Platform Pricing in Matching Markets

Abstract

This paper develops a simple model of monopoly platform pricing accounting for two pertinent features of matching markets. 1) The trading process is characterized by search and matching frictions implying limits to positive cross-side network e¤ects and the presence of own-side congestion. 2) Matched agents bargain over the division of the match surplus depending on the qualitative characteristics of both parties. We …nd that, compared to the frictionless benchmark typically analyzed in the classic platform pricing literature, the harms of monopoly market power are mitigated by frictions. However, when the platform is allowed to make investments in the reduction of frictions, a private platform is likely to under-invest compared to a Pigouvian platform. In addition, accounting for user quality di¤erentiation further reduces classic harms of monopoly market power when user quality types are complements in creation of the match surplus. In this case it is socially desirable to attract smaller groups of users with higher average quality to maximize the aggregate match surplus, resulting in a downward price distortion. This result is reversed when quality types are substitutes and the distortion disappears when they are strategically independent.

Managing competition on a two‐sided platform

“…For analysis with general distribution functions, see Hagiu (). See also Goos, Van Cayseele, and Willekens () for a related analysis of a monopoly platform matching two groups, in the presence of positive cross‐group and negative within‐group external effects.…”

For-Profit Search Platforms

“…Caillaud and Jullien () and Goos et al. () do provide a random‐matching microfoundation, but gains from trade are public information and always positive, whereas in our analysis the focus of attention is on the heterogeneity of agents' private types, private information bargaining, and the platform's strategy to avoid a breakdown of bargaining in equilibrium . Rochet and Tirole () mostly deal with reduced form utility functions, but also consider an extension with a random‐matching microfoundation with two‐sided private information (see Rochet and Tirole, , pp.…”

“…Goos et al. () additionally assume nontransferable utility in the main part of their paper, in which they analyze participation, per match, and transaction fees. In an extension, they consider transferable utility, but only participation fees.…”

A Business Model Framework to Characterize Digital Multisided Platforms

“…Such interdependencies often lead to a feedback loop in which the number of participants on both sides affect each other recursively. [3,4,40,50,51] Pricing…”

“…Moreover, platform managers can choose among different kinds of fees to be applied. [3,10,44,51] Integration and control Platforms involve several users on each side. It is important to control the behavior of the users participating in the platform and contents provided.…”