When Banks Leave Town: Deposit-Driven Branch Consolidation and Its Local Economic Consequences

Abstract

A cornerstone of banking infrastructure, U.S. bank branches have been closing at a rapid pace after decades of expansion. This paper studies how banks choose which branches to close and whether these closures affect local real economic activity. First, using branch-level deposit data and small business lending data, I show that although physical branch presence is strongly associated with deposit collection and loan origination, closure decisions are driven primarily by deposit- rather than lending-performance. Second, to estimate the effects of closures on local business dynamism, I use post-merger overlap between acquirer and target branch networks to produce likely exogenous variation in branch closures. The effects of a closure are economically significant: closing all branches in a tract (about two on average) reduces the business entry rate by one-fourth of a standard deviation. I propose that these effects are driven by reduced credit supply following the loss of local soft-information production by branches. Consistent with this mechanism, the effects of branch closures are stronger for smaller, more opaque firms, weaker during credit booms when screening is less important, and weaker in areas with greater FinTech penetration, providing preliminary evidence that technology can help offset the role of physical branches in supporting local economic activity.