Opendata, web and dolomites


Teaser, summary, work performed and final results

Periodic Reporting for period 1 - MISSINGMIDDLE (The Causal Effect of Public Policy and Income on Child Health and Human Capital)


A recent literature in economics on the “fetal origins hypothesis” documented that a range of early prenatal and postnatal shocks and interventions can have substantial effects on long-term human capital formation (e.g. adult health, wages). However, we still know little...


A recent literature in economics on the “fetal origins hypothesis” documented that a range of early prenatal and postnatal shocks and interventions can have substantial effects on long-term human capital formation (e.g. adult health, wages). However, we still know little about the years in between early infancy and adulthood, referred to as the “missing middle”. How do early shocks affect health and human capital formation during childhood? How do the effects of different types of interventions or shocks at different ages compare? What are the most cost-effective ways of improving young children’s future outcomes?
I aim to fill this gap in the literature by taking advantage of several natural experiments in a country, Spain, for which high quality administrative data are available for the past 35 years. State of the art econometric techniques, combined with large sample sizes, will allow me to evaluate credibly and precisely the causal effects of a number of different public policies and shocks on child development.
I consider 5 different early “shocks” in early childhood (at different ages), affecting: i) Household material resources (an unconditional mother’s allowance); ii) Parental time (subsidized paternity leave); iii) Medical treatments around birth (elective delivery); iv) The availability of family planning services (access to abortion); and v) Aggregate demand shocks to different sectors of the economy.
I evaluate their impact on health and cognitive development at ages 0-15 as measured in hospital and primary health care records, school grades, and standardized test scores, among other. I also study the potential channels linking treatments to child outcomes, including family size, parental time use and labor market outcomes, expenditure patterns, etc.
Results will help us understand how shocks in early life can have long-term effects on human capital.

Work performed

The project evaluates the effect of a range of different “early interventions” on child outcomes. I proposed to study 5 different “shocks”, and was able to make progress on 3 of them:
1) Legalizing abortion
After abortion was legalized in Spain (1985), birth rates fell, especially among young women as did the number of marriages involving very young women. These effects were more pronounced in provinces where abortion services were provided in the early years after legalization.
We follow the affected cohorts of women over time across several data sources, and find that the short-term fall in births did not translate into lower completed fertility. Births and marriages were delayed, but eventually women caught up, and we actually find that women who had early access to abortion were less likely to divorce, probably because they married later and formed couples that were better matched.
The evidence also suggests that women who were able to avoid an early birth thanks to legal abortion were more likely to complete high school, and their long-term earnings were higher.
Results show that limiting access to abortion may have important negative short- and long-term consequences for women, without increasing completed fertility.
2) Paternity leave
We analyzed the effects of the introduction of paternity leave in Spain on parents’ outcomes. The main and surprising result was that families where the father was eligible for paternity leave after the birth of a child in 2007 were less likely to have another child in the future, compared with families where the father was not eligible. This may be because mothers go back to work earlier in households where the father can take paternity leave, but also because affected fathers have lower desired fertility.
We also find that affected fathers spend more time on childcare even several years after the birth of the child. We then analyze the effect of paternity leave on child development. We find that children whose fathers were eligible for paternity leave have higher grades in math in primary school. We also find some evidence suggesting that affected children may have more progressive views towards gender roles.
3) Family subsidies on children
First we compare children who were born just before and just after the introduction of a universal child benefit in Spain in 2007. We find no significant difference in health outcomes for children whose mothers received the benefit, compared with those who didn’t, at ages 0 to 8. We also find no significant differences in grades in primary school. We conclude that receiving the benefit after the birth of the child had no short or medium-term effects on health, and no medium-term effects on cognitive development.
We then study effects on health for the younger siblings of the children born around the benefit introduction date in 2007. The question is now whether benefit receipt (in 2007) had any effect on the health at birth of the next child. A recent literature suggests that targeting pregnant women may be more effective than subsidies after the birth of the child, given the importance of fetal development. We hypothesize that prenatal effects may be even stronger if vulnerable women are targeted even before they are pregnant. We find that low-income women who received the benefit after having a child in 2007 were less likely to have a low-birthweight baby in their following pregnancy.

Final results

Considering the abovementioned results, progress beyond the start of the art is :
1) The effects of legalizing abortion: our contribution to the existing literature derives from the fact that we are able to provide credible estimates of the causal effect of a major national reform in the legal regulation of abortion on long-term outcomes for women. Previous studies for other countries had focused on short-term fertility effects. We are able to combine sound identification with high-quality data to shed light on a very policy-relevant question.
2) The effects of paternity leave: The existing literature on the effects of paternity leave is still limited, given that these types of policies are relatively recent. We are the first to show that more generous paternity leave can have negative effects on fertility. Regarding effects on children, there is only one previous study documenting effects on school outcomes in Norway. Providing additional evidence from other countries is also a relevant contribution to this recent literature.
3) The effects of family subsidies on children: To our knowledge, we are the first to provide strong causal evidence comparing the effects on child health of a benefit paid after the birth of a child, versus before the mother is pregnant.

Website & more info

More info: