The Gendered Spillover Effect of Young Children's Health on Human Capital: Evidence from Turkey

Reducing morbidity among preschool children may improve educational outcomes for their adolescent sisters by lessening their childcare duties and allowing them to attend school, thereby increasing their educational attainment and literary rates. This is defined as a “spillover effect” because those who significantly benefited from a policy (i.e. adolescent sisters) were not the ones that the policy primarily targeted (i.e. children under five years old).

• ·       The Turkish vaccination campaign was associated with a 5% decline in disability, a 3% increase in literacy and a 2% increase in educational attainment overall in all age-eligible children.

The effect of the vaccination campaign on educational returns was found to be highest for girls with siblings, with an additional 0.0058-point increase in literacy and 0.0130 unit increased in educational attainment for this group.

• For adolescent girls, the vaccination campaign was also associated with an increase the female to male ratio in school enrollment by 0.016.
• The spillover effect was found to be stronger if the mother worked outside the home, especially in the cotton fields.
• The greater number of young children in the household, the stronger the spillover effect was.
• In other words, the adolescent girls showing the greatest educational benefits from the vaccination policy were also from households where there were many young children vaccinated under the policy.
• The spillover effect was stronger for girls from households with an older brother.
• The spillover effect was absent if an elder sister or grandmother was present in the household. In other words, the effect of this campaign on adolescent girls was minimized if her household already had someone that would perform the maternal labor for the sick younger child instead.

The researchers used an economic framework for analyzing domestic labor—in this case, childcare to interpret this data. When younger siblings get sick, older sister often substitute for maternal labor, especially if the mother works outside the home. The Turkish vaccination policy led to fewer children becoming sick from the immunized diseases and potentially other diseases given the link between measles infection and immune system function, and thus adolescent girls with vaccinated younger siblings spent less time caring for sick younger siblings and more on schooling.

In summary, those most responsible for childcare duties are particularly benefitted when policies address the health of the children they care for. This case highlights the importance of examining healthcare interventions for potential education and gender impacts. Furthermore, this study shows how improving child health, in addition to affirmative action and economic growth programs, is an important policy tool for closing gender gaps in developing countries.

In this study, the researcher conducted a quasi-experimental analysis on data from the Census, Demographic and Health Surveys (DHS), and National Education Statistics of Formal Education in Turkey in order to assess the impact of the 1985 Turkish National Immunization Campaign on childhood health and educational attainment. A difference-in-differences (DD) estimator was employed to assess these metrics. The cohort impacted by the vaccination campaign—which targeted children under five years old—were those between the ages of 0 and 60 months at the time of the campaign, i.e. born between 1980 and 1986.

The “spillover effect” of this vaccination policy was isolated by measuring the health and educational attainment of the older siblings of these under-five-years-old children who were affected by the vaccination campaign. This older sibling cohort is defined as children who were 11 to 14 years old at the time of the 1985 and 1990 census and had younger siblings in the household. This age range was selected because this group was young enough to still have lived in the same household as their younger siblings, but old enough to not be eligible for the vaccination campaign themselves.

A placebo test was conducted for the health and educational attainment of households where there were no young children. The placebo test showed that the health and educational attainment spillover effects were not observed in these control groups. One potential concern about the use of these statistics is that often “vital events,” such as birth of a child, are underreported in low- and middle-income countries. This problem was addressed by the researcher using a correction factor (CF) for the under-reporting using a unique question asked on the 1985 Turkish census.