RT Dissertation/Thesis
T1 Social performance of microfinance institutions : theory and empirical evidence 
A1 Milan,Florence Marie
WP 2012/11/08
AB Inspired by the innovative efforts of pioneers in non-governmental organizations, notably the Grameen Bank, financial services delivery to the poor has grown since the 1980?s and has been advocated as an important component of development interventions. These financial services, also known as microfinance, are believed to have positive outcomes on production, income, and consumption at household and macro-economic levels. The provocative idea that alleviating poverty can be profitable to microfinance institutions (MFIs) and donors had an enticing effect on the private sector hence private capital sources increased. The paradigm of microfinance shifted to MFIs emphasizing on cost effectiveness and financial viability in their operations. The microfinance industry has made considerable progress in financial performance measurement and evaluation. Of late, the microfinance industry is shifting from its emphasis on financial sustainability to a renewed concern on social performance. There is heightened interest in developing a social performance measurement tool with a common set of key social indicators to go hand in hand with the financial performance. Among others, measurement of social performance will encourage institutions to be more mindful in maintaining their social mission especially now that an increasing number of microfinance programs are gearing towards privatization. It allows MFIs to demonstrate social performance, transparency and credibility which can lead to donors and investors reallocation of funds towards socially-oriented MFIs.  
Empirical data were collected while the researcher was working with the Research Department of AMK1 in Cambodia.  Quantitative panel data were collected annually from 2006 to 2008 covering AMK clients and non-clients in 55 rural villages in nine provinces of Cambodia. Qualitative data were collected from 2007 to 2009. 
The first purpose of the study is to review the important theoretical frameworks on social performance and examine ongoing social performance initiatives in the microfinance industry in view of the synthesized frameworks. The concept of social performance borrows elements from different established concepts that we find in business, ethics and society. The study shows that social performance measurement pulls together three dominant frameworks ? the corporate social performance model, the stakeholder theory, and the accountability theory- into one theme. Drawing from the review of some of the well-known social performance initiatives in the microfinance industry, the study finds it necessary to integrate both process and outcome approaches in social performance measurement. The process approach reviews how an MFI identifies, integrates, and manages its social goals. Complimentarily, the outcome approach measures stakeholder satisfaction (client, staff, etc), outreach and impact of MFI policies and financial services. The study shows that there is an overemphasis of the outcome approach on one stakeholder, the clients, and is further limited on the depth of outreach of MFIs. 
Second, the study describes and reviews a practitioner?s approach in social performance measurement with empirical analysis. AMK provided an example of how an MFI can integrate the process and outcome approaches in their operations. While most outcome approaches used by other MFIs focus on social performance to clients with emphasis on the depth of outreach, AMK takes on a broader outcome approach to its stakeholders by also devising a system to measure satisfaction of clients and staff.   AMK?s social performance measurement tools include an annual staff satisfaction survey, financial procedures and operations audit, client protection audit, client profile, depth of outreach, client satisfaction, exit client survey, and competition analysis. AMK?s depth of outreach is measured using the principal component analysis (PCA) on cross-sectional data (based on CGAP?s Poverty Assessment Tool). The PCA method creates a poverty index for each household and poverty groups (poor, less poor and better off) are created using a tercile analysis based on the poverty indexes of the control group (non-clients). Using the same method, the statistical results of the study indicate that more AMK clients fall under the poor group as compared to non-clients. The PCA method can be used by MFIS to show that they have properly targeted poor clients and to report on the depth of their outreach. The PCA method also paves the way for the measurement of poverty changes over time for individuals or households.
Inherent in evaluating social programs such as microfinance is to measure its end result and impact. Impact of microfinance on clients is an important indicator in the measurement of social performance. Therefore, the third purpose of the study is to contribute to the empirical literature on the impact of group lending on clients. Different tests and models were applied on the panel data. Thus far, little attention has been given to the time-varying effects of microfinance. The research contributes to filling this research gap by considering 1-year and 2-year intervals between panel rounds to compare group lending effects on new entrants and long-term clients. Also, many impact studies faced incomplete sample bias. This research tackles this issue by including client dropouts in the sample. 
By applying the PCA weights of the base years (2006 and 2007 cross-sectional data) to panel data collected in 2008, the study was able to identify households that moved in and out of poverty. With two time-variant poverty indexes per household, a transition matrix identified those households that are transiently poor and chronically poor. Contrary to other studies that apply PCA over pooled data, the study shows the advantage of using different set of weights on different time periods of the unpooled data. Using different set of weights accounts for the changes in poverty characteristics and the different time intervals of the panel data. The study shows that household?s movement to a wealthier group has been significant among AMK clients, notably among the chronically poor. The statistical analysis on the changes in the mean of the clothing and footwear expenditure of households shows considerable increase in AMK clients compared to non-clients among the chronically poor. Among the transiently poor, clients have significantly accumulated savings while no significant change could be found among non-clients. 
As there is weak empirical evidence on the impact of microfinance on clients, the other component of the panel data analysis was to understand the determinants of credit participation using probit regression; analyze the factors that affect the likelihood of the client household becoming chronically poor, transiently poor (worsen and improved, separately), and never poor using multinomial logistic model; and test the impact of microfinance on food consumption in rural Cambodia with different statistical models. The findings suggest that households which are more economically stable and have the ability to build assets have less demand for small loans. Households with liquid assets are less likely to borrow. Because the panel data includes new entrants and long-term clients collected at 1- and 2-year intervals, it shed light on the importance of long term participation in group lending. An important finding of the fixed effect model with interaction variables and the conditional change score model is that access to group loans has a negative impact on food consumption of new clients but a positive impact on long-term clients. The negative impact on new clients may be temporary. This finding is supported by results of a multinomial logistic regression which shows that for a long-term client the odds of belonging to the ascended transient poor group rather than the descended transient poor group are 6 percent. 
The fourth purpose of the study is to review the different social performance measurement tools in existence today, draw lessons and select common criteria for selecting a social performance measurement tool. In general, the weaknesses on the social performance initiatives lie on its comparability of results and missing analysis on some social issues. A practical measurement method should have quantifiable indicators to minimize subjectivity and establish credibility. The challenge is to quantify indicators or assign values to qualitative information. One of the starkest contrasts between the social performance tools is the design and requirements in terms of technical knowledge, time and cost. Rigorous tools will be rejected by MFIs if they see it as a financial burden and too time consuming. Practicality is important in standardizing social performance measurement tools if it hopes to gain industry-wide acceptance.
K1 Mikrofinanzierung
PP Hohenheim
PB Kommunikations-, Informations- und Medienzentrum der Universität Hohenheim
UL http://opus.uni-hohenheim.de/volltexte/2012/774