| Poverty maps are highly disaggregated databases of poverty and inequality and there is a growing demand for them by policy-makers and researchers who want to learn more about targeted programs, local outcomes, and decentralization. Unfortunately, challenges remain in producing accurate, localized poverty maps since most household surveys are too small to be representative at such levels of disaggregation, and census data do not contain the required information to calculate poverty. On November 14, 2002, World Bank research economists, Peter Lanjouw, Berk Özler, and Johan Mistiaen presented an overview of their work in developing a methodology to estimate welfare indicators for small areas. They also described the preliminary results from pilot projects in Ecuador, Madagascar, Kenya, and South Africa.
Lanjouw addressed the difficulties they have encountered while attempting to produce poverty maps. Afterwards, he described several options including collecting household data at a highly disaggregated level, producing a basic needs index from census and other data sources, and applying a methodology of combining survey and census data. Lanjouw described the method of combing census and survey data in greater detail and explained how the approach assigns a measure of welfare from household surveys into the census using statistical prediction methods. He noted that this method builds on the foundation of traditional poverty analysis and provides a good measure of welfare, adjustments of spatial price variation, and a poverty line.
Afterwards, he described the three stages of the methodology and provided preliminary results from Ecuador. He was followed by Johan Mistiaen who covered results on urban poverty maps in Nairobi and a decomposition of urban poverty in Madagascar. Berk Özler described results from his research on crime and inequality in South Africa.
|
Home
> Learning
& Knowledge > BSPAN
> Presentation
High Speed
Low Speed
Audio 
