A growing body of evidence shows that adverse environmental conditions affect human capital and socio-economic outcomes. However, little is known about whether the experience of such events early in life has effects on skill formation that are long-lasting and persist into adulthood, and about whether the demographic context mediates these effects. Here, we shed first light on these questions by combining data from a large cognitive skills survey spanning more than 30,000 individuals from birth cohorts over five decades in 11 countries, with retrospective information about weather conditions at the local level during early childhood. The results show that drought exposure during childhood has a negative effect on skill formation that persists later in life. This effect is mainly restricted to a high-fertility context, mitigated in families with high socio-economic standing, and by parents actively involved in their children’s school performance.

This paper investigates how much bilateral trade is affected by temporary disruptions of shipping networks caused by storms. Specifically, we examine how much trade is directed to other shipping routes or does not take place at all. We unpack the aggregate impact of oceanic cyclones by examining transportation volume and freight costs by a major container ship company. In event-studies nested in a gravity-style equation, we find that a cyclone reduces trade by 1.3% between countries across all modes of transport, while this effect more than doubles for affected port-pairs. After a storm, shipping firms charge higher freight costs on affected routes amplifying the trading impact. Adjacent shipping routes step in as substitutes for minerals and chemicals. Data on geo-located ship voyages show that following a storm, ships travel at slower speed incurring delay by up to 30 hours depending on the shipping company.

Have discriminatory housing policies contributed to today’s pollution and climate hazard disparities? We examine the impact of Redlining during the 1930s in the US, which assigned risk grades to neighborhoods according to housing characteristics and ethnic composition, on spatial patterns of urban environmental disparities. An extensive literature on environmental justice documents strong correlations between a worse neighborhood risk grade and higher air pollution and climate risks today. However, whether these disparities have been caused by redlining or merely coincide with pre-existing discrimination is unclear. Our analysis exploits an exogenous city size cutoff i.e. only neighborhoods in cities above 40,000 residents received risk grades from HOLC. We compare areas that received a particular grade with neighborhoods that would have received the same grade if their city had been treated. The control neighborhoods are defined using a machine learning algorithm trained to classify HOLC-like grades using full-count census records. Using local measures of environmental hazards, we find that the disparities exhibit the same pattern in treated and comparison cities, with meagre differences across the same grade in treatment and control cities. Instead, our results suggest that sorting, and alternative forms of discrimination drive contemporary environmental and climate disparities.

A growing body of literature indicates that heat stress and precipitation deficiencies can pose a critical threat to human health, particularly in less developed countries with low coping capacities and high exposure. The aims of this study are twofold. First, we shed light on the recall of drought events in rural Thailand by linking longitudinal survey data with objective meteorological data. Here, an anomaly in the survey design serves as a natural experiment. We find that a shorter time interval between surveys has a large positive effect on households correctly reporting a drought event. Second, we shed light on the health effects of droughts using the objective drought measure. In our panel over seven waves, we find a strong effect on diseases, which is slightly stronger than when using the reported drought measure.