Association for Public Policy Analysis & Management

Multiple analyses of historical weather and socio-economic data have now produced a substantial body of robust evidence that high temperature anomalies lead to a range of adverse social and economic impacts, including reductions in economic productivity and growth in both the agricultural and non-agricultural sectors, increases in morbidity, crime and conflict. Another, independent body of evidence establishes the long-term impacts of early life stress on adult socio-economic indicators, health and well-being. In particular, the fetal origins hypothesis posits that in-utero circumstances can have substantial long-term impacts on human development. Numerous studies have provided evidence in support of this hypothesis, finding that economic, environmental or disease-related stress in infancy or in-utero lead to long-term impacts on physical and cognitive health, educational attainment and wages. Taken together, these two sets of findings suggest that higher temperatures at the time of birth may have long-term productivity consequences: individuals who are exposed to high ambient temperatures in-utero or in infancy may experience life-long negative consequences through a number of possible channels, both physiological and economic (for example, declines in income or economic hardship can reduce consumption of crucial nutritional inputs by pregnant women). To the best of our knowledge and, despite the obvious importance to climate change impacts, we are not aware of studies that have investigated this link.

In this paper, we investigate the effect of high temperature anomalies around the time of birth on formal earnings as an adult. Economic theory suggests that in well-functioning markets, wages provide an accurate indicator of economic productivity and human capital, including physical and cognitive function. A relationship between temperature anomalies in-utero and adult earnings would therefore measure the total economic losses associated with long-term human capital losses resulting from stress in-utero. Our analysis makes use of a unique data set on the 2010 earnings of all 1.5 million formal sector workers in Ecuador, born between 1950 and 1989, that was merged with civil registry data to identify the place and time of birth of these individuals, and then merged with historical weather data sets to identify temperature and precipitation levels around the time of birth (including in-utero). A regression analysis of the relationship between adult earnings and temporal temperature anomalies revealed that higher temperatures in-utero lead to significantly lower adult earnings for women, with a 1°C increase in average monthly temperature in-utero leading to a 1.1%-1.7% decrease in adult earnings. These results are highly robust to the inclusion of fine geographic controls, localized annual cycles and time trends, and to various falsification tests. Even though the reduced-form analysis does not allow us to identify which of the established negative impacts of high temperatures is driving the association, the random nature of temperature variations over time within a geographical locality facilitates causal inference.