Anti-poverty programs often seek to improve their impact by targeting households for assistance according to one or more criteria. These are, however, often based upon measurements of key welfare indicators, such as income or consumption, in a single time period. This paper investigates whether it is possible to improve the accuracy of targeting by distinguishing between the chronically and transitorily poor on the basis of household characteristics.
Using the IFPRI household food security panel data for 686 households in rural Pakistan between 1986/87 and 1990/91, the authors show that most of the characteristics which distinguish the chronically poor from the transitorily poor are similar to those which distinguish the poor from the non-poor.
he paper then compares the poverty impact of policies designed to increase mean incomes and those designed to even out fluctuations of income over time. It concludes that large reductions in transitory poverty can be achieved by interventions designed to even out incomes whereas reducing chronic poverty in the long term will require large and sustained growth in household incomes.
Authors
