In communities highly dependent on rainfed agriculture for their livelihoods, the common occurrence of climatic shocks such as droughts can lower the opportunity cost of having children, and raise fertility. Using longitudinal household data from Madagascar, we estimate the causal effect of drought occurrences on fertility, and explore the nature of potential mechanisms driving this effect. We exploit exogenous within-district year-to-year variation in rainfall deficits, and find that droughts occurring during the agricultural season significantly increase the number of children born to women living in agrarian communities. This effect is long lasting, as it is not reversed within four years following the drought occurrence. Analyzing the mechanism, we find that droughts have no effect on common underlying factors of high fertility such as marriage timing and child mortality. Furthermore, droughts have no significant effect on fertility if they occur during the non-agricultural season or in non-agrarian communities, and their positive effect in agrarian communities is mitigated by irrigation. These findings provide evidence that a low opportunity cost of having children is the main channel driving the fertility effect of drought in agrarian communities.
Sylvain Dessy : Department of Economics and CRREP, Université Laval, Quebec, Canada; sdes@ecn,ulaval.ca
Francesca Marchetta : Université Clermont Auvergne, CNRS, IRD, CERDI, Clermont-Ferrand, France; firstname.lastname@example.org
Roland Pongou : University of Ottawa, ONT, Canada and Harvard University, Boston, USA; email@example.com
Luca Tiberti : Department of Economics, Université Laval, and PEP; firstname.lastname@example.org
Acknowledgments: Francesca Marchetta acknowledges the support received from the Agence Nationale de la Recherche of the French government through the program "Investissements d’avenir" (ANR-10-LABX-14-01)"; Luca Tiberti acknowledges the support from the Partnership for Economic Policy (PEP), which is financed by the Department for International Development (DFID) of the United Kingdom (UK Aid) and the International Development Research Centre (IDRC) of Canada. We are grateful to seminar participants at the University of Ottawa, theWorld Bank, and the CRREP/CDR conference (01-11-2019)