Testing for Complementarity between Stores and E-Commerce

Abstract

We empirically investigate the relevance of demand-side complementarity between electronic and traditional provision of banking services. Since no systematic data on prices for the two types of service is available, it is not possible to estimate cross-elasticities of demand. We resort to two indirect tests. The first test is based on estimating the relationship between branches and the diffusion of e-banking services in local markets, controlling for individual bank and market characteristics. We test three alternative hypotheses on the relationship between the two delivery channels: complementarity, substitution or segmentation. We find that banks expanded relatively more in the e-business in those local markets where they had relatively fewer branches, with the exclusion of markets where the banks were chartered. The second test is based on measuring the impact of the joint provision of banking services - electronically and at traditional branches - on banks' revenues per customer. We estimate a non-standard revenue function that relates revenues from asset management, brokerage and payment services to the share of customers employing e-banking, given the total number of bank customers. Our results show that a high share of e-customers is associated with a reduction in revenues per customer. This evidence is not consistent with complementarity because banks did not extract substantial consumer surplus from the joint provision of electronic services and traditional services at the branch.