Scientific Publication

An operational tool for evaluating poverty outreach of development policies and projects

Abstract

Development institutions and projects frequently seek to target poorer segments of the population. Yet, existing methods for evaluating their outreach are generally unsuited to most operational settings, since they are either too costly and cumbersome (e.g., detailed income or household surveys), or they produce results that are not comparable between villages or regions within a country (e.g., participatory poverty appraisals). This paper presents a new and operationally suitable method to measure the poverty of clients of development projects in relation to the general population of nonclients. The method was developed in response to demands by donors and development practitioners for a low-cost evaluation instrument that could be used as a regular operational tool for assessing the poverty outreach of a development project or institution. While the method was originally developed for the purpose of assessing the poverty outreach of microfinance institutions (MFIs), we believe the method can be used for any development policy or project that pursues an explicit objective of reaching poorer people. The paper begins by discussing existing methods of poverty assessment. Next, the paper presents heuristic steps for identifying indicators of poverty to be tested in the case studies, including the questionnaire that was field tested in four countries with large differences in poverty-level, socioeconomic, and cultural contexts, and with MFIs that worked either in urban, rural, or mixed areas with different target clientele and financial products. The authors then describe the method of principal component analysis used to construct a poverty score as the measure of relative poverty. The paper concludes with a summary of results from four country case studies (two in Sub-Saharan Africa, one in South Asia, and one in Central America).