 2.14). Punjab has the lowest stunt- 80 ing rate at 30  percent, while half the children under 5 in Sindh are stunted. 60 Punjab generally has higher expected years of 40 schooling for boys and girls, and Balochistan the lowest (figure 2.14, panel c). The gender gap 20 FIGURE 2.13  Human Capital Index trend in Pakistan, by province 0 Punjab Sindh Khyber Balochistan Human Capital Index value Pakhtunkhwa 0.6 c. Expected years of schooling Total Male Female 12 0.5 Punjab 0.42 0.4 0.40 Khyber Pakhtunkhwa 9 0.36 Chapter 2 | Pakistan Human Capital Review 0.34 0.31 Sindh 0.32 0.3 0.31 0.27 Balochistan 6 0.2 0.1 3 0.0 1990 2012 2017 2019 0 Source: Demographic and Health Surveys 2017; Trends in Inter- Punjab Sindh Khyber Balochistan Pakhtunkhwa national Mathematics and Science Study (TIMSS) Science test 2019. Source: Demographic and Health Surveys 2017; Trends in International Mathematics and Note: Only one round of TIMSS has been conducted in Pakistan. Science Study (TIMSS) Science test 2019. Note: Only one round of TIMSS has been conducted in Pakistan. 57 for enrollment is the smallest in Punjab at about gender is much larger, at 38 TIMSS points, than 3 percentage points, against 15–19  percentage that for urban–rural location. points for the other three provinces. The decomposition of the changes in HCI values WHY DOES PAKISTAN UNDERUTILIZE into its components shows that the apparent ITS HUMAN CAPITAL? provincial inequality stemmed from changes in both health (stunting) and education (enroll- Women’s participation in the labor force in- ment and dropout). creased from the 1990s through the 2000s. In the past few years, close to half of women with Urban–rural inequalities postsecondary degrees consider themselves to be a full part of the labor market. But growth Gaps in human capital outcomes between urban stalled during the past decade. And women and rural areas are large, though they have de- with primary and secondary education do not clined a little over the past three decades. From have better (nonagricultural) employment. While 1990 to 2019, the HCI value rose from 0.38 women are gradually increasing their level of ed- to 0.43 in urban areas and from 0.30 to 0.38 ucation, this trend also goes hand in hand with in rural areas (figure 2.15). The difference in increasing unemployment, except for women HCI values between urban and rural areas has with higher education (figures 2.17 and 2.18). shrunk faster in the most recent decade, due to accelerated gains in rural areas. While education reduces the gender gap in em- ployment, structural constraints in culture, in- The main sources of the urban–rural inequality cluding conservative attitudes and gendered in human capital are the differences in under-5 division-of-labor norms, also strongly influence stunting rates and expected years of school- women’s labor outcomes. Some women are ing (figure 2.16). For TIMSS, the difference for willing to work if they can find employment and have safe transport to and from their jobs, and FIGURE 2.15  In three decades, rural areas reducing physical mobility constraints has a have closed the Human Capital Index gap with large impact on women but not men.16 urban areas a little Human Capital Index value So, by directly addressing labor market con- 0.6 straints for women, such as through empha- sizing higher education and increasing physical 0.5 mobility and safety to and from school and work, Urban 0.43 Pakistan can reduce gender employment gaps Chapter 2 | Pakistan Human Capital Review 0.4 0.38 0.38 and improve the utilization of its human capital. 0.30 Rural 0.3 CONCLUSION AND 0.2 RECOMMENDATIONS 0.1 Pakistan has made some progress in its human 0.0 capital over the past three decades, but its 1990 2012 2017 2019 overall levels of human capital remain low, and its human capital growth is not fast enough to Source: Demographic and Health Surveys, Household Income and Expenditure Surveys, and Pakistan Social and Living Stan- catch up with other lower-middle-income coun- dards Measurement Surveys. tries. Moreover, inequalities persist, with human 58 FIGURE 2.16  Human Capital Index components in Pakistan by economic development Percent Percent a. Probability of survival to age 5 b. Children under 5 not stunted 100 100 Low-income country average 80 80 Low-income country average 60 60 40 40 20 20 0 0 Urban Rural Urban Rural c. Expected years of schooling d. Learning outcomes 12 400 Low international benchmark 10 Low-income country average 300 8 6 200 4 100 2 0 0 Urban Rural Urban Rural Source: Demographic and Health Surveys, Household Income and Expenditure Surveys, and Pakistan Social and Living Standards Mea- surement Surveys. capital gaps remaining similar over time between hard to undo and will require a more systematic rich and poor households, men and women, and approach to put improving human capital out- rural and urban areas. Among women, human comes back on track. The 2022 floods further Chapter 2 | Pakistan Human Capital Review capital is severely underutilized due to low invest- limited access to health and education services ments in education and to societal norms that and likely increased malnutri­ tion and learning hinder their participation in the labor market. poverty. The COVID-19 pandemic and the 2022 floods Investing in girls’ education and training is an im- have been serious disruptions to human capital portant precondition to increasing female labor formation in Pakistan. Delays in key health serv- force participation. Addressing both demand ices such as immunization and losses in learning and supply constraints that limit girls’ education and dropouts due to school closures are esti- should be a priority. Social norms are among the mated to have reversed eight years of hard-won most powerful factors in determining women’s gains on the HCI. Without quick and sustained interactions in the public sphere. Possible inter- action, this reversal will become increasingly ventions to influence norms include strategic 59 FIGURE 2.17  Labor force participation of FIGURE 2.18  For women especially, having women continues to grow for women with post- a postsecondary education means a higher secondary education chance of employment Percent Percent a. Labor force participation, age 15–65 a. Male unemployment, age 15–65 100 50 Male 80 40 60 30 Total 40 20 Female Postsecondary Lower secondary 20 10 No education 0 0 1990 1995 2000 2005 2010 2015 2018 2000 2004 2008 2011 2014 2018 b. Female labor force participation, age 15–65 b. Female unemployment, age 15–65 50 50 Postsecondary Postsecondary 40 40 30 30 No education 20 20 Primary and Lower secondary secondary 10 10 No education 0 0 2000 2004 2008 2012 2016 2018 2000 2004 2008 2011 2014 2018 Source: World Development Indicators; World Bank calculations Source: World Bank calculations using data from Labor Force using data from Labor Force Surveys. Surveys. Note: Respondents with only primary education and upper-­ secondary education are excluded to avoid overcrowding the graphs. Chapter 2 | Pakistan Human Capital Review use of positive messaging about strong female accelerating progress in human capital devel- role models. And engaging men is crucial in opment is essential. The COVID-19 situation changing norms surrounding women’s econom- appears to be easing, but with macro-fiscal —­ ic activities. Informational nudges­ particularly risks growing, there is a risk that investments in those emphasizing the difference in earnings human development may be cut back. However, between females and males­ —­can encourage this is not a time to take the foot off the accel- women to enroll in training programs to enter erator. Continued and deliberate multisectoral more male-dominated occupations. and whole-of-government efforts are essential to build strong education, health, and social pro- As Pakistan seeks to become an upper-middle- tection systems to prepare for and mitigate po- income country by its centennial in 2047, tential future shocks. 60 NOTES 1. The evidence, primarily from developed countries, 8. Chandir et al. 2020. suggests a potential mismatch between the require- 9. WHO 2021. ments of new tasks and technologies and the skills 10. See chapter 3 for a snapshot of the household-­ of the current workforce (Autor and Dorn 2013). level impacts of the COVID-19 pandemic on young 2. No historical data on the quality of schooling are families. available for Pakistan. The first time Pakistan par- 11. Azevedo et al. 2022. ticipated in an internationally benchmarked student assessment was in 2019 (TIMSS). 12. Government of Pakistan 2022. 3. Provincial disaggregation of LAYS is not possible. 13. Baez, de la Fuente, and Santos 2010. 4. The probability of survival to age 5 is calculated as 14. NDMA 2022; UNICEF 2022. the complement of the under-5 mortality rate. The 15. Changes in HCI values over a short period are typi- under-5 mortality rate is the probability of a child cally small and might simply reflect updates to com- born in a specified year dying before reaching the ponents that are measured sporadically, rather than age of 5 if subject to current age-specific mortality actual changes in underlying outcomes. In contrast, rates. the analysis of longer-term trends has a more solid 5. World Bank 2020. basis, given the scope for smoothing out short-run idiosyncrasies. 6. Bau and Das 2020. 16. Field et al. 2020. 7. Andrabi, Daniels, and Das 2020. REFERENCES Andrabi, T., B. Daniels, and J. Das. 2020. “Human Capital Baez, J., and I. V. Santos. 2007. “Children’s Vulnerability Accumulation and Disasters: Evidence from the Pakistan to Weather Shocks: A Natural Disaster as a Natural Ex- Earthquake of 2005.” Journal of Human Resources 57 (2). periment.” Working paper, Active Learning Network for Accountability and Performance in Humanitarian Action, Andrabi, T., B. Daniels, and J. Das. 2021. “Human Capital London, UK. Accumulation and Disasters: Evidence from the Pakistan Earthquake of 2005.” The Journal of Human Resources Baron, J., I. Trako, N. Ejaz, and M. Bend. Forthcoming. (June 2021): 1–67. http:/ /jhr.uwpress.org/content/early​ “Constraints to Girls’ Education in Pakistan.” Washington, /2021/06/02/jhr.59.2.0520-10887R1. DC: World Bank. Autor, D. H., and D. Dorn. 2013. “The Growth of Low-Skill Baron, J., M. Bend, E. M. Roseo, and I. Farrakh. 2022. Service Jobs and the Polarization of the US Labor Mar- “Floods in Pakistan: Human Development at Risk.” Wash- ket.” American Economic Review 103 (5): 1553–97. ington, DC: World Bank. https:/ /openknowledge.world​ Chapter 2 | Pakistan Human Capital Review bank.org/handle/10986/38403. Azevedo, J. P., A. Hasan, D. Goldemberg, K. Geven, and S. A. Iqbal. 2022. “Simulating the Potential Impacts of Bau, N., and J. Das. 2020. “Teacher Value Added in a COVID-19 School Closures on Schooling and Learning Low-Income Country.” American Economic Journal: Eco- Outcomes: A Set of Global Estimates.” The World Bank nomic Policy 12 (1): 62–96. Research Observer 36 (1): 1–40. https://doi.org/10.1093​ Beegle, K., R. H. Dehejia, and R. Gatti. 2003. “Child Labor, /wbro/lkab003. Income Shocks, and Access to Credit. Policy Research.” Baez, J., A. de la Fuente, and I. V. Santos. 2010. “Do Nat- Working Paper, 3075, World Bank, Washington, DC. https:/ / ural Disasters Affect Human Capital? An Assessment openknowledge.worldbank.org/handle/10986​/18222. Based on Existing Empirical Evidence.” IZA Discus- Chandir S., D. A. Siddiqi, H. Setayesh, and A. J. Khan. sion Paper 5164, Institute for the Study of Labor, Bonn, 2020. “Impact of COVID-19 Lockdown on Routine Germany. 61 Immunisation in Karachi, Pakistan.” Lancet Global Health Jacoby, H., and E. Skoufias. 1997. “Risk, Financial Mar- 8 (9): e1118–e1120. kets, and Human Capital in a Developing Country.” Review of Economic Studies 64 (3): 311–335. De Janvry, A., F. Finan, E. Sadoulet, and R. Vakis. 2006. “Can Conditional Cash Transfer Programs Serve as Safe- Jensen, R. 2000. “Agricultural Volatility and Investments ty Nets in Keeping Children at School and From Working in Children.” American Economic Review 90 (2): 399–404. When Exposed to Shocks?” Journal of Development Eco- Kousky, C. 2016. “Impacts of Natural Disasters on Chil- nomics 79 (2): 349–373. dren.” The Future of Children 26 (1): 73–92. http://www​ Dong, Y., and H. Yang. 2020. “Natural Disaster and Child .jstor.org/stable/43755231. Labor: Evidence from Indonesian Earthquake,” February Lai, B. S., B. A. Auslander, S. L. Fitzpatrick, and V. Pod- 21, 2020. https://ssrn.com/abstract=3542086 or http:/ /​ kowirow. 2014. “Disasters and Depressive Symptoms in dx.doi.org/10.2139/ssrn.3542086. Children: A Review.” Child and Youth Care Forum 43 (4): Field, E., S. U. Junaid, H. Majid, A. Malik, A. Shahid, and K. 489–504. https:/ /doi.org/10.1007/s10566-014-9249-y. Vyborny. 2020. Transport and Urban Labour Market Inte- Lai, B. S., and A. La Greca. 2020. “Understanding the gration: Evidence on Travel Time and Congestion from a impacts of natural disaster on children.” Society for Re- Mass Transit Quasi-experimental Evaluation and Evidence search in Child Development (SRCD) Child Evidence Brief on Firms from a Randomised Control Trial in Pakistan, 3ie 8, August. Grantee Final Report. New Delhi: International Initiative for Impact Evaluation (3ie). NDMA (National Disaster Management Authority). 2022. “NDMA Floods (2022) SITREP–2022.” Islamabad, Paki- Global Financing Facility. 2020. “Preserve Essential stan: NDMA. http:/ /cms​.ndma​.gov​.pk​/storage​/app​/public​/ Health Services during the Covid-19 Pandemic: Pakistan.” situation​-reports​/October2022​​/9WVaA9l0lwTcmtxz​P1fa.pdf. https://www.globalfinancingfacility.org/sites/gff_new​ /files/documents/Pakistan-Covid-Brief-GFF.pdf. Takasaki, Y. 2017. “Do Natural Disasters Decrease the Gender Gap in Schooling?” World Development 94 (June Government of Pakistan. 2021. “PSLM — ­ ­2019–20: Pa- 2017): 75–89. kistan Social and Living Standards Measurement Sur- vey.” Pakistan Bureau of Statistics, Ministry of Planning Tang, B., X. Liu, Y. Liu, C. Xue, and L. Zhang. 2014. “A Me- Development and Special Initiatives: Government of ta-Analysis of Risk Factors for Depression in Adults and Pakistan. https://www.pbs.gov.pk/sites/default/files//​ Children after Natural Disasters.” BMC Public Health 14 pslm/publications/pslm_district_2019-20/Key_Finding​ (1): 623. https://doi.org/10.1186/1471-2458-14-623. _Report_of_PSLM_District_Level_Survey_2019-20.pdf. UNICEF (United Nations Children’s Fund). 2022. “Pakistan: Government of Pakistan. 2022. “Pakistan Floods 2022: Humanitarian Situation Report, 3,” September 20, 2022. Post-Disaster Damage Assessment (PDNA).” Ministry https://www.unicef.org/media/127711/file​/ Pakistan​ of Planning Development and Special Initiatives: Gov- -Humanitarian-SitRep-No.3-(Floods)-20-Sep​-2022.pdf. ernment of Pakistan. https:/ /www.pc.gov.pk/uploads​ UNICEF (United Nations Children’s Fund) Pakistan. /downloads/PDNA-2022.pdf 2018. “National Nutrition Survey: Key Findings Report.” Chapter 2 | Pakistan Human Capital Review Hallegatte, S., A. Vogt-Schilb, J. Rozenberg, M. Bangalore, UNICEF, Islamabad, Pakistan. https:/ /www​.unicef​.org​ and C. Beaudet. 2020. “From Poverty to Disaster and Back: /pakistan ​ /media ​ / 1871 ​ /file ​ / KeyFindings​​ - --National​ a Review of the Literature.” EconDisCliCha 4: 223–247. Nutrition​Survey2018.pdf. Hicks, J. H., M. Kremer, I. Mbiti, and E. Miguel. 2011. “Voca- UNOSAT (United Nations Satellite Centre). 2022. “Pre- tional Education Voucher Delivery and Labor Market Re- liminary Satellite Derived Flood Evaluations Assessment turns: A Randomized Evaluation among Kenyan Youths.” –Islamic Republic of Pakistan” (map), October 12, 2022. Washington, DC: World Bank. https://reliefweb.int/map/pakistan/preliminary-satellite​ -derived-flood-evolution-assessment-islamic-republic​ Husted, L., I. M. Opper, and R. J. Park. 2022. “The Impact of -pakistan-11-oct-2022. Natural Disasters on Human Capital.” EdWorkingPaper, 22– 548, Annenberg Institute for School Reform at Brown Uni- Ureta, M. 2006. “Hurricane Mitch, Family Budgets versity, London, UK. https://doi.org/10.26300/bkah​-we19. and Schooling in Nicaragua.” Paper prepared for the 62 Inter-American Development Bank conference, Washing- https://www.who.int/publications/i/item/WHO​ - 2019​ ton, DC., February 10, 2005. https://www.cerge-ei.cz/pdf​ -nCoV-EHS-continuity-survey-2021.1. /events/papers/051114_t.pdf. World Bank. 2020. The Human Capital Index 2020 Up- WFP (World Food Programme). 2022. “WFP Pakistan date: Human Capital in the Time of COVID-19. Washing- Situation Report,” September 19, 2022. https://reliefweb​ ton, DC: World Bank. .int/report/pakistan/wfp-pakistan-situation-report-19​ World Bank. 2022. Inflation and the Poor. Washington, DC: -september-2022. World Bank. https://reliefweb.int/report/pakistan​/pakistan​ WHO (World Health Organization). 2021. Second Round of -development-update-october-2022-inflation​-and-poor. the National Pulse Survey on Continuity of Essential Health Services during the COVID-19 Pandemic. Geneva: WHO. Chapter 2 | Pakistan Human Capital Review 63 Fostering early childhood development SUMMARY increase. Across provinces, only 39–59 percent 3 of children are reported by their parents to be Inadequate early childhood investments and developmentally on track by age 3–4, well below outcomes undermine Pakistan’s human capi- the 75  percent average for low- and middle- tal development over the life cycle. From birth income countries. to age 8, brain development is at its most rapid and is highly sensitive to inputs related to nur- Parenting and ECE are among the critical areas turing care and stimulation. Due to inadequate to address. Early stimulation is low, and harsh supports, young Pakistani children’s health and parenting is common in Pakistan. Across prov- learning levels have been persistently low, espe- inces, between 28  percent and 48  percent of cially for children in poor households and rural parents of children age 2–5 engage in early areas. stimulation activities such as singing, talking, or reading with children. The use of harsh parenting­ Policy and program environments that should —­psychological or physical aggression­ —­is high, support early childhood development (ECD) are ranging from 57  percent to 87  percent across uneven across four core sectors that underpin provinces. Moreover, only 19 percent of children ECD services: health and nutrition, water and age 3–5 were enrolled in preprimary school in sanitation, education, and social protection. 2019. Enrollment is especially low for children In addition to varying capacity and resources in the poorest quintile, for children living in rural across provinces, challenges include the multi- areas, and for children with a disability. Gender sectoral nature of early childhood services, insuf- disparities in enrollment emerge as early as pre- ficient financing, inadequate workforce devel- primary school. Among children age 5, when opment, unequal parenting capacity, and lack of compulsory education begins, 58 percent were quality assurance systems. Data on early child- out of school in 2019–20. Enrollment declined hood development and relevant resources and for all ages of young children during 2018–20. services have not been readily available, con- sistent, or highlighted as a key determinant of Parental distress and low access to resourc- human capital. However, Pakistan is undertak- es can impair young children’s development Chapter 3 | Pakistan Human Capital Review ing a policy dialogue to create a national early trajectories. Almost two-thirds of households childhood development framework. reported losing income during the COVID-19 pandemic, and almost half ran out of food more Young children in Pakistan are not developing frequently. Higher distress­ —­measured by care- as well as their low- and middle-income country givers’ inability to be affectionate to their chil- peers. Several risk factors impede their develop- dren and problems with sleep, worry, irritability ment and learning in the early years, including or anger, and nervousness or anxiety­ —­was sig- poverty, stunting, living in a rural area, disabil- nificantly correlated with lower childhood devel- ity, low maternal education, harsh parenting, opment levels among children age 0–2.1 inadequate early stimulation at home, and lack of early childhood education (ECE). Child de- The costs of continued inaction on ECD in Paki- velopment outcomes fall as these risk factors stan are very high. Consider ECE as an example: 65 failing to achieve what peer countries have al- in various sectors support or undermine ECD ready managed to achieve­ —­preprimary educa- and highlights gaps that need to be redressed. tion for roughly two-thirds of eligible children­ It closes with recommendations on ways to ele- —­ percent of its GDP for will cost Pakistan 1.8  vate and accelerate progress in ECD. every cohort that continues to be deprived of quality, age-appropriate ECE. This conservative- Expanded and effective public investment in ly amounts to US$4.7 billion for each cohort of human development starting before birth is children age 3–5. needed. The early childhood years, which begin at conception and continue to age 8, are a unique window of opportunity for such invest- INTRODUCTION ment. Among Pakistan’s priorities for the ac- cumulation of human capital are reductions in Investing in high-quality programs that support stunting and in learning poverty (being unable young children’s health, nutrition, and early to read and understand a simple text by age learning improves learning outcomes and ulti- 10), which require increasing the amount and mately increases adult wages and productivity. quality of investment in ECD through program- The first five years of life are the fastest period ming that promotes nurturing care, including of human growth; indeed, as much as 90 per- early stimulation, in children’s first 1,000 days, cent of an individual’s brain development oc- universal access to high-quality ECE, and the curs by age 5.2 Smart investments in the phys- provision of healthcare, nutrition, and safe en- ical, cognitive, linguistic, and social-emotional vironments. If the attention paid to young chil- development of young children are critical to dren and the investments in ECD remain at their put them on the path to greater well-being and current low levels, despite the additional shocks prosperity and to help countries break the cycle created by the COVID-19 pandemic and the of poverty, address inequality, and be more pro- 2022 floods, human development in Pakistan ductive and competitive in a rapidly changing may be undermined even more than before the global economy.3 pandemic. This chapter presents a multisectoral examina- tion of the national and provincial landscapes EARLY CHILDHOOD DEVELOPMENT for ECD in Pakistan. It explains the critical im- OUTCOMES ARE LOW portance of improving investments and services for young children and families if Pakistan is to Outcomes for child health, nutrition, and educa- realize its goal of building human capital. The tion are persistently low in Pakistan relative to chapter describes Pakistan’s status quo for outcomes in other countries and have especial- Chapter 3 | Pakistan Human Capital Review young children’s health, well-being, and learning ly disadvantaged children in poor households outcomes and discusses how ECD outcomes and rural areas. Health and nutrition outcomes intersect with risk factors that pose threats to in early childhood, which constrain or support child development and human capital. Analy- school readiness and longer-term cognitive out- ses from a phone survey present a snapshot of comes and human capital development, are dire family resource reductions, parental distress, for many young Pakistani children. Only 66 per- and early childhood outcomes as experienced cent of 1-year-old children have received all during the COVID-19 pandemic. Cost of inac- basic vaccinations, 40 percent of children under tion analyses show that continuing the current 5 are stunted, and 17 percent under 5 are wast- low investment will be costly for Pakistan in the ed.4 Pakistan has the highest rate of stunting long run. The chapter summarizes how policy among similar low- and middle-income coun- environments and program implementation tries (see chapter 4). 66 Data on early learning and primary school read- FIGURE 3.1  Parents report that many children iness, while limited, are concerning. Without age 3 and 4 are not developmentally on track large-scale data collection systems that regu- Children developmentally on track (percent) larly measure early childhood development and 100 school readiness levels, there have been gaps in the knowledge base. Some datasets use grade 80 1 outcomes as a proxy for preprimary learning, Low- and middle-income country average: 75 while others rely on parent reporting, and data 60 on development for children under 3 have been 59.4 54.6 missing. Information on social-emotional de- 47.5 40 velopment is particularly limited for all ages, 39.1 clouding the picture relative to children’s ability to follow instructions, get along with peers, solve 20 problems, and demonstrate curiosity and cre- ativity, which facilitate future achievement and 0 human capital development.5 Punjab Sindh Khyber Balochistan (2017–18) (2018–19) Pakhtunkhwa (2019–20) (2019) To better understand how preschool-age chil- Source: Multiple Indicator Cluster Surveys. dren are developing beyond physical growth, Note: Based on parent reports of whether children age 3–4 are developmentally on track in at least three of the following four analyses were run using data from the Multi- domains: literacy-numeracy, physical, social-emotional, and ap- ple Indicator Cluster Survey. Parents reported proaches to learning. whether their child was developmentally on track in various domains (literacy-numeracy, physi- According to the Annual Status of Education cal, social-emotional, and approaches to learn- Report–Pakistan 2019, more than a quarter ing) judged by the presence of age-appropriate (26.8 percent) of children in grade 1 cannot read knowledge and skills. In the literacy domain, for in their local language, and more than a third example, to be considered developmentally on (33.7 percent) cannot read in English.7 From track, a child’s parent reports that the child has 2018 to 2019, the percentage of children with at least two of these three skills: she can identify/­ knowledge of basic arithmetic decreased from name at least 10 letters of the alphabet, can read 79 percent to 74 percent, although rates had im- at least four simple words, and knows the names proved from the 70 percent in 2015. and recognizes the symbols of numbers 1 to 10. Although not a nuanced measure of ECD, this re- Substandard opportunities and development in veals roughly where preschool-age Pakistani chil- the early years undermine children’s capacity to dren stand developmentally according to their take full advantage of later schooling because Chapter 3 | Pakistan Human Capital Review parents (figure 3.1). Results in each province are they lack the strong foundation needed for fur- below the 75 percent average across 60 low- and ther development and learning. Three-fourths middle-income countries6: 59.4 percent in Pun- (75 percent) of children cannot comprehend a jab (2017–18), 47.5 percent in Sindh (2018–19), simple paragraph by age 10. Regional inequities 54.6 percent in Khyber Pakhtunkhwa (2019), and show even higher rates based on where a child 39.1 percent in Balochistan (2019–20). Rates lives; learning poverty rates are approximately of being developmentally on track are lower for 50 percent in Punjab, Pakistan’s most populous children exposed to various risk factors. province, but they soar to 80 percent in Baloch- —­ istan.8 Learning poverty in Pakistan is high­ the Without adequate early childhood development average among low- and middle-income coun- and school readiness, low education achieve- tries is 53  —­ percent­ and forecast to become ment continues into early primary school. higher (79 percent) in the context of COVID-19 67 and the 2022 floods.9 Failure to increase in- detrimental lifelong and intergenerational vestments, service coverage, and programming impacts on their physical and mental health, quality to bolster ECD will perpetuate the condi- learning, behavior, productivity, and ultimate- tions that result in children in Pakistan achieving ly their country’s human development.11 ECD only an estimated 40 percent of their potential encompasses a young child’s cognitive, lan- in adulthood. guage, motor, and social-emotional develop- ment. Biological and psychosocial risks (such ECD is a key area for reform in the World Bank’s as maternal illness and malnutrition, maternal forthcoming Country Partnership Framework depression, violence, poverty, and inadequate for Pakistan. Improving policy environments care and early learning opportunities) expose and implementing effective programs at scale­ the developing infant to stressors. The onset, —­ both within sectors and as multisectoral severity, and chronicity of these stressors can —­ initiatives­ will help meet three key objectives influence early brain development, which is for ECD. First is ensuring that young children most rapid during this period. are nourished and kept physically healthy, par- ticularly in the first 1,000 days. Second is that The widespread risks to poor early development they receive appropriate stimulation from re- in Pakistan and the potential gains to invest- sponsive caregivers to support and enrich their ments in early childhood warrant more atten- cognitive and social-emotional development. tion to ECD in Pakistan’s human development And third is that young children need strong, agenda through multisectoral, equitable, and positive emotional bonds with parents and inclusive actions. Early interventions are vital to protection from stress to thrive. Meeting these mitigate risks and promote protective factors basic needs culminates in positive outcomes that shape healthy brain development, yielding for children and contributes to Pakistan’s gains for future health, learning, and produc- broader goals to improve education outcomes, tivity (figure 3.2).12 They also reduce inequities health metrics, inclusive economic growth, within and between populations.13 Evidence and governance systems­ —­which all depend on clearly shows that early childhood interventions human capital development. yield economic benefits for individuals14 and for countries,15 but child outcomes in Pakistan are less favorable than they could be in spite of ef- MULTIPLE RISK FACTORS fective interventions (box 3.1). IMPEDE EARLY CHILDHOOD DEVELOPMENT AND LEARNING Risk factors often overlap and threaten early childhood Early childhood risk factors and development and learning protective factors Chapter 3 | Pakistan Human Capital Review Children in Pakistan are exposed to an array ECD is critical to human capital development of risk factors, which include living in poverty, and is affected by both risk factors and pro- being stunted, living in a rural area, living with tective factors. Evidence from neuroscience a disability, having a mother with low education points to the need for nurturing environments attainment, receiving inadequate early stimu- and early intervention to promote childhood lation at home, harsh parenting that includes development and establish a foundation for psychological or physical aggression, and not future human development. An estimated participating in ECE. Each risk factor is associ- 250 million children in low- and middle-income ated independently with lower rates of being de- countries are not achieving their developmen- velopmentally on track, but multiple risks often tal potential in the first five years of life,10 with co-occur. 68 Lack of access to early childhood education FIGURE 3.2  Differing trajectories of human development as a Enrollment in ECE among children age 3–5 is function of exposure to protective and risk factors low by international standards and in relation to Optimum Recovery Below potential Optimum Pakistan’s own goals. As a signatory to the Sus- tainable Development Goals (SDGs), Pakistan Behavioral supports universal access to at least one year of competence Protective factors > preprimary education (SDG 4.2), but enrollment trajectories risk factors still falls far short of this goal. The national aver- Reduction in risk factors, increase in protective factors, or age enrollment rate of children age 3–5 in early intervention during a sensitive period childhood education is only 19  percent (figure 3.3). Enrollment is even lower for some groups Brain function of children, such as low-income children, rural children, children with a functional disability, Risk factors > protective factors and girls, as discussed below. Below potential Prenatal Birth Early Adolescence Adulthood Living in poverty childhood age Living in poverty predicts a lower likelihood Source: Reprinted from The Lancet, 378/9799, Walker, S.P., Wachs, T.D., Grantham-McGregor, of being developmentally on track. In Punjab, S., Black, M.M., Nelson, C.A., Huffman, S.L., Baker-Henningham, H., Chang, S.M., Hamadani, J.D., Lozoff, B. and Gardner, J.M.M., Inequality in early childhood: risk and protective factors only 46.9 percent of children in the lowest in- for early child development, pp.1325–1338, Copyright (2011), with permission from Elsevier. come bracket are reported to be developmen- DOI: 10.1016/S0140-6736(11)60555-2. tally on track, assuming no other risk factors. Note: The cumulative effect of exposure over time is represented by the darkening lines, This is better than the 26 percent in Sindh but which indicate the increasing strength of each trajectory. BOX 3.1 Acting early to promote protective factors for Pakistani children: What the evidence tells us Despite the challenges and gaps in service one study met the criteria for major postpartum provision, there is evidence that effective in- depression.1 Pilot interventions have shown that terventions can be successfully implemented low-cost mental health interventions­ —­such as in Pakistan. Several interventions and effective the Thinking Healthy Programme, delivered by programming have been shown to reduce the community health workers­ —­have multiple ben- Chapter 3 | Pakistan Human Capital Review risks that threaten healthy child development. efits, including reduced maternal depression, improved knowledge about play and develop- Maternal health and nutrition. Maternal under- ment for young children, and improved breast- nutrition (see chapter 4), which leads to infants feeding practices.2 who are small for their gestational age, remains a persistent challenge globally and in Pakistan, Young children’s health and nutrition. Early with risks for developmental delays and disabil- deprivation leading to stunting remains a per- ities. In addition to the physical well-being of sistent national challenge (see chapter 4) and mothers, poor maternal mental health is a rec- is associated with poor health, learning, and ognized risk for poor child development. In Pa- productivity in later life. Yet several studies of kistan, up to 53 percent of untreated women in (continued) 69 BOX 3.1, continued interventions with potential for scaling up in behavioral problems, increased aggression, Pakistan have found that community-based and poor self-regulation5 and early cognitive health interventions can promote newborn development delay,6 which may affect ed- survival and infant and young child health and ucation attainment. In addition to violence reduce stunting.3 against children, exposure to domestic vio- lence threatens children’s safety, develop- Early learning opportunities and responsive ment, and care as well as women’s safety care. High-quality, responsive interactions at and well-being.7 In Pakistan, up to 65 percent home and at school contribute to young chil- of women in rural areas report experiencing dren’s development and learning, whereas in- domestic violence.8 The INSPIRE report on adequate early learning and responsive care preventing violence provides examples of par- undermine young children’s development and enting programs that successfully integrate school readiness. Early learning, at home and key messages on gentle discipline and pre- in early childhood care and education centers, venting violence at home.9 In addition, training provides opportunities for young children to programs for preschool and primary school explore their world, build relationships with teachers that have promoted improved be- others, and acquire the skills and tools for life- havioral management with young children in long learning. Research in Pakistan has shown classrooms in low- and middle-income coun- that stimulation and responsive care in parent- tries have potential in Pakistan.10 ing programs delivered through community health services and family physicians benefit Notes children’s development and care.4 1. Rahman et al. 2008 2. Rahman et al. 2008; 2012. 3. Bhutta et al. 2011; Khan et al. 2020. Safe, violence-free environments. Violence 4. Khan et al. 2018; Yousafzai et al. 2014. against children, commonplace in homes in 5. Cuartas et al. 2021; Pace et al. 2019. 6. Cuartas et al. 2020. Pakistan, undermines young children’s devel- 7. WHO 2016. opment and results in difficulties throughout 8. Zakar, Zakar, and Abbas 2016. the child’s life, including social-emotional and 9. WHO 2016. 10. Baker-Henningham and Walker 2018. still puts about half of children living in pover- risks associated with living in poverty: reduced Chapter 3 | Pakistan Human Capital Review ty at risk. Moreover, living in poverty is associ- access to ECE further reduces the probability of ated with many other threats to development, developing the motor, cognitive, language, and including lack of access to ECE. There is a social-emotional skills necessary to achieve a 12.1  percentage point gap nationally in enroll- child’s potential. ment in ECE for children age 3–5 between the wealthiest quintile (27.2 percent) and the poor- Living in a rural setting est quintile (15.2 percent) (figure 3.4). Enroll- Living in a rural setting raises the probabili- ment gaps are 22.6 percentage points in Sindh, ty that a child will experience multiple risks 8.2  percentage points in Punjab, 6.3  percent- and will not develop to potential. In Punjab, age points in Balochistan, and 5.0 percentage 56.3 percent of rural young children were re- points in Khyber Pakhtunkhwa. This pattern ported to be on track, while in Sindh, only shows the compounding effect of the multiple 44.0 percent were, assuming no exposure to 70 other risk factors. Whereas just 6.5  percent FIGURE 3.3  Early childhood education enrollment of children of urban children experience four risk fac- age 3–5 falls far short of universal access, in Pakistan and by tors, 20.0 percent of rural children in Pakistan province, 2019–20 do.16 Rural children have less access to the in- Percent frastructure and service delivery they need Out of school Currently enrolled in preprimary Currently enrolled in primary to thrive. In addition to having less access to 100 3.9 4.4 5.1 1.6 2.7 13.3 5.2 healthcare, early stimulation at home, and 18.9 27.0 10.9 92.1 piped water, rural children also have reduced 80 84.0 85.2 access to ECE. More urban children age 3–5 77.1 than rural children were enrolled in ECE in 60 68.7 2019 in Punjab (33.1 percent versus 23.9 per- cent), Sindh (19.7 percent versus 4.1 percent), 40 Khyber Pakhtunkhwa (21.9  percent versus 11.8 percent), and Balochistan (7.7 percent ver- 20 sus 4.4 percent) (figure 3.5). Balochistan sees the lowest ECE enrollment rates among rural children, but Sindh has the greatest inequality. 0 Overall Punjab Sindh Khyber Balochistan Among rural children, poor children have the Pakhtunkhwa least access to ECE programs: while 21.0 per- cent of children in the wealthiest rural families Source: World Bank calculations using data from the Pakistan Social and Living Standards Measurement Survey 2019–20. are enrolled in ECE nationally, only 14.2 percent of poor rural children are enrolled. Having a disability FIGURE 3.4  Gaps in enrollment in early childhood education The prevalence of functional disabilities in Pa- between children age 3–5 in the lowest and highest income quintiles, kistan, which appears similar to that in compa- in Pakistan and by province rable countries, correlates with a reduced like- Bottom 20% Top 20% lihood of developing on track compared with Overall peers without a disability. Nearly 5  percent of 15.2 27.2 Pakistani children age 3–5 have a functional disability, according to parent reports, which Punjab 25.3 33.4 indicates atypical development related to sight, hearing, difficulties with movement, memory Sindh 5.1 27.6 problems, or verbal communication problems.17 Pakistan seems on par with estimates global- Khyber Pakhtunkhwa 12.4 17.4 Chapter 3 | Pakistan Human Capital Review ly and regionally.18 The prevalence of moderate and severe disabilities for children in low- and Balochistan 2.7 9.0 middle-income countries from birth to age 14 is estimated at 5.1 percent globally and 5.2 per- 0 10 20 30 40 50 cent in South-East Asia; the prevalence in high- Percent of early childhood education enrollment income countries is 2.8 percent.19 Source: World Bank calculations using data from the Pakistan Social and Living Standards Measurement Survey 2019–20. Young children reported to have a functional disability are considerably less likely to be devel- opmentally on track than the overall population of young children. In Pakistan, there is a large gap in enrollment in ECE between children with 71 FIGURE 3.5  Gaps in enrollment in early childhood education a known functional disability (10.8 percent) and between children age 3–5 in rural areas and those in urban areas, in those without (20.1 percent) (figure 3.6). There Pakistan and by province is also a large gap in development outcomes. Rural Urban For instance, Sindh reports that 49.4 percent of children without a disability are developmentally Overall 15.6 26.3 on track but just 26.0 percent of children with a disability are developmentally on track.20 Punjab 23.9 33.1 Lack of early stimulation Sindh 4.1 19.7 Whereas two-thirds (69  percent) of parents in low- and middle-income countries provide ade- Khyber 11.8 21.9 quate early stimulation to their young children, Pakhtunkhwa less than half of parents in every province pro- vide adequate early stimulation in Pakistan Balochistan 4.4 7.7 (figure 3.7).21 Early stimulation of cognitive 0 10 20 30 40 50 and social-emotional skills, which takes place Percent of early childhood education enrollment through reading books, telling stories, sing- ing songs, going outside together, playing, or Source: World Bank calculations using data from the Pakistan Social and Living Standards naming or counting, is low across the board: Measurement Survey 2019–20. 35.3  percent in Punjab, 45.0  percent in Sindh, 35.0  percent in Khyber Pakhtunkhwa, and 49.3 percent in Balochistan (figure 3.8). Access FIGURE 3.6  Gaps in enrollment in early childhood education are to learning materials at home is also low. Chil- large between children age 3–5 with a functional disability and those dren under 3, whose brain development is at without one, in Pakistan and by province its most rapid and sensitive, receive low rates Functional disability No functional disability of stimulation and access to learning resourc- es as well (figure 3.9). This seriously hinders Overall 10.8 20.1 human development given the importance of early stimulation for brain development in the Punjab 16.9 28.5 first few years of life. Lack of adequate social- emotional and cognitive support intersects with Sindh 5.8 11.6 lack of proper nutrition and other critical inputs to increase the likelihood of stunting and other Khyber 7.8 13.6 long-term risks to development. Pakhtunkhwa Chapter 3 | Pakistan Human Capital Review Harsh discipline Balochistan 4.0 5.4 Closely tied to lack of early stimulation are com- 0 10 20 30 40 50 mon practices of harsh discipline and a lack of Percent of early childhood education enrollment supervision, both deterrents to healthy growth and positive development. Harsh parenting Source: World Bank calculations using data from the Pakistan Social and Living Standards measures, including psychological aggression Measurement Survey 2019–20. such as yelling and name calling and physical violence such as hitting and pinching, are prev- alent across Pakistan. Between 43 percent and 73 percent of children age 0–2 experience harsh discipline, based on parent reports for the prior 72 month (figure 3.10). Among children age 3–5, FIGURE 3.7  Rates of early stimulation by caregivers in provinces the rates are even higher: 84.9 percent in Pun- in Pakistan are low, children age 2–5 jab, 81.5 percent in Sindh, 87.1 percent in Khyber Percent of children 24–59 months with whom an adult has engaged in four or more stimulating activities in the past 3 days Pakhtunkhwa, and 57.0 percent in Balochistan. 100 About 10  percent of children under 3 are left without supervision in Sindh and Khyber Pakh- 80 Low- and middle-income country average: 69 tunkhwa, 23 percent in Balochistan, and about 6 percent in Punjab, as measured by parent re- 60 ports of an hour or more of unsupervised time during the previous week. The percentage of 40 45.3 48.2 children age 3–5 left without adult supervi- 33.4 sion is 7.6  percent in Punjab, 12.5  percent in 27.9 20 Sindh, 6.1 percent in Khyber Pakhtunkhwa, and 27.7 percent in Balochistan (figure 3.11). Lack of 0 supervision reduces children’s well-being and Punjab Sindh Khyber Pakhtunkhwa Balochistan safety in the absence of responsive caregiving (2017–18) (2018–19) (2019) (2019–20) for children’s cues, such as crying, accidents, lack of protection from harm and injury, oppor- Source: Statistics as reported in Multiple Indicator Cluster Survey reports. Note: Early stimulation is based on parent report of child’s experience of four or more stim- tunities to ingest soiled materials, and increas- ulating activities from the following list: reading books, telling stories, singing songs, taking ing diarrhea and disease. the child outside, playing with the child or naming/counting ) with adult caregivers at home in the past three days. Data are from different points in time and as such are not strictly comparable. Gender disparities Gender is not included as a risk factor in the cur- rent analyses because it does not emerge as a key determinant of exposure to risk factors or FIGURE 3.8  Rates of early stimulation by caregivers and access developmental outcomes prior to school entry. to playthings in the household are low across provinces in Pakistan, Although older girls and women face a large children age 3–5 number of barriers to developing and deploy- Has engaged with an adult in 4 or more stimulating activities in the past 3 days ing their human capital in Pakistan, in the early Plays with 2 or more types of playthings Has 1 or more children’s books years, girls and boys are relatively similar in their Punjab experiences of risks and developmental out- (2017–18) comes (figure 3.12).22 Sindh (2018–19) However, gender disparities start to emerge as Chapter 3 | Pakistan Human Capital Review early as preprimary enrollment. While these dif- Khyber ferences are small in comparison to the gaps Pakhtunkhwa (2019) by income and other factors, it is important to note that gender gaps in education attainment Balochistan are evident even before children start primary (2019–20) school. These early hurdles in access to human capital development lead to a persistent lack of 0 20 40 60 80 100 parity across the life cycle for girls and women Percent of children age 36–59 months in Pakistan. Punjab has the smallest gap in en- Source: World Bank calculations using data from the Multiple Indicator Cluster Survey. Note: Data are from different points in time and as such are not strictly comparable. rollment between boys (27.4 percent) and girls (26.6 percent), at less than 1 percentage point 73 FIGURE 3.9  Rates of early stimulation by caregivers and access (figure 3.13). The gap is 1.7 percentage points in to resources in the household are low across provinces in Pakistan, Sindh (11.7 percent for boys versus 10.0 percent children age 0–2 for girls), 2.3 percentage points in Khyber Pakh- Has engaged with an adult in 4 or more stimulating activities in the past 3 days a tunkhwa (14.4 percent versus 12.1 percent), and Plays with 2 or more types of playthings Has 1 or more children’s books 1.9 percentage points in Balochistan (6.0  per- Punjab cent versus 4.1 percent)­ —­ with low overall rates (2017–18) the pressing concern for both genders. Sindh Accumulating risks (2018–19) While any one risk factor is independently as- sociated with a reduced probability of positive Khyber Pakhtunkhwa outcomes, an accumulation of risks further (2019) decreases a child’s chances of growing and learning to potential. Exposure to multiple risks Balochistan (2019–20) reduces children’s chances of staying on track, and the more risks, the lower the chances of op- 0 20 40 60 80 100 timal development. Risk factors tend to cluster, Percent of children age 0–35 months so that children experience them simultane- Source: World Bank calculations using data from the Multiple Indicator Cluster Survey. ously, sequentially, or both. For instance, there  a. Data are for children age 24–35 months. is a considerable overlap among children who Note: Data are from different points in time and as such are not strictly comparable. lack early stimulation, those who lack access to ECE, and those whose mothers have low edu- cation attainment. These children are at risk of FIGURE 3.10  Use of harsh discipline is common in Pakistan not experiencing the responsive and engaging among children age 0–2 and 3–5 interactions they need to thrive, both from par- Age 12–35 months Age 36–59 months ents or guardians at home and from teachers in preschool programs. Gaps between groups of Punjab (2017–18) children start early and widen over time. In Punjab and Sindh, 94  percent of children Sindh (2018–19) experienced at least one of the top three risk factors in that province, while rates of multiple Khyber exposure were also high. Among the risk fac- Pakhtunkhwa (2019) tors examined, the three most prevalent among 3- and 4-year-olds in Punjab were failing to ex- Chapter 3 | Pakistan Human Capital Review Balochistan perience early stimulation at home (72 percent), (2019–20) living in a rural area (67 percent), and not being 0 20 40 60 80 100 enrolled in ECE (66 percent). Approximately half Experienced psychological aggression or physical punishment of these children experienced at least two of the during the past month (percent of age group) risk factors (45 percent lived in a rural area and Source: World Bank calculations using data from the Multiple Indicator Cluster Survey. had no ECE, 49  percent experienced no early Note: Harsh discipline includes psychological aggression (name calling, yelling) and physical stimulation at home and were not enrolled in punishment (hitting, pinching). Parents were asked about practices in the previous month. Data are from different points in time and as such are not strictly comparable. early childhood education, and 53 percent lived in a rural area and experienced no early stimu- lation at home). More than a third (37 percent) of children in Punjab experienced all three risks 74 (figure 3.14). In Sindh, the three most preva- FIGURE 3.11  Even very young children often lack supervision lent risk factors were not being enrolled in early Lack of child supervision in the past week in provinces in Pakistan, children age 0–2 and 3–5 childhood education (86  percent), low mater- Age 0–35 months Age 36–59 months nal education (74 percent), and a lack of early Punjab stimulation (49 percent). The rates of multiple (2017–18) exposure were even higher than in Punjab, with 42 percent of 3- and 4-year-olds experiencing all Sindh three risk factors. Stepwise regressions showed (2018–19) that a lack of early childhood education was the greatest single predictor of being found devel- Khyber Pakhtunkhwa opmentally off track among children age 3 and 4 (2019) in both Punjab and Sindh. Balochistan (2019–20) As exposure to risk factors increases, children’s well-being and performance decline. For ex- 0 5 10 15 20 25 30 ample, in Sindh, 83.0 percent of children age 3 Left alone or under the supervision of another child younger than age 10 for and 4 who are exposed to zero or one risk fac- more than one hour at least once in the past week (percent of age group) tor are likely to be reported by their parents as Source: World Bank calculations using data from the Multiple Indicator Cluster Survey. Note: Data are from different points in time and as such are not strictly comparable. being developmentally on track compared with 69.8 percent of children exposed to two risk fac- tors (living in poverty, being stunted, living in a rural area, having a disability, having a mother FIGURE 3.12  At age 3–5, girls and boys are equally likely to with less than a primary education, not receiv- experience multiple risks to on-track developmental outcomes, ing early stimulation, receiving harsh punish- Punjab province ment, and not being enrolled in ECE). However, Percent of children as the number of risks mounts, children’s likeli- Male Female 25 hood of staying on track drops commensurate- ly: the likelihood of being developmentally on track drops to 57.8 for three risks, to 44.7 per- 20 cent for four risks, to 43.2 percent for five risks, to 37.4 percent for six risks, and to 26.5 percent 15 for seven or more risks. 10 This association is disquieting because many young children experience multiple risks: in 5 Chapter 3 | Pakistan Human Capital Review Sindh, 1 child in 5 experiences three risk factors, almost 1 in 4 experiences four risk factors, and 0 about 1 in 10 experiences six or more risk fac- 0 1 2 3 4 5 6 7+ tors. The trend is similar for the other provinces Number of risk factors (figure 3.15). Source: World Bank calculations using data from the Multiple Indicator Cluster Survey in Punjab. Note: Risk factors include being stunted, having a mother with less than primary education, Helping households and communities facilitate not being enrolled in early childhood education, being in the poorest wealth quintile, living nurturing care is critical for healthy ECD, and it in a rural area, having a functional disability, and not experiencing adequate stimulation at is equally imperative to provide access to inter- home (engaged in less than four stimulating activities with any household caregiver over ventions and protective factors to enable young three days). children to develop to their potential. While it is 75 FIGURE 3.13  Gender gaps in education attainment disadvantaging ECE­ —a­ re needed as buffers against the multiple girls begin to emerge in early childhood education enrollment in risks that children face. For example, children in Pakistan and across provinces, children age 3–5 Pakistan who enroll in ECE, even when quality is Female Male not known, display higher performance in being developmentally on track at age 3 and 4 than Overall children who never enroll (figure 3.16). In Pun- jab, 74 percent of children age 3 and 4 who at- Punjab tended an ECE program were rated as develop- mentally on track compared with 52 percent of Sindh children not enrolled. This pattern holds across the other provinces. Khyber Pakhtunkhwa The benefits of investing in ECD far outweigh the costs. As important as investing in ECE is invest- Balochistan ing in widespread high-quality services in health 0 10 20 30 40 50 and nutrition, water and sanitation, and social Percent of early childhood education enrollment protection. Capacity building of parents and other caregivers is a powerful tool for improving Source: World Bank calculations using data from the Pakistan Social and Living Standards child outcomes and long-term human capital Measurement Survey 2019–20. accumulation. This requires multisectoral coor- dination for optimal service delivery, as well as critical to increase young children’s experienc- the development of beneficial knowledge and es of nurturing care and mitigate exposure to skills about feeding, early stimulation, immuni- risks, broader interventions at scale­—s­ uch as zation and healthcare, sanitation practices, su- parenting programs and access to high-quality pervision, and positive discipline. FIGURE 3.14  Rates of multiple exposure to the most prevalent risk factors are high among children age 3 and 4 in Punjab and Sindh Punjab Sindh Not enrolled in Not enrolled in early childhood education early childhood education 66% 86% 9% 14% Chapter 3 | Pakistan Human Capital Review 12% 8% 2% 28% 37% 42% 6% 5% 3% 2% 16% 2% No stimulation Living in a No stimulation Low maternal at home rural area at home education 72% 67% 49% 74% Source: World Bank calculations using data from the Multiple Indicator Cluster Survey 2017–18 for Punjab and 2018–19 for Sindh. 76 FIGURE 3.15  The likelihood of being developmentally on track declines with the number of risk factors, children age 3 and 4 Percent of children Percent of children developmentally on track Percent of children Percent of children developmentally on track a. Punjab (2017–18) b. Sindh (2018–19) 50 100 50 100 Percent of children developmentally on track 40 80 40 80 Percent of children developmentally on track 30 60 30 60 20 40 20 40 10 20 10 20 0 0 0 0 0 1 2 3 4 5 6 7+ 0–1 2 3 4 5 6 7+ Number of risk factors Number of risk factors c. Khyber Pakhtunkhwa (2019) d. Balochistan (2019–20) 50 100 50 100 Percent of children developmentally on track 40 80 40 80 Percent of children developmentally on track 30 60 30 60 20 40 20 40 10 20 10 20 0 0 0 0 0–1 2 3 4 5 6 7+ 0–1 2 3 4 5 6 7+ Number of risk factors Number of risk factors Source: World Bank calculations using data from the Multiple Indicator Cluster Survey. Note: Risk factors include being in the lowest income quintile, stunted, living in a rural area, having a functional disability, mother with less than a primary school education, low early stimulation at home (less than four activities in the previous 24 hours), exposed to harsh parenting, and Chapter 3 | Pakistan Human Capital Review not enrolled in early childhood education. HOUSEHOLD-LEVEL STRESS AND increased levels of poverty and food insecurity, EARLY CHILDHOOD DEVELOPMENT disruptions in education, decreased healthcare IN THE ERA OF COVID-19 and social support systems, elevated stress levels, and decreased well-being for parents.24 The COVID-19 pandemic has disrupted early Results from a nationally representative phone childhood services and increased exposure survey of 3,021 households in 2021–22 across to risks for young children around the world. the country shed light on the stressors experi- Studies have found increases in mental health enced by Pakistani parents during the pandemic concerns among caregivers and young chil- and the current status of early childhood devel- dren and increases in violence against children,23 opment. (See Annex 3 for survey details.) 77 FIGURE 3.16  Across Pakistan, children age 3 and 4 who enroll provinces (0.06, 0.14, and 0.26 standard devi- in early childhood education are much more likely to be rated ation above, respectively) and below the mean developmentally on track for the latter (0.26 standard deviation below). In Percent of children age 3 and 4 rated as developmentally on track short, children age 0–2 in Khyber Pakhtunkhwa Does not attend early childhood education program and in rural areas nationally were particularly vul- Attends early childhood education program nerable during the COVID-19 pandemic. 100 Levels of early stimulation at home were strongly 80 79 associated with levels of development for chil- 74 74 dren from birth through age 2. Parents’ engage- 60 ment in early stimulation activities was a positive 56 52 53 53 and significant predictor of infants’ and young 40 children’s development. Early stimulation was 42 evaluated using children’s exposure to child-ori- 20 ented learning resources, such as play objects and picture books, as well as responsive and 0 engaging interactions between caregivers and Punjab Sindh Khyber Pakhtunkhwa Balochistan adults, such as telling stories, singing lullabies, (2017–18) (2018–19) (2019) (2019–20) clapping, talking to the child, and visiting friends. Source: World Bank calculations using data from the Multiple Indicator Cluster Survey. The more stimulating resources and interactions Note: Data are from different points in time and as such are not strictly comparable. parents provided, the better the child’s out- comes. Even a single stimulating interaction in Child outcomes are affected by the prior three days doubled z-scores relative to location, parent capacity, and none (figure 3.17). The value of early stimulation preprimary enrollment for children age 0–2 was evident for children in both urban and rural settings, regardless of the Children age 0–3 number of people in the household, and among Other than nutrition data, which are often used as children with less educated mothers. a proxy, no data were available to measure ECD as an independent construct for children age 0–2 Survey results also showed that caregiver ca- in Pakistan, so the phone survey provided an op- pacity correlates with child development. Moth- portunity to measure child development for this ers’ education attainment predicted develop- age group, using the Caregiver-­ Reported Early mental outcome scores among children age Development Instrument short form. Results 0–2, such that children with mothers who had showed that location matters for early childhood either not attended or not completed primary Chapter 3 | Pakistan Human Capital Review development in Pakistan, whether by urban– school exhibited significantly lower levels of de- rural divide or by province. Significant differences velopment (0.08 standard deviation below the were found as expected across urban–rural geo- mean) compared with those whose mothers graphical divides, with urban children age 0–3 had completed primary school (0.21 standard showing slightly higher developmental outcomes deviation above the mean) (p < .001). Differenc- on average (0.16 standard deviation above the es in children’s developmental outcome scores mean) than their rural peers (0.05 standard de- at age 0–2 were not predicted by paternal ed- viation below the mean). Irrespective of urban– ucation attainment, although those whose fa- rural differences, children in Punjab, Sindh, and thers had a primary school education or less Balochistan exhibited higher developmental exhibited lower levels of child development than outcomes than children in Khyber Pakhtunkhwa. their peers whose fathers had a secondary or Scores were above the mean for the former three higher level of education. 78 Children age 3–5 children should enroll in school, on average, Access to opportunities to learn at home and in parents reported that by around age 4 or 5 preprimary school are key predictors of school (mean of 4.5), children should be enrolled in readiness in Pakistan. Items from the Measuring school. Several factors predicted earlier age Early Learning Quality and Outcomes–Measure of entry beliefs: younger child age, living in the of Development and Early Learning Teacher/ Punjab province, a mother with a secondary Caregiver report were adapted for the phone education, early stimulation at home, house- survey to evaluate school readiness among chil- hold size, and current enrollment status. For dren age 36–72 months, and three subdomains example, a parent of a 3-year-old in Punjab were evaluated, language knowledge, math who is currently enrolled in school, receives knowledge, and social-emotional skills. Summa- maximum stimulation at home and has an ry scores, not age standardized, were created for educated mother is predicted to believe their 25 items. Beyond child age, strong predictors of child should be enrolled in school at 3.7 years. school readiness­ —­that is, better literacy, numer- By contrast, a parent of a 5-year-old in Khyber acy, social-emotional, and approaches to learn- Pakhtunkhwa who is not currently enrolled —­ ing skills­ included early stimulation at home in school, receives no stimulation at home, and enrollment in ECE (figure 3.18). Moreover, and has a mother with less than a primary having a mother with secondary education or school education is predicted to believe their higher was associated with a 1.7 point increase in child should enroll at 5.6 years. This pattern school readiness scores compared with having a is meaningful because the age the parent be- mother with less than primary school education. lieves their child should begin school is close- ly correlated with actual age of enrollment. Caregiver beliefs were a Among children of parents who believe chil- key determinant of a child’s dren should enroll at age 3, 72.4 percent were enrollment age currently enrolled in school; among children of parents who believe children should enroll at Beliefs about school entry predicted actual age 6, 23.0 percent were currently enrolled in enrollment patterns. When asked at what age school. FIGURE 3.17  Children under 3 in high-stimulation households have higher levels of child development than children in low-stimulation households in Pakistan Percent of households sampled Child development (z-score) 10 1.0 Child development z-score 8 0.5 Chapter 3 | Pakistan Human Capital Review 6 0.0 4 –0.5 2 –1.0 0 –1.5 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Number of stimulating activities Source: World Bank calculations using Gallup phone survey data 2021–22. 79 FIGURE 3.18  School readiness scores increase with access current ability to be affectionate to their child, to early stimulation, maternal education and enrollment in early get enough sleep, control their worry, feelings of childhood education among children age 5 irritability or anger, and feelings of nervousness or anxiety. There were significant positive cor- No early stimulation, less than relations between income loss and self-report- primary school educated 10.9 mother, does not attend ECE ed distress among parents surveyed for every dimension of well-being assessed (figure 3.19). No early stimulation, secondary school educated 11.6 The rate of irritability or anger was twice as high mother, does not attend ECE for those who had lost income as for those who High early stimulation, had not. About 19.2 percent of those who had not secondary school educated mother, not enrolled in ECE 18.3 lost income during the COVID-19 pandemic re- ported feeling more irritated or angry than usual High early stimulation, in the past 15 days compared with 37.9 percent secondary school educated 23.2 mother, enrolled in ECE of those who had lost income. Similarly, 19.3 per- cent of those who had not lost income report- 0 5 10 15 20 25 ed feeling nervous or anxious, compared with School readiness sum score 37.3 percent of those who had lost income. Men Source: World Bank calculations using phone survey data and Early Learning Partnership reported slightly lower capacity than women to caregiver response survey items for school readiness measured by language, math, and be affectionate to children and to control their social-emotional scores (range = 0–25). worry, and women reported slightly higher inci- Note: ECE is early childhood education. Analyses included only urban children and controlled for type of toilet and household size. dence than men of feeling irritated or angry, dif- ficulty getting enough sleep at night, and being Reduced household resources, nervous or anxious; but the differences between parenting distress, and lower child men and women were not significant. outcomes Parental distress independently predicts child Income loss during the COVID-19 pandemic af- development from age 0–2. Regression analyses fected almost two-thirds of households with that controlled for relevant confounding vari- young children. The majority of households in —­ ables at the p < .05 level­ province, income loss, the survey sample lost income (63.1  percent). monthly mobile phone expenditure, and mater- Based on parent perception of degree of loss, nal education levels­ —­showed that levels of pa- some reported experiencing a small (13.3 per- rental distress are strongly negatively correlated cent) or medium (25.1 percent) loss, but most with child development outcomes at age 0–2. felt they had experienced a large loss (61.6 per- Maximum parental distress levels, where a parent cent). Losses occurred across all areas and ge- self-reported the incidence of all stressors out of Chapter 3 | Pakistan Human Capital Review ographies and regardless of parental education a maximum of five, predict an early childhood de- levels. Almost half of respondents (48.8 percent) velopment z-score 0.4 standard deviation lower said they had run out of food more frequently than a household with no parental distress. In than before the pandemic, and 59.4 percent had contrast, parental distress was not a significant skipped or cut down on the size of meals. predictor of school readiness levels at age 3–5. Income loss was associated with greater paren- tal distress, and higher parental distress was HIGH COST OF INACTION: EARLY associated with poorer child development levels CHILDHOOD EDUCATION from age 0–2. A five-item measure of parental distress/well-being provided insight into parents’ There are multiple entry points for enhancing perceptions of their well-being based on their ECD outcomes, so interventions should ideally 80 be multisectoral, but addressing even one major FIGURE 3.19  Parents who had lost income since 2020 report area of intervention can generate large returns. higher distress than parents who had not lost income Consider ECE as an example. Resources cur- Percent rently allocated by the education sector for Did not lose income since March 2020 Lost income since March 2020 40 ECE are inadequate, despite overwhelming ev- 37.9 idence showing high returns on investment in 37.3 37.3 36.5 ECE. In 2020, national investment in education 30 32.5 30.2 represented just 2.5 percent of GDP, down from 28.6 27.2 2.9 percent in 2017, and considerably below the global average of 4.5  percent. While the exact 20 amount allocated to ECE is unclear due to the 19.2 19.3 combining of preprimary and primary educa- tion in budget reports, Investment in ECE is es- 10 timated to be very low. In addition to overall low spending, provincial disparities disadvantage 0 children in some provinces more than others, Found it di cult Feeling more Found it di cult Been nervous Not been able limiting opportunities for children in Baloch- to be a ectionate irritated to get enough or anxious to stop or control to child or angry sleep at night worrying istan in particular. There was an almost 30 per- centage point difference across provinces in the Source: World Bank calculations using phone survey data 2021–22. Note: Parental distress was based on self-report of feelings during the previous 15 days. share of recurrent education spending allocat- ed to primary schooling (including preprimary) in 2020, from 35.2  percent in Balochistan to the impact of increased preschool participation 40.8  percent in Sindh, 46.4  percent in Khyber on school attainment and subsequent chang- Pakhtunkhwa, and 64 percent in Punjab . es in wages using data for 73 low- and middle- income countries.25 The cost that society pays Raising enrollment in ECE has immediate costs for failing to implement such an intervention was and long-term benefits. The cost of investment estimated for three scenarios in Pakistan: in ECE is immediate. Boosting enrollment has a • Low coverage: Equalizing enrollment rates wide range of direct and indirect benefits that between rich (highest income quintile) and are more difficult to quantify but will be realized poor (bottom four quintiles) households to throughout a person’s life through improve- achieve 27.2  percent enrollment across in- ments in child development, school readiness come levels (based on the Pakistan Social and performance, and ultimately wages. Cost-­ and Living Standards Measurement Survey effectiveness analyses that quantify the impact 2019–20 enrollment rates for children in the of such programs on long-term wages can be highest income quintile). Chapter 3 | Pakistan Human Capital Review used to estimate benefits. • Medium coverage: Achieving the enrollment rates of a regional peer (Nepal, at 61.9 per- Conservative estimates using a benefit–cost cent enrollment). ratio of 6.4:1 reveal a high cost to inaction. Be- • Full coverage: Achieving one year of universal cause enrollment in ECE is a key predictor of ECE enrollment, meeting Pakistan’s commit- being developmentally on track, failing to im- ment to SDG 4.2 and 2009 National Educa- prove enrollment rates is economically expen- tion Policy objectives. sive, as well as harmful at the individual level. The cost of inaction (forgone benefits) for Pakistan Continuing coverage at present levels, and not was calculated using the most conservative pa- raising coverage levels, would be expensive. The rameters (a benefit–cost ratio of 6.4:1 and a dis- cost of inaction would be 0.34 percent of GDP count rate of 6 percent) from a study modeling for the low-coverage scenario (failing to boost 81 enrollment from 18.9  percent to 27.2  percent cost the country 5.65 percent of GDP for each of all 3- to 5-year-olds in preprimary school), cohort. 1.79  percent of GDP for the medium-coverage scenario (failing to enroll 61.9 percent of 3- to 5-year-olds), and 3.39  percent of GDP for the THE EXISTING POLICY AND full-coverage scenario (figure 3.20). The cost of PROGRAMMING ENVIRONMENT inaction on medium coverage is conservatively IS UNDERMINED BY estimated at US$4.7 billion for each cohort that IMPLEMENTATION SHORTFALLS fails to receive medium coverage. Pakistan’s policy environment for supporting A less conservative approach using a benefit–­ young children to develop their potential often cost ratio of 10:1 shows even higher costs of falls short when it comes to implementation. inaction (maintaining the status quo). With a Pakistan’s policies point in the right direction benefit–cost ratio of 10:1, failing to raise ECE in multiple sectors that deeply impact child enrollment rates from the current 18.9 percent development in the early years. Health and nu- to 27.2 percent across income quintiles (low trition policies have a long and broad history, coverage) would cost Pakistan 0.57  percent of and policy improvements are apparent in the GDP (see figure 3.20) annually. Failing to achieve water, sanitation, and hygiene sector in the medium-level coverage of 61.9  percent would past decade, with substantial infrastructure in- cost the country 2.98 percent of GDP for each vestments. The education sector has the am- cohort and failing to achieve full coverage would bitious goal of one year of universal preprimary education. Social protection policies include parental leave for federal employees, for exam- FIGURE 3.20  Not raising early childhood education enrollment ple. But good intentions and knowledge of best rates is costly in the long run practices are insufficient to reach young chil- Percent of GDP dren, especially vulnerable populations, with 6 the services they need for a strong and equal Cost of failing to close gap at low coverage rate start in life if implementation remains flawed 5 (for example, low-quality toilets increase soil and water contamination), insufficient (a vast 4 majority of workers have no paid leave), or of Cost of failing to close gap at medium coverage rate unknown quality (the quality of ECE services is 3 not evaluated). Even in areas with deep policy roots, such as health and nutrition, child out- 2 comes fall far short (low immunization rates, Chapter 3 | Pakistan Human Capital Review 1 high stunting rates). Cost of failing to close gap at full coverage rate 0 In addition, service delivery for young children BC ratio = 3.0 BC ratio = 6.4 BC ratio = 8.0 BC ratio = 10.0 is fragmented. Pakistan’s policies and pro- grams recognize that early childhood spans Source: World Bank estimates using the benefit–cost ratio from Engle et al. (2011), data on population age 3–5 from the 2020 Pakistan Census, government expenditure per primary the ages from birth to age 8, but service de- student from World Bank (2015), and GDP from World Bank (2020). livery is often split between infants and tod- Note: BC is benefit–cost.The green line depicts the cost of failing to equalize preschool en- dlers (age 0–2 years, including prenatal care) rollment rates between rich (top quintile) and poor (bottom four quintiles) households (27.2 percent enrollment from the Pakistan Social and Living Standards Measurement Sur- and preschool-age children (age 3–5 years). vey 2019–20). The blue line depicts the cost of failing to achieve moderate coverage of pre- Whereas children in the younger age group school enrollment using Nepal’s 61.9 percent enrollment rate as a comparison. The red line are the primary focus of health and nutrition depicts the cost of failing to achieve universal coverage of preschool enrollment. policies and programs, the older age group 82 receives more attention in education. Sec- Realizing improvements in children’s health to toral approaches to ECD remain siloed and boost human capital in Pakistan in the next 30 inadequate. years will require substantial increases in pub- lic financing, massive information campaigns, Globally, ECD experts are clear about what and new platforms to gather and integrate services and experiences young children need feedback.30 to thrive, across the four core sectors of health and nutrition, water and sanitation, education, Broad-based packages of services for pregnant and social protection.26 The global Nurturing women and children in the first 1,000 days of life Care Framework (NCF) calls for every child to are essential. Comprehensive services aimed at experience good health and adequate nutrition, mothers and young children in the first 1,000 security and safety, early learning opportunities, days are needed, such as a family support pack- and responsive caregiving.27 International ex- age for the early years with parental support, perts designate these interventions not as op- including family planning and parenting educa- tional services, but as necessary foundations for tion; a pregnancy package with pre- and ante- the development of children’s physical, social- natal care and information on nutrition; a birth emotional, and cognitive development and long- package, including delivery attended by skilled term human capital accumulation. providers, birth registration, and information on the benefits of exclusive breastfeeding; and a Health and nutrition children’s health and development package, in- cluding immunizations, information on deworm- Policies and programs at the national level ing, and identification and treatment of acute promote an integrated approach to health and malnutrition. Once children reach preschool nutrition in the first three years of life, but Pa- age, an ECE package featuring high-quality pre- kistan has a long way to go for providing ade- primary programs that foster sharing of infor- quate services for child health, a prerequisite mation on health and hygiene, nutrition, and diet to improved ECD and future achievement and at the school level would also improve health socioeconomic inclusion. Policies have gener- outcomes. ally aimed to tackle stunting and malnutrition, with stunting reduction goals being laid out in Water, sanitation, and hygiene the National Health Action Plan 2019–2023 and prioritized in the provincial agendas, but Pakistan has greatly improved and expand- progress has been slow relative to peers, and ed water, sanitation, and hygiene infrastruc- stunting rates (40.2 percent) remain the high- ture and services and made progress toward est in the region. Investments and service de- achievement of the SDGs, but improvements Chapter 3 | Pakistan Human Capital Review livery for mothers and young children are inad- have not translated into adequate ECD prog- equate, and uptake of recommended practices ress. Between 2015 and 2020, Pakistan enacted is low, for example, only half of women have the policies and delivered services that improved recommended number of antenatal care vis- conditions across provinces. It increased ac- its, and only one-third take iron and folic acid cess to basic hygiene by 16 percentage points supplements during pregnancy.28 Despite high (to 80  percent) and basic sanitation by 9 per- food insecurity, only 48.0  percent of children centage points (to 68 percent).31 Over the past under 6 months are exclusively breastfed, and two decades, the national rate of open defeca- 21.0 percent of children age 6–23 months ex- tion (13 percent) has been more than halved.32 perience minimum dietary diversity.29 Poor For some SDG targets, such as universal basic health outcomes have been exacerbated by hygiene, Pakistan is one of the few countries the COVID-19 pandemic and the 2022 floods. on track to achieve universal coverage by 2030 83 FIGURE 3.21  Progress toward universal basic hygiene from 2015– children’s health: Pakistan’s national stunting 20, among countries with less than 99 percent coverage in 2020 level (40.2 percent) remains alarming. Systemic Rate of change, 2015–20 (percentage points per year) factors include insufficient fecal waste manage- 4 ment and wastewater systems and the expan- Bangladesh Pakistan sion of low-quality toilets that have raised levels Progress is too slow 6 countries of soil and water, contributing to weak nutrition- in 23 countries are on track al outcomes for children.33 An estimated 70 per- 2 Indonesia cent of households drink bacterially contami- Guinea-Bissau nated water,34 posing a serious hazard for young children. Consumption of E. coli-contaminated water can lead to irreversible intestinal damage 0 and hinder children’s long-term growth. Coverage is decreasing in 4 countries ECD outcomes remain stagnant because of Sudan –2 undefined institutional responsibilities, un- 0 20 40 60 80 100 even coverage, and varied household practices. Basic hygiene services, 2020 (percent) Planning frameworks are weak at the provincial Source: WHO and UNICEF 2021. level, and local governments lack the technical capacity and tools to target resource alloca- (figure 3.21). Accelerated progress will be need- tion. Despite the Local Government Act of 2015, ed to achieve other SDG targets, such as provid- considerable overlap and confusion remain re- ing universal access to safely managed drinking garding the roles and responsibilities at each water and ending open defecation by 2030. tier of government, resulting in weak planning, coordination, and monitoring.35 Inequalities in Even so, positive developments in water and water, sanitation and hygiene services by wealth sanitation infrastructure and services have and location are stark.36 For example, there are not translated into adequate improvements in wide urban–rural gaps in access, with piped water and flush-to-sewer latrines mainly being FIGURE 3.22  Inequalities by wealth quintile in water, sanitation, concentrated in urban areas.37 Similar inequal- and hygiene coverage in Pakistan ities persist according to wealth (figure 3.22), Top 20% Bottom 20% although even the richest households living in poorer regions remain unprotected against Basic 78 the higher incidence of diarrhea and stunting.38 hygiene 31 Household health behaviors and hygiene prac- tices are important determinants of levels of Chapter 3 | Pakistan Human Capital Review contamination in the environment and unhy- 95 Basic sanitation gienic practices remain prevalent, including un- 47 safe stool disposal, bare feet in households, and having animals in dwellings.39 Basic 97 drinking water 62 Education 0 20 40 60 80 100 Pakistan has articulated policy goals for young Population with access (percent) children’s education, but failure to monitor in- vestments, child outcomes, and system progress Source: WHO and UNICEF 2021. obscures evaluation. The National Plan of Action Note: Comparing highest wealth quintile and lowest quintile groups in 2018. 2001–15 set 50  percent ECE enrollment as a 84 target, and the 2009 National Education Policy FIGURE 3.24  After rising between 2004 and 2018, preprimary declared universal access to one year of ECE a enrollment began to decline in Pakistan among children age 3–8 target in its 10-year plan. The country has further Percent articulated policy and standards through several 50 acts in the past decade, including by establishing 5-year-olds Pakistan’s National Curriculum starting at age 4. 40 However, financing data combine public prima- ry and preprimary education spending, so how 6-year-olds 30 much Pakistan and the provinces invest in ECE cannot be precisely determined. Program quality 4-year-olds and ECE participation patterns, including intensi- 20 ty, duration, and sequencing, are also unknown. 7-year-olds Terminology for preprimary is ambiguous, which 10 3-year-olds challenges monitoring and coordination efforts. 8-year-olds And, there is almost no information on the qual- 0 ity of ECE programs in either public or private 2004 2006 2008 2010 2012 2014 2018 2020 settings. In Sindh, more than half (55.6 percent) Source: World Bank calculations using Pakistan Social and Living Standards Measurement of preprimary provision occurs through private Surveys 2004–20. programs, which are not necessarily bound by public regulations (figure 3.23). yet fewer than half of 5-year-olds in Paki- stan are enrolled in school (figure 3.25). Only Fewer than one-fifth of Pakistan’s preschool-age 42.3  percent of children age 5 are enrolled children enroll in ECE, and rates are falling. (9.1 percent in primary school and 33.1  per- Despite its goals, Pakistan’s coverage  of ECE cent in preprimary), leaving 57.7  percent out remains extremely low­ —o­nly 18.9  percent of of school altogether­ ­nd far fewer 3- and —a preschool-age children were enrolled in a pre- 4-year-olds were enrolled in preprimary school primary program in Pakistan in 2019–20, as (3.2  percent and 17.2  percent respectively). previously noted­—­and the rates declined for all Among children age 5 enrolled in an ECE pro- ages of young children between 2018 and 2020 gram, only 11 percent are in a primary school (figure 3.24). preparation or “prep” class; the remainder go to nursery school (16 percent) or a playgroup Enforcement of existing policy is not priori- (6 percent) (figure 3.26). The content and tized: compulsory education begins at age 5, skills taught in these programs are unclear. FIGURE 3.23  Private providers of early childhood education serve large shares of children age Chapter 3 | Pakistan Human Capital Review 3–5 in Pakistan provinces, 2019–20 Balochistan Khyber Pakhtunkhwa Punjab Sindh Private Private 35% 45% Public Public Private 44% Public 49% 51% Private Public 55% 56% 65% Source: World Bank calculations using Pakistan Social and Living Standards Measurement Surveys 2004–20. 85 FIGURE 3.25  Enrollment among children age 3–8 rises by age but physically separate classrooms for preprimary is very low for children age 3–5 in Pakistan, 2019–20 children within public primary schools. Added Percent barriers are the insufficient number of teachers, Out of school Currently enrolled in preprimary Currently enrolled in primary inadequate professional development for teach- 100 ers of children age 3–5, and insufficient financial resources. Barriers to demand for ECD services 80 include cost, lack of information, and, during the COVID-19 pandemic, fear of infection.40 Fees that 60 are prohibitive include nontuition costs such as uniforms, book fees, and transportation. “Too expensive” was the top-cited reason for not en- 40 rolling a preschool-age child in 2019 by almost all parents (89 percent), whereas “child too young” 20 was the top-cited reason (77 percent) for not en- rolling children in the same age group the year 0 before. In addition, parents lacked information 3 years 4 years 5 years 6 years 7 years 8 years about the quality of programs, their rights, and Age school accountability. Source: World Bank calculations using Pakistan Social and Living Standards Measurement Surveys 2019–20. Social protection FIGURE 3.26  More than half of 5-year-old Pakistan has ratified international treaties that children in Pakistan are out of school, 2019–20 guarantee children special protections, but no Grade 2 national laws directly address child protection, 1% and provinces have administrative authority for Grade 1 child protection programs. Although neither the 8% Prep Constitution of Pakistan nor any law at the na- 11% tional level directly addresses child protection, Play group 6% Out of Pakistan has ratified key international treaties school 58% and enacted several national policies relevant Nursery to child protection. The government created the 16% National Commission for the Rights of the Child in 2017, with the mandate of examining existing laws relating to children and proposing new leg- Source: World Bank calculations using Pakistan Social and Living islation related to child rights and abuse. Prov- Chapter 3 | Pakistan Human Capital Review Standards Measurement Surveys 2019–20. inces have a key role in administering protection and care programs for children, leaving poli- Multigrade classrooms in which children range cy and program implementation fragmented. in age from 3 to 8 and overcrowding are com- While several initiatives are in place across the mon at both the preprimary and primary levels, provinces, coordination, capacity, and coverage creating difficult conditions for teachers. challenges persist. Provincial social welfare de- partments provide education facilities for chil- Though barriers to supply and demand have dren with special needs and homes for rehabil- not been thoroughly analyzed, infrastructure, itating orphaned children. workforce issues, and costs are clearly salient. Supply-­side constraints tend to focus on in- Social assistance transfer programs provide frastructure, and key barriers are the lack of critical poverty reduction support. While there 86 are some social assistance programs provid- focused on children under 5. Pakistan tends to ing targeted support for children, most pro- use a siloed approach to service delivery for grams are targeted at reducing vulnerability ECD rather than developing multisectoral poli- and poverty at the household level. In 2007, cies and establishing coordinated approaches the government developed the National Social to program implementation. Although there Protection Strategy (NSPS) with the vision of have been some efforts to implement multi- developing a comprehensive social protection sectoral programs, initiatives have been small system covering the poorest and most vulnera- in scale and have not been managed by an ble populations. NSPS served as the foundation overarching coordinating body working across for launching the Benazir Income Support Pro- districts or provinces. Multisectoral approach- gramme (BISP) in 2011, Pakistan’s flagship safe- es applied through the nutrition sector have ty net program, providing unconditional cash shown success in terms of multisectoral buy- transfers to the most vulnerable. This interven- in, planning, and establishing governance plat- tion has been key because young children are forms.43 In some cases, outcomes cannot be overrepresented among the poor, who account improved without multisectoral efforts, such as for 21.9 percent of the population as a whole stunting, which requires support from nutrition but 27.0 percent of children from birth to age 2 programs, parenting education initiatives, and and 29.6 percent of children age 3–5. The BISP improved access to high-quality water and sani- resulted in a paradigm shift in social protection tation services. and a large increase in pro-poor social protec- tion spending, with beneficiaries amounting to There are substantial gaps in Pakistan between 52 percent of the population.41 the country’s policy goals and actual delivery of services in terms of coverage, quality, and sub- The BISP provides important protections for sequent human development. Denboba et  al. young children. It supports access to education (2014) outlined 25 key interventions considered by providing conditional cash transfers (CCTs) by global experts to be essential for favorable to eligible children starting at age 5; these sti- child development, aligned with four core early pends cover one year of preprimary enrollment childhood-relevant sectors: health and nutri- as well as primary education. And recogniz- tion, water and sanitation, education, and so- ing stunting as a key challenge in the country’s cial protection. Across sectors and provinces in human capital development, BISP incentivizes Pakistan, policy environments are fair or good, uptake of health services among pregnant and whereas program implementation and uptake lactating women and children under 2 through of recommended practices are lacking­ —i­ n some its Health and Nutrition CCT program. The BISP cases, dramatically so. Intervention coverage in has laid out several policies and strategies to as- the two most populous provinces in Pakistan, Chapter 3 | Pakistan Human Capital Review sist the work of ministries and social protection Punjab and Sindh, show that policy intentions agencies and avoid duplication of efforts, and are not yet matched by access (figures 3.27 the Emergency Cash program for COVID-19 has and 3.28). Challenges include unavailable or been ranked among the top four global initia- unaffordable programming, small or informal tives for social protection.42 scale of programming, and weak data collec- tion and monitoring and evaluation systems, Using multisectoral approaches to which obscure rates of coverage, inputs and out- improve ECD outcomes comes and, in turn, the use of data for program improvements. Multisectoral policies and initiatives for ECD have been limited, and there is no coordinat- Multisectoral policies and programs would im- ing body to oversee multisectoral interventions prove ECD but require enabling environments 87 FIGURE 3.27  Access to essential interventions among young children in Punjab Ratings are indicated by shading: ■ ■ ■ ■ Full coverage ■ ■ ■ ■ More than half ■ ■ ■ ■ Less than half ■ None/unknown Pregnancy Birth 12 months 24 months 36 months 54 months 72 months Counseling on adequate diet during Exclusive Complementary feeding Adequate, nutritious, and safe diet pregnancy breast­ feeding Iron and folic Therapeutic zinc supplementation for diarrhea acid for pregnant mothers Prevention and treatment for acute malnutrition (moderate and severe) Health and Micronutrients Micronutrients: supplementation and fortification nutrition Antenatal visits Immunizations Attended delivery Deworming Planning for family size and spacing Access to healthcare Prevention and treatment of parental depression Access to safe water Water and Adequate sanitation sanitation Hygiene/handwashing Maternal education Education about early stimulation, growth, and development Education Early childhood and preprimary programs Continuity to quality primary education Birth registration Social Parental leave and adequate childcare protection Child protection services Social assistance transfer programs Transfer programs Source: World Bank elaboration based on Government of Punjab reports and Multiple Indicator Cluster Survey data. and programs that support caregivers as well actions for caregivers as well as children include as children. Coordination across sectors better access to affordable, high-quality childcare ser- supports children and families.44 Multisectoral vices,45 parental leave for both mothers and Chapter 3 | Pakistan Human Capital Review approaches are promoted through the NCF de- fathers, and breastfeeding and family-friendly veloped by the World Health Organization, the policies. Improved infrastructure and service United Nations Children’s Fund (UNICEF), and delivery in the water and sanitation sector would the World Bank (2018) to improve young chil- need to converge with parenting education pro- dren’s access to good health, nutrition, respon- grams promoting good health and hygiene prac- sive care, safety and security, and early learning. tices. Also at the program level, multisectoral The NCF recognizes that each sector plays a and multigenerational parenting education pro- role in supporting ECD; however, existing inter- grams would empower families by increasing ventions being delivered by sectors may not be access to services supporting caregiver well- sufficient, there may be gaps, or services may being, such as caregiver mental health services, require strengthening and coordination to am- while promoting nurturing care for children. plify benefits. At the policy level, critical policy Effective program investments in Pakistan, 88 FIGURE 3.28  Access to essential interventions among young children in Sindh Ratings are indicated by shading: ■ ■ ■ ■ Full coverage ■ ■ ■ ■ More than half ■ ■ ■ ■ Less than half ■ None/unknown Pregnancy Birth 12 months 24 months 36 months 54 months 72 months Counseling on adequate diet during Exclusive Complementary feeding Adequate, nutritious, and safe diet pregnancy breast­ feeding Iron and folic Therapeutic zinc supplementation for diarrhea acid for pregnant mothers Prevention and treatment for acute malnutrition (moderate and severe) Health and Micronutrients: supplementation and fortification nutrition Antenatal visits Immunizations Attended delivery Deworming Planning for family size and spacing Access to healthcare Prevention and treatment of parental depression Access to safe water Water and Adequate sanitation sanitation Hygiene/handwashing Maternal education Education about early stimulation, growth, and development Education Early childhood and preprimary programs Continuity to quality primary education Birth registration Social Parental leave and adequate childcare protection Child protection services Social assistance transfer programs Source: World Bank elaboration based on Government of Sindh reports and Multiple Indicator Cluster Survey data. including training early childhood educators, an ECD service delivery mapping has been con- would strengthen frontline services for young ducted, national ECD policy dialogue is under children and their caregivers and continuity of way, an ECD policy framework is in progress, Chapter 3 | Pakistan Human Capital Review services to ensure successful transitions for a national steering committee and a technical young children from their home to preschool to working group for ECD have been established, a early primary school environments. joint planning commission of the National SUN Secretariat and the Nutrition Section of the Min- Multisectoral initiatives in process provide in- istry of Planning, Development and Special Ini- sights and synergistic pathways forward. In tiatives is being set up, and a PC-II (pro forma 2013, Pakistan joined the Scaling Up Nutrition for development projects) has been developed (SUN) Movement through the Planning and De- to secure government resources for ECD. There velopment division in the Prime Minister’s office may be potential to further build the capacity of to improve development partner coordination this ministry and its provincial counterparts to and multisectoral coordination for improved nu- govern such initiatives. The Inter-Provincial Ed- trition at the provincial level. Under this initiative, ucation Ministers’ Conference may also provide 89 a forum for experience sharing and developing protection and education platforms,53 which offer multisectoral ways forward.46 new opportunities for program implementation in Pakistan. In addition, fatherhood engagement Parenting as a key multisectoral issue is also emerging as a key focus area for parent- ing to promote development, learning, and the A child-centric and cost-effective approach to prevention of violence in global contexts.54 Be- service delivery in early childhood would be cause parenting is a constant throughout the cross cutting, including by improving service early years, multiple parent-oriented platforms delivery to parents. By leveraging the potential at different stages of parenting and child devel- synergy of existing platforms such as the Lady opment can improve developmental outcomes Health Worker Program or ECE programs, stake- for children as newborns, infants, toddlers, pre- holders have opportunities to multiply the bene- schoolers, and primary school–age children. fits of comprehensive programs through parents and guardians, who have continuous opportuni- ty to provide for children’s holistic needs. ACCELERATING PROGRESS TO IMPROVE EARLY CHILDHOOD Parenting programs increase parents’ capacity DEVELOPMENT OUTCOMES to nurture child development so that children can reach their potential. The rights of the young- Early childhood development is a key means of est citizens to receive nurturing, sensitive, and accelerating Pakistan’s progress toward human emotionally responsive care, to play, to explore capital development and becoming a middle-in- their environment, and to learn are noted in the come country. The Pakistan@100: Shaping the United Nations Convention on the Rights of the Future report recommends easing constraints Child,47 which Pakistan ratified in 1990. The im- to human capital development to accelerate portant role parents and guardians play in sup- growth in the medium term in three ways. First porting the provision of care and the obligation of is to increase parents’ awareness of child de- states to support parenting are recognized in the velopment, health, nutrition, and stimulation. Convention and by researchers.48 However, the Second is to reduce stunting through ECD pro- provision of nurturing care in Pakistan is under- grams. Third is to improve the efficiency of pub- mined by the multiple risks reviewed above. Par- lic investment in human capital development, enting programs provide information, resources, particularly health and education.55 These rec- and skills to support parents in optimizing nur- ommendations directly align with policies to turing care so their young children can develop improve services for young children and their healthily and be protected from harm. caregivers through access to high-quality, com- prehensive programs that will directly affect Chapter 3 | Pakistan Human Capital Review There is consistent global evidence that parent- ECD and learning outcomes, particularly for vul- ing programs are effective in promoting ECD. nerable populations. Protecting and augment- The effectiveness of parenting programs starting ing children’s experiences of nurturing care, before and at birth is well established through in- mitigating risks, and expanding early interven- ternational studies.49 In Pakistan, parenting pro- tions to close gaps in children’s growth and de- grams have been successfully delivered through velopment will advance Pakistan’s human capi- a range of platforms, including community health tal accumulation. services,50 health clinics,51 and parenting pro- gram research in Pakistan has addressed both Pakistan has opportunities to improve ECD child inputs and caregiver well-being.52 Global ev- through multiple entry points corresponding to idence also suggests that there are opportunities national priorities. In addition to aligning with the to implement parenting programs through social recommendations of immediate reform in the 90 Pakistan@100 report, strategies to support ECD CONCLUSION AND align directly with the forthcoming World Bank– RECOMMENDATIONS Pakistan Country Partnership Framework, which lays out the country’s priorities according to Efficient and meaningful progress toward five anchoring areas of need and potential, col- human capital development in Pakistan will not lectively referred to as the 5 Gs: girls’ and boys’ occur without efficient and meaningful invest- education, growing healthy, green and clean Pa- ments in young children. Investments in young kistan, growing inclusively, and governance. ECD children’s healthy development from gestation is one of eight areas singled out for reform, and through the primary years are a prerequisite for investments in nurturing care and education for improved human capital outcomes for the coun- young children through each of the 5Gs would try’s growth. The quality of Pakistan’s human promote improved parenting, ECD and learning capital has been uneven and lower than in com- outcomes (figure 3.29). ECD strategies also align parable countries. Gaps between groups are with the steps and priorities of the 2020 Govern- wide, particularly disadvantaging those in low- ment of Pakistan, SUN Secretariat, and UNICEF income households, rural areas, and girls and mapping exercise, which are organized around women. As discussed throughout the chapter, the components of the NCF. risks for vulnerable children intersect and lead FIGURE 3.29  Investing across sectors to promote early childhood development will accelerate progress in human capital development in Pakistan Improved early Improved parenting Improved learning childhood development Girls’ and boys’ Growing Green and Growing Governance education healthy clean Pakistan inclusively Children are well nourished, Children receive early stimulation Children are nurtured and especially in the rst 1,000 days and learning opportunities protected from stress • Good nutritional status of mothers • Responsive and engaging • Avoid household and community and expecting mothers interactions with parents/caregivers stressors (neglect, violence, • Exclusive and continued • Opportunities for age-appropriate displacement, household shocks) breastfeeding and play-based learning through • Positive emotional connections with Chapter 3 | Pakistan Human Capital Review • Proper feeding of children under 5; quality preschool programs parents/caregivers micronutrients • Supportive discipline • Immunization for and treatment of childhood illnesses • Good hygiene practices Delivering on the essential interventions above requires e orts across sectors Quality, Educated Family leave Safety Water Reduced diverse, and and and quality, nets and and income a ordable empowered a ordable response to sanitation poverty food women childcare shocks Source: Adapted using the World Bank Investing in the Early Years framework and the upcoming Pakistan Country Partnership Framework. 91 to a development trajectory that fails to support Expand cash transfer support to achieving full potential. It is estimated that set- young families from pregnancy backs resulting from COVID-19-related setbacks through preschool may be equivalent to eight years of forgone • Supplement cash transfers, where they exist, progress­ —­and outcomes among young children with programs such as the Lady Health Work- in Pakistan, although measured imperfectly, er Program, and develop training and aware- are already showing declines. Given that invest- ness programs to empower mothers at the ments in human capital are a multiplier for pro- community level to support each other. ductivity and that investments in ECD provide • Use community-based mother support groups some of the best returns among human capital to monitor child development outcomes and investments, a coordinated, multisectoral ap- identify early markers for interventions. proach to improving outcomes for young chil- dren will lead to immense benefits for individual Improve the supply of clean water children, their families, and the country’s eco- and adequate sanitation with piped nomic prospects. water delivery wherever feasible • Improve existing tariff structures with a view to Investments must focus on caregivers, program enhancing sustainability, treatment of human coverage, and program quality. More efficient waste, water treatment, and water quality and effective investments in young children in tracking. Pakistan require multisectoral efforts, including • Ensure all public facilities such as schools investing in parents and educators. Integrating and basic health units have adequate WASH high-quality, contextually relevant parenting facilities. programs into existing health, social protection, and education platforms from gestation to age Encourage behavior change at the 8 will substantially improve child development household and community levels and learning outcomes. Motivating parents to • Develop and implement parenting programs engage with their young children and enroll that teach caregivers how to engage in respon- them in high-quality childcare and ECE would go sive interactions with their children and pro- far in reducing learning poverty. Access to high- vide early stimulation. These programs would quality early care and education programs will encourage breastfeeding, teach parents how to also boost women’s labor force participation. provide stimulation, and underscore the bene- While this chapter has focused more on access fits of proper nutrition and positive parenting. to programs than on their quality, the impor- • Encourage reduction in the incidence of open tance of quality cannot be overstated. Programs defecation and encourage good hygiene that fail to achieve minimal quality standards practices such as regular handwashing. Chapter 3 | Pakistan Human Capital Review are ineffective, waste scarce resources, and undermine demand for services, which com- Ensure the convergence and promises coverage over time. Building capacity geographic co-location of sectoral among frontline health workers and cultivating interventions for nutrition a qualified, appropriately compensated ECE and early development at the workforce are the foundations of high-quality household level programming. • Empower district health and education au- thorities, and ensure they have the capacity Policy and programming options to provide op- and resources commensurate with their re- portunities and accelerate progress in ECD in sponsibilities to implement early childhood Pakistan include: development and nutrition programs. 92 • Strengthen monitoring and evaluation sys- will go far toward realizing Pakistan’s ambitions tems for ECD and nutrition by incorporat- to meaningfully reduce stunting and learning ing children’s developmental milestones poverty and enable children to grow up physical- into health information systems while also ly healthy, and intellectually sound in safe, clean strengthening education information sys- environments with responsive, nurturing care- tems to collect key health markers. givers. Coordinated programming across sec- • Regularly evaluate the reach and quality of tors will lead to improved parenting, better child such multisectoral programs to ensure they outcomes, and higher learning levels today and remain fit-for-purpose. tomorrow. In addition to the enormous econom- ic benefits to improved human development re- Multiple entry points exist for supporting better sulting from increased performance, innovation, development outcomes for young children in and productivity, effective investments in ECD Pakistan that fully align with the country’s goals will reap incalculable returns on investment for and a life cycle approach to human capital devel- each young child who is able to receive nurturing opment. Recommendations related to the NCF care and thrive during childhood and beyond. Chapter 3 | Pakistan Human Capital Review 93 NOTES 1. The term caregiver in this chapter includes par- 23. Proulx et al. 2021. ents; the terms caregiver and parent are used 24. Yoshikawa et al. 2020. interchangeably. 25. Engle et al. 2011. 2. World Bank 2022a. 26. Denboba et al. 2014. 3. World Bank 2022b. 27. WHO, UNICEF, and World Bank 2018. 4. Government of Pakistan, UNICEF, and SUN 2021. 28. NIPS and ICF 2019. 5. Sarfraz 2020. 29. MNHSRC Nutrition Wing 2018; NIPS and ICF 2019. 6. Lu et al. 2020. 30. World Bank 2019. 7. The data reflect children’s performance at the end 31. WHO and UNICEF 2021. of grade 1 rather than in preschool, and the survey data are not weighted or nationally representative, 32. World Bank 2018. but they point to low achievement levels. 33. World Bank 2018. 8. Geven 2020. 34. UNICEF Pakistan 2021. 9. Geven and Hasan 2020. 35. World Bank 2018. 10. Black et al. 2017. 36. WHO and UNICEF 2021. 11. Berens and Nelson 2015; Folger et al. 2018. 37. World Bank 2018. 12. Campbell et al. 2014; Gertler et al. 2014; Walker et al. 38. World Bank 2018. 2011a. 39. NIPS and ICF 2019. 13. Walker et al. 2011b. 40. Mureed 2021. 14. Campbell et al. 2014; Gertler et al. 2014. 41. Gentilini et al. 2020. 15. Engle et al. 2011. 42. Gentilini et al. 2020. 16. Multiple Indicator Cluster Survey (MICS) data. 43. Zaidi et al. 2018. 17. Pakistan Social and Living Standards Measurement 44. Britto et al. 2018; Richter et al. 2017. Survey data. 45. Devercelli and Beaton-Day 2020. 18. It is challenging to compare prevalence of disabil- 46. National Curriculum Council 2020. ities across countries—attempts to standardize question sets for disability status are under way— 47. UN CRC 1989. and country rankings are not currently available. 48. Britto et al. 2017. These data do not account for children with undiag- 49. Jeong et al 2021. nosed intellectual or physical disabilities or children with developmental delays. 50. Yousafzai et al. 2014. 19. WHO and World Bank 2011. 51. Khan et al. 2018. Chapter 3 | Pakistan Human Capital Review 20. Sindh MICS 2018–2019. 52. Husain et al. 2021. 21. Lu et al. 2020. 53. Britto et al. 2017. 22. In the MICS data for all four provinces, there were 54. Kruk et al. forthcoming. very small practical differences in the share of boys 55. World Bank 2019. and girls experiencing multiple risks. These small dif- ferences were statistically significant in Punjab and Balochistan but not in Sindh or Khyber Pakhtunkhwa. 94 REFERENCES Baker-Henningham, H., and S. Walker. 2018. “Effect of health problems for the next generation.” Psychological Transporting an Evidence-based, Violence Prevention In- Trauma: Theory, Research, Practice, and Policy 12(S1): tervention to Jamaican Preschools on Teacher and Class- p. S195. wide Child Behaviour: A Cluster Randomised Trial.” Global Cuartas, J., D. C. McCoy, C. Rey-Guerra, P. R. Britto, E. Mental Health 5. Beatriz, and C. Salhi. 2019. “Early Childhood Exposure Berens, A. E., and C. A. Nelson. 2015. “The Science of to Non-violent Discipline and Physical and Psychological Early Adversity: Is There a Role for Large Institutions in Aggression in Low- and Middle-income Countries: Na- the Care of Vulnerable Children?” The Lancet 386 (9991): tional, Regional, and Global Prevalence Estimates.” Child 388–398. Abuse & Neglect 92: 93–105. Bhutta, Z. A., S. Soofi, S. Cousens, S. Mohammad, Z. A. Cuartas, J., K. P. Ward, J. Ma, and A. Grogan-Kaylor. 2020. Memon, I. Ali, A. Feroze, F. Raza, A. Khan, S. Wall, and J. “Physical Punishment and Colombian Children and Ado- Martines. 2011. “Improvement of Perinatal and Newborn lescents’ Cognitive and Behavioral Outcomes.” Journal of Care in Rural Pakistan through Community-based Strat- Applied Developmental Psychology 68: 101140. egies: A Cluster-randomised Effectiveness Trial.” The Lan- Cuartas, J., D. G. Weissman, M. A. Sheridan, L. Lengua, cet 377 (9763): 403–412. and K. A. McLaughlin. 2021. “Corporal Punishment and Black, M. M., and F. E. Aboud. 2011. “Responsive Feeding Elevated Neural Response to Threat in Children.” Child Is Embedded in a Theoretical Framework of Responsive Development 92 (3): 821–832. Parenting.” The Journal of Nutrition 141 (3): 490–494. Denboba, A. D., L. K. Elder, J. Lombardi, L. B. Rawlings, Black, M. M., S. P. Walker, L. C. Fernald, C. T. Andersen, A. R. K. Sayre, Q. T. Wodon. 2014. Stepping Up Early Child- M. DiGirolamo, C. Lu, D. C. McCoy, G. Fink, Y. R. Shawar, hood Development: Investing in Young Children for High J. Shiffman, and A. E. Devercelli. 2017. “Early Childhood Returns. Washington, DC: World Bank. http:/ /documents​ Development Coming of Age: Science through the Life .worldbank.org/curated/en/868571468321240018​ Course.” The Lancet 389 (10064): 77–90. /Stepping-up-early-childhood-development-investing-in​ -young-children-for-high-returns. Britto, P. R., S. J. Lye, K. Proulx, A. K. Yousafzai, S. G. Mat- thews, T. Vaivada, R. Perez-Escamilla, N. Rao, P. Ip, L. C. Devercelli, A., and F. Beaton-Day. 2020. Better Jobs and Fernald, and H. MacMillan. 2017. “Nurturing Care: Pro- Brighter Futures: Investing in Childcare to Build Human moting Early Childhood Development.” The Lancet 389 Capital. Washington, DC: World Bank. (10064): 91–102. Early Learning Partnership. 2019. Early Learning Part- Britto, P. R., M. Singh, T. Dua, R. Kaur, and A. K. Yousafzai. nership Systems Research: Phase I (2017–19) Summary 2018. “What Implementation Evidence Matters: Scal- Brief. ing-up Nurturing Interventions that Promote Early Child- Engle, P. L., L. C. Fernald, H. Alderman, J. Behrman, hood Development.” Annals of the New York Academy of C. O’Gara, A. Yousafzai, M. C. De Mello, M. Hidrobo, N. Sciences 1419 (1): 5–16. Ulkuer, I. Ertem, and S. Iltus. 2011. “Strategies for Reduc- Chapter 3 | Pakistan Human Capital Review Britto, P.R., H. Yoshikawa, and K. Boller. 2011. “Quality of ing Inequalities and Improving Developmental Outcomes Early Childhood Development Programs in Global Con- for Young Children in Low-income and Middle-income texts: Rationale for Investment, Conceptual Framework Countries. The Lancet 378 (9799): 1339–1353. and Implications for Equity and Commentaries.” Social Eshel, N., B. Daelmans, M. C. D. Mello, and J. Martines. Policy Report 25 (2): 1–31. 2006. “Responsive Parenting: Interventions and Out- Campbell, F., G. Conti, J. J. Heckman, S. H. Moon, R. Pinto, comes.” Bulletin of the World Health Organization 84: E. Pungello, and Y. Pan. 2014. “Early Childhood Invest- 991–998. ments Substantially Boost Adult Health.” Science 343 Fesseha, E., and A. Pyle. 2016. “Conceptualising Play- (6178): 1478–1485. based Learning from Kindergarten Teachers’ Perspec- Cuartas, J., 2020. “Heightened risk of child maltreatment tives.” International Journal of Early Years Education 24 amid the COVID-19 pandemic can exacerbate mental (3): 361–377. 95 Fisher, J., M. C. D. Mello, V. Patel, A. Rahman, T. Tran, S. Low-Resource Setting: Cluster Randomized Controlled Holton, and W. Holmes. 2012. “Prevalence and Determi- Trial.” Depression and Anxiety 38 (9): 925–939. nants of Common Perinatal Mental Disorders in Women Jeong, J., E. E. Franchett, C. V. Ramos de Oliveira, K. Reh- in Low- and Lower-middle-income Countries: A System- mani, and A. K. Yousafzai. 2021. “Parenting Interventions atic Review.” Bulletin of the World Health Organization 90: to Promote Early Child Development in the First Three 139–149. Years of Life: A Global Systematic Review and Meta-­ Folger, A.T., E. A. Eismann, N. B. Stephenson, R. A. Sha- analysis.” PLoS Medicine 18 (5): e1003602. piro, M. Macaluso, M. E. Brownrigg, and R. J. Gillespie. Khan, G. N., S. Kureishy, S. Ariff, A. Rizvi, M. Sajid, C. 2018. “Parental Adverse Childhood Experiences and Off- Garzon, A. A. Khan, S. de Pee, S. B. Soofi, and Z. A. Bhu- spring Development at 2 Years of Age.” Pediatrics 141 (4). tta. 2020. “Effect of Lipid-based Nutrient Supplement­ —­ Gentilini, U., M. Almenfi, I. Orton, and P. Dale. 2020. Social Medium Quantity on Reduction of Stunting in Children Protection and Jobs Responses to COVID-19: A Real-Time 6–23 Months of Age in Sindh, Pakistan: A Cluster Ran- Review of Country Measures. Washington, DC: World Bank. domized Controlled Trial. PloS One 15 (8): e0237210. https://openknowledge.worldbank.org/handle​/10986​ Khan, M. A., S. S. Owais, S. Maqbool, S. Ishaq, H. J. Khan, /33635. F. A. Minhas, J. Hicks, M. A. Khan, and J. D. Walley. 2018. Gertler, P., J. Heckman, R. Pinto, A. Zanolini, C. Ver- “Is Integrated Private-clinic–based Early Child Develop- meersch, S. Walker, S. M. Chang, and S. Grantham-Mc- ment Care Effective? A Clustered Randomised Trial in Pa- Gregor. 2014. “Labor Market Returns to an Early Child- kistan.” BJGP Open 2 (2). hood Stimulation Intervention in Jamaica.” Science 344 Kruk, M. E., T. P. Lewis, C. Arsenault, Z. A. Bhutta, G. Irimu, (6187): 998–1001. J. Jeong, Z. S. Lassi, S. M. Sawyer, T. Vaivada, P. Waiswa, Geven, K. 2020. “Learning Poverty and ASER 2019.” In A. and A. K. Yousafzai. Forthcoming. “Making Health and P. Idara-e-Taleem-o-Aagahi (ed.), Annual Status of Edu- Social Systems Work For All Children in Low-income and cation Report (ASER) Pakistan, National (Rural). Lahore, Middle-income Countries: Structural Innovations to De- Pakistan: ASER Pakistan Secretariat and ITA. liver High-Quality Services.” The Lancet. Geven, K., and A. Hasan. 2020. Learning Losses in Paki- Lu, C., J. Cuartas, G. Fink, D. McCoy, K. Liu, Z. Li, B. Dael- stan Due to COVID-19 School Closures: A Technical Note mans, and L. Richter. 2020. “Inequalities in Early Child- on Stimulation Results. Washington, DC: World Bank. hood Care and Development in Low/middle-income Countries: 2010–2018.” BMJ Global Health 5 (2): Government of Pakistan, UNICEF (United Nations Chil- e002314. dren’s Fund), and SUN. 2021. Pakistan Country Experi- ence on Early Childhood Development: Integrated Early Malik, I., and L. Pop. 2013. “Consolidating Social Protec- Childhood Development (Early Childhood Development) tion.” Pakistan Policy Note 11, World Bank, Washington, from Policy to Programming. Presentation at the WHO DC. SEARO Regional Meeting, April 2021. McCoy, D. C., J. Cuartas, J. Behrman, C. Cappa, J. Hey- Haq, W., S. H. Raza, and T. Mahmood. 2020. “The Pan- mann, F. López Bóo, C. Lu, A. Raikes, L. Richter, A. Stein, Chapter 3 | Pakistan Human Capital Review demic Paradox: Domestic Violence and Happiness of and G. Fink. 2021. “Global Estimates of the Implications Women.” PeerJ 8: e10472. of COVID-19-related Preprimary School Closures for Children’s Instructional Access, Development, Learning, Howard, L. M., E. Molyneaux, C. L. Dennis, T. Rochat, A. and Economic Well-being.” Child Development 92 (5): Stein, and J. Milgrom. 2014. “Non-psychotic Mental Dis- e883–e899. orders in the Perinatal Period.” The Lancet 384 (9956): 1775–1788. MNHSRC (Ministry of National Health, Services, Regu- lations and Coordination) Nutrition Wing. 2018. Nation- Husain, N., T. Kiran, B. Fatima, I. B. Chaudhry, M. Husain, al Nutrition Survey 2018 Volume I. http://phkh.nhsrc.pk​ S. Shah, P. Bassett, N. Cohen, F. Jafri, S. Naeem, Z. Zadeh, /sites/default/files/2021-03/National%20Nutrition​%20 C. Roberts, A. Rahman, F. Naeem, M. I. Husain, and N. Survey%20Key%20Findings%20Volum%201%20 Chaudhry. 2021. “An Integrated Parenting Intervention UNICEF%202018.pdf. for Maternal Depression and Child Development in a 96 Moya, A., P. Serneels, A. Desrosiers, V. Reyes, M. J. Torres, Rahman, A., Z. Haq, S. Sikander, I. Ahmad, M. Ahmad, and A. Lieberman. 2021. “The COVID-19 Pandemic and and A. Hafeez. 2012. “Using Cognitive-behavioural Tech- Maternal Mental Health in a Fragile and Conflict-affected niques to Improve Exclusive Breastfeeding in a Low-liter- Setting in Tumaco, Colombia: A Cohort Study.” The Lan- acy Disadvantaged Population.” Maternal & Child Nutri- cet Global Health 9 (8): e1068–e1076. tion 8 (1): 57–71. Munir, M. M., U. Rubaca, M. H. Munir, and B. Munir. 2021. Rahman, A., A. Malik, S. Sikander, C. Roberts, and F. “An Analysis of Families Experiences with Young Children Creed. 2008. “Cognitive Behaviour Therapy-based Inter- with Intellectual and Developmental Disabilities (IDDs) vention by Community Health Workers for Mothers with during COVID-19 Lockdown in Pakistan.” International Depression and Their Infants in Rural Pakistan: A Clus- and Multidisciplinary Journal of Social Sciences 10 (1): ter-randomised Controlled Trial.” The Lancet 372 (9642): 81–103. 902–909. Mureed, S. 2021. Web-based Poll Survey to Assess Impact Richter, L. M., B. Daelmans, J. Lombardi, J. Heymann, F. L. of COVID-19 on RMNCAH&N Services: Perceptions of Key Boo, J. R. Behrman, C. Lu, J. E. Lucas, R. Perez-Escamil- Informants. Islamabad: Health Services Academy NIH. la, T. Dua, and Z. A. Bhutta. 2017. “Investing in the Foun- dation of Sustainable Development: Pathways to Scale National Curriculum Council. 2020. Single National Cur- up for Early Childhood Development.” The Lancet 389 riculum Early Childhood Care and Education Grade Pre I (10064): 103–118. 2020. Islamabad: Ministry of Federal Education and Pro- fessional Training. Sarfraz, H. 2020. “Early Learning Progress in Pakistan: Recent Evidence from ASER 2015–19.” In A. P. Idara-e- Neuman, M. J., and S. Powers. 2021. “Political Pri- Taleem-o-Aagahi (ed.), Annual Status of Education Re- oritization of Early Childhood Education during the port (ASER) Pakistan, National (Rural). Lahore, Pakistan: COVID-19 Pandemic: A Comparative Policy Analysis of ASER Pakistan Secretariat & ITA. Low- and Middle-Income Countries.” Policy Research Working Paper 9872, World Bank, Washington, DC. UNICEF (United Nations Children’s Fund) Pakistan. 2021. https://openknowledge.worldbank.org/handle/10986​ WASH: Water, Sanitation and Hygiene. https://www.unicef​ /36648. .org/pakistan/wash-water-sanitation-and-hygiene-0. Ac- cessed 18 February 2022. NIPS (National Institute of Population Studies) and ICF. 2019. Pakistan Demographic and Health Survey 2017–18. Unwin, H. J. T., S. Hillis, L. Cluver, S. Flaxman, P. S. Goldman, Islamabad, Pakistan, and Rockville, MD, USA: NIPS and A. Butchart, G. Bachman, L. Rawlings, C. A. Donnelly, O. Rat- ICF. mann, and P. Green. 2022. “Global, Regional, and National Minimum Estimates of Children Affected by COVID-19-as- Pace, G. T., S. J. Lee, and A. Grogan-Kaylor. 2019. “Spank- sociated Orphanhood and Caregiver Death, by Age and ing and Young Children’s Socioemotional Development in Family Circumstance Up to Oct 31, 2021: An Updated Mod- Low- and Middle-income Countries.” Child Abuse & Ne- elling Study.” The Lancet Child & Adolescent Health. https:/ / glect 88: 84–95. doi.org/10.1016/S2352-4642(22)​00005-0. Pakistan Bureau of Statistics. 2021. “Pakistan Social and Walker, S. P., S. M. Chang, M. Vera-Hernández, and S. Living Standard Measurement Survey (PSLM), 2019– Grantham-McGregor. 2011a. “Early Childhood Stimula- Chapter 3 | Pakistan Human Capital Review 2020.” Government of Pakistan. https:/ /www.pbs.gov.pk​ tion Benefits Adult Competence and Reduces Violent Be- /content/pslm-district-level-survey-2019-20-microdata. havior.” Pediatrics 127 (5): 849–857. Accessed 17 February 2022. Walker, S. P., T. D. Wachs, S. Grantham-McGregor, M. M. Patel, V. 2021. “Scale up Task-sharing of Psychological Black, C. A. Nelson, S. L. Huffman, H. Baker-Henningham, Therapies.” The Lancet 399 (10322): 343–345. https:/ /​ S. M. Chang, J. D. Hamadani, B. Lozoff, and J. M. M. Gard- doi.org/10.1016/S0140-6736(21)02736-7. ner. 2011b. “Inequality in Early Childhood: Risk and Pro- Proulx, K., R. Lenzi-Weisbecker, R. Rachel, K. Hackett, V. tective Factors for Early Child Development.” The Lancet Cavallera, B. Daelmans, and T. Dua. 2021. “Responsive 378 (9799): 1325–1338. Caregiving, Opportunities for Early Learning, and Chil- WHO (World Health Organization). 2016. INSPIRE: Seven dren’s Safety and Security during COVID-19: A Rapid Re- Strategies for Ending Violence against Children. Geneva: view.” MedRxiv. WHO. 97 WHO (World Health Organization) and UNICEF (United World Bank. 2022b. Early Childhood Development. World Nations Children’s Fund). 2021. Progress on Household Bank: Washington, DC. https:/ /www.worldbank.org/en​ Drinking Water, Sanitation and Hygiene 2000–2020: Five /topic/earlychildhooddevelopment#1. Years into the SDGs. Geneva: WHO and UNICEF. Yoshikawa, H., A. J. Wuermli, P. R. Britto, B. Dreyer, J. F. WHO (World Health Organization), UNICEF (United Na- Leckman, S. J. Lye, L. A. Ponguta, L. M. Richter, and A. Stein. tions Children’s Fund), and World Bank. 2018. Nurturing 2020. “Effects of the Global Coronavirus Disease-2019 Care for Early Childhood Development: A Framework for Pandemic on Early Childhood Development: Short- and Helping Children Survive and Thrive to Transform Health Long-term Risks and Mitigating Program and Policy Ac- and Human Potential. Geneva: WHO. https:/ /nurturing​ tions.” The Journal of Pediatrics 223: 188–193. -care​.org​/resources​/Nurturing​_Care​_Framework​_en​.pdf. Yousafzai, A. K., M. A. Rasheed, and Z. A. Bhutta. 2013. WHO and World Bank. 2011. World Report on Disability. “Annual Research Review: Improved Nutrition: A Pathway Geneva: WHO. to Resilience.” Journal of Child Psychology and Psychiatry 54 (4): 367–377. World Bank. 2018. When Water Becomes a Hazard. A Di- agnostic Report on the State of Water Supply, Sanitation, Yousafzai, A. K., M. A. Rasheed, A. Rizvi, R. Armstrong, and and Poverty in Pakistan and Its Impact on Child Stunting. Z. A. Bhutta. 2014. “Effect of Integrated Responsive Stim- WASH Poverty Diagnostic. Washington, DC: World Bank. ulation and Nutrition Interventions in the Lady Health Worker Programme in Pakistan on Child Development, World Bank. 2019. Pakistan@100: Shaping the Future. Growth, and Health Outcomes: A Cluster-randomised Washington, DC: World Bank. https:/ /openknowledge​ Factorial Effectiveness Trial.” The Lancet 384 (9950): .worldbank.org/handle/10986/31335. Accessed 18 Feb- 1282–1293. ruary 2022. Zaidi, S., Z. Bhutta, S. S. Hussain, and K. Rasanathan. World Bank. 2021. Government Expenditure on Education 2018. “Multisector Governance for Nutrition and Early (Total % GDP). World Bank Data. https://data.worldbank​ Childhood Development: Overlapping Agendas and Dif- .org/indicator/SE.XPD.TOTL.GD.ZS. Accessed 18 Febru- fering Progress in Pakistan.” BMJ Global Health 3: 1–8. ary 2022. Zakar, R., M. Z. Zakar, and S. Abbas. 2016. “Domestic World Bank. 2022a. Early Years: The Foundation for Violence against Rural Women in Pakistan: An Issue of Human Capital. Washington, DC: World Bank. https://​ Health and Human Rights.” Journal of Family Violence 31 www.worldbank.org/en/programs/earlyyears. (1): 15–25. Chapter 3 | Pakistan Human Capital Review 98 Boosting nutrition SUMMARY and extends through the first two years of life. 4 So, to tackle child stunting, a comprehensive Malnutrition is one of Pakistan’s biggest devel- multisectoral approach is needed during this opment challenges, with long-term implications critical window of opportunity. Focusing on for human capital development and economic only a subset of factors will not achieve the growth. The proportions of young children stunt- expected results and is likely to leave a large ed (40  percent), underweight (30  percent), majority of the children with increased risk of wasted (18  percent), and overweight (10  per- illness and death, poor cognitive development, cent) in Pakistan are higher than in any other slower education attainment, and diminished country in South Asia. These national averages lifelong income potential. mask wide subnational disparities. For example, stunting­—a­ n indicator of chronic malnutrition While there has been progress in reducing child- that is strongly associated with brain develop- hood malnutrition, the pace has been slow, de- ment, physical growth, and human capital­ —­ is spite decades of national and provincial efforts. worse among rural and poor households. In Analysis of data from several years of National Khyber Pakhtunkhwa, 48  percent of children Nutrition Surveys puts the average annual re- under 5 are stunted, and in Sindh, 50 percent. duction rate of stunting at 0.5  percent­—­too Even in the wealthiest income quintile, about sluggish to achieve a significant reduction. The 23 percent of children under 5 suffer from stunt- major constraints to nutrition program effec- ing, so poverty is not the only factor. Making tiveness lie in implementation, supervision, matters worse, the repeated waves of COVID-19 monitoring, and evaluation. At both the federal since March 2020 and the floods in 2022 have and provincial levels, there is commitment to reduced incomes, threatened livelihoods, exac- improve nutrition outcomes. Recent strategic erbated food insecurity, and disrupted the pro- directions recognize the multisectoral nature of vision and use of basic health and nutrition serv- nutrition and seek to engage multiple stakehold- ices. And they have set back Pakistan’s efforts to ers and sectors to improve nutrition outcomes. reduce malnutrition. But across programs, inadequate leadership, governance, coordination, staffing, technical ca- Chapter 4 | Pakistan Human Capital Review Three key factors are associated with stunting pacity, procurement, and financing plague their in Pakistan. The first is insufficient food intake, effectiveness. Appropriate stewardship and due to inadequate breastfeeding and comple- oversight bodies and institutional arrangements mentary feeding practices. The second is a for implementation are not always set up in time range of environmental risks such as unclean to drive strategy and implementation. And dis- water and deficient sanitation and hygiene. trict institutions (where they exist) often are not And the third is the inadequacy of services empowered to make local implementation deci- for children and women, including those for sions and resolve issues expeditiously. Although reproductive, maternal, newborn, and child there have been improvements in data collec- health and for immunization, nutrition, and tion and tracking of indicators, their reliability, family planning. Stunting starts in the womb completeness, and use in decisionmaking re- due to poor maternal nutrition and infection main weak. 101 Based on the analyses in this study and re- stunted, 18 percent wasted, 54 percent anemic, view of the experiences of countries that have and 52 percent deficient in vitamin A. The prev- improved nutrition, the chapter recommends alence of overweight among children under 5 strengthening stakeholder momentum and co- almost doubled between 2011 and 2018, from ordination to make nutrition a national priority­ 5 percent to 9.5  percent. Malnutrition is also —­as well as mobilizing additional financing and prevalent in women of reproductive age: more tracking expenditures. It also recommends for- than 14  percent are underweight, 24  percent mulating a national nutrition policy and costed overweight, 14  percent obese, 27  percent de- provincial sector plans to address bottlenecks ficient in vitamin A, and 43  percent anemic. and strengthen interventions. Among adolescents, about 7  percent of boys and 5 percent of girls are obese.1 INTRODUCTION All forms of malnutrition affect individual and national growth and productivity. Stunting, an Pakistan has a high burden of malnutrition in all indicator of chronic malnutrition, is most close- its forms (box 4.1). One-fifth of children are born ly associated with brain development, physical with low birthweight, and among children under growth, and thus human capital development. 5, 30  percent are underweight, 40  percent Stunting has lifelong and even intergeneration- al consequences for individuals­ —­and nations. BOX 4.1 Childhood stunting leads to increased mortal- ity, increased morbidity (in childhood and later in adulthood), decreased cognitive ability, a de- layed start and reduced attainment at school, Malnutrition: and far lower individual earnings and slower Key definitions national economic growth. Stunted children are a third less likely than other children to es- Underweight: Low weight-for-age, indicates cape poverty as adults. Childhood stunting has a combination of long-term and immedi- intergenerational implications: malnourished ate-term malnutrition. mothers are more than twice as likely as well-­ Stunting: Low height-for-age, indicates nourished mothers to have stunted children. chronic malnutrition. Wasting: Low weight-for-height, indicates Micronutrient deficiencies have substantial im- acute malnutrition. pacts on health and human capital. Iodine de- Micronutrient deficiencies: A lack of essen- ficiency, particularly in pregnancy, affects fetal tial vitamins and minerals—required in small development and child intelligence quotient Chapter 4 | Pakistan Human Capital Review amounts by the body for proper growth and (IQ). An average loss of 12.5–13.5 IQ points was development—includes deficiencies in io- found in children born to iodine-deficient moth- dine, iron, vitamin A, zinc, and folic acid. ers. Children with iodine deficiency lose 13 IQ Overweight and obesity: Abnormal or ex- points on average, making them less educable. cessive fat accumulation that presents a risk Iron deficiency is a common cause of anemia, to health. The prevalence of overweight and which contributes to increased maternal mor- obesity is commonly assessed by the body bidity and mortality as well as premature birth mass index (BMI, ­ the weight in kilograms di- and low birthweight. The deficiencies have im- vided by the square of the height in meters paired brain development and reduced produc- [kg/m2]). A BMI over 25 kg/m2 is defined as tivity in adults. Vitamin A deficiency increases overweight, a BMI over 30 kg/m2 as obese. the severity of measles, malaria, and diarrhea in children and increases morbidity, including 102 preventable pediatric blindness, and death. Vi- lipoproteins, high fasting blood glucose levels, tamin A supplementation for children has been and high blood pressure­ —a­ re among the prime linked to a 23 percent reduction in child mortal- contributors to death and disabilities in Paki- ity. Zinc deficiency is associated with reduced stan. The health and economic impacts of obe- immunity and with the increased incidence, se- sity are profound, particularly for child obesity, verity, and duration of diarrhea; and it has a neg- which puts children at risk of poor dietary and ative effect on child growth.2 physical activity habits early on and of develop- ing noncommunicable diseases later in life. Obesity among children, adolescents, and adults is a public health issue with serious im- Countries that have greatly reduced stunting plications for Pakistan’s human capital. Sever- show the potential of implementing a prov- al parameters associated with obesity­ high —­ en package of interventions through policy basal metabolic index, high levels of low-density change and programmatic action (box 4.2).3 For BOX 4.2 Stunting reduction in exemplar countries This box briefly summarizes the experiences of six coun- Key elements tries that have made substantial progress in achieving • Advocacy and nutrition championship, political will, health outcomes, including reducing stunting: Brazil, Ethi- and a targeted, data-driven, multisectoral approach. opia, Nepal, Peru, Senegal, and Thailand. The common • Civil society involvement. elements in their approaches include commitment and • Robust and comprehensive set of nutrition and food se- political will, donor engagement, multisectoral approach- curity indicators to monitor progress and guide planning. es, community mobilization, improved service delivery, strong data generation with rigorous monitoring, and Ethiopia strong communications. The experiences of these coun- Stunting rate: 67 percent in 1992, 38 percent in 2016. tries offer valuable lessons for Pakistan. Approach Brazil • Strategic pro-poor investments in agriculture, educa- tion, health, and sanitation: Stunting rate: 37.1 percent in 1974, 7.1 percent in 2006/07. • In agriculture, expanded extension services, training Approach centers for farmers, technical hotline for farmers, Chapter 4 | Pakistan Human Capital Review • Nutrition and pro-poor social development policies be- and social safety nets. ginning in the 1980s. • In education: new schools and improvements in ac- • During years of economic growth, policies to invest in cess, enrollment, literacy, and gender parity. education, health, sanitation, and food security. • In health: extension worker program to broaden • In 2002, launch of Fome Zero (Zero Hunger) policy, reach, driving improvements in maternal and child consolidating several health, education, and nutrition health and in reducing open defecation. conditional cash transfer (CCT) programs and scaling • Adoption of a “One Plan, One Budget, One Report” them up into Bolsa Família. As a result, improvements slogan, under which the country successfully coordi- across multiple social indicators, with maternal educa- nated donor and partner support, while maintaining tion the largest contributor to improved nutrition. (continued) 103 BOX 4.2, continued high control over strategic design and implementation Senegal decisions. Stunting rate: 34 percent in 1992, 17 percent in 2017. • Program design management at the national level and implementation at the community level. Approach • In health: A network of “health huts” and trained com- Key elements munity health workers set up to provide basic preven- • Focus on social sector spending, multisectoral pro- tive and curative care for mothers and children and ac- gramming, and data-driven decisionmaking. cess to contraceptives. • Use of high-quality data to inform implementation. • In education: Initiatives to improve universal access to education and the quality of education and teach- Nepal er training, leading to substantial increases in primary Stunting rate: 68 percent in 1995, 25 percent in 2022. school enrollment, gender parity, and adult literacy. • In water, sanitation, and hygiene (WASH): Donor-­ Approach supported reforms privatizing provision of WASH • Investments in health, including the Female Community services, which increased access to and improved the Health Volunteer program, under which 50,000 trained quality of water and sanitation services for the urban volunteers promote family planning, provide health population. Access to piped water had a notable im- counselling, and deliver basic services and products, and pact on stunting reduction. programs to promote antenatal care and safe delivery. • Investments in education: 98 percent of children en- Key elements rolled in primary schools by 2011. • Political commitment (presidential leadership), an in- • Sanitation efforts leading to decreased open defecation. novative multisectoral approach, donor support that allowed piloting innovative approaches, and communi- Key elements ty mobilization. • Multisectoral investments, large-scale donor engage- • Improvements in the economy due to remittance in- ment, and community engagement and women’s flows and increases in agricultural yields, which helped empowerment. reduce household food insecurity. Peru Thailand Stunting rate: 29.8 percent in 2005, 14.6 percent in 2014. Stunting rate: Over 22 percent in 1987, about 10 percent in Approach 2016. • Introducing a nutrition-focused poverty-reduction Approach strategy targeting the 880 poorest districts. • Use of community-based nutrition programs since the • Consolidating about 80 nutrition-related programs to 25. 1970s, followed by a “basic minimum needs” approach • Shifting from mostly feeding programs to a multi- under which communities were mobilized to plan for sectoral package focusing on the first 1,000 days. and monitor interventions. • A health insurance program, CCTs, national maternal, • By the early 1980s, nutrition integrated into several de- newborn, and child health programs, nutrition counsel- velopment policies. ling, family planning, growth monitoring, and improved • By the 1990s, visible thrust to achieve health for all, targeting of health services to underserved areas and Chapter 4 | Pakistan Human Capital Review with community mobilization instrumental in helping sanitation and hygiene efforts. achieve this. Key elements Key elements • Civil society advocacy, political will (political candi- • Advocacy and nutrition championship, political will and dates committed to a 5x5x5 pledge: reducing malnu- effective policies, community mobilization for health trition of children under 5 by 5 percent in 5 years), and service delivery, counselling, and data generation. evidence-based policy setting. • Strong multisectoral coordination and collaboration. • Targeting, multisectoral approaches, and improved service delivery and demand generation. Source: Brazil, Ethiopia, Nepal, Peru, and Senegal: Exemplars in Global • Strong data generation with rigorous monitoring and Health (https://www.exemplars.health/). Thailand: Heaver and Kachon� - a strong social and behavior change communication dam 2002; Ruel 2008; Tontisirin and Gillespie 1999. 104 campaign. example, the Kyrgyz Republic and Peru achieved The data covered fiscal years 2017–18, 2018–19, a more than 5 percent annual reduction in stunt- and 2019–20. The analysis followed Scaling Up ing. How? They increased the coverage of their Nutrition (SUN) guidelines to measure nutri- direct healthcare sector nutrition interventions. tion expenditures, to identify relevant minis- They addressed immediate determinants that tries, departments, cost centers, and financing could improve maternal nutrition and newborn documents, and to classify nutrition-relevant outcomes. They promoted early and exclusive expenditures as either nutrition-specific or nu- breastfeeding. And they coupled complemen- trition-sensitive.4 It then assigned to each inter- tary feeding practices with indirect nutrition in- vention a weight that represents the proportion terventions implemented by nonhealth sectors. of the line item that is truly contributing to im- The lessons from these countries point the way proving nutritional outcomes. for Pakistan to build more effective programs and reduce stunting faster. MALNUTRITION SITUATION The chapter’s findings are based on desk re- AND TRENDS views and consultations with key stakeholders to collect and analyze data pertaining to nutri- Malnutrition is the top driver of death and dis- tion interventions and their quality, coverage, abilities in Pakistan (figure 4.1). And while there service delivery and utilization, financing, and has been progress in reducing childhood mal- organizational arrangements. The World Health nutrition, the pace has been slow, despite two Organization (WHO) Stewardship for Health decades of national and provincial efforts. For Systems framework was used to assess specific example, the 2018 National Nutrition Survey topics, such as policies and programs. Financing (NNS) estimated the average annual reduction for nutrition was assessed at the federal level, rate of stunting at 0.5 percent­ —­too sluggish to using integrated financial management informa- significantly reduce the prevalence of malnutri- tion system (IFMIS) datasets, and at the provin- tion. Faster change will be needed if Pakistani —­ cial level­ Punjab and Sindh­ —­using IFMIS data- children are to attain their full growth and devel- sets and the Annual Development Plan (ADP). opment potential. FIGURE 4.1  Top 10 risk drivers of deaths and disabilities in Pakistan, 2019 Behavioral risks Environmental/occupational risks Metabolic risks Percent change, 2009 2019 2009–19 Malnutrition 1 1 Malnutrition –17.0 Air pollution 2 2 Air pollution –9.0 Chapter 4 | Pakistan Human Capital Review Wanter, sanitation, and hygiene 3 3 High blood pressure 38.4 Tobacco 4 4 Dietary risks 28.6 High blood pressure 5 5 Tobacco 3.1 Dietary risks 6 6 Water, sanitation, and hygiene –28.6 High fasting plasma glucose levels 7 7 High fasting plasma glucose levels 41.0 High body mass index 8 8 High body mass index 53.0 High LDL 9 9 High LDL 31.1 Kidney dysfunction 10 10 Kidney dysfunction 33.3 Source: Institute for Health Metrics and Evaluation. 105 Among South Asian countries, Pakistan has one to health and nutrition services; water, sanitation, of the highest rates of malnutrition (see box and hygiene (WASH) conditions; education; nu- 4.1). And South Asia, with stunting prevalence of trition diversity; and maternal, infant, and young 34.0 percent among children under 5, exceeds child feeding practices. Nutrition diversity may the global average of 21.3 percent. On key nutri- be considered a particularly low hanging fruit as tion indicators, Pakistan’s rates are the highest it can be addressed by fortification policies, both in South Asia (figure 4.2). On stunting, Pakistan voluntary and mandatory, for staples including stands at the higher end among all countries salt fortified with iodine and iron, wheat flour and that have conducted Demographic and Health rice fortified with iron, Vitamin B12, and folate, and Surveys since 2010. In Pakistan, stunting, un- milk and oil fortified with Vitamin A and Vitamin D. derweight, and wasting rates remain consider- ably high (figure 4.3). Stunting in Pakistan is not just a poverty issue. While malnutrition is higher in children under High national stunting rates mask large 5 from lower wealth quintiles, it is high even variations­ among provinces, and for poor, —­ in the middle and upper wealth quintiles (fig- marginalized, and extremely vulnerable groups ure 4.5). Knowledge of such variations is im- versus the rest of the population. Stunting rates portant for targeting analysis, advocacy, and range from 33 percent in Islamabad Capital Ter- communications. ritory and 36 percent in Punjab to over 46 per- cent in Balochistan and 48  percent in Khyber Stunting in children under 5 also varies with age, Pakhtunkhwa and the Newly Merged Districts. peaking at 24–36 months. Around 17  percent Within provinces, stunting rates across districts of children are born with low birthweight,5 and vary considerably; prevalence in several dis- the prevalence of stunting then rises to nearly tricts exceeds 50 percent. 50  percent at 24–36 months, remains almost at that level until 48 months, and then shows a The prevalence of stunting is high (31 percent) in small decline (figure 4.6). Pakistan’s urban areas and even higher (41 per- cent) in rural areas (figure 4.4). The higher rate in Micronutrient deficiencies among children under rural areas reflects such factors as poorer access 5 remain persistently high and vary widely across FIGURE 4.2  Key nutrition indicators for children under 5 in South Asian countries Percent Average GDP per capita (current US$) Stunted Underweight Wasted Overweight Anemia in women 15–49 years 60 12,000 Average GDP per capita 50 10,000 Chapter 4 | Pakistan Human Capital Review 40 8,000 30 6,000 20 4,000 10 2,000 0 0 Afghanistan Nepal Pakistan Bangladesh India Bhutan Sri Lanka Maldives Source: World Bank calculations using data from World Development Indicators 2020; Nepal Multiple Indicator Cluster Survey 2019; Pakistan National Nutrition Sur- 106 vey 2018. FIGURE 4.3  Trends in prevalence of stunting, FIGURE 4.4  Stunting prevalence in Pakistan, rural versus urban underweight, and wasting in children under 5 in areas Pakistan, 1990–2018 Percent Percent 100 Stunting 60 80 Demographic and Health Surveys 50 60 62 40 40 37 52 38 48 47 National Nutrition Surveys 40 42 36 34 32 20 26 24 20 0 Urban Rural Urban Rural Urban Rural Urban Rural Islamabad FATA 0 Punjab Sindh Khyber Balochistan 1990 1995 2000 2005 2010 2015 2018 Pakhtunkhwa Underweight Source: World Bank calculations using data from Demographic and Health Survey 2017–18. 50 Note: FATA is Federally Administered Tribal Areas. 40 40 38 National Nutrition Surveys provinces. Over half of children are deficient in vitamin A (12  percent severely deficient), 6 out 30 of 10 children are deficient in vitamin D (over 29 Demographic and Health Surveys 13 percent severely deficient), and 18 percent suf- 23 20 fer from zinc deficiency. According to the 2018 NNS, over the past two decades more than half 10 of children age 6–59 months have been anemic, with iron-deficiency anemia accounting for more than half the total. Like other indicators of mal- 0 1990 1995 2000 2005 2010 2015 2018 nutrition, childhood anemia also varies across provinces, with rates as high as 70.5 percent in Wasting 20 Balochistan and 68.3  percent in Khyber Pakh- 17 tunkhwa and the Newly Merged Districts. Ma- National Nutrition Surveys ternal anemia­ —­which can contribute to stunting Chapter 4 | Pakistan Human Capital Review 15 because anemia can lead to babies of low birth- 13 weight that are small for their gestational age­—­is high in Pakistan: nearly 43 percent of women of 10 reproductive age (pregnant and not pregnant) 9 Demographic and Health Surveys are anemic, and the proportions are higher in Bal- 7 ochistan (61 percent), and Khyber Pakhtunkhwa 5 and the Newly Merged Districts (53  percent). Among pregnant women in Pakistan 35 percent 0 are anemic, with the highest rate of 54 percent in 1990 1995 2000 2005 2010 2015 2018 Balochistan. Anemia in women from the richest Source: Pakistan Demographic and Health Surveys 2001, 2013, quintile is also high (37 percent), but less than in 2018; Pakistan National Nutrition Survey 2018. the poorest quintile (52 percent).6 107 FIGURE 4.5  Stunting by wealth quintiles among children under 5 the number of deaths had reached 30,333.7 in Pakistan During the pandemic, lockdowns and restric- 2018 2013 tions on movement reduced incomes, increas- 22 ing poverty and limiting vulnerable households’ Richest 23 economic and physical access to nutritious and adequate food. Disruptions in the provision and 30 Fourth use of health and nutrition services, including re- 38 productive, maternal, and other healthcare serv- 32 Middle ices, put women and their children at higher risk 41 of malnutrition. Food insecurity disproportion- Second 45 ately affects poor and socially disadvantaged 56 people­ —­the very people who both are at higher 57 risk of COVID-19 infection and have poorer ac- Poorest 62 cess to healthcare services.8 The analyses in this 0 10 20 30 40 50 60 70 report show that close to 36.4 million people in Percent Pakistan are highly vulnerable to food insecurity because of natural and human-made hazards, Source: World Bank calculations using data from Demographic and Health Survey 2017–18. including the pandemic. Another 49 million peo- ple (25 percent) are moderately food insecure, Impact of the COVID-19 pandemic and 21 million households (10 percent) are se- on malnutrition verely food insecure.9 Since March 2020, five increasingly destructive The disruption of routine health and nutrition waves of COVID-19 have swept through Paki- services increased the risk of all forms of mal- stan, pushing millions into poverty and great- nutrition. During the COVID-19 pandemic, health er food insecurity, and thus vulnerability to all facilities have had to switch to isolation, quar- forms of malnutrition. As of March 2022, con- antine, and high-dependency units to meet the firmed COVID-19 cases totaled 1,522,4194, and increasing patient load, limiting their ability to FIGURE 4.6  Stunting by age in months in Pakistan Percent Severe stunting Moderate stunting 50 40 Chapter 4 | Pakistan Human Capital Review 30 20 10 0 0–5 6–8 9–11 12–17 18–23 24–35 36–47 48–59 Age (months) Source: Demographic and Health Survey 2017–18. 108 deliver routine services. Outpatient therapeutic had reduced the physical activity of 67 percent program services were limited, and admissions of respondents and increased the weight of of severely malnourished children at stabiliza- 39 percent.14 tion centers were reduced, reflecting transport constraints and patients’ fear of visiting health facilities.10 Similar fears restricted home vis- DETERMINANTS OF STUNTING its and community-based services provided by Lady Health Workers in rural areas.11 The limited Using a well-accepted nutrition causal frame- distribution of vitamin A and iron supplementa- work (figure 4.7), the analysis here explores the tion left some children and pregnant women un- key factors associated with stunting in Pakistan. served. An early estimate in May 2020 was that In particular, it focuses on the association be- the disruptions in all essential services could tween the prevalence of stunting and the imme- result in a 22  percent increase in both mater- diate determinants of nutrition­ —a ­ dequate feed- nal and child mortality in Pakistan by the end of ing, adequate care, and adequate environmental 2020.12 Reduced numbers of antenatal care vis- health (box 4.3). The association holds true at its alone brought an increased risk to the birth- the national, provincial, and district levels. Coun- weights, nutrition, and health of children born terfactual simulations show that stunting prev- during this period. However, as the pandemic alence could decrease if a higher proportion of has eased, the number of antenatal care visits children benefited from adequate feeding, care, has increased to close to the previous level. and environmental health. Healthy feeding habits have been disrupted. Food intake, environmental health, and care for COVID-19-related constraints on the purchase children and women are the main determinants of diverse and nutritious foods (due to loss of of malnutrition at the most immediate level.15 livelihoods and reduced income) may well have Factors such as income poverty, gender, and kept mothers and young children from consum- education underpin all three, alongside the larg- ing healthful diets. Breastfeeding rates dropped er political, economic, social, and cultural envi- among mothers with COVID-19 infection, who ronments. The widely accepted United Nations feared transmitting the disease to their babies. Children’s Fund Nutrition Causal Framework In addition, lockdowns and limited mobility re- (adapted in figure 4.7) is a useful construct: duced access to clean water and safe sanita- children can receive adequate feeding, environ- tion services, conditions that could exacerbate mental health, and care­—­or none, or anything in malnutrition. between­ —­and thus can fall into eight mutually exclusive categories that represent the various The COVID-19 pandemic may have contribut- combinations (figure 4.8). Chapter 4 | Pakistan Human Capital Review ed to another form of malnutrition­ —­increased obesity and overweight. Physical inactivity and Very few of Pakistan’s children under 2 have ad- lifestyle changes due to stay-at-home orders, equacy in all three dimensions, and the propor- along with disruptions in consumption patterns, tion of children with multiple adequacies is low. may have contributed to Pakistan’s already high More than 26 percent of children lack adequacy rates of obesity and overweight. In a May 2020 on any of the three dimensions, and fewer than study to assess the level of physical activity of 2 percent of children have adequacy on all three the population of Pakistan during the pandem- dimensions. The highest percentages are those ic, 89.6 percent of respondents (more females for whom only environmental health is adequate than males) reported inadequate physical activ- (45.6 percent), and less than 10 percent of chil- ity.13 A web-based survey conducted in Punjab dren benefit from adequate environmental in late 2020 found that the COVID-19 lockdown health plus either adequate feeding or adequate 109 FIGURE 4.7  The Nutrition Causal Framework Maternal and child undernutrition The 1,000-day window of opportunity Inadequate IMMEDIATE CAUSES Disease dietary intake Unhealthy household Household Inadequate care environment and lack food insecurity of health services UNDERLYING CAUSES Income poverty Employment, dwelling, assets, remittances, transfers Lack of capital Financial, human, physical, social, and natural BASIC CAUSES Social, economic, and political context Source: Adapted from UNICEF (1998). care (figure 4.8). Given these inadequacies, the food security were important, including the the country’s huge malnutrition burden is no food security term would cause the feeding ade- enigma. quacy term to lose its significance. Food insecurity has an independent effect on Pakistan fares poorly on all of the known deter- Chapter 4 | Pakistan Human Capital Review the probability of being stunted. If a family is minants of stunting: food secure, it has a lower probability of being • Infant and young child feeding practices are stunted, holding constant for the above ade- very poor; only a very small share of children quacies, mother’s education, age of child, and under 2 are fed as recommended by WHO. wealth category, regardless of the degree of • Children born with low birthweight are at food insecurity (mild, moderate, or severe). This higher risk of stunting. effect on the probability of being stunted is in- • The coverage of critical healthcare indicators dependent of benefiting from adequate feeding. for mothers and children in the first 1,000- And the fact that the feeding adequacy remains day window (conception to age 2) is poor. significant even when the food security control • Although WASH practices have improved in is included indicates that the knowledge of what the past two decades, open defecation per- to provide the child with is also important. If only sists, and water quality remains an issue. 110 BOX 4.3 Indicators used for assessing the adequacy of food, environmental health, and care Adequate feeding • Living in cluster with fewer than 25 percent prac- • For children younger than 6 months, exclusive ticing open defecation. breastfeeding. • For children between 6 and 9 months, breastfeed- Adequate care ing and Minimum Acceptable Diet for age (as de- • For children under 12 months, any three of the fined by the World Health Organization [WHO]). following four: received any vitamin A or micro- • For children between 9 and 24 months: Breast- nutrient supplementation; mother received an- feeding and Minimum Acceptable Diet for age tenatal care (more than two, with the first in the (as defined by WHO). first trimester); mother received any iron and • For children between 6 and 24 months, not folic acid supplements; birth attended by trained breastfeeding and Minimum Acceptable Diet for personnel. age (as defined by WHO). • For children between 12 and 24 months, any four of the following five: received any vitamin A or Adequate environmental health micronutrient supplementation; mother received • Improved water source. antenatal care (in the first trimester); mother re- • Improved sanitation. ceived any iron and folic acid supplements; birth • Handwashing (before preparing food, before eat- attended by trained personnel; and full vaccina- ing, before feeding child, after handling feces, and tion with or without card. after defecating). • Mother’s education is strongly associated Ministry of National Health Services, Regula- with malnutrition. tions and Coordination’s total expenditure that • Food security, however, improved between year and to 2.5 percent of the federal govern- 2011 and 2018. ment’s total expenditure (table 4.1). It was more than what the government spent in FY17/18 Annexes 4A, 4B, and 4C summarize, respec- (1.68 percent) and FY18/19 (1.45  percent). tively, Pakistan’s policies and strategies tackling However, overall nutrition-specific expenditure Chapter 4 | Pakistan Human Capital Review child malnutrition, its programs tackling malnu- decreased over the period, mainly because of trition, and the federal and provincial nutrition the end of some donor-funded federal nutrition stewardship functions. programs, and nutrition-sensitive expenditure increased (figure 4.9), largely because of the enhanced financing for social safety nets.16 The NUTRITION FINANCING period of the trend analysis includes the COVID- 19 pandemic, so that spending patterns and National financing trends may not be fully representative of the longer term. The data include on-budget nutri- Analysis for this chapter shows that, in FY19/20, tion financing only. There was no granular-lev- the federal government spent PKR  80.4 billion el information from which to assess whether on nutrition­ —­equivalent to 11 percent of the this nutrition spending was in line with national 111 FIGURE 4.8  Distribution of eight categories of adequacies in children younger than 24 months in Pakistan Food intake, environmental health, and care 1.6 Food intake and environmental health 9.1 Environmental health and care 8.5 Food intake and care 0.5 Food intake only 5.6 Environmental health only 45.6 Care only 2.5 None 26.2 0 10 20 30 40 50 Percent Source: World Bank calculations using data from National Nutrition Survey 2018. TABLE 4.1  Federal expenditure on nutrition interventions, FY19/20 Indicators Nutrition specific Nutrition sensitive Nutrition total In PKR, millions (current prices) 1,892.58 78,549.87 80,442.45 Per capita PKR (current prices) 236 9,804 10,040 Per capita USD (current prices) 1.50 62.05 63.54 Share of government expenditure (percent) 0.06 2.43 2.49 Share of expenditure of the Ministry of National Health Services, Regulations and Coordination (percent) 10.36 0.65 11.01 Source: World Bank calculations using data from federal Ministry of Finance and provincial Departments of Finance. nutrition plans, whether it was distributed eq- programs for immunization, maternal and child uitably across geographic areas, and whether it health, and population welfare. has benefited the poor. Budget absorption capacity­ —­the share of actu- While data limitations do not allow tracking ex- —­ al expenditure in the original budget­ increased. penditure at the program and activity levels, it Final nutrition-related expenditure as a pro- is useful to examine relative spending by minis- portion of the original nutrition-related bud- Chapter 4 | Pakistan Human Capital Review tries and departments. The Ministry of Finance, get exceeded 100 percent in FY19/20, up from Revenue, and Economic Affairs accounted for 82  percent in FY17/18 (figure 4.11). More than the largest shares (more than 50  percent) of 100  percent of budget absorption in FY19/20 federal nutrition expenditure in FY17/18 and is due to additional funding for cash grants, not FY18/19. In FY19/20, however, more than two- recorded in the original budget for FY19/20, to thirds of nutrition spending was implement- people impacted by the COVID-19 pandemic in ed by the Poverty Alleviation and Social Safety the second half of that financial year. Net Division, a wing of the Ministry of Finance, Revenue, and Economic Affairs, as it scaled up With the expected launch of the federal Tack- pro-poor services during the COVID-19 pan- ling Malnutrition Induced Stunting in Paki- demic. The Ministry of National Health Services, stan (TMIS) nutrition program, approved in Regulations and Coordination implemented May 2021, federal and provincial financing for 112 FIGURE 4.9  Expenditure trends for nutrition- nutrition is expected to increase. Of the total specific and nutrition-sensitive interventions TMIS amount (PKR  310 billion over five years, Billion PKR July 2021–June 2026), the federal government Nutrition speci c Nutrition sensitive proposes to contribute about PKR  159.2  bil- 100 lion, with the remainder coming in as contri- butions from provinces/regions. The proposed 80 distribution of federal funding to the prov- inces is as follows: Punjab, PKR  73.2  million; 60 Sindh, PKR  million; Khyber Pakhtunkhwa, 34.1  PKR 25.4 million; Balochistan, PKR 10 million; 40 and Islamabad and other areas, PKR 7.9 million. 20 Financing in Punjab and Sindh Nutrition finance tracking exercises supported 0 FY17/18 FY18/19 FY19/20 by the World Bank in Punjab and Sindh serve as models that all other provinces can build on.17 Source: World Bank calculations using data from the 2017–20 in- tegrated financial management system of federal and provincial Similar exercises in all provinces can inform al- departments. locations for nutrition, track expenditure rates, Note: In FY19/20, among the thematic sectors, social protection and examine funds flows and other parameters had the highest share of total nutrition spending at the federal level, and the health sector had a small share (figure 4.10). So- to determine financing-related bottlenecks and cial protection spending accounted for more than three-quarters constraints. of overall nutrition spending, a rise due to increased spending on social safety programs, through which the federal government provided COVID-19 relief packages. While Punjab’s total nutrition expenditure for FY19/20 is almost twice that of Sindh, in per capita terms both provinces spend similar FIGURE 4.10  Federal nutrition spending by amounts (table 4.2). Both provinces have con- thematic sectors, FY17/18–19/20 sistently increased their nutrition spending over Percent the past three fiscal years (figure 4.12). Like Water and Social Health Education Agriculture federal spending, the two provinces’ nutrition-­ sanitation protection specific spending is lower than their nutrition-­ 100 sensitive spending. Total nutrition spending translates to more than 4 percent of Punjab gov- 80 ernment spending and a bit less than 4 percent of Sindh government spending. Chapter 4 | Pakistan Human Capital Review 60 Unlike relative sectoral expenditure at the 40 federal level, in both provinces, the health sector accounts for the largest share of total 20 related expenditure (figures 4.13 nutrition-­ and 4.14). Most of the expenditure under the 0 health thematic sector is for nutrition-related FY17/18 FY18/19 FY19/20 counseling and treatment and for immuniza- Source: World Bank calculations using data from the 2017–20 in- tion, a nutrition-­specific intervention. In Pun- tegrated financial management system of federal and provincial jab, health accounted for almost half of the departments. total nutrition-related expenditure for FY19/20 and agriculture for another one-third­ —­ mostly 113 FIGURE 4.11  Nutrition budget absorption social protection for only 0.6 percent. The De- rate by nutrition category, federal level, partments of Health made the largest share of FY17/18–19/20 nutrition-related spending in both Punjab and Percent Sindh in FY19/20. The next-highest share was, 120 in Punjab, the Food Department (more than 25 percent), and in Sindh, the Education De- 110 partment (around 20 percent). In both provinc- es, nutrition budget absorption capacity is high 100 (see figure 4.14). 90 Nutrition sensitive 80 NUTRITION STEWARDSHIP AND Nutrition total INSTITUTIONAL ARRANGEMENTS 70 Nutrition speci c Stewardship remains one of the most criti- 60 cal elements for improving nutrition. Drawing FY17/18 FY18/19 FY19/20 a parallel between nutrition stewardship and WHO’s concept of health system stewardship Source: World Bank calculations using data from the 2017–20 in- tegrated financial management system of federal and provincial (see below), it is the government’s responsibili- departments. ty to promote the nutrition and well-being of the population. The concept also assumes that the incurred to ensure food security in the prov- quality of stewardship exercised can influence ince but also for research, development, exten- overall outcomes. Stewardship responsibilities sion services, and so on. The social protection could be shared between national and sub­ and education sectors together accounted for national authorities and local government, and 20 percent of total nutrition expenditure in across ministries and even some providers, but FY19/20 in Punjab. In Sindh, spending in the the national government remains the “steward health sector accounted for 62 percent of total of stewards.” This does not necessarily mean, nutrition-related spending, spending in edu- however, that the government needs to fund cation for around 25 percent, and spending in and provide all interventions.18 Leadership and TABLE 4.2  Nutrition public spending in Punjab and Sindh (FY19/20) Nutrition specific Nutrition sensitive Nutrition total Punjab Chapter 4 | Pakistan Human Capital Review  In PKR, millions (current prices) 16,537.31 56,245.76 72,783.07  Per capita PKR (current prices) 141.53 481.38 622.91  Per capita USD (current prices) 0.90 3.05 3.94  As share of government expenditure (%) 0.99 3.36 4.35 Sindh  In PKR, millions (current prices) 7,768.37 24,914.87 32,683.24  Per capita PKR (current prices) 151.09 484.58 635.68  Per capita USD (current prices) 0.96 3.07 4.03  As share of government expenditure (%) 0.87 2.78 3.65 Source: World Bank calculations using data from the 2017–20 integrated financial management system of federal and provincial departments. 114 FIGURE 4.12  Nutrition spending in Punjab and Sindh, FY17/18–19/20 Billion PKR Nutrition speci c Nutrition sensitive Nutrition total 80 60 40 20 0 Punjab Sindh Punjab Sindh Punjab Sindh FY17/18 FY18/19 FY19/20 Source: World Bank calculations using data from the 2017–20 integrated financial management system of federal and provincial departments. FIGURE 4.13  Nutrition spending by thematic sector in Punjab and Sindh, FY17/18–19/20 Percent Percent Punjab Sindh Agriculture and food systems Education Health Agriculture Education Health Social protection Social protection Water and sanitation Multiple Water and sanitation Others Multiple 100 100 80 80 60 60 40 40 20 20 0 0 FY17/18 FY18/19 FY19/20 FY17/18 FY18/19 FY19/20 Chapter 4 | Pakistan Human Capital Review Source: World Bank calculations using data from the 2017–20 integrated financial management system of federal and provincial departments. sustained political commitment are among the The WHO conceptual framework for the stew- top factors in every successful program and ardship functions of health systems identifies initiative that has improved nutrition at scale.19 six key functions/subfunctions that are critical Case studies of countries that have been suc- for effective stewardship. While many frame- cessful in reducing malnutrition, such as Brazil works and notions for stewardship have been and Peru, point to the crucial role of political proposed, the WHO framework for health sys- championing at the highest level in leveraging tem stewardship is particularly useful to assess nutrition as a cross-sectoral agenda. the effectiveness of nutrition stewardship. 115 FIGURE 4.14  Nutrition budget absorption BOX 4.4 capacity, Punjab and Sindh, FY17/18–19/20 Percent 100 Sindh The Scaling Up Punjab Nutrition Movement 90 The Scaling Up Nutrition (SUN) Move- ment, a global effort to build institutional 80 capacity for nutrition within governments, has played a seminal role in steering nutri- 70 tion coordination efforts. In Pakistan, the SUN Secretariat was established under the Planning Commission, with the Chief of 60 Nutrition serving as the SUN focal person. FY17/18 FY18/19 FY19/20 The SUN Core Group, the National Nutri- tion Committee, consisting of high-level Source: World Bank calculations using data from the 2017–20 in- tegrated financial management system of federal and provincial representatives from key government min- departments. istries and development partners, has been instrumental in building nutrition steward- The advocacy and engagement of development ship at the highest levels and in the creation partners have contributed to the development of multisectoral and multistakeholder plat- of stewardship structures and capacities in Pa- forms at the federal and provincial levels.1 kistan. Development partners’ funding and tech- Provincial SUN secretariats work at the nical support for nutrition have been significant subnational level. The provincial SUN units in building Pakistan’s early nutrition programs, in are at different levels of operationalization. developing nutrition strategies at the federal and The unit in Khyber Pakhtunkhwa takes an provincial levels, and in building commitment for especially active role in steering and track- nutrition and positioning it as a new public pol- ing implementation of the Khyber Pakh- icy agenda.20 Partners’ technical support and tunkhwa-Multisectoral Integrated Nutrition engagement allowed Pakistan to consider global Strategy, and the SUN Punjab Unit leads evidence to inform evolving strategic directions the coordination of a multisectoral nutrition for nutrition. Development partners continue response plan. The SUN units in Sindh and with their participation in several stewardship Balochistan are not functional, however. bodies and with their technical and financial Chapter 4 | Pakistan Human Capital Review support for programs, innovations, and monitor- Note ing and evaluation. The SUN Movement in 2013 1. SUN Movement Compendium 2014. brought in further technical support at the fed- eral and provincial levels and advanced the stew- ardship bodies and agenda in Pakistan (box 4.4). effectiveness. (Annex 4C outlines federal and provincial stewardship structures.) Pakistan’s Pakistan’s stewardship functions significant progress in policy and strategy formu- lation, its commitment building, and the overall Analysis of the six key functional domains of enabling environment for promoting programs stewardship of the WHO framework (table 4.3) and interventions to improve nutrition outcomes indicates several that are functioning well in Pa- is mainly due to the stewardship for nutrition kistan and others that need attention for greater that the country has developed over the years. At 116 TABLE 4.3  Key functions of the World Health Organization health system stewardship 1. Generating • Gaining a good understanding of: intelligence • Current and future trends and the performance of the existing system and institutional arrangements. • Information on important contextual factors and actors. • Possible policy options, based on national and international evidence and experience. 2. Formulating strategic • Articulating goals and objectives (medium- and longer-term), based on reliable policy direction intelligence and governing values, ethics, principles, etc. • Defining the roles of public, private, and voluntary sector actors in financing, resource generation, and stewardship functions. • Identifying the policy instruments and institutional arrangements required to achieve desired outcomes. • Outlining feasible strategies for making required changes. • Providing guidance for prioritizing expenditure, based on realistic resource and needs assessment. • Outlining arrangements to monitor performance and effects of change. 3. Ensuring tools for • Ensuring implementation of policies and strategies to achieve goals. This requires that implementation stewards: • Have powers commensurate with their responsibilities, powers that they use properly for policy implementation. • Set and ensure enforcement of fair rules, incentives, and sanctions that are in line with health system goals, for actors involved in provision, financing, and resource generation. • Ensure that the rights and responsibilities of users/consumers are defined and that mechanisms to protect consumers are exercised fairly. 4. Building coalitions • Build and maintain a wide variety of bilateral and multilateral relationships (from legally and partnerships binding to formal or loose affiliations). • The purpose of the partnership determines the parties involved (professional associations, government ministries/departments, private sector, international agencies, donors, nongovernmental organizations, regulatory bodies). • The purposes of the partnership determine whether it is a “one-off” (development of new policy or legislation, a media campaign, or a reform initiative) or is for regular/repeated tasks (e.g., planning or budget-setting, routine monitoring of service quality). 5. Creating a fit • Ensure appropriate institutional arrangements to meet policy objectives. between policy • Ensure a supportive management culture. objectives and • Establish clear linkages and lines of communication. organizational • Remove “structural” constraints to equitable and efficient resource use. structure and culture 6. Ensuring • Ensure the accountability of all actors. accountability • Be accountable to the population. Source: World Health Report 2000. the same time, some areas remain where stew- • Formulating strategic policy direction. At the ardship could play a greater role and institutional national and provincial levels, several strate- arrangements could be strengthened to further gy documents provide strategic direction to Chapter 4 | Pakistan Human Capital Review policy and program implementation. actions to improve nutrition outcomes. These documents were developed with broad stake- Pakistan’s application of three subfunctions holder engagement, are informed by evidence of the WHO framework has demonstrated on effective interventions, take into consid- strengths eration the country’s nutrition situation and These subfunctions are generating intelligence, challenges, and take a multisectoral approach. formulating strategic policy direction, and build- Most outline clear goals and objectives and ing coalitions and partnerships. Still, a few weak prioritize vulnerable groups. In some, however, spots remain in each area. These domains have the targets appear to be either too ambitious been instrumental in the increased commit- or too modest. The strategies are comple- ment to and priority given to nutrition in the past mented by guidelines that provide technical decade at the federal and provincial levels. and operational guidance on a specific topic 117 or intervention. The strategies do not assign function in Pakistan today, the picture is responsibilities for resource mobilization and mixed. Information on specific things such do not provide direction for resource prioriti- as cost-effective interventions, on possible zation based on realistic resource and needs institutional arrangements for different func- assessment­ —­ an important stewardship func- tions, and on recent international evidence tion, as there are rarely enough resources to is available in most strategies and guidelines do everything, and strategic choices are nec- approved by stewardship bodies. In other essary. Further, arrangements to monitor per- areas, however, intelligence is not gathered. formance, assess the effects of a new strate- For example, results from three household gic direction, and make changes informed by surveys have become available only after sev- experience are largely either not fully defined eral years, and there is no system to generate or not covered. Because there is no nation- reliable information more frequently.21 al nutrition policy or any provincial nutrition Although management information sys- plans, various piecemeal strategies provide tems to track nutrition indicators have been direction in different areas. strengthened and data collection has im- Stewardship action is required to ensure proved, data quality and reliability are gen- that the multiple strategies come together erally weak, and the data are not regularly under the policy umbrella and include stra- used for decisionmaking. Vital national and tegic directions for areas not covered by the provincial financing information is not gath- individual strategies. A unified policy would ered and tracked and is not available to stew- help harmonize the multiple strategies, es- ards; efforts in this area have only just begun. tablish relative priorities, guide funding al- This critical gap affects other stewardship locations, and establish a common results domains as well­ —­ for example, effective re- framework with defined indicators and moni- source planning and allocations in line with toring mechanisms, including core indicators policy commitments, tracking expenditure, for sectors other than health, thus serving as and ensuring accountability for spending. the reference for harmonization with other Further, updated information on human re- sector policies. The policy development and sources, their placements and vacancies, and approval process itself would bring several their capacity and performance are not read- stakeholders together; create greater sen- ily and systematically available. And there is sitization for nutrition among senior policy- no systematic sharing of key information with makers, the public, and civil society organiza- stewardship bodies at the district and sub- tions; and build consensus. district levels. • Building coalitions and partnerships. The stewardship bodies are some examples of Three key stewardship functions require Chapter 4 | Pakistan Human Capital Review the many coalitions and partnerships formed attention and strengthening for specific purposes­ —­some to accomplish a These functions are ensuring tools for imple- particular task (for example, a working group mentation, creating a fit between policy objec- to develop strategies or guidelines), and some tives and organizational structure and culture, to exist on an ongoing basis for technical en- and ensuring accountability. Policies and strat- gagement, reviews, advocacy, program de- egies are only as effective as their implementa- sign, and operational support and financing. tion. Policy and program analyses conclude that • Generating intelligence. A well-­functioning now that enabling policy-equivalent strategies intelligence generation domain gives all are generally in place, and several of them have those with stewardship responsibilities the been translated into programs (Annexes 4A and reliable, up-to-date information they need 4B), the overarching challenge across programs for decisionmaking. For this stewardship is implementation (box 4.5). 118 BOX 4.5 The biggest impediment to nutrition program effectiveness and impact is implementation Common issues plague programs: coordination, lead- for decisionmaking are weak areas. Program evaluations ership and staffing, capacity, procurement, and fiscal are sorely lacking, and except for some donor-support- issues. The program designs reflect attention to steward- ed evaluation efforts, such as the Food Fortification Pro- ship bodies and institutional arrangements for implemen- gramme, no evaluation or program results were available tation, but the lack of timely creation and operationaliza- for this review. Consultations for the review found that a tion of these arrangements undermines the stewardship wealth of knowledge exists in the government system necessary for providing strategic direction and driving im- but is not documented and shared. Given the numerous plementation. Irregular meetings of key decisionmaking approaches to nutrition programming across provinces bodies, long delays in approvals, frequent changes in and the number of smaller programs piloted, knowledge leadership, long-standing vacancies, procurement de- sharing could be of critical importance to guide smart and lays, and funds flow problems tend to be critical issues effective investments and implementation. across provinces and programs, significantly impeding implementation. Despite an almost decade-long engagement by the na- tional and provincial governments and various donors, Capacity constraints at various levels, including among the results of the 2018 National Nutrition Survey (NNS) functionaries and structures created to coordinate and show that neither stunting nor malnutrition has been facilitate implementation, are major problems. Moreover, significantly reduced. Several major provincial programs, district institutions (where they exist) are often not em- including multisectoral ones, have been initiated and/or powered to make local implementation decisions to resolve expanded since the conclusion of the NNS data collection issues expeditiously. In addition to the many common im- for its 2018 round, and with the upcoming Tackling Malnu- plementation challenges across programs, multisectoral trition Induced Stunting in Pakistan nutrition program and efforts also face challenges of sectoral convergence, includ- the expansion of the Benazir Income Support Program ing convergent planning and implementation to leverage Nashonuma, they have the potential to lead to improve- the full multisectoral impacts of the interventions. ments in nutrition outcomes, provided implementation issues are addressed in earnest and objective program Although there has been some progress in monitoring evaluations are conducted to capture impact. This activity and evaluation, tracking indicators, and collecting data, is all the more urgent, given the heightened risk of stunt- data reliability and completeness and the use of data ing due to the COVID-19 pandemic. Chapter 4 | Pakistan Human Capital Review • Ensuring tools for implementation. The WHO fairly. In Pakistan, there appear to be many framework is based on the premise that mismatches between the responsibilities and stewards need to have powers commensu- the powers of the stewardship bodies. And in rate with their responsibilities, and that they several instances where authority is given, it use these powers­ —­setting and ensuring en- is not used­ —­one of the main reasons for de- forcement of fair rules, incentives, and sanc- lays in key decisions. tions in line with the system goals. It also as- In a federal system like Pakistan’s, where sumes that the rights and responsibilities of the ultimate stewardship responsibility rests users/consumers are defined and that mech- with the national government, the steward- anisms to protect consumers are exercised ship responsibilities are dispersed across 119 provinces, districts, and subdistricts, and and operational capacities. The issue of or- implementation responsibilities are assigned ganizational and management culture­ —­the to the district and local levels. The program lack of continuity and institutional memo- analysis identified as one of the significant ry associated with the frequent leadership challenges to implementation the fact that changes and staff turnover, delayed decision- the bodies responsible for nutrition at the making, long-standing staff vacancies, and provincial department, district, and local lev- procurement delays­ —­has emerged as a els to which implementation responsibilities salient constraint to program implementa- are assigned lack capacity and powers to tion and reflects deep-rooted organizational carry out those responsibilities. To illustrate, culture issues. Several stewardship bodies the district nutrition committees in Sindh and do not meet at the frequency specified in Punjab, which are responsible for oversee- their terms of reference; this not only affects ing implementation, lack technical capaci- decisionmaking, timely reviews, and problem ty, do not have the powers to hold staff and solving, but also causes loss of momentum. functionaries accountable, and have neither A detailed assessment of the most appropri- access to the information they need to ob- ate organizational structures is required, es- jectively assess funding needs and priorities, pecially given the multiple departments and nor control over funds and funding priorities actors involved. to meet the real needs. Another aspect of im- • Ensuring accountability, responsibility, and an- plementation policy has to do with stewards swerability to the population. Stewards must having and exercising the powers to guide be accountable for their own actions to the the behavior of different actors. For example, other actors in the system and to the popula- while the Federal Nutrition Wing in the Minis- tion, and they must ensure that all the other try of National Health Services, Regulations actors in the system at all levels are also held and Coordination has the ultimate respon- accountable for their actions. The accountabil- sibility for national stewardship on nutrition, ity of stewards includes, among other things, it does not appear to have the powers to accountability for policy/strategy implementa- demand accountability for results or perfor- tion and achievement of goals, for efficient and mance from the nutrition units in the provin- effective use of resources to produce the de- cial planning and development departments. sired outcomes, and for timely decisionmaking • Creating a fit between policy objectives and and provision of adequate resources to the organizational structure and culture. To the other actors to enable them to deliver their re- credit of the high-level stewardship bodies, sponsibilities, and then to hold them account- there has been an effort to create new in- able for their performance and achievement stitutional structures aligned with the new of results. While this may be linked to a larger Chapter 4 | Pakistan Human Capital Review strategic directions­ —t­he multisectoral and systemic change across several levels, depart- Integrated reproductive, maternal, newborn, ments, and programs, the higher-level stew- adolescent, and child health and nutrition ardship bodies can institute several measures approaches at various levels. However, these and mechanisms to enhance accountability structures are largely unable to function as and lead to better outcomes. envisaged not only because of their lack of authority commensurate with responsibil- Pakistan has made good progress in ity but also because of other constraints. establishing stewardship bodies Funding and staffing continue to be based At the highest levels, the prime minister exercis- on vertical programs or sectoral programs es stewardship at the national level and senior with their own lines of communication and re- leaders exercise it in several provinces, and the porting, and they lack appropriate technical commitment to improve nutrition outcomes is 120 high. There is a recognition of the multifacet- development issue high on the national and ed nature of nutrition and the need for multi- provincial agenda. That entails building and sectoral actions to improve nutrition outcomes. committing to a sense of urgency for action, To make the strategic shift in direction from engaging all stakeholders, and mounting large- nutrition as a responsibility of the health sec- scale campaigns to generate greater public tor alone to one that brings in other sectors, awareness, concern, and engagement around several strategy documents and multisectoral critical nutrition issues.22 To drive the agenda stewardship structures have been developed forward, it will be extremely important to follow and are operational to varying degrees across through on the commitments in planning, pri- provinces. An assessment of the stewardship oritization (including allocation of resources), functions of these bodies using a WHO health and monitoring. stewardship function framework indicates many strengths and well-functioning areas (although It should protect and earmark nutrition funding, some weak spots remain to be addressed): pro- mobilize additional financing for nutrition, and viding strategic direction, building coalitions introduce and institutionalize systems to track and partnerships, and generating intelligence nutrition financing and expenditure. Financial for decisionmaking. Functional domains that re- allocations in line with commitments are criti- quire further attention point to the fact that the cal to translating policy intent into action and to stewardship bodies at the different levels need ensuring that nutrition programs and interven- powers commensurate with their responsibili- tions are adequately funded. ties, including those for ensuring enforcement, • Costing nutrition interventions and mobiliz- incentives, and sanctions; transforming organi- ing resources. Costed action plans to assess zational and management culture; and demand- real funding needs are essential to mobilizing ing greater accountability. resources, including from domestic sources, donors, the private sector, and specific levies (of course, with due consideration of conflicts ADDRESSING CONSTRAINTS TO of interest).23 PROGRAM EFFECTIVENESS • Earmarking nutrition funding and developing and institutionalizing a nutrition financing Pakistan is committed to meeting its Sustainable and expenditure tracking system. Institution- Development Goals stunting target, and it has alizing the assessment of nutrition financ- made encouraging progress on building commit- ing is critical to protect nutrition funding ment and developing strategies. But to protect and ensure the availability of reliable data and promote human capital accumulation, it will on nutrition financing to inform plans, guide need to take urgent and effective action, build- resource allocation, prioritize and monitor Chapter 4 | Pakistan Human Capital Review ing on its past work, to accelerate the reduction expenditure, enhance accountability, and in stunting and address the other forms of mal- make the case for investments by donors nutrition. The following recommendations are and government. In addition, a comprehen- drawn from Pakistan’s experiences, successes, sive system that tracks budget and expen- and disappointments and from lessons learned diture (domestic, external aid, on-budget, in countries that have reduced malnutrition. and off-budget), using standard formats and consistent codes and wordings across Prioritizing nutrition, including its years and government levels and various financing financial documents, such as the ADP and the integrated financial management infor- Pakistan needs to make nutrition an important mation system, will help in tracking nutrition national priority, framing it as an urgent human expenditure. 121 Crafting a comprehensive national with stunting to maximize the nutrition impacts nutrition policy of those policies. It would help to use the SUN system of nutrition markers to assess and Pakistan needs to develop and adopt a compre- then boost sectoral contributions to stunting hensive national nutrition policy and provincial reduction. plans. Such policy needs to address nutrition holistically and articulate the government’s Improving the policy environment commitment to improve nutrition outcomes. until the new nutrition policy is While several strategies exist and have been implementable serving as policy equivalents, an overarching policy, approved by Parliament, is a more pow- Pakistan needs to ensure timely approval and erful instrument to provide strategic direction updating of strategies, particularly those that to government ministries and departments and are cross-sectoral. Existing strategies and poli- lay important groundwork for funding priorities, cies should be updated to clearly reflect a com- laws, and implementation. mitment to nutrition­ —­ better defining the most effective actions in each sector to improve nu- Pakistan should address, through the nutrition trition outcomes, setting out the indicators to be policy, the immediate and underlying determi- tracked by sectors, and developing harmonized nants of nutrition, and prioritize the most vul- criteria across sectors for beneficiary selection nerable areas and populations with high-impact to ensure the convergence of sectoral inputs nutrition-specific interventions. Using a phased and benefits. Updating is also warranted to inte- approach, all pregnant and nursing women and grate new knowledge, analysis, or implementa- children under 2 could eventually be targeted tion lessons and to strengthen the institutional with convergent interventions so that the num- structures proposed in a strategy, particularly to bers of children who receive adequate feeding, ensure that they have authority commensurate environmental health interventions, and health- with their responsibility. care increase exponentially. It should pay close attention to the sustainability It should integrate nutrition services as part of of interventions in project design. Too often, in- primary care, to promote the co-location and terventions end when the project (that is, donor convergence of service delivery and reduce the funding) ends. Short-term interventions cannot vertical silos of nutrition services. An empha- make long-term progress in reducing malnutri- sis on behavior change communication and tion. In designing and implementing projects, it community-­ focused interventions could help is important to ensure that the interventions can address social norms around feeding and WASH be continued without the donor’s involvement. Chapter 4 | Pakistan Human Capital Review and generate demand for nutrition services. Ap- proaches for improving nutrition may need to be It should assess and articulate financing needs tailored, given that stunting rates are high even and relative priorities, an important aspect often in educated households of the middle and upper missing in strategies. It is critical to assess and wealth quintiles. articulate the real resource needs to translate the strategy into action, prioritize the financing It should require, through the nutrition policy, of the most cost-effective interventions across that policies and strategies in the health sector sectors, and outline the approach to fully cover and other relevant sectors should be examined high-burden areas. This approach helps build a through a nutrition lens. The focus should be common understanding of needs and priority on prioritizing the most effective actions to ad- interventions, and it can also help stewards mo- dress the determinants most closely associated bilize resources. 122 It should improve communication. Wide dis- to ensure that children under 2 receive three semination of strategy and policy documents things simultaneously: adequate food and nutri- within ministries and departments and to all ent intake, adequate environmental health inter- stakeholders is important to ensure that all par- ventions, and adequate care. Thus, convergence ties benefit from the direction afforded by these at the village and household levels to reach ben- documents. In several instances policy/strategy eficiaries (each mother-child dyad) with the full documents needed for this review were not ac- range of interventions is essential.27 cessible in the public domain and were not even available in the relevant departments. It should increase demand and uptake. Initia- tives to increase the uptake of services­ —­for Boosting program outcomes by example, behavior change communication, scaling up the most cost-effective social mobilization, and even conditional cash interventions transfers­—h­ave typically started late in proj- ect life and have often not been sustained after Pakistan needs to expand program coverage the project’s end. Demand-side efforts should and outreach for impact in high-burden dis- go hand in hand with supply-side efforts and tricts, ensuring convergent delivery of all sec- mechanisms for the sustainability of communi- tors’ interventions. Today, scattered program ty structures should be well supported. The con- initiatives do not cover entire districts, even in tinued role of Lady Health Workers as champi- the highest-burden areas. Outreach to the most ons of community-focused interventions should remote and underserved areas is essential.24 Of- be explored. fering incentives for nutrition tasks,25 building capacity, and more robustly integrating nutrition Easing major implementation indicators into the program monitoring system bottlenecks through the should all contribute to Lady Health Workers’ intervention of stewards and ability to deliver high-quality nutrition inter- leaders across a range of structures ventions, particularly counseling and behavior change communication. Lady Health Workers Pakistan needs to ensure timely decisionmaking have the potential to contribute to changing so- and problem resolution. Regular meetings cial norms around feeding and WASH and to in- of high-level decisionmaking bodies to make duce demand among community members for timely decisions that are crucial to program primary care and nutrition services.26 implementation­ —­such as appointing leaders and staff, ensuring leadership continuity, fill- It should prioritize the most cost-effective in- ing vacancies, initiating and moving procure- terventions and sectors to expand the coverage ment processes, ensuring adequate funding Chapter 4 | Pakistan Human Capital Review of direct nutrition interventions. Other sectors’ and smooth funds flow­ are important to move —­ contributions are important to improve nutri- programs. Stewards must hold themselves ac- tion outcomes but shifting resources from di- countable for discharging these functions, and rect nutrition interventions to other sectoral they must hold others accountable by setting interventions has to be carefully weighed. The and enforcing standards for these critical opera- importance of nutrition-specific investment tional tasks. Instituting mechanisms, such as ap- among the most vulnerable populations cannot pointing independent watchdogs and conduct- be stressed enough. ing self-audits, can also enhance accountability. It should ensure the geographic co-location of It should provide powers commensurate with re- sectoral interventions and their convergence at level sponsibility, especially to implementation-­ the household level. To dent stunting, it is critical bodies, and build their stewardship capacity. 123 Institutional structures can fulfill their responsi- nutrition messaging and BCC to improve di- bilities well only if they are empowered to exer- etary diversity. Every effort needs to be made cise their authority and use it appropriately. to build a BCC component into agriculture programs and devise suitable food fortifica- It should build technical and operational ca- tion policies to complement it. pacity across the sectors that contribute to • BCC. Behavior change is a critical yet com- nutrition. Because many sectoral stewards and plex aspect of improving nutrition outcomes. officials lack an understanding of nutrition and And at the design stage, BCC is a prominent of the most appropriate and cost-effective ways feature of most of the programs reviewed, their sectors could advance it, it is strongly sug- but BCC outcomes need to be improved. BCC gested that they be sensitized.28 materials from different programs should be assessed for their impact and viability. Criti- It should sharpen sectoral interventions to bet- cal time and resources go into creating BCC ter leverage their effect and increase efficien- materials, often leading to implementation cy through sharing common behavior change delays; therefore, materials from different communication (BCC) resources. programs should be centrally pooled then • Education. Given the strong association of adapted as needed. The target audiences mothers’ education with reduced stunting, should be considered when these materials one of the most important contributions are developed, in a way to provide simple, the education sector could make to improve consistent messaging. Materials for children nutrition is to focus on girls’ education and in particular should be age-appropriate and keeping girls in schools, which is also an ef- engaging. fective measure to delay the age of marriage and childbearing. Further, school-based nu- Strengthening stewardship trition programs have tremendous potential functions to reach vulnerable children. Training teach- ers in screening and nutrition messages is To inform and institute change, Pakistan should recommended, with active coordination with undertake an in-depth study of the three key the health department for further referrals as functions that have constrained effective nutri- necessary. tion stewardship: ensuring tools for implemen- • WASH. As an important contributor to ad- tation, creating a fit between policy objectives dressing diarrhea and other noncommuni- and organizational structure and culture, and cable diseases, WASH remains a priority area ensuring accountability (see table 4.3). The for intervention. However, there is a critical three are interrelated. Change on the margins need to evaluate WASH programming with will not bring about the transformation neces- Chapter 4 | Pakistan Human Capital Review a focus on its nutrition effects. A 2018 World sary to respond with the urgency required. An Bank WASH diagnostic recommends that in-depth and comprehensive study is strongly sanitation infrastructure be improved and recommended to examine the structure, func- better regulated; additionally, it is critical that tions, powers, capacities, and accountability these issues be adequately discussed and re- mechanisms­­ of the stewardship bodies and in- solved through multisectoral planning. 29 stitutional arrangements for nutrition. • Food, agriculture, livestock, fisheries. Most programs and interventions in this area are Bolstering monitoring, evaluation, concerned primarily with yields and growth and knowledge generation rates. Successful agriculture programs have emphasized the importance of food as a path Pakistan needs to develop systems to track to improved nutrition, with accompanying the multisectoral impacts of nutrition not only 124 to monitor progress and identify target pop- children receive adequate feeding, environmen- ulations for intervention, but also to advance tal health interventions, and health and nutrition multisectoral implementation. Monitoring and care. Thus, policies and programs to reach chil- evaluation is particularly weak in most program- dren under 2 with food, care, and environmental ming in Pakistan, and monitoring for nutrition health interventions will be critical to improve is no exception, despite moderately strong tar- nutrition outcomes. The COVID-19 pandemic get-setting for most nutrition projects. Although and the 2022 floods, which disrupted education there appears to be significant understanding and health systems including nutrition interven- within the government regarding the different tions, threaten to erase any gains made in re- programs and projects, little substantive doc- cent years. umentation has been developed at the govern- ment level. Most reviews of government pro- Based on the analysis in this chapter and the grams have been donor funded and limited to experiences of several other countries that have funded activities. So, there is a need for greater successfully reduced stunting (including Brazil, ownership of, capacity building for, and empha- Ethiopia, Nepal, Peru, Senegal, and Thailand), sis on monitoring and evaluation at the govern- the following are the main recommendations for ment level. tackling malnutrition in Pakistan. Pakistan should also form and revitalize bodies Prioritize cost-effective nutrition for knowledge exchanges, like the interprovincial interventions informed by Lady Health Worker Program forum, to facilitate evidence learning, especially in key areas where provincial • Raise the national profile of stunting, which approaches differ. Knowledge generation is crit- is a major human capital catastrophe requir- ical for informing additional programming and ing national and local efforts. Ensure political strategizing for challenging contexts. Program championing for stunting reduction at the na- data on previous implementation experiences tional and highest provincial levels. Strength- should also be recorded. en nutrition-focused development partners’ platforms at the national and provincial levels to improve cross-sectoral nutrition gover- CONCLUSION AND nance and coordination bodies and oversee RECOMMENDATIONS multisectoral nutrition actions. Systematize monitoring and evaluation for debottleneck- Malnutrition has long-term implications for a ing and course adjustments. nation’s human capital and growth. The per- • Remove existing constraints to program im- sistently high levels of malnutrition in Pakistan plementation, including long delays in approv- Chapter 4 | Pakistan Human Capital Review jeopardize human capital accumulation and als, frequent changes in leadership, procure- economic growth. The prevalence of stunting, ment delays, and funds flow issues. Ensure a measure of chronic malnutrition that has life- timely decisions and problem resolution. long impacts on physical growth and cognition, Provide powers commensurate with respon- remains at 40 percent. The slow pace of decline sibility, especially to implementation-level calls for major efforts to accelerate progress. bodies, and strengthen their stewardship The stunting situation varies across and within capacity. Build technical and operational ca- provinces, is higher in rural areas and in lower pacity across the sectors that contribute to wealth quintiles, but is not insignificant even in nutrition. the middle and upper wealth quintiles. The risk • Support “kangaroo” mother care to keep the factors that drive high rates of stunting remain newborn skin to skin with the mother. Re- rampant; fewer than 2  percent of Pakistan’s place iron and folic acid supplementation for 125 pregnant women with multiple micronutrient that shape healthy brain development, yield- supplementation. ing gains for future health, learning, and productivity. For example, healthy and well-­ Coordinate nutrition interventions nourished children learn better and earn and advocacy with those for early more as adults, and higher levels of educa- childhood development tion help improve both health outcomes and • Address constraints affecting the effective- access to better employment opportunities. ness of existing strategies. Ensure timely ap- • Identify priority districts to implement inter- proval and updating of strategies, particularly ventions based on hands-on global knowledge those that are cross-sectoral. Give more at- on convergence and sequencing of interven- tention to the sustainability of interventions. tions. Map out interventions in water, sanita- Assess and articulate financing needs and tion, and hygiene; social protection; educa- relative priorities. Improve communication. tion; nutrition; and maternal, neonatal, and • Integrate high-quality, contextually relevant child health that might be able to converge. parenting and behavioral change programs Empower district health authorities, and en- into health, social protection, and education sure that they have the capacity and resourc- platforms. Early interventions, including ac- es commensurate with their responsibilities cess to early childhood education, are vital to to implement nutrition and early childhood mitigate risks and promote protective factors development programs. Chapter 4 | Pakistan Human Capital Review 126 NOTES 1. National Nutrition Surveys. 18. WHO 2000. 2. Shekar et al. 2016. 19. See Exemplars in Global Health at https://www.ex- 3. Bhutta et al. How countries can reduce child stunt- emplars.health/. ing at scale: lessons from exemplar countries. Am J 20. Mejia Acosta and Fanzo 2012. Clin Nutr 2020;112(Suppl):894S–904S. 21. Demographic and Health Survey 2018, NNS 2018, 4. According to the Lancet Framework, nutrition-spe- and Multiple Indicator Cluster Survey 2018. cific interventions target the immediate determi- 22. Executive leadership, agenda-setting, and mes- nants of malnutrition—inadequate dietary intake saging all played key roles in Brazil, Peru, and Ethi- and ill health—and nutrition-sensitive interventions opia. Peru, for example, effectively mobilized pub- target the intermediate determinants. In this chap- lic awareness with its “5 by 5 by 5” program, with ter, unless otherwise specified, the term nutrition President Garcia’s public commitment to reduce interventions includes both nutrition-specific and chronic malnutrition rates in children under 5, by 5 nutrition-sensitive interventions. percent in 5 years. In Pakistan, the prime minister 5. Demographic and Health Survey 2017–18. already leads the Pakistan National Nutrition Coor- 6. NNS 2018. dination Council, demonstrating high-level political commitment. 7. Government of Pakistan n.d. 23. Zambia’s 1 percent Medical Levy suggests that there 8. Klassen and Murphy 2020. is potential for channeling tax revenue from mining 9. OCHA 2020; WFP 2020. companies to fund a nutrition strategy. See Mejia 10. In addition, statistics from the Ministry of National Acosta and Fanzo (2012). Health Services, Regulations, and Coordination in- 24. Nepal has demonstrated the key role community- dicate a decline in antenatal care visits and in the based workers can play in outreach to improve number of facility births during the COVID-19 wave, nutrition outcomes. Community-based workers in which could be due to disruption in services and the Pakistan—such as the Lady Health Workers and difficulty of traveling during a lockdown. Community Health Workers in the health sector— 11. For more information, see the Nutrition Internation- are among the mainstays for delivering nutrition al website at https://nutritionintl.org. services at the community level. More focused at- tention to nutrition in the training of Lady Health 12. Government of Pakistan n.d. Workers and of their supervisors is strongly recom- 13. Ijaz et al. 2020. mended, with updated training curricula providing 14. Ali et al. 2021. appropriate training modules and programs. 15. Environmental health refers to the availability of 25. J-PAL 2014. drinking water and sanitation facilities; care includes 26. TMIS proposes to support expansion of the Lady such factors as antenatal care, nutritional supple- Health Worker platform and investment in Lady mentation during pregnancy, and appropriate treat- Health Workers’ training and capacity building in 67 Chapter 4 | Pakistan Human Capital Review ment of childhood diarrhea. high-burden districts, with complementary provin- 16. Nutrition-specific interventions refer to interven- cial investments (Annex 4B.) tions that address the immediate determinants of 27. District nutrition committees established under foetal and child nutrition and development, whereas Sindh’s Accelerated Action Plan for Reduction of nutrition-sensitive interventions are those in com- Stunting and Malnutrition are a promising platform plementary sectors such as agriculture, health, so- through which to coordinate convergent actions and cial protection, early child development, education, targeting; however, experience from the plan also and water and sanitation that affect the underlying demonstrated the need for the health sector to take determinants of nutrition. the lead in village-level joint work. Ensuring integra- 17. In Sindh, the nutrition financing OPM was contract- tion of nutrition services into primary care service ed by the World Bank to Oxford Policy Management, delivery has the potential to have synergistic effects. and the information is taken from OPM (2017). Sindh has demonstrated good practice in using 127 health sector and BISP data to inform the rollout of 29. It found that open-defecation-free policies and pro- social protection and related interventions, and this gramming have overlooked waste management model can be built on to achieve the combined im- and toilet quality, and have not resulted in improved pacts of multisectoral nutrition interventions. health outcomes, and that interventions in such sec- 28. The example from Khyber Pakhtunkhwa, where sys- tors as agriculture and livestock overlook the WASH tematic sensitization of over 500 officials and joint impacts of their programming (World Bank 2018). planning across sectors was conducted, stands out. REFERENCES Ali, A., M. Sohaib, S. Iqbal, K. Hayat, A. U. Khan, and M. F. Ra- Maternal and Child Malnutrition: An Update of the Evi- sool. 2021. “Evaluation of COVID-19 Disease Awareness and dence.” Lancet Child and Adolescent Health 5 (5). Its Relation to Mental Health, Dietary Habits, and Physical Khan, N. 2019. “Reaching the Last Mile in Universal Salt Io- Activity: A Cross-sectional Study from Pakistan.” The Amer- dization in Pakistan.” Nutrition International, 4 November. ican Journal of Tropical Medicine and Hygiene 104 (5): 1687. https:/ /www.nutritionintl.org/news/all-blog-posts​/reaching​ Bhutta, Z., N. Akseer, E. Keats, T. Vaivada, S. Baker, S. -the​-last​-mile​-in​-universal​-salt​-iodization​-in​-pakistan/. Horton, J. Katz, P. Menon, E. Piwoz, M. Shekar, C. Victora, Klassen, S., and S. Murphy. 2020. “Equity as Both a and R. Black. 2020. “How Countries Can Reduce Child Means and an End: Lessons for Resilient Food Systems Stunting at Scale: Lessons from Exemplar Countries.” The from COVID-19.” World Development 136: 1–4. American Journal of Clinical Nutrition 112 (Supplement 2). Mejia Acosta, A., and J. Fanzo. 2012. “Fighting Maternal Government of Pakistan. 2020. “Poverty Alleviation and and Child Malnutrition: Analysing the Political and Institu- Social Safety Division Curtain Raiser, October 2020.” tional Determinants of Delivering a National Multisectoral Government of Pakistan. n.d. “Coronavirus in Pakistan.” Response in Six Countries.” Mimeo, Institute for Develop- http://covid.gov.pk/. Accessed 3 March 2022. ment Studies, Brighton, UK. Heaver, R., and Y. Kachondam. 2008. “Thailand’s National OCHA (United Nations Office for the Coordination of Hu- Nutrition Program­ Lessons in Management and Capac- —­ manitarian Affairs). 2020. “Pakistan Humanitarian Re- ity Development.” HNP Discussion Paper, World Bank, sponse Plan for COVID-19 Pandemic 2020.” Washington, DC. OPM (Oxford Policy Management). 2017. “Government Horton, S., M. Shekar, C. McDonald, A. Mahal, and J. K. of Sindh Nutrition Expenditure Review (FY 2017–18 to FY Brooks. 2010. Scaling Up Nutrition: What Will it Cost? Di- 2019–20).” rections in Development: Human Development. Washing- Ruel, M. T. 2008. “Addressing the Underlying Determi- ton, DC: World Bank. nants of Undernutrition: Example of Successful Integra- Ijaz, S., S. Ismail, L. Fayyaz, A. Ashraf, Y. Razzaq, and M. tion of Nutrition in Poverty Reduction and Agriculture Chapter 4 | Pakistan Human Capital Review A. Chaudhry. 2020. “Physical Activity, Screen Time and Strategies.” SCN News 36: 21–29. Emotional Wellbeing of Population during COVID-19 Pan- Shekar, M., J. Condo, M. A. Pate, and S. Nishtar. 2021. demic in Pakistan.” Biomedica 36 (4). “Maternal and Child Malnutrition: Progress Hinges on J-PAL (Abdul Latif Jameel Poverty Action Lab). 2014. “Pay Supporting Women and More Implementation Research.” for Performance Incentives for Healthcare Workers in The Lancet 397 (10282): 1329–1331. Pakistan.” Lahore, Pakistan: J-PAL. https:/ /www​.poverty​ Shekar, M., J. Kakietek, J. D. Eberwein, and D. Walters. action ​ l ab​ .org ​ /evaluation ​ /pay​ - perfomance ​ - incentives​ 2016. An Investment Framework for Nutrition: Reaching -heathcare​-workers-pakistan. the Global Targets for Stunting, Anemia, Breastfeeding, Keats, E., J. Das, R. Salam, Z. Lassi, A. Imdad, R. Black, and Wasting. Directions in Development: Human Devel- and Z. Bhutta. 2021. “Effective Interventions to Address opment. Washington, DC: World Bank. 128 Shekar, M., and B. Popkin (eds.). 2020. Obesity: Health WHO (World Health Organization). n.d. “Pakistan Inte- and Economic Consequences of an Impending Global grated Nutrition Strategy Operational Framework Plan.” Challenge. Human Development Perspectives series. Geneva: WHO. https://www.who.int/hac/crises/pak​/ Washington, DC: World Bank. pakistan​_​nutrition_strategy.pdf. SUN Movement Compendium 2014. Pakistan. World Bank. 2017. An Investment Framework for Nutrition. Washington, DC: World Bank. Tontisirin, K., and S. Gillespie. 1999. “Linking Communi- ty-Based Programs and Service Delivery for Improving World Bank. 2018. When Water Becomes a Hazard. Wash- Maternal and Child Nutrition.” Asian Development Review ington, DC: World Bank. 7 (1/2): 33–65. World Bank. 2020. “South Asia Economic Update.” Wash- Ullah, I., A. Pillai, S. Dey, S. Raza, and N. A. Khan. 2019. ington, DC: World Bank. https://pubdocs.worldbank.org​ “Pakistan’s Journey towards Universal Coverage of Io- /en/375781554478871276/mpo-pak.pdf. dised Salt: A Non-Systematic Review.” Pakistan Journal of World Bank. 2021. “Poverty and Equity Brief: Pakistan.” Public Health 9 (4): 195–197. Washington, DC: World Bank. https:/ /databank​.world​ UNICEF (United Nations Children’s Fund). 1998. The bank.org/data/download/poverty/987B9C90 ​ - CB9F​ State of the World’s Children 1998. New York: UNICEF. -4D93​-AE8C​-750588BF00QA​/AM2020​/Global​_POVEQ​ _PAK.pdf. WFP (World Food Programme). 2020. “Comprehen- sive Food Security and Livelihood Assessment (CFSLA) Zaidi, S., S. K. Mohmand, N. Hayat, A. M. Acosta, and Z. 2020.” Rome: WFP. https:/ /reliefweb.int/sites/reliefweb​ A. Bhutta. 2013. “Nutrition Policy in the Post-Devolution .int/files/resources/WFP-0000118584.pdf. Accessed 1 Context in Pakistan: An Analysis of Provincial Opportuni- October 2020. ties and Barriers.” IDS Bulletin 44 (3): 86–93. WHO (World Health Organization). 2000. World Health Report 2000: Health Systems: Improving Performance. Geneva: WHO. Chapter 4 | Pakistan Human Capital Review 129 Providing education to all children SUMMARY is challenging, but it can be done. Pakistan can 5 create supportive environments for increased Pakistan has one of the world’s highest rates retention by training teachers and school ad- of children out of school, with an estimated ministrators to identify children at risk of drop- 20.3 million of its 63.3  million school-age chil- ping out and provide in-school remediation for dren out of school.1 In addition, Pakistan’s learn- low-achieving children. It can also scale up the ing poverty rate­ —­the percentage of children more effective and efficient existing approach- unable to read and understand a short age-ap- es and expand them to schools in districts with propriate text by age 10­ —­at 75 percent before high dropout rates. And it can support house- the COVID-19 pandemic and the 2022 floods is holds in reducing dropouts by expanding the more than 16 percentage points above the aver- education conditional cash transfer (CCT) pro- age for South Asia and more than 19 percentage grams, increasing benefit amounts for second- points above the average for lower-middle-in- ary school children, and providing dedicated come countries. The high number of primary and free transport for secondary school girls and secondary school–age children who are and female teachers. not in school and the low quality of education­ —­about 65  percent of children perform below Bringing all children to school while ensuring minimum proficiency in reading­ —­explain Pa- a higher level of learning would cost at least kistan’s high learning poverty rate. So, simply 5.4 percent of GDP, up from the current invest- bringing all children to school will not be enough ment of 2.5  percent. This would require large to end Pakistan’s learning poverty and build its efficiency gains in access and improvements in human capital. quality. Employing business as usual to bring out-of-school children into school in each prov- Among the top barriers to children’s enrollment ince and improve the quality of education would and progression in school, the most common cost around 4.3 percent of GDP. are high cost, distance to schools, perceived poor quality of education, and shortage of teachers. These barriers are particularly rele- INTRODUCTION Chapter 5 | Pakistan Human Capital Review vant in rural schools and persist across educa- tion levels. Therefore, an important first step is With nearly one-third of its children out of school to recognize that out-of-school children are not and three-quarters of them trapped in learn- a homogenous group. Policies need to be tai- ing poverty, Pakistan is experiencing a human lored to the characteristics of distinct groups to development crisis. Although enrollment of maximize impact. The longer children are out of Pakistan’s children age 5–16 rose from 50 per- school, the less likely they are to go to school, cent in 1990 to 70  percent in the school year especially children who have dropped out. Tack- 2018/19, learning challenges persist through- ling dropouts requires targeted, aggressive, out the system, and barriers to enrollment and and innovative solutions. Bringing all children learning increase as children grow older. For ex- into school while also improving learning out- ample, while 85 percent of boys and 79 percent comes and overall education system efficiency of girls are enrolled at age 9, just 56 percent of 131 boys and 40 percent of girls remain in school by bring all children into school. This chapter de- age 16.2 Each year, therefore, Pakistan’s educa- velops a framework that, along with higher and tion system loses a huge number of students. better targeted funding, responds to the varying In 2018/19, 8  percent of students (15  million) needs of different groups of children to sustain- dropped out.3 Due to school closures and the ably reduce learning poverty and enable Paki- socioeconomic impact of the COVID-19 pan- stan to realize its human capital potential. demic and the 2022 floods on families, an esti- mated 2–3.5 million additional children will drop out of school in Pakistan, erasing years of enroll- CHILDREN OUT OF SCHOOL ment gains.4 The most recent estimate puts the number of Pakistan invests only 2.5 percent of its GDP in —­ out-of-school children at 20.3 million­ just over education, far less than the international median 11 million girls and 9 million boys (table 5.1).7 of 4.4 percent.5 The government invests roughly This estimate, based on data from Pakistan So- 10 percent of its annual budget in education­ —­ cial and Living Standards Measurement Survey half the international benchmark recommended (PSLM) 2019–20, is a substantial reduction by the Education 2030 Framework for Action.6 since the Alif Ailaan estimate of 25  million in Increasing funding and the efficiency of that 2014.8 To contextualize these numbers, this sec- funding remain key challenges for providing tion presents seven stylized facts about out-of- every child a quality education. school children in Pakistan. Little is known about the costs of bringing out- A majority of Pakistan’s out-of- of-school children to school, and how to do it school children are girls efficiently. This chapter develops back-of-the- envelope estimates of what it would cost to pro- During the 2019/20 school year, before the vide all children with a “life raft”: schooling at a onset of the COVID-19 pandemic, 37 percent of minimum quality level to achieve literacy and girls and 27 percent of boys age 5–16 were not develops a framework of interventions to bet- in school. At 10 percentage points, the gender ter cater to the needs of out-of-school children gap in enrollment has narrowed slightly, from in Pakistan. It begins with a set of stylized facts 13 percentage points in 2007 (figure 5.1). on the out-of-school population and then offers estimates of how much it will cost Pakistan to Girls’ school attendance is impeded by supply have all of its children in school under different and demand constraints.9 Obstructions on the assumptions, including the added cost of a high- supply side include distance to schools and er-quality education. Although the debate over low quality of schools (low-quality teaching and Chapter 5 | Pakistan Human Capital Review out-of-school children usually lumps children learning materials, lack of water and sanitation with different life stories and characteristics facilities and boundary walls, and shortage of under a single label, a single solution will not female teachers). Demand constraints include TABLE 5.1  Distribution of out-of-school children age 5–16 in Pakistan, by gender, 2019/20 Gender Population of children age 5–16 In school Out of school Percentage out of school Girls 30,135,715 19,075,908 11,059,807 37 Boys 33,187,662 24,127,430 9,060,232 27 Total 63,324,209 43,060,462 20,263,747 32 Source: World Bank calculations using data on age structure from Pakistan Census 2017 and on education participation from Pakistan Social and Living Standards Measurement Survey 2019–20. 132 FIGURE 5.1  More girls than boys are out of school Percent Male Female 50 47 40 43 40 39 37 36 36 34 30 30 27 27 25 25 25 20 10 0 2004–05 2006–07 2008–09 2010–11 2012–13 2014–15 2018–19 Source: World Bank calculations using data from Pakistan Social and Living Standards Measurement Surveys 2004–19. poverty, concerns about girls’ security to and several major barriers to girls’ education. These from school and work, and social norms and at- included poverty, Pashtunwali (tribal code), reli- titudes about girls’ education. gion, accessibility (transport), resources, short- age of female teachers, curriculum, and political The likelihood that a girl will not be in school in- apathy and corruption.12 Among these barriers, creases with age.10 Only 1 woman in 5 in Pakistan social norms favoring boys’ education over girls’ has completed her secondary education.11 The were most frequently mentioned. These findings dropout rate for girls rises from 34 percent in pri- are in line with those of the World Bank’s recent mary school to 73 percent in secondary school. “Barriers to Girls’ Education in Pakistan” and the In recent years, the cohort survival rate in pub- Human Rights Watch Report, “Shall I Feed my lic schools has fallen for girls but increased for Daughter, or Educate Her?”13 boys. For example, in 1996, 53 percent of boys and 63 percent of girls reached grade 5, where- Out-of-school children are more as in 2016, 60 percent of boys and 53 percent of likely to live in rural areas girls reached grade 5, a drop of 10 percentage points for girls against a gain of 7  percentage In 2018/19, 35 percent of Pakistan’s rural chil- points for boys. In 1996, 22 percent of boys and dren (15 million) age 5–16 were out of school, 33 percent of girls reached grade 10, whereas in compared with 20 percent (4.4 million) of urban Chapter 5 | Pakistan Human Capital Review 2016, 30 percent of boys and 29 percent of girls children (figure 5.2). This sizable gap in basic ed- reached grade 10, a drop of 4 percentage points ucation participation rates14 has remained large for girls and a gain of 8 percentage points for over the past two decades, varying only by about boys. 3 percentage points (from 15 to 18 percentage points). Growth in participation rates has been Little analytical work has been done on the low in both urban and rural areas, up 3 percent- impact of social norms on families’ schooling age points among urban children (77 percent in decisions for girls in Pakistan. In a qualitative 2007/08 to 80 percent in 2018/19) and 4 per- study of men’s perceptions of girls’ education centage points among rural children (61 percent in Khyber Pakhtunkhwa province, where girls’ in 2007/08 to 65 percent in 2018/19). In rural enrollment is among the lowest in the country, areas with low school density, increasing school Pashtun men of diverse backgrounds agreed on supply can reduce the share of out-of-school 133 FIGURE 5.2  Rural children are more likely to be out of school In 2019/20, these two provinces accounted for Basic school participation, by geographic location, 2007–19 almost 14 million of Pakistan’s 20.3 million out- 100 of-school children­—­69  percent. Although Pun- jab has seen the greatest drop in its share since Urban 1998 (down 15 percentage points), when it had 80 more than half of all out-of-school children, the Rural province still has the largest number (6.6  mil- 60 lion) and share (35 percent) (figure 5.3). 40 While Punjab shows that it is possible to make good progress in a fairly short time, challeng- 20 es remain there and in the rest of the country. Almost 62  percent of Pakistan’s out-of-school 0 children lived in the other three provinces in 2007–08 2010–11 2013–14 2015–16 2018–19 2019/20, most of them in Sindh (32  percent, 6.5 million) and Khyber Pakhtunkhwa (18.7 per- Source: World Bank calculations using data from Pakistan Social and Living Standards Mea- surement Surveys 2007–19. cent, 3.8 million), along with 10 percent (2 mil- Note: The education participation rate is the number of children attending any type of school lion) in Balochistan. Balochistan and Sindh have (public, private, madrassas, other) expressed as a percentage of the total number of children struggled to increase enrollment in step with in the 5–16 age group during a given school year. population growth. Between 1998 and 2017, the number of out-of-school children doubled in children and boost learning. For example, con- Balochistan, with almost 1.2  million more chil- struction of new schools in rural areas increased dren out of school, while the number rose by al- enrollment rates and learning outcomes for most 1 million in Sindh. all students, especially for girls, in Afghanistan and increased school attendance in Indonesia.15 Balochistan and Sindh have the However, the cost of providing access to quality highest rates of out-of-school schooling is higher in rural areas, where children children are dispersed across wide areas and many are not in school. The main factor limiting access to Higher shares of children are out of school in education for girls, but also for boys, is distance Balochistan (59  percent) and Sindh (42  per- to school, which points to the lack of adequate cent) than in the other two provinces (table 5.2). infrastructure in rural areas.16 In Pakistan, this Punjab’s rate is roughly one-third that of Baloch- is confounded by parental perceptions of inse- istan and half that of Sindh. If the policy objective curity for girls at school and on their way to and is to reduce the total number of out-of-school Chapter 5 | Pakistan Human Capital Review from school and work. children, focusing on Sindh and Khyber Pakh- tunkhwa is the right approach; if the objective is Among provinces, Punjab made to reduce prevalence of out-of-school children the greatest strides in reducing per province, the focus should be on Baloch- out-of-school children over the istan and Sindh, and to a lesser degree Khyber past two decades Pakhtunkhwa. Clarifying the objective matters for funding, planning, and targeting pedagogi- The number and share of out-of-school children cal approaches. Although this chapter argues differ across provinces and are highest in Pun- that all children should be able to attend school, jab, followed by Sindh. More than three-quarters there are important policy questions about how (76 percent) of the country’s population live in to start, what to focus on, and what the long- Punjab (53  percent) and Sindh (23  percent). term strategy should be. The answers matter for 134 equity, funding, implementation capacity, and, FIGURE 5.3  The distribution of out-of-school children by province especially important, education outcomes for changed between 1998 and 2017 girls and boys. Number of out-of-school children per province (millions) 25 Out-of-school children tend to be older 20 Sindh 5.1 6.0 More children are out of school at higher school Balochistan, 1.2 levels. During the 2018/19 school year, more 15 Punjab secondary school–age children (14–15) were 10.5 2.5 out of school (40 percent) than middle school– 10 6.6 age children (11–13; 25  percent, 3.6  million) or primary school–age children (6–10; 23 percent, 5 6.7 million). Between 2007 and 2019, participa- Khyber Pakhtunkhwa 3.6 4.0 tion rates rose slowly for children at each school 0 level. The largest drop-off in the percentage of 1998 2017 children enrolled in school occurred between Share of out-of-school children by province (percent) middle and secondary school, fluctuating be- tween 13 and 17 percentage points (figure 5.4). 100 Sindh 32.0 24.6 These facts are important for the design of pro- grams that can address the needs and charac- 80 Balochistan, 5.9 teristics of out-of-school children to reduce out- 13.5 of-school rates. It is important to pay particular 60 Punjab 50.5 attention to expanding access in middle schools, 35.2 where the number of education establishments 40 drops dramatically from primary levels. 20 Also important to the design of programs to re- Khyber Pakhtunkhwa 19.3 19.1 duce out-of-school rates are school completion 0 and repetition rates. Limited public and private 1998 2017 resources are not being used efficiently if chil- dren enroll but do not complete their studies Source: World Bank calculations using data on age structure from Pakistan Census 2017 and on education participation from Pakistan Social and Living Standards Measurement Survey 2017–18. (dropout) or take longer than expected to com- plete their studies (repetition). Increasing com- pletion rates and decreasing repetition rates Only one student in three entering the school Chapter 5 | Pakistan Human Capital Review and improving pedagogy must be part of any ef- system completes secondary school on time. fort to reduce the number of out-of-school chil- The cohort survival rate measures an education dren across Pakistan. system’s holding power and internal efficiency TABLE 5.2  Balochistan has the country’s highest concentration of out-of-school children, 2018/19 Khyber Statistic Punjab Pakhtunkhwa Sindh Balochistan School-age children (5–16) (% of national total) 51 19 23 7 Province rate of out-of-school children (5–16) (%) a 21 31 42 59 Source: World Bank calculations using data from Pakistan Social and Living Standards Measurement Survey 2018–19. a. The number of school-age children in the province who are not enrolled in school as a percentage of total school-age children. 135 FIGURE 5.4  The older the child, the less likely to be in school dropped out. Both boys and girls tend to com- School participation, by school level and age group, 2007–19 (percent) plete primary school, but then more girls drop 100 out in middle school, and more boys drop out in secondary school. Primary school age (6–10) 80 Dropout rates are higher among older children. Middle school age (11–13) One reason is that there are far more primary 60 Secondary school age (14–15) schools than secondary schools in Pakistan. With longer distances to travel to middle and 40 secondary schools, fewer children will make the trip. This is true particularly for girls be- 20 cause of parental concern for their safety, which seems, at least in part, to drive the decision to 0 keep them at home. Paired with prevailing gen- 2007/08 2010/11 2013/14 2015/16 2018/19 der norms and low labor market participation of women, girls face increasing barriers to edu- Source: World Bank calculations using data from Pakistan Social and Living Standards Mea- surement Surveys 2007–19. cation as they age. Well-targeted international initiatives have found that paying older children by following a group of students as they prog- to attend school can reduce dropout rates. In ress through the primary and secondary levels. South Africa, for example, expanding the Child In the 2007 cohort, of the 3.4  million children Support Grant raised teenage enrollment by enrolled in grade 1 in 2007/08, almost half had at least 10 percent among beneficiaries.17 This dropped out by grade 5 in 2011/12 (48 percent is important because demand-side initiatives, of girls and 46 percent of boys) (figure 5.5). By such as the Waseela-e-Taleem program are grade 10 in 2016/17, only 30 percent of children being expanded to middle and secondary edu- (0.9 million) remained in school, 70 percent had cation, with a premium for girls’ enrollment. FIGURE 5.5  Only 1 student in 3 entering the education system Many children never make it to finishes secondary school school Cohort survival rate (percent) 100 The proportion of children who have never at- tended school fell considerably in 2004–19, 80 particularly among girls, but the rate is still high. In 2004, 39 percent of girls and 26 percent of boys had never attended school. By 2019, these 60 Chapter 5 | Pakistan Human Capital Review numbers had dropped 10 percentage points for girls (to 29  percent) and 7  percentage points 40 Male for boys (to 19 percent) (figure 5.6). While en- Female couraging, this still means that nearly one girl 20 in three and one boy in five has never attended school. Dropout rates remained stable over the 0 same period (around 7 percent). In addition, the Class 1 Class 2 Class 3 Class 4 Class 5 Class 6 Class 7 Class 8 Class 9 Class 10 average age of children in first grade is 7, mean- (2007/08) (2008/09) (2009/10) (2010/11) (2011/12) (2012/13) (2013/14) (2014/15) (2015/16) (2016/17) ing that many children are already overage by Source: World Bank calculations using Pakistan Education Statistics 2007–17. Note: The cohort survival rate is calculated by dividing the total number of children originally first grade. Taken together, these four facts­ —­ enrolled in the first grade of primary school who reached each successive grade by the num- never attended, high dropouts, late entry, and ber of children in the same cohort and multiplying by 100. —­ repetition­ mean that the stock of out-of-school 136 FIGURE 5.6  Many children have never been to school, and large numbers drop out Percent Never attended school Male Female 50 40 30 20 10 0 2004 2006 2008 2010 2012 2014 2019 Dropped out of school Male Female 10 8 6 4 2 0 2004 2006 2008 2010 2012 2014 2019 Source: World Bank calculations using data from Pakistan Social and Living Standards Measurement Surveys 2004–19. children increases every year, implying an op- budgets. The responsibilities of the federal gov- portunity to improve the efficiency of education ernment for education consist largely of coordi- Chapter 5 | Pakistan Human Capital Review resource use. nation with provincial governments, the higher education system, and school education with- Total public investment in in areas under its direct jurisdiction. The total education matters for academic provincial budget for education in 2020/21 was achievement, and low investment PKR 831.18 billion (US$5,437.88 million),18 and levels and poor efficiency lead to the federal government’s education budget poor outcomes was PKR 15.52 billion (US$101.53 million), for a total annual expenditure per child of PKR 35,801 Most public investment decisions on school ed- (US$235). Per child spending varied widely by ucation are made and executed by the provinc- province, from PKR  27,839 (US$182) in Pun- es. In 2020/21, nearly all financial allocations jab to PKR  57,559 (US$377) in Balochistan. to education (98 percent) came from provincial These figures reflect the financing and planning 137 challenges in providing education in a country insecurity, conflict, and emergencies (including where efficiency levels are determined by the wars and pandemics). The importance of these varied management quality and geographic issues also differs across districts. characteristics of each province. Poor and rural children, especially girls, are the Increasing funding and spending efficiency re- most likely to be outside the education system. main important challenges. Education spending As children get older, the pressure to work, ei- per child differs considerably across districts, af- ther at home or in the labor market, increases, fecting years of schooling, school satisfaction, and the perceived low quality of education does and learning outcomes.19 With nearly 30 percent not incentivize them to continue in school. of children out of school, it is not surprising that Pakistan’s capacity to transform spending into Multivariable analysis pinpoints the top barri- outcomes is low. The results are similar when ers to children’s enrollment and progression in the outcomes are education quality and enroll- school. Analyses of PSLM Survey 2014–15 data ment in secondary education. show that the most common problem that par- ents identified in schools is the shortage of teach- In terms of efficiency, Pakistan is far from the ers, followed by the poor quality of education, the international frontier for education spending, distance to schools, and the high cost (figure as measured by the vertical distance between 5.7). These problems are particularly relevant in the country and the efficiency frontier, which is rural schools and persist across education levels. based on the outcomes (in this case gross pri- mary enrollment) that the most efficient coun- tries could achieve given their utilization of re- PROVIDING EDUCATION sources for a given amount of resources.20 TO ALL CHILDREN Six scenarios were developed to estimate what FACTORS AFFECTING it would cost Pakistan to bring and keep all of its SCHOOL ENROLLMENT children in school. The first three scenarios con- sider what it would cost to expand education to Following the framework developed in the World all children age 5–16, without addressing quality. Development Report 2018: Learning to Realize The scenarios consider reductions in the number Education’s Promise,21 this study identified fac- of out-of-school children while incorporating effi- tors in the school system (supply-side deter- ciency improvements in public investments and minants) and in households and communities alternative costing drawn from service delivery (demand-side determinants) that drive differ- mechanisms in the private sector, as well as the Chapter 5 | Pakistan Human Capital Review ences in education outcomes among children. differences in cost of providing education in each Supply-side constraints include limited supply province’s rural and urban sectors. These three of schools and poor quality of learning environ- scenarios are then estimated again, reflecting the ments, shortage and poor quality of teachers, cost to improve quality. All estimates are present- and weak leadership and school management ed as percentages of GDP to allow comparisons. practices, as well as structural factors, such as corruption and lack of enforcement of compul- Educating all children without sory education laws. Demand-side constraints improving quality include household characteristics (such as pov- erty, education attainment of the household The first three scenarios focus only on how head, and child labor); gender discrimination much it would cost to expand access to educa- and social norms (such as early marriage); and tion for all out-of-school children (table 5.3). 138 FIGURE 5.7   Why so many children do not go to school in Pakistan Rural schools Urban schools All schools School-related reasons Shortage of teachers School too far Poor quality of education Education is expensive Shortage of books Lack of toilet and water Other 0 10 20 30 40 Percent of responses Household-related reasons Too expensive Too far away Poor teaching Had to help at home Had to help with work Parents did not allow No female sta No male sta Child sick Child too young Child not willing Lack of documents Education not useful Education completed Chapter 5 | Pakistan Human Capital Review Marriage Service Other reasons 0 10 20 30 40 Percent of responses Source: World Bank calculations using data from Pakistan Social and Living Standards Measurement Survey 2014–15. 139 TABLE 5.3  The sticker price of having all children in school under different scenarios—lower bounds Scenario 1 Scenario 2 Scenario 3 Scenario 4 Scenario 5 Scenario 6 Current public BaU + BaU + expenditure in Big efficiency out‑of‑school high efficiency + primary and Business gains in public children at the BaU + improved quality Sticker price and secondary as usual education cost of low‑fee improved quality of for out-of-school Higher quality assumptions education (BaU) (25%) private schools public education children for all children Sticker price using national parameters (% of GDP) 2.5 4.4 3.7 3.3 4.9 3.8 5.4 Sticker price using provincial parameters (% of GDP) 2.1 3.8 3.1 2.9 4.3 3.4 4.8 Assumptions Average cost Bringing all Public schools Public schools Public schools absorb Public schools Public schools during 2017–19 out-of-school absorb absorb all out-of- all out-of-school retain current absorb all out-of- children 25 percent of school children at children and retain students at the school children into public out-of-school the average cost current students current unit cost and retain current schools at children at the of low-fee private with an investment of US$139.85 and students at the the average same current schools (US$107 in quality estimated absorb all out-of- current unit cost cost for the total cost and per child per year) to have a unit cost at school children plus the quality country of absorb the rest the midpoint of the at the unit cost premium calculated US$240 per of out-of-school difference between of highly efficient as the difference child per year children at the the cost of a low-fee and effective between the per- more efficient private school and nongovernmental unit cost of highly cost of US$202 that of highly efficient organization efficient and effective per child per and effective private schools of nongovernmental year schools of US$163 US$163 organization (such as those run by schools and low-fee the Citizen Foundation private schools of and others) US$240 + US$28) Source: World Bank calculations based on multiple data sources. These include annual budget statements, Pakistan Education Statistics, and Pakistan Social and Living Standards Measurement Survey data. • Scenario 1. Business as usual (BaU): The to provide the same quality of education public education system is expanded to in- for children already in school at the current clude all out-of-school children at the average public school cost per child of US$240 but cost of providing public education (US$240 absorbs all out-of-school children at the unit per child per year, calculated as national ed- cost of low-fee private schools of US$107.22 ucation expenditure divided by the number of This scenario is more of a thought exper- students in the system). iment to analyze cost and does not imply • Scenario 2. Big efficiency gains in public edu- endorsement. It could also be seen as an ef- cation: The underlying assumption is that the ficiency gain scenario, but because of fiscal public school system allocates resources inef- constraints for the public education sector ficiently. Using this assumption, the total per to provide education to all students at rough- Chapter 5 | Pakistan Human Capital Review student annual cost drops to US$202. The ly the same level of quality and at the same public school system becomes 25  percent cost, this scenario is better considered as more efficient, providing education to 25 per- using the private sector to provide education cent of out-of-school children with no addi- to all out-of-school children. This is not unre- tion to total expenditure and incorporates all alistic as the experience of Punjab has shown other out-of-school children into the system in mobilizing the private sector to accompa- at the more efficient average cost per stu- ny the efforts of the public sector. dent per year of US$202. This cost provides a benchmark for increased efficiency. Employing BaU to bring all out-of-school chil- • Scenario 3. BaU + out-of-school children en- dren into school is expensive and would require rolled at the average cost of low-fee private boosting annual education investment from schools: The public school system continues percent to 4.4  2.5  percent of GDP (see table 140 5.3, scenario 1)­—­a cost virtually identical to The estimates do not capture other factors, in- the global median. However, bringing all out-of- cluding that many out-of-school children are in school children into school at the prevailing unit parts of the country where the average cost of cost would be an expensive and inefficient way reaching them is higher than the unit cost used of expanding a system that is already weak in in these calculations. These first three scenari- enrollment and retention. os also assume that functioning schools already exist where they are needed (public, private, or A big push to enhance public sector efficien- both) or that provincial education systems take cy by 25 percent (see table 5.3, scenario 2) to bold measures to increase access, perhaps by educate out-of-school children (see table 5.3, maximizing school infrastructure through dou- scenario 3) would cost around 3.5  percent of ble shifts, shorter school days, or other mea- GDP, a 1  percentage point increase from the sures that would allow the current supply to re- current cost. This implies that under prevailing spond to the increased demand. Moreover, the fiscal constraints expanding access to all out-of- large financial push needed for the first three school children would be suitable only in scenar- scenarios would result in limited reductions ios that enhance the efficiency of money allocat- in learning poverty as they do not include im- ed within the system. provements in education quality (discussed just below). Under the first three scenarios, learning Improved targeting of pockets of out-of-school poverty would fall modestly from 75 percent to children at provincial and district levels to inform 68 percent, despite substantial increases in the evidence-based budgeting, using relevant met- education budget. rics and results, will be a critical first step toward a more efficient public education system. Current Educating all children while budgeting processes do little to take into account improving quality local conditions related to out-of-school children. Mapping targeted pockets of out-of-school chil- The second three scenarios consider how much dren in all provinces would align the budgeting it would cost Pakistan to bring all children to process with local needs and result in more effi- school while ensuring a higher level of learning cient allocation of money against targets, leading (see table 5.3). to better enrollment results at lower cost. • Scenario 4. BaU + improved quality of public education: All out-of-school children are ab- Employing BaU to bring all out-of-school children sorbed into the public school system at the into school in each province would require boost- average annual cost per student (US$240) ing annual education investment from 2.5  per- plus a premium of US$28 per child per year cent of GDP to 3.8  percent of GDP; enhancing to improve the quality of public education. Chapter 5 | Pakistan Human Capital Review efficiency by 25 percent or using low-fee private This premium is calculated as the midpoint schools would cost around 3  percent of GDP. of the difference between the cost of a low- Provinces operate at different marginal-cost effi- fee private school (US$107) and the cost of ciencies. The amount of additional budget alloca- highly efficient and effective schools run by tions required is greatly reduced by accounting nongovernmental organizations such as the for budgets, out-of-school children, enrollment, Citizens Foundation (US$163). and unit cost estimates at the provincial level. • Scenario 5. BaU + high efficiency + improved This shows that targeting at the subnational level quality for out-of-school children: Students in is critical for greater effectiveness and efficiency. public schools remain at the current cost level (US$240 per child per year), and all out-of- These estimates are indicative only and should school children are incorporated into the pub- be interpreted cautiously as lower bounds. lic school system at the cost associated with 141 schools run by a highly efficient and effective 2.5 percent). This scenario would require large nongovernmental organization (US$163). efficiency gains in access and improvements in • Scenario 6. Higher quality for all children: quality. Employing BaU to bring out-of-school Students currently attending public schools children into school in each province and im- and all out-of-school children receive an edu- prove the quality of education would cost cation at the current unit cost plus an added around 4.8 percent of GDP for scenario 6. This amount calculated as the difference between is 0.6 percentage point lower than a national low-fee private schools and schools run by a estimate that does not take different provincial highly effective nongovernmental organiza- parameters into account. Similarly, using pro- tion (US$28). vincial parameters would reduce the cost of sce- nario 4 to 4.3 percent of GDP and of scenario 5 Under the BAU + improved quality scenario 4, to 3.4 percent of the GDP. costs rise by at least an additional 0.5 percent- age point of GDP over the simple BaU case. Improving education quality, Bringing all children into school under the BaU not just expanding access, is scenario with an annual per-child quality pre- essential mium raises needed investment in education to 4.9 percent of GDP, up from 2.5 percent (see Just bringing all children to school but not im- table 5.3, scenario 4). The additional cost to im- proving quality will involve significant financial prove quality is lower than the cost to provide cost, but learning poverty would not decline very access for all children. much because of the poor quality of current ed- ucation. Reductions in learning poverty would In the unlikely scenario 5 of BaU + high efficiency be limited by the high proportion of students + improved quality for all out-of-school children, unable to achieve minimum proficiency in read- total expenditure would be 3.8 percent of GDP. ing, which is currently 65 percent in primary ed- ucation. Under the BaU scenario, investment in Under scenario 6, the cost would rise to at education goes from 2.5  percent to 4  percent least 5.4 percent of GDP (up from the current of GDP, reducing out-of-school children to, say, 1 percent, and learning poverty from 75 percent FIGURE 5.8  Improving both access and quality reduces learning to 65 percent. poverty faster Learning poverty The cost of reducing learning poverty by 10 per- 0.8 centage points could be offset by efficiency gains in public education. How much learning With xed quality poverty could potentially be decreased is af- Chapter 5 | Pakistan Human Capital Review 0.7 fected by the reading proficiency that students can achieve. For example, reducing the per- Minimum pro ciency level With improving quality centage of students performing below the min- 0.6 (business-as-usual scenario) imum proficiency level in reading by 10 percent- age points can reduce learning poverty faster Minimum pro ciency level (improved) than bringing more out-of-school children into 0.5 school (figure 5.8). In short, bringing all chil- dren into school will not greatly reduce learning 0.4 poverty because of the low education quality in 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 Pakistan. Investments in reducing the number Percentage of out-of-school children in primary education of out-of-school children must be accompa- Source: World Bank calculations using data from simulation in table 5.3. nied by investments in improving the quality of 142 education to achieve meaningful reductions in GIVING CHILDREN A learning poverty. BETTER EDUCATION The COVID-19 pandemic is This section identifies time-bound priority poli- undermining education gains cies and outlines a framework for them. The pol- icies are designed to support current students The COVID-19 pandemic has likely increased the in continuing their education and acquisition of number of out-of-school children and worsened —o life skills­ ­ ffering “stepping stones” along their learning poverty in Pakistan, increasing the cost education path; and to focus on different groups of bringing all children into school and improving of out-of-school children­—­providing “bridges” the quality of education. As part of a nationwide back into school. These policies are encapsu- lockdown beginning in March 2020, Pakistan lated in a framework to sustainably reduce the closed all of its schools, which will have a pro- number of out-of-school children and learning found effect on the education system, further poverty more widely. The recommendations eroding already low learning levels. In particular, below are grounded not only in the realities and it is estimated that 1–2.5 million additional chil- experience in Pakistan but also in internation- dren will drop out of school, learning-­ adjusted al evidence as per the Recommendations of years of schooling will decline by one-third to the Global Education Evidence Advisory Panel two-thirds of a school year, and learning poverty reports.25 will rise to at least 79 percent.23 Building stepping stones and Disruption of education during the COVID-19 bridges pandemic will disproportionately affect disad- vantaged and hard-to-reach children, including Pakistan needs to educate all its children to a girls and young women. Education inequities level that promises them a more productive fu- are likely to increase due to unequal access ture. An important first step is to recognize that to remote learning technology. The full depth out-of-school children are not a uniform group. and scope of the impacts of the pandemic will Policies need to be tailored to the characteris- depend on its duration, its seasonal charac- tics of their different groups to increase impact. teristics, government and education author- Bringing out-of-school children into education ities’ resolve, and the financial and technical or training is a huge challenge. Research sug- response to mitigate its effects. Incorporating gests that the longer children are out of school, the estimated impacts of the COVID-19 pan- the less likely they are to go to school, especially demic on enrollment is likely to substantially children who have dropped out.26 Tackling drop- raise the cost of enhancing the education sys- out requires targeted, aggressive, and innova- Chapter 5 | Pakistan Human Capital Review tem to include more children and provide them tive approaches. Bringing all children into school with a better education. The increases reflect while also improving learning outcomes and primarily the additional learning losses, which overall education system efficiency is challeng- will increase the number of students who do not ing, but it can be done. achieve minimum proficiency in reading, and, to a lesser extent, the increase in the number of Stepping stone 1. Create supportive out-of-school children. Preliminary data show schools for increased retention learning losses in grades 1–5 in math and lan- • Focus on foundational skills, in particular guage in rural Pakistan due to school closures.24 early literacy, numeracy, and socioemotional Data for the whole country on both dropouts skills. Learning other subjects or skills is and learning losses has just started to be col- harder if students are still struggling to un- lected and analyzed. derstand what they read, and communicate 143 appropriately what they know, think, and can International evidence is clear that teachers do. Conduct a rapid assessment of all early are the key to student learning outcomes27 grade reading support initiatives in the coun- and that strengthening their ability to connect try and develop training materials for care- with at-risk students is crucial for decreasing givers of young children, using existing early dropout rates.28 Low-performing students are childhood education curriculum and mate- often the first to drop out of school.29 Offering rials (see chapter 2). The US Agency for In- in-school remediation to lagging students in ternational Development, the United Nations districts with the lowest primary–secondary Children’s Fund, and various nongovernmen- school transition rates will reduce dropout tal organizations have offered early childhood hot spots. Additional resources, in the form of education programs throughout Pakistan for teachers and teaching and learning material, years, and provide important resources and may be needed to support these efforts. experience with what works, including struc- tured lessons plans with linked materials, and Stepping stone 2. Support households in some monitoring of and support to teachers. reducing dropouts • Scale the more effective and efficient ap- • Continue the expansion and improve the im- proaches and expand them to schools in plementation and verification of the educa- districts with high dropout rates. Implement tion CCT program, increase benefit amounts programs in the most disadvantaged districts for secondary school children, and provide of the country first, to increase school readi- dedicated and free transport for middle and ness among the most vulnerable. secondary school girls and female teachers. • Plan for the systematic introduction of formal The transition to secondary school is a key preschool classrooms with minimum quality attendance drop-off point for boys and girls standards for later phase-in. Hiring women in Pakistan. Providing monetary support and in the community as caregivers and training school transport resources directly to fam- them in early childhood education can in- ilies of secondary school–age children can crease community buy-in while providing in- ease these concerns and stem secondary come and social status for women. school dropout rates.30 • To support early literacy, establish school- • Reduce time to school by refurbishing exist- based libraries and read-at-home kits (for ing spaces, create and support community example, Read@Home) that include low-cost schools (with local teachers), and provide story and chapter books in local languages. materials that have shown to be impactful. Prioritizing early childhood education, with a Expanding public–private partnerships to focus on early literacy, through these imme- middle school and across the country, with diate actions will prepare Pakistan’s youngest strong accountability mechanisms for private Chapter 5 | Pakistan Human Capital Review learners to enter school on time and ready to schools and ensuring teachers’ conditions, learn, which can boost enrollment rates and can help improve supply of education serv- overall education system efficiency. ices wherever public provision is inadequate. • Train teachers and school leaders to iden- • Improve the overall availability and targeting tify children at risk of dropping out and pro- of CCT programs.31 CCTs are most effective vide in-school remediation for low-achieving when payments are directly linked to desired primary school children. To better equip behavior change,32 for example, predicated teachers to support all students, pre- and on children’s enrollment and regular atten- in-service training curricula should instruct dance at school.33 Expanding the CCT to offer teachers how to identify at-risk children and nutritious meals and fresh produce in the common learning disorders (such as dyslex- most disadvantaged villages is a highly effec- ia) and inform them about available support. tive means of ensuring that children have the 144 energy to show up in school and learn. Com- Set up computer centers that children can ac- munity members can be employed to help cess, where they are able to learn in a safe en- with food preparation, including local farmers vironment. Local area networks can be set up to provide the food and local women to pre- to keep children safe within the digital space. pare it.34 For example, Idara e-Taleem o-Aagahi’s Chalo Parho Barho (Let’s Read and Grow) program Bridge 1. Undertake remedial actions for shows learning gains for children in under- recent dropouts performing districts of Sindh, Punjab, and • Provide additional learning support for stu- Balochistan. The program covers grades 2–5 dents who are re-enrolling after a moderate and is open to children in school and out of period of being out of school (less than two school. It enables children out of school to years). Support can be provided after school, enroll (or re-enroll) in an appropriate grade through one-on-one or small group tutoring and it helps ensure that children at risk of during the school day in parallel with regu- dropping out of school (because their learn- lar classes, and through bridging programs ing is not progressing) stay in school.36 held when school is not in session (summer • Introduce basic literacy and numeracy pro- school). grams, coupled with skills development pro- • Partner with school committees and teach- grams, to prepare older children who have ers to identify recent dropouts and support never attended school or are longer-term their return to school. Identify common rea- dropouts for integration into the labor mar- sons for dropout (for example, for girls living ket. Partner with local community organiza- in rural areas it might be a lack of transport to tions and potential employers to meet the school or lack of latrines in the school). Next­ needs in the locality. For example, in rural —­and this is key­ —­give school committees agricultural areas, focus on agricultural skills the authority and resources to remediate (such as no-till farming and agri-marketing). identified issues, with teacher input, perhaps In urban centers, partner with humanitarian through teacher representatives appointed organizations (such as the Edhi Foundation) to the school committees. For example, offer to provide safe living conditions for street and oversee the use of dedicated grants for children, with accelerated learning opportu- two years for students returning to school, nities, including life skills. predicated on consistent attendance. Re-­ • Build schools in areas where difficult or no enrollment campaigns and working with all access to secondary school is the key reason stakeholders in the community to send chil- for dropout. Couple it with the CCTs and a dren to school can pay off as it did in Punjab program on safe transportation services for with SMS messaging are cost-effective ways girls and female teachers. Chapter 5 | Pakistan Human Capital Review of enhancing re-enrollment, particularly after • Hire female teachers to increase enrollment the school closures caused by the COVID-19 and attendance among adolescent girls. pandemic.35 Reducing the number of out- Bridge 2. Offer alternative learning of-school children and learning pathways for children who have never been poverty more widely to school and for long-term dropouts • Provide accelerated-learning programs for The framework proposed below identifies three children who have never been to school or types of out-of-school children (never in school, who dropped out. For example, offer digital dropped out for less than two years, dropped learning options to educate youth who have out for more than two years) and in-school chil- some reading skills, primarily in urban areas. dren who are falling behind and recommends 145 policy packages targeted to the specific needs that includes student materials and teacher of each group (table 5.4). Programs are need- lesson plans (including scripted), training, and ed not only to bring out-of-school children into ongoing support, which are shown to be effec- school, but also to prevent more children from tive in international studies (Annex 5). Student being out of school. For each group of out-of- report cards paired with parent–teacher confer- school students, proper targeting (by age, gen- ences can also improve student learning. Struc- der, household socioeconomic status, urban– tured pedagogy and school report cards appear rural location, and so on) is fundamental for to be cost-effective means of improving student devising and implementing solutions that will learning once children are in school. increase learning. Older children who are in school may benefit For children of all ages who are already in from blended-learning options that pair face-to- school, structured pedagogy and report cards face classroom learning with online curriculum and parent–teacher conferences can help keep components. These programs work well when them there. Pakistan might consider offering its delivered to students through structured online teachers access to structured pedagogy, a coor- and in-class settings by teachers trained to fa- dinated approach to teaching to the right level cilitate this interaction and when they include TABLE 5.4  Matrix to guide sustainably reducing out-of-school children and learning poverty Proposed policies, by age Age 5–9 Age 10–12 Age 13 and older Out of school Never in school School outreach to family School outreach to family + in‑school Accelerated skills program + in‑school remediation remediation + bridge program (e.g., + technical and vocational education + multigrade classrooms summer school) or accelerated skills program Students who dropped out Less than one year ago School outreach to family School outreach to family (identify dropout School outreach to family (identify dropout reasons) reasons) (identify dropout reasons) + in‑school remediation + in‑school remediation + in‑school remediation One to two years ago School outreach to family School outreach to family (identify dropout School outreach to family (identify dropout reasons) reasons) (identify dropout reasons) + at-home reading materials + in-school remediation or accelerated + in-school remediation or + in‑school remediation skills program accelerated skills program More than School outreach to family School outreach to family Accelerated skills program + two years ago + at-home reading materials + in-school remediation technical and vocational education + in‑school remediation + bridge program (e.g., summer school) or Chapter 5 | Pakistan Human Capital Review accelerated skills program At-risk students in school Started on time Structured pedagogy Teacher training to identify at-risk students Teacher training to identify at-risk + structured pedagogy + school report students cards/parent–teacher meetings + structured pedagogy + school report cards/parent–teacher meetings Overage by Multigrade classrooms Structured pedagogy Structured pedagogy one to two years + structured pedagogy + blended learning options + blended learning options + at‑home reading materials Overage by Multigrade classrooms Structured pedagogy Structured pedagogy two or more years + structured pedagogy + blended learning options + blended learning options + at‑home reading materials Source: Developed based on insights from Evans and Popova (2016), Evans and Yuan (2019), and World Bank (2018, 2020e). 146 some student control over time, place, path, different needs, and policies need to be tai- or pace. With blended learning, classroom and lored to maximize impact. The youngest out- online experiences are tailored to reinforce one of-school children (age 5–9), around half of another.37 the total (9.5 million), have the most straight- forward needs: access to schools, a caring For children who dropped out less than two teacher in front of the classroom, and perhaps a years ago, in-school remediation can support nudge to parents from the community to enroll curricular catch-up, including one-on-one and their children in school. But children who are small-group tutoring. One-on-one and small- in school and at risk of dropping out (typically group support are proven catalysts for acceler- age 8–16) often require additional learning sup- ated learning. Tutoring has the greatest impact port, to make sure that parents see the value of on reading abilities in the early years (especially keeping their child in school. As the labor and in kindergarten and first grade) and on math in marriage markets start pulling older children later grades. Tutoring conducted during school away from school, parents may need incen- hours is more effective than tutoring after tives that reduce their cost of sending children school and tutoring by teachers or paraprofes- to school, such as stipends. Older children who sionals is more effective than tutoring by volun- have never been to school require specific liter- teers or parents.38 acy interventions and short-term skills training to help them become literate and obtain job-­ For children who have been out of school for related skills. more than two years, bridge programs may be necessary to help them catch up on the cur- Based on the analysis in this chapter and global riculum. Key considerations for policymakers best practices, policy and programming recom- include the need to plan early, provide pro- mendations include: gram and staffing continuity from year to year, and integrate summer teaching with staff Expand the supply of safe schools so that development.39 every child has a guaranteed seat in school • Prioritize public sector provision and public–­ For older children and youth who have never private partnerships. Strong controls on pro- been to school, accelerated skills programs fo- viders to guarantee school outcomes, along cusing on functional literacy and numeracy, and with the safety and welfare of students and on professional and life skills training, are impor- teachers, are paramount to get and keep chil- tant for developing basic human capital. Training dren in school. Clean water and basic WASH programs such as the Jóvenes programs in Latin facilities, including safe bathrooms, must be America have demonstrated success in reach- present in all schools. Chapter 5 | Pakistan Human Capital Review ing vulnerable youth and in improving their for- • Rehabilitate and build schools where they mal employment opportunities and earnings.40 are needed. School and classroom con- If needed, accelerated skills programs could be struction is particularly important in areas followed by more formal, longer-term technical where difficult or no access to school is a and vocational education and training options. key reason for dropout. This has been suc- cessful in parts of Balochistan, where aban- doned public buildings have been adapted CONCLUSION AND and made adequate for service provision of RECOMMENDATIONS education. • Hire more qualified (especially female) teach- The population of out-of-school children broad- ers based on merit to increase enrollment and ly comprises three distinct groups, each with attendance among adolescent girls. Like girls, 147 female teachers may require safe and dedi- Prioritize literacy to increase retention, and cated transport to reach schools, particular- enable human capital accumulation ly in remote areas. Parents often see female • Develop literacy training materials for care- teachers as an indicator of security for young givers and teachers of young children, using girls. existing curriculum and materials. • Conduct a rapid assessment of all reading ini- Support households to increase enrollment tiatives across Pakistan. and reduce dropouts • Draw on existing initiatives. Government, • Expand the education CCT program to both nongovernmental organizations, and devel- boys and girls and provide dedicated and opment partners have offered early reading free transport for secondary school girls. programs throughout Pakistan for years that Direct provision of monetary support and can provide important resources and imple- safe school transport can ease families’ mentation experiences. concerns, increasing enrollment and reduc- • Direct support to teachers. Structured peda- ing dropouts. CCTs are most effective when gogy and training in teaching at the right level payments are directly linked to children’s en- have shown positive impacts on children’s lit- rollment and regular attendance at school. In eracy levels. the most disadvantaged districts, adding a • Train teachers and school leaders to identify premium for food or food vouchers can off- children at risk of dropping out and provide in- set families’ need for children to participate school remediation centered on foundational in household or paid work, freeing resources learning and numeracy. Low-performing stu- for the education required to build human dents are often the first to drop out of school. capital. International evidence is clear that teachers • Raise awareness about the importance of are key to student learning outcomes and education, particularly for girls. Commu- strengthening their ability to connect with at- nity approaches that bring parents, and risk students is crucial for decreasing drop- other household and community members out rates. To better equip teachers to support into advocating for and facilitating girls’ ac- students, preservice and in-service training cess to education and security from home curricula should instruct teachers how to to school can increase education demand. identify at-risk children and common learn- More research is needed on measuring ing disorders (such as dyslexia), informing and shifting social norms around girls’ ed- them of available support. ucation. This work includes sharing infor- • Introduce basic literacy and numeracy pro- mation on returns to education and the grams, coupled with skills development pro- value of education for all for human capital grams, to prepare older children who have Chapter 5 | Pakistan Human Capital Review development. never attended school or are longer-term dropouts for integration into the labor mar- ket. Partner with local community organiza- tions and potential employers to meet the needs in the locality. 148 NOTES 1. Estimates of children out of school range between 20. Herrera and Ouedraogo (2018) used both Free Dis- 18.7 million and 20.7 million depending on different posable Hull and Data Envelopment Analysis meth- sources of data and different simulation exercises. ods in their analysis. The estimates used here are based on census data 21. World Bank 2018. from 2017. 22. Information from Punjab suggests that low-fee pri- 2. PSLM 2018–2019. vate schools provide roughly the same level of edu- 3. PSLM 2018–2019. cation quality as public schools but at a much lower 4. This estimation comes from information from a cost per student. This is a thought exercise. Paki- phone survey carried out by the World Bank in Janu- stan’s current private school sector cannot achieve ary 2021. this at today’s human and physical capacity. It would need to expand. In essence, neither public nor pri- 5. UNESCO 2019. vate schools, or a combination of the two, have the 6. The Education 2030 Framework for Action proposes ability to absorb all of Pakistan’s out-of-school chil- two benchmarks as crucial reference points: allo- dren. More capacity will need to be created. cate at least 4–6 percent of GDP to education and/ 23. World Bank 2020c. or allocate at least 15–20 percent of public expendi- ture to education. 24. ITA 2021. 7. This is triangulated with the age cohort findings of 25. Word Bank 2020d, 2022e. the national population census conducted in 2017. 26. OECD 2020. The results are based on the population age 5–16. 27. See, for instance, Hanushek and Rivkin (2006). 8. Alif Ailaan 2014. Alif Ailaan was a national education 28. Dinu 2015. campaign in Pakistan that worked on mainstream- 29. Rumberger 2001. ing education data to raise education quality and de- crease the incidence of out-of-school children with 30. Large distances significantly limit enrollment and support from the UK Foreign, Commonwealth and attendance, particularly for girls (Theunynck and Development Office. Rabakoson 2017). In a review of 58 randomized con- trolled trial studies of presecondary schools in 28 9. Trako et al. 2020. low- and middle-income countries in Africa, Asia, 10. Pakistan Education Statistics 2016–2017. and Latin America, J-PAL (2017, Roll Call: Getting 11. UNESCO Institute for Statistics 2020. Children into School. p. 11) found that “when school 12. Jamal (2016) led participants through a two-round is far away, reducing travel time can help boost par- Delphi exercise, followed by in-depth qualitative ticipation. This can be particularly important for interviews. girls and in areas where security is an issue.” 13. Human Rights Watch 2018; Trako et al. 2020. 31. Evans and Yuan 2019. 14. The education participation rate is the number of 32. In Malawi, conditional and unconditional cash transfers to girls who had dropped out of school Chapter 5 | Pakistan Human Capital Review children attending any type of school (government, private, madrassas, other) expressed as a percent- raised enrollment and English reading comprehen- age of the total number of children in the 5–16 age sion. Enrollment effects were almost twice as large group during a given school year. among CCT recipients as among unconditional cash transfer recipients. Reading comprehension also in- 15. Burde and Linden 2013; Duflo 2001. creased more in the CCT group (Baird et al. 2016). 16. Human Rights Watch 2018; Trako et al. 2020. 33. Baird et al. 2016. 17. Eyal, Woolard, and Burns 2014. 34. Drake et al. 2016; WFP 2020. 18. Numbers in this paragraph are based on April 20, 35. Hasan et al. 2021. 2021, exchange rates. 36. https://www.itacec.org/CPB​-(Accelerated​-Learning​ 19. World Bank 2019a. -Programme). 149 37. Horn and Staker 2011, 2012. statistics, socioeconomic data, and poverty mapping. 38. Nickow, Oreopoulos, and Quan 2020. Qualified private firms, nongovernmental organiza- tions, public institutions, and training agencies then 39. Harris 2001. provide training on a competitive basis. Before re- 40. The Jóvenes programs offer poor youth training in ceiving government funds for training, providers are professional and life skills, followed by workplace in- required to arrange internships for trainees and to ternships. Based on a pilot in Chile in the early 1990s, ascertain what kinds of skills local employers need. this comprehensive approach to training has spread Life skills training focuses on problem-solving skills, throughout Latin America, with countries tailoring the workplace behavior, conflict management, job search program to their needs. Usually, disadvantaged youth techniques, and self-esteem (González-Velosa, Ripan, are identified through out-of-work and out-of-school and Shady 2012; Ibarrarán and Shady 2009). REFERENCES Adukia, A. 2017. “Sanitation and Education.” American Eco- Behrman, J., S. Khan, D. Ross, and R. Sabot. 1997. “School nomic Journal: Applied Economics 9 (2): 23–59. https:/ / Quality and Cognitive Achievement Production: A Case www​.aeaweb​.org​/articles​?id=10.1257/app​.20150083. Study for Rural Pakistan.” Economics of Education Review 16 (2): 127–142. Aikman, S., and P. Pridmore. 2001. “Multigrade Schooling in ‘Remote’ Areas of Vietnam.” International Journal of Ed- Benhassine, N., F. Devoto, E. Duflo, P. Dupas, and V. ucational Development 21 (6): 521–536. Pouliquen. 2015. “Turning a Shove into a Nudge? A ‘La- beled Cash Transfer’ for Education.” American Economic Alif Ailaan. 2014. “25 Million Broken Promises: The Crisis of Journal: Economic Policy 7: 86–125. https:/ /www.nber​ Pakistan’s Out-of-school Children.” Islamabad: Alif Ailaan. .org/papers/w19227. Andrabi, T., J. Das, and A. Khwaja. 2013. “Students Today, Bruns, B., and J. Luque. 2014. Great Teachers: How to Teachers Tomorrow: Identifying Constraints on the Provi- Raise Student Learning in Latin America and the Caribbe- sion of Education.” Journal of Public Economics 100 (1): an. Washington, DC: World Bank. 1–14. Burde, D., and L. Linden. 2013. “Bringing Education to Andrabi, T., J. Das, and A. Khwaja. 2017. “Report Cards: Afghan Girls: A Randomized Controlled Trial of Village-­ The Impact of Providing School and Child Test Scores on based Schools.” American Economic Journal: Applied Educational Markets.” American Economic Review 107 Economics 5: 27–40. https:/ /www.aeaweb.org/articles​ (6): 1535–63. https:/ /www.aeaweb.org/articles?id=10​ ?id=​10.1257​/app.5.3.27. .1257/aer.20140774. Charnes, A., W. W. Cooper, and E. Rhodes. 1978. “Measur- Baird, S., E. Chirwa, J. De Hoop, and B. Özler. 2016. “Girl ing the Efficiency of Decision Making Units.” European Power: Cash Transfers and Adolescent Welfare. Evidence Chapter 5 | Pakistan Human Capital Review Journal of Operational Research 2 (6): 429–444. from a Cluster-randomized Experiment in Malawi.” In S. Edwards, S. Johnson, and D. N. Weil (eds.), African Suc- Chetty, R., J. N. Friedman, and J. E. Rockoff. 2011. “The cesses, Volume II: Human Capital. Chicago, IL: University Long-Term Impacts of Teachers: Teacher Value-Added of Chicago Press. https:/ /www.nber.org/papers/w19479. and Student Outcomes in Adulthood.” Working Paper 17699, National Bureau of Economic Research, Cam- Banerjee, A., P. Glewwe, S. Powers, and M. Wasserman. bridge, MA. 2013. “Expanding Access and Increasing Student Learn- ing in Post-primary Education in Developing Countries: A Conn, K. 2014. Identifying Effective Education Interven- Review of the Evidence.” Post-Primary Education Initia- Saharan Africa: A Meta-analysis of Rigorous tions in Sub-­ tive Review Paper, Abdul Latif Jameel Poverty Action Lab, Impact Evaluations. Unpublished manuscript, Columbia Cambridge, MA. University, New York. 150 Dinu, B. 2015. “Impact of Teacher-Student Communica- Filmer, D., and N. Schady. 2011. “Does More Cash in Con- tion on ‘High-Risk Dropout’ Students.” Human Develop- ditional Cash Transfer Programs Always Lead to Larger ment and Sciences, Ochanomizu University, Bunkyoku, Impacts on School Attendance?” Journal of Development 112–8610, 2–1-2 Otsuka, Tokyo, Japan. https://core.ac.uk​ Economics 96 (1): 150–157. /download/pdf/234682592.pdf. Garcia, S., and J. Saavedra. 2017. “Educational Impacts Drake, L., A. Woolnough, C. Burbano, and D. Bundy. and Cost-Effectiveness of Conditional Cash Transfer Pro- 2016. Global School Feeding Sourcebook: Lessons grams in Developing Countries: A Meta-analysis.” Work- from 14 Countries. Singapore: World Scientific. https://​ ing Paper 23594, National Bureau of Economic Research, documents.wfp.org/stellent/groups/public/documents​ Cambridge, MA. /communications/wfp284904.pdf. Garn, J., L. Greene, R. Dreibelbis, S. Saboori, R. Rhein- Duflo, E. 2001. “Schooling and Labor Market Conse- gans, and M. Freeman. 2013. “A Cluster-randomized Trial quences of School Construction in Indonesia: Evidence Assessing the Impact of School Water, Sanitation and Hy- from an Unusual Policy Experiment.” American Economic giene Improvements on Pupil Enrolment and Gender Par- Review 91 (4): 795–813. ity in Enrolment.” Journal of Water Sanitation and Hygiene for Development 3: 592–601. https:/ /iwaponline​.com​ Duflo, E., P. Dupas, and M. Kremer. 2017. “The Impact of /washdev​/article​-abstract​/3​/4​/592​/30334​/A​-cluster​ Free Secondary Education: Experimental Evidence from -randomized​-trial​-assessing​-the​-impact​-of ​?redirected​ Ghana.” Working Paper, Massachusetts Institute of Tech- From​=fulltext. nology, Cambridge, MA. http:/ /www.poverty-action.org​ /publication/returns-secondary-education-unpacking​ Glewwe, P. W., E. A. Hanushek, S. D. Humpage, and R. -delivery-senior-secondary-schooling-ghana. Ravina. 2014. “School Resources and Educational Out- comes in Developing Countries: A Review of the Liter- Edmonds, E., and M. Shrestha. 2014. “You Get What You ature from 1990 to 2010.” In P. Glewwe (ed.), Education Pay for: Schooling Incentives and Child Labor.” Journal of Policy in Developing Countries. Chicago, IL: University of Development Economics 111: 196–211. https:/ /www.nber​ Chicago Press. https://doi/org/10.3386/w17554. .org/papers/w19279. Global Education Evidence Advisory Panel. 2020. Eridani, T. 2020. “Online Tutoring Improves Disadvan- “Cost-Effective Approaches to Improve Global Learning: taged School Pupils Performance and Wellbeing in What Does Recent Evidence Tell Us Are ‘Smart Buys’ for Lockdown.” News Release, EurekAlert, 15 July. https://​ Improving Learning in Low- and Middle-income Coun- www.eurekalert.org/news-releases/806598. tries?” World Bank. https://documents1.worldbank.org​ Evans, D., and A. Mendez Acosta. 2020. “Education in Af- /curated/en/719211603835247448/pdf/Cost-Effective​ rica: What Are We Learning?” Working Paper 542, Center -Approaches-to-Improve-Global-Learning-What-Does​ for Global Development, Washington, DC. https:/ /​www​ -Recent-Evidence-Tell-Us-Are-Smart-Buys-for-Improving​ .cgdev​.org​/sites​/default​/files​/education​-africa​-what​-are​ -Learning-in-Low-and-Middle-Income-Countries.pdf. -we​-learning.pdf. González-Velosa, C., L. Ripan, and D. R. Shady. 2012. “How Evans, D., and A. Popova. 2016. “What Really Works to Can Job Opportunities for Young People in Latin America be Improved?” Labor Markets and Social Security Unit Chapter 5 | Pakistan Human Capital Review Improve Learning in Developing Countries? An Analysis of Divergent Findings in Systematic Reviews.” The World (SCL/LMK), Technical Notes IDB-TN-345, Inter-American Bank Research Observer 31 (2). https:/ /openknowledge​ Development Bank, Washington, DC. .worldbank.org/handle/10986/29308. Graham, J., and S. Kelly. 2018. How Effective Are Early Evans, D., and F. Yuan. 2019. “What We Learn about Girls’ Grade Reading Interventions? A Review of the Evidence. Education from Interventions that Do Not Focus on Washington, DC: World Bank. https:/ /doi.org/10.1596​ Girls.” Policy Research Working Paper 8944, World Bank, /1813​-9450​-8292. Washington, DC. https:/ /openknowledge.worldbank.org​ Hanushek, E., and S. Rivkin. 2006. “Teacher Quality.” /handle/10986/32128. In Handbook of the Economics of Education, Volume 2. Eyal, K., I. Woolard, and J. Burns. 2014. “Cash Transfers Elsevier. and Teen Education: Evidence from South Africa.” Unpub- Hasan, Amer; Geven, Koen; Tahir, Ayesha. 2021. SMS lished Report. Girl Data Insights: How Has COVID-19 Affected Support 151 for Girls’ Education in Punjab, Pakistan?. World Bank, Abdul Latif Jameel Poverty Action Lab, Cambridge, MA. Washington, DC. © World Bank. https:/ /openknowledge​ https://www.povertyactionlab.org/publication/roll-call​ .worldbank.org/handle/10986/35477 License: CC BY -getting-children-school. 3.0 IGO Kim, J., H. Alderman, and P. Orazem. 1999. “Can Private Harris, C. 2001. “Summer School: Research-Based Rec- School Subsidies Increase Enrollment for the Poor? The ommendations for Policymakers.” SERVE (South Eastern Quetta Urban Fellowship Program.” The World Bank Eco- Regional Vision for Education) Policy Brief, National Insti- nomic Review 13: 443–65. https:/ /elibrary.worldbank.org​ tute on the Education of At-Risk Students, Office of Edu- /doi/10.1093/wber/13.3.443. cational Research and Improvement, Department of Ed- Kremer, M., E. Miguel, and R. Thornton. 2009. “Incentives ucation, Washington, DC. https://files.eric.ed.gov/fulltext​ to Learn.” The Review of Economics and Statistics 91: /ED456557.pdf. 437–56. https://dash.harvard.edu/handle/1/3716457. Herrera, S., and A. Ouedraogo. 2018. “Efficiency of Pub- Krishnaratne, S., H. White, and E. Carpenter. 2013. “Qual- lic Spending in Education, Health, and Infrastructure: An ity Education for All Children: What Works in Education in International Benchmarking Exercise.” Policy Research Developing Countries. 3ie Working Paper 20, Internation- Working Paper 8586, World Bank, Washington, DC. https:/ / al Initiative for Impact Evaluation. https://www.3ieimpact​ openknowledge.worldbank.org/handle/10986​/30431. .org/evidence-hub/publications/working-papers/quality​ Horn, M., and H. Staker. 2011. The Rise of K-12 Blended -education-all-children-what-works-education. Learning. Lexington, MA: Innosight Institute. Le, H. M. 2018. “Another Textbook Project? The Imple- Horn, M., and H. Staker. 2012. Classifying K-12 Blended mentation of Escuela Nueva in Vietnam.” Educational Learning. Lexington, MA: Innosight Institute. Research for Policy and Practice 17: 223–239. https://doi​ .org/10.1007/s10671-018-9230-x. Human Rights Watch. 2018. “Shall I Feed my Daughter, or Educate Her? Barriers to Girls’ Education in Pakistan.” Lim, B., K. Lee, and C. Lee. n.d. “Free Disposal Hull (FDH) 12 November. https:/ /www.hrw.org/report/2018/11/12​ Analysis for Efficiency Measurement: An Update to DEA.” /shall-i-feed-my-daughter-or-educate-her/barriers-girls​ The Stata Journal. https:/ /www.cgdev.org/sites/default​ -education-pakistan. /files/archive/doc/stata/MO/DEA/free_disposal_hull​ .pdf. Ibarrarán, P., and D. Shady. 2009. “Evaluating the Impact of Job Training Programs in Latin America: Evidence from Lim, J., and J. Meer. 2017a. “The Impact of Teacher–Stu- IDB Funded Operations.” Journal of Development Effec- dent Gender Matches: Random Assignment Evidence from tiveness 1 (2): 195–216. South Korea.” Journal of Human Resources 52 (4): 979– 997. http://jhr.uwpress.org/content/52/4/979​.abstract. International Education News. 2019. “Expanding to Say “Yes”: The Ongoing Work of The Citizens Foundation in Lim, J., and J. Meer. 2017b. “Persistent Effects of Teach- Pakistan (Part 1 of 2).” 1 May. https://internationalednews​ er–Student Gender Matches.” Working Paper 24128, Na- .com/2019/05/01/expanding-to-say-yes-the-ongoing​ tional Bureau of Economic Research, Cambridge, MA. -work-of-the-citizens-foundation-in-pakistan-part-1-of​-2/. Macdonald, K., and B. T. Vul. 2018. “A Randomized Evalu- Chapter 5 | Pakistan Human Capital Review Islam, A. 2019. “Parent–Teacher Meetings and Student ation of a Low-Cost and Highly Scripted Teaching Method Outcomes: Evidence from a Developing Country.” Euro- to Improve Basic Early Grade Reading Skills in Papua New pean Economic Review 111: 273–304. https://www.theigc​ Guinea.” Education Global Practice, Policy Research Work- .org/wp-content/uploads/2016/10/Islam-2016-working​ ing Paper 8427, World Bank, Washington, DC. https:/ / -paper.pdf. open​knowledge​.worldbank​.org​/bitstream​/handle​/10986​ /29833​/WPS8427​.txt​?sequence​=2&​isAllowed​=y. Jamal, A. 2016. “Why He Won’t Send His Daughter to School­—­Barriers to Girls’ Education in Northwest Paki- Maluccio, J., A. Murphy, and F. Regalia. 2010. “Does Sup- stan: A Qualitative Delphi Study of Pashtun Men.” SAGE ply Matter? Initial Schooling Conditions and the Effective- Open. doi:10.1177/2158244016663798 ness of Conditional Cash Transfers for Grade Progres- sion in Nicaragua.” Journal of Development Effectiveness J-PAL (Abdul Latif Jameel Poverty Action Lab). 2017. “Roll 2: 87–116. https:/ /www.researchgate.net/publication​ Call: Getting Children into School.” J-PAL Policy Bulletin, 152 /228877017_Does_supply_matter_Initial_schooling​ Rumberger, R. 2001. Why Students Drop Out of School _conditions_and_the_effectiveness_of_conditional_cash​ and What Can be Done. University of California: Santa _transfers_for_grade_progression_in_Nicaragua. Barbara, CA. McEwan, P. 2012. “Improving Learning in Primary Schools Sperling, G. B., and R. Winthrop. 2015. What Works in of Developing Countries: A Meta-analysis of Randomized Girls’ Education: Evidence for the World’s Best Invest- Experiments.” Review of Educational Research 20 (10): 1–42. ment. Washington, DC: Brookings Institution Press. Murnane, R. J., and A. J. Ganimian. 2014. “Improving Ed- Suryadarma, D., A. Suryahadi, S. Sumarto, and H. Rogers. ucational Outcomes in Developing Countries: Lessons 2006. “Improving Student Performance in Public Primary from Rigorous Evaluations.” Working Paper 20284, Na- Schools in Developing Countries: Evidence from Indone- tional Bureau of Economic Research, Cambridge, MA. sia.” Education Economics 14 (4): 401–429. https://www.nber.org/papers/w20284.pdf. The Economist. 2020. “England’s Catch-up Tutoring Nickow, A., P. Oreopoulos, and V. Quan. 2020. “The Im- Programme Has Bold Ambitions.” 5 November. https://​ pressive Effects of Tutoring on PreK–12 Learning: A Sys- www.economist.com/britain/2020/11/05/englands​ tematic Review and Meta-analysis of the Experimental -catch-up-tutoring-programme-has-bold-ambitions. Evidence.” Working Paper 27476, National Bureau of Eco- Theunynck, S., and H. Rabakoson. 2017. “School Construc- nomic Research, Cambridge, MA. https:/ /www.nber.org​ tion for Basic Education in Burundi.” School Construction /papers/w27476. Study, World Bank, Washington, DC. http:/ /documents1​ OECD (Organisation for Economic Co-operation and Devel- .worldbank.org/curated/en/751581498759557305/pdf​ opment). 2020. “The Impact of COVID-19 on Student Equi- /P161127-06-29-2017-1498759550576.pdf. ty and Inclusion: Supporting Vulnerable Students during Thomas, C., and C. Shaw. 1992. “Issues in the Develop- School Closures and School Re-openings.” Paris: OECD ment of Multigrade Schools.” Technical Paper 172, World Publishing. https://www.oecd.org/coronavirus​/policy​ Bank, Washington, DC. -responses​/the​-impact​-of​-covid​-19​-on​-student​-equity​ -and​-inclusion​-supporting​-vulnerable​-students​-during​ Trako, E., J. Baron, M. Bend, and N. Ejaz. 2020. Barriers -school​-closures​-and​-school​-re​-openings​-d593b5c8/. to Girls’ Education in Pakistan: Special Section on the Po- tential Impact of COVID-19. Washington, DC: World Bank. Oreopoulos, P. 2020. “Scale Up Tutoring to Combat COVID Learning Loss for Disadvantaged Students.” Scientific UNESCO (United Nations Educational, Scientific and American, 24 November. https:/ /www.scientificamerican​ Cultural Organization) Institute for Statistics. 2018. “Out- .com/article/scale-up-tutoring-to-combat-covid-learning​ of-School Children and Youth.” Fact Sheet 48, UNESCO, -loss-for-disadvantaged-students/. Paris. http://uis.unesco.org/en/topic/out-school​-children​ -and​-youth. Pakistan Education Statistics. 2016–17. National Educa- tion Management Information System (NEMIS), Acade- UNESCO (United Nations Educational, Scientific and Cul- my of Educational Planning and Management (AEPAM), tural Organization). 2019. Global Education Monitoring Re- Pakistan. port. Paris: UNESCO. https:/ /en.unesco.org/gem​-report/. Chapter 5 | Pakistan Human Capital Review Piper, B., and K. Medina. 2010. “Early Grade Reading As- UNESCO (United Nations Educational, Scientific and Cul- sessment (EGRA) Plus: Liberia. Program Evaluation Re- tural Organization) Institute for Statistics. 2020. “Educa- port.” RTI International. https://files.eric.ed.gov/fulltext​ tional Attainment, at Least Completed Upper Secondary, /ED516080.pdf. Population 25+, Female (%) (Cumulative) – Pakistan.” Paris: UNESCO. https:/ /data.worldbank.org/indicator/SE​ Psacharopoulos, G., C. Rojas, and E. Velez. 1993. “Achieve- .SEC.CUAT.UP.FE.ZS?locations=PK. ment Evaluation of Colombia’s Escuela Nueva: Is Multi- grade the Answer?” Comparative Education Review 37 UNICEF (United Nations Children’s Fund). 2022. Are Chil- (3): 263–276. dren Really Learning? Exploring foundational skills in the midst of a learning crisis. New York: UNICEF. https:/ /​www​ RTI. 2020. “Structured Pedagogy Literature Review.” https:// .unicef​.org​/press​-releases​/23​-countries​-yet​-fully​-reopen​ scienceofteaching.s3.eu-west-3.amazonaws.com​/index​ -schools​-education​-risks​-becoming​-greatest​-divider. .html#​/lessons​/qRcSj0b2yqminRXJh7cSd_2tkJc​-0blC. 153 WFP (World Food Programme). 2020. State of School World Bank. 2020c. Learning Losses in Pakistan Due Feeding Worldwide 2020. Rome: WFP. https:/ /www.wfp​ to COVID-19 School Closures: A Technical Note on Sim- .org/publications/state-school-feeding-worldwide-2020 ulation Results. Washington, DC: World Bank. https://​ openknowledge.worldbank.org/handle/10986/34659. World Bank. 2018. World Development Report 2018: Learning to Realize Education’s Promise. Washington, World Bank. 2020d. Cost​ -Effective Approaches to Im- DC: World Bank. https://openknowledge.worldbank.org​ prove Global Learning. Recommendations of the Global /handle/10986/28340. Education Advisory Panel. Washington, DC: World Bank. World Bank. 2019a. Ending Learning Poverty: What Will It World Bank. 2022e. Prioritizing Learning During Covid-19: Take? Washington, DC: World Bank. The Mos Effective Ways to Keep Children Learning During the Post-Pandemic. Recommendations of the Global Edu- World Bank 2019b. South Asia Economic Focus, Fall 2019: cation Advisory panel. Washington, DC: World Bank. Rethinking Decentralization. Washington, DC: World Bank. https://doi.org/10.1596/978-1-4648-1515-7. World Bank. Forthcoming. Barriers to Girls’ Education in Pakistan. Washington, DC: World Bank. Chapter 5 | Pakistan Human Capital Review 154 Improving labor market outcomes of the poor SUMMARY only slightly less educated than men, and for- 6 mally wage-employed women are far better ed- There is a close relationship between human ucated than formally wage-employed men. capital and labor market outcomes of the poor. Poorer households not only lag in human capi- Boosting the human capital of the poor, espe- tal outcomes but also invest far less in absolute cially against shocks like the COVID-19 pan- terms in education and health than wealthi- demic, requires interventions over the life cycle er households. Investments in poor people’s to ensure long-term livelihood gains. Safety net human capital and labor market outcomes are programs that augment cash transfers with self-reinforcing. Boosting incomes and liveli- other services, often called cash-plus programs­ hoods of the poor is thus critical for strengthen- —­ in addition to other poverty-targeted econom- ing human capital. In Pakistan, women, youth, ic inclusion programs that combine social assis- and the less-educated generally have far poorer tance with measures to boost productivity­ —­are labor market outcomes, and women and vulner- essential for improving the livelihoods of the able groups tend to suffer more from shocks, poor, their human capital, and their productiv- such as the COVID-19 pandemic and climate ity. For households not poor enough to receive change. cash transfers, seed capital, microfinance, and technical and business management skills In Pakistan, women’s labor force participation training could be more relevant. Support to the rate (LFPR) is low, at just over 20 percent (and self-employed could include business manage- even lower for women age 15–24), while men’s ment training, paired with other complementa- is 80  percent. And 60  percent of working-age ry interventions. Tackling additional constraints women are not in employment, education, or faced by women will need to be carefully inte- training (NEET), against only 6 percent of work- grated into the design of interventions and with ing-age men. Social norms, limited mobility, and flexibility to address the challenges women may widespread safety concerns are associated with face. In this regard, new programs need to clear- women’s inability to be in the labor force. Other ly identify specific targets. Identifying potential factors that hinder women’s economic activities beneficiaries can be done using tried and tested Chapter 6 | Pakistan Human Capital Review include biases in hiring decisions and addition- existing national systems such as the National al costs that employers may pay for segregated Socio-Economic Registry and poverty score- spaces and restrooms. cards used by other public programs such as the Benazir Income Support Program (BISP). Gaps are substantial in the earnings of women and of informal sector workers. Formally wage- employed females, informally wage-employed INTRODUCTION males, and formally wage-employed males earn substantial premia for being formally employed Investments in human capital and labor mar- or for being male (regardless of formality sta- ket outcomes are self-reinforcing. For example, tus). The gender gaps are even more glaring an additional year of school generates higher since informally wage-employed women are earnings on average. Ensuring access to quality 157 education boosts cognitive and socioemotional indicators and illustrates their broad variation skills, which enlarges human capabilities. And across different segments of the labor force. It good health increases workers’ productivity. describes some of the key employment patterns These individual returns to human capital gener- of the poor and puts a spotlight on the promi- ate substantial benefits for the wider economy­—­ nence of the informal sector and the impact of greater affluence and thus accelerated poverty the COVID-19 pandemic (box 6.1). The chapter reduction­ —­helping break the intergenerational then reviews the global evidence on programs transmission of poverty. But Pakistan’s Human that could both boost the human capital of the Capital Index (HCI) value of 0.41 suggests that poor and improve their earnings to enable them a child born in Pakistan today will, by age 18, be to invest more in human capital. It also reviews only 41 percent as productive as she would have the few prominent programs that may be rel- been if she had received adequate investments evant in this area and illustrates some entry in education and health.1 That puts the human points for new programs in Pakistan. capital of the next generation of Pakistani work- ers lower than the global average of 56 percent and the average for South Asia of 48 percent. LOW UNEMPLOYMENT BUT POORER LABOR MARKET The problem is that poorer households are OUTCOMES FOR YOUTH, WOMEN, often unable to invest enough in human capital, AND THE LESS EDUCATED so they lag in human development outcomes, contributing to the persistence of intergenera- Low unemployment masks tional poverty. For example, households in the structural challenges for the bottom consumption quintile in Pakistan spend vulnerable groups PKR 196 per month on education, those in the top quintile, PKR 3,728; the equivalent figures Pakistan’s unemployment rate, at 3.1 percent, is for health are PKR 765 and PKR 1,623.2 As their relatively low and similar to those in developing incomes rise, households tend to invest more countries with high poverty rates (figure 6.1). in health and education, with high income elas- Average unemployment rates are substantially ticities of education spending, especially for lower in developing countries than in rich coun- poorer households, and associations between tries (2.5 percent in the poorest quartile of the household wealth and out-of-pocket (OOP) world income distribution compared with 8 per- health spending.3 OOP health spending in Pa- cent on average).6 Since people are too poor to kistan accounts for 56.2 percent of total health have the luxury of not working, most less-skilled spending on average.4 In a study of OOP health workers in developing countries choose self- spending in katchi abadis (urban informal settle- employment activities and cannot afford to be Chapter 6 | Pakistan Human Capital Review ments) in Islamabad, 86 percent of households unemployed.7 had to bear such expenses, either by borrowing money from a neighbor or a relative or by selling Gender gaps are wide household belongings to treat children under 5.5 Such expenses can drain the finances of poor Despite similar unemployment rates among households. women and men, Pakistan’s labor market shows large gender gaps—­ a low LFPR among The chapter builds on the relationship between women and high rates of women NEET (figure human capital and labor market outcomes for 6.2). The LFPR among men, at 80 percent, is a the poor and reviews some of the few programs startling 60 percentage points higher than that currently active for addressing the needs of the for women. On the NEET rate, the gender gap poor. It describes the headline labor market remains huge, with 60  percent of working-­ age 158 BOX 6.1 The COVID-19 pandemic has adversely affected Pakistan’s labor market through a sudden drop in labor demand and an increase in unemployment Against the same period in 2019, the number of job group, 46  percent experienced permanent layoffs, advertisements in Pakistan plunged by 76 percent 13 percent temporary layoffs, and 26 percent cuts in mid-March 2020, when the national lockdown in salary or working hours. About 68 percent of fe- was introduced.1 Traditionally male-dominated in- males and 61 percent of males faced income reduc- dustries were more resilient than industries where tions due to work loss. female employment is concentrated. Of business- es that responded to the survey, 32  percent ex- A Gallup survey in 2020 found that 57  percent of perienced business closures, and of the firms that Pakistani workers saw their wages cut, 54 percent remained open, 51  percent reduced their working had their working hours lowered, and 52  percent hours. stopped working temporarily.4 A COVID-19 impact survey by the Pakistan Bureau These sweeping job losses had disproportional im- of Statistics suggests that almost half of the work- pacts on women and other vulnerable groups in ing population was severely affected by the busi- the informal sector. The Pakistan Institute of De- ness closures and lockdown.2 Notably, 74  percent velopment Economics in April 2020 estimated that of the affected were daily wage earners, mainly in the share of vulnerable employment­ —­among own- construction, and own-account workers in non- account workers, daily wage earners, and unpaid agriculture who either lost their jobs or saw their family workers­ —­was around 56  percent (52  per- income reduced, suggesting that social protection cent among males and 71 percent among females). programs need to cover these workers. By industry, It inevitably hit those working in agriculture, where 80 percent of workers in construction and 72 per- more than 80  percent are in vulnerable employ- cent in manufacturing were hit the most by tempo- ment. Most of the layoffs in all provinces were like- rary business closures in April–July 2020. ly be among daily wageworkers and workers paid piece rates.5 The employment share fell from 65  percent to 51 percent, while the share who were looking for Notes jobs increased by 20  percentage points, after the 1. Tas et al. 2021. 2. Pakistan Bureau of Statistics 2020. onset of the COVID-19 pandemic.3 About 52 percent 3. Ray 2021. Chapter 6 | Pakistan Human Capital Review of those who reported working before the pandem- 4. Tas et al. 2021 ic were not working in July–August 2020. Of this 5. Nasir and Faraz 2020. women being NEET against only 6  percent of (age 15–24). The LFPR for Pakistan as a whole working-age men. These gaps persist among remains comparable to the South Asia average the younger age cohort (age 15–24). (figure 6.3). However, Pakistan’s female LFPR is only 22.2 percent for women age 15–64 and Pakistan’s female LFPR ranks among the low- even lower for women age 15–24, in contrast est globally. The overall LFPR has been rela- to the female LFPR in other Asian economies. tively stable since 2001, at 52–55 percent for For example, the LFPR for young Bangladeshi all age cohorts and 41–43  percent for youth women age 15–24 was 26.3  percent in 2018, 159 FIGURE 6.1  Pakistan’s unemployment before the COVID-19 FIGURE 6.2  Pakistan’s labor force pandemic was generally much lower than the regional average participation rate shows wide gender gaps Unemployment rate, by age cohort, Pakistan and South Asia (percent) LFPR and NEET rates, by age cohort and gender, 2019 (percent) 25 Total Male Female 100 South Asia average, age 15–24 20 80 15 60 10 Pakistan, age 15–24 40 South Asia average, age 15–64 5 20 Pakistan, age 15–64 0 0 2001 2004 2006 2008 2010 2012 2014 2016 2019 Age 15–64 Age 15–24 Age 15–64 Age 15–24 Source: World Bank calculations using data from Pakistan Labor Force Survey 2018–19. LFPR NEET Source: World Bank calculations using data from Pakistan Labor Force Survey 2018–19. FIGURE 6.3  While Pakistan’s overall labor force participation rate Note: NEET is not in employment, education, or training. LFPR is is similar to the South Asia average, the youth cohort is less so labor force participation rate. LFPR, by age cohort for Pakistan and South Asia, regional average (percent) 70 women. In Pakistan, women are often employed in nonstandard forms of work such as unpaid South Asia average, age 15–64 60 family work or home-based work. Pakistan, age 15–64 Gender gaps in the LFPR and NEET rates are 50 wide in all provinces but are greatest in Baloch- Pakistan, age 15–24 istan (figure 6.4). The LFPR and NEET rates in Balochistan are 9.5  percent and 85.5  percent, 40 South Asia average, age 15–24 respectively, for women, and 83.9 percent and 5.1 percent for men. Sindh and Khyber Pakh- 30 tunkhwa have similar female LFPRs (at around 2001 2004 2006 2008 2010 2012 2014 2016 2019 13 percent), but Sindh’s female NEET rate Chapter 6 | Pakistan Human Capital Review Source: World Bank calculations using data from Pakistan Labor Force Survey 2018–19. is 57.1 percent, against 5.9  percent for men. Note: LFPR is labor force participation rate. Among provinces, Punjab, the most populous, has the smallest gender gaps in the LFPR. increasing to 38 percent once all women age 15 and older are considered. The equivalent figures The female LFPR in Pakistan is constrained on were 35 percent and 53 percent in Malaysia and both the supply and demand sides. On the sup- 55 percent and 79  percent in Vietnam. These ply side, the decision for a working-age female to LFPRs should be interpreted with caution be- become economically active depends on a range cause of a potential downward bias, particularly of factors, including social norms and early mar- for women.8 An important driver of this down- riage.9 There is evidence that women not in the ward bias is measurement error arising from a labor force transit straight from being a student lack of clarity about what constitutes “work” for to a homemaker.10 Marriage for women brings 160 FIGURE 6.4  Gender gaps in labor force participation and not in employment, education, or training rates are similar across provinces, but greatest for Balochistan LFPR and NEET rates, by gender (percent) Balochistan Khyber Pakhtunkhwa 100 100 80 80 60 60 40 40 20 20 0 0 Total Male Female Total Male Female Punjab Sindh 100 100 80 80 60 60 40 40 20 20 0 0 Total Male Female Total Male Female LFPR NEET rate Source: World Bank calculations using data from Pakistan Labour Force Survey 2018–19. Note: NEET is not in employment, education, or training. LFPR is labor force participation rate. Chapter 6 | Pakistan Human Capital Review more barriers, including increased childcare Social norms limit the type of jobs women take­ duties and housework, and less ability to make —­undercutting their potential income­—­and independent decisions: 61 percent of women in partly explain why about 75 percent of female urban areas and 45 percent in rural areas work employment is in agriculture13 and the high from their dwelling because of housework and share of home-based work. reproductive duties.11 Men’s support for wom- en’s employment is a crucial decision factor in Limited mobility and widespread safety con- the female LFPR. Given such limited empower- cerns are associated with women’s inability to ment of women in Pakistan, the education levels be in the labor force. A significant proportion of husband and parents are also potential de- of women respondents to the World Bank’s terminants of a woman joining the labor force.12 2013 Pakistan Labor Skills Survey reported 161 that they could not travel alone to essential FIGURE 6.5  Unemployment rates are higher health services, for social reasons, or to the for better educated workers in Pakistan local market. About 17 percent of women who a. Years of education attainment, by employment status could travel to the local market alone were in Years Total Male Female the labor force compared with 9 percent of 10 women who reported that they could not go to the market alone. Women who felt safe walking 8 alone outside in their communities or neigh- borhoods were more likely to work (17 percent) 6 than those who did not feel safe (11 percent). These concerns also partly explain why women 4 in Pakistan are disproportionately engaged in home-based work.14 2 On the demand side, job opportunities may not be available because of possible biases in hir- 0 ing and additional costs that employers pay for Employed Unemployed women’s conveniences.15 It is not uncommon to b. Unemployment rates, by education status find job postings based on gender in Pakistan’s Percent leading job portal, Rozee.pk.16 And employers Less than primary Primary Lower secondary Secondary School Higher Secondary Post- may face added expenses to provide segregat- Certi cate School Certi cate secondary ed spaces, restrooms, and transport for accom- 25 modating female workers. Without dedicated spaces for females, women might not be willing 20 to work. 15 The low quality of jobs and gendered occu- pational pattern among female workers also 10 seem to be demand factors in determining women’s economic activities. For example, in Peshawar, a sizable proportion of the already 5 small female workforce (21 percent) is unpaid as family workers compared with 3.1 percent 0 Male Female of the male workforce. The World Bank’s Quet- ta Urban Household Survey 2021 reveals that Source: World Bank calculations using data from Pakistan Labour Force Survey 2018–19. Chapter 6 | Pakistan Human Capital Review only 16  percent of working-age women par- ticipate in the labor market, and most of the employed women in Quetta city perform low The COVID-19 shocks have worsened the cur- value-added, socially accepted occupations as rent low rate of economic activity for Pakistani self-employed or own-account workers, partic- women. Recent administrative data from the ularly among low-educated women: 61.7  per- largest job-posting site reveal notable job losses cent of the employed women working in the in the sectors where women are more likely to textile sector and 9.9  percent in education.17 be employed: education and health, particularly This occupational profile of working women re- in urban areas. About 59  percent of previous- flects the gender gaps in earnings and as result ly employed female respondents reported that offers few incentives for women to participate they lost their jobs once the pandemic hit com- in the labor market. pared with 50  percent of male respondents.18 162 The pandemic has clearly led to a dispropor- FIGURE 6.6  In Pakistan’s youth age cohort, unemployment are tionate increase in women’s unpaid care work, higher in urban areas and climbed faster than in rural areas regardless of their employment status. Unemployment rates, based on time-use for 15- to 24-year-old age cohort (percent) 2010 2019 Unemployment rates vary widely 20 by education and location 15 Unemployment rates are generally lower for less-educated workers (figure 6.5). Unemploy- ment for men and women with at least a post- 10 secondary degree is more than 21 percent but less than 4 percent for those with less than pri- mary education. 5 Unemployment and underemployment among 0 youth (age 15–24) are higher in urban than National Rural Urban National Rural Urban National Rural Urban rural areas.19 This pattern partly reflects the Total Male Female higher education attainment of urban workers. Over 2010–19, the unemployment rate grew Source: World Bank calculations using data from the International Labour Organization, ­ LABORSTA modeled estimates. by 7.4 percentage points for urban males and by 6.6 percentage points for rural males (fig- ure 6.6). For urban females, the equivalent fig- FIGURE 6.7  Four-fifths of wageworkers are ures were increases of 10.4 and 3.7 percentage informally employed points. For underemployment (fewer than 35 Distribution of wage-employed workers (percent) hours a week worked), there is some evidence Male Female that the key drivers in Pakistan are mismatch- Informal Formal Informal Formal 100 es between labor supply and demand.20 Al- though education attainment has improved in 80 Pakistan, growing numbers of workers cannot use their education background on the job for 60 work. 40 EMPLOYMENT IN PAKISTAN IS OVERWHELMINGLY INFORMAL 20 Chapter 6 | Pakistan Human Capital Review Informality predominates in Pakistan, with 0 94.9 percent of wageworkers not having a for- Micro Small Medium Large All rms mal contract:21 83  percent of all wageworkers Source: World Bank calculations using data from Pakistan Labour are informally employed male workers, and in- Force Survey 2018–19. Note: Microenterprises are firms with fewer than 10 workers, formally employed female workers account for small firms are those with 10–99 workers, medium firms are another 12 percent (figure 6.7). The absence of those with 100–250 workers, large firms are those with more formal contracts for wageworkers is greater, the than 250 workers. See Annex 6A for the definition of formal wage smaller the firm: 95  percent, 89  percent, and employment. 82 percent in micro-, small, and medium enter- prises, respectively. Of wage employees in large firms, 38 percent have formal contracts (see the 163 FIGURE 6.8  Almost 90 percent of (households engaged in self-employment and wageworkers, formal and informal, are micro­ enterprises with fewer than 10 employ- employed in microenterprises ees) more than doubled. Microenterprises in- Distribution of wage-employed workers (percent) creased by 60  percent, while the number of Micro Small Medium and large self-employed grew by 53 percent. About four- 100 fifths of informal establishments are in serv- ices. Employment in the informal sector and in 80 microenterprises­—­both traditionally associated with low productivity­—­ has thus been a major 60 driver of poverty reduction in recent years.23 40 In 2015, according to household survey data, 24  percent of Pakistan’s population was liv- ing below the national poverty line, down from 20 64 percent in 2001.24 The observed progress in poverty reduction is accounted for largely by the 0 Informal, Formal, Informal, Formal, All types expansion of male off-farm economic opportu- male male female female nities in the informal sector and by the increase Source: World Bank calculations using data from Pakistan Labour in cross-border migration and associated re- Force Survey 2018–19. mittances.25 The growth of entrepreneurship Note: Microenterprises are firms with fewer than 10 workers, and wage work has coincided with reductions small firms are those with 10–99 workers, medium firms are those with 100–250 workers, large firms are those with more in unpaid work in household enterprises (con- than 250 workers. See Annex 6A for the definition of formal wage tributing family workers), though once again employment. wide gender gaps are apparent (table 6.1)­ —a­s are gaps among age cohorts for wagework- note to figure 6.7 for number of employees by ers (figure 6.9). Unpaid family members con- firm size). Microenterprises account for 88 per- tributing to household enterprises accounted cent of all employment, formal and informal, for for 21.4  percent of employment in 2001, and men and women, with small firms accounting this rate fell by 2.2 percentage points by 2019. for another 12 percent (figure 6.8). Small firms Own-account work also fell by 6.5 percentage are more important for formal workers than me- points over the same period, from 42.2  per- dium and large firms. cent in 2001. Labor from unpaid contributions to household enterprises and other generally The informal sector and microenterprises were, low-productivity own-account work was divert- before the COVID-19 pandemic, vital for main- ed into higher-paying wage work or entrepre- Chapter 6 | Pakistan Human Capital Review taining labor demand growth.22 Formal private neurship. On the shifts by gender, the aggregate sector development in Pakistan often faced phenomenon is driven by shifts among men. constraints, including an unfavorable business For example, the share of male labor who were environment, the lack of integration with glob- wageworkers increased from 36.0  percent to al value chains, and the sector’s own failure to 47.2  percent over 2001–19. In contrast, there innovate. The informal sector was thus crucial was virtually no change in the percentage of for absorbing the growing labor force. Three working women contributing to household en- of four jobs outside agriculture were generat- terprises: the share was 49.5 percent in 2001 ed in the informal sector over 2001–17, with and 50.2 percent in 2019. The share of women 95 percent in construction. Between 2001 and who were wageworkers or entrepreneurs actu- 2015, the number of informal establishments ally fell. 164 TABLE 6.1  Distribution of workers by employment type, FIGURE 6.9  Informality rates are higher among 2001 and 2019 youth wageworkers in urban areas than among urban wageworkers generally Change Share of wageworkers (youth and full working-age) informally employed, by 2001 2019 (percentage location and gender (percent) Employment type (percent) (percent) points) Total Age 15–64 Age 15–24 Urban Rural Urban Rural  Employees 35.58 43.68 8.10 100  Employers 0.78 1.41 0.63  Own-account workers 42.23 35.73 –6.50 80  Contributing family workers 21.41 19.17 –2.23 Male 60  Employees 35.96 47.20 11.24  Employers 0.90 1.74 0.84 40  Own-account workers 46.78 39.66 –7.11  Contributing family workers 16.36 11.39 –4.96 20 Female  Employees 33.42 29.64 –3.77 0 Total Male Female  Employers 0.12 0.10 –0.01 Source: World Bank calculations using data from Pakistan Labor Force Survey  Own-account workers 16.96 20.07 3.12 2018–19.  Contributing family workers 49.51 50.18 0.67 Source: World Bank calculations using data from the International Labor Orga- nization, LABORSTA. GENDER AND HOUSEHOLD labor productivity. Generally, the labor produc- STATUS IN INFORMAL SETTINGS tivity of informal activities has been steadily DAMPEN THE BENEFITS OF increasing: that of both the self-employed and HUMAN CAPITAL INVESTMENTS microenterprises was about three times higher in 2015 than 2001. Growth rates slowed during Education, employment sector, the Global Financial Crisis, particularly among location, and work experience all microenterprises, but remained positive. The affect earnings economy as a whole recorded a 3.1 percent de- cline in labor productivity over 2007–11. The impact of education on earnings is consis- Chapter 6 | Pakistan Human Capital Review tent with findings on informal establishment An individual’s monthly earnings closely reflect owners’ productivity and education in Pakistan. their education attainment, employment sec- The productivity of establishments whose own- tor, work experience, and gender consistently ers have at least upper-secondary education across provinces and at the national level. The is about 1.3 times that of those who do not. Dif- returns to education attainment and employ- ferences in productivity persist, and labor pro- ment in a particular sector in Pakistan are con- ductivity is 10 percent higher for owners with sistent with cross-country evidence, after a upper-secondary education, after differences in range of characteristics, such as hours of work, work experience, sector, location, and income are controlled for. Higher education attainment are controlled for. Over time, a 1  percentage is associated with higher earnings, as is greater point increase in the share of educated own- work experience proxied by age (though the re- ers is associated with a 0.5 percent increase in turns to experience decline over time). Working 165 FIGURE 6.10  Gaps in income by gender and formality status are outside agriculture is associated with higher in- wide in Pakistan comes, as is being in an urban area. Average monthly income (PKR) 20,000 Gaps are substantial in the earnings of women and informal 16,809 sector workers 15,000 14,885 Reflecting gender gaps, being female is associ- 12,430 ated with substantially lower earnings: informal- 10,000 ly wage-employed females in Pakistan have an average monthly income of only PKR 6,500. For- 6,523 mally wage-employed females, informally wage- 5,000 employed males, and formally wage-employed males earn 1.9, 2.3, and 2.6 times as much per 0 month­ —­large premia for being formally em- Informally wage Informally wage Formally wage Formally wage ployed or for being male (regardless of formality employed, male employed, female employed, male employed, female status). The gender gaps are even more glaring Source: World Bank calculations using data from Pakistan Labor Force Survey 2018–19. Note: See Annex 6A for the definition of formal wage employment. given that informally wage-employed women are only slightly less educated than men (4.8 years versus 5.1 years of education attainment) FIGURE 6.11  Formally wage-employed women in Pakistan are the and that formally wage-employed women are far best educated among the four groups better educated than formally wage-employed Education attainment (years) men (10.2 years as opposed to 7.1 years of edu- 12 cation attainment). Education attainment does not correlate with gender gaps in incomes (fig- 10 10.2 ures 6.10 and 6.11). 8 Household status affects earnings 7.1 and employment type 6 5.1 4.8 The extent of the contributing factors to income 4 depends on whether an individual is from a 2 poor household (table 6.2). There is a substan- tial negative association between earnings and 0 being from a household in the bottom 40 per- Chapter 6 | Pakistan Human Capital Review Informally wage Informally wage Formally wage Formally wage cent (B40) of the consumption distribution. employed, male employed, female employed, male employed, female This relationship is understandable since some- Source: World Bank calculations using data from Pakistan Labor Force Survey 2018–19. one from a poorer household would likely have Note: See Annex 6A for the definition of formal wage employment. lower wages. The correlation between educa- tion attainment and monthly wages is also well understood from the literature­ —­education at- tainment (whether primary, secondary, or post- secondary) always improves wages, highlighting the returns to human capital investment. How- ever, when a household’s B40 consumption status interacts with education attainment, the overall association between education and 166 TABLE 6.2  Mincerian regressions with log of TABLE 6.3  Probit estimations of the probability of being in unpaid monthly earnings as the dependent variable, work or self-employment, Pakistan Pakistan Unpaid family worker = 1 Self-employment = 1 Variable versus paid wage = 0 versus paid wage = 0 Dependent variable = log of Education (base: no education) Variable monthly wages  Primary 0.013*** 0.028*** Education (base: no education)  Secondary 0.027*** 0.034***  Primary 0.163***  Postsecondary –0.005** –0.018***  Secondary 0.307*** Employment sector (base: agriculture)  Postsecondary 0.789***  Industry –0.168*** –0.186*** Employment sector (base: agriculture)  Service –0.149*** –0.109***  Industry 0.281*** Urban 0.001 –0.011***  Service 0.217*** Female 0.067*** –0.059*** Urban 0.123*** Work experience (proxied by age) –0.279*** 0.240*** Female –1.316*** From bottom 40 –0.044*** –0.023*** Work experience (proxied by age) 0.078*** Primary education * Bottom 40 –0.009*** 0.004 From bottom 40 –0.344*** Secondary education * Bottom 40 –0.009*** 0.003 Primary education * Bottom 40 –0.050** Postsecondary education * Bottom 40 –0.001 0.015*** Secondary education * Bottom 40 –0.159*** Postsecondary education * Bottom 40 –0.528*** Source: World Bank estimates using data from the Household Income and Expenditure Sur- vey 2018–19. Source: World Bank estimates using data from the Household In- Note: Standard errors in parentheses, *** p < .01, ** p < .05, * p < .1. Bottom 40 refers to bot- come and Expenditure Survey 2018–19. tom two quintiles of the consumption distribution. Note: Standard errors in parentheses, ***  p < .01, ** p < .05, * p < .1. Bottom 40 refers to households in the bottom two quin- tiles of the consumption distribution. Since the dependent vari- able is the log of monthly wages, the coefficients on the explanato- employment in Pakistan. And being from a B40 ry variables can also be interpreted as percentage point changes to month wages. For example, being in an urban area would im- household increases the probability of being in prove wages by 12.3 percent (relative to someone in a rural area), paid wage employment or self-employment, as while achievement of primary education would improve wages by determined by Probit estimations (table 6.3). 16.3 percent (relative to someone with no education). People with primary and secondary school com- pletion (against no education) are more likely higher earnings is substantially lower. These re- to be in paid wage employment than in unpaid sults for Pakistan are consistent with evidence employment and to be self-employed than in from other countries. After other variables are paid wage employment. Women are more likely Chapter 6 | Pakistan Human Capital Review controlled for, education had an impact on to be unpaid than paid wageworkers and less household incomes in Vietnam, but the returns likely to be self-employed. B40 individuals are to education were very heterogeneous across more likely to be in paid wage employment, but the income distribution and tended to be higher the interactions between household wealth sta- for better-off households, which could have had tus and education attainment vary. B40 individ- impacts on income inequality.26 Similarly, Min- uals who have primary or secondary education cerian regressions for households in rural China attainment only (but not necessarily those with found that returns to education were higher for postsecondary education) are more likely to be better-off households.27 wage-employed than in unpaid work; however, B40 individuals with postsecondary education These same factors also affect an individu- are more likely to be in self-employment than in al’s status in paid wage employment and self- wage employment. 167 Socioemotional skills matter for evidence on the positive correlation between some labor market outcomes cognitive abilities and labor market outcomes is several decades old, there is growing evidence Cross-country evidence suggests that pover- of how socioemotional skills can have as large ty status can influence cognitive and socio­ an impact as cognitive skills, if not a larger one, emotional skills acquisition. The income status on various labor market outcomes.31 For exam- of a child’s family influences socioemotional ple, extraversion is positively correlated with skills acquisition during school and subse- earnings in a service economy such as the Unit- quently in later economic life.28 Youth from ed States.32 The relationships between socio- poor households are thus likely to have lower emotional abilities and labor market outcomes socioemotional abilities than youth from have also been found in many developing econ- wealthier households. Socioemotional abili- omies. 33 In Bangladesh, for instance, emotional ty influences the acquisition of cognitive skills stability has a consistently positive correlation and vice versa. Foundational cognitive skills with wages, but grit has a negative correlation, include literacy and numeracy and are crucial particularly among manufacturing workers and for the ability to attain higher-level cognitive nonprofessionals.34 and job-specific technical skills in low- and middle-income countries at least.29 There is In Pakistan, socioemotional skills, such as extra- also evidence that high cognitive test scores version, conscientiousness, openness, emotion- are associated not just with high cognitive skills al stability, and agreeableness, are correlated but also with socioemotional attributes such as with high education attainment (figure 6.12). high motivation.30 Both openness and extraversion are positive- ly correlated with monthly earnings, based on Cross-country evidence also suggests that correlations with data from the Pakistan Labor socioemotional abilities may be important Skills Survey 2013.35 These findings are con- for some labor market outcomes. While the sistent with evidence from other countries. For FIGURE 6.12  Socioemotional skills by education level in Pakistan Strength (1–4, with higher numbers indicating stronger traits) Extraversion * Conscientiousness* Openness* Emotional stability* Agreeableness* Grit Hostility Decisionmaking 4 3 Chapter 6 | Pakistan Human Capital Review 2 1 Less than High school Vocational Bachelor’s degree high school education or more Source: Ahmed et al. (2022), based on data from the Pakistan Labor Skills Survey 2013 (World Bank 2013). Note: The box-whisker plots describe the distribution. In a given box, the horizontal line in the box describes the median score, the edges of the box describe the interquartile range, the whiskers identify the range (excluding outliers), and dots outside the range represent statistical outliers. * indicates a personality trait. 168 example, US high school students with better TABLE 6.4  Heckman selection model results to illustrate the social skills and work habits complete more relationship between monthly wages and various characteristics years at school.36 Though these students also including personality traits and cognitive skills exhibit higher cognitive abilities, the grades Outcome equation Selection equation awarded by middle school teachers are affected Marginal effects by attendance, disruptiveness, and work hab- on monthly its.37 In rural China, even in poor and rural con- IHS-transformed earnings texts, socioemotional skills affect the education conditional on attainment.38 being observed Probability of Variable (selected) observed earnings Education (reference: no education) In Pakistan, most of the returns to socio-  Primary 0.283 (0.293) 0.009 (0.041) emotional skills are not statistically significant,  Secondary 0.459* (0.276) −0.142*** (0.0361) except for agreeableness (table 6.4). After cog- nitive skills (measured by Raven scores), edu-  Postsecondary 1.363*** (0.325) 0.049 (0.066) cation attainment, parental education, wealth, Parental education, years work experience, nature of the employment,  Father 0.003 (0.029) −0.005 (0.004) access to finance, and location are controlled  Mother 0.105*** (0.036) −0.011 (0.008) for, only the impact of agreeableness remains Raven total score 0.023** (0.011) 0.001 (0.001) statistically significant. However, it is associated Male 1.092** (0.468) 1.079*** (0.035) with negative returns to earnings. The impact Age 0.159*** (0.057) 0.074*** (0.008) of agreeableness on earnings may vary by cul- Age-squared −0.002** (0.001) −0.001*** (0.0001) ture. For example, Japanese men who reported Urban 0.344 (0.235) −0.105*** (0.030) a higher level of agreeableness earned more, Personality intensity whereas American men with a higher level of agreeableness experienced a penalty.39 How-  Extroversion 0.064 (0.205) 0.009 (0.027) ever, in both countries, agreeableness had a  Conscientiousness −0.087 (0.209) 0.034 (0.027) positive impact on earnings among individuals  Openness 0.033 (0.187) −0.045* (0.025) working in companies with more than 1,000 em-  Emotional stability −0.042 (0.186) −0.018 (0.027) ployees. This also suggests that agreeableness  Agreeableness −0.313* (0.189) 0.007 (0.027) improves job performance and productivity  Grit −0.013 (0.221) 0.117*** (0.029) directly rather than indirectly through occupa-  Hostility −0.159 (0.142) −0.034* (0.020) tional choice. Because most of the jobs in Pa-  Decisionmaking 0.223 (0.204) −0.005 (0.028) kistan are created by small and medium enter- Agricultural land ownership −0.062 (0.047) 0.080* (0.046) prises, it is possible that personality traits linked Having at least one disability 0.032 (0.040) −0.041 (0.049) with competitiveness may be valued more than Currently married −0.119** (0.057) 0.153*** (0.049) Chapter 6 | Pakistan Human Capital Review agreeableness. Household with a bank account 0.035 (0.034) −0.045 (0.043) There is a strong association between extra­ Observations 3,333 3,333 version and openness to earnings for the Heckman is valid; Wald test (rho = 0): (Prob > chi2) < 0.05) Yes Yes self-employed in Pakistan (see table 6.4). This finding is consistent with the intuition that Source: Reproduced from Ahmed et al. (2022). personal relationships are important for en- Note: The dependent variable is IHS-transformed monthly wage in local currency units. Mar- trepreneurial activities. More generally, socio- ginal effects are repohisd. IHS is inverse hyperbolic sine. Numbers in parentheses are stan- dard errors; *** p < .01, ** p < .05, * p < .1. emotional skills are mostly statistically insignif- icant for wageworkers in formal and informal sectors as well as for the self-employed. 169 IMPROVING LABOR MARKET and education and no effects on cognitive and OUTCOMES­—­ECONOMIC INCLUSION socioemotional skills.44 Bangladesh’s second- PROGRAMS AND BEYOND ary education stipend program, which has been documented to have had substantial effects on Social safety nets can help, but female education outcomes, was also found do not necessarily improve labor to have mixed impacts on female labor market market outcomes outcomes. The program’s impacts on educa- tion were associated not only with an increased Globally, cash transfer programs focusing on the female LFPR but also with lower wages, as the poorest segments of the income and wealth dis- stipend beneficiaries often ended up in low- tribution can help support consumption when productivity self-employment.45 An evaluation appropriately targeted. Substantial research of the well-known and long-running PROGRE- has observed that such programs reduce the in- SA CCT in Mexico found that the cash transfers cidence of household poverty and its intensity.40 could not be associated with any evidence of im- Cash transfer programs have also been found to provements in wages, employment, or intergen- improve human development indicators, such erational mobility.46 as education, health, and nutrition.41 Beyond cash transfers stand other Pakistan already has poverty- and gender-­ entry points for improving labor targeted cash transfer programs. The BISP is market outcomes the country’s flagship unconditional cash trans- fer program targeting poor women. In addition, CCTs are useful core, “anchor” interventions for there are two conditional cash transfer (CCT) complementary services, but to have an impact programs: the education-focused Waseela-e- on labor market outcomes, notably productivity, Taleem, and a health and nutrition CCT program they need to be paired with other interventions. for mothers and children up to age 2. The one- Cash transfers made without an explicit employ- time Emergency Cash program was launched ment focus tend to result in little to no change in after the COVID-19 pandemic hit Pakistan, to labor market outcomes.47 In contrast, transfers mitigate economic hardship and risks of hunger, made explicitly for job search assistance or busi- and the Tahafuz program is for those who have ness startup tend to increase adult labor supply suffered a catastrophic event. There is some ev- and earnings. The likely main channels are the al- idence that the BISP has supported consump- leviation of liquidity and risk constraints. tion.42 Beyond consumption, impact evaluations have shown that such transfers in Pakistan have Programs that augment cash transfers with improved the intrahousehold decisionmaking other services are now being referred to as Chapter 6 | Pakistan Human Capital Review power of women, primarily by increasing the “cash-plus” programs. Worldwide, such pro- resources under their control and by leading to grams are increasingly packaging services for greater coverage of women through the Com- greater impact. Of youth employment programs puterized National Identity Card system.43 spanning skills development, entrepreneurship support, subsidized employment, public infor- However, standalone cash transfer interventions mation systems, and comprehensive programs, have mixed outcomes on labor market per- ones that integrate multiple interventions are formance, which of course is not their focus. A more likely to have positive impacts, especially cross-country review of CCT programs that had in low- and middle-income countries.48 substantial impacts on education outcomes and affected children both in utero and in their early The collection of programs with bundled serv- education found mixed impacts on employment ices often described as “economic inclusion 170 programs” include many that are anchored on or other productivity-enhancing measures (for cash transfers when targeting lower-income example, microfinance). These households groups.49 A recent meta-analysis of 97 impact may not be poor enough to need immediate evaluations of economic inclusion programs consumption support, but they tend to be vul- found that most programs with combined inter- nerable to shocks, are at risk of poverty, tend ventions had significant impacts, and not only to be liquidity constrained, and typically have on measures of livelihoods and poverty (figure limited precautionary savings. At the other end 6.13). This meta-analysis covered programs of the income distribution are richer and better- with a single anchoring intervention­—­whether educated households for which services such as cash transfers, livelihood support, or financial commercial finance, formal tertiary education, —­ inclusion­ but were augmented by complemen- and formal skills development may be appro- tary services. The latest generations of these priate. Still, financial inclusion and entrepre- cash-plus programs are increasingly incorpo- neurship support could be core interventions in rating innovative design features, such as psy- Pakistan for a package of services for the B40 if chosocial support services.50 (See the sections the services are targeted well and tailored to the “Moving beyond cash transfers” and “World specific needs of the poor. Bank–supported multipronged economic inclu- sion programs in the Sahel” in Annex 6B.) For households not poor enough to need cash transfers, other core services could be more The package of services needed varies by socio- relevant.52 They include such measures as seed economic status. In a framework that delineates capital, microfinance, and technical and busi- a continuum of services across the income dis- ness management skills training. Support to tribution, most households at the bottom of the the self-employed could include business man- distribution need consumption support.51 Mov- agement training, paired with other interven- ing up the distribution, nonpoor but still low- tions.53 For technical skills development, pro- income households in the informal sector may grams geared toward those with lower formal need other services such as social insurance education attainment could be the basis of a FIGURE 6.13  Economic inclusion programs boost not just livelihoods of the poor but also education, health, and nutrition outcomes Number of impact evaluations out of 97 reporting a statistically significant impact on a particular indicator Positive and signi cant impact No impact (not signi cant, with a positive or negative sign) Income and revenue 51 17 Assets 50 14 Chapter 6 | Pakistan Human Capital Review Consumption 37 15 Savings 37 8 Employment/income diversi cation 22 11 Women's empowerment 16 12 Child education 11 9 Psychosocial well-being 8 3 Child health and nutrition 10 5 0 20 40 60 80 100 Percent of studies reporting on given outcomes Source: Andrews et al. 2021. 171 core intervention. Pakistan’s entrepreneurship-­ poor women provides an important lesson given focused interventions tend to be geared toward that Pakistan has few  programs for women segments of the labor market that have far micro­entrepreneurs focused on poor women or higher levels of education attainment, are from those with less work experience and education households with greater access to services, or attainment. For example, the Pakistan Women are involved in small and medium enterprises. Entrepreneurship Program targets women with Important for developing a dynamic private sec- an existing business and 14 years of formal ed- tor no doubt, these programs are generally less ucation, making it (and similar programs) less relevant for the poor or the mass of the wage- relevant for the vast majority of the female employed, because only 12.4 percent of workers working-­age population. in Pakistan are employed in small and medium enterprises. CONCLUSION AND An apprenticeship approach for on-the-job learn- RECOMMENDATIONS ing, like the ustad-shagrid (master–­apprentice) model could be poverty-targeted and augmented Poorer households in Pakistan are trapped in with additional services (see the section on “The a vicious cycle of low human capital and poor Skills Training for Advancing Resources program labor market outcomes, the most serious of in Bangladesh” in Annex 6B). The program fea- which is low earnings. Boosting human capital tures an informal apprenticeship program that through means such as education attainment places youth in pairs under a master crafts per- would enhance earnings, but households must son who is usually an experienced shop-­ owner also increase their earnings to be able to have or worker within a trade. On average, these youth the resources to invest in themselves. are 16–17 years old and have had five or six years of education. The apprentices receive hands-on The review of evidence from several countries training five days a week for six months.54 In ad- suggests multipronged interventions, including dition to the workplace-based practical training, those anchored on cash transfer programs, can apprentices receive classroom-­ based soft-skills be powerful mechanisms to support the labor training once a week on issues including finan- market (and human capital) outcomes of the cial literacy, market assessment, and English-­ poor. But they must clearly identify the segment language skills. Once the training is completed, of the labor market that the program will target participants are linked with potential employers and coordinate the package of interventions to and are certified through the National Training be bundled together. and Vocational Qualifications Framework. The on-the-job training increased labor market par- Based on the analyses in this chapter and global Chapter 6 | Pakistan Human Capital Review ticipation of program participants by 23 percent- best practices, policy and programming recom- age points (up to 59 percent) and earnings by mendations for Pakistan may include: 44  percent through both self-employment and • Build the human capital of the poor, especial- wage employment.55 ly against shocks like the COVID-19 pandem- ic and climate change, through interventions Beyond the core elements of economic inclu- over the life cycle to ensure long-term liveli- sion programs, tackling additional constraints hood gains. Cash transfers and other safety that women face should be included with pur- net programs protect the human capital of poseful design, depending on individual women. the poor by supporting their consumption, The result of the randomized controlled trial of especially when faced with adverse shocks. Tunisia’s Community Works and Local Participa- When these programs are complemented tion program targeting highly marginalized and with additional services, often collectively 172 called economic inclusion programs, they consumption, and time use­ —­and its package can improve human capital investments, of services significantly increased incomes such as children’s education, health, and and revenues. The National Poverty Grad- nutrition, as well as household earnings. For uation Program, in addition to transfers for households not poor enough to receive cash consumption, packages coaching, business transfers, seed capital, microfinance, and capital, financial service facilitation, market technical and business management training links, and skills training. As poor households are more relevant. begin to expand or diversify their business • Ensure that economic inclusion programs that under the program, they become eligible combine safety nets with complementary serv- to apply for loans; asset beneficiaries and ices fit the needs of different segments of the loan recipients are eligible to receive basic poor and vulnerable. Productive inclusion pro- training to use the assets and loans effec- grams that go beyond cash can help improve tively. And the Kamyab Pakistan Program, household resilience to withstand shocks in launched in 2021, is using financial services the long term. For programs with a focus on for entrepreneurship­ —­rather than asset women, interventions must account for social transfers­—a­ s the core intervention for bun- norms, which have a major impact on wom- dling other services. en’s social interactions by restricting access • Leverage existing national systems for ef- to services or certain occupations deemed ficient and effective delivery of programs. not suitable for women. Possible interventions Pakistan has a strong starting point to de- to influence norms include strategic use of liver relevant and comprehensive services positive messaging about strong female role given its existing strategic policy initiatives, models and engaging to change norms sur- and its investments in systems. It has made rounding women’s economic activities. Sim- notable progress in building and strength- ilarly, employing women in public leadership ening the social protection system through positions can support the acceptance of am- the flagship BISP. It has invested in systems bitions and career aspirations among women. such as the National Socio-Economic Reg- Youth, women, and other marginalized groups istry and the poverty scorecard, as well as may need to engage in self-employment but a robust national identification system (civil may lack capital, technical skills, and experi- registration, digital and biometric, or voter ence running an enterprise. Depending on the identification). A key factor in the potential combination of needs they face, programs for effectiveness of support programs­ —­ to both them could include a combination of class- protect the human capital of the poor and room or even on-the-job learning (such as improve their productivity­ —­is whether the apprenticeships), classroom-based training services reach the intended segments of the Chapter 6 | Pakistan Human Capital Review in business management, socioemotional poor, by location and gender. While policy- skills development, and access to finance, as makers may still face key challenges­ —­such through collateral-­free microfinance. Youth as identifying the appropriate beneficiaries, entrepreneurship programs in other emerging the right mix of interventions for the target economies show that successful programs group, and the best mix of implementing pair business management training with other agencies in their design of new programs­ —­ services such as access to capital and coun- leveraging these major strengths will stand seling to enhance noncognitive ability. them in good stead and provide a stable • Use existing programs as a basis for econom- and high-quality starting point for new pro- ic inclusion interventions. An asset transfer– grams, whether implemented by federal or based pilot in Sindh spanned assets, finance, provincial entities. 173 NOTES 1. Based on World Bank (2020a). Pakistan’s HCI value 20. Farooq and Ahmed 2007; Farooq, Ahmed, and Ali is 0.41, putting it in the bottom quintile of countries. 2008. The methodology for the HCI is documented in 21. Pakistan’s high rates of informality are not uncom- Kraay (2018, 2019). mon in developing countries. Close to 90 percent of 2. Estimated from the Household Integrated Economic all employment in South Asia is informal (ILO 2018; Survey 2018–19. Loayza and Meza-Cuadra 2018). 3. See Azevedo et al. (2020) for elasticities. Geven and 22. World Bank 2020b. Hasan (2020) note that the high elasticity is driven 23. World Bank 2020b. by several socioeconomic factors, including high 24. World Bank 2020b. poverty levels, which lead families to push their chil- dren into labor or marriage from an early age, and 25. World Bank 2020b. the cost of private schooling. For OOP health spend- 26. Tran, Tran, and Nguyen 2020. ing, Khalid et al. (2021) estimated that expenditures 27. Wei et al. 1999. increased with increasing wealth, decreased with 28. Fletcher and Wolfe 2016; Kautz et al. 2014. increasing household size, and differed by province and region. 29. Valerio et al. 2016. 4. Based on 2018 data from the World Development 30. Brunello and Schlotter 2011. Indicators. For comparison, the South Asia average 31. Deming 2017; Heckman, Stixrud, and Urzua 2006. is 62.4 percent, and the middle-income country av- 32. Fletcher 2013. erage is 36.5 percent. OOP spending is the dominant mode of healthcare financing in developing countries 33. This literature is reviewed in Nomura and Adhikari (Malik and Syed 2012). In Pakistan, it is 67 percent (2017). of total health spending. Analysis of determinants 34. Nomura and Adhikari 2017. of OOP health spending is a key aspect of equity in 35. World Bank 2013. healthcare financing, as it helps formulate effective 36. Lleras 2008. health policy. Evidence on OOP spending in Pakistan is sparse, but there is evidence that economic status 37. Farkas 1996. and number of old-age family members are signifi- 38. Glewwe, Huang, and Park 2017. cant, positive predictors of OOP spending. 39. Lee and Ohtake 2016. 5. Rehman, Shaikh, and Ronis 2014. 40. See, for instance, Ferreira et al. (2009) and Moly- 6. Feng, Lagakos, and Rauch 2018. neux, Jones, and Samuels (2016). 7. Fields 2004. 41. Afzal, Mirza, and Arshad 2019; Cotto 2018; Ferreira 8. Amir et al. 2018. et al. 2009. 9. Solotaroff et al. 2019. 42. Afzal, Mirza, and Arshad 2019; Amrin and Ashfaq 2020; Iqbal, Padda, and Farooq 2020; Nayab and Fa- Chapter 6 | Pakistan Human Capital Review 10. Amir et al. 2018. rooq 2014; Shehzad 2011. 11. Cho and Majoka 2020. 43. Ambler and de Brauw 2017. 12. Shaheen, Sial, and Awan 2011. 44. Millán et al. 2019. 13. Sarwar and Abbasi 2013. 45. Shamsuddin 2015. 14. Amir et al. 2018. 46. Rodriguez-Oreggia and Freije 2012. 15. World Bank 2022, forthcoming. 47. Baird, McKenzie, and Özler 2018. 16. Matsuda et al. 2019. 48. Kluve et al. 2019. 17. World Bank 2022a. 49. World Bank (2021) defines economic-inclusion pro- 18. Tas et al. 2021. grams as those that include a combination of cash 19. See Annex 6A for definitions of key concepts. or in-kind transfers, skills training, coaching, access 174 to finance, and links to market support. 50. Bossuroy et al. 2022. and counseling to enhance socioemotional ability 51. Guven, Himanshi, and Joubert 2021. (Honorati 2015; McKenzie 2015). Traditional busi- ness training programs in other countries that teach 52. Andrews et al. 2021. financial recordkeeping, marketing, and other busi- 53. Youth entrepreneurship programs from other ness management skills have been found to yield countries, such as Kenya and Nigeria, show that mixed results (McKenzie and Woordruff 2014). successful programs pair business management 54. Bhattacharjee and Kamruzzaman 2016; Das 2018. training with other services like access to capital 55. Das 2018. REFERENCES Afzal, A., N. Mirza, and F. Arshad. 2019. “Conditional vs Baird, S., D. McKenzie, and B. Özler. 2018. “The Effects Unconditional Cash Transfers: A Study of Poverty Demo- of Cash Transfers on Adult Labor Market Outcomes.” IZA graphics in Pakistan.” Economic Research/Ekonomska Journal of Development and Migration 8 (1): 1–20. Istraživanja 32 (1): 3360–3377. Bhattacharjee, A., and M. Kamruzzaman. 2016. “Towards Ahmed, S.A., P. Chen, Z. Majoka, and J. Saha. 2022. “What Employability and Better Livelihood: An Evaluation of Is the Role of Socio-emotional Skills in Supporting South BRAC’s Skills Development Initiative.” Skills Development Asia’s Informal Sector Poor? Insights from Pakistan and Working Paper 1, BRAC Institute of Governance and De- Sri Lanka.” Chapter 7 in Bussolo, M. and S. Sharma (eds). velopment, Dhaka. 2022. Rethinking Informality in South Asia. Washington, Bossuroy, T., M. Goldstein, B. Karimou, D. Karlan, H. Ka- DC: World Bank. zianga, W. Parienté, P. Premand, C. C. Thomas, C. Udry, J. Ambler, K., and A. De Brauw. 2017. “The Impacts of Cash Vaillant, and K. A. Wright. 2022. “Tackling Psychosocial Transfers on Women’s Empowerment: Learning from Pa- and Capital Constraints to Alleviate Poverty.” Nature 605: kistan’s BISP Program.” Social Protection and Labor Dis- 291–297. cussion Paper 1702, World Bank, Washington, DC. Brunello, G., and M. Schlotter. 2011. “Non-cognitive Skills Amir, S., A. Kotikula, R. P. Pande, L. L. Y. Bossavie, and U. and Personality Traits: Labor Market Relevance and Their Khadka. 2018. “Female Labor Force Participation in Paki- Development in Education & Training Systems.” Discus- stan: What Do We Know?” World Bank, Washington, DC. sion Paper 5743, Institute for the Study of Labor, Bonn, Germany. Amrin, S., and M. Ashfaq. 2020. “The Impact of Benazir Income Support Program (BISP) on Household Food Ex- Cho, Y., and Z. Majoka. 2020. “Pakistan Jobs Diagnostic: penditures in Faisalabad.” Pakistan Journal of Agricultural Promoting Access to Quality Jobs for All.” Jobs Working Sciences 57 (1). Paper 20, World Bank, Washington, DC. Andrews, C., A. de Montesquiou, I. Arevalo Sanchez, P. V. Cotto, S. 2018. “The Effect of Cash Transfer Programs on Chapter 6 | Pakistan Human Capital Review Dutta, B. V. Paul, S. Samaranayake, J. Heisey, T. Clay, and S. Poverty Reduction.” Business and Public Administration Chaudhary. 2021. The State of Economic Inclusion Report Studies 12 (1): 1. 2021: The Potential to Scale. Washington, DC: World Bank. Das, N. C. 2018. “Training the Disadvantaged Youth and https://open​knowledge​.world​bank​.org/handle​/10986​ Labor Market Outcomes: Evidence from a Randomized /34917. Controlled Trial in Bangladesh.” Dhaka: BRAC. Azevedo, J. P. W. D., A. Hasan, D. Goldemberg, S. A. Iqbal, Deming, D. J. 2017. “The Growing Importance of Social and K. M. Geven. 2020. “Simulating the Potential Impacts Skills in the Labor Market.” Working Paper 21473, Nation- of COVID-19 School Closures on Schooling and Learning al Bureau of Economic Research, Cambridge, MA. Outcomes: A Set of Global Estimates.” Policy Research Working Paper Series 9284, World Bank, Washington, DC. Farkas, G. 1996. Human Capital or Cultural Capital? Eth- nicity and Poverty Groups in an Urban School District. New York: Routledge. 175 Farooq, S., and U. Ahmed. 2007. “Underemployment, Ed- Pakistan.” Pakistan Journal of Applied Economics 30 (2): ucation, and Job Satisfaction.” The Pakistan Development 153–180. Review 46 (4): 895–907. Kautz, T., J. J. Heckman, R. Diris, B. Ter Weel, and L. Farooq, S., U. Ahmed, and R. Ali. 2008. “Education, Under- Borghans. 2014. “Fostering and Measuring Skills: Improv- employment, and Job Satisfaction.” Pakistan Journal of ing Cognitive and Non-cognitive Skills to Promote Life- Commerce and Social Sciences 1: 83–91. time Success.” Working Paper 20749, National Bureau of Economic Research, Cambridge, MA. Feng, Y., D. Lagakos, and J. E. Rauch. 2018. “Unemploy- ment and Development.” Working Paper 25171, National Khalid, F., W. Raza, D. R. Hotchkiss, and R. H. Soelaeman. Bureau of Economic Research, Cambridge, MA. 2021. “Health Services Utilization and Out-of-pocket (OOP) Expenditures in Public and Private Facilities in Pa- Ferreira, F. H. G., A. Fiszbein, M. Grosh, K.N. Ellen, P. Olin- kistan: An Empirical Analysis of the 2013–14 OOP Health to, N. Schady, and E. Skoufias. 2009. Conditional Cash Expenditure Survey.” BMC Health Services Research 21 Transfers: Reducing Present and Future Poverty. Washing- (1): 1–14. ton, DC: World Bank Group. Kluve, J., S. Puerto, D. Robalino, J. M. Romero, F. Rother, J. Fletcher, J. M. 2013. “The Effects of Personality Traits Stöterau, F. Weidenkaff, and M. Witte. 2019. “Do Youth Em- on Adult Labor Market Outcomes: Evidence from Sib- ployment Programs Improve Labor Market Outcomes? A lings.” Journal of Economic Behavior & Organization 89: Quantitative Review.” World Development 114: 237–253. 122–135. Kraay, A. 2018. “Methodology for a World Bank Human Fletcher, J. M., and B. Wolfe. 2016. “The Importance Capital Index.” Policy Research Working Paper, 8593, of Family Income in the Formation and Evolution of World Bank, Washington, DC. Non-cognitive Skills in Childhood.” Economics of Educa- tion Review 54: 143–154. Kraay, A. 2019. “The World Bank Human Capital Index: A Guide.” The World Bank Research Observer 34 (1): 1–33. Geven, K., and A. Hasan. 2020. Learning Losses in Paki- stan Due to COVID-19 School Closures: A Technical Note Lee, S. Y., and F. Ohtake. 2016. “Is Being Agreeable a Key on Simulation Results. Washington, DC: World Bank. to Success or Failure in the Labor Market?” ISER Discus- sion Paper 960 (February), Institute of Social and Eco- Glewwe, P., Q. Huang, and A. Park. 2017. “Cognitive Skills, nomic Research, Osaka University, Osaka, Japan. Noncognitive Skills, and School-to-work Transitions in Rural China.” IZA Discussion Paper 10566, Institute of Lleras, C. 2008. “Do Skills and Behaviors in High School Labor Economics, Bonn, Germany. Matter? The Contribution of Noncognitive Factors in Ex- plaining Differences in Educational Attainment and Earn- Guven, M., J. Himanshi, and C. Joubert. 2021. “Social Pro- ings.” Social Science Research 37 (3): 888–902. tection for the Informal Economy: Operational Lessons for Developing Countries in Africa and Beyond.” World Bank, Loayza, N., and C. Meza-Cuadra. 2018. “A Toolkit for Infor- Washington, DC. mality Scenario Analysis.” Washington, DC: World Bank. http://www.worldbank.org/informality. Accessed 3 March Heckman, J. J., J. Stixrud, and S. Urzua. 2006. “The Ef- 2022. Chapter 6 | Pakistan Human Capital Review fects of Cognitive and Noncognitive Abilities on Labor Market Outcomes and Social Behavior.” Journal of Labor Malik, A. M., and S. I. A. Syed. 2012. “Socio-economic Economics 24 (3): 411–482. Determinants of Household Out-of-pocket Payments on Healthcare in Pakistan.” International Journal for Equity in Honorati, M. 2015. “The Impact of Private Sector Intern- Health 11 (1): 1–7. ship and Training on Urban Youth in Kenya.” Policy Re- search Working Paper 7404, World Bank, Washington, DC. Matsuda, N., T. Ahmed, and S. Nomura. 2019. “Labor Market Analysis Using Big Data: The Case of a Pakistani ILO (International Labor Organization). 2018. Women and Online Job Portal.” Policy Research Working Paper 9063, Men in the Informal Economy: A Statistical Picture. Gene- World Bank, Washington, DC. va: International Labor Office. McKenzie, D. 2015. “Identifying and Spurring High- Iqbal, T., I. U. H. Padda, and S. Farooq. 2020. “Sustain- Growth Entrepreneurship: Experimental Evidence from able Impacts of Social Safety Nets: The Case of BISP in a Business Plan Competition.” Policy Research Working 176 Paper 7391, Impact Evaluation Series, World Bank, Wash- Shamsuddin, M. 2015. “Labor Market Effects of a Female ington, DC. Stipend Programme in Bangladesh.” Oxford Development Studies 43 (4): 425–447. McKenzie, D., and C. Woodruff. 2014. “What Are We Learn- ing from Business Training and Entrepreneurship Evalu- Shehzad, I. 2011. “Benazir Income Support Programme ations around the Developing World?” World Bank Re- (BISP) and Its Impact on Women’s Empowerment.” search Observer 29 (1): 48–82. SAARC Journal of Human Resource Development 7 (1): 71–81. Millán, T. M., T. Barham, K. Macours, J. A. Maluccio, and M. Stampini. 2019. “Long-term Impacts of Conditional Solotaroff, J. L., G. Joseph, A. Kuriakose, and J. Sethi. Cash Transfers: Review of the Evidence.” The World Bank 2019. Getting to Work: Unlocking Women’s Potential in Sri Research Observer 34 (1): 119–159. Lanka’s Labor Force. Directions in Development: Coun- tries and Regions. Washington, DC: World Bank. Molyneux, M., W. Jones, and F. Samuels. 2016. “Can Cash Transfer Programmes Have ‘Transformative’ Effects?” Tas, E. O., T. Ahmed, N. Matsuda, and S. Nomura. 2021. The Journal of Development Studies 52 (8): 1087–1098. “Impacts of COVID-19 on Labor Markets and Household Well-Being in Pakistan: Evidence from an Online Job Plat- Nayab, D., and S. Farooq. 2014. “Effectiveness of Cash form.” Policy Brief, World Bank, Washington, DC. Transfer Programmes for Household Welfare in Pakistan: The Case of the Benazir Income Support Programme.” Tran, T. A., T. Q. Tran, and H. T. Nguyen. 2020. “The Role of The Pakistan Development Review 53 (2): 145–174. Education in the Livelihood of Households in the North- west Region, Vietnam.” Educational Research for Policy Nasir, M., and N. Faraz. 2020. “PIDE COVID-19 Bulletin No. and Practice 19 (1): 63–88. 13.” Lahore: Pakistan Institute of Development Economics. Valerio, A., M. L. Sanchez Puerta, N. Tognatta, and S. Mon- Nomura, S., and S. Adhikari. 2017. “The Influence of roy-Taborda. 2016. “Are There Skills Payoffs in Low- and Non-Cognitive Skills on Wages within and between Firms: Middle-Income Countries? Empirical Evidence Using Evidence from Bangladesh’s Formal Sector.” Policy Re- STEP Data.” Policy Research Working Paper 7879, World search Working Paper 8053, World Bank, Washington, DC. Bank, Washington, DC. Pakistan Bureau of Statistics. 2020. “Special Survey for Wei, X., M. C. Tsang, W. Xu, and L. K. Chen. 1999. “Educa- Evaluating Socio-Economic Impact of COVID-19 on Well- tion and Earnings in Rural China.” Education Economics 7 being of People.” (2): 167–187. Ray, J. 2021. “COVID-19 Put More Than 1 Billion Out of World Bank. 2013. Pakistan Labor Skills Survey 2013. Work.” Gallup, 3 May. https:/ /news.gallup.com/poll​/348722​ Washington, DC: World Bank. /covid​-put​-billion​-work​.aspx. Accessed 3 March 2022. World Bank. 2020a. The Human Capital Index 2020 Up- Rehman, A., B. T. Shaikh, and K. A. Ronis. 2014. “Health date: Human Capital in the Time of COVID-19. Care Seeking Patterns and Out of Pocket Payments for Children under Five Years of Age Living in Katchi Abadis World Bank. 2020b. “Islamic Republic of Pakistan: Lev- (Slums), in Islamabad, Pakistan.” International Journal for eling the Playing Field.” Systematic Country Diagnostic, Chapter 6 | Pakistan Human Capital Review Equity in Health 13 (1): 1–6. World Bank, Washington, DC. Rodríguez-Oreggia, E., and S. Freije. 2012. “Long Term Im- World Bank. 2021. Women in the Workforce in Peshawar: pact of a Cash-Transfers Program on Labor Outcomes of Pakistan Gender and Social Inclusion Platform and the the Rural Youth.” Working Paper 230, Center for Interna- Pakistan Poverty and Equity Program. Washington, DC: tional Development at Harvard University, Cambridge, MA. World Bank. Sarwar, F., and A. S. Abbasi. 2013. “An In-Depth Analysis World Bank. 2022a. Women in the Workforce in Quetta: of Women’s Labor Force Participation in Pakistan.” Middle Results from the Quetta Urban Household Survey. Wash- East Journal of Scientific Research 15 (2): 208–215. ington, DC: World Bank. Shaheen, S., M. Sial, and M. Awan. 2011. “Female Labor World Bank. 2022b. Swimming through Sand: Pakistan’s Force Participation in Pakistan: A Case of Punjab.” Journal Country Economic Memorandum. Washington, DC: World of Social and Development Sciences 2 (3): 104–110. Bank. 177 Annexes ANNEX 1A DATA AVAILABILITY Indicator PSLM HIES DHS LFS Years Expected years of school ✔ ✔ ✔ 1990, 1998, 2001, 2004–08, 2010–15, 2017–19 Under-5 mortality rate ✔ 1990, 2006, 2012, 2017 Prevalence of stunting ✔ 1990, 2012, 2017 Labor force participation rate ✔ 2000, 2002, 2004, 2006, 2008–09, 2011, 2013–15, 2018 Human Capital Index value ✔ ✔ 1990, 2012, 2017 Utilization-adjusted Human Capital Index value ✔ ✔ ✔ 1990, 2012, 2017 Note: All indicators can be disaggregated by province, age, gender, wealth, and urban–rural location. PSLM is Pakistan Social and Living Standards Measurement Survey (https:/ /www.pbs.gov.pk/content/pakistan-social-and-living-standards-measurement). HIES is Household Integrated Economic Survey (https:/ /www.pbs. gov.pk/form/data-information-request-form). DHS is Demographic and Health Survey. LFS is Labor Force Survey. ANNEX 1B DATA SOURCES AND METHODOLOGY FOR MEASURING HUMAN CAPITAL Data sources savings, liabilities, and consumption expendi- ture. Its data can be used to analyze consump- Multiple rounds of the Pakistan Social and Living tion patterns at the national and provincial lev- Standards Measurement Survey (PSLM), the els and by urban–rural location. Household Integrated Economic Survey (HIES), a Demographic and Health Survey (DHS), and DHS: DHSs are nationally representative house- the Pakistan Labor Force Survey (LFS) were hold surveys that collect information on a wide combined to calculate human capital outcomes. range of population, health, and nutrition indica- For learning-adjusted years of schooling, Paki- tors for monitoring and impact evaluation. stan’s assessment results from the 2019 Trends in International Mathematics and Science Study Labor Force Survey (LFS): A major aim of the Annexes | Pakistan Human Capital Review (TIMSS) were used. Data from the Human Cap- LFS is to collect a set of comprehensive statis- ital Project and the World Development Indica- tics on the various dimensions of the country’s tors were also used to benchmark outcomes. civilian labor force to pave the way for develop- ing skills, planning, generating employment, as- PSLM: The PSLM is a district-­level survey that sessing the role and importance of the informal collects information on key social indicators. It sector, and measuring the volume and charac- is the main source of multidimensional poverty teristics of employment. estimates. TIMSS: TIMSS, conducted by the International HIES: The HIES is a national household survey Association for the Evaluation of Educational that collects information on household income, Achievement, measures the mathematics and 178 science achievement of students across the alone complete the full cycle of preprimary, pri- world and is typically conducted in grades 4 and mary, and secondary education that is the norm 8. In 2019, TIMSS’s seventh cycle was adminis- in more advanced countries. The time she does tered in 64 countries and eight benchmarking spend in school may translate unevenly into systems. Data for Pakistan cover grade 4.1 learning. When she reaches age 18, she carries with her lasting effects of poor health and nutri- Definitions of indicators tion in childhood that limit her physical and cog- nitive abilities as an adult. Expected years of school: The number of years that a child of school entrance age is expected Traditionally, human capital has been measured to spend at school, including years spent on using proxies for the quantity of education, such repetition. It is the sum of the age-specific en- as years of schooling. However, a large body of rollment ratios for primary, secondary, post- evidence built over the past decades has re- secondary nontertiary, and tertiary education.2 vealed other equally important dimensions of human capital. The quality of education­ ­ ot —n Learning-adjusted years of schooling: A measure just its quantity­—i­s an important predictor of of data quality calculated by multiplying expect- individual earnings. Health in the first years of ed years of school by the ratio of the most re- —­ life­ and perhaps even earlier­ —h­ as a long-term cent harmonized test score to 625, which corre- impact on individual productivity. Children who sponds to advanced attainment on the TIMSS are in good health have better labor market test.3 outcomes not only because they tend to spend more time in school and learn more but also be- Under-5 mortality rate: The probability of dying cause they have higher productivity for a given between birth and age 5, expressed per 1,000 level of education attainment. Beyond health live births.4 and cognition, skills have a strong association with better labor market outcomes. One of Stunting: The percentage of girls or boys under the most striking findings of the literature that 5 whose height-for-age is more than two stan- has improved understanding of human capital dard deviations below the median for the inter- during the first years of life is that all the dimen- national reference population age 0–59 months. sions of human capital are complementary and For children up to age 2, height is measured by start interacting with each other early. recumbent length. The Human Capital Index (HCI), developed as Labor force participation rate: The proportion part of the World Bank’s Human Capital Project, of the population age 15 and older that is eco- measures the amount of human capital that a —a nomically active­ ­ ll people who supply labor to child born today can expect to attain by her 18th Annexes | Pakistan Human Capital Review produce goods and services during a specified birthday, given the risks of incomplete education period.5 and poor health that prevail in her birth country. It highlights how investments that improve edu- Methodology behind human cation and health outcomes today will affect the capital measures productivity of the next generation of workers. The HCI measures current education and health Imagine the trajectory from birth to adulthood outcomes since they can be influenced by cur- of a child born in Pakistan today. There is a risk rent policy interventions to improve their quan- that the child may not survive to her fifth birth- tity and quality. Unlike traditional measures, day. Even if she does reach school age, there is the HCI quantitatively illustrates the key stag- a further risk that she will not start school, let es in the trajectory described above and their 179 consequences for the productivity of the next 0 to 1, and a value of x means that a worker of generation of workers, in a summary measure. the next generation will be only x × 100 percent The HCI consists of three components: as productive as she would have been under • Survival: This component reflects the fact the benchmark of complete education and full that children born today need to survive until health. Equivalently, the gap between x and 1 human capital accumulation through formal measures the shortfall in worker productivity education can begin. Survival is measured due to gaps in education and health relative to using the under-5 mortality rate. the benchmark. • Education: This component combines infor- mation on the quantity of education a child The components are combined into a single can expect to obtain by age 18 with a measure index by first converting them into contribu- of quality: how much children learn in school tions to productivity relative to the benchmark based on countries’ relative performance on of complete education and full health. Multiply- international student achievement tests. By ing these contributions to productivity together adjusting for quality, this component reflects yields the overall HCI value: the reality that children in some countries learn far less than those in other countries, HCI = Survival × Education × Health. despite being in school for a similar amount of time. This measure is the same as learn- For survival, the relative productivity interpreta- ing-adjusted years of schooling. tion is stark: children who do not survive child- • Health: This component uses two indicators hood never become productive adults. As a re- of a country’s overall health environment: the sult, expected productivity as a future worker of prevalence of stunting among children under a child born today is reduced by a factor equal 5 and the adult survival rate, defined as the to the survival rate, relative to the benchmark proportion of 15-year-olds who will survive where all children survive: until age 60. The first indicator reflects the health environment experienced during pre- 1 – Under-5 Mortality Rate Survival = . natal, infant, and early childhood develop- 1 ment; the second reflects the range of health outcomes that a child born today may experi- The benchmark of complete high-quality educa- ence as an adult. tion corresponds to 14 years of school and a har- monized test score of 625. The relative produc- The education and health components of tivity interpretation for education is anchored human capital described above have intrinsic in the large empirical literature measuring the value that is undeniably important but difficult returns to education at the individual level. A to quantify. This in turn makes it challenging to rough consensus from this literature is that Annexes | Pakistan Human Capital Review combine them into a single measure. One solu- an additional year of school raises earnings by tion is to interpret each component in terms of about 8 percent. The parameter φ = 0.08 mea- its contribution to worker productivity, relative sures the returns to an additional year of school to a benchmark corresponding to complete and is used to convert differences in learn- education and full health. Accordingly, the HCI ing-adjusted years of schooling across countries transforms each component into a measure of into differences in worker productivity: this contribution. The HCI uses returns to edu- Harmonized Test Score cation and health to convert the related indica- Education = eφ(Expected Years of School × 625 . – 14) tors into worker productivity. The resulting over- all index measures the productivity of a worker Compared with a benchmark where all children relative to this benchmark. The HCI ranges from obtain a full 14 years of school by age 18, a child 180 who obtains only 10 years of education can ex- The utilization-adjusted Human Capital Index pect to be 32  percent less productive as an (UHCI) adjusts the HCI for labor market un- adult (a gap of 4 years of education, multiplied derutilization of human capital, based on the by 8 percent a year). employment to population ratio (the fraction of people age 15–65 who are employed) or the bet- For health, the relative productivity interpreta- ter employment rate (the fraction of people who tion is based on the empirical literature measur- are in the type of job where they might be better ing the economic returns to better health at the able to use their skills and abilities to increase individual level. With the HCI, the estimated con- their productivity): tributions of health to worker productivity are based on the prevalence of stunting and adult UHCI (Basic) = ETP × HCI survival rate averaged together. The contribu- UHCI (Full) = BER × HCI + (1 – BER) × HCImin tion of health to productivity is expressed rela- tive to the benchmark of full health, defined as where ETP is the employment to population the absence of stunting and a 100 percent adult ratio, BER is the better employment rate, and survival rate: HCImin is the productivity of raw labor (with 0 years of schooling, 100 percent stunting, and 0 Health = e(γASR × (Adult Survival Rate – 1) + γStunting × (Not Stunted Rate – 1))/2. chance of adults surviving to age 60), or 0.2. Source: World Bank 2020b. Annexes | Pakistan Human Capital Review 181 ANNEX 2 METHODOLOGY FOR THE PHONE SURVEY ON IMPACTS OF THE 2022 FLOODS The data on the impacts of the 2022 floods sample was stratified by gender of the child. come from a nationally representative phone The survey also oversampled households that survey carried out by Gallup Pakistan from Sep- reported any impact of floods, aiming for a sam- tember 29 to October 28, 2022. Random digit ple of at least 1,000 households that suffered dialing of mobile phones was used to reach effects of floods in their area. The survey collect- active numbers across Pakistan from all four ed information on the education status of chil- telecom providers. Approximately 93  percent dren, food security, child work, health, environ- of households have access to a mobile phone.6 ment, and household composition. A limited set Despite the high penetration of mobile phones, of sociodemographic characteristics was also lower-income households are overrepresented collected, including education of parents, as- in the remaining 7  percent of households that sets, gender, family composition, rural or urban do not have access to mobile technology. Thus area, district, and province. the results likely underestimate the true im- pacts of the floods on families and their children The survey randomly called 40,800 numbers, in Pakistan. reaching 15,750 individuals who answered the phone, 5,420 of whom agreed to the interview, Each random number was called until a call and 4,044 of whom completed the survey. At was answered (with a maximum of three at- the beginning of the survey, enumerators of- tempts). To maximize the response rate, calls fered a PKR 200 phone credit for completing were placed at different times on different days the survey to 75 percent of those who answered of the week. Once individuals were contact- the phone (selected randomly). To better cap- ed by mobile phone, consent was obtained, a ture the aggregate impact, weights were created screening questionnaire was administered, and for the selection of province, urban–rural, sex, a unique study identification number was gener- and education of household head. ated for each respondent. Interviewers entered data into a tablet with Survey CTO software that Phone surveys are known to miss households at had the preloaded questionnaire with automat- the bottom of the income distribution because ic skipping patterns (computer assisted phone those households often lack mobile phones or interviewing). live in hard-to-reach areas (where mobile phone coverage is limited). Further, flood damage may The target population of the survey was parents have exacerbated those issues. In most instanc- or caregivers of children age 3–17. If more than es, the results likely underestimate the magni- one child lived in the household, one child was tude of effects given the characteristics of those Annexes | Pakistan Human Capital Review randomly selected as the subject of inquiry. The who might not be reachable via phone. 182 ANNEX 3 DESCRIPTION AND DEMOGRAPHICS OF RESPONDENTS TO THE GALLUP SURVEY Gallup Pakistan was commissioned to con- whereas the actual composition of households duct a phone survey from November 2021 to in Pakistan is more urban (68.6  percent) than February 2022. The process relied on random rural (31.4 percent). Monthly household expen- digit dialing of mobile phones using all four diture distribution in the sample ranged from telecom providers with active numbers across PKR 1,000 (US$5.65) to PKR 90,000 (US$509), the country. (Landline penetration is extreme- with a median of PKR 25,000 (US$142). In 2019, ly limited in Pakistan, at less than 3  percent, based on PSLM data, median household expen- whereas 90 percent of households use a mobile diture was PKR 21,750 (US$122). In the month phone, according to the 2019–20 Pakistan So- prior to the survey, 34.6 percent of respondents cial and Living Standards Measurement Survey had not worked for pay, including 26.9 percent [PSLM].) Respondents were caregivers age 18 of men and 63 percent of women, and a major- and older in households with at least one child ity of households had lost income during the age 72 months or younger. Roughly half the se- COVID-19 pandemic, discussed further below. lected children were in the 0–35 months (0–2 years) age group, and roughly half were in the Mothers, fathers, and other caregivers took part 36–72 months (3–5 years) age group, inclu- in the survey. Respondents included parents sive of children age 72 months (that is, who just (70 percent) and other caregivers (30 percent). turned 6 years old). Gender-wise, the sample More specifically, 16  percent were mothers, largely reflects the population (55.9  percent 54 percent were fathers, 8 percent were grand- boys in the sample versus 52.2  percent boys parents, and 22 percent were other caregivers. in the population and 44.1 percent girls in the Overall, 56 percent of respondents were male, sample versus 47.8 percent girls in the popula- and 44 percent were female. tion). Of 3,907 potential respondents reached, 448 declined to be interviewed (11  percent), Respondents were asked about their edu- 433 (11  percent) partly completed the inter- cation and water and sanitation conditions. view, and 3,021 (77 percent) fully completed the About 46 percent of mothers and 28.9 percent interview. of fathers had not completed primary school, 20.2 percent of mothers and 21.5  percent of All areas and types of families were represent- fathers had completed primary school, and ed in the survey. Households in all provinces 33.8 percent of mothers and 49.6 percent of fa- and territories across Pakistan were represent- thers had completed secondary school or high- ed: Punjab (n = 1,301), Sindh (n = 425), Khyber er. Most households (90.2 percent) had a toilet Pakhtunkhwa (n = 903), Balochistan (n = 223), in the house. While most had access to an im- Annexes | Pakistan Human Capital Review and Islamabad Capital Territory and other areas proved water source, less than half (42 percent) (n = 169). The share of respondents from Sindh had piped water inside the house. The rest had was smaller than the province’s share of the total access to bore water (22.2  percent), a neigh- population (14.1 percent versus 24.5  percent), borhood tap (12.4 percent), well water (9.2 per- whereas the share from Khyber Pakhtunkhwa cent), or bottled or sachet water (9.5 percent); was larger (29.9  percent versus 19.1  percent). 4.8 percent of the sample relied on river water The shares of respondents from Punjab and as their primary source of drinking water. Balochistan were within 4 percentage points of each province’s share of the national popula- To measure child development for children tion. The sample included more rural (57.6 per- under 3, the Caregiver-Reported Early Develop- cent) than urban (42.4  percent) households, ment Instrument (CREDI) short form was used. 183 CREDI is a globally validated, population-­ level, child. Items represent developmentally appro- developmental assessment tool for children age priate skills aligned with the rapid changes that 0–3. It has been used in more than 17 countries occur from birth to age 3, and forms are stan- and is designed to be culturally and linguistical- dardized within subgroups by age. Therefore, a ly neutral. CREDI captures development across question about social-emotional development four domains, motor skills, language, cogni- looks different for a 6-month-old child than a tion, and social-emotional development. The 30-month-old child, and so forth. The current short form questionnaire has 20 items to be survey provides the first CREDI measurements completed by parents or caregivers about their taken in Pakistan. Annexes | Pakistan Human Capital Review 184 ANNEX 4A TACKLING CHILD MALNUTRITION: A REVIEW OF POLICIES AND STRATEGIES Pakistan’s development blueprint, Pakistan Vi- spending and planning. It targets socioeconomi- sion 2014–2025, takes a comprehensive view cally vulnerable populations, pregnant and nurs- of the country’s development challenges across ing mothers, adolescent girls, and children up sectors. Nutrition is also highlighted in overarch- to age 2 and prioritizes areas with high stunting ing provincial development visions. and other malnutrition indicators. Geographic convergence of the multisectoral interventions Emergence of the multisectoral is a key guiding principle. The PMNS tasks the approach federal government with coordination and mon- itoring as provinces implement their nutrition The Pakistan Intersectoral Nutrition Strategy strategies. 2011 was seminal in framing nutrition as a multi- sectoral issue, especially at the planning level. Nutrition policies and strategies in It emphasized the roles of health, social protec- the health sector tion, agriculture (including livestock and fisher- ies), water and sanitation, education, and family National health vision documents advocate for planning in contributing to nutrition. an integrated and multisectoral approach to health and nutrition. The National Integrated Provincial nutrition strategies, which are rooted Reproductive, Maternal, Newborn, Child and in the local context, attempt to improve nutri- Adolescent Health and Nutrition (RMNCAH&N) tion through the convergent action of multiple Strategy 2016–2020 and the provincial work- sectors, expanding nutrition actions beyond plans following the strategy set the stage to the health sector. The strategies emphasize mainstream nutrition services within existing geographic convergence to varying degrees. health programs. Sindh’s strategy most comprehensively plans for it, prioritizing districts with a high burden of Provinces have operationalized the integration stunting, with all sectoral interventions converg- into their health programs and projects to differ- ing at the union council level. In Punjab, health ent degrees. Khyber Pakhtunkhwa and Punjab and water, sanitation, and hygiene interventions have integrated their vertical health programs, target high-risk populations but with likely diver- including the Expanded Programme on Immu- gent geographic rollout. Khyber Pakhtunkhwa’s nization; Maternal, Newborn, and Child Health strategy proposes nutrition interventions and Nutrition; and the Lady Health Worker Pro- through health and education in all high-burden gram. Sindh has mainstreamed nutrition into districts and a “model district” approach where health sector programs as part of its Nutrition Annexes | Pakistan Human Capital Review nutrition interventions related to agriculture, Support Program in nine districts, and in 2020, livestock, and fisheries converge. it integrated eight vertical health programs to operationalize the strategy. Sindh’s integrated Based on the provincial strategies, the compre- RMNCAH&N strategy is also a part of its Ac- hensive Pakistan Multisectoral Nutrition Strat- celerated Action Plan for Reduction of Stunting egy 2018 (PMNS) was developed. The strategy and Malnutrition. sets clear, achievable goals for nutrition and emphasizes nutrition-sensitive programming, The National Health Action Plan 2019–2023 and including advocating for all sectors to adopt individual provincial health sector reform strate- a “nutrition lens” in planning. It also proposes gies emphasize nutrition. The national plan out- common results frameworks to guide provincial lines health sector priorities, envisages universal 185 access to high-quality healthcare, and sets tar- From policies to programs to impacts gets to reduce stunting and malnutrition. How- ever, it does not adequately link to the nutrition Pakistan has a comprehensive enabling policy strategies, particularly the multisectoral ap- environment to provide direction and guidance proach, nor does it specify the implementation for nutrition actions and programs. It does not and oversight arrangements needed to achieve have a nutrition policy, but several vision and its nutrition targets. strategy documents at the federal and provin- cial levels are used as policy equivalents. These With devolution, provincial governments started documents were developed by engaging stake- to steer the health sector through independent holders across provinces, sectors, civil society, strategies. They developed these strategies in and academia as well as development and nu- response to challenges related to service deliv- trition partners. They were also informed by the ery quality and coverage (especially of poor and global evidence base of effective interventions vulnerable people), health workforce compe- and consider direct or nutrition-specific inter- tence, and health sector governance and regu- ventions implemented through the health sector lation. Most of the strategies identified nutrition and nutrition-related interventions by nonhealth as important within health, and each adopted an sectors that can improve nutrition outcomes. integrated approach through essential health services packages at all levels. The strategies Most of the strategies outline clear goals and clearly identified the macro issues and goals of objectives and explicitly prioritize vulnera- the health sector, but few presented monitoring ble groups. In several instances, however, the targets or detailed specific implementation or targets appear either too ambitious­ —a ­lmost targeting methods. They guided positive steps aspirational­—o­r too modest and overly cau- toward reforming the health sector, but struc- tious. The strategies are complemented by tural and operational challenges limited their guidelines that further define and provide tech- impact. nical and operational guidance on a specific topic or intervention­ ­ uch as food fortification —s Policies and strategies in nonhealth sectors and dietary guidelines. Going forward, a phased vary in their attention to nutrition. Although approach­ —­prioritizing the most vulnerable several sectoral actions in and of themselves­ people with the highest-impact integrated and —­ for example, educating women, keeping girls co-located interventions­ may yield the largest —­ in school longer, making communities free of impact in the shortest timeline. open defecation, and providing safe water­ —­ contribute to better nutrition outcomes, a co- The strategies and their development process ordinated and convergent program, focusing have introduced sensitization and built commit- on actions that leverage each sector’s vantage ment for nutrition at the national and provincial Annexes | Pakistan Human Capital Review points, is crucial to reap the full benefits of levels and provide an enabling framework. How- the multisectoral approach. To create such a ever, for several years they remained mostly on program, the sectoral policies and strategies paper, and several still do; operationalization must explicitly identify their sector’s roles, ac- into programs has lagged, remaining weak and tions, coordination mechanisms, and resource limited. It will take commitment, stewardship, requirements to contribute to nutrition out- capacity, and financial and human resources to comes, along with meaningful indicators to implement the strategies at scale and harness track progress. their full benefit to improve nutrition outcomes. 186 ANNEX 4B TACKLING MALNUTRITION: A REVIEW OF PROGRAMS National nutrition programs and for therapeutic foods), ensuring emergency and projects inpatient services for severe acute malnutrition at nutrition stabilization centers, and building While devolution provided provinces with pro- system capacity and protocols to respond to gram leadership opportunities, including in emergency nutrition situations. nutrition, national nutrition programs remain important. National programs are being rolled TMIS champions pooled procurement and ad- out across the country, focusing on provinces vocacy for legislative change for the Food Forti- and territories with poorer nutrition outcomes. fication Programme (FFP). It leverages existing These programs­ —­providing mainly financial, social mobilization structures such as mothers’ technical, and oversight support for direct nu- groups to trigger behavior change at the house- —­ trition interventions­ include donor-funded pro- hold level and increase demand for nutrition grams such as micronutrient supplementation services. The program proposes implementa- and food fortification programs, federal govern- tion arrangements at multiple levels with clear ment–funded programs such as the Benazir roles and responsibilities. The Federal Nutrition Income Support Program (BISP) Nashonuma Wing coordinates among stakeholders; provides and Tackling Malnutrition Induced Stunting in technical assistance and guidance; builds ca- Pakistan (TMIS), continued federal support for pacity; prepares and implements the monitor- health and nutrition programs in territories and ing and evaluation framework; advocates with regions where responsibilities were not devolved policymakers; oversees a mass media cam- or not fully devolved, nutrition support as part of paign; establishes nutrition guidelines, frame- the response to natural or human-caused emer- works, and mechanisms; handles financial gencies, and stewardship for programmatic and procurement; and manages the supply chain. operational research and knowledge generation. Provinces and territories develop and review nu- trition plans and implement and monitor project TMIS includes evidence-based interventions interventions. Districts bear the responsibility in the 67 districts in Pakistan with the highest for implementation and coordination. prevalence of stunting, targeting the first 1,000 days of life (pregnant and nursing mothers and BISP Nashonuma provides conditional cash children under 2) and adolescents. The inter- transfers to pregnant and lactating women and ventions include nutrition counseling, supple- children under 2 from ultrapoor households in mentary feeding, food fortification, systems nine districts with high stunting. It incentivizes strengthening, increased coverage and capaci- the uptake of health and nutrition services such Annexes | Pakistan Human Capital Review ty of the Lady Health Worker Program (LHWP), as counseling; immunization; awareness ses- and capacity building. Close to 40  percent of sions on breastfeeding, sanitation, hygiene, and funds (most of the federal contribution) is ear- dietary diversity; antenatal care; and consump- marked for lipid-based nutrition supplements, tion of specialized nutritious foods. It awards ready-to-use supplementary food, and ready- a higher cash transfer for girls and, as of Sep- to-use therapeutic food. TMIS also supports tember 2021, had more than 66,000 registered clear referral mechanisms to avoid counterpro- beneficiaries. ductive competition between the LHWP and the Community Midwife Program. At the health fa- The LHWP is an important delivery platform cility level, TMIS supports updating guidelines, for nutrition services at the community level. It strengthening supply chains (including those provided primary healthcare outreach to about 187 54 percent of Pakistan’s population in 2019, par- for undernourished children up to age 5. These ticularly in rural communities and urban slums. interventions are delivered through the health Its mandate has evolved to include infant and sector. Food fortification efforts were initiated young child feeding; immunization; diarrhea in the 1980s with salt iodization but have not management; antenatal care; micronutrient been continuous. Strong technical and financial supplementation; and water, sanitation, and hy- support from donors, particularly the National giene. The LHWP is integrated with other health Fortification Alliance, has been instrumental in sector verticals in Khyber Pakhtunkhwa, Punjab, advocating for renewing, expanding, and institu- and Sindh, while in Azad Jammu and Kashmir, tionalizing the country’s two large food fortifica- Balochistan, and Gilgit-Baltistan it continues as tion initiatives­—­Universal Salt Iodization (USI) a separate program. Maximizing the program’s and the FFP. potential to deliver its nutrition mandate is im- portant. Lady Health Workers are typically re- USI, a long-standing initiative to protect people garded with respect and considered influential, from iodine deficiency disorders, demonstrates particularly in infant and young child feeding robust design features and responsiveness to practices,7 which are crucial to improving nu- implementation issues. Revitalized in 2002, the trition outcomes. There is evidence that pro- program was successfully piloted in 20 districts grammatic focus on specific functions has led and by 2012 covered all 110 salt-producing dis- to improvements (as with poliomyelitis8); thus, tricts in the country.10 It is slated to continue bolstering the status of nutrition within the pro- until at least 2025. USI has many strengths: its gram could have a marked impact. Enhancing design drew on lessons from similar initiatives in the focus on nutrition will entail building capac- Pakistan and across Asia; it provides technical ity and support, which can include offering in- assistance, support for equipment, fortificant centives for nutrition tasks.9 The nutrition indi- supplies, capacity building, advocacy for legisla- cators tracked by the program vary by province; tion, and public awareness activities to promote developing a core set of meaningful indicators consumption; it is attentive to the need for mon- to track the program’s progress and using the itoring, building in mechanisms both for quality information to drive future improvements would assurance and internal monitoring at the pro- be useful. Knowledge exchange can be especial- cessor level and for community-based monitor- ly valuable in key areas where provincial imple- ing; and it is attentive to implementation issues mentation differs. TMIS can play a valuable role and responsive in finding appropriate solutions. in strengthening these crucial areas to improve For example, when small- and medium-scale the LHWP’s nutrition performance. processors faced financial issues with purchas- ing the fortificant, a revolving fund was created Programs to address micronutrient to procure it, and it was repackaged into smaller deficiencies units that those processors could afford.11 Annexes | Pakistan Human Capital Review To address micronutrient deficiencies, Pakistan According to the 2011 and 2018 rounds of the has both large-scale food fortification programs National Nutrition Survey, household use of io- for its entire population and micronutrient sup- dized salt rose substantially at the national level­ plementation programs for specific groups. —­from 17.0  percent to 79.6  percent. However, Micronutrient supplementation interventions use remains much lower in Balochistan (60 per- include iron-folate and multiple micronutrient cent in 2018) and Khyber Pakhtunkhwa (31 per- supplementation for pregnant and nursing cent), and additional efforts to expand coverage women, vitamin A supplementation for children and consumption in those two provinces are re- age 6–59 months, and micronutrient powder quired. USI is not yet backed by the necessary sachets for all children age 6–23 months and legislation in every province,12 making adequate 188 iodization challenging and leaving program sus- eight of Balochistan’s drought-affected districts tainability at risk. TMIS incorporates objectives are receiving humanitarian responses to mal- and interventions to increase household con- nutrition: seven are serviced through a United sumption of iodized salt to 88 percent by 2026 Nations Children’s Fund (UNICEF) intervention and proposes providing iodized oil capsules in and one through a Central Emergency Response areas with high incidence of iodine deficiency Fund. Through UNICEF, emergency nutrition disorders and low supplies of adequately iodized support has also been extended to drought-af- salt. It is important that the federal and provin- fected areas in Khyber Pakhtunkhwa and cial governments invest in and support the USI Sindh. Through the National Disaster Manage- to ensure program continuity, expansion, and ment Agency, nutrition clusters for coordinated strengthening. responses to emergency situations already exist in Pakistan. TMIS includes additional measures The FFP aims to reduce micronutrient deficien- for integrating nutrition response during emer- cies by fortifying wheat flour and edible oil. It gencies, including building capacity for disaster supports fortifying roller mill wheat flour with risk reduction, preparedness, and coordination. iron, folic acid, zinc, and vitamin B12 and forti- fying edible oil/ghee with vitamins A and D. The The need to improve implementation program has set clear targets to reduce iron and vitamin A deficiency in women and children Pakistan has seen a concerted move in recent under 5, as well as targets for population cover- years to develop programs that deliver direct age of fortified wheat/flour and edible oil/ghee. nutrition interventions across all provinces and Its four key components to support fortification territories. While commitment to developing in- efforts­—­technical assistance to federal, provin- tegrated, multisectoral programs is strong and cial, and district governments for monitoring strategies and approaches for multisectoral pro- and enforcement; technical assistance and in- grams have been developed by all provinces, the centives to the relevant industries; consumer health sector remains the primary sector deliv- awareness and advocacy; and targeted studies ering nutrition interventions, and multisectoral to better inform implementation­ —­ reflect global programs are few. The new national TMIS pro- best practices in food fortification programs and gram, while committed to multisectoral conver- responsiveness to country needs. gence, is also rooted in the health sector. Going forward, a more multisectoral view will likely Emergency nutrition responses yield larger impacts on nutrition outcomes. across provinces The design, scope, and coverage of provincial To protect the nutrition status of the most vul- programs vary considerably, but overall, pro- nerable people during natural or human-caused grams and projects are designed well, align- Annexes | Pakistan Human Capital Review emergencies, including during the ongoing ing with national and provincial policies and COVID-19 pandemic, nutrition interventions are strategies. Projects and programs are typically incorporated as key elements of the govern- designed consultatively and draw on lessons ment’s response. In the affected areas nutrition from earlier programs and pilots. Most set clear, support efforts, including donor support, typi- measurable, and reasonable goals for reduc- cally focus on establishing outpatient therapeu- ing stunting and demonstrate a shift from only tic programs, screening children and improving screening for and treating severe acute malnu- the coverage of operational health sector nu- trition to incorporating approaches and inter- trition interventions (such as providing micro- ventions to prevent stunting. Most programs nutrient powder and iron and folic acid), and target adolescent girls, pregnant and nursing promoting breastfeeding. For example, at least mothers, and children under 2­ —­individuals at 189 the crucial times when appropriate nutrition in- vulnerable and at-risk people are not reached terventions can have the most impact, not only or are minimally reached by nutrition interven- during childhood but throughout life. Most pro- tions. TMIS aims to fully cover the districts with grams also target the most vulnerable and dis- the highest stunting, addressing this issue in a advantaged people. However, some areas with phased manner. ANNEX 4C ENSURING FEDERAL AND PROVINCIAL STEWARDSHIP OF NUTRITION Federal stewardship structures practice of confining nutrition to the health de- partment. Despite being overtaken by the devo- The recently established interministerial Paki- lution of health under the 18th Constitutional stan National Nutrition Coordination Council, Amendment, the strategy served as a strategic chaired by the prime minister, is the country’s framework for guiding provinces to define nutri- highest stewardship body. A first in the history of tion in their postdevolution development agen- Pakistan, the council was created to lead nation- das.14 The wing has been the chief coordinating al nutrition policy and strategy development, body for all stakeholders in nutrition, including build synergies, and ensure cross-country, provinces and development partners, setting cross-ministerial, intersectoral, and interprovin- technical guidelines that are aligned with inter- cial collaboration to prevent childhood stunting national commitments and global strategies,15 and malnutrition. It has four core functions­ —­ supporting their implementation with the help formulating strategy and policy; developing of development partners, and coalescing alli- human capital; monitoring and assessing im- ances and technical committees. It was instru- pact; and advocating and raising awareness­ —­ mental in developing TMIS, which is expected and will be supported by the Planning Com- to strengthen and better coordinate nutrition mission and the Ministry of National Health —­ leadership at all levels­ provincial, regional, and Services, Regulations and Coordination. Nutri- district. tion is a decentralized subject and a provincial responsibility, but this high-level stewardship at The National Nutrition Cluster (NC) coordi- the federal level will be critical to supporting and nates with national and international partners providing direction to provinces, leading syner- on agreed priorities to ensure coherent, strate- gistic and collaborative multisectoral policy and gic, and effective provision of nutrition services program action for nutrition.13 during emergencies. It focuses on coordination, capacity building, emergency preparedness, as- The Federal Nutrition Wing at the Ministry of sessment, and response for nutrition. It devel- Annexes | Pakistan Human Capital Review National Health Services, Regulations and Co- oped a common framework, the NC Prepared- ordination plays a pivotal role in moving the nu- ness and Response Plan, to guide the actions of trition agenda forward. As the dedicated home all partners in the nutrition sector against disas- for nutrition both before and after devolution, ter. It does not replace the need for planning by the wing has served as the primary advocate for individual agencies in relation to their mandate nutrition, including the development of multi- and responsibilities within clusters, but it pro- sectoral strategies. In 2009, with United Nations vides focus and coherence to the various plan- Children’s Fund support, it initiated the devel- ning levels that are required to respond effec- opment of the Pakistan Intersectoral Nutrition tively. It is designed to be a flexible plan that is Strategy 2011 to facilitate cross-sectoral action updated to incorporate lessons and strengthen on nutrition­—­a major shift from the previous future emergency response. In cooperation with 190 the Provincial Disaster Management Authority units have been established in each province’s and the Department of Health, provincial NCs planning and development department. are expected to develop their own preparedness and response plans aligned with the national All four provinces have stewardship bodies for plan.16 nutrition. While capacity for stewardship var- ies, each province has developed its own inter- Over time, several institutions­ alliances, tech- —­ sectoral nutrition strategy and stewardship and nical groups, advisory bodies, and steering institutional structures to support its imple- committees­ —­have been created for specific mentation and other nutrition programs. Each interventions and areas. Each group has a man- has a Nutrition Steering Committee with rep- date and has served a valuable purpose in draw- resentation by the sectors identified in the rele- ing attention to its issue area; advocating for it; vant Multi-sectoral Nutrition Strategies, chaired collaborating to develop appropriate strategies, by senior leadership­ —­by the chief minister in interventions, and guidelines for addressing Punjab, by the additional chief secretaries in it; and supporting and, to an extent, monitor- Balochistan and Khyber Pakhtunkhwa, and by ing implementation. They have worked closely the chairperson of the Planning and Develop- with the Federal Nutrition Wing and the provin- ment Board in Sindh. The steering committees cial nutrition stewardship institutions, helping are supported by technical working groups that bring in new knowledge and international good provide technical advice. Stewardship and insti- practices. tutional arrangements for planning, implemen- tation, coordination, and monitoring have also Provincial stewardship structures been set up at the district levels. Provincial food fortification alliances exist in all provinces. The devolution of nutrition to the provinces brought the role of provincial stewardship to the Federal-provincial coordination fore. Provinces’ previous role had been confined to delivering health services, but with devolution Formal coordination mechanisms are not de- they received greater autonomy, with responsi- fined, but the Federal Nutrition Wing lends bility for financing17 and stewarding the health technical support to provinces and coordinates and nutrition functions. A look at strategic op- periodic reviews. Joint annual reviews of pro- portunities for and barriers to action on nutri- vincial programs include a review of provincial tion in Pakistan’s four provinces in the first few strategies, with cross-learning by all provinces. years after devolution reveals that the provinces As part of the Ministry of National Health Ser- faced common constraints: lack of a compre- vices, Regulations and Coordination, the Fed- hensive policy and minimal allocations by the eral Nutrition Wing has key provincial counter- state, driven by the low visibility of nutrition, parts in departments of health, but this makes Annexes | Pakistan Human Capital Review siloed working of sectors, and weak coalitions. coordinating with the provincial nutrition units Provinces also faced the common contextual in the planning and development departments impediments of poverty, patriarchy, and inade- somewhat challenging. The technical bodies at quate health and water, sanitation, and hygiene the national and provincial levels coordinate as coverage.18 With increased attention to nutrition required. In Balochistan, Punjab, and Sindh the and with advocacy and support from donors provincial health departments coordinate the and UN agencies, provinces developed steward- activities of the technical working group; how- ship structures, including for both horizontal co- ever, in Khyber Pakhtunkhwa there is a techni- ordination and vertical integration. These struc- cal working group for each sector, and they in- tures have evolved in recent years, and nutrition volve public officials. 191 ANNEX 5 INTERNATIONAL EXPERIENCE IN REDUCING THE NUMBER OF OUT-OF-SCHOOL CHILDREN Reducing the number of out-of-school children Building schools­ —­or providing transport to in Pakistan requires understanding what works schools­ —i­ncreases enrollment. Tellingly, in Paki- to get more children into school and to improve stani villages with a secondary school for girls, retention, learning outcomes, and attainment there are twice as many educated women as in once they are there. This annex summarizes the villages without one.19 In areas with sufficient en- international literature on what works to improve rollment to populate new schools, building more schooling, focusing on interventions that have schools can increase enrollment. In Quetta, Bal- undergone rigorous evaluation through random- ochistan, opening new private girls’ schools in ized controlled trials and quasi-experimental poor urban neighborhoods led to a 33 percent studies. Both supply- and demand-side issues increase in enrollment.20 Pakistan’s experience limit children’s access to education. is not unique: constructing new schools raised enrollment rates and learning outcomes for all Supply-side issues students, especially girls, in rural areas of Af- ghanistan and boosted school attendance in Among the explanatory variables examined in Indonesia.21 the literature, those most relevant to Pakistan are distance to school; teacher quality and gen- Teacher quality der; water, sanitation, and hygiene facilities; structured pedagogy; and complete information Research on rural schools in Pakistan has found to parents in the form of engaging report cards. that improving the quality of teachers is likely to have higher returns than improving the qual- Distance to school ity of school buildings.22 In Punjab, Pakistan’s wealthiest province, only 56 percent of teachers Pakistan’s bottleneck in basic education infra- demonstrated basic mastery of the curriculum. structure translates into long walks to school for A study in Indonesia of a nationally representa- many students. In 2016/17, 80 percent of pub- tive sample of primary students found, unsur- lic schools in Pakistan were primary schools, prisingly, that students learn more when teach- 11 percent were middle schools, and 9 percent ers are present, schools and their teachers are were secondary schools (table A5.1). of higher quality, and class sizes are smaller.23 Given the extensive evidence from many coun- tries showing the importance of teachers in in- TABLE A5.1  Most public schools in Pakistan creasing student learning,24 improving the qual- Annexes | Pakistan Human Capital Review are for primary education ity of Pakistan’s teachers should be an urgent priority. Students who receive a high-quality Number of School level schools Percentage education are more likely to stay and succeed in Primary 119,149 80 school. Middle 16,428 11 Teacher gender High 12,576 8 Higher secondary 1,698 1 Female teachers are helpful in creating a sup- Total 149,851 100 portive environment for girls’ education.25 In Source: Pakistan Education Statistics 2016–17, chapter 3, pp. 13–17. Pakistan the Citizens Foundation employs only female teachers in its schools.26 After finding 192 that an all-female staff could increase girls’ en- schooling in just one year.35 Grade 3 students’ rollment and recognizing challenges to women’s reading skills improved by 0.6–0.7 standard de- mobility in Pakistan, it began providing trans- viation, with greater gains for girls than for boys. port to its teachers. The schools are coeduca- Also in Liberia an evaluation of the Early Grade tional, and their female-only staff has helped Reading Assessment Plus found that targeting increase the number of girls, allowing many the quality of reading instruction in grades 2 schools to achieve gender parity in education. and 3 can substantially raise student literacy. This is impressive given that across Pakistan’s Students’ reading comprehension more than public schools nearly 60  percent of girls are doubled, rising an average of 33.6  percentage out of school by grade 6.27 Hiring female teach- points over baseline scores.36 And in Papua New ers may improve girls’ learning too. In middle Guinea teachers learned a low-cost (US$60 per schools in the Republic of Korea, a study of ran- child) structured pedagogy method to improve dom assignment of students to classes with fe- early-grade reading skills that they could apply male teachers found that “female students per- in just one hour a day.37 form substantially better on standardized tests when assigned to female teachers,” with results Tutoring persisting five years later.28 Students can make as much progress in 12 Water, sanitation, and hygiene facilities weeks of tutoring as in three to five months of normal schooling.38 In Italy middle school stu- The availability of water, sanitation, and hygiene dents who received three hours of online tu- facilities can improve enrollment and gender toring a week achieved a 4.7  percent boost in parity.29 About 26 percent of Pakistan’s prima- performance in mathematics, English, and Ital- ry and secondary schools lack clean drinking ian.39 Programs pioneered by Match Education water and toilets.30 Improved water and sani- in Boston, Massachusetts, and scaled by Saga tation facilities support school attendance and Education in Chicago, Illinois, provide students performance, particularly among adolescent who are below grade level in mathematics with girls.31 Providing sex-specific latrines in India an individualized 50 minute class each day. Tu- raised girls’ enrollment across districts, though tors work with two students at a time and cover it had no impact on test scores.32 content appropriate to their current skills while also linking to what is being taught in the regu- Structured pedagogy lar classroom. Costs are kept (relatively) low by using paraprofessionals (for example, recent Structured pedagogy­ —­ a coordinated, com- college graduates) to provide the tutoring. The bined approach that includes student materials results are impressive: participating students and teacher lesson plans, training, and ongoing advanced one to two years in math learning in Annexes | Pakistan Human Capital Review support­—­appears to be a cost-effective means one year of instruction. of improving student learning. In several coun- tries structured pedagogy improved student Effective multigrade classrooms outcomes in functional literacy and numeracy.33 Structured pedagogy programs lead to learning Out of necessity, some countries have man- improvements in literacy and numeracy in low- aged to create well-functioning multigrade and middle-income countries, and this type of classrooms on a larger scale. Such classrooms intervention can be more effective than many have appropriate texts and materials, effective others.34 For example, an early-grade reading teaching practices, and teachers who are sup- program accelerated literacy among Liberi- ported through ongoing training programs.40 In an children by the equivalent of three years of remote areas with low population density, such 193 as those served by Colombia’s Escuela Nueva drops as poorer families struggle to pay for and in mountainous areas of Vietnam, multi- schooling.44 This reality is even starker for girls, grade classrooms have proven effective for in- as many lower-income families prioritize boys’ creased enrollment and positive student learn- education. For households in the poorest in- ing outcomes.41 come quintile in Pakistan, the probability of attending school is about 8  percentage points Report cards lower for a girl than for a boy.45 When parents know how their children are doing Transfers to lower the cost of education in school, children begin to do better. An ex- periment in 112 Pakistani villages through the Reducing school fees boosts enrollment. “Stu- Learning and Educational Achievements in Paki- dent participation is sensitive to the perceived stan Schools initiative found that issuing report costs and benefits of education. Because the cards increased enrollment by 4.5 percent and costs of education are immediate and easy to test scores by 42 percent.42 A survey confirmed observe, even small changes in costs can have that parents shifted their perceptions of school important impacts on participation.”46 The most quality to reflect their children’s higher scores. effective way to increase enrollment is to reduce The report card intervention­ —i­ncluding testing, the cost of schooling by offering cash directly to printing, and distribution­—­ cost US$1 per child. poor families.47 Cash can be in the form of con- In Bangladesh a randomized field experiment ditional or unconditional transfers, merit-based encouraging parents to discuss report cards scholarships, or criteria-based grants. Reduc- with teachers showed that doing so improved ing the cost of schooling is an effective way to learning outcomes by 0.26 standard deviation bring low-income children, especially girls, into in the first year and 0.38 standard deviation in school, even with supply-side constraints and the second year, with no statistically significant even when social norms favor boys’ education. difference between girls and boys.43 Cash transfers Demand-side factors Cash transfers spur enrollment. Conditional Families that can afford to send their children to cash transfers have a significant, positive ef- school often do. Pakistan’s government offers fect on the likelihood of school participation at free schooling and textbooks across the coun- both the primary and secondary levels.48 The try and stipends and cash transfers in some evidence is especially strong for girls, with en- districts, but households prefer private schools, rollment increases as large as 20  percentage due possibly to negative perceptions of public points.49 For girls in Malawi who had dropped school quality. Rigorous studies in many coun- out of school, both conditional and uncondition- Annexes | Pakistan Human Capital Review tries have examined demand-side initiatives to al cash transfers raised enrollment and English reduce the number of out-of-school children. reading comprehension. However, the enroll- Among those most relevant to Pakistan are ment effects were almost twice as large among household-level challenges tied to poverty: the recipients of conditional cash transfers as solution is to decrease the cost of schooling among recipients of unconditional cash trans- through a range of targeted interventions. fers, and reading comprehension also increased more in the conditional cash transfer group.50 Household-level challenges tied to poverty Pakistan has countrywide conditional cash transfers implemented through the Waseela-e- Poverty is negatively related to school enroll- Taleem program under the Benazir Income Sup- ment. When income falls, school enrollment port Program umbrella. 194 Conditional cash transfers also promote educa- points more likely to enroll in college.52 Offering tion progression. Half the improvement in grade merit-based scholarships in Kenya increased girls’ progression in Nicaragua over 1999–2003 was learning by 0.19 standard deviation.53 accounted for by lower dropout and repetition rates among beneficiary children who were Grants already in school when the conditional cash transfer program began.51 The program was es- Grants increase school enrollment and, if con- pecially effective in areas with poor initial school tinued, school progression. In South Africa ben- supply conditions, suggesting that low initial eficiaries of an unconditional grant had enroll- school supply is not an insurmountable obsta- ment rates at least 10 percentage points higher cle, as long as such constraints are identified than nonbeneficiaries. Expanding the South in planning and dealt with in implementation. African Child Support Grant raised teenage en- This finding has important implications for Pa- rollment at least 10 percentage points among kistan, where inadequate school supply is a seri- beneficiaries.54 ous barrier to continuing education, particularly middle and secondary schooling. Reducing the cost of schooling can even alter social norms. In Morocco a small cash transfer Merit-based scholarships to fathers of school-age children in poor rural communities that was explicitly labeled an edu- Merit-based scholarships reward students for cation support program but was not conditional studying while making school more affordable and on school attendance led to large gains in school providing an incentive to value it more highly. In participation.55 Adding conditions and targeting Ghana, for example, secondary school scholarship mothers made almost no difference. The pro- recipients were 26 percentage points more likely gram seemed to increase parents’ belief that ed- to complete secondary school and 3 percentage ucation was a worthwhile investment. Annexes | Pakistan Human Capital Review 195 ANNEX 6A KEY LABOR MARKET CONCEPTS AND DEFINITIONS Employment: Employment covers any work for who in the week prior to the survey worked at wage, salary, profit, or family gain, including the least an hour for wage or salary paid in cash or production of goods for own consumption. The in kind or who temporarily did not work during employed comprise all people age 15–64 who the reference period but had a formal attach- in the week prior to the survey worked at least ment to a job. an hour for wage, salary, profit, or family gain or who temporarily did not work during the refer- Self-employment: Self-employment covers any ence period but had a formal attachment to a work where people engage in work to earn profit job or enterprise. or family gain, including the production of goods for own consumption. The self-employed com- Unemployment: People who are unemployed prise people age 15–64 who in the week prior are those who simultaneously are without work, to the survey worked at least an hour to earn currently available for work, and seeking work. profit or family gain or who temporarily did not The unemployed comprise all people age 15–64 work during the reference period but were with who in the week prior to the survey were not an enterprise. Unpaid family work is considered in paid or self-employment, were available for self-employment. work, and were seeking work. Economically active: The economically ac- Underemployment: People in time-related tive comprise people age 15–64 who in the underemployment are those who are willing to week prior to the survey were either employed­ work additional hours, are available to work ad- —­ engaged in paid or self-employment­ or —­ ditional hours (that is, they are ready, within a unemployed. specified subsequent period, to work additional hours given opportunities for additional work), Labor force: The labor force comprises both the and have worked less than a threshold relating employed and unemployed. to working time. Formal wage employment: Wage employment is Wage employment: Wage employment covers formal if the employee has a written contract. any employment where the people employed re- —­ ceive payment­ —­ salary or wage­ for their work. Source: International Labor Organization, The wage employed comprise people age 15–64 Pakistan Bureau of Statistics. Annexes | Pakistan Human Capital Review ANNEX 6B PACKAGE OF PROGRAMS TO IMPROVE LABOR MARKET OUTCOMES Moving beyond cash transfers search, and placement, the augmented pro- gram increased the probability of a formal job While cash transfer programs are an important and higher wages, especially for young men.56 instrument to support poor people, they need Peru’s Haku Winay program, which seeks to to be paired with other interventions to im- improve social and economic outcomes among prove labor market outcomes. When Argentina’s the rural population, highlights the impacts Plan Jefes conditional cash transfer program on incrementing productivity of beneficiaries, was paired with support for skill upgrading, job asset accumulation, income generation, and 196 market access through capacity development. Focusing on a program in Niger, Bossuroy et al. The programs reviewed were characterized by (2022) conducted an impact evaluation of a a package of services for formal employment multi-intervention program anchored on a na- that featured socioemotional support, labor tional cash transfer program targeting extreme- market intermediation, case management, and ly poor women. The additional services includ- internships. ed group savings promotion, coaching, and entrepreneurship training and added a lump- Cash transfers made without an explicit em- sum cash grant, psychosocial interventions, or ployment focus tend to result in little to no both. Packages of services with psychosocial change in labor market outcomes.57 In contrast, interventions were the most cost-effective, high- transfers explicitly made for job search assis- lighting the value of including well-designed psy- tance or business startup tend to increase adult chosocial interventions. labor supply and earnings. The main channel is likely the alleviation of liquidity and risk con- The Skills Training for Advancing Resources straints. Programs that augment cash transfer program in Bangladesh with other services are now referred to as “cash- plus” programs. One example of an apprenticeship approach is Skills Training for Advancing Resources by BRAC Impact evaluations of several labor market in Bangladesh. It features an informal appren- programs in Latin America supported by the ticeship program that places young people in Inter-American Development Bank found that pairs under a master craftsperson who is usu- comprehensive and demand-driven programs ally an experienced shop owner or worker in a can have high impact when focused on poor (or trade. On average, these young people are age otherwise) disadvantaged youth who have diffi- 16–17 and have five or six years of education. culty in labor market insertion.58 The apprentices receive hands-on training five days a week for six months.59 In addition to the World Bank–supported multipronged workplace-based practical training, apprentic- economic inclusion programs in the Sahel es receive classroom-based soft-skills training once a week on such issues as financial literacy, In six countries in the Sahel, the World Bank market assessment, and English language skills. supports multipronged economic inclusion Once the training is completed, the project links programs that prominently feature psychoso- the participants with potential employers, and cial interventions. Bossuroy et al. (2021) evalu- they are certified through the National Training ated the impact of national cash transfer pro- and Vocational Qualifications Framework. On- grams, identifying a core package that includes the-job training increased participants’ labor a lump-sum cash grant, a second package that market participation by 23  percentage points Annexes | Pakistan Human Capital Review substitutes the cash grant with psychosocial in- (up to 59  percent) and earnings by 44  per- terventions, and a third that includes both the cent through both self-employment and wage cash grant and the psychosocial interventions. employment.60 All three packages showed large impacts on consumption and food security and increased A randomized control trial of a program participation and profits in women-led off-farm targeting poor women in Tunisia business and livestock activities. The benefit–­ cost ratio for the “pure” psychosocial pack- Tunisia’s Community Works and Local Partici- age was 126  percent; the complete package, pation program shows the need to design pro- 95 percent; and the core cash-grant package, grams targeting women entrepreneurs.61 The 58 percent. effects of cash grant intervention had positive 197 impacts on the economic well-being of benefi- For example, impoverished women need a more ciaries, highly marginalized and poor women, intensive package of services than nonpoor but raised concerns that the effects may not women do to break out of subsistence produc- persist in the long run. A cash grant interven- tion and grow their businesses.63 tion paired with financial training through sim- ple exercises and videos had positive results for Pilot of an asset transfer–based economic income-generating activities of other members inclusion program of the household but not beneficiaries’ entre- preneurship. Findings related to a subinterven- The pilot was implemented through the Paki- tion adding “gender dialogue” that invited male stan Poverty Alleviation Fund and its local part- partners to the training were inconclusive on ner network of nongovernmental organizations whether the gender dialogue added value relat- in Sindh. Beneficiary households had to meet ed to labor market participation, income-gener- three of five eligibility criteria that established ating activities, asset accumulation, or women’s their status as a vulnerable and poor house- autonomy. These results show the potential and hold. The package of services included a sav- limitation of capital injection interventions to ings component, a health component, an asset promote women’s entrepreneurship and sug- transfer, consumption support, and follow-up gest designing more country-specific programs. visits from the participating nongovernmental Moreover, psychology-based training was par- organization. The savings component encour- ticularly effective for woman who owned busi- aged households to save with the help of rotat- nesses, for whom traditional training has often ing savings and credit associations; the health been ineffective. Women who received personal component used the services of Lady Health initiative training saw their profits increase by Workers for frequent checkups, health and hy- 40 percent, compared with 5 percent for tradi- giene training, and other basic health services; tional business training.62 the asset transfer was a transfer of a produc- tive asset equivalent to PKR  15,000 (com- A review of a wide array of women-targeted monly livestock); consumption support was interventions elsewhere found that the same in the form of monthly transfers of PKR 1,000 type of interventions had significantly different to the household; and nongovernmental orga- outcomes depending on the profiling of women nization follow-up involved weekly visits from (poor versus nonpoor and young versus adult). partner staff. Annexes | Pakistan Human Capital Review 198 NOTES 1. The World Bank’s Harmonized Learning Outcome collection and generation, in addition to the previous puts learning data from international and region- purely population-based model. al assessments on a comparable scale. The data 18. Zaidi et al. 2013. can be accessed at https://datacatalog.worldbank. 19. Andrabi, Das, and Khwaja 2013. org/dataset/harmonized​ - learning​ - outcomes​ - hlo​ -database. 20. Kim, Alderman, and Orazem 1999. 2. World Bank n.d. 21. Burde and Linden 2013; Duflo 2001. 3. Filmer et al. 2018. 22. Behrman et al. 1997. 4. https://childmortality.org/methods. 23. Suryadarma et al. 2006. 5. Pennings 2020. 24. For example, Bruns and Luque (2014) and Chetty, Friedman, and Rockoff (2011). 6. Government of Pakistan. 25. Sperling and Winthrop 2015. 7. National Nutrition Survey 2018. Evidence from Nepal shows that a similar community health work- 26. For more information, see https://www.tcf.org.pk. er program with wide coverage has resulted in major 27. Human Rights Watch 2018; International Education improvements in breastfeeding outcomes, which News 2019. in turn contributed to the country’s substantial re- 28. Lim and Meer 2017a, 2017b. duction in stunting over two decades­—­from close to 29. Garn et al. 2013. 70 percent in 1995 to 36 percent in 2016. 30. Pakistan Education Statistics 2016–17. 8. National Nutrition Survey 2018. 31. Sperling and Winthrop 2015. 9. J-PAL 2014. 32. Adukia 2017. 10. Khan 2019. 33. RTI 2020. 11. Ullah et al. 2019. 34. Banerjee et al. 2013; Conn 2014; Evans and Mendez 12. Only Sindh has legislated mandatory salt iodiza- Acosta 2020; Evans and Popova 2016; Glewwe et al. tion (in 2013); three other provinces have enacted 2014; Global Education Evidence Advisory Panel amendments to provincial food rules to the same 2020; Graham and Kelly 2018; Krishnaratne, White, effect, and the remaining provinces are enacting leg- and Carpenter 2013; McEwan 2012; Murnane and islation for USI. Ganimian 2014. 13. Government of Pakistan 2020. 35. Piper and Medina 2010. 14. WHO n.d. 36. Piper and Medina 2010. 15. Some examples include the development of multi- 37. Macdonald and Vul 2018. sectoral nutrition strategies, the Food Fortification Strategy, the IYCF and IYCF Communication Strat- 38. The Economist 2020; Nickow, Oreopoulos, and egy, and several guidelines, such as those for ado- Quan 2020; Oreopoulos 2020. lescent nutrition and supplementation, community- Annexes | Pakistan Human Capital Review 39. Eridani 2020. based management of acute malnutrition, vitamin A 40. Thomas and Shaw 1992. supplementation, and infant and young child feed- ing in emergencies. 41. Aikman and Pridmore 2001; Le 2018; Psacharopou- los, Rojas, and Velez 1993. 16. https://www.nutritioncluster.net/country/pakistan. 42. Andrabi, Das, and Khwaja 2017. 17. Although the 2010 devolution shifted financing re- sponsibility for devolved ministries to provincial 43. Islam 2019. governments, the seventh National Finance Com- 44. Evans and Popova 2016; Evans and Yuan 2019. mission increased the financial share of resources to 45. World Bank forthcoming. 56 percent and introduced a more equitable distri- 46. J-PAL 2017, p. 1. bution formula that factored in economic underde- velopment, inverse population density, and revenue 199 47. Benhassine et al. 2015; Duflo, Dupas, and Kremer 54. Eyal, Woolard, and Burns 2014. 2017; Edmonds and Shrestha 2014; Evans and Yuan 55. Benhassine et al. 2015. 2019; Eyal, Woolard, and Burns 2014; Maluccio, Mur- 56. Mourelo and Escudero 2017. phy, and Regalia 2010. 57. Baird, McKenzie, and Özler 2018. 48. Garcia and Saavedra 2017. 58. Ibarrarán and Rosas Shandy 2009. 49. Filmer and Schady 2011. 59. Bhattacharjee and Kamruzzaman 2016; Das 2018. 50. Baird et al. 2016. 60. Das 2018. 51. Maluccio, Murphy, and Regalia 2010. 61. Ferrah et al. 2021. 52. Duflo, Dupas, and Kremer 2017. 62. Campos et al. 2017. 53. Kremer, Miguel, and Thornton 2009. 63. Buvinic and Furst-Nichols 2014. REFERENCES Adukia, A. 2017. “Sanitation and Education.” American Review Paper, Abdul Latif Jameel Poverty Action Lab, Economic Journal: Applied Economics 9 (2): 23–59. Cambridge, MA. https://www.aeaweb.org/articles?id=10.1257/app​ Behrman, J., S. Khan, D. Ross, and R. Sabot. 1997. “School .20150083. Quality and Cognitive Achievement Production: A Case Aikman, S., and P. Pridmore. 2001. “Multigrade Schooling Study for Rural Pakistan.” Economics of Education Review in ‘Remote’ Areas of Vietnam.” International Journal of Ed- 16 (2): 127–142. ucational Development 21 (6): 521–536. Benhassine, N., F. Devoto, E. Duflo, P. Dupas, and V. Andrabi, T., J. Das, and A. Khwaja. 2013. “Students Today, Pouliquen. 2015. “Turning a Shove into a Nudge? A ‘La- Teachers Tomorrow: Identifying Constraints on the Provi- beled Cash Transfer’ for Education.” American Economic sion of Education.” Journal of Public Economics 100 (1): Journal: Economic Policy 7: 86–125. https:/ /www.nber​ 1–14. .org/papers/w19227. Andrabi, T., J. Das, and A. Khwaja. 2017. “Report Cards: Bhattacharjee, A., and M. Kamruzzaman. 2016. “Towards The Impact of Providing School and Child Test Scores on Employability and Better Livelihood: An Evaluation of Educational Markets.” American Economic Review 107 BRAC’s Skills Development Initiative.” Skills Development (6): 1535–63. https:/ /www.aeaweb.org/articles?id=10​ Working Paper 1, BRAC Institute of Governance and De- .1257/aer.20140774. velopment, Dhaka. Baird, S., E. Chirwa, J. De Hoop, and B. Özler. 2016. “Girl Bruns, B., and J. Luque. 2014. Great Teachers: How to Power: Cash Transfers and Adolescent Welfare. Evidence Raise Student Learning in Latin America and the Caribbe- from a Cluster-randomized Experiment in Malawi.” In S. an. Washington, DC: World Bank. Annexes | Pakistan Human Capital Review Edwards, S. Johnson, and D. N. Weil (eds.), African Suc- Burde, D., and L. Linden. 2013. “Bringing Education cesses, Volume II: Human Capital. Chicago, IL: University to Afghan Girls: A Randomized Controlled Trial of Vil- of Chicago Press. https:/ /www.nber.org/papers/w19479. lage-based Schools.” American Economic Journal: Ap- Baird, S., D. McKenzie, and B. Özler. 2018. “The Effects plied Economics 5: 27–40. https:/ /www.aeaweb.org​ of Cash Transfers on Adult Labor Market Outcomes.” IZA /articles?id=10.1257/app.5.3.27. Journal of Development and Migration 8 (1): 1–20. Buvinic, M., and R. Furst-Nichols. 2014. “Promoting Wom- Banerjee, A., P. Glewwe, S. Powers, and M. Wasserman. en’s Empowerment: What Works?” Policy Research Work- 2013. “Expanding Access and Increasing Student Learn- ing Paper 7087, World Bank, Washington, DC. ing in Post-primary Education in Developing Countries: A Campos, F., M. Frese, M. Goldstein, L. Iacovone, H. Review of the Evidence.” Post-Primary Education Initiative C. Johnson, D. McKenzie, and M. Mensmann. 2017. 200 “Teaching Personal Initiative Beats Traditional Training Washington, DC. https://openknowledge.worldbank.org​ in Boosting Small Business in West Africa.” Science 357 /handle/10986/32128. (6357): 1287–1290. Eyal, K., I. Woolard, and J. Burns. 2014. “Cash Transfers Chetty, R., J. N. Friedman, and J. E. Rockoff. 2011. “The and Teen Education: Evidence from South Africa.” Unpub- Long-Term Impacts of Teachers: Teacher Value-Added and lished Report. Student Outcomes in Adulthood.” Working Paper 17699, Ferrah S, J. Gazeaud, N. Khan E. Mvukiyehe, V. Shrotri, O. National Bureau of Economic Research, Cambridge, MA. Sterck, and S. B. Zineb. 2021. Enhancing Female Entre- Conn, K. 2014. Identifying Effective Education Interven- preneurship through Cash Grants. Washington, DC: World Saharan Africa: A Meta-analysis of Rigorous tions in Sub-­ Bank. Impact Evaluations. Unpublished manuscript, Columbia Filmer, D., H. Rogers, N. Angrist, and S. Sabarwal. 2018. University, New York. “Learning Adjusted Years of Schooling (LAYS): Defining Das, N.C. 2018. “Training the Disadvantaged Youth and a New Macro Measure of Education.” Policy Research Labor Market Outcomes: Evidence from a Randomized Working Paper 8591, World Bank, Washington, DC. Controlled Trial in Bangladesh.” Dhaka: BRAC. Filmer, D., and N. Schady. 2011. “Does More Cash in Con- Duflo, E. 2001. “Schooling and Labor Market Conse- ditional Cash Transfer Programs Always Lead to Larger quences of School Construction in Indonesia: Evidence Impacts on School Attendance?” Journal of Development from an Unusual Policy Experiment.” American Economic Economics 96 (1): 150–157. Review 91 (4): 795–813. Garcia, S., and J. Saavedra. 2017. “Educational Impacts Duflo, E., P. Dupas, and M. Kremer. 2017. “The Impact of and Cost-Effectiveness of Conditional Cash Transfer Pro- Free Secondary Education: Experimental Evidence from grams in Developing Countries: A Meta-analysis.” Work- Ghana.” Working Paper, Massachusetts Institute of Tech- ing Paper 23594, National Bureau of Economic Research, nology, Cambridge, MA. http:/ /www.poverty-action.org​ Cambridge, MA. /publication/returns-secondary-education-unpacking​ Garn, J., L. Greene, R. Dreibelbis, S. Saboori, R. Rhein- -delivery-senior-secondary-schooling-ghana. gans, and M. Freeman. 2013. “A Cluster-randomized Trial Edmonds, E., and M. Shrestha. 2014. “You Get What You Assessing the Impact of School Water, Sanitation and Pay for: Schooling Incentives and Child Labor.” Journal of Hygiene Improvements on Pupil Enrolment and Gender Development Economics 111: 196–211. https:/ /www.nber​ Parity in Enrolment.” Journal of Water Sanitation and Hy- .org/papers/w19279. giene for Development 3: 592–601. https:/ /iwaponline​ .com/washdev/article-abstract/3/4/592/30334​ Eridani, T. 2020. “Online Tutoring Improves Disadvan- /A-cluster-randomized-trial-assessing-the-impact-of​ taged School Pupils Performance and Wellbeing in Lock- ?redirectedFrom=fulltext. down.” News Release, EurekAlert, 15 July. https://www​ .eurekalert.org/news-releases/806598. Glewwe, P. W., E. A. Hanushek, S. D. Humpage, and R. Ravina. 2014. “School Resources and Educational Out- Evans, D., and A. Mendez Acosta. 2020. “Education in Af- comes in Developing Countries: A Review of the Liter- rica: What Are We Learning?” Working Paper 542, Cen- ature from 1990 to 2010.” In P. Glewwe (ed.), Education ter for Global Development, Washington, DC. https:/ /​ Annexes | Pakistan Human Capital Review Policy in Developing Countries. Chicago, IL: University of www.cgdev.org/sites/default/files/education-africa​-what​ Chicago Press. https://doi/org/10.3386/w17554. -are​-we-learning.pdf. Global Education Evidence Advisory Panel. 2020. Evans, D., and A. Popova. 2016. “What Really Works to “Cost-Effective Approaches to Improve Global Learning: Improve Learning in Developing Countries? An Analysis What Does Recent Evidence Tell Us Are ‘Smart Buys’ for of Divergent Findings in Systematic Reviews.” The World Improving Learning in Low- and Middle-income Coun- Bank Research Observer 31 (2). https:/ /openknowledge​ tries?” World Bank. https://documents1.worldbank.org​ .worldbank.org/handle/10986/29308. /curated/en/719211603835247448/pdf/Cost-Effective​ Evans, D., and F. Yuan. 2019. “What We Learn about Girls’ -Approaches-to-Improve-Global-Learning-What-Does​ Education from Interventions that Do Not Focus on -Recent-Evidence-Tell-Us-Are-Smart-Buys-for-Improving​ Girls.” Policy Research Working Paper 8944, World Bank, -Learning-in-Low-and-Middle-Income-Countries.pdf. 201 Graham, J., and S. Kelly. 2018. How Effective Are Early Initiative for Impact Evaluation. https://www.3ieimpact​ Grade Reading Interventions? A Review of the Evidence. .org/evidence-hub/publications/working-papers/quality​ Washington, DC: World Bank. https:/ /doi.org/10.1596​ -education-all-children-what-works-education /1813-9450-8292. Le, H. M. 2018. “Another Textbook Project? The Imple- Government of Pakistan. 2020. “Poverty Alleviation and mentation of Escuela Nueva in Vietnam.” Educational Social Safety Division Curtain Raiser, October 2020.” Research for Policy and Practice 17: 223–239. https://doi​ .org/10.1007/s10671-018-9230-x. Human Rights Watch. 2018. “Shall I Feed My Daughter, or Educate Her? Barriers to Girls’ Education in Pakistan.” Lim, J., and J. Meer. 2017a. “The Impact of Teacher–Stu- 12 November. https:/ /www.hrw.org/report/2018/11/12​ dent Gender Matches: Random Assignment Evidence from /shall-i-feed-my-daughter-or-educate-her/barriers-girls​ South Korea.” Journal of Human Resources 52 (4): 979– -education-pakistan. 997. http://jhr.uwpress.org/content/52/4/979​.abstract. Ibarrarán, P., and D. Rosas-Shady. 2009. “Evaluating the Lim, J., and J. Meer. 2017b. “Persistent Effects of Teach- Impact of Job Training Programmes in Latin America: Ev- er–Student Gender Matches.” Working Paper 24128, Na- idence from IDB Funded Operations.” Journal of Develop- tional Bureau of Economic Research, Cambridge, MA. ment Effectiveness 1 (2): 195–216. Macdonald, K., and B. T. Vul. 2018. “A Randomized Eval- International Education News. 2019. “Expanding to Say uation of a Low-Cost and Highly Scripted Teaching “Yes”: The Ongoing Work of The Citizens Foundation in Method to Improve Basic Early Grade Reading Skills in Pakistan (Part 1 of 2).” 1 May. https://internationalednews​ Papua New Guinea.” Education Global Practice, Policy .com/2019/05/01/expanding-to-say-yes-the-ongoing​ Research Working Paper 8427, World Bank, Washington, -work-of-the-citizens-foundation-in-pakistan-part-1-of​-2/. DC. https://openknowledge.worldbank.org/bitstream​ /handle/10986/29833/WPS8427.txt?sequence=2&​ Islam, A. 2019. “Parent–Teacher Meetings and Student isAllowed=y. Outcomes: Evidence from a Developing Country.” Euro- pean Economic Review 111: 273–304. https://www.theigc​ Maluccio, J., A. Murphy, and F. Regalia. 2010. “Does Sup- .org/wp-content/uploads/2016/10/Islam-2016-working​ ply Matter? Initial Schooling Conditions and the Effective- -paper.pdf. ness of Conditional Cash Transfers for Grade Progres- sion in Nicaragua.” Journal of Development Effectiveness J-PAL (Abdul Latif Jameel Poverty Action Lab). 2014. 2: 87–116. https:/ /www.researchgate.net/publication​ “Pay for Performance Incentives for Healthcare Work- /228877017_Does_supply_matter_Initial_schooling​ ers in Pakistan.” Lahore, Pakistan: J-PAL. https:/ /www​ _conditions_and_the_effectiveness_of_conditional_cash​ .poverty ​ a ction ​ l ab​ .org ​ /evaluation​ /pay​ - perfomance​ _transfers_for_grade_progression_in_Nicaragua. -incentives-heathcare-workers-pakistan. McEwan, P. 2012. “Improving Learning in Primary Schools of Khan, N. 2019. “Reaching the Last Mile in Universal Salt Io- Developing Countries: A Meta-analysis of Randomized Ex- dization in Pakistan.” Nutrition International, 4 November. periments.” Review of Educational Research 20 (10): 1–42. https://www.nutritionintl.org/news/all-blog-posts​/reaching​ -the-last-mile-in-universal-salt-iodization-in​-pakistan/. MNHSRC (Ministry of National Health, Services, Regu- lations and Coordination) Nutrition Wing. 2018. Nation- Annexes | Pakistan Human Capital Review Kim, J., H. Alderman, and P. Orazem. 1999. “Can Private al Nutrition Survey 2018 Volume I. http://phkh.nhsrc.pk​ School Subsidies Increase Enrollment for the Poor? The /sites/default/files/2021-03/National%20Nutrition​%20 Quetta Urban Fellowship Program.” The World Bank Eco- Survey%20Key%20Findings%20Volum%201%20 nomic Review 13: 443–65. https:/ /elibrary.worldbank.org​ UNICEF%202018.pdf. /doi/10.1093/wber/13.3.443. Mourelo, E. L., and V. Escudero. 2017. “Effectiveness of Kremer, M., E. Miguel, and R. Thornton. 2009. “Incentives Active Labor Market Tools in Conditional Cash Transfers to Learn.” The Review of Economics and Statistics 91: Programs: Evidence for Argentina.” World Development 437–56. https://dash.harvard.edu/handle/1/3716457. 94: 422–447. Krishnaratne, S., H. White, and E. Carpenter. 2013. “Qual- Murnane, R. J., and A. J. Ganimian. 2014. “Improving Ed- ity Education for All Children: What Works in Education in ucational Outcomes in Developing Countries: Lessons Developing Countries. 3ie Working Paper 20, International 202 from Rigorous Evaluations.” Working Paper 20284, Na- Schools in Developing Countries: Evidence from Indone- tional Bureau of Economic Research, Cambridge, MA. sia.” Education Economics 14 (4): 401–429. https://www​ https://www.nber.org/papers/w20284.pdf. .researchgate.net/profile/Sudarno_Sumarto/publication​ /24079203_Improving_Student_Performance_in_Public​ Nickow, A., P. Oreopoulos, and V. Quan. 2020. “The Im- _Primary_Schools_in_Developing_Countries_Evidence​ pressive Effects of Tutoring on PreK–12 Learning: A Sys- _from_Indonesia/links/0c96051a514c5e3df4000000​ tematic Review and Meta-analysis of the Experimental /Improving-Student-Performance-in-Public-Primary​ Evidence.” Working Paper 27476, National Bureau of Eco- -Schools-in-Developing-Countries-Evidence-from​ nomic Research, Cambridge, MA. https:/ /www.nber.org​ -Indonesia.pdf. /papers/w27476. The Economist. 2020. “England’s Catch-up Tutoring Pro- Oreopoulos, P. 2020. “Scale Up Tutoring to Combat COVID gramme Has Bold Ambitions.” 5 November. https:/ /www​ Learning Loss for Disadvantaged Students.” Scientific .economist.com/britain/2020/11/05/englands-catch​ American, 24 November. https:/ /www.scientificamerican​ -up-tutoring-programme-has-bold-ambitions. .com/article/scale-up-tutoring-to-combat-covid-learning​ -loss-for-disadvantaged-students/. Thomas, C., and C. Shaw. 1992. “Issues in the Develop- ment of Multigrade Schools.” Technical Paper 172, World Pennings, S. 2020. “The Utilization-Adjusted Human Cap- Bank, Washington, DC. ital Index.” Policy Research Working Paper 9375, World Bank, Washington, DC. Ullah, I., A. Pillai, S. Dey, S. Raza, and N. A. Khan. 2019. “Pakistan’s Journey towards Universal Coverage of Io- Piper, B., and K. Medina. 2010. “Early Grade Reading As- dised Salt: A Non-Systematic Review.” Pakistan Journal of sessment (EGRA) Plus: Liberia. Program Evaluation Re- Public Health 9 (4): 195–197. port.” RTI International. https://files.eric.ed.gov/fulltext​ /ED516080.pdf. WHO (World Health Organization). n.d. “Pakistan Inte- grated Nutrition Strategy Operational Framework Plan.” Psacharopoulos, G., C. Rojas, and E. Velez. 1993. “Achieve- Geneva: WHO. https:/ /www.who.int/hac/crises/pak​ ment Evaluation of Colombia’s Escuela Nueva: Is Multi- /pakistan_nutrition_strategy.pdf. grade the Answer?” Comparative Education Review 37 (3): 263–276. World Bank. n.d. World Development Indicators data- base. Washington, DC: World Bank. https://databank​ RTI. 2020. “Structured Pedagogy Literature Review.” .worldbank.org/metadataglossary/world-development​ https://scienceofteaching.s3.eu-west-3.amazonaws.com​ -indicators/series. /index.html#/lessons/qRcSj0b2yqminRXJh7cSd_2tkJc​ -0blC. World Bank. Forthcoming. Barriers to Girls’ Education in Pakistan. Washington, DC: World Bank. Sperling, G. B., and R. Winthrop. 2015. What Works in Girls’ Education: Evidence for the World’s Best Invest- Zaidi, S., S. K. Mohmand, N. Hayat, A. M. Acosta, and Z. ment. Washington, DC: Brookings Institution Press. A. Bhutta. 2013. “Nutrition Policy in the Post-Devolution Context in Pakistan: An Analysis of Provincial Opportuni- Suryadarma, D., A. Suryahadi, S. Sumarto, and H. Rogers. ties and Barriers.” IDS Bulletin 44 (3): 86–93. 2006. “Improving Student Performance in Public Primary Annexes | Pakistan Human Capital Review 203 About the authors and contributors AUTHORS early childhood development and for disabili- ty-inclusive education. He co-led the 2018 flag- Lire Ersado, the Task Team Leader for the re- ship report on the quality of education in East port, is the World Bank’s Human Development Asia and Pacific, “Growing Smarter: Learning Practice Leader for Pakistan, based in Islam- and Equitable Development in East Asia and Pa- abad. In this capacity, Lire oversees the Bank’s cific.” Amer holds a Ph.D. in public policy from programs and engagements covering the Ed- the University of Chicago as well as a bachelor’s ucation; Health, Nutrition, and Population; and degree in history from Yale University. Social Protection and Jobs Global Practices. Prior to his current role, he held a similar posi- Ashi Kohli Kathuria, a Senior Nutrition Spe- tion in the South Caucasus, covering Armenia, cialist with the World Bank until her retirement Azerbaijan, and Georgia, based in Tbilisi. After in January 2022, provided technical leadership joining the World Bank as a Young Profession- to the Bank’s efforts to improve nutrition in the al, he worked in the Poverty and Equity Global South Asia and East Asia and Pacific Regions. Practice and the Human Development Practice Having led large-scale public health and nu- Group in Eastern and Southern Africa, Europe trition programs for decades, she made many and Central Asia, and Middle East and North Af- contributions to the development, testing, docu- rica, based in Washington, DC­ —­ and in Sana’a, mentation, and scale-up of innovative approach- Republic of Yemen, covering Armenia, Azerbai- es to improve women’s and children’s nutrition jan, Bulgaria, Egypt, Georgia, Jordan, Kyrgyz and health. A champion of multisectoral actions Republic, North Macedonia, Oman, Serbia, and and integrated approaches to improve nutri- the Republic of Yemen. Prior to joining the World tion, she has supported agriculture; water, san- Bank Group, Lire was a postdoctoral research itation, and hygiene; women’s self-help groups; fellow at the International Food Policy Research and food system platforms to identify entry Institute in Washington, DC, and taught at Ha- points and implement actions to address early ramaya University, Ethiopia. He holds a Ph.D. in nutrition and development. As the Bank’s lead in About the authors | Pakistan Human Capital Review economics from Virginia Polytechnic Institute South Asia for the Investing in Early Years pro- and State University. gram, she led efforts in Sindh, Pakistan, to inte- grate early learning and stimulation for children Amer Hasan  is a Senior Economist in the Educa- under 3 with government health and nutrition tion Global Practice, focusing on the South Asia outreach services. Region, where he is working on several multi- sectoral human capital investment projects. His Juan Baron  is a Senior Economist in the Educa- most recent assignment before this was with tion Global Practice, where he works on analyti- the Human Capital Project. He has also been a cal and lending activities in the South Asia and part of the East Asia and Pacific education team, Latin America and Caribbean Regions. Juan has where he worked on China and Indonesia and worked on analytical and operations in Domin- served as Task Team Leader for lending opera- ican Republic, Haiti, Jordan, and Pakistan and tions and analytical activities. He served as the has supported teams in the Middle East and East Asia and Pacific regional focal point for North Africa, Latin America and the Caribbean, 204 South Asia, and Sub-­ Saharan Africa Regions. He S. Amer Ahmed i s a Lead Economist and Pro- uses quantitative and qualitative research meth- gram Leader in Human Development in the ods, including impact evaluation, to explore a South Asia Region. He oversees the World range of topics on labor, gender, and education Bank’s human development engagements in economics, among others. Prior to his current Bangladesh and Bhutan. Amer was previous- role, Juan worked in the Poverty Unit at the Bank ly in the World Bank’s Prospects Group and in Africa and at the Central Bank of Colombia. has published on climate change, fiscal policy, Juan holds a Ph.D. in economics from the Aus- labor, international trade, growth, and demo- tralian National University. Juan is also the The- graphics. Aside from peer-reviewed research, he matic Lead of the Management Capacity and has co-authored flagship products such as the Service Delivery Group, is an EdTech fellow, and World Bank’s biannual Global Economic Pros- has led the implementation of teach at the right pects report and the World Bank’s final Global level approaches using adaptive software. Monitoring Report (2015/2016) on the Millenni- um Development Goals. Amer has a Ph.D. in ap- Koen Geven i s an Economist in the Education plied economics from Purdue University, where Global Practice. Koen leads research on the he specialized in international trade. learning crisis in education, the COVID-19 pan- demic’s impact on education in Pakistan, girls’ education, inequality in higher education, and CONTRIBUTORS inequality in family dynamics. He also leads and co-leads education investment projects in Aisha K. Yousafzai i s an Associate Professor of Bangladesh and Pakistan. Koen holds a Ph.D. in Global Health in the Department of Global Health social and political sciences from the European and Population at Harvard T.H. Chan School of University Institute. Public Health and is Visiting Faculty in the De- partment of Pediatrics and Child Health at Aga May Bend is a Consultant in the Education Khan University. She completed her Ph.D. at the Global Practice, where she coordinates literacy Institute of Child Health, University College Lon- and data improvement projects and co-leads don, with a focus on international child health, analytical work on girls, out-of-school children, nutrition, and development. Her research has and children with disabilities. Prior to this, she focused on understanding integrated early child- worked with the Education Global Practice in hood interventions. She has 20 years of field the Middle East and North Africa and East Asia research experience in low- and middle-income and Pacific Regions and co-led the Skills Global countries, having lived and worked in South Asia About the authors | Pakistan Human Capital Review Solutions Group. May joined the Bank in 2009, and East Africa and led program evaluations in when she began working in the Education Glob- Central and Eastern Europe. The goal of her re- al Practice in Latin America and the Caribbean. search is to promote early childhood develop- Her areas of expertise include early childhood ment and to support capacity to develop, evalu- development, educational technology, blended ate, and improve early childhood interventions. learning, out-of-school children, socioemotional Aisha serves on several global advisory groups skills, and teacher quality. May has authored and on early childhood development for such agen- co-authored several World Bank education pub- cies as the United Nations Children’s Fund, the lications, including the Middle East and North World Bank, and the World Health Organization. Africa Region’s Flagship Report on Education. She holds a master’s degree in economic and Ali Mirza is a Health Specialist currently work- political development from Columbia University ing in the World Bank Pakistan Country Office. and a bachelor’s degree in Latin American histo- He has been with the Bank for the past three ry from Brown University. years and is engaged in supporting various 205 health, nutrition, and population operations in she works on service delivery and learning en- the country. He has worked on development vironments in fragile contexts, with a special policy operations, investment project financing, focus on community-based management, gen- and program-for-results instruments, covering der, infrastructure, remote monitoring, and op- human capital, universal heatlh coverage, and erations optimization. She has worked across pandemic preparedness. Dr. Mirza is leading various regions, including Latin America, Africa, several advisory service and analytics activities the Middle East, and South Asia. Prior to her and is contributing to ongoing dialogue on cli- current role, Elena worked at the Inter-American mate change. More recently, he has been devel- Development Bank. She holds two master’s de- oping a health financing framework and review- grees in international economic policy and eco- ing the existing social health insurance models nomic development from Columbia University in Pakistan to support the country’s efforts to and Sciences Po Paris. achieve universal health coverage. Elizabeth Hentschel is a Ph.D. candidate in the Aliya Kashif is a Senior Health Specialist in the Global Health and Population Department at the World Bank. Trained as a public health special- Harvard T.H. Chan School of Public Health. With ist, she has been working in Pakistan for the a specific focus in early childhood development past 19 years in various capacities with the Gov- in low- and middle-income contexts, Elizabeth ernment, United Nations, international nongov- uses both quantitative and qualitative methods ernmental organizations, and, for the past 10 to evaluate and develop techniques to better un- years, with the Bank. She has worked on various derstand measurement pathways that influence health, nutrition, and population projects and children’s early development. Elizabeth has con- analytical and advisory services in the Bank in ducted research in India, Kenya, Pakistan, Tan- addition to providing cross-sectoral support to zania, Uganda, and Vietnam; holds a master’s other Global Practices such as Education; Social degree in global health and population from Protection; Disaster Risk Management; Water, Harvard University, and a bachelor’s degree in Sanitation, and Hygiene; and Agriculture and biomolecular science and international studies Livestock. A doctor by profession, Aliya holds a from the University of Michigan. additional master’s degree in business admin- istration and a master’s degree in public health. Heather Tomlinson is a Child Development Spe- cialist with experience in research, policy, non- Amna Ansari holds a Ph.D. in development stud- profit, and international development. She has ies from the University of Cambridge and has worked for the World Bank, the United Nations About the authors | Pakistan Human Capital Review more than 10 years of experience in education. Children’s Fund, US Senator Edward M. Kenne- She has served dynamic roles across a range of dy, the US National Association for the Educa- multilateral, public, and private organizations, tion of Young Children, the US National Institute including consulting, research, and training and of Child Health and Human Development, and has led the development of various education Vanderbilt University. Heather is co-founder and policies and sector plans in Punjab (Pakistan). former executive director of the Roshan Learn- Amna previously served the World Bank as an ing Center in Jakarta. She has authored two Early Years Fellow (Pakistan) and continues to books on early childhood and dozens of schol- provide technical and operational support for arly articles, chapters, and reports. Heather large-scale education and human capital invest- was an invited member of the Editorial Board of ment projects in the country. the 2016 SAGE Encyclopedia of Contemporary Early Childhood Education. She holds a Ph.D. in Elena Roseo is a Senior Operations Officer in developmental psychology from the University the South Asia Region at the World Bank, where of Georgia. 206 Izza Farrakh is a Senior Education Specialist in quantitative research and impact evaluation work the South Asia region at the World Bank, where of several large-scale randomized controlled tri- she works on strengthening institutions for als. He has co-authored several scientific arti- service delivery, with a special focus on educa- cles in peer-reviewed journals and presented at tion equity and quality. Before joining the World several international conferences and seminars. Bank, Izza worked as the Head of Pakistan for He holds a master’s degree in economics from the global advisory firm Adam Smith Interna- North South University and a bachelor of science tional and with Save the Children, the Research degree and a master of science degree in applied Triangle Institute, and the American Institutes statistics from the University of Dhaka. for Research. Izza has a master’s degree in pub- lic policy and management from the School of Mahreen Tahir-Chowdhry, a Consultant in the Oriental and African Studies and a bachelor’s World Bank since 2015, has more than 10 years of degree in economics and English from the Uni- experience in the field of education with national versity of Virginia. and international organizations. Her work has fo- cused primarily on early childhood development, Jahanzaib Sohail is a Health Economist with basic education, skills development, and higher the World Bank. His main areas of interest in- education. She has been actively involved in sup- volve economic evaluation of healthcare proj- porting the government in policy, project design, ects, health economic modeling and data an- implementation, and evaluation. Mahreen holds a alytics, national health accounts, domestic master’s degree in international educational de- resource mobilization, and analyzing the impact velopment from Columbia University. of country’s health financing policies in the con- text of universal health coverage. Prior to join- Mamata Ghimire is a Health Economist with the ing the Bank in 2019, Jahanzaib worked with World Bank, with expertise in health financing, different government bodies, UN agencies, and nutrition financing, public expenditure reviews, international nongovernmental organizations on and surveys. She has worked in the South Asia, a local and global basis regarding public health, West Africa, and East Asia and Pacific Regions. health management, and health financing. Jah- Mamata’s experience spans high-quality ana- anzaib has a bachelor of medicine–bachelor of lytical work and internationally visible research, surgery degree, a master’s degree in business advisory services, and lending operations. She administration, and a master of science degree holds a Ph.D. in health economics. in public health from the renowned universities of Pakistan; has a master of health economics Manav Bhattarai, a Senior Health Specialist in About the authors | Pakistan Human Capital Review degree from the University of Queensland, Aus- the World Bank, is a medical professional spe- tralia; and is currently enrolled as DrPH in health cializing in public health. Since joining the Bank, economics candidate at the London School of he has worked on various health and nutrition Hygiene and Tropical Medicine. projects and analytical activities in Bhutan, India, Nepal, and Pakistan. Jyotirmoy Saha is a Consultant in the Social Protection and Jobs Global Practice. His current Mina Zamand is an Economist in the Educa- responsibilities include designing and supporting tion Global Practice, where she works primarily supervision of urban poverty and social protec- on Pakistan. She has more than seven years of tion surveys and statistical analyses in support professional experience working with aid agen- of multiple safety net operations. Previously, he cies on interventions and research for improving worked with the Population Council in Bangladesh human development outcomes. Her thematic on projects related to reproductive health and experience includes early childhood education, poverty, gender, and youth programs, overseeing foundational learning, higher education, adaptive 207 social protection, and safety net delivery systems. works to improve the accessibility, comparabil- Mina holds a master’s degree in economics for ity, and scalability of education data. Her work development from the University of Oxford and a focuses on automated data quality assurance. master’s degree in economics from the National Tianshu is a PhD candidate in economics and University of Sciences and Technology, Pakistan. education at the Teachers College, Columbia University, and she holds a bachelor’s degree in Soofia Yunus is a Senior Health Specialist at the operations research from Columbia University. World Bank Pakistan Country Office who recent- ly joined the country team. She is a medical doc- Yi Ning Wong  is a Consultant in the Education tor by qualification and holds a fellowship degree Global Practice, where she focuses on improv- in obstetrics and gynecology and a master’s de- ing data access for policy. Her work includes gree in public health. Soofia brings more than learning loss simulations and the Learning Data 20 years of experience working with the public Compact. She holds a master’s degree in pub- sector in various positions in hospital settings, lic policy from the University of Chicago and such as deputy director in the Ministry of Pop- a bachelor of arts degree in political science ulation Welfare, the Ministry of National Health and screenwriting from Loyola Marymount Services, Regulations and Coordination and University. Expanded Program on Immunization, and has been engaged in policy-level decisionmaking in Ziauddin Hyder is a Senior Nutrition Specialist Pakistan as a director-general. She also served in at the World Bank based in Washington, DC. In Saudi Arabia for several years. Her recent contri- his current role, he supports health and nutri- bution was in planning and promoting COVID-19 tion operations and assists in designing and vaccination roll-out, managing the immunization implementing multisectoral nutrition and health supply chain across Pakistan, and coordinating system reforms. He guided the preparation of with all donors and international partners. At the first Bangladesh National Plan of Action the World Bank, she is part of teams working on for Nutrition. His hard work took micronutrient health, nutrition, and pandemic preparedness. sprinkles from research to policy to scaled up programs across countries. He contributed to Thelma Seoeun Choi  is an Economist in the So- strengthening the SUN movement by offering cial Protection and Jobs Global Practice. She technical support to countries in Africa in de- supports lending operations and undertakes an- veloping costed nutrition action and investment alytical activities in Bangladesh, Bhutan, and Pa- plans. He was trained as a medical doctor and kistan, focusing on labor markets and jobs, eco- has a master’s degree in food and nutrition plan- About the authors | Pakistan Human Capital Review nomic inclusion, social insurance systems, and ning and a PhD in nutritional epidemiology. human capital development. Before joining the Bank, Thelma was a Research Analyst at the In- Zohair Zaidi i s Consultant in the World Bank’s ternational Monetary Fund, where she focused on Education Global Practice. He works on analyt- economic diversification, financial markets, and ical and operational aspects of education pro- macroeconomic analysis. Thelma has a bachelor gramming in Pakistan. His research interests of arts degree in political science from Yonsei Uni- include education governance systems and versity and a master’s degree in international eco- politics of public service delivery in low-income nomics and development from the Johns Hopkins country contexts. He previously worked with School of Advanced International Studies. Mott Macdonald, Development Alternatives In- corporated, Oxfam Great Britain, and the Gov- Tianshu Guo is a data scientist on the World ernment of Pakistan. Zohair holds a master’s de- Bank’s Education Analytics Team, where she gree in public affairs from Princeton University. 208