Joseph Florent Feulefack, Manfred Zeller:
How Accurate Is Participatory Wealth Ranking (PWR) in Targeting the Poor? A Case Study from Bangladesh


Georg-August-University Göttingen, Institute of Rural Development, Germany

The UN Millennium Development Goals seek to halve the number of people living below a dollar per day by 2015. Many development policies and projects therefore seek to target the (dollar-)poor in the provision of goods, capital, and services, and use a number of poverty targeting methods. PWR is a participatory poverty assessment method that relies on local reference groups rating the relative poverty status of households in their community. This paper assesses the validity of PWR, and compares PWR scores with other poverty measures: (1) per capita daily expenditures, as measured with the Living Standard Measurement Survey (LSMS) method of the World Bank, and (2) the subjective assessment of the household's poverty status by the interviewer. We examine three questions. (1) Is there any significant correlation between PWR scores, assessment by the interviewer, and LSMS-type per-capita expenditures? (2) How accurate is a calibrated PWR-tool in predicting a household being below or above the international poverty line of one-dollar a day? (3) How does the accuracy performance of PWR change if the scores are calibrated at higher different geographic levels (e.g. district instead of community)? Using a village census in 8 villages located in four of the five divisions of Bangladesh, 1660 households have been scored using the PWR method. A randomly selected sub-sample of 320 households (40 households in each village) were interviewed with an LSMS-type questionnaire. The data allow identifying households that have per-capita expenditures below 1 dollar a day at purchasing power parity rate. Our results show that calibrated PWR scores can achieve an overall accuracy of 70 to 79 percent, i.e. up to 8 out of 10 households are correctly predicted. As expected, the accuracy is higher if the PWR scores are calibrated at lower administrative level, and highest if calibrated at the community level. For the case of Bangladesh, the results confirm the validity of PWR as a poverty targeting method for development policies and projects if used at the community level.

Keywords: Bangladesh, measurement, Participatory Wealth Ranking (PWR), poverty, targeting

Full paper:


Contact Address: Joseph Florent Feulefack, Georg-August-University Göttingen, Institute of Rural DevelopmentAlbrecht-Thaer-Weg 12a/08, Göttingen, Germany, e-mail:
Andreas Deininger, November 2005