Résumé

Floods are among the most common and destructive natural disasters, and climate change is increasing flood risk across the world. But identifying exposure to floods is challenging and the evidence on how these disasters affect livelihood decisions of farm households in developing countries is limited. This paper uses nationally-representative panel household survey data from Nigeria in a difference-in-differences framework to analyze how exposure to floods in 2012 affects household labor supply and livelihood outcomes in agricultural communities over the following seven years. We first show that identification of community flood exposure varies depending on whether household survey reports or satellite imagery are used, with measurement and definition issues in both sources leading them to capture different types of flood events. These differences affect conclusions about impacts of floods identified by survey or by satellite only. Focusing on impacts of flooding identified by both measures, we find persistent negative effects on household food security and net farm production income on average, but positive effects on participation in wage employment, household farm labor hours, and cultivation of staple crops. In particular, households exposed to floods are more likely to cultivate rice, driving the increases in farm labor and area planted and contributing to decreased production of more valuable non-staple crops. The results highlight the importance of flood measurement decisions, and indicate that households in agricultural communities respond to floods by both attempting to diversify income sources and adapting agricultural production.

Informations pratiques