Research

Merchants rarely price discriminate by payment method. This enables card networks like Visa to fund consumer rewards with merchant fees. I develop and estimate a model of platform competition to compare how regulation, increased network competition, and public entry affect U.S. consumer payments. I find that the current U.S. payment system is regressive and inefficient, and that network competition exacerbates these problems. I model consumer adoption and merchant acceptance of multiple cards, merchant pricing, and network competition. I estimate the model by matching data on the effects of debit rewards reductions and a large grocer’s decision to accept credit cards. The estimated model matches external evidence on networks’ costs, merchants’ margins, and the effects of AmEx’s OptBlue program on merchant acceptance. Uniform caps on merchant fees are progressive and increase annual welfare by $31 billion by reducing rewards and credit card use. Because higher income households are more likely to use credit cards, fee caps save the median household around $50 every year but cost high-income households $1000. Because consumers are reward-sensitive, but merchants are fee-insensitive, competition has the opposite effects. Few consumers adopt public options without rewards, limiting their benefits.

We study the equilibrium consequences of differences in mortgage shopping behavior between majority and minority borrowers. We identify minority-specialized lenders, who disproportionately lend to minority borrowers and originate one-fifth of minority mortgages. These smaller lenders charge high mortgage rates and borrowing from them is partially responsible for the minority interest rate gap. Minority-specialized lenders are more likely to employ minority employees and have higher market shares in areas with more non-English speakers.  Borrowers are also less likely to withdraw mortgage applications from these lenders. These facts suggests that minority-specialized lenders provide costly minority-specialized services, rather than discriminate against these borrowers. To quantify the effect of minority-specialized service provision in equilibrium, we estimate a model in which minority specialized lenders compete with mainstream lenders, and majority and minority borrowers differ in loan demand as well as the types of lenders they consider. Our novel identification strategy uses withdrawn mortgage applications to separately identify borrower consideration sets and preferences. The estimated model can rationalize the minority gap in rates as well as consideration set size across groups. Minority-specialized lending attracts minorities by providing services valued by minorities, and by lowering search frictions. Minorities gain from a broader diffusion of minority-specialized lending, and these gains are large relative to potential gains from eliminating residual racial discrimination in interest rates. Our model suggests fair lending laws can disincentivize mainstream lenders' investments in minority-specialization, reducing competition and welfare for minority borrowers.

Financial frictions can overturn conventional antitrust analysis of startup acquisitions. I extend Myers-Majluf to include the option to be acquired. Low types are acquired, medium types issue equity, and high types do not invest. Blocking acquisitions lowers the average type of equity issuers and raises the cost of capital for standalone startups. The welfare loss from lower investment can overwhelm the welfare gains from blocking anticompetitive acquisitions. A case study from the pharmaceutical industry suggests antitrust policy can have a large effect on the valuations of startups who are unlikely to be acquired for anticompetitive reasons.