Research Findings

Viet Nam Publication Abstracts

Generating Poverty Maps: An Application to Viet Nam

Over the past 15 years, researchers and policymakers have become increasingly interested in using geographic targeting to improve the cost-effectiveness of programs to alleviate poverty and food insecurity. Poverty maps are useful to target spending on infrastructure, health, education, and nutrition. Previous research shows that such targeting is not very accurate unless the geographic units are small. It is not easy, however, to obtain information on poverty for a large number of small regions (e.g. districts or villages) throughout a country. Sample size constraints generally prevent the use of household surveys for estimating poverty at the neighborhood, city, or even district level. Most household budget surveys allow estimates of poverty for just 5 to 10 regions.

This study develops a method for generating disaggregated poverty maps by combining survey and census data and applies the method to Viet Nam. First, the relationship between rural poverty and 25 household indicators was estimated using household survey data taken from the 1992-93 Viet Nam Living Standards Survey. Then, census data on those same indicators were used to estimate poverty rates for each of the 543 rural districts in Viet Nam. The results were presented in the form of district-level poverty maps using geographic information system (GIS) software.

The results demonstrate that many household characteristics are, individually, fairly weak predictors of poverty. When combined using probit regression analysis, however, they are much better at identifying poor households. A program targeting households according to a poverty index combining 19 household characteristics and 6 regional dummy variables could reduce leakage and undercoverage rates to 17 percent. Poverty is concentrated in the north and in districts furthest from the coast and cities. However it is unclear whether this is attributable to lack of access to markets or rather geographic factors not included in this analysis. The study further demonstrates the possibility of combining household survey data and census data to generate highly disaggregated maps of poverty. The method was used to produce a map classifying rural districts in Viet Nam, although the same method could be used to classify rural communes (of which there are close to 8,800).

Key words: poverty, poverty mapping, rural poverty, safety nets, targeting, Viet Nam

last updated: February 25, 1999