ognitive, language, and social- emotional skills. Policy recommendations This chapter establishes that children and adolescents in some types of homes do not receive sufficient care and that resources alone are not sufficient to keep human capital accumulation on track. Because both resources and the care environment are important in human capital accumulation among children and adolescents, a policy agenda to promote human capital would address the immediate resource constraints faced by poor households and seek to improve the care children receive. While most countries implement some form of income redistribution, even if the goal is not to augment the economy’s human capital, improving the care environment for the benefit of children is rarely a policy objective. Increasing the resources of poor families Table 2.1 lists two recommendations to increase the resources available to poor families and assesses the magnitude of the effects estimated in studies and the strength of the evidence. The table also details known challenges associated with implementation and the design features that still require experimentation. Evidence on several countries shows that raising parental income through job programs or income support may significantly increase children’s participation in school, academic achievement, and social-emotional development. In Uganda, for example, the Youth Opportunities Program raised the earnings of young adults by 38 percent and household consumption by 10 percent. Yet, because of the program, young people were also less likely to work and more likely to continue in school.23 Likewise, in Nepal, programs offering vocational training and microcredit to women led to a 10 percentage point rise in maternal nonfarm self-employment and household earnings and higher school participation among their children.24 Human Capital Accumulation in the Home 33 TABLE 2.1 Evidence on increasing resources at home among poor families Pro-poor job programs Public works, self-employment support, incentives for skill development, and job-matching services. Estimates of impact on human capital: Mediuma Strength of evidence: Mediumb Summary: Increasing parental earnings translates into increased investments in children’s human capital, particularly in terms of school participation. Implementation challenges: On employment and earnings, the general macroeconomic context may exert more influence than publicly funded programs. Still to learn: What types of employment support and for whom translate the most into human capital investments in children and adolescents? Cash transfers to poor families Payments directly to poor households made by the government. Estimates of impact on human capital: Mediuma Strength of evidence: Highb Summary: Cash transfers for poor households tend to increase food expenditures and the use of education and health services, particularly when payments are conditional on a minimum level of service usage. Their effects on skill formation and malnutrition of children and the mental health of adults have been negligible to modest.c Implementation challenges: Monitoring adherence to conditionalities can be complex and expensive. In some contexts, but not in others, the influx of cash increases price levels, and households just below eligibility thresholds have to reduce essential expenditures.d Still to learn: Should transfers target a specific household member (for example, mothers)? Should transfers be conditional or labeled for a specific purpose? How long should households have access to transfers? Is there a threshold value above which transfers may also more sizably and consistently improve skills, nutrition, and mental health?e Source: Original table for this publication. a. The estimated impact is high if the estimated impacts on the human capital accumulation of children persist over multiple stages of the life cycle or generate immediate impacts equivalent to at least 0.5 standard deviations in low- and middle-income countries. It is medium if the immediate impact is between 0.25 and 0.50 standard deviations. b. The strength of evidence is high if there are multiple meta-analyses focused on experimental or quasi-experimental evidence. At least one meta-analysis must include studies in low- and middle-income countries that find a significant average size in the effects. It is medium if there are more than five experimental or quasi-experimental studies in low- and middle-income countries. It is low if there are fewer than five studies in low- and middle-income countries or if all evidence is limited to high-income contexts. c. Baird et al. (2013); Manley et al. (2022); McGuire et al. (2022). d. Cunha et al (2019); Egger et al. (2022); Filmer et al. (2023). e. Baird et al. (2016). There is evidence that, if women earn more in the labor market, the human capital of their children increases.25 An expansion in female labor force participation, however, need not occur at the expense of declines in care for children (refer to chapter 4, box 4.3). 34 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces By contrast, declines in labor income adversely affect the human capital of children. In Brazil, for instance, job losses among parents raised the risk of criminal behavior among both parents and children, although these effects were fully offset if unemployment insurance cushioned the income shock.26 Cash transfer programs, through which the government assigns payments directly to households, represent explicit attempts to augment the resources available in the home. Similar direct transfers of goods and money have been used by governments for thousands of years to (re)distribute income or to assist the vulnerable, such as children and individuals with disabilities or war-related injuries.27 Between 2010 and 2019, they covered more than 400 million people in low- and middle-income countries. They reached a peak coverage of 900 million people during the COVID-19 pandemic.28 These programs can provide regular, automatic payments to all households that meet certain socioeconomic criteria or they may condition payments based on household investments in children’s human capital (for instance, attaining a minimum school attendance rate of 80 percent) or on the participation of adult household members in the labor market. Cash transfers to poor families may increase human capital investments in children in the short term.29 In a pilot trial in Malawi, for instance, cash transfers that were equivalent to approximately 10 percent of annual household expenditures and were conditional on school attendance by adolescent girls raised enrollment by 11 percentage points (or 16 percent). They also generated modest gains in reading and cognitive scores.30 An unconditional variant of the program that did not require school attendance did not lead to these gains in education, but was associated with a reduction in teenage marriage (by 8 percentage points or 44 percent) and teenage pregnancy (by 7 percentage points or 27 percent). While cash transfers may translate into human capital investments, the impact of cash transfers on outcomes among children—their physical growth and skill development—has been much more limited.31 The available evidence shows that there is still much to learn about the design features in programs that lead to success, such as who is eligible, who within the household actually receives the transfers, and how long do they receive the transfers. Eligibility rules may become important if ineligible households are still vulnerable or if eligible households are not vulnerable. In the Philippines, for example, households that were ineligible for the transfers were priced out of some foods when the growth in demand associated with households that received the transfers led to higher prices. As a result, chronic malnutrition rose by 34 percent among some ineligible households.32 Likewise, little is known about the optimal duration or size of transfers, and few governments in low- and middle-income countries would be able to finance transfers for years if the transfers represent substantial changes in income. Human Capital Accumulation in the Home 35 Improving the care environment of children and adolescents What happens if programs or policies are aimed explicitly at altering the care environment children experience without augmenting resource availability? Table 2.2 lists recommendations for these types of programs, illustrates the magnitude and strength of the evidence base, and documents known implementation challenges and what remains to be learned. Parenting programs Parenting programs counsel parents on activities they can undertake in the home to provide early cognitive and social-emotional stimulation to their children, often with an element of demonstration. Parents may learn how to make simple toys using materials commonly found in the home and use them to promote early TABLE 2.2 Evidence on improving care environments for children and adolescents Parenting programs Counseling and demonstrations among parents to encourage early cognition stimulation and social-emotional support, typically implemented during home visits or small group sessions. Estimates of impact on human capital: Higha Strength of evidence: Highb Summary: Parenting programs tend to improve children’s early cognitive, motor, and social-emotional skills and may help reduce violence in the home. Longitudinal evidence shows long-term benefits in adulthood on earnings, proclivity to commit crime, and mental health.c Implementation challenges: The estimated effects of parenting programs tend to decline with scale. Implementation fidelity tends to be weak. Home visits occur at a lower frequency than anticipated. Front-line workers often need to shift attention to other tasks, such as nutrition counseling.d Still to learn: • Should households be targeted according to poverty status, maternal education, or a measure of child development? • What qualifications do front-line staff require? Should they dedicate their time to parenting programs, or can workers take on delivery as an additional task? Should personnel be paid through a salary or stipend? • How effective are parenting programs for school-age children and adolescents? Preprimary education Center- or school-based education among children ages 3–6. Estimates of impact on human capital: Higha Strength of evidence: Highb Summary: Preprimary education immediately improves children’s behavior and their skills in literacy, mathematics, and social-emotional learning. Benefits persist into adulthood. As adults, the individuals eventually complete more education, earn more, and commit less crime.e Implementation challenges: If preprimary expansion causes students to switch schools (because of a subsidy, school construction closer to home, and so on), children’s skills may decline if they were previously attending higher-quality schools. Improving quality in existing schools does not appear to be a challenge, however.f Still to learn: • Can preschools be used to improve nutrition and health among children?g • Do the services offered to parents (such as job training or mental health counseling) augment impacts? (Table continues on next page) 36 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces TABLE 2.2 Evidence on improving care environments for children and adolescents (continued) Mental health care among adults Estimates of impact on human capital: Mediuma Strength of evidence: Highb Summary: Psychotherapy tends to improve the mental health of adults, who then invest more in their children’s education. While mental health impacts can be large and persistent, the effects on children’s education are modest.h Implementation challenges: Delivery first requires the screening and diagnosis of mental health disorders, but mental health care personnel tend to be scarce in low- and middle-income countries.i Source: Original table for this publication. a. The estimated impact is high if the estimated impacts on the human capital accumulation of children persist over multiple stages of the life cycle or generate immediate impacts equivalent to at least 0.5 standard deviations in low- and middle-income countries. It is medium if the immediate impact is between 0.25 and 0.50 standard deviations. b. The strength of evidence is high if there are multiple meta-analyses focused on experimental or quasi-experimental evidence. At least one meta-analysis must include studies in low- and middle-income countries that find a significant average size in the effects. It is medium if there are more than five experimental or quasi-experimental studies in low- and middle-income countries. It is low if there are fewer than five studies in low- and middle-income countries or if all evidence is limited to high-income contexts. c. Gertler et al. (2014, 2021); Grantham-McGregor et al. (1999); Hamadani et al. (2006); Jensen et al. (2021); Jeong et al. (2020); Jervis et al. (2023); Walker et al. (2022); Yousafzai et al. (2014). d. Andrew et al. (2018); Araujo et al. (2021); Bos et al. (2024); Ganimian et al. (2024); Kirkwood et al. (2023). e. Bailey et al. (2021); Holla et al. (2021). f. Berkes et al. (2024); Dean and Jayachandran (2020); Ganimian et al. (2024); Gray-Lobe et al. (2022); Wolf et al. (2019). g. Carneiro and Ginja (2014); Sommer et al. (2024). h. Baranov et al. (2020); Bhat et al. (2022); Lund et al. (2024); Tol et al. (2019). i. WHO (2021). numeracy skills among children under age 3, or they may learn how to reduce their reliance on physical violence to maintain discipline. Though most parents do not endorse the use of violent punishment, many report that they use such measures in the home (refer to box 2.1). Evidence from numerous small-scale randomized controlled trials suggest that parenting programs are effective in immediately improving children’s cognitive, motor, and social-emotional skills.33 Longitudinal evidence from randomized controlled trials shows impacts that persist into adulthood (refer to figure 2.8). Children benefiting from early stimulation programs exhibit higher educational attainment, earn more when they enter the labor market as adults, commit fewer crimes, and face a lower probability of suffering from mental health disorders.34 In an experiment in Jamaica, for instance, community workers made weekly home visits to households with stunted toddlers to demonstrate the way to provide cognitive and social-emotional stimulation to the young children. These children grew up to have 43 percent higher wages as adults and lower rates of depression and substance abuse compared with children who had not benefited from the program.35 Human Capital Accumulation in the Home 37 FIGURE 2.8 Lifelong effects of preschool and parenting programs a. Parenting programs try to change care inside the home. A program in Jamaica led to increases in early measures of cognitive, language, motor, and social-emotional skills. This is consistent with evidence from randomized controlled trials around the world. Improved cognitive and language skills were still evident when children were adolescents, as was a decrease in self-reported depression and anxiety. As they became older, there was a decrease in violent behavior and an increase in educational attainment, as well as an earnings advantage of 25 percent at age 22. Early childhood School age Adolescence and youth Adulthood These early improvements persisted, manifesting as better cognitive and language skills. There were persistent advantages in health and less substance abuse. By age 31, earnings had improved by 37 percent. b. Preschools can be a scalable alternative for altering care environments. Around the world, preschool programs increase both cognitive and social-emotional skills during the preprimary period, and these effects persist into primary school. Benefits continue into adulthood in the form of increased likelihood of employment and higher earnings, as well as decreases in poverty, reliance on social assistance, and engagement in crime. Early childhood and school age Adolescence and youth Adulthood Second generation Available longitudinal data suggest preschool can decrease behavioral problems and obesity in early adolescence and engagement in criminal activities and idleness among young adults. Children benefiting from preschool attain more education overall. When they are adults, children who benefited from a preschool program have children who attain more education, bear fewer children as teenagers, and commit less crime compared with the offspring of adults who did not benefit from preschool as children. Source: Original figure for this publication. 38 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces Parenting programs have proven difficult to scale, however.36 Home visits involve one-on-one interactions between parents and facilitators, and programs designed to have frequent visits can encounter difficulty in achieving implementation fidelity.37 The large documented effects have been estimated in small-scale studies evaluating programs that cater to as few as 70 children and that achieve a high frequency of home visits (weekly) over a long period (24 months). Evidence is emerging that group sessions in which 8 to 12 parents in a community meet a facilitator can achieve impacts equivalent to individual home visits at a much lower cost.38 Little evidence is available on whether parenting programs can be virtual or implemented using text, voice recordings, or phone calls. Some evidence indicates that virtual interventions can improve parenting practices and children’s outcomes, but other studies find no impact on child development.39 More research is required to learn about optimal design beyond scale. For example, should these programs target households with few resources, similar to the way social assistance may be targeted? Or should the programs rely on other markers to indicate the need for improvement in the care environment, such as maternal education or characteristics of children, including their nutritional status or early vocabulary skills? Many programs operate through existing service delivery infrastructure and add parenting programs to the tasks of existing community health workers, many of whom are part-time and not paid regular salaries. It is an open question whether implementation fidelity could improve with a dedicated, professionalized workforce. Most trials have been focused on young children, typically younger than age 6. Parents, however, likely need advice on promoting the cognitive development and mental health of their children during later stages of childhood, including adolescence. Evidence shows that these programs can also improve parenting practices and children’s outcomes if they are implemented during adolescence, but this is an emerging field of research.40 Mental health care for adults A randomized controlled trial in rural Pakistan among pregnant women and mothers of newborns who exhibited depressive symptoms demonstrates how parental mental health can interfere with human capital investments in children. When these women participated in the cognitive behavioral therapy delivered in their homes by community health workers through the Thinking Healthy Program, they exhibited lower probability of postpartum depression over the next year.41 They were 26 percentage points (or 60 percent) more likely to report that they spent time playing with their children than mothers who had not received treatment through the program. The children of the treated mothers were also Human Capital Accumulation in the Home 39 10 percent more likely to have completed their scheduled immunizations by the time they were 12 months old. The mental health benefits among the mothers persisted even seven years after the program. The children benefited because their parents invested more time and money in education by, for example, paying for more expensive, higher-quality schools and helping their children study at home.42 Nonspecialist counselors have delivered treatments in high-impact trials. This type of worker may be needed to deliver programs addressing mental health because there is a severe shortage of mental health personnel around the world. Globally, there are 13 mental health workers for every 100,000 people, and fewer than 2 for every 100,000 in low-income countries, compared with more than 60 in high-income countries. Preprimary education Given the challenges related to scale and the inherent difficulty of changing behavior within the home, it may be easier and more cost-effective to use preschools to provide an alternative care environment for children. Preschools exhibit wide coverage in upper-middle-income and high-income countries and have been used to deliver other services, such as basic health care services for children. Global evidence shows that children’s health and the development of both cognitive and social-emotional skills improve if children attend preschool, even in low-resource settings.43 A community preschool scale-up in Mozambique, for instance, improved skills in the preprimary period, increased school attendance during the primary period a few years later, and expanded the study time of older children who would have otherwise had to care for their preschool- age siblings.44 Longitudinal evidence available only on high-income countries indicates that preschool, similar to the effects of parenting programs, confers lifelong benefits (refer to figure 2.8). Children grow up to complete more education, commit less crime, earn more in the labor market, and rely less on social assistance.45 Their own children also accumulate more human capital.46 If preschool services are to be effective, however, they must provide care and stimulation that is better relative to what is provided to children at home or through other care arrangements. As the evidence presented in this chapter demonstrates, the average home environment in many contexts in low- and middle-income countries may not be conducive to the development of the early numeracy and literacy skills children will need to be prepared to learn in primary school. If preschool expansion leads children already participating in preprimary or primary education to switch to new schools, their skill development may not accelerate or may even decline if they switch out of schools of equal or higher quality.47 Thus, expansion would ideally first target those children who would not otherwise attend preschool. 40 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces In contrast to parenting programs, preschool has been successfully scaled up in many countries. Teachers need not have as many qualifications or as much training as teachers in higher grades. In India, secondary-school graduates with only two weeks of training were able to accelerate skill development.48 Teachers in Kenya with fewer qualifications than standard civil service teachers were able to generate skill improvements equivalent to more than one year of learning if they followed a structured curriculum and established recommendations for pedagogy.49 Preschool pupils also need not have as much instructional time as children in higher grades. In Bangladesh, two hours a day was sufficient to improve literacy, numeracy, and social-emotional skills among four-year-old children.50 The learning agenda in preprimary education policy involves complementary programs that may provide other services for children and their families, such as health screenings, counseling, cash transfers, or job training. Such a two-generation approach would simultaneously supply services to support skill development among children and interventions to bolster the livelihoods and well-being of parents and their relationships with their children.51 Increasing the education of future parents Increasing the resources of parents, parenting programs, and preprimary education can immediately address shortfalls in the resources and care required for human capital accumulation at home. Given the large differences observed across the world among children whose mothers vary in educational attainment, and given the evidence establishing a causal link between the human capital of individuals and the subsequent human capital accumulation among their children, investments in education today might be considered investments in the human capital of future parents.52 They are thus likely to increase the resources and care invested in human capital accumulation in the home. In a randomized controlled trial in Ghana, for example, girls who were offered a scholarship for secondary school were 28 percentage points (63 percent) more likely to complete secondary school and 7.1 percentage points (36 percent) more likely as adults to have partners who had completed tertiary education. They also went on to have children who were healthier—they were much less likely to die as infants or toddlers—and who exhibited more advanced cognitive skills than the children of women who had not been offered a scholarship.53 These effects do not appear to operate through increases in earnings or purchases of inputs, which were similar among mothers who had not been offered secondary-school scholarships when they were younger. Instead, the mothers who had won scholarships earlier talked more with their children and provided more stimulation. Thus, education confers parents with skills that are useful in producing human capital within the home. Over the long term, increasing and Human Capital Accumulation in the Home 41 improving the education of future parents may therefore be critical to enhancing human capital accumulation. In sum, human capital accumulation starts early and starts in the home. The family exerts a powerful influence on child and adolescent health and skill development through the resources available at home for purchasing important inputs, such as nutrient-rich foods or children’s books, and through the amount of nurturing care household members provide to stimulate learning and keep children safe. Policies need to target the entire home, both the resources and care environments that are essential for human capital accumulation. While increasing resources among poor households is already a policy goal to keep families from falling into poverty and to protect them from income shocks, policies that increase these resources, such as jobs programs or cash transfers, are typically not considered instruments for improving learning and educational attainment or mental health among adolescents. This chapter provides evidence that these policies do indeed yield the types of impacts that are often more closely associated with schools and clinics. Increasing resources for poor families will not be enough, however. Policies will need to address the care environment as well. While influencing how parents raise their children may seem an intrusion of policy into the private sphere of the household, the evidence presented in this chapter indicates that programs offering parents the tools to stimulate their children, provide social-emotional support, and rely on nonviolent behavior management yield large benefits throughout the life cycle, such as higher educational attainment and earnings and less reliance on social assistance and criminal activity. While the school is often considered the proper place to integrate the delivery of many services, such as education, nutrition, and health care, the home may also be viewed as an appropriate place to undertake such service integration. Likewise, the evidence shows that the home is not important solely for young children. The resources and care supplied within the home also affect human capital accumulation among adolescents. Increased support for human capital accumulation at home may require an expansion in the scope of activities in some sectors. While social protection programs sometimes deploy social workers for home visits, and the health system may rely on outreach by community health workers, current services reaching the home tend not to reflect how critical the home is in the development of cognitive and social-emotional skills. Integrating programs that represent an acknowledgment of this important role of the family will prepare children and adolescents to make the transitions to higher stages of school and to work that will determine health and labor productivity in adulthood. These transitions will occur when other places, such as the neighborhood and the workplace, begin to exercise greater relevance in human capital accumulation. 42 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces Notes 1. Sonuga-Barke et al. (2017). 2. Almond and Currie (2011); Heckman (2006). 3. Rao et al. (2013). 4. Hargrave and Sénéchal (2000); Knauer et al. (2020). 5. Baird, Friedman, et al. (2011). 6. Adhvaryu et al. (2019). 7. Akee et al. (2010); Akee et al. (2018). 8. Akee et al. (2010). 9. Akee et al. (2018); Akee et al. (2024). 10. Akee et al. (2018). 11. Dewey et al. (2022). 12. Cattaneo et al. (2009). 13. Cuijpers et al. (2015); Lund et al. (2024). 14. Pierce et al. (2020). 15. Rege et al. (2011). 16. Data of GBD 2021 (Global Burden of Disease Study 2021) Data Resources, Global Health Data Exchange, Institute for Health Metrics and Evaluation (accessed June 4, 2025), https://ghdx .healthdata.org/gbd-2021. Mental disorders include depressive disorders, anxiety disorders, schizophrenia, autism spectrum disorders, bipolar disorder, conduct disorder, idiopathic developmental intellectual disability, eating disorders, and attention-deficit hyperactivity disorder. 17. Refer to CFPS (China Family Panel Studies) (dashboard), Institute of Social Science Survey, Peking University, https://www.isss.pku.edu.cn/cfps/en/. 18. These income advantages relative to children who have never been left behind persist among all categories of children who have ever been left behind if income is measured on a per capita basis across all categories, except for children who are chronically left behind. 19. Barcellos et al. (2014). 20. Jayachandran and Pande (2017). 21. Martinez et al. (2017). 22. Jakiela et al. (2024). 23. Blattman et al. (2014). 24. Chakravarty et al. (2019). 25. Jayachandran and Voena (2025). 26. Britto et al. (2022). 27. Gentilini (2024). 28. Gentilini et al. (2022). 29. Baird, McIntosh, et al. (2011); Fernald and Hidrobo (2011). 30. Baird, McIntosh, et al. (2011). 31. Baird et al. (2014); Manley et al. (2022). 32. Filmer et al. (2023). 33. Jeong et al. (2021); Jervis et al. (2023). 34. Garcia et al. (2016); Heckman et al. (2013). Human Capital Accumulation in the Home 43 35. Gertler et al. (2021); Walker et al. (2022). 36. Araujo et al. (2021). 37. Andrew et al. (2018); Bos et al. (2024); Kirkwood et al. (2023). 38. Grantham-McGregor et al. (2020). 39. Arteaga et al. (2023); Arteaga et al. (2025); Dinarte-Díaz et al. (2023); Francis et al. (2024); Rafla et al. (2024). 40. Cluver et al. (2018); Goagoses et al. (2023). 41. Rahman et al. (2008). 42. Baranov et al. (2020). 43. Holla et al. (2021). 44. Martinez et al. (2017). 45. Bailey et al. (2021); Gray-Lobe et al. (2023); Havnes and Mogstad (2011); Heckman et al. (2010). 46. Rossin-Slater and Wüst (2020). 47. Bouguen et al. (2018); Dean and Jayachandran (2020). 48. Ganimian et al. (2024). 49. Gray-Lobe et al. (2022). 50. Spier et al. (2020). 51. Sommer et al. (2024). 52. Heckman and Karapakula (2019); Rossin-Slater and Wüst (2020). 53. Duflo et al. (2024). References Adhvaryu, Achyuta, James Fenske, and Anant Nyshadham. 2019. “Early Life Circumstance and Adult Mental Health.” Journal of Political Economy 127 (4): 1516–49. Akee, Randall K. Q., William E. Copeland, E. Jane Costello, and Emilia Simeonova. 2018. “How Does Household Income Affect Child Personality Traits and Behaviors?” American Economic Review 108 (3): 775–827. Akee, Randall K. Q., William E. Copeland, Gordon Keeler, Adrian Angold, and E. Jane Costello. 2010. “Parents’ Incomes and Children’s Outcomes: A Quasi-Experiment Using Transfer Payments from Casino Profits.” American Economic Journal: Applied Economics 2 (1): 86–115. Akee, Randall K. Q., William E. Copeland, and Emilia Simeonova. 2024. “Child Mental Health, Family Circumstance, and Long-Term Success: The Effect of Household Income.” Journal of Human Resources 59 (April): S77–S107. Almond, Douglas, and Janet Currie. 2011. “Human Capital Development Before Age Five.” In Handbook of Labor Economics, vol. 4, part B, edited by Orley C. Ashenfelter and David E. Card. North-Holland. Andrew, Alison, Orazio Pietro Attanasio, Emla Fitzsimons, Sally M. Grantham-McGregor, Costas Meghir, and Marta Rubio-Codina. 2018. “Impacts 2 Years After a Scalable Early Childhood Development Intervention to Increase Psychosocial Stimulation in the Home: A Follow-Up of a Cluster Randomised Controlled Trial in Colombia.” PLoS Medicine 15 (4): e1002556. Araujo, María Caridad, Marta Dormal, Sally M. Grantham-McGregor, Fabiola Lazarte, Marta Rubio- Codina, and Norbert R. Schady. 2021. “Home Visiting at Scale and Child Development.” Journal of Public Economics Plus 2 (5): 100003. Arteaga, Irma, Andreas de Barros, and Alejandro J. Ganimian. 2025. “The Challenges of Scaling Up Effective Child-Rearing Practices Using Technology in Developing Settings: Experimental Evidence from India.” Journal of Research on Educational Effectiveness. Published ahead of print, March 10, 2025. https://doi.org/10.1080/19345747.2025.2450318. 44 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces Arteaga, Irma, Julieta Trias, and Miriam Martinez. 2023. “Can Technology Narrow the Early Childhood Stimulation Gap in Rural Guatemala?” April, World Bank. https://thedocs.worldbank .org/en / doc/2155ea93406ba0692d191c2b0dea8ec7-0090052023/original/Irma-SIEF-panel -April12 -2023 -vf .pdf. Bailey, Martha J., Shuqiao Sun, and Brenden Timpe. 2021. “Prep School for Poor Kids: The Long-Run Impacts of Head Start on Human Capital and Economic Self-Sufficiency.” American Economic Review 111 (12): 3963–4001. Baird, Sarah Jane, Francisco H. G. Ferreira, Berk Özler, and Michael Woolcock. 2013. “Relative Effectiveness of Conditional and Unconditional Cash Transfers for Schooling Outcomes in Developing Countries: A Systematic Review.” Campbell Systematic Reviews 8 (September), Campbell Collaboration. Baird, Sarah Jane, Francisco H. G. Ferreira, Berk Özler, and Michael Woolcock. 2014. “Conditional, Unconditional, and Everything in Between: A Systematic Review of the Effects of Cash Transfer Programmes on Schooling Outcomes.” Journal of Development Effectiveness 6 (1): 1–43. Baird, Sarah Jane, Jed Friedman, and Norbert R. Schady. 2011. “Aggregate Income Shocks and Infant Mortality in the Developing World.” Review of Economics and Statistics 93 (3): 847–56. Baird, Sarah Jane, Craig T. McIntosh, and Berk Özler. 2011. “Cash or Condition? Evidence from a Cash Transfer Experiment.” Quarterly Journal of Economics 126 (4): 1709–53. Baird, Sarah Jane, Craig T. McIntosh, and Berk Özler. 2016. “When the Money Runs Out: Do Cash Transfers Have Sustained Effects on Human Capital Accumulation?” Policy Research Working Paper 7901, World Bank. Baranov, Victoria, Sonia Bhalotra, Pietro Biroli, and Joanna Maselko. 2020. “Maternal Depression, Women’s Empowerment, and Parental Investment: Evidence from a Randomized Controlled Trial.” American Economic Review 110 (3): 824–59. Barcellos, Silvia Helena, Leandro Siqueira Carvalho, and Adriana Lleras-Muney. 2014. “Child Gender and Parental Investments in India: Are Boys and Girls Treated Differently?” American Economic Journal: Applied Economics 6 (1): 157–89. Berkes, Jan, Adrien Bouguen, Deon Filmer, and Tsuyoshi Fukao. 2024. “Improving Preschool Provision and Encouraging Demand: Evidence from a Large-Scale Construction Program.” Journal of Public Economics 230 (February): 105050. Bhat, Bhargav, Jonathan de Quidt, Johannes Haushofer, et al. 2022. “The Long-Run Effects of Psychotherapy on Depression, Beliefs, and Economic Outcomes.” NBER Working Paper 30011 (May), National Bureau of Economic Research. Blattman, Christopher, Nathan Vincent Fiala, and Sebastián Martínez. 2014. “Generating Skilled Self- Employment in Developing Countries: Experimental Evidence from Uganda.” Quarterly Journal of Economics 129 (2): 697–752. Bos, Johannes M., Abu Syeid Mohammad Parve Shonchoy, Saravana Ravindran, and Akib Khan. 2024. “Early Childhood Human Capital Formation at Scale.” Journal of Public Economics 231 (March): 105046. Bouguen, Adrien, Deon Filmer, Karen Macours, and Sophie Naudeau. 2018. “Preschool and Parental Response in a Second Best World: Evidence from a School Construction Experiment.” Journal of Human Resources 53 (2): 474–512. Britto, Diogo G. C., Paolo Pinotti, and Breno Sampaio. 2022. “The Effect of Job Loss and Unemployment Insurance on Crime in Brazil.” Econometrica 90 (4): 1393–423. Carneiro, Pedro Manuel, and Rita Ginja. 2014. “Long-Term Impacts of Compensatory Preschool on Health and Behavior: Evidence from Head Start.” American Economic Journal: Economic Policy 6 (4): 135–73. Cattaneo, Matias D., Sebastian Galiani, Paul J. Gertler, Sebastian Martinez, and Rocio Titiunik. 2009. “Housing, Health, and Happiness.” American Economic Journal: Economic Policy 1 (1): 75–105. Chakravarty, Shubha, Mattias K. A. Lundberg, Plamen Nikolov, and Juliane Zenker. 2019. “Vocational Training Programs and Youth Labor Market Outcomes: Evidence from Nepal.” Journal of Development Economics 136 (January): 71–110. Human Capital Accumulation in the Home 45 Cluver, Lucie D., Franziska Meinck, Janina I. Steinert, et al. 2018. “Parenting for Lifelong Health: A Pragmatic Cluster Randomised Controlled Trial of a Non-Commercialised Parenting Programme for Adolescents and Their Families in South Africa.” BMJ Global Health 3 (1): e000539. Cuijpers, Pim, Erica Weitz, Eirini Karyotaki, Judy Garber, and Gerhard Andersson. 2015. “The Effects of Psychological Treatment of Maternal Depression on Children and Parental Functioning: A Meta- Analysis.” European Child and Adolescent Psychiatry 24 (2): 237–45. Cunha, Jesse M., Giacomo De Giorgi, and Seema Jayachandran. 2019. “The Price Effects of Cash Versus In-Kind Transfers.” Review of Economic Studies 86 (1): 240–81. Dean, Joshua T., and Seema Jayachandran. 2020. “Attending Kindergarten Improves Cognitive Development in India, but All Kindergartens Are Not Equal.” Working paper (July 23), Northwestern University; Booth School of Business, University of Chicago. Dewey, Kathryn G., Charles D. Arnold, K. Ryan Wessells, et al. 2022. “Preventive Small-Quantity Lipid- Based Nutrient Supplements Reduce Severe Wasting and Severe Stunting Among Young Children: An Individual Participant Data Meta-Analysis of Randomized Controlled Trials.” American Journal of Clinical Nutrition 116 (5): 1314–33. Dinarte-Díaz, Lelys Ileana, María Marta Ferreyra, Sergio Urzúa, and Marina Bassi. 2023. “What Makes a Program Good? Evidence from Short-Cycle Higher Education Programs in Five Developing Countries.” World Development 169 (September): 106294. Duflo, Esther, Pascaline Dupas, Elizabeth Spelke, and Mark P. Walsh. 2024. “Intergenerational Impacts of Secondary Education: Experimental Evidence from Ghana.” NBER Working Paper 32742 (July), National Bureau of Economic Research. Dunn, Lloyd M., and Leola M. Dunn. 1997. Peabody Picture Vocabulary Test, 3rd ed. American Guidance Service. Egger, Dennis, Johannes Haushofer, Edward Andrew Miguel, Paul Niehaus, and Michael Walker. 2022. “General Equilibrium Effects of Cash Transfers: Experimental Evidence From Kenya.” Econometrica 90 (6): 2603–43. Fernald, Lia C. H., and Melissa Lucia Hidrobo. 2011. “Effect of Ecuador’s Cash Transfer Program (Bono de Desarrollo Humano) on Child Development in Infants and Toddlers: A Randomized Effectiveness Trial.” Social Science and Medicine 72 (9): 1437–46. Filmer, Deon, Jed Friedman, Eeshani Kandpal, and Junko Onishi. 2023. “Cash Transfers, Food Prices, and Nutrition Impacts on Ineligible Children.” Review of Economics and Statistics 105 (2): 327–43. Francis, Taja, Lelys Ileana Dinarte-Díaz, Shawn Powers, and Helen Baker-Henningham. 2024. “The Virtual Irie Homes Toolbox: Adaptation and Remote Delivery of an Early Childhood, Violence Prevention, Parenting Program in Jamaica.” Journal of Research in Childhood Education 38 (Supplement 1): S92–S111. Ganimian, Alejandro J., Karthik Muralidharan, and Christopher R. Walters. 2024. “Augmenting State Capacity for Child Development: Experimental Evidence from India.” Journal of Political Economy 132 (5): 1565–602. García, Jorge Luis, James J. Heckman, Duncan Ermini Leaf, and María José Prados. 2016. “The Life-Cycle Benefits of an Influential Early Childhood Program.” NBER Working Paper 22993, National Bureau of Economic Research. Gentilini, Ugo. 2024. Timely Cash: Lessons from 2,500 Years of Giving People Money. Oxford University Press. Gentilini, Ugo, Mohamed Almenfi, Hrishikesh T. M. M. Iyengar, et al. 2022. Social Protection and Jobs Responses to COVID-19: A Real-Time Review of Country Measures. Living Paper Version 16 (February 2, 2022). World Bank. http://hdl.handle.net/10986/37186. Gertler, Paul J., James J. Heckman, Rodrigo Pinto, et al. 2014. “Labor Market Returns to an Early Childhood Stimulation Intervention in Jamaica.” Science 344 (6187): 998–1001. Gertler, Paul J., James J. Heckman, Rodrigo Pinto, et al. 2021. “Effect of the Jamaica Early Childhood Stimulation Intervention on Labor Market Outcomes at Age 31.” NBER Working Paper 29292 (September), National Bureau of Economic Research. 46 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces Goagoses, Naska, Tijs Bolz, Jule Eilts, et al. 2023. “Parenting Dimensions/Styles and Emotion Dysregulation in Childhood and Adolescence: A Systematic Review and Meta-Analysis.” Current Psychology 42 (22): 18798–8822. Grantham-McGregor, Sally M., Akanksha Adya, Orazio Pietro Attanasio, et al. 2020. “Group Sessions or Home Visits for Early Childhood Development in India: A Cluster RCT.” Pediatrics 146 (6): e2020002725. Grantham-McGregor, Sally M., Lia C. H. Fernald, and Kavita Sethuraman. 1999. “Effects of Health and Nutrition on Cognitive and Behavioural Development in Children in the First Three Years of Life, Part 1: Low Birthweight, Breastfeeding, and Protein-Energy Malnutrition.” Food and Nutrition Bulletin 20 (1): 53–75. Gray-Lobe, Guthrie, Anthony Keats, Michael R. Kremer, Isaac Mbiti, and Owen Ozier. 2022. “Can Education Be Standardized? Evidence from Kenya.” Working Paper 2022-68 (September), Becker Friedman Institute, University of Chicago. Gray-Lobe, Guthrie, Parag A. Pathak, and Christopher R. Walters. 2023. “The Long-Term Effects of Universal Preschool in Boston.” Quarterly Journal of Economics 138 (1): 363–411. Hamadani, Jena D., Syed N. Huda, Fahmida Khatun, and Sally M. Grantham-McGregor. 2006. “Psychosocial Stimulation Improves the Development of Undernourished Children in Rural Bangladesh.” Journal of Nutrition 136 (10): 2645–52. Hargrave, Anne C., and Monique Sénéchal. 2000. “A Book Reading Intervention with Preschool Children Who Have Limited Vocabularies: The Benefits of Regular Reading and Dialogic Reading.” Early Childhood Research Quarterly 15 (1): 75–90. Havnes, Tarjei, and Magne Mogstad. 2011. “No Child Left Behind: Subsidized Child Care and Children’s Long-Run Outcomes.” American Economic Journal: Economic Policy 3 (2): 97–129. Heckman, James J. 2006. “Skill Formation and the Economics of Investing in Disadvantaged Children.” Science 312 (5782): 1900–02. Heckman, James J., and Ganesh Karapakula. 2019. “Intergenerational and Intragenerational Externalities of the Perry Preschool Project.” NBER Working Paper 25889 (May), National Bureau of Economic Research. Heckman, James J., Seong Hyeok Moon, Rodrigo Pinto, Peter A. Savelyev, and Adam Yavitz. 2010. “The Rate of Return to the HighScope Perry Preschool Program.” Journal of Public Economics 94 (1–2): 114–28. Heckman, James J., Rodrigo Pinto, and Peter Savelyev. 2013. “Understanding the Mechanisms Through Which an Influential Early Childhood Program Boosted Adult Outcomes.” American Economic Review 103 (6): 2052–86. Holla, Alaka, Maria Magdalena Bendini, Lelys Ileana Dinarte-Díaz, and Iva Trako. 2021. “Is Investment in Preprimary Education Too Low? Lessons from (Quasi) Experimental Evidence Across Countries.” Policy Research Working Paper 9723, World Bank. Jakiela, Pamela, Owen Whitfield Ozier, Lia C. H. Fernald, and Heather Ashley Knauer. 2024. “Preprimary Education and Early Childhood Development: Evidence from Government Schools in Rural Kenya.” Journal of Development Economics 171 (October): 103337. Jayachandran, Seema, and Rohini P. Pande. 2017. “Why Are Indian Children So Short? The Role of Birth Order and Son Preference.” American Economic Review 107 (9): 2600–29. Jayachandran, Seema, and Alessandra Voena. 2025. “Women’s Power in the Household.” CEPR Discussion Paper DP20938 (December), Centre for Economic Policy Research. Jensen, Sarah K. G., Matias Placencio-Castro, Shauna M. Murray, et al. 2021. “Effect of a Home-Visiting Parenting Program to Promote Early Childhood Development and Prevent Violence: A Cluster- Randomized Trial in Rwanda.” BMJ Global Health 6 (1): e003508. Jeong, Joshua, Emily E. Franchett, Clariana V. Ramos de Oliveira, Karima Rehmani, and Aisha K. Yousafzai. 2021. “Parenting Interventions to Promote Early Child Development in the First Three Years of Life: A Global Systematic Review and Meta-Analysis.” PLoS Medicine 18 (5): e1003602. Human Capital Accumulation in the Home 47 Jeong, Joshua, and Zhihui Li. 2020. “The Association Between Fathers’ Depression and Children’s Socioemotional Development: Evidence from a Longitudinal Household Survey in China.” Prevention Science 21 (5): 672–80. Jervis, Pamela, Jacqueline Coore-Hall, Helen O. Pitchik, et al. 2023. “The Reach Up Parenting Program, Child Development, and Maternal Depression: A Meta-Analysis.” Pediatrics 151 (Supplement 2): e2023060221D. Kirkwood, Betty R., Siham Sikander, Reetabrata Roy, et al. 2023. “Effect of the SPRING Home Visits Intervention on Early Child Development and Growth in Rural India and Pakistan: Parallel Cluster Randomised Controlled Trials.” Frontiers in Nutrition 10 (June): 1155763. Knauer, Heather Ashley, Pamela Jakiela, Owen Ozier, Frances E. Aboud, and Lia C. H. Fernald. 2020. “Enhancing Young Children’s Language Acquisition Through Parent-Child Book-Sharing: A Randomized Trial in Rural Kenya.” Early Childhood Research Quarterly 50 (1): 179–90. Lund, Crick, Kate Orkin, Marc Witte, et al. 2024. “The Effects of Mental Health Interventions on Labor Market Outcomes in Low- and Middle-Income Countries.” NBER Working Paper 32423 (May), National Bureau of Economic Research. Manley, James, Harold Alderman, and Ugo Gentilini. 2022. “More Evidence on Cash Transfers and Child Nutritional Outcomes: A Systematic Review and Meta-Analysis.” BMJ Global Health 7 (4): e008233. Martinez, Sebastian, Sophie Naudeau, and Vitor Azevedo Pereira. 2017. “Preschool and Child Development Under Extreme Poverty: Evidence from a Randomized Experiment in Rural Mozambique.” Policy Research Working Paper 8290, World Bank. McGuire, Joel, Caspar Kaiser, and Anders M. Bach-Mortensen. 2022. “A Systematic Review and Meta- Analysis of the Impact of Cash Transfers on Subjective Well-Being and Mental Health in Low- and Middle-Income Countries.” Nature Human Behaviour 6 (3): 359–70. Pierce, Matthias, Holly F. Hope, Adekeye Kolade, et al. 2020. “Effects of Parental Mental Illness on Children’s Physical Health: Systematic Review and Meta-Analysis.” British Journal of Psychiatry 217 (1): 354–63. Rafla, Joyce, Kate Schwartz, Hirokazu Yoshikawa, et al. 2024. “Cluster Randomized Controlled Trial of a Phone-Based Caregiver Support and Parenting Program for Syrian and Jordanian Families with Young Children.” Early Childhood Research Quarterly 69 (4th Quarter): 141–53. Rahman, Atif, Abid Malik, Siham Sikander, Christopher Roberts, and Francis Creed. 2008. “Cognitive Behaviour Therapy–Based Intervention by Community Health Workers for Mothers with Depression and Their Infants in Rural Pakistan: A Cluster-Randomised Controlled Trial.” Lancet 372 (9642): 902–09. Rao, Mayuree, Ashkan Afshin, Gitanjali Singh, and Dariush Mozaffarian. 2013. “Do Healthier Foods and Diet Patterns Cost More Than Less Healthy Options? A Systematic Review and Meta-Analysis.” BMJ Open 3 (12): e004277. Rege, Mari, Kjetil Telle, and Mark Votruba. 2011. “Parental Job Loss and Children’s School Performance.” Review of Economic Studies 78 (4): 1462–89. Rossin-Slater, Maya, and Miriam Wüst. 2020. “What Is the Added Value of Preschool for Poor Children? Long-Term and Intergenerational Impacts and Interactions with an Infant Health Intervention.” American Economic Journal: Applied Economics 12 (3): 255–86. Sommer, Teresa Eckrich, Emily E. Franchett, Hirokazu Yoshikawa, and Joan Lombardi. 2024. “A Global Call for Two-Generation Approaches to Child Development and Caregivers’ Livelihoods.” Child Development Perspectives 18 (4): 204–14. Sonuga-Barke, Edmund J. S., Mark Kennedy, Robert Kumsta, et al. 2017. “Child-to-Adult Neurodevelopmental and Mental Health Trajectories After Early Life Deprivation: The Young Adult Follow-Up of the Longitudinal English and Romanian Adoptees Study.” Lancet 389 (10078): 1539–48. Spier, Elizabeth, Kevin Kamto, Adria Molotsky, Azizur Rahman, Najmul Hossain, Zannatun Nahar, and Hosneara Khondker. 2020. Bangladesh Early Years Preschool Program Evaluation: Baseline Report for the World Bank Strategic Impact Evaluation Fund. April. American Institutes for Research. 48 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces Tol, Wietse A., Sarah McIvor Murray, Crick Lund, et al. 2019. “Can Mental Health Treatments Help Prevent or Reduce Intimate Partner Violence in Low- and Middle-Income Countries? A Systematic Review.” BMC Women’s Health 19 (1): 34. Walker, Susan P., Susan M. Chang, Amika S. Wright, Rodrigo Pinto, James J. Heckman, and Sally M. Grantham-McGregor. 2022. “Cognitive, Psychosocial, and Behaviour Gains at Age 31 Years from the Jamaica Early Childhood Stimulation Trial.” Journal of Child Psychology and Psychiatry 63 (6): 626–35. WHO (World Health Organization). 2021. Mental Health Atlas 2020. WHO. Wolf, Sharon, J. Lawrence Aber, Jere Richard Behrman, and Morgan Peele. 2019. “Longitudinal Causal Impacts of Preschool Teacher Training on Ghanaian Children’s School Readiness: Evidence for Persistence and Fade-Out.” Developmental Science 22 (5): e12878. Yousafzai, Aisha K., Muneera A. Rasheed, Arjumand Rizvi, Robert Armstrong, and Zulfiqar A. Bhutta. 2014. “Effect of Integrated Responsive Stimulation and Nutrition Interventions in the Lady Health Worker Programme in Pakistan on Child Development, Growth, and Health Outcomes: A Cluster- Randomised Factorial Effectiveness Trial.” Lancet 384 (9950): 1282–93. Chapter 3 Human Capital Accumulation in Neighborhoods Andres Yi Chang, Patrick Hoang-Vu Eozenou, and Ildo Lautharte Summary Human capital accumulation is influenced by where people are born, grow up, and live. Neighborhoods provide access to schools, health care, the community, safe streets, a clean environment, and job opportunities—all shaping human capital. A good neighborhood can help an individual thrive; a bad neighborhood can restrict and entrap. This chapter examines the effect of the neighborhoods on human capital. Families at similar incomes may vary in the opportunities they offer to their children depending on where they live. Even at identical household incomes, children who grow up in a wealthier neighborhood tend to earn significantly more as adults than individuals who grow up in a poorer neighborhood. While improving schools and health clinics in a struggling neighborhood is necessary, it may not be sufficient if other problems, such as violence and pollution, are undermining human development. The challenges are often interconnected, requiring coordinated solutions across an entire neighborhood. Addressing the challenges may require the targeting of struggling neighborhoods to improve service quality, address environmental hazards, and strengthen social capital. A reproducibility package is available for this book in the Reproducible Research Repository at https://reproducibility.worldbank.org/catalog/461. 52 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces The neighborhood matters Imagine two groups of children born on the same day and possessing similar individual and family characteristics. One group is raised in a neighborhood where the opportunities to build human capital are plentiful, and the other group is raised in a neighborhood where the opportunities are scarce. Children in the first group will grow up playing in safe spaces, breathing clean air, and attending good local schools. They will be treated by well-trained doctors and benefit from strong social networks that will influence their life choices and the likelihood they will obtain good first jobs. Children in the second group, meanwhile, will grow up in an area with pollution and poor infrastructure and will attend schools with poorly performing teachers. The neighborhood may be controlled by a gang. As these children grow up, they might have limited job opportunities and fewer resources to invest in themselves and their families. By the time the two groups of children reach adulthood, their health, education, skills, job prospects, and life trajectories will appear quite different. This chapter discusses why. What’s so special about neighborhoods? They provide access to schools and health facilities. They are characterized by distinct environments (refer to box 3.1). Local streets may not be safe, or the air and water may not be clean. The job opportunities and social dynamics may differ, ranging from cohesion to exposure to crime. BOX 3.1 Defining neighborhoods as a geographic concept In this chapter, neighborhood refers to a physical space that is defined geographically rather than based on the identity of a community. Neighborhoods and villages are microareas within larger administrative units, such as municipalities or districts. The practical definition of neighborhood often depends on the availability of the data or topic relevance. In the study of labor markets, researchers may use commuting zones; in education, school districts; and in health care, clinic service areas. Unlike definitions based on shared identity, culture, or social network, the geographic framework focuses on physical proximity. People living in the same neighborhood share a common geographic environment (the same local services, infrastructure, environmental conditions, and job opportunities), regardless of whether they identify as part of the same community. As the ability of people to move freely increases, their concept of a neighborhood may expand. While the neighborhood of a child might be a couple of blocks around the child’s home and school, it expands later when the child reaches adulthood and attends college or commutes to work. Throughout this chapter, neighborhoods and villages refer to the places where people live, interact, and influence one another’s behaviors and decisions. Human Capital Accumulation in Neighborhoods 53 FIGURE 3.1 Neighborhoods affect human capital through distinct channels Access to and quality of education and health facilities Environmental factors Social dynamics Local economic conditions Source: Original figure for this publication. Consider four factors (refer to figure 3.1). First, access to services: in some neighborhoods, children may be obliged to walk several kilometers to reach an overcrowded school with few trained teachers, and the only health clinic often lacks medical staff, vaccines, or medications. Second, environmental conditions: in flood-prone areas, repeated exposure to contaminated water may lead to chronic illness and stunting among young children. Third, social dynamics: in neighborhoods with high crime and weak community attachments, parents may be discouraged from allowing their children to play outside or walk to school because of persistent conflict between gangs and police. Fourth, job opportunities: in poor urban neighborhoods, only informal low-wage jobs are available, imposing a limit on household resources and children’s exposure to professions where they may continue learning. Neighborhoods shape human capital outcomes beyond individual and household characteristics.1 Recent evidence shows that up to half of the variation in earnings, educational attainment, and employment among individuals may be explained by the neighborhood where people grew up.2 Moving children from disadvantaged to advantaged neighborhoods, regardless of their family or personal characteristics, can have a long-term effect on their life chances. Figure 3.2 uses data from Brazil to compare children in households with identical incomes who grew up in neighborhoods where most people were poor (the points on the right in each panel) or who grew up in neighborhoods where few people were poor (the points on the left in each panel). The children who grew up in wealthier neighborhoods complete 2.3 more years of schooling, are 25 percentage points more likely to have a formal job, 31 percentage points more likely by age 29 to earn 54 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces more than their parents, and earn twice as much, on average, as the individuals who grew up in poorer neighborhoods. These differences are large, and the bulk of the differences appear to be causal effects of neighborhoods, rather than the sorting of particular kinds of households into better or worse neighborhoods.3 FIGURE 3.2 Neighborhood poverty in childhood and outcomes in adulthood, Brazil a. Years of schooling b. Probability of formal employment Number of years Probability (%) 11.7 12.0 11.5 11.0 10.5 10.0 9.5 9.0 9.4 0 25 50 75 100 Share of low-income parents in childhood neighborhood (%) 80.5 90 80 70 60 50 0 55.1 25 50 75 100 Share of low-income parents in childhood neighborhood (%) c. Probability of earning more than parent d. Income at ages 25–29 Probability (%) Income (Brazilian reais) 60.6 65 60 55 50 45 40 35 30 25 R$26,500 R$30,000 R$25,000 R$20,000 R$15,000 R$10,000 29.9 R$13,500 0 25 50 75 100 0 25 50 75 100 Share of low-income parents in childhood neighborhood (%) Share of low-income parents in childhood neighborhood (%) Source: Original figure for this publication, based on Britto et al. 2025. Note: The figure shows the relationship between the share of low-income parents in the neighborhood during childhood and the following measures of average adulthood outcomes of children from low-income families growing up in these neighborhoods: years of schooling, probability of formal employment, probability of earning more than parent, and income at ages 25–29. The scatterplots use neighborhoods as the observation unit, divided into 10 equal groups based on the percentage of low-income parents in each. Low income is defined as income at or below the 33rd percentile of the national income distribution. The share of low-income parents in the childhood neighborhood is used as a proxy for neighborhood characteristics growing up. Human Capital Accumulation in Neighborhoods 55 Why neighborhoods matter Neighborhoods affect human capital accumulation through numerous channels. From early childhood to adulthood, and from child mortality to lifetime income, where people live matters because it influences access to services and local markets and the amount of exposure to environmental hazards and social dynamics. Neighborhoods shape access and quality in education and health care Most people live near school and health care facilities. According to recent estimates based on the geolocation of road networks, at least 82 percent of the world population lives no more than 30 minutes away by motorized vehicle from a health facility.4 The share is similar for primary schools. Household surveys and patient interviews reinforce this conclusion. Results indicate a median travel time of around 18 minutes to a school and 34 minutes to a location offering medical treatment (refer to figure 3.3). Recent studies in various parts of India find that most of the rural and urban population use health services that are within two to four kilometers of their homes, and, in Pakistan, 92 percent of children attend schools that are within a 15-minute walk.5 Overall, access to basic social services does not appear to be the main local impediment to human capital accumulation. Steady gains in access to schools and health facilities have not been matched by improvements in service quality, and the quality of services depends largely on the place where people live. For instance, figure 3.4, panel a, illustrates the average quality of health facilities across 817 rural villages in 19 of India’s most populous states (excluding New Delhi), grouped into 10 quality bins. It shows that fewer than 50 percent of the health providers in the lowest-quality facilities could correctly diagnose tuberculosis, compared with 90 percent in the top 10 percent of villages. Figure 3.4, panel b, is based on comparisons of school quality across 112 rural villages in Punjab, Pakistan. Villages are grouped into 10 quality bins. Moving a child from a village in the bottom 10 percent in school quality to one in the top 10 percent would raise the mean school quality by 0.65 standard deviations, or the learning equivalent of 44 percent of a school year. 56 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces FIGURE 3.3 Households use local schools and local health facilities a. Travel time to schools b. Travel time to health facilities China Nigeria Kenya Congo, Dem. Rep. Nepal Peru Guinea-Bissau Moldova Bhutan India Malawi Ethiopia Viet Nam Average one-way travel time (minutes) Average one-way travel time (minutes) Low income Lower middle income Upper middle income High income Sources: Original figure for this publication, based on Bautista-Hernández 2023; CBS 2011; Ding and Feng 2022; Fink et al. 2022; Malone and Rudner 2011; NBS 2022; 2019 data of SAEB (Basic Education Assessment System, Brazil) (dashboard), National Institute of Educational Studies and Research, https://data-basis.org/dataset/e083c9a2 -1cee-4342-bedc -535cbad6f3cd?table=d429a79a-eca1-461c-9c1f-ce65d61048a1; SDI (Service Delivery Indicators) (dashboard), World Bank, https://www.worldbank.org/en/programs/service-delivery-indicators; 2006 data of Young Lives Study (dashboard), Oxford Department of International Development, University of Oxford, https://www.younglives.org.uk/. Note: The data are derived from nationally representative surveys and smaller-scale studies. School coverage varies by grade and depends on data availability. Health care coverage estimates are based on patient exit interviews, for example, patients seeking family planning services in Kenya, parents of children under age 5 in Malawi, and adult patients in Nigeria. Median: 18 minutesNepalMexicoTanzaniaCongo,Dem. Rep.BrazilEthiopiaBurkina FasoChinaViet NamIndiaPeruSouth AfricaJapanGhanaAustralia0153045604332302524201918171717161510015304560Median:34 minutes605953494635343430252219118 Human Capital Accumulation in Neighborhoods 57 FIGURE 3.4 School and health care quality varies substantially across villages a. Quality of health facilities, India b. Quality of schools, Pakistan Measure of provider competence Measure of school value added On average, a doctor in the bottom 10% of villages is 46 percentage points less likely to correctly diagnose TB than one in the top 10% of villages. On average, a child in the bottom 10% of villages learns 44% less per year than a child in the top 10% of villages. Bottom 10% Villages, by quality bin Top 10% Bottom 10% Villages, by quality bin Top 10% Sources: Andrabi et al. 2025; Das et al. 2022. Note: Each panel shows the 20 percent–80 percent range (vertical line) and mean (dot) of health care provider competence and school quality across villages in India and Pakistan, grouped into 10 equal bins based on quality. Health care provider competence is computed using item response theory on the ability of a provider to diagnose and treat tuberculosis, preeclampsia, diarrhea, and dysentery. School quality is measured according to school value added. The sample in panel a includes 817 rural villages in the 19 most populous states in India (excluding New Delhi) that are representative of more than 90 percent of India’s rural population based on the 1991 census. The sample in panel b includes 112 rural villages in three districts in Punjab, Pakistan, that are representative of more than 60 percent of the provincial population by the time of the survey in 2003. Bins are constructed using village averages. TB = tuberculosis. Neighborhoods determine job opportunities in a local economy How much does the place where people are born, grow up, and live matter in productivity and life income? In the United States, half the variation in mean earnings across commuting zones is attributable to place effects.6 People cannot accumulate human capital at work if they do not work, and higher incomes allow parents to make more investments in children. So, differences in employment and earnings translate into large differences in future human capital.7 In addition, local unemployment increases the proportion of single-parent households, harms mental health, raises the likelihood of substance abuse, and increases crime, all factors that result in less human capital accumulation.8 Neighborhoods determine the exposure to environmental conditions There are large differences across neighborhoods in air quality, access to clean water and sanitation, and waste disposal. Individuals may take protective measures, such as using masks or air filters, but local environmental conditions 12345678910–0.2–0.400.20.4123456789100–0.2–0.40.20.4 58 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces and community behavior may matter more than individual action. An individual is less likely to become ill if others nearby are not ill. Poor water, sanitation, and hygiene infrastructure contributes to the spread of waterborne diseases, increasing childhood illness and mortality.9 Figure 3.5 illustrates the health benefits of sanitation in India. It shows that reductions in diarrhea in households with improved sanitation are quite modest unless others in the village also have access to improved sanitation. Similar patterns have been reported in Indonesia, Mali, and Tanzania, where overall child health gains occur only if village sanitation coverage reaches 50 percent.10 The benefits are not limited to health. In Indonesia, for example, children living in open defecation–free communities during their first two years of life are more than 10 percentage points less likely to be stunted, and they exhibit higher cognitive test scores than children in communities where all other households defecate in the open.11 FIGURE 3.5 Village sanitation coverage is key to controlling waterborne disease in India Local externalities start at 30% coverage. 23% There is some benefit if a household gains access to improved sanitation within a same village. There is a much greater benefit if a household moves to a village with full coverage of improved sanitation. 77% Household has unimproved sanitation. Household has improved sanitation. 0 Share of the village with access to improved sanitation (%) Source: Original figure for this publication, based on Andrés et al. 2017. Note: The figure plots the predicted likelihood of diarrhea prevalence across different shares of village coverage. The vertical dashed line at 30 percent marks the threshold at which local externalities begin to emerge. The vertical dashed line at 50 percent is a reference line for the decomposition example of a household in a village at 50 percent improved sanitation and a household that moves from such a village to a village at 100 percent improved sanitation. The arrows and highlighted percentage labels are associated with these examples. 4268101214Predicted likelihood of diarrhea (%)255075100 Human Capital Accumulation in Neighborhoods 59 Exposure to pollution leads to worsening health outcomes, impairments in cognitive development, declines in long-term educational attainment, and rises in the probability of poverty among adults who, as children, have been so exposed.12 In Mexico, for instance, exposure to lead emanating from industrial activities restricts cognitive development and school performance among children residing close to the factories emitting the toxins.13 Children attending a school 1 kilometer farther from areas polluted with toxic waste in Chile exhibit significantly higher mathematics and language scores than children living closer.14 The lifetime cost associated with each affected child is estimated at more than US$60,000. In India, children living near a coal plant are 0.1 standard deviations shorter than unexposed children, and, the closer they live to the coal plant, the larger is the observed effect.15 Data linked to more than 2,000 industrial mines in 26 African countries show that, after a mine opens in a river watershed, there is a 25 percent rise in child mortality rates in villages downriver relative to villages upriver.16 Neighborhood effects on human capital: Social dynamics and interaction Neighborhoods are a basis of social dynamics. Decisions to focus on studying, stay in school, seek medical care, find a good job, or commit a crime are influenced by interactions with friends, peers, classmates, coworkers, and professional contacts. Often, these interactions occur in neighborhoods. In Chile, potential college applicants who are eligible for a student loan are significantly more likely to attend and complete university if their closest neighbors were also eligible for a student loan and enrolled in university.17 In India, exposure to women’s leadership in village councils influences the career aspirations and educational attainment of adolescent girls.18 In similar fashion, a local environment where there is widespread violence can depress human capital. This is evident in historical data on San Salvador, El Salvador, where gangs restricted mobility, limited the access of residents to better job opportunities, and substantially reduced secondary-school graduation rates (refer to map 3.1 and figure 3.6). Similarly, in Peru, children who grew up in villages where coca was cultivated, typically to produce cocaine, and who were exposed at an early age to illegal labor markets were around 26 percent more likely to drop out of school to participate in illegal farming.19 In sum, neighborhoods not only shape the quality of schools and health services and the access of individuals to them, but also expose individuals to localized conditions, such as environmental factors, infrastructure quality, safety, and social cohesion, that have a substantial impact on human capital accumulation. 60 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces MAP 3.1 Gang control, by neighborhood, San Salvador, El Salvador FIGURE 3.6 Probability of secondary-school graduation, by gang territory, San Salvador, El Salvador Probability of graduation after 10 years of gang control (%) 60 50 40 30 Natural barriers Source: Melnikov et al. 2025. Note: Neighborhoods under the control of two of El Salvador’s most prominent gangs, Mara Salvatrucha (MS-13) and Barrio 18 (the 18th Street Gang), are highlighted. A major government crackdown launched in 2022 drastically reduced crime, but dramatically increased the incarceration rate in El Salvador. –400 –200 0 200 400 Distance to nearest gang territory boundary (meters) Neighborhood inside gang territory Neighborhood outside gang territory Source: Melnikov et al. 2025. Note: Negative distance values indicate individuals living inside gang territory. Positive values indicate individuals living outside gang territory. Policy implications This chapter highlights that neighborhoods are important in human capital not only because of the schools and health care they provide, but also because of the community, safe streets, clean environment, and job opportunities they offer. While improving schools and clinics in struggling neighborhoods is necessary, it may not be sufficient if other problems, such as violence and pollution, are holding back human development. Some of these challenges are interconnected and require coordinated solutions across the entire neighborhood as a unit. What can be done? Policy should provide resources and incentives to enhance the quality of services, the environment, and the social capital in struggling neighborhoods. From a human capital perspective, this requires three concrete approaches (refer to table 3.1). 1,000 metersNeighborhoodsunder gang controlMS-1318th Street Human Capital Accumulation in Neighborhoods 61 TABLE 3.1 Human capital policies for struggling neighborhoods Policy approach Initiatives Improving service • Financial incentives to encourage local governments to improve outcomes in quality low-performing facilities Improving environmental • Investments in and support for qualified staff to serve in remote communities (for example, community health workers) • Joint program design and targeting across sectors that focus on human capital improvement as a shared priority conditions that affect • Programs that reduce air and water pollution and improve waste collection human capital • Programs that reduce crime and violence in neighborhoods Improving social capital • Programs that foster positive social interactions • Mobilization of local actors to ensure relevance, accountability, and collective action • Reductions in neighborhood violence Source: Original table for this publication. Improving service quality: Policies implemented in the same way everywhere can still lead to substantial inequality in the access to good schools and health care. Some policies must be spatially targeted to allocate resources and services to fill gaps. Improving environmental conditions that affect human capital: Struggling neighborhoods often lag in multiple dimensions, including air pollution, contamination associated with inadequate waste disposal, and inadequate access to clean water and sanitation. These shortcomings may depress human capital accumulation. Moreover, policies to improve environmental conditions are frequently associated with positive local externalities. This underscores the importance of acting at the neighborhood level. Improving social capital: Neighborhood social dynamics can be used to enhance policy effectiveness. Social dynamics—norms, peer effects, and exposure to role models—shape the incentives for individuals to invest in education, skills, health, and work, thereby directly influencing human capital formation. They are also a key factor if collective action is needed to achieve human development goals. Making progress in human capital, whether by improving service quality, environmental conditions, or social capital, requires policy makers to think about neighborhoods in a comprehensive manner. Creating incentives for local governments to improve lagging schools and clinics Incentivizing and equipping local governments to improve education and health outcomes in neighborhoods served by underperforming facilities may be an 62 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces effective policy. Such an approach was successfully implemented in the state of Ceará, Brazil, in the early 2000s. The state government focused on lagging areas and relied on a results-based financing system as part of a broader national education reform. Anchored in regular student assessments and performance- based funding, the system motivated municipalities to prioritize advances in learning. Stakeholders in Ceará aligned the mechanism with an equity agenda. Thus, schools in disadvantaged neighborhoods received additional technical support and were rewarded if they made progress. Low-performing school principals were paired with peers from more well performing schools in similar socioeconomic settings. Local education teams also received technical assistance tailored to the needs of their schools and communities. By concentrating on enhancing the quality of services in these underperforming areas, the authorities in Ceará reduced spatial disparities.20 Beyond promoting equity, targeting lagging neighborhoods may also be more efficient. Consider two hypothetical villages. Village A has one of the best health facilities in the country, and its residents are generally healthy. Village B has one of the worst health facilities in the country, and its residents experience poor health outcomes. While investing additional resources in village A might lead to marginal improvements, the potential for impact is limited because residents are already receiving adequate and timely care from qualified professionals and have access to vaccines and essential medicines. In village B, however, even modest investments would probably lead to substantial health gains. Providing basic services such as vaccinations or access to a doctor can dramatically reduce preventable illnesses and improve early treatment before health problems become severe. Raising the standard of care in village B may also be more feasible. Policy makers may replicate initiatives that have already produced good results in places such as village A. Targeting underperforming neighborhood schools or health centers, therefore, not only enhances equity, but also raises the likelihood of propelling these communities closer to their production possibility frontier, that is, their full potential. Investing in and supporting staff to serve in remote communities Lagging neighborhoods are often in remote rural villages. Targeting these areas through sustained investment in and support for professional staff can help deliver better services. In health care, training and deploying community health workers have been shown to be effective in enhancing health outcomes in remote villages, including by reducing malaria, asthma, and infant and child mortality and by improving maternal health, breastfeeding rates, and child nutrition.21 In Sub- Saharan Africa, community health programs delivering curative treatments for malaria, diarrhea, and pneumonia have achieved significant reductions in child mortality, especially if they are combined with other preventive measures.22 Human Capital Accumulation in Neighborhoods 63 In education, financial incentives and behavioral strategies targeted at teachers have been effective in expanding recruitment and reducing turnover in schools facing difficulties in attracting staff, while also increasing teacher quality.23 Programs offering nonfinancial support to teachers in remote areas have also yielded positive outcomes. For example, the standardization of pedagogy and school management have produced positive results on learning in Kenya and elsewhere.24 Combining standardized lesson plans with paraprofessional teachers delivering supplementary after-school classes and frequent monitoring has also generated substantial learning gains among children in rural villages in The Gambia and India.25 Together, these examples show that well-designed community health worker and teacher policies can play a critical role in improving education and health in underserved and remote communities. Mobilizing local actors to ensure relevance, accountability, and action Policy makers can enhance the impact of health and education services by working closely with key local actors, such as community members, parents, school committees, and health user groups, to monitor services and hold providers accountable. If key actors within a neighborhood are informed and organized, they are well positioned to spot problems and advocate for better results.26 For instance, providing local actors and communities in Uganda with information about the poor quality of health services and facilitating meetings between them and providers reduced provider absenteeism, increased utilization, and improved health outcomes.27 Similarly, providing school and child test scores to communities in Pakistan led parents to choose better schools, and it led schools to strive to be more competitive and efficient, resulting in higher test scores, lower private school fees, and greater enrollment.28 In some cases, however, programs that seek to mobilize local actors and empower communities fail to meet their objectives because of design flaws, elite capture, limited capacity, or unintended social consequences. For instance, in India, village education committees that sought to improve school quality through oversight had no impact on community involvement, teacher effort, or learning outcomes because the process was dominated by elites, while other committee members had limited authority and capacity.29 Such examples show that neighborhoods are not merely the places where people live. They may also be powerful platforms for improving public service delivery. Accounting for local externalities and general equilibrium effects Some interventions are more effective if they are adopted widely in a community, such as vaccinations, deworming, or removing stagnant water to prevent malaria.30 This is because they benefit not only those people who are directly targeted, but also individuals in close proximity who benefit from the lower rate of illness and disease. In practice, this means that these types of interventions need to target the neighborhood as a whole. 64 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces Consider again the case of India illustrated in figure 3.5. Imagine a policy that aims to reduce child diarrhea through improved sanitation. The effect of the policy on a particular household would be a 12 percent reduction in the incidence of diarrhea if no one else in the village has improved sanitation, compared with a 32 percent reduction in a village that has reached full coverage. This shows the large difference in policy effectiveness based on community-wide adoption. Increased benefits from wider coverage can also encourage individual households to adopt improved sanitation, creating a positive reinforcement loop.31 Programs with strong neighborhood effects should prioritize achieving high coverage within selected communities to maximize impact and efficiency. Unlike traditional models that favor spreading investments thinly, the concentration of resources in fewer areas can yield greater returns because collective uptake enhances outcomes. Policies that account for broader market responses within a neighborhood or village—that is, local general equilibrium effects—can also substantially amplify the overall impact. For instance, in Pakistan, providing grants to all public schools in a village spurred competition, prompting private schools to improve quality and raising student test scores across both types of schools, making the program 85 percent more cost-effective than another program variant that provided grants to a single school in a village.32 Similarly, in rural villages in Mexico, an in-kind transfer program that supplied basic food items to poor households drove down the prices of transferred food items by 4 percent by increasing the availability of food items in the market. This generated an extra indirect transfer to consumers of about 14 percent of the direct transfer.33 As this demonstrates, policies can leverage local market responses to increase their impact.34 Making human capital gains a shared priority in program design and targeting Maximizing the impact of human development policies requires deliberate coordination across sectors. Such coordination is often most effective if carried out within neighborhoods. If actors in local education, health care, and social protection coordinate with infrastructure, transport, and urban planning teams, interventions can be more effective, equitable, and sustainable. For instance, designing transport systems and policies to promote safety within the context of human development goals—such as ensuring safe, affordable, and reliable connections to schools, health clinics, and workplaces—can remove critical access barriers. In Chile and Mexico, investments in transportation infrastructure allowed some families to send their children to better schools.35 In Kenya, better roads have been linked to improved health service delivery because they have helped attract more highly qualified staff and eased supply logistics.36 Similarly, in El Salvador, the Safe Schools Program reduced dropouts and gang recruitment by placing local police officers near schools during arrival and dismissal times.37 These examples underscore the importance of integrated neighborhood planning to help ensure that Human Capital Accumulation in Neighborhoods 65 local actors and agencies across sectors work together to improve human capital outcomes. The most effective way to support human capital accumulation may sometimes involve prioritizing a sector that is not traditionally the focus of social policy. A compelling example emerged during consultations for this report with education authorities in Rio de Janeiro. In Brazil, all graduating secondary-school students are required to take a national college entrance examination, which is administered on designated days. In the favelas of Rio de Janeiro, however, police operations targeting criminal groups can sometimes coincide with examination dates, creating serious safety risks that prevent students from reaching test centers and potentially derailing their chances of gaining access to higher education. Only after education authorities had coordinated with law enforcement to avoid operations during examination days could students safely sit for the tests, demonstrating how aligning security policy with human development goals can sometimes remove critical barriers to educational opportunity. Delivering programs that foster positive social interactions Providing opportunities for positive neighborhood interactions has been beneficial among some at-risk populations, such as youth or adults vulnerable to violent behavior. Various school-based group counseling, mentoring, and cognitive behavioral therapy programs have resulted in a reduction in violence and antisocial behaviors and improvements in educational outcomes in violent contexts, including in Chicago, Liberia, and San Salvador.38 These programs are relatively low in cost, but they have all been delivered by local community members, such as school tutors, mentors, and the staff of local nongovernmental organizations, who were in a unique position to connect and engage with youth at risk of violent behavior. Similarly, a successful program in Türkiye among academically strong but socially disruptive adolescents reduced their antisocial behavior and assisted them in gaining admission to more highly selective secondary schools, thereby improving schools and communities.39 Conclusion: Putting it all together The place where people live—their neighborhood or village—plays a crucial role in shaping human capital outcomes and should be central to policy design. Beyond individual or household traits, neighborhood conditions influence access to quality education and health care, exposure to environmental risks, the availability of jobs, and social dynamics. In many cases, neighborhood characteristics explain a large share of the differences in earnings, educational attainment, and health outcomes observed across communities. Recognizing the neighborhood as a central space for human capital development has important implications for policy. In some cases, improvement in outcomes requires that the entire neighborhood is targeted, not just individual families or 66 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces service facilities. For instance, interventions focusing on sanitation, public safety, or pollution control are only effective if they are implemented broadly, with collective participation. In other cases, certain neighborhoods may need to be prioritized because they are unable to deliver basic quality services on their own. Even well- designed national policies can lead to large disparities in school and health care quality if local conditions are ignored. A neighborhood-focused approach encourages coordination across traditionally separate sectors. Only if they work together can sectors adequately address the complex and interrelated barriers that limit human capital accumulation. Notes 1. Chyn and Katz (2021); Sharkey (2016). 2. Britto et al. (2025); Chetty and Hendren (2018a), (2018b); Chetty et al. (2026). 3. The evidence that neighborhoods causally affect human capital outcomes is provided by studies that isolate place effects (that is, the benefits deriving from better services, environments, and networks) from sorting effects (that is, households self-selecting into neighborhoods). Researchers rely on three main strategies, as follows: (a) rich administrative data to compare children—often siblings—who move at different ages; (b) randomized relocation through programs such as housing lotteries (for example, the Moving to Opportunity for Fair Housing randomized social experiment in the United States); and (c) forced relocation associated with events, such as natural disasters or housing demolitions, which offer quasi-experimental variation. 4. Weiss et al. (2020). However, not everyone has access to a motor vehicle. Only 57 percent of the world population is within a 60-minute walk of a health facility. The shares are smaller in low-income countries and among households at lower socioeconomic status within each country. Prioritizing access is therefore still relevant in many places. 5. Andrabi et al. (2017); Das and Hammer (2007); Das and Mohpal (2016). 6. Card et al. (2025). 7. Moretti (2024). 8. Autor et al. (2019); Diette et al. (2018); Pierce and Schott (2020). 9. Lepault (2023). 10. Cameron et al. (2022). 11. Cameron et al. (2021). 12. Aizer and Currie (2019); Clay et al. (2025); Heissel et al. (2022); Persico (2024); Persico et al. (2020). 13. Tanaka et al. (2022). 14. Rau et al. (2015). 15. Vyas (2025). 16. Gittard and Hu (2024). 17. Barrios-Fernández (2022). 18. Beaman et al. (2012). 19. Sviatschi (2022). 20. Loureiro et al. (2020). 21. Nkonki et al. (2017). 22. Christopher et al. (2011). 23. Ajzenman et al. (2026); Evans and Mendez Acosta (2023). Human Capital Accumulation in Neighborhoods 67 24. Gray-Lobe et al. (2022). 25. Eble et al. (2021); Lakshminarayana et al. (2013). 26. World Bank (2003). 27. Björkman-Nyqvist and Svensson (2010). 28. Andrabi et al. (2017). 29. Banerjee et al. (2010). 30. Ahuja et al. (2017). 31. Deutschmann et al. (2024). 32. Andrabi et al. (2024). 33. Cunha et al. (2019). Caution is warranted because the effects had the opposite implication for food- producing households in the recipient village. Overall welfare effects may thus depend on the composition of the local economy. 34. Local market responses can sometimes also work against policy objectives. For instance, in the Dominican Republic, the expansion of public schools resulted in the closure of some private schools of relatively higher quality (Dinerstein et al. 2020). In other contexts, if a large share of households in a village receive cash transfers, increased demand for locally traded goods can drive up prices, including food prices, making essential nutrition less affordable for transfer-ineligible households. In the Philippines, this dynamic led to higher malnutrition and stunting rates among children not eligible for cash transfers (Filmer et al. 2023). 35. Dustan and Ngo (2018); Herskovic (2020). 36. Becerra Luna et al. (2022). 37. Castro et al. (2025). 38. Blattman et al. (2023); Dinarte-Díaz et al. (2024); Heller et al. (2017). 39. Alan and Kubilay (2025). References Ahuja, Amrita, Sarah Baird, Joan Hamory Hicks, Michael R. Kremer, and Edward Miguel. 2017. “Economics of Mass Deworming Programs.” In Child and Adolescent Health and Development, edited by Donald A. P. Bundy, Nilanthi de Silva, Susan Horton, Dean T. Jamison, and George C. Patton. Disease Control Priorities Series, vol. 8, 3rd ed. World Bank. Aizer, Anna, and Janet Currie. 2019. “Lead and Juvenile Delinquency: New Evidence from Linked Birth, School, and Juvenile Detention Records.” Review of Economics and Statistics 101 (4): 575–87. Ajzenman, Nicolás, Gregory Elacqua, Luana Marotta, and Anne Sofie Westh Olsen. 2026. “Order Effects and Teachers’ Labor Supply: A Nationwide RCT in Ecuador.” Preview ahead of publication, American Economic Journal: Applied Economics. https://www.aeaweb.org/articles?id=10.1257 / app.20230525&from=f. Alan, Sule, and Elif Kubilay. 2025. “Empowering Adolescents to Transform Schools: Lessons from a Behavioral Targeting.” American Economic Review 115 (2): 365–407. Andrabi, Tahir Raza Shah, Natalie Bau, Jishnu Das, Naureen Karachiwalla, and Asim Ijaz Khwaja. 2024. “Crowding in Private Quality: The Equilibrium Effects of Public Spending in Education.” Quarterly Journal of Economics 139 (4): 2525–77. Andrabi, Tahir Raza Shah, Natalie Bau, Jishnu Das, and Asim Ijaz Khwaja. 2025. “Heterogeneity in School Value Added and the Private Premium.” American Economic Review 115 (1): 147–82. Andrabi, Tahir Raza Shah, Jishnu Das, and Asim Ijaz Khwaja. 2017. “Report Cards: The Impact of Providing School and Child Test Scores on Educational Markets.” American Economic Review 107 (6): 1535–63. 68 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces Andrés, Luis Alberto, Bertha Briceño, Claire Chase, and Juan Augustin Echenique. 2017. “Sanitation and Externalities: Evidence from Early Childhood Health in Rural India.” Journal of Water, Sanitation and Hygiene for Development 7 (2): 272–89. Autor, David H., David Dorn, and Gordon H. Hanson. 2019. “When Work Disappears: Manufacturing Decline and the Falling Marriage Market Value of Young Men.” American Economic Review: Insights 1 (2): 161–78. Banerjee, Abhijit Vinayak, Rukmini Banerji, Esther Duflo, Rachel Glennerster, and Stuti Khemani. 2010. “Pitfalls of Participatory Programs: Evidence from a Randomized Evaluation in Education in India.” American Economic Journal: Economic Policy 2 (1): 1–30. Barrios-Fernández, Andrés. 2022. “Neighbors’ Effects on University Enrollment.” American Economic Journal: Applied Economics 14 (3): 30–60. Bautista-Hernández, Dorian Antonio. 2023. “Determinants and Metropolitan Patterns of School Travel Time: Is Location Associated with Average Years of Schooling? The Case of Mexico City.” Cities 139 (August): 104408. Beaman, Lori A., Esther Duflo, Rohini P. Pande, and Petia Topalova. 2012. “Female Leadership Raises Aspirations and Educational Attainment for Girls: A Policy Experiment in India.” Science 335 (6068): 582–86. Becerra Luna, Laura Natalia, Mathilde Sylvie Maria Lebrand, Nino Pkhikidze, and Andres Yi Chang. 2022. “Infrastructure Matters: Complementarities with the Quality of Health Service Delivery in Kenya.” Policy Research Working Paper 10220, World Bank. Björkman-Nyqvist, Martina, and Jakob Svensson. 2010. “When Is Community-Based Monitoring Effective? Evidence from a Randomized Experiment in Primary Health in Uganda.” Journal of the European Economic Association 8 (2–3): 571–81. Blattman, Christopher, Sebastian Chaskel, Julian C. Jamison, and Margaret A. Sheridan. 2023. “Cognitive Behavioral Therapy Reduces Crime and Violence over Ten Years: Experimental Evidence.” American Economic Review: Insights 5 (4): 527–45. Britto, Diogo G. C., Alexandre de Andrade Fonseca, Paolo Pinotti, Breno Sampaio, and Lucas Warwar. 2025. “Do CCTs Create Conditions to Thrive? Bolsa Família and Social Mobility in Brazil.” WIDER Working Paper 96/25 (December), United Nations University–World Institute for Development Economics Research. Cameron, Lisa Ann, Claire Chase, Sabrina Haque, George Joseph, Rebekah Pinto, and Qiao Wang. 2021. “Childhood Stunting and Cognitive Effects of Water and Sanitation in Indonesia.” Economics and Human Biology 40 (January): 100944. Cameron, Lisa Ann, Paul J. Gertler, Manisha Shah, María Laura Alzúa, Sebastian Martinez, and Sumeet Patil. 2022. “The Dirty Business of Eliminating Open Defecation: The Effect of Village Sanitation on Child Height from Field Experiments in Four Countries.” Journal of Development Economics 159 (November): 102990. Card, David E., Jesse Rothstein, and Moises Yi. 2025. “Location, Location, Location.” American Economic Journal: Applied Economics 17 (1): 297–336. Castro, Eleno, Felipe Coy, Carlos Schmidt-Padilla, and Maria Micaela Sviatschi. 2025. “Breaking the Gang: A Preventive Approach to Reduce Recruitment in Schools.” Working paper, September 18. https://www.micaelasviatschi.com/wp-content/uploads/2025/09/SchoolGangs-82.pdf. CBS (Central Bureau of Statistics, Nepal). 2011. Nepal Living Standards Survey 2010/11: Statistical Report. 2 vols. November. National Planning Commission, CBS. Chetty, Raj, John N. Friedman, Nathaniel Hendren, Maggie R. Jones, and Sonya R. Porter. 2026. “The Opportunity Atlas: Mapping the Childhood Roots of Social Mobility.” American Economic Review 116 (1): 1–51. Chetty, Raj, and Nathaniel Hendren. 2018a. “The Impacts of Neighborhoods on Intergenerational Mobility I: Childhood Exposure Effects.” Quarterly Journal of Economics 133 (3): 1107–62. Human Capital Accumulation in Neighborhoods 69 Chetty, Raj, and Nathaniel Hendren. 2018b. “The Impacts of Neighborhoods on Intergenerational Mobility II: County-Level Estimates.” Quarterly Journal of Economics 133 (3): 1163–1228. Christopher, Jason B., Alex Le May, Simon Lewin, and David A. Ross. 2011. “Thirty Years After Alma-Ata: A Systematic Review of the Impact of Community Health Workers Delivering Curative Interventions Against Malaria, Pneumonia and Diarrhoea on Child Mortality and Morbidity in Sub-Saharan Africa.” Human Resources for Health 9 (1): 27. Chyn, Eric, and Lawrence F. Katz. 2021. “Neighborhoods Matter: Assessing the Evidence for Place Effects.” Journal of Economic Perspectives 35 (4): 197–222. Clay, Karen, Edson R. Severnini, and Xiao Wang. 2025. “The Hidden Toll of Airborne Lead: Infant Mortality Impacts of Industrial Lead Pollution.” NBER Working Paper 33447 (August), National Bureau of Economic Research. Cunha, Jesse M., Giacomo De Giorgi, and Seema Jayachandran. 2019. “The Price Effects of Cash Versus In-Kind Transfers.” Review of Economic Studies 86 (1): 240–81. Das, Jishnu, Benjamin Daniels, Monisha Ashok, Eun-Young Shim, and Karthik Muralidharan. 2022. “Two Indias: The Structure of Primary Health Care Markets in Rural Indian Villages with Implications for Policy.” Social Science and Medicine 301 (May): 112799. Das, Jishnu, and Jeffrey S. Hammer. 2007. “Location, Location, Location: Residence, Wealth, and the Quality of Medical Care in Delhi, India.” Health Affairs 26 (Supplement 2): 338–51. Das, Jishnu, and Aakash Mohpal. 2016. “Socioeconomic Status and Quality of Care in Rural India: New Evidence from Provider and Household Surveys.” Health Affairs 35 (10): 1764–73. Deutschmann, Joshua W., Molly Lipscomb, Laura Schechter, and Jessica Zhu. 2024. “Spillovers Without Social Interactions in Urban Sanitation.” American Economic Journal: Applied Economics 16 (3): 482–515. Diette, Timothy Maurice, Arthur H. Goldsmith, Darrick Hamilton, and William Alexander Darity, Jr. 2018. “Race, Unemployment, and Mental Health in the USA: What Can We Infer About the Psychological Cost of the Great Recession Across Racial Groups?” Journal of Economics, Race, and Policy 1 (2): 75–91. Dinarte-Díaz, Lelys Ileana, Pablo Egaña del Sol, Claudia Martínez, and Cindy Jacqueline Rojas Alvarado. 2024. “When Emotion Regulation Matters: The Efficacy of Socio-Emotional Learning to Address School-Based Violence in Central America.” IZA Discussion Paper DP 16831 (February), Institute of Labor Economics. Dinerstein, Michael, Christopher Neilson, and Sebastián Otero. 2020. “The Equilibrium Effects of Public Provision in Education Markets: Evidence from a Public School Expansion Policy.” IRS Working Paper 645 (November), Industrial Relations Section, Princeton University. Ding, Pengxiang, and Suwei Feng. 2022. “How School Travel Affects Children’s Psychological Well-Being and Academic Achievement in China.” International Journal of Environmental Research and Public Health 19 (21): 13881. Dustan, Andrew, and Diana K. L. Ngo. 2018. “Commuting to Educational Opportunity? School Choice Effects of Mass Transit Expansion in Mexico City.” Economics of Education Review 63 (April): 116–33. Eble, Alex, Chris Frost, Alpha Camara, et al. 2021. “How Much Can We Remedy Very Low Learning Levels in Rural Parts of Low-Income Countries? Impact and Generalizability of a Multi-Pronged Para-Teacher Intervention from a Cluster-Randomized Trial.” Journal of Development Economics 148 (January): 102539. Evans, David K., and Amina Mendez Acosta. 2023. “How to Recruit Teachers for Hard-to-Staff Schools: A Systematic Review of Evidence from Low- and Middle-Income Countries.” Economics of Education Review 95 (August): 102430. Filmer, Deon, Jed Friedman, Eeshani Kandpal, and Junko Onishi. 2023. “Cash Transfers, Food Prices, and Nutrition Impacts on Ineligible Children.” Review of Economics and Statistics 105 (2): 327–43. 70 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces Fink, Günther, Eeshani Kandpal, and Gil Shapira. 2022. “Inequality in the Quality of Health Services: Wealth, Content of Care, and the Price of Antenatal Consultations in the Democratic Republic of Congo.” Economic Development and Cultural Change 70 (3): 1295–1336. Gittard, Mélanie, and Irène Hu. 2024. “MiningLeaks: Water Pollution and Child Mortality in Africa.” PSE Working Papers 2024–24, Paris School of Economics. Gray-Lobe, Guthrie, Anthony Keats, Michael R. Kremer, Isaac Mbiti, and Owen Ozier. 2022. “Can Education Be Standardized? Evidence from Kenya.” Working Paper 2022-68 (September), Becker Friedman Institute, University of Chicago. Heissel, Jennifer Ann, Claudia Persico, and David Simon. 2022. “Does Pollution Drive Achievement? The Effect of Traffic Pollution on Academic Performance.” Journal of Human Resources 57 (3): 747–76. Heller, Sara B., Anuj K. Shah, Jonathan Guryan, Jens Ludwig, Sendhil Mullainathan, and Harold A. Pollack. 2017. “Thinking, Fast and Slow? Some Field Experiments to Reduce Crime and Dropout in Chicago.” Quarterly Journal of Economics 132 (1): 1–54. Herskovic, Luis. 2020. “The Effect of Subway Access on School Choice.” Economics of Education Review 78 (October): 102021. Lakshminarayana, Rashmi, Alex Eble, Preetha Bhakta, et al. 2013. “The Support to Rural India’s Public Education System (Stripes) Trial: A Cluster Randomised Controlled Trial of Supplementary Teaching, Learning Material and Material Support.” PLOS ONE 8 (7): e65775. Lepault, Claire. 2023. “Is Urban Wastewater Treatment Effective in India? Evidence from Water Quality and Infant Mortality.” HAL Sciences de l’Homme et de la Société, Centre pour la Communication Scientifique Directe. https://hal.science/hal-04232407v1/file/JMP_CLepault _ Oct6_23.pdf. Loureiro, Andre, Louisee Cruz, Ildo Lautharte, and David K. Evans. 2020. “The State of Ceará in Brazil Is a Role Model for Reducing Learning Poverty.” Report 34156 (June), World Bank. https://hdl .handle .net/10986/34156. Malone, Karen, and Julie Rudner. 2011. “Global Perspectives on Children’s Independent Mobility: A Socio-Cultural Comparison and Theoretical Discussion of Children’s Lives in Four Countries in Asia and Africa.” Global Studies of Childhood 1 (3): 243–59. Melnikov, Nikita, Carlos Schmidt-Padilla, and María Micaela Sviatschi. 2025. “Gangs, Labor Mobility and Development.” Econometrica 93 (6): 2083–2121. Moretti, Enrico. 2024. “Place-Based Policies and Geographical Inequalities.” Oxford Open Economics 3 (Supplement 1): i625–33. NBS (National Bureau of Statistics, Tanzania). 2022. Tanzania National Panel Survey, Wave 5, 2020–2021. November. NBS. Nkonki, Lungiswa, Aviva Tugendhaft, and Karen J. Hofman. 2017. “A Systematic Review of Economic Evaluations of CHW Interventions Aimed at Improving Child Health Outcomes.” Human Resources for Health 15 (1): 19. Persico, Claudia. 2024. “Can Pollution Cause Poverty? The Effects of Pollution on Educational, Health and Economic Outcomes.” Working paper (September), Department of Public Administration and Policy, School of Public Affairs, American University. Persico, Claudia, David Figlio, and Jeffrey Roth. 2020. “The Developmental Consequences of Superfund Sites.” Journal of Labor Economics 38 (4): 1055–97. Pierce, Justin R., and Peter K. Schott. 2020. “Trade Liberalization and Mortality: Evidence from US Counties.” American Economic Review: Insights 2 (1): 47–64. Rau, Tomás, Sergio S. Urzúa, and Loreto Reyes. 2015. “Early Exposure to Hazardous Waste and Academic Achievement: Evidence from a Case of Environmental Negligence.” Journal of the Association of Environmental and Resource Economists 2 (4): 527–63. Sharkey, Patrick. 2016. “Neighborhoods, Cities, and Economic Mobility.” RSF 2 (2): 159–77. Sviatschi, Maria Micaela. 2022. “Making a NARCO: Childhood Exposure to Illegal Labor Markets and Criminal Life Paths.” Econometrica 90 (4): 1835–78. Human Capital Accumulation in Neighborhoods 71 Tanaka, Shinsuke, Kensuke Teshima, and Eric A. Verhoogen. 2022. “North-South Displacement Effects of Environmental Regulation: The Case of Battery Recycling.” American Economic Review: Insights 4 (3): 271–88. Vyas, Sangita. 2025. “The Child Health Impacts of Coal: Evidence from India’s Coal Expansion.” Journal of Human Resources 60 (2): 496–537. Weiss, Daniel J., Andy Nelson, C. A. Vargas-Ruiz, et al. 2020. “Global Maps of Travel Time to Healthcare Facilities.” Nature Medicine 26 (12): 1835–38. World Bank. 2003. World Development Report 2004: Making Services Work for Poor People. World Bank; Oxford University Press. Chapter 4 Human Capital Accumulation at Work Joana Silva Summary Human capital formation does not stop at school. It continues at work. Work is not only where skills are used, but also where they are built— through practice, exposure to technology, interactions with peers and managers, and training. This chapter examines how people learn at work, how much skill development occurs in the workplace, why investment in human capital in these settings is low, and what policies can help. It shows that learning at work varies systematically across job types. The same increase in experience generates only about half as much learning among the self- employed relative to wage workers, and, among wage workers, there is twice as much learning in large firms as in small firms. One challenge is that around 70 percent of workers in low- and middle-income countries are employed in jobs with limited learning potential, such as small-scale agriculture, low-quality self-employment, and microenterprises. Unlocking learning in current jobs and expanding the availability of jobs associated with substantial human capital formation are therefore essential to raising productivity and incomes. A second challenge is the misallocation of talent. About 50 percent of women are out of the labor force, and 20 percent of youth are neither working nor in education. Removing barriers to their participation in labor markets would enable more skill development at work. These challenges call for policies that expand learning on the job, ease transition to employment, and create more jobs with strong learning potential, supported by broader reforms that reduce market failures and misallocation. A reproducibility package is available for this book in the Reproducible Research Repository at https://reproducibility.worldbank.org/catalog/461. 74 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces The workplace matters People accumulate significant human capital at work through on-the-job experience, training, and advancing to jobs with stronger potential for learning (refer to table 4.1). New evidence indicates that work experience and education may be equally important in explaining income gaps between low-income, middle- income, and high-income economies.1 Learning at work is critical for countries to move up the value added ladder, sparking productivity and innovation that lead to job creation. Together with education, learning at work shapes what countries produce and how they produce it. The notion that learning at the current job occurs solely through explicit training is incorrect. Learning occurs through on-the-job practice (as one performs a task, one learns how to improve the performance), exposure to technology, the acquisition of more responsibilities, and interactions with peers and managers. This shapes both technical and soft skills, such as teamwork and decision- making. Joint tasks with coworkers foster collaboration, and the quality of workplace interactions influences knowledge diffusion and productivity.2 Building human capital varies by type of job and workplace (refer to figure 4.1). Among farmers, this means learning new agricultural techniques or improving current ones to boost yields. Farmers learn not only how much input to use, such as fertilizer or improved seeds, but also how to use more productive technologies. Among the self-employed, building human capital involves acquiring know-how and business skills that enhance the production, profitability, and sustainability of their businesses. Among wage workers, it entails acquiring experience that can improve performance. This knowledge can be general (and therefore transferable) or firm-specific (and therefore lost if a worker changes firms). TABLE 4.1 How human capital accumulates at work Channels Sources of learning Be employed. if they do not have a job). Participate in the workforce (no one can learn on the job Learn at the current job. On-the-job learning: Learn through practice, technology, and peers and managers. Explicit training: Participate in training programs. Move to a new job with Advance to a job entailing more learning and greater stronger potential for productivity. learning. Source: Original table for this publication. Human Capital Accumulation at Work 75 FIGURE 4.1 How learning occurs in various contexts Farmers Farmers learn at work by experimenting with new techniques and new technologies to boost yields. Self-employed The self-employed develop skills to expand their business and increase their earnings. Wage workers Wage workers acquire (general or firm-specific) experience that can improve their performance. Source: Original figure for this publication. Learning mechanisms also vary by job. Farmers refine techniques through trial and error, peer exchanges, and extension services. The self-employed build skills through hands-on experience, business networks, and structured training. Wage workers learn by practicing advanced techniques, adopting technology, solving problems, engaging with peers and managers, and receiving formal or informal training. What limits human capital accumulation at work? There are three main challenges in building human capital at work in low- and middle-income countries: (1) low learning through jobs, (2) low labor force participation, and (3) a shortage of jobs that entail high skill development at work (that is, high human capital jobs). Low learning through jobs. On-the-job learning is often more restrained in these countries. When businesses lack resources, they may invest little or nothing in workplace learning. Market failures, such as limited access to credit or uncertain business conditions, further constrain investment. Even when resources are available, businesses may underinvest in training if the expected returns are low. One key reason is that trained workers may leave the firm before the investment is 76 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces recouped, especially in high-turnover environments, which are common in low- and middle-income countries.3 Another reason is the absence of complementary investments and assets. Skills and technology are complements, and exposure to the latter accelerates skill development at work, particularly among more educated workers.4 Weak management may also be a constraint. When management quality is low, workers learn less from supervisors, and the returns to training diminish.5 The same holds for coworkers. When peers lack human capital, opportunities to learn from them are limited. Learning from peers and managers at work also requires an organizational structure and information flows that enable such interactions (refer to table 4.2). The constraints are even greater in firm-sponsored external training. Many businesses consider training an expense rather than an investment. A lack of expertise and resources, combined with the lack of high-quality, relevant, and flexible programs, often because of coordination failures and misaligned incentives, discourages participation. In small firms, where every worker matters, releasing staff for training is especially challenging. Even if resources are available, which is more unlikely in economies dominated by small firms, training systems often fail to deliver the skills firms need, resulting in low returns. Low labor force participation. Many women and young people are not in education, employment, or training (NEET), limiting human capital accumulation through work. If individuals are matched with jobs for which they lack the required skills, or if high-talented individuals are not matched to roles that foster learning, aggregate human capital accumulation will be lower. According to recent estimates, reductions in the misallocation of talent, largely driven by women starting to work, accounted for 20 percent to 40 percent of US economic growth between 1960 and 2010.6 Recent analysis focusing on the effects of removing barriers to women’s entrepreneurship in India shows that the labor force participation costs among women are about twice what they are among men. TABLE 4.2 Opportunities for human capital accumulation, by job type Job type Learning through practice Technology exposure Learning from peers and managers Explicit training Out of the labor force None (no job, no experience to apply) None None None On-farm employment High potential, but current setup Low technology use Limited peer interaction Dependent on limits scope extension services Self-employment Limited tasks and innovation; requires Mostly low-skill Fewer or no peers from Dependent on business expansion for growth manual jobs whom to learn access to business network Wage employment More diverse tasks; career Varies; more intensive Dependent on manager Dependent on progression opportunities in large firms quality, internal organization, profitability workforce composition Source: Original table for this publication. Human Capital Accumulation at Work 77 Removing these barriers would double female labor force participation in India and increase real GDP by 43 percent.7 A shortage of jobs that entail high skill development at work. A high human capital job is one in which substantial skill acquisition takes place. These jobs are typically wage jobs in the modern sector. As the subsequent section will show, the problem is that most workers in low- and middle-income countries are not working in these jobs. In what follows, we examine what is constraining learning at work in low- and middle-income countries. Employment is concentrated in jobs in which less learning occurs Where do people work? In low- and middle-income countries, around 70 percent of workers are employed in small-scale agriculture, low-quality self-employment, or microfirms (fewer than five workers). These jobs typically offer limited opportunity for on-the-job learning. The pattern is even more pronounced in low- income countries, where these types of jobs employ about 80 percent of workers, while only 20 percent work in positions that support learning on the job. In high-income countries, the pattern is nearly reversed: approximately 80 percent of workers hold high-learning jobs, while only 20 percent are in jobs that offer little opportunity for learning at work (refer to figure 4.2, panel a). In terms of occupations, in low- and lower-middle-income countries, 71 percent of all employed people work in agricultural, domestic, or manual occupations. Only 16 percent are clerks or sales or services workers, and 13 percent are technicians, professionals, or managers. In contrast, in high-income countries, the shares are reversed, with the latter two groups making up 71 percent of total employment. Self-employment is prevalent in low- and middle-income countries (79 percent and 53 percent, respectively) and significantly higher than in high-income countries (12 percent).8 Within middle-income countries there are large differences. In lower- middle-income countries, 67 percent of workers are self-employed, but this share drops to 41 percent in upper-middle-income countries (refer to figure 4.2, panel b). Most self-employment in low- and middle-income countries takes place in low-skill occupations, such as farmers working on small plots or street vendors selling phone chargers, rather than high-skill occupations, such as doctors in private medical practice. In low- and middle-income countries, more than 90 percent of self- employed workers are in low-skill employment (many in subsistence farming), while fewer than 10 percent are in high-skill employment (mostly self-employed professionals). Meanwhile, the share of the self-employed working in professional or other high-skill occupations in low- and middle-income countries is only about one-quarter that observed in high-income countries (refer to figure 4.2, panel c). 78 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces FIGURE 4.2 Most people are in jobs that offer little opportunity for learning at work b. Share of workers, by self-employment or wage job (%) HIC LIC MIC 98 92 92 90 2 8 8 10 HIC 59 41 HIC LIC MIC HIC Source: Original figure for this publication, based on labor statistics of ILOSTAT (dashboard), International Labour Organization, https://ilostat.ilo.org/. Note: Percentages refer to share of the employed population. Low-learning jobs include small-scale agriculture, low-quality self- employment, or microfirms. Low-skill occupations include clerical support workers; service and sales workers; skilled agricultural, forestry, and fishery workers; craft and related trade workers; plant and machine operators and assemblers; and elementary occupations. High-skill occupations include managers, professionals, technicians, and associate professionals. Microfirms are firms with up to five workers. Small and medium firms are firms with up to 50 workers. Large firms are firms with 50 or more workers. HIC = high-income country; LIC = low-income country; LMIC = lower-middle-income country; MIC = middle-income country; UMIC = upper-middle-income country. MIC = LMIC + UMIC. For data by region and country, refer to the interactive figures online at https://humancapital.worldbank.org/en/building-human-capital-where-it-matters. Evidence on Latin America shows that there is a marked decline in the likelihood of self-employment along the income distribution. The probability of self-employment is three times greater in the bottom (poorest) quintile of the income distribution than in the top quintile, whereas, in the United States, self-employment is slightly U-shaped over income, which means that both the lowest and highest earners are more likely to be self-employed.9 Most wage workers in low-income countries are employed in microfirms, an average of 78 percent, and, in middle-income countries, 57 percent (refer to figure 4.2, panel d). In sharp contrast, the share is only 21 percent in high-income countries. This reflects the underlying firm-size distribution. A disproportionately large share of firms in low- and middle-income countries are small, even after accounting for the differences in production structure across industries.10 In Africa, small and informal firms, including self-employed workers and subsistence farmers, account for about a. Share of workers, by low orhigh learning job (%)19677981517149293321UMICLMICMICLICLow learning jobs High learning jobsSelf-employedWage workers12537988472141675933UMICLMICMICLICc. Share of self-employed workers,by skill level of occupation (%)d. Share of wage workers,by firm size (%)MicrofirmsSmall and medium firmsLarge firms2178341746UMICLMIC5064262024174572320Low skillHigh skillUMICLMIC Human Capital Accumulation at Work 79 86 percent of all jobs and roughly 50 percent of economic output.11 Meanwhile, large highly productive firms remain scarce.12 Manufacturing employment in enterprises with fewer than 10 employees exceeds 50 percent in many African countries. It reaches 80 percent in Ghana. This compares with less than 5 percent in the United States.13 On-the-job learning by job type Among the three main types of employment—farming, self-employment, and wage work—how much learning is occurring at work? Measuring learning on the job is not straightforward, as discussed in box 4.1, but much less of it happens in poorer countries than in richer countries. Learning in on-farm employment is a known driver of growth in agriculture.14 Yet, in poorer countries, a large productivity gap persists in agriculture relative to other sectors. Using plot-level data from the Living Standards Measurement Study, figure 4.3 illustrates the relationship between agricultural productivity—proxied by total factor productivity—and years of farm experience among smallholder farmers in six Sub-Saharan African countries. The evidence suggests that farmer productivity does not rise significantly with experience in these settings. BOX 4.1 Measuring learning at work How should learning on the job be estimated across types of jobs in the absence of direct measures of cognitive and noncognitive job skills? Ideally, researchers would use a measure that captures the progression of job-related skills. However, such data are not available in most countries. The literature therefore relies on indirect measures, such as improvements in productivity and yields among farmers, business practices, performance and profits (that is, output, net of input costs) among the self-employed, and wage growth among salaried workers. The underlying assumption is that, while indirect, these measures largely reflect increases in the marginal product of labor, thereby capturing the skills and knowledge gained through experience on the job. Among farmers and the self-employed, studies proxy learning by tracking how revenues evolve, accounting for the fact that the incomes of farmers and the self-employed represent payment to labor income and capital income.a Among wage workers, studies use wage growth as an indicator of increases in the marginal product of labor, capturing skills and knowledge gained through experience. While this growth may not represent human capital accumulation in a narrow sense, it does signal improvements in labor productivity. a. Gollin (2002). 80 Building Human Capital Where It Matters: Homes, Neighborhoods, and Workplaces FIGURE 4.3 Farmer productivity on small plots in Sub-Saharan Africa does not increase much with experience a. Ethiopia b. Malawi c. Mali Total farm productivity, adjusted for inputs Total farm productivity, adjusted for inputs Total farm productivity, adjusted for inputs 1.0 0.5 0 –0.5 –1.0 –1.5 1.0 0.5 0 –0.5 –1.0 –1.5 1.0 0.5 0 –0.5 –1.0 –1.5 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 Farm experience (years) Farm experience (years) Farm experience (years) d. Niger e. Nigeria f. Tanzania Total farm productivity, adjusted for inputs Total farm productivity, adjusted for inputs Total farm productivity, adjusted for inputs 1.0 0.5 0 –0.5 –1.0 –1.5 1.0 0.5 0 –0.5 –1.0 –1.5 1.0 0.5 0 –0.5 –1.0 –1.5 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 Farm experience (years) Farm experience (years) Farm experience (years) Sources: Original calculations for this publication, based on Bentze and Wollburg 2024; LSMS-ISA (Living Standards Measurement Study: Integrated Surveys on Agriculture) (dashboard), World Bank, https://www.worldbank.org/en/programs /lsms/initiatives/lsms-ISA. Note: Productivity is measured using total factor productivity, estimated as the residual from a Cobb-Douglas production function, with log yield (kilograms per hectare) as the dependent variable. Regressors include inputs (capital, labor, land), plot, and manager characteristics. Plot sizes (small, medium, large) are defined by country-specific terciles of plot size distribution. The plot-level data are harmonized. Among the self-employed, evidence indicates that returns to experience are limited in low- and middle-income countries. On average, in low- and middle- income countries, one additional year of self-employment experience yields