Do prenatal child benefits improve infant health outcomes?

A pregnant woman sits on her bed, holding her hands over her belly

Recent research has found that cash transfers that disadvantaged women in the United States or in Latin America received during pregnancy had a positive effect on the health of their newborn children.

In Barcelona School of Economics Working Paper 1261, “Prenatal Transfers and Infant Health: Evidence from Spain,” Libertad González and Sofia Trommlerová study the effect of a cash transfer targeting new mothers in Spain on their subsequent children’s health. In contrast to previous research, this cash transfer took place even before the woman was pregnant with the next child.

In July 2007, Spanish prime minister Zapatero announced that a generous, lump-sum child benefit of 2500 euros would be paid to all women giving birth from the day of his announcement onwards. The fact that the new policy was introduced unexpectedly creates an excellent natural experiment which is exploited by the authors through a Regression Discontinuity Design (RDD) strategy. 

The main research question of this paper is whether the health of subsequent newborns (i.e. the next child after the one that made women eligible for the subsidy) improved as a consequence of the cash transfer.

Methodology and results

The authors use birth weight of singleton children as the main marker of fetal health. Birth weight has been shown to predict long-term health and economic success. In their identification strategy, a woman is considered “treated” if she gave birth to a child soon after the policy announcement, and thus became a potential recipient of the benefit. Women who gave birth just before the announcement belong to the “control” group. The identifying assumption is that women who give birth just before or just after the unexpected announcement are very similar. Therefore, any changes in the health of their next child should not be affected by any other factor other than eligibility for the benefit.

Armed with population-wide, administrative microdata on newborns’ health and parents’ socio-demographic characteristics, and with survey data on household income and maternal health, the authors seek to answer the main research question. They are also interested in exploring possible channels behind the observed changes in children’s health. 

Before moving on to the main analysis, one important aspect needs to be checked. If the subsidy produced an increase in fertility among treated women (via an income effect) and/or changes in family characteristics among women who decided to have another child, the results of the study would be questionable. To rule out this possibility, the authors study whether there were any changes in fertility patterns and in family characteristics (among those women who did have another child) between treated and control women. There is no evidence that either of these dimensions were affected by the child benefit. These results support the validity of the identification strategy.

Following this validation exercise, the authors move on to the main analysis. They find that there is a statistically significant relationship between the treatment (i.e. having received the child benefit) and the birth weight of the next child. This result holds only for very low birth weight (VLBW), i.e. children born with less than 1,500 grams. Women who were entitled to receive the baby bonus in 2007 were 0.36 percentage points less likely to give birth to a baby with VLBW in the next five years, which is a decrease by 83%.

The results suggest that children that would have been born under 1,500 grams in the absence of the child benefit, were born with 2,000-2,500 grams instead, as a result of the cash transfer. This is robust to the inclusion of day and region fixed effects, parental characteristics, and seasonality adjustments. 

The next intuitive question to ask is whether this effect varies depending on the socio-economic status of the family. As expected, low income families (the lowest two quintiles of the income distribution) are the main drivers of the results, see Figure 1. The authors also consider maternal education (low-educated versus high-educated) and marital status (unmarried versus married women) for robustness, and the more disadvantaged group always drives the results.

Figure 1: Probability that the subsequent child has birth weight below 1,500 grams

Notes: Sample of low-income households, i.e. the lowest two quintiles of the income distribution. Displayed are weekly averages 8 weeks before and 8 weeks after the policy announcement.

As the vast majority of children born with VLBW are premature, the authors also distinguish between babies born prematurely and babies born on term. The child benefit led to a significant and substantial improvement among premature babies, whose birth weight increased by 100 grams on average. 

Possible mechanisms

What can explain the observed improvements in health at birth among babies conceived and born after their mother received a generous cash transfer? Among the two possible biological causes for higher birth weight – faster intrauterine growth and longer gestation – the authors can confirm the first one but not the second one. This means that babies did not stay longer in the womb, but they did grow faster in-utero.

Adding another layer to the analysis, the authors explore the underlying channels that might be behind faster intrauterine growth. On the one hand, eligible mothers seem to have eaten healthier, reduced their drinking and smoking, and increased their healthy habits. On the other hand, they find no significant changes in cohabitation or employment patterns, meaning that women did not work more (or less) and were not more (or less) likely to live with a partner after they received the benefit.


This paper is the first one to explore the effects of a cash transfer granted to women in the preconception period. There are three main findings. First, the benefit had a positive impact on children’s health – it decreased the fraction of very low birth weight babies. Second, this effect was higher in disadvantaged families. Third, the improvements seem to be a consequence of better maternal health. 

The present work may have relevant implications for policy. First, cash transfers to women seem to matter not only during pregnancy but also prior to it. Second, the main beneficiaries of such transfers are disadvantaged women and their children. Hence, policies that improve the economic situation in low-income households may have far-reaching positive impacts on the health and wellbeing of the next generation.