In "Do Defaults on Payday Loans Matter?" author Robert Mann uses a difference-in-difference regression-based analysis to analyze "harm" in the payday lending market. He finds that little difference in changes in credit scores between payday defaulters and non-defaulters and uses this as evidence that payday loans do not cause harm. However, this study suffers from significant conceptual and technical flaws.
Even without these flaws, Mann is testing an incoherent hypothesis. Any harm to credit scores would be caused, not by default, but by the financial drain resulting from paying the exorbitant fees charged by payday lenders. This harm would be incurred most acutely by the payday borrowers who pay multiple fees- these borrowers may or may not be the defaulters. Furthermore, even if credit score is serving as a crude proxy for the financial distress caused by an inability to repay a payday loan, Mann's model is flawed because he assumes inability to repay always leads to default when, in fact, more often than not it leads to serial loan refinancing (i.e. "rollovers").