Drawing on survey data from Young Lives, an international study of childhood poverty involving 12,000 children in four countries, this paper examines the effects of environmental shocks on food insecurity and children's development. The data, from children and their families living in rural and urban locations in Ethiopia, the Indian state of Andhra Pradesh, Peru, and Vietnam, provide information on the same individuals over time, allowing consideration of how earlier incidences of food insecurity and exposures to environmental shocks shape later outcomes. Regression analysis is used to estimate the relationships between these and other relevant factors. After introducing the data and methods, this report first considers the reported incidence of environmental shocks in the studied communities, and, second, the reported incidence of household food insecurity.
Following a review of previous Young Lives evidence on how households cope with these events we then analyse the effects of environmental shocks on households‟ food security and on children‟s nutritional outcomes, measured in terms of their height-for-age, or stunting. Environmental shocks are commonly reported, with droughts common in Ethiopia and Andhra Pradesh, and flooding common in Peru and Vietnam. Restricting analysis to reporting of these respective types of shocks in 2009 it is shown that residing in a rural location, living in poverty, and being engaged in agricultural work all increase the likelihood of reporting an environmental shock. The disparities in risks within any given location appear to be larger in urban areas; in rural locations there is more community-wide, covariate, reporting of shocks.
The report then moves from descriptive analysis to multivariate regression techniques, which highlight underlying associations by controlling for other explanatory risk factors. Evidence from Ethiopia shows that within communities, greater wealth and greater shares of income from non-farm sources reduce the chances of reporting drought, but that household wealth levels have less impact on reporting environmental shocks in the other countries. By and large, though, much of the variation in the reporting of shocks is down to where people live. In terms of the frequency of shocks, those households which experience recurrent shocks tend to be in rural areas, be more agricultural, and have lower wealth and maternal education levels than either those experiencing one-off shocks or those who never report a shock.
