Climate Shocks and Economic Growth in Sub-Saharan Africa

Swarthmore College, Swarthmore PA

Investigators

Abstract

Droughts are widely cited in both contemporary and historical accounts of African growth. The now massive empirical literature on economic growth has grappled with the puzzle of low and volatile African growth, but has had very little to say about the impact of climate shocks. This study develops empirical bounds on the impact of rainfall fluctuations on African growth and on the main channels of this impact. The project brings a widely useful rainfall dataset to the attention of researchers focusing on economic growth in developing countries, on African development in particular, and on macroeconomic performance in individual African countries. The project uses an historical dataset on national African precipitation patterns constructed for the purposes of the project by the principal investigators. The underlying data are historical monthly weather-station observations that have been geographically gridded by climatologists for use in large-scale climatology models. The data are rich enough to allow the construction of climate norms predating the period of analysis, which coincides with the availability of economic growth data beginning in the 1960s. The investigators use estimated rainfall norms to calculate annual country-level rainfall shocks, which they then use in country-by-country time-series regressions that relate economic growth, investment, and agricultural output to these shocks. Alternative transformations of the rainfall variable allow a sequence of readily-interpreted specifications to be implemented, including nonlinear specifications that allow for threshold effects, wet and dry outliers, and cumulative effects.