Poverty and natural disasters: A meta-analysis

We conduct a meta-regression analysis of the existing literature on the impacts of disasters on households, focusing on the poor and on poverty measures. We find much heterogeneity in these impacts, but several general patterns, often observed in individual case-studies, emerge. Incomes are clearly impacted adversely, with the impact observed specifically in per-capita measures (so it is not due to the mortality caused by the observed disaster). Consumption is also reduced, but to a lesser extent than incomes. Importantly, poor households appear to smooth their food consumption by reducing the consumption of non-food items; the most significant items in this category are spending on housing, health, and education. This suggests potentially long-term adverse consequences as consumption of these services is often better viewed as long-term investment. We do not find consistent patterns in long-term impacts; it appears the limits of the meta-regression methodology prevent us from observing patterns in the relatively few heterogeneous research projects that examine these long-term effects. The importance of addressing risk within the context of sustainable development and poverty alleviation is clear. The impact of disasters on the poor may be increasingly worrying considering the climate variations we anticipate.