Should Firms Share Information About Expensive Customers? An Equilibrium Analysis

Advances in information technology increasingly allow firms to identify expensive, high-cost customers, who are not only individually less profitable for firms but also raise the average marginal cost incurred by firms and thus impose a negative externality on inexpensive customers. Should competing firms share information that identifies such customers? The answer to this question has important implications for firm profitability, consumer welfare, and privacy laws that currently constrain firms’ ability to share information. Considering consumers to be heterogeneous in terms of the cost they impose on firms, this paper presents an analytical model to examine the conditions in which two differentiated Bertrand competitors prefer to share information. The firms’ incentives to share information differ according to the degree of product differentiation, the relative proportion of expensive customers in the market, the relative marginal cost of selling to expensive customers and the level of noise in the information. When firms sell substitutable products a Prisoner’s Dilemma results. The competing firms unilaterally benefit from sharing information, but the benefits from reneging on an information-sharing agreement are even higher; paradoxically, in equilibrium, both firms therefore keep their information private. A third-party agency such as an industry trade association might serve to supervise and coordinate information-sharing agreements between competing firms. In contrast, when the firms sell complementary products, they always have an incentive to share information. Importantly, this paper establishes that information sharing decreases the welfare of expensive customers but increases that of inexpensive customers. Privacy laws thus protect expensive customers more than inexpensive customers. In certain conditions, the aggregate consumer welfare might increase, when firms share information identifying expensive customers. This research recommends relatively weaker, not stronger, privacy laws, which is counterintuitive to the recommendations of the popular press.