The long-run impacts of paid maternity leave on height and educational attainment

The paper examines whether and to what extent paid maternity leave has long-run impacts on human capital, focusing on adult educational attainment and height. Contextually, the International Labour Organization sets a global standard recommending at least 14 weeks of paid maternity leave, and nearly all countries have some form of paid leave, though details vary. Prior research emphasizes short- and medium-run benefits of maternity leave on maternal and child health, labor market outcomes, and child development, but evidence on adult outcomes is limited. This study aims to fill that gap by leveraging multi-country data over nearly three decades to assess the persistent effects of early-life exposure to paid maternity leave on adult outcomes, thereby informing policy on women’s well-being and sustainable development.

The review highlights two strands of literature. First, studies show maternity leave improves child health (lower hospitalization and mortality) and educational outcomes (higher test scores, lower dropout), and benefits mothers’ labor market and mental health outcomes. Evidence in developing countries indicates increases in vaccination and reductions in infant mortality with longer paid leave. Second, research on early-life conditions shows nutrition, disease exposure, and socioeconomic status shape adult height and education. Height reflects cumulative net nutrition and correlates with cognition and health. Education is associated with lower mortality, better intergenerational health, and higher earnings and financial outcomes. Government policies in early life (e.g., preschool, food stamps, Medicaid) can improve long-term outcomes. This study connects these literatures by assessing whether paid maternity leave—an early-life policy—affects adult educational attainment and height.

Data come from the Demographic and Health Surveys (DHS) Household Members files (1990–2019) for background characteristics, educational attainment (years), and height (cm), and from the World Bank’s Women, Business and the Law 2021 database for country-year maternity leave laws (availability, payment, administration, and duration). Using birth year from DHS, each individual is assigned the duration of paid maternity leave in their birth year. The sample is restricted to individuals aged 18+ and to countries where maternity leave benefits are fully administered by the government (public funding, mandated private insurance, social insurance, or reimbursed employer-paid leave) to avoid endogeneity from optional employer policies. The main outcomes are years of education and height in centimeters; robustness checks use high school completion (indicator) and height-for-age Z-scores. The primary specification is an OLS regression of outcomes on paid maternity leave duration (weeks) at birth, controlling for individual characteristics (gender, marital status, age, age-squared, rural residence, household head), with birth-year and survey-year fixed effects. Preferred models include residential cluster fixed effects (neighborhoods) to control for unobserved local heterogeneity; country fixed effects are also examined. Sampling weights are applied and standard errors are clustered at the residential cluster level. Identification exploits within-location, across-cohort variation in leave duration induced by policy reforms; the key assumption is that reform timing is uncorrelated with residential cluster-specific factors at birth that affect adult outcomes. Robustness analyses include: restricting to cohorts within 5- and 2-year windows around reforms; unweighted regressions; alternative outcome measures; adding country-specific quadratic trends; and raising the age cutoff to 22+.

• Main effects (preferred specification with residential cluster fixed effects): Each additional week of paid maternity leave at birth increases educational attainment by about 0.007 years and adult height by about 0.056 cm.
• Simpler specifications show larger correlations that attenuate with controls and fixed effects: Education effect declines from ~0.095 (bivariate) to 0.067 (with controls), 0.016 (country FE), and 0.007 (cluster FE). Height effect declines from ~0.079 (bivariate) to 0.043 (with controls), 0.044 (country FE), and 0.056 (cluster FE).
• Robustness checks:
  • Cohort windows around reforms: 5-year window yields +0.010 years of education and +0.031 cm height per week; 2-year window yields +0.008 years and +0.029 cm, all statistically significant.
  • Unweighted regressions: +0.007 years and +0.050 cm per week.
  • Alternative outcomes: +0.4 percentage points in high school completion probability and +0.010 SD in height-for-age Z-score per week.
  • Country-specific quadratic trends included: effects remain ~+0.007 years and +0.056 cm per week.
  • Higher age cutoff (22+): +0.008 years and +0.051 cm per week.
• Sample: ~1.23 million individuals for education and ~326,000 for height across 29 developing countries, 1990–2019. Average education 8.04 years; average height 157.2 cm; average paid leave at birth 11.54 weeks.
• Magnitude perspective: With a ~20 cm difference between the tallest and shortest countries, the estimated height effect per additional week is about 2.8% of that gap when scaled across policy differences discussed in the paper. Effects are comparable to those from early-life rainfall shocks and class size reductions documented in prior research.

Findings indicate that longer paid maternity leave has persistent positive effects on adult human capital—more schooling and taller stature—suggesting that early-life policy interventions can yield long-term benefits. Proposed mechanisms include: (1) increased maternal time with infants, improving breastfeeding, immunization, health, behavior, and cognitive development; (2) reduced reliance on non-parental care, which is associated with adverse cognitive and behavioral outcomes; and (3) alleviation of financial constraints via statutory pay, enabling investment in child health inputs. The magnitudes align with broader literature on early-life conditions (e.g., rainfall shocks, class size) and reinforce the importance of prenatal and early postnatal environments. Results are robust to multiple specifications, alternative outcomes, and additional controls for trends, supporting the interpretation that policy-driven extensions of leave contribute to long-run human capital accumulation.

The study demonstrates that extensions in paid maternity leave in developing countries produce lasting gains in adult educational attainment and height. Leveraging DHS and World Bank legal data and within-location, across-cohort variation from policy reforms, the authors find that each additional week of paid leave at birth raises schooling and adult stature, with effects robust across numerous checks. These results add to evidence that early-life policies have enduring benefits and suggest that benefits of supporting women’s well-being through paid maternity leave extend into the next generation’s human capital. The authors advocate for the enforcement and extension of paid maternity leave as a pathway to sustainable development.

• Identification relies on the assumption that the timing of maternity leave reforms is exogenous to residential-cluster-specific factors influencing later-life outcomes; unobserved confounding cannot be entirely ruled out despite extensive fixed effects and trend controls.
• Windowed analyses (2- and 5-year cohorts) provide sharper identification around reforms but reduce sample size, potentially lowering estimate efficiency.
• The sample is restricted to countries where maternity leave benefits are fully administered by governments or reimbursed employer schemes, which helps avoid endogeneity but may limit generalizability to settings with employer-discretionary leave policies.
• DHS-based measures and cross-country legal data may not capture actual take-up or full utilization heterogeneity of leave across locales within countries, though residential cluster fixed effects mitigate this concern.