Does rainfall matter for economic growth? Evidence from global sub-national data (1990–2014)

This paper addresses a seldom recognized empirical puzzle. Much micro-econometric evidence suggests that precipitation has wide ranging impacts on vital economic indicators such as agricultural yields, human capital and even conflict. And yet paradoxically most macro-econometric evidence (especially in the climate economics literature) finds that precipitation has no robust and significant impact on various measures of aggregate economic output. Here, we argue that spatial aggregation of weather at the country level explains this result. We use annual sub-national GDP data to show a concave relationship between precipitation and local GDP growth between 1990 and 2014. We demonstrate that when data are aggregated at larger spatial scales, the significance of the impact decreases and eventually it vanishes. The impact of precipitation on aggregate economic activity is predominantly felt in developing countries and is insignificant in developed countries. Agriculture is found to be a key pathway. Our results have substantial consequences for measuring the economic impact of climate change.