Consumer Valuation of Network Convenience: Evidence from the Banking Industry

This paper develops a structural demand model for US retail banking services in which consumers have preference over the geographical convenience of their banks' networks. The model accounts for the fact that (i) consumers spend time in both home and workplace and may need to access their bank's services at any time of the day, and (ii) consumers do not consider services provided by different bank networks as perfect substitutes, therefore prefer to use branches of their own banks. I estimate the model using a detailed nationwide worker flow dataset that tracks consumers by residence and workplace. My results confirm that consumers value bank networks for their proximity to multiple locations. I conduct a counterfactual experiment to quantify the effect of this demand motive on bank deposits. After breaking the compatibility of services among branches within the same network, choosing a multi-branch bank no longer necessarily reduces consumers' expected travel distance, which takes away the potential gain from larger networks. For the average city and bank in the sample, each additional branch implies an increase of 0.1 percentage points in the extra market share generated from the network effect, which amounts to roughly one million dollars in deposits. These results on firm performance, along with the corresponding consumer welfare analysis, provide valuable insights to anti-trust regulators in evaluating within-market horizontal mergers in the presence of spatial network competition.