Mobile termination and collusion, revisited

Abstract

The standard model by Laffont et al. (RAND Journal of Economics, 29(1): 1-37, 1998a; 38-56, 1998b) treats termination fees as an instrument to increase market power in a one-shot game of horizontal product differentiation. A prediction (Gans and King, Economics Letters, 71: 413-420, 2001) within this framework is that, with non-linear tariffs, firms should be interested in low termination fees. This seems to be at odds with regulatory experience in many countries. We offer an alternative approach, using an infinitely repeated Bertrand competition. We focus on symmetrical calling patterns and investigate simple two-part tariffs for two customer types, as well as general non-linear tariffs for two types and for a continuum of types. In this framework, when looking also at collusion in retail prices, termination fees make deviations from the collusive outcome less attractive. The optimum deviation strategy is usually to try to attract the high valuation customers since they exhibit the highest profits. Thus, a deviator will have a pool of heavy users which will have more outgoing than incoming calls, implying net termination payments. A cooperatively chosen termination rate can increase the deviator's cost and thereby always stabilizes collusion.