After a long and rather exhausting review process, the Sex and Credit paper has been finally accepted for publication at the Journal of Banking and Finance.
The paper and its findings fit relatively well with the recent prominence given to behavioural economics and finance (best illustrated by the Noble Prize for Richard Thaler). Using a large loan-level dataset from an microfinance bank in Albania we gauge whether
the assignment of a first-time borrower to a loan officer of the same or the opposite gender has repercussion for loan conditionality and, ultimately, for the likelihood of the borrower to return for a second loan. We find that borrowers that are assigned
to a loan officer of the opposite gender (where the assignment is a random process) are less likely to return for a second loan than first-time borrowers that are assigned to a same-gender loan officer, with one possible reason being that they have to pay
higher interest rates and receive shorter-maturity and smaller loans. Is this statistical discrimination, i.e. do opposite-gender loan officer impose more stringent loan conditionality because of higher risks in lending relationship between borrowers and loan
officers of different genders? We find that no, as there is no significant difference in the performance of such loans. So it is taste-based discrimination? Yes and no! We find that the gender effect comes exclusively from loan officers
with limited experience with opposite-gender loan officers (for all loan-officers it is the first job after college). Once loan officers gain sufficient experience with opposite gender borrowers, they overcome this initial bias. But it is not only the
experience, but also the incentive structure. The bias comes from loan officers with limited opposite-gender experience in districts where the bank faces limited competition from other financial institutions and where branches are small, i.e. where internal
competition is limited and discretion for loan officers larger. Our results are consistent with the existence of an initial gender bias and learning effects that lead to the disappearance of the bias, but also stress the importance of competition in limiting
the effects of such bias. Finally, does the bias come from male or female loan officers? While our identification strategy does not allow us to answer this question conclusively, additional estimations suggest that the bias comes from both
genders towards the respective other gender.