Do Bank Branches Still Matter? The Effect of Closings on Local Economic Outcomes (Job Market Paper)
Abstract: I study the relationship between bank-specific capital and credit access in a new setting: bank branch closings in markets where the branch network is dense. Existing regulation in the U.S. is targeted toward areas with few branches where closings inhibit physical access to the branch network. I show that, even in crowded markets, closings can have large effects on local credit supply. To generate plausibly exogenous variation in the incidence of closings, I use Census tract level data paired with a novel identification strategy that exploits within-county variation in exposure to post-merger consolidation. This instrument identifies the effect of closings that occur in close proximity to other branches. I find that closings have a prolonged negative impact on credit supply to local small businesses, but only a temporary effect on local mortgage lending. The number of new small business loans is 13% lower for several years, and this decline persists even after the entry of new banks. The decline in lending is highly localized, dissipating 8 miles out, and is concentrated in low-income and high-minority neighborhoods. These results show closings have large effects on local credit supply when lending is information-intensive and lender-specific relationships are difficult to replace. I provide a framework for discussing the welfare implications, which depend on the characteristics of the marginal borrower.
Do Credit Market Shocks Affect the Real Economy? Quasi-Experimental Evidence from the Great Recession and 'Normal' Economic Times (with Michael Greenstone and Alexandre Mas)
Abstract: We estimate the effect of the reduction in credit supply that followed the 2008 financial crisis on the real economy. We predict county lending shocks using variation in pre-crisis bank market shares and estimated bank supply-shifts. Counties with negative predicted shocks experienced declines in small business loan originations, indicating that it is costly for these businesses to find new lenders. Using confidential microdata from the Longitudinal Business Database, we find that the 2007-2009 lending shocks accounted for statistically significant, but economically small, declines in both small firm and overall employment. Predicted lending shocks affected lending but not employment from 1997-2007.
Research in Progress
A Choice Amongst Many: Household Borrowing in a Setting with Multiple Providers (with Robert Townsend)
Abstract: We seek an in-depth understanding of the choices that households and small businesses make when choosing among several sources of credit in an emerging market country, Thailand. Borrowers in this setting face a number of different financial service providers. Even within the formal and informal categories, these lenders offer contracts that differ along many dimensions including the interest rate, collateral requirement, and duration. To understand how households choose optimally amongst these different options, we make use of a rich monthly panel dataset that tracks income, cash flow, and balance sheet activity at the level of the individual household or enterprise, in addition to complementary annual data with a larger geographic coverage. Within these data, we observe variation by region in terms of the level of wealth and industrialization, urban and rural status, and distance from financial providers, which generates heterogeneity in the set of potential lenders faced by different households. We also exploit various supply-side regulations, including the introduction of a large government microcredit program, that induce plausibly exogenous variation in credit supply over time and across villages. Combining these with data on each household's realized portfolio of loans, we estimate a model of household decision-making among multiple providers and explore policy implications.