Vice President, Salvatore Piccolo, alongside professors Gülen Karakoç and Marco Pagnozzi, wrote an article for the International Journal of Industrial Organization. The article investigates incentives and anticompetitive practices of incumbents in vertically related markets, as well as the effects of information sharing on consumer surplus and total welfare.
A manufacturer designs a long-term contract with a retailer who is privately informed about demand, and they face future competition by an entrant. When demand is correlated across periods, information about past sales affects firms’ behavior after entry. We analyze the incentives of the incumbent players to share this information with the entrant and show that the manufacturer and the retailer have contrasting preferences: when the retailer wants to disclose information, the manufacturer does not, and vice versa. Although transparency harms consumers and reduces total welfare, incumbent players jointly benefit from selling information to the entrant.
In vertically related markets, incumbents facing the threat of future entry may engage in anticompetitive practices to protect their market power — e.g., exclusive dealings, limit pricing, most-favored-nation clauses, and other forms of vertical restraints. Incumbents may also use information disclosure as a strategic tactic to discipline future competitors. However, although information sharing among rivals has been extensively studied in static oligopoly models (see, e.g., Vives, 2006, for a survey), little is known on firms’ incentives to share information in dynamic environments, where incumbents may strategically disclose or hide information to potential entrants. Even less is known about the interplay between these incentives and vertical contracting.
We analyze a dynamic vertical contracting environment in which a manufacturer interacts with an exclusive retailer for two periods. In the first period, the incumbent firms are monopolists, while in the second period they face competition by an integrated entrant. The incumbent firms may share information about their past sales with the entrant. In the baseline model we assume that firms compete à la Cournot by selling a homogeneous product whose demand is uncertain and, in every period, is privately observed by the downstream players — i.e., by the retailer in both periods and by the entrant in the second period. The manufacturer designs a long-term contract to elicit the retailer’s private information, specifying the quantity that the retailer distributes and the transfer that it pays to the manufacturer in every period. Since demand is correlated over time, the retailer’s second-period production depends on its report about demand and on its production in the first period (that the manufacturer uses to update its beliefs about demand) — i.e., the optimal dynamic contract features memory (e.g., Baron, Besanko, 1984, Laffont, Tirole, 1996, Battaglini, 2005 among others).
Within this setting we examine the following questions. Do incumbents want to share their private information with the entrant? How is information sharing affected by vertical contracting? What is the price of information? What are the effects on final consumers?
We show that firms’ profits depend on the entrant’s information about first-period production. One may wonder why the entrant should care about this information, since it observes demand in the second period, when it chooses production. The reason is that, because the entrant does not observe demand in the first period, past production conveys information on the agency (organizational) costs of the incumbents and hence on the retailer’s production in the second period (i.e., its future competitive conduct), thus affecting firms’ choices and market competition post entry.1 In other words, the long-term interaction between the incumbent players creates a contractual link between periods, and the role of information sharing hinges on this link.
Our main result is that the manufacturer and the retailer have diverging incentives to share information about past production (or sales). When the downstream firm is willing to disclose this information to the entrant, the upstream firm does not want to do so, andvice versa. The reason is that the manufacturer’s profit and the retailer’s information rent are affected by the entrant’s production in opposite ways. Moreover, the incentive to share information depends on the degree of demand uncertainty, which reflects the relevance of information asymmetry. Specifically, we find that when uncertainty about demand is small, the retailer wants to inform the entrant about its first-period production, while the manufacturer has no incentive to do so. The opposite incentives arise when uncertainty about demand is large: the manufacturer wants to disclose first-period production, while the retailer has no incentive to do so.
Information sharing has two effects on the incumbent firms’ payoffs. First, disclosing information decreases their joint profit. The reason is that, when the entrant knows that demand was low in the first period, it knows with certainty that its downstream rival will sell a low quantity in the second period (due to the distortion needed to minimize its intertemporal rent). Therefore the entrant becomes aggressive, thus reducing the equilibrium price and, in turn, the total profit that incumbent firms can share. The second effect of disclosing information has to do with the retailer’s information rent and can be decomposed as follows. On the one hand, facing an aggressive (informed) entrant in the low-demand state makes it less attractive for the retailer to misrepresent a high state as a low one — i.e., the manufacturer can use the entrant to ‘punish’ its retailer for misrepresenting demand. Hence, other things being equal, the manufacturer would like to face an informed entrant. On the other hand, however, when the entrant is very aggressive (i.e., because it is informed), it is more difficult for the manufacturer to control its retailer’s incentive to misreport demand through the distortion of output: a high distortion triggers a strong reaction by an informed entrant, which reduces the equilibrium price and, in turn, the upstream incumbent’s profit.
The degree of demand uncertainty affects these effects as follows. When uncertainty about demand is high, the retailer’s information rent is high, hence the second effect makes it worthwhile for the manufacturer to share information in order to decrease the retailer’s information rents, even though doing so reduces revenues. For the same reason, the retailer does not want to face an informed entrant. By contrast, when uncertainty about demand is low, facing an informed entrant is costly for the manufacturer because it cannot distort the quantity in the low-demand state too much since the informed entrant would respond aggressively. Of course, this effect benefits the retailer, who therefore prefers to share information and face lower distortions. Clearly, for intermediate demand uncertainty, the effect on revenues dominates the effects on rents for the manufacturer, so that both incumbent firms do not want to share information.
Interestingly, although the two incumbent firms never agree on whether to disclose information to the entrant at no cost, the entrant is always willing to pay a sufficiently high price to induce incumbents to jointly sell such information. The reason is that information sharing maximizes total firms’ profits in the market. Therefore, a market for information may arise in our environment.2
The effects of information sharing on consumer surplus and total welfare depend on its impact on the overall efficiency of the industry. Sharing information reduces the incumbent’s production because it induces the entrant to increase production when the incumbent distorts it for rent extraction reasons. On balance, however, aggregate production is lower with information sharing because, holding constant the incumbent’s production, the entrant’s production decision is always efficient regardless of its information (since the entrant equalizes marginal revenue to marginal cost). In other words, although information sharing rebalances production between the incumbent and the entrant, it reduces market efficiency because it increases the retailer’s rent. As a result, consumer surplus and welfare are always lower with information sharing in our model and, contrary to what is commonly believed, a welfare maximizing policy should reduce transparency and forbid incumbents to disclose information to entrants.
Hence, in our environment, the emergence of a market for information always harms consumers, because it always induces incumbents to share information with entrants. Prohibiting a market for information, however, is not sufficient to protect consumers. In fact, our analysis suggests that welfare depends on the proper assignment among incumbents of property rights on information about past contractual arrangements and sales data, since this determines whether information is shared or not (in the absence of a market for information).3
Our results are of interest for competition policy because they challenge the traditional conjecture that sharing ‘historical’ information is consumer-welfare neutral (as opposed to exchanges of current information) because it does not reveal future competitor strategies. The new channel analyzed in our paper suggests that, since long-term contracts feature memory, sharing historical sales data may indeed harm consumers, and more so when demand uncertainty is persistent. Interestingly, this contrasts with some regulatory trends intended to promote transparency and level paying field competition. Several national and international regulatory agencies are considering implementing new codes of conduct based on transparency.4
Finally, in the online Appendix, we consider price competition with differentiated products and show that, in this case, transparency may actually increase consumer surplus and welfare. The reason is that prices are strategic complements and information sharing allows the entrant to react more accurately, thereby enabling the entrant to charge, on average, a lower price in equilibrium. Interestingly, in contrast to what happens under quantity competition, sharing information benefits consumers but does not always benefit the entrant because it reduces expected prices. On the whole, information sharing under price competition makes it more costly for the manufacturer to elicit information from its retailer, which lowers the entrant’s retail prices and tends to increase market efficiency.
Read the full article
This article was originally published by the International Journal of Industrial Organization and also available on ScienceDirect. The views expressed are those of the authors only and do not necessarily represent the views of Compass Lexecon, its management, its subsidiaries, its affiliates, its employees, or clients.