Under the Weather: Health, Schooling, and Economic Consequences of Early-Life Rainfall.

Life in rural areas of developing countries is prone to many kinds of risk, such as illness or mortality of household members, crop or other income loss due to natural phenomena (weather, insect infestations, or fire, for example), and civil conflict. In addition to their contemporaneous effects, the effects of certain types of shocks may still be felt many years or even decades later. From a public policy standpoint, it is particularly important to identify shocks that have large long-run effects. Moreover, the mechanics underlying the persistence of shocks may be of considerable interest. For example, a health shock may have a long-run effect simply because the health shock itself persists over time. Alternately, the health shock may not directly affect longrun outcomes, but it could affect some other outcome—such as educational attainment—that helps determine long-run well-being. In this paper, we focus on shocks that occur at the very beginning of life. We ask how sensitive long-run individual well-being is to environmental conditions around the time of birth. In particular, we examine the effect of weather shocks around the time of birth on the adult health, education, and socioeconomic outcomes of Indonesian women and men born between 1953 and 1974. In addition, we attempt to shed light on the intervening pathways connecting early-life rainfall to adult outcomes, illuminating the roles of health and educational attainment in determining adult socioeconomic status. This investigation has considerable data requirements. It necessitates information on weather shocks experienced by individuals several decades earlier, as well as detailed current information on adult outcomes. We use information in the Indonesian Family Life Surveys (IFLS) on an individual’s year and location of birth, and link each individual in that survey to localityspecific rainfall data for their birth year. For individuals born in rural areas (on whom we focus), rainfall variation across space and time should generate corresponding variation in agricultural output and thus household income. To deal with measurement error in the rainfall data, we run instrumental variables regressions where variables for rainfall measured at slightly more distant rainfall stations serve as instruments for rainfall in the individual’s birthplace and birth year. We examine the impact of early-life rainfall on a range of adult outcomes observed in 2000. We find that higher early-life rainfall leads to improved health, schooling, and socioeconomic status for women. Women with 20 percent higher rainfall (relative to normal local rainfall) in their year and location of birth are 3.8 percentage points less likely to self-report poor or very poor health. They attain 0.57 centimeters greater height, attain 0.22 more completed grades of Under the Weather: Health, Schooling, and Economic Consequences of Early-Life Rainfall