Complements integration and foreclosure: the case of joint consumption.

• Author: Garmon, Christopher

1. Introduction

   In 1995, the Comdata Corporation purchased Trendar, a supplier of fuel desk point-of-sale (POS) devices specializing in sales to truck stops. (1) At the time, Comdata was the dominant provider of fleet card services to the long-haul, irregular route trucking industry. Fleet cards and fuel desk POS devices are used in conjunction to facilitate the sale of diesel fuel between truck stops and truckers. Fleet cards are similar to credit cards but have additional features that enable trucking companies to protect against fraud by restricting what their drivers buy and where they buy it. Fuel desk POS devices process fleet card transactions and often provide other services to truck stops and gasoline stations (e.g., controlling the gas/diesel pumps). Soon after the merger, competitors of Trendar, such as Flying-J/ROSS, complained that they were having difficulty processing Comdata fleet card transactions. In 1999, the Federal Trade Commission signed a consent order with Comdata to allow fleet cards access to the Trendar machine and to allow other POS systems access to ComChek cards. (2)

   At first glance, the acquisition of Trendar by Comdata seems to be yet another example of a dominant firm integrating into a complementary market and foreclosing competitors in this market. With the recent antitrust trial pitting the Department of Justice against Microsoft, the Telecommunications Act of 1996, which set guidelines for local exchange carrier entry into long-distance telephony, and the many recent high-profile vertical mergers (e.g., AOL-Time Warner), there has been a renewed interest in the economic effects of vertical and complementary product integration. A number of recent articles have investigated the incentive of a dominant firm to integrate into a complementary market and foreclose rivals and the associated welfare effects of this foreclosure (e.g., Economides 1998; Riordan 1998; Sibley and Weisman 1998; McAfee 1999). While the results of these studies are mixed, the general message of the recent literature is that vertical integration can reduce welfare if the integrating firm is dominant in its original market. (3) This contrasts with the traditional view that vertical integration, with some exceptions (e.g., regulated industries and variable proportions production), is most likely beneficial if it has any welfare effect at all. (4)

   Despite the wide array of results in the literature, all of the articles analyzing vertical or complementary product integration have one assumption in common--an assumption that does not hold in the case of Comdata and Trendar. They all assume that both complementary products are purchased and used in conjunction by a particular consumer (just as a vertical relationship concerns the inputs and outputs of a particular firm). In some cases, such as fleet cards and fuel desk POS devices, complementary products are sold to different consumers or firms and are used in conjunction to aid transactions between them. This is sometimes referred to as joint consumption. While these cases are somewhat unusual, they are certainly not rare. Other examples of joint consumption include credit cards and retail POS devices, ATMs and their associated card networks, and photographic film and film processing equipment, which are used in conjunction to produce photographs. In this last example, film-processing equipment is sold to retail firms that develop film (e.g., Wal-Mart, Eckerd, etc.), while film is sold to consumers through a variety of outlets.

   In what cases does a dominant firm have an incentive to merge with the maker of a complementary product as Comdata did with Trendar? If an incentive exists, would the merger lead to foreclosure of competitors and, if so, would it increase or reduce social welfare? As with traditional vertical and complementary product mergers, there are many factors that can determine the answers to these questions, but there are two factors unique to joint consumption markets that can provide an increased incentive for a dominant firm to integrate and foreclose.

   First, entry into either half of a joint consumption market (or both simultaneously) requires a level of coordination that is not necessary for entry into standard complementary product markets. As shown by Aghion and Bolton (1987), Nalebuff (1999), Choi and Stefanadis (2001), and Carlton and Waldman (2002), the parties involved in a standard complementary products merger may have an incentive to create a closed system where the products of each cannot be used by competitors. To successfully enter when facing a dominant closed system, the entrant must provide both complements simultaneously (i.e., the entrant must enter with a competitive system of complements). As shown in the literature, this can be difficult when network effects or diseconomies of scope are present. In joint consumption markets, this simultaneous entry might be nearly impossible when facing the closed system of a dominant firm because the complements must be successfully introduced simultaneously to two separate types of consumers who use the products together. Returning to the trucking industry example, what truck stop would install a POS device incompatible with the cards used by most truckers? Likewise, what trucking company would use a fleet card incompatible with the POS devices in most truck stops? This catch 22 was the impetus for the FTC's requirement that Comdata open its fleet card and POS device system to competitors. (5)

   The second factor that provides an increased incentive for a dominant firm to vertically integrate and foreclose in joint consumption markets is the potential ability to price discriminate over consumers of the complementary product (referred to hereafter as the new consumers). This is the focus of the current article. To isolate this factor, which is unique to joint consumption markets, it is useful to abstract from the many factors that can lead to an incentive to integrate and foreclose in standard complementary products markets, including the increased barriers to entry of a closed system as discussed above. For instance, in this article, I assume the merger will not result in any cost savings on the margin. To abstract from the reduction of the double margin, it is assumed that one product is supplied monopolistically while the other is supplied competitively. (This also roughly matches the situation in fleet cards and fuel desk POS devices before the Comdata/Trendar merger). Finally, to facilitate comparisons with the previous literature, it is assumed that the complementary products are always used together in a one-to-one ratio. Under these circumstances, it has been shown on numerous occasions that a monopolist of one product has no incentive to integrate into the sale of the other product if they are both sold to the same consumer. Even if the monopolist merged with one of the complementary product retailers and foreclosed other retailers, the monopolist could not increase its profits because the consumer only cares about the combined price of the products.

   Does this result change if the complementary products are sold to different consumers to aid in transactions between them? The answer depends on whether the monopolist has the ability to price discriminate in the previously competitive market for the complement. If a uniform price must be charged in both complementary product markets (e.g., because of arbitrage), then a monopolist of one does not gain by integrating into the other market and foreclosing other retailers to gain another monopoly. The reason is that a price increase on one set of consumers will be passed on to the other set of consumers via the transactions between them. (For instance, if Comdata were to raise the price of Trendar services, truck stops would pass this price increase onto truckers through the fuel price.) However, a monopolist can gain by integrating into the other market and foreclosing other retailers if price discrimination is feasible in the second market (e.g., if Comdata could charge a price for the Trendar machine that varies nonlinearly with its use). With quantity-dependent prices, the monopolist can capture rents from the new consumers without reducing the quantity demanded by the monopolist's current customers. In fact, the monopolist has an incentive to lower the price charged to his current customers in order to increase the rents that can be gained in the complementary market. Thus, the monopolist's current customers are made better off when the monopolist integrates and forecloses to gain a monopoly in a complementary market with different customers. These new customers are obviously made worse off, but the overall effect on welfare is less clear. In many, but not all cases, overall...