Inter-generational and Within-generation Spillovers in Human Capital Formation

McQuade, Gerry and Porter, Catherine and Singhal, Saurabh (2025) Inter-generational and Within-generation Spillovers in Human Capital Formation. PhD thesis, Lancaster University.

Text (2025McQuadePhD)
2025McQuadePhD.pdf - Published Version
Available under License Creative Commons Attribution-NonCommercial-NoDerivs.
Download (3MB)

Abstract

This thesis consists of three pieces of applied work, each combining multiple data sources to investigate the short, medium and long-term determinants of health and human capital formation in Peru. Chapter 2 examines the multigenerational effects of maternal grandmothers’ exposure to drought during pregnancy. Matching longitudinal data on the birthdate and location of the mother of Young Lives respondents in Peru with external climate data, I find that exposure to drought has a persistent negative impact on the health stock of the daughter and grandchild. Grandchildren display lower height-for-age, beginning in early childhood and persisting through adolescence, with the height gap widening as they enter puberty. Additionally, grandchildren have lower early-life weight-for-age, however this effect diminishes with age. The effect is strongest for grandsons, and isolated to grandmothers living in rural areas during exposure, with exposure in early pregnancy having the largest impact. The first generation are also affected, with mothers being shorter in stature in adulthood. Estimating the average controlled direct effect (ACDE), I find that the mother’s long-term health is the primary mediator for transmission between the first and second generation, although I cannot fully rule out other unobserved mechanisms. Chapter 3 also matches Young Lives data to historical climate data to assess how exposure to rainfall shocks in early life can have a persistent influence on the development of personality traits in adolescence and adulthood. I find high rainfall exposure in the second and third years of life negatively affects scores. Additionally, high prenatal rainfall, specifically in the 3rd trimester, has a positive impact on scores, driven by girls and those in the poorest households. Examining underlying mechanisms, I find that parents increase labour supply in response to shocks, which has a negative impact on parent-child interaction in early life, with no effects on measures of material investments or child physical development. Chapter 4 combines school administrative data with national standardised tests to identify siblings and estimate spillover effects in attainment arising from a public school-day extension reform. Using a regression discontinuity design based on school eligibility criteria, we estimate a positive effect of older siblings’ schooling on younger sibling attainment in reading and mathematics. Positive spillover effects are driven by the effect on girls, with the largest effect amongst sister-sister pairs, compared to null effects for younger brothers. Our results indicate that evaluations which consider only the benefits for the targeted child could systematically understate the benefit-cost ratio of educational reforms.