Staff rotation in Microfinance - a field experiment

Individual behavior is largely driven by incentives, yet those faced by employers and employees are often misaligned - the classic principal-agent problem. This project will study management practices in the financial sector, in particular in relation to policies for staff rotation. In lending, decisions motivated by the incentives of loan officers can make the difference between good and bad loans, with far reaching consequences both for borrowers and for market outcomes. In developing counties, where a large share of loans is extended by microfinance institutions, contact between borrowers and lenders is personal and frequent. This leads to staff building relation-specific knowledge that can improve efficiency of credit activities, but these personal relationships also increase the probability of collusion occurring between staff and borrowers. Microfinance is thus a particularly relevant context for the study of staff rotation on institutional outcomes and those relating to both staff and the clientele.

The project will examine the effect of loan officer rotation on portfolio quality, as well as staff and borrower behavior, types and welfare. We partner with BRAC Microfinance in Uganda and implement a randomized field experiment affecting all group lending personnel in 106 of its local branches. Beginning in January 2019, 53 randomly selected branches rotate borrower groups between their loan officers every six months while the remaining 53 branches froze rotation during 2019. Baseline data collected in late 2018, and administrative data from BRAC will allow us to evaluate the rotation policy on key outcomes. The ISF funding will enable the collection of more detailed follow up data through additional surveys on both staff and borrowers, as well as the compilation of richer administrative data. 

Our project is anticipated to bridge three literatures: the literature on agency problems and contract design, a growing empirical literature studying contracts in real world labor markets, and the empirical and theoretical literature on microfinance, which has to date mainly focused on borrowers. Whereas previous works have utilized pre-existing organizational policies of rotation to draw conclusions of its impact on behavior and outcomes, our randomized experiment allows for a more convincing study of the causal effect of a rotation policy. Our primary outcomes are loan portfolio growth and quality, but our setting also allows us to collect additional data on characteristics, preferences and behavior of both staff and borrowers over time. This enables us to better isolate mechanisms at play and improve our understanding of the multiple potential effects of staff rotation. More generally, a better understanding of staff incentives and management practices in microfinance is warranted, and this study hopes to contribute in addressing this gap.