Working Paper: CEPR ID: DP9222
Authors: Masayuki Kudamatsu; Torsten Persson; David Strömberg
Abstract: We estimate how random weather fluctuations affected infant mortality across 28 African countries in the past, combining high-resolution data from retrospective fertility surveys (DHS) and climate-model reanalysis (ERA-40). We find that infants were much more likely to die when exposed in utero to much longer malaria spells than normal in epidemic malaria regions, and to droughts in arid areas, especially when born in the hungry season. Based on these estimates, we predict aggregate infant deaths in Africa, due to extreme weather events and to maternal malaria in epidemic areas for 1981-2000 and 2081-2100.
Keywords: climate change; maternal malaria; maternal malnutrition; natural experiments
JEL Codes: I15; O13; O15; O55; Q54
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Weather shocks (Q54) | Infant mortality (J13) |
| Prolonged malaria spells (C41) | Infant mortality (J13) |
| Droughts in utero (J13) | Infant mortality (J13) |
| Droughts during the hungry season (Q54) | Infant mortality (J13) |
| Extreme malaria episodes (E65) | Infant deaths (J13) |
| Droughts in arid areas (Q54) | Infant deaths (J13) |
| Climate change scenarios (Q54) | Infant mortality risks (J13) |