Identifying the poor for efficient targeting: Results for papua new guinea

Many countries try to protect their poor during structural adjustment. Targeting is needed to prevent benefits leaking to the non-poor but screening for direct transfers may be too costly for developing countries. One solution is indirect targeting, based on the characteristics of the poor. Household survey data for urban areas of Papua New Guinea are used to show how a poverty index and probit estimation can identify multiple characteristics of poor households. The results suggest that interventions should be targeted towards larger households, where the head is unemployed and has lower levels of schooling. Female headship and age are not useful characteristics for targeting.