D’Amore McKim School of Business, Finance Group
Department of Economics
370 Huntington Ave.
Boston, MA 02215
Primary: Household Finance, Public Economics
Secondary: Banking, Nonprofit Organizations
“Are (Nonprofit) Banks Special? The Economic Effects of Banking With Credit Unions” [Draft]
Nonprofit banks in the U.S. are primarily organized as credit unions (CUs) and have grown steadily over the last two decades, increasing their share of total lending to U.S. households. This paper studies the economic effects of banking with CUs using consumer credit report data merged to administrative data on originated mortgages and detailed data on the locations and balance sheets of CUs. To estimate causal effects, I construct a novel instrument for banking with a CU using a distance-weighted density measure of nearby CUs. I find that banking with a CU causes borrowers to have fewer unpaid bills, higher credit scores, and a lower risk of bankruptcy several years later. I find support for several mechanisms behind these results: CUs charge lower interest rates, price in less risk-sensitive ways, and are less likely to resell their originated mortgages in the secondary market. These results are inconsistent with CUs behaving as “for-profits in disguise”, and suggest that many consumers experience better outcomes with CUs than with for-profit banks.
“Credit Information Sharing: Implications for Credit Card Competition and Household Credit Access” (with Benedict Guttman-Kenney)
We study the consequences of a credit reporting innovation that enhanced the value of credit file data to US credit card lenders. The reporting innovation allowed lenders to observe a history of repayment behavior on each credit card, as opposed to only one month’s repayment amount. This paper analyzes lenders’ incentives to share data around this innovation and the consequences of resulting changes in data sharing upon individuals’ finances.
We document that this credit reporting innovation led to the unravelling of the data sharing equilibrium, with a large fraction of US credit cards no longer reporting payment information to credit bureaus. The unraveling of information sharing is inefficient because it increases informational asymmetries between borrowers and lenders. The selection of credit card lenders who stop reporting appears consistent with them having more market power and therefore more to lose from increased competition than other lenders. Lenders who stop reporting have a portfolio of borrowers with $90 (13%) higher spending each month and 13% more likely to repay debt in full than lenders who kept reporting. Without payment information, the borrowers at lenders who stop reporting appear less profitable and riskier than they actually are.
Using a difference-in-difference design, we test the effects of this unraveling on household credit access by comparing individuals whose information stopped being reported with individuals whose information was always or never reported. While the choice of lenders to stop reporting is endogenous, it yields an exogenous source of exposure across cardholders because they cannot anticipate the event. Our preliminary results suggest that the effects of the unraveling of data sharing on credit card credit limits and household borrowing are small. Ongoing work is investigating effects on implied costs of borrowing.
“Nonprofit Price Competition in Banking” (with Jordan van Rijn)
Price theory models predict that nonprofit and for-profit firms should respond differently to changes in market power. We test this prediction in small geographic banking markets using quasi-experimental variation in market power induced by mergers of large national banks. Linking list price data on loans and deposits from S&P’s Ratewatch to the FDIC’s summary of deposits, we compare Credit Unions’ price response to for-profit banks’ response. Findings are interpreted in the context of theories of firm pricing, nonprofit behavior, and debates on tax policy.