A growing body of research argues that early adverse experiences have lasting effects not only on later health outcomes, but also on human capital accumulation. This interest has been inspired by the modern literatures in epidemiology and economics, which demonstrate that genetics do not exclusively explain the evolution of health, cognitive and socio-emotional skills. In contrast, family investments and the environment play a major role in determining skill development and subsequent inequalities (Almond and Currie, 2011; Barker, 1995; Cunha and Heckman, 2007; Gluckman and Hanson, 2005; Heckman, 2007).

In this paper, I investigate the persistent effect of negative shocks early in life on children's health and cognitive outcomes, and explore whether shocks at certain periods matter more than others. I exploit the geographic intensity of extreme floods during the 1997-1998 El Niño phenomenon in Ecuador as a source of exogenous variation in children's exposure to a negative shock at different periods early in life.  Extreme weather events are a relevant type of shock because recent trends in global climate change suggest that they may become more frequent and their intensity less predictable (Kovats et al., 2003). Income losses, consumption fluctuations, and infectious diseases are among the consequences of weather shocks (Baez et al., 2010). These changes can disrupt development before birth and during early childhood, both of which are critical periods for future human capital accumulation.

I show that children who were in-utero during the El Niño 1997-98 are shorter in stature, tend to be more anemic, and score lower in vocabulary tests five and seven years after the shock. Average effect sizes fall between 0.09 and 0.13 standard deviations. Furthermore, the timing of exposure results are in line with the medical findings on sensitive periods for growth and brain development. The negative effect on height comes from exposure during the third trimester in-utero, while the deficit in cognition comes from exposure during the first trimester.

To explore potential mechanisms behind these persistent effects, I look at the impact of El Niño floods on inputs to human capital production: birth endowments and family inputs (income, consumption, and breastfeeding). Children exposed to El Niño floods, especially during the third trimester in utero, were more likely to be born with low birth weight. Furthermore, households affected by El Niño 1997-98 suffered a decline in income, total consumption, and food consumption in the aftermath of the shock. Moreover, exposure to El Niño floods decreased the duration of exclusive breastfeeding and increased the duration of non-exclusive breastfeeding. Falsification exercises suggest that selection concerns such as selective fertility, mobility, and infant mortality do not drive the results. 

This paper highlights relevant implications for developing policies that improve the coping mechanisms against negative shocks that affect both income, and parental stress and health. Those policies should target disadvantaged families with children and pregnant women not just before giving birth, but also early in pregnancy to prevent the negative consequences on children's development.