This article introduces a new methodology to estimate climate exposure at the
household’s level with the standardized precipitation evapotranspiration index (SPEI)
as its building block. As the probability distribution of the SPEI is known, one can
easily recover the marginal probability distribution of expected consumption.
Furthermore, the approach is simple enough to accommodate quantile regressions and
hence offer the opportunity to broaden the scope of the analysis to different categories
of the population. I illustrate the methodology with a case study on Ethiopia. I find
notably that while poor households in the most remote villages are almost as resilient
to a 10-year return period drought as poor households living in the vicinity of a town
(up to 20 km), the contrary is true for richer households: the ones living in remote
parts of Ethiopia are much more at risk than their suburban counterparts.

Keep in touch with the Grantham Research Institute at LSE
Sign up to our newsletters and get the latest analysis, research, commentary and details of upcoming events.