Microfinance institutions (MFIs) which provide financial services in developing economies are hybrid enterprises in the sense that they balance social and financial objectives. It is therefore not surprising that empirical researchers have studied characteristics, conditions and managerial practices related to increased social or financial performance1 or sometimes both.2 While the financial dimension of an MFI is straightforward to quantify given the abundance of accounting-based and comparable metrics, the social performance is harder to grasp. Most empirical studies resort to the framework of Schreiner (2002) who distinguishes the breadth of outreach referring to the number of poor clients reached from the depth of outreach referring to the poverty-level of the clients. Consequently, most studies use the (growth in) number of clients as proxies for outreach-breadth and average loans and proportion of vulnerable clients (female clients and rural clients) as proxies for outreach-depth. There is, however, discussion whether these MFI-level univariate proxies are reliable indicators to capture one or more dimensions of social performance, which is an inherently complex and multidimensional concept. Interestingly, the concept of social performance is viewed very differently by researchers and practitioners. Whereas practitioners advocate an encompassing view involving several institutional layers, empirical researchers typically zoom in on a single institutional outcome. It remains to be seen whether both views are reconcilable and offer complementary insights or whether they are discordant.
Browse by title
Microfinance (MF) has been long heralded as a silver bullet to alleviate poverty and empower women. It enables the poor to access low cost credit, from which they were previously excluded, to undertake profitable entrepreneurialism. However, the impact of MF has become a highly contested arena, with critical (often systematic) reviews tending to controvert earlier claims that there existed good evidence of highly beneficent impacts (for example Bateman, 2010; Roy, 2010; Stewart et al., 2010; Duvendack et al., 2011; Roodman, 2012). Until the late 1990s most of the evidence was based on either practitioner contracted qualitative evaluations and commentaries (for example Counts, 1996; Todd, 1996), or a few quantitative studies that did not adequately control for selection bias (for example Gaile and Foster, 1996; Sebstad and Chen, 1996). A widely acknowledged notable exception was the complex analysis of a quasi-experimental design by Pitt and Khandker (1998) which became known as the most convincing evidence in favour of the headline claims for MF. Confidence in the likely positive impacts of MF was boosted by theoretical model based “typical findings” derived from empirical studies; these models purported to explain the causal mechanisms embedded in the innovative practices of microfinance institutions (MFIs) which enabled them to reach the poor and maintain financial viability (for example Varian, 1990; Besley and Coate, 1995). If it is the case that this combination of qualitative and quantitative studies backed up by convincing theoretical modelling was misleading, a better understanding of how best to measure MF impact is warranted to assess the quality and rigour of the empirical evidence to date. This chapter begins with a chronological overview of the key microfinance impact evidence with a focus on the methodological approaches adopted and how these evolved over time. Next, MF-specific evaluation challenges are discussed and linked to the recent methodological debates on experimental versus quasi-experimental studies. Finally, this chapter concludes with reflections on what we may have learnt from the discourse on MF impact.