The Borrowers Who Lose Access

The subject here is the bottom risk-score decile in the states that adopted consumer rate caps. The specific question is how much their realized credit changed, how quickly the change appeared, and whether losing access to high-cost borrowing improved their measurable financial stress. Keeping those questions separate matters: a policy can remove an expensive loan, reduce the number of open accounts, and still fail to improve the repayment outcome used to assess borrower harm.

A persistent contraction in balances and accounts

The New York Fed study identifies high-risk borrowers as households in the bottom decile of credit scores before a state’s limit took effect. It then compares their outcomes with those of other borrowers in the treated setting and with corresponding households in states that did not adopt a consumer usury limit during the study period. The comparison is not a claim that every household below a score threshold was denied a loan. It estimates whether the bottom group moved differently after the policy than borrowers who were less exposed to a ceiling that could bind at the margin.

The result is a substantial contraction in measured credit. Table 4 reports a triple-difference coefficient of -$1,669.70 for debt, against a reported outcome mean of $11,308.04. Dividing the coefficient by that mean gives a decline of about 14.8 percent. The body of the report describes the result as a 14.7 percent decline, while the abstract and introduction use 16.9 percent. Those presentations do not support false precision. The defensible reading is that debt fell by roughly 15 percent in the table-based comparison, or approximately 15–17 percent depending on the report’s presentation.

The account result is similarly meaningful. The reported coefficient is -0.65 accounts against a mean of 3.30, which is about a 19.7 percent reduction and is described in the paper as a 19.6 percent decline. This is not just a smaller balance on the same set of accounts. It indicates that high-risk households had fewer open credit relationships after the limits took effect. Some borrowers may have closed accounts, while others may have opened fewer accounts or failed to renew a relationship that would previously have continued. The data cannot distinguish each of those paths, but they all represent a contraction on the extensive margin of measured credit.

That distinction is important for interpreting a rate cap. A balance can fall for benign reasons: a household can pay down debt, borrow less because a need has passed, or shift to a source outside the measured category. A fall in account counts is more directly consistent with fewer available or active credit relationships, although it still does not reveal whether a lender rejected an application, whether the borrower chose not to apply, or whether another source replaced the account. The paired movement in balances and accounts makes the high-risk result harder to explain as only a change in repayment timing or average loan size.

The timing strengthens the interpretation. In the event study, debt falls by close to $1,000 per borrower immediately after the cap is enacted and the reduction approximately doubles over the next five quarters. Account counts fall by more than 20 percent of the sample mean over the five quarters after the law. The pattern is therefore persistent and develops beyond the first post-policy observation. If the result were only a short-lived reporting disturbance or a one-quarter pause in originations, the later path would look different. Instead, the measured contraction continues.