아프리카ㆍ중동

Abstract

Poverty estimates based on cross-section data provide static snapshots of poverty rates. Although a time series of cross-section data can offer some insights into poverty trends, it does not allow for an assessment of dynamics at the household level. Such a dynamic perspective on poverty generally calls for panel data and this kind of analysis can usefully inform poverty reduction policy, notably the design of social protection interventions. Absent actual panel data for Senegal, this paper applies new statistical methods to construct synthetic panel data from two rounds of cross-section household surveys in 2005 and 2011. These data are used to study poverty transitions. The results suggest that, in marked contrast to the picture obtained from cross-section data, there exists a great deal of mobility in and out of poverty during this period. More than half the population experiences changes in its poverty status and more than two-thirds of the extreme (food) poor move up one or two welfare categories. Factors such as rural residence, disability, exposure to some kind of natural disaster, and informality in the labor market are associated with a heightened risk of falling into poverty. Belonging to certain ethnicities and factors such as migration, working in the non-agriculture sector, and having access to social capital are associated with a lower risk of falling into poverty.