Is There Statistical Evidence that the Oregon Payday-Loan Rate Cap Hurts Consumers?

1. A recent unpublished manuscript whose conclusions were widely circulated in the electronic media (Zinman, 2009) claimed that Oregon 2007 payday loan (PL) rate-limiting regulations (hereafter, “Cap”) have hurt borrowers. 2. The report’s main conclusion, phrased in cause-and-effect language in the abstract - “…restricting access caused deterioration in the overall financial condition of the Oregon households…” - relies on a single, small-sample survey funded by the payday-lending industry (PLI). The survey is fraught with methodological flaws. 3. Moreover, survey results do not support the claim that Oregon borrowers fared worse than Washington borrowers, on any variable that can be plausibly attributed to the Cap. 4. In fact, Oregon respondents fared better than Washington respondents on two key variables: on-time bill payment rate and avoiding phone-line disconnects. On all other relevant variables they fared similarly to Washington respondents. In short, the reported claim is baseless.

💡 Research Summary

The paper provides a systematic critique of Zinman’s 2009 unpublished manuscript, which claimed that Oregon’s 2007 payday‑loan interest‑rate cap (“the Cap”) caused a deterioration in borrowers’ overall financial condition. The authors begin by outlining the policy context: Oregon imposed a statutory ceiling on payday‑loan APRs in 2007, while Washington, with a similar demographic profile but no cap, served as a comparison state. Zinman’s analysis relied on a single, industry‑funded survey of roughly 300 borrowers, employing a difference‑in‑differences (DiD) framework to compare pre‑ and post‑Cap outcomes between the two states.

The critique focuses on several methodological flaws. First, the sample size is small and non‑random; the survey was financed by the payday‑lending industry, raising concerns about selection bias because respondents were likely recruited through channels favorable to lenders. Second, the study suffered a high attrition rate (about 30 %), yet the handling of drop‑outs was not transparently reported. If financially distressed borrowers were more likely to drop out, the remaining sample would be biased toward better‑off participants, inflating any apparent “no‑harm” findings. Third, the DiD approach did not adequately control for concurrent macro‑economic shocks—most notably the 2008 financial crisis—and for state‑specific trends unrelated to the Cap. Consequently, the estimated treatment effect conflates the Cap with other unobserved factors.

Fourth, the key outcome variables (on‑time bill payment, phone‑line disconnections, income changes, debt levels) were all self‑reported and not cross‑validated with administrative records, introducing measurement error and potential social‑desirability bias. Fifth, the authors point out that the language used in Zinman’s abstract (“restricting access caused deterioration…”) implies causality that the data cannot support; the analysis merely shows correlation.

Empirically, the re‑analysis presented in the paper shows that Oregon borrowers actually performed better than Washington borrowers on two crucial metrics: the rate of on‑time bill payments and the incidence of phone‑line disconnects. On all other examined variables—total debt, income volatility, and self‑reported financial stress—there were no statistically significant differences between the states. Thus, the evidence does not substantiate the claim that the Cap harmed consumers; on the contrary, it suggests that the cap may have had neutral or even modestly positive effects on certain financial behaviors.

In conclusion, the authors argue that Zinman’s claim lacks statistical support due to flawed sampling, inadequate control of confounders, high attrition, and reliance on self‑reported data. They call for future research employing larger, randomly selected samples, robust longitudinal designs, and objective administrative data to properly assess the impact of payday‑loan regulation on borrower welfare.