Gender, Credit Risk and Performance in Sub-Saharan African Microfinance Institutions

ABSTRACT

The involvement of women in business in developing countries has become a subject of great interest for many researchers. In particular, female involvement in microfinance institutions has received special attention from governments and development institutions given its potential impact on poverty alleviation. This paper assesses the effect of gender on the credit risk and performance of microfinance institutions in sub-Saharan Africa. A sample of 43 microfinance institutions from 19 sub-Saharan African countries was selected and data was collected over the period 2010–2016. Seemingly unrelated regressions (SURs) were performed to examine how gender affects the credit risk and performance of microfinance institutions. The findings do not show any significant impact of female loan officers on credit risk, financial performance or social performance. Thus, all else being equal in the countries analyzed, female loan officers do not impact the credit risk and performance differently compared to male credit officers. The contribution of this paper is to shed light on the debate on the impact of gender on the performance of microfinance institutions.

KEYWORDS:

• Microfinance
• gender
• performance
• credit risk
• sub-Saharan africa

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Different reasons are usually advanced to explain why the gender of agents with decision-making power may influence the outcome of economic transactions. Relative to risk-taking or credit risk, a non-exhaustive list includes differences between men and women with respect to risk-aversion, overconfidence, social preferences, tolerance for inequality, negotiation skills, information processing, experiencing of emotions, competitiveness and career patterns. Due to space limitations, we do not review all these reasons here. See Bellucci, Borisov, and Zazzaro (2010b) for a description and literature review about these reasons.

2 We thank an anonymous referee for suggesting the addition of this short sub-section.

3 Even though we use 43 MFIs over the period 2010–2016, we are far from having 301 observations due to missing values. Only rows with valid data on all variables are kept. For many countries, due to missing values in one or the other variable used, we end up with just one observation instead of 7. For example, if for ROA, we have observations from 2010 to 2016, we are supposed to have 7 relevant observations. If the independent variable SIZE has missing observations for years 2014 to 2015, two rows are deleted (leaving only 5 observations for this institution). Furthermore, if another independent variable DEPOSIT has missing observations for years 2010 to 2013, four rows are deleted (leaving only one valid observation for this institution).