Climate and poverty in Africa South of the Sahara. (January 2020)

• Record Type: Journal Article
• Authors: Azzarri, Carlo; Signorelli, Sara
• Is Part Of: World development. Volume 125(2020)
• Publication Date: 2020-01
• Subjects: Mapping -- Poverty -- Climate change -- Weather shocks -- Sub-Saharan Africa -- Spatial models; Economic history -- 1990- -- Periodicals; Economic assistance -- Developing countries -- Periodicals; 330.9
• DOI: 10.1016/j.worlddev.2019.104691
• Languages: English
• ISSNs: 0305-750X

Abstract

Highlights: The study uses a novel, Africa-wide dataset of nationally-representative, consumption-based household surveys. In SSA long-term climate conditions yield differential effects on welfare compared to short-term climatic shocks. Excess rainfall disproportionately hinders welfare in selected SSA regions while drought shows uncertain spatial effects. Exposure to flood shock is associated to 35% decrease in consumption and 17 percentage points increase in extreme poverty. Smallholder farmers appear to be the most vulnerable to weather variability, both in case of floods and droughts. Abstract: To estimate the effects of weather conditions on welfare globally, cross-country comparisons need to rely on international poverty lines and comparable data sources at the micro-level. To this end, nationally representative household surveys can offer a useful instrument, also at the sub-national level. This study seeks to expand the existing knowledge on the determinants of poverty in Africa south of the Sahara (SSA), examining how long-term climatic conditions and year-specific weather shocks affect expenditure per capita. We take advantage of a novel and unique dataset combining consumption-based household surveys for 24 SSA countries -representative of more than half of the African population and two thirds of SSA- and geospatial information on agro-climatic conditions, market access and other spatial covariates of poverty. To our knowledge, it is the first time that a welfare-based, multidisciplinary, micro-level dataset with such wide spatial coverage has been assembled and examined. Our analysis relies on a linear and spatial model at the household- and district-level, respectively, both controlling for socio-economic, demographic, and geographic confounding factors. Results are consistent across econometric approaches, showing that living in more humid areas is positively associated with welfare, while the opposite occurs living in hotter areas, as existing literature shows. Flood shocks -defined as annual rainfall higher than one standard deviation from the 50-year average- are associated to a 35% decrease in total and food per-capita consumption and 17 percentage point increase in extreme poverty. On the other hand, extreme shortages of rain and heat shocks show an uncertain effect, even when estimates control for spatial correlation between welfare and weather conditions using the spatial error correction model. Given the heterogeneous effects of climatic events across SSA macro-regions, local-specific adaptation and mitigation strategies are suggested to help bringing households on a sustainable path.