Expert economists Enrique Andreu, Damien Neven and Salvatore Piccolo recently co-authored a paper for the International Journal of Industrial Organization on the link between manufacturers’ incentives to enter an information-sharing agreement and the strategies according to which they delegate price authority to their downstream retailers.
In this paper, the authors offer new insights into how competing manufacturers delegate price authority to their privately informed retailers and how an information-sharing agreement affects such decisions, equilibrium prices, and consumer surplus.
We characterize the degree of price discretion that two competing manufacturers grant their retailers in a framework where demand is uncertain and privately observed by the retailers, while manufacturers only learn it probabilistically. In contrast with the consolidated vertical contracting literature, we assume that manufacturers cannot use monetary incentives to align the retailers’ incentives to pass on their unverifiable distribution costs to consumers. Our objective is to study how, in this context, an information-sharing agreement according to which manufacturers share their demand information affects prices, profits and consumer surplus. While equilibria with full price delegation never exist, regardless of whether manufacturers share information, partial delegation equilibria may exist with and without the exchange of information. These equilibria feature binding price caps (list prices) that prevent retailers from passing on their distribution costs to consumers, and are more likely to occur when manufacturers exchange demand information than when they do not share this information. Manufacturers profit from exchanging demand information when products are sufficiently differentiated, and retailers’ distribution costs are high enough. Yet, expected prices are unambiguously lower when manufacturers exchange demand information than when they don’t, making the information exchange beneficial to consumers.
Information sharing agreements are common in many prominent sectors and have been under close antitrust scrutiny on both sides of the Atlantic for many years. An established literature has extensively studied the competitive and welfare effects of communication in oligopoly games where firms first decide whether to share their private information (about demand or costs) and then compete à la Cournot or Bertrand (see, e.g., Kühn and Vives (1995), and Vives, 2006, for surveys of this literature).
These models rest on the traditional hypothesis that firms are profit-maximizing black boxes and are thus silent on the interplay between firms’ internal delegation decisions and their incentives to exchange information. Yet, when firms are viewed as organizations composed of different individuals with dispersed information, responsibilities, and non-congruent interests, the allocation of decision rights becomes an integral component of their internal architecture and competitive strategies.
To investigate the interaction between information sharing and internal organization, in this paper, we study the link between manufacturers’ incentives to enter an information-sharing agreement, wherewith they exchange demand information, and the strategies according to which they delegate price authority to their downstream retailers. However, in contrast to the existing vertical contracting literature, where similar questions have been addressed under the hypothesis that monetary incentives are enforceable, we consider a competitive environment where manufacturers cannot internalize retailers’ objectives through monetary transfers but can only decide what they are entitled to do by designing permission sets from which retailers can select prices. Hence, manufacturers face the so-called ‘delegation dilemma’: giving up vertical control to gain flexibility, or imposing rigid rules unresponsive to changes of the environment?
We consider a stylized two-stage game, with linear demand functions, in which two upstream manufacturers compete by producing differentiated products and choose simultaneously, in the first stage, how much price authority to grant their downstream retailers who are privately informed about an aggregate, additive and binary demand shock — i.e., being closer to the final market, these retailers are better informed than manufacturers about consumers’ willingness to pay. In the second stage, after demand has been realized, retailers simultaneously set actual prices given the constraints (if any) imposed by the manufacturers in the first stage. We assume that retailers incur an observable but not verifiable distribution cost to introduce a simple wedge between upstream and downstream objectives. The presence of this cost, together with retailers’ private information, creates a natural misalignment of preferences that is the core of our theory, and makes it different from traditional vertical contracting models with privately informed retailers. Since by assumption manufacturers cannot internalize the non-verifiable distribution cost through monetary transfers, retailers will pass on this cost to consumers at the expense of sale volumes, profits and consumer surplus. Therefore, although retailers are better informed than manufacturers on demand and can tailor prices to this information, they tend to charge excessive prices compared to what upstream competition mandates. Yet, in contrast to other delegation models, manufacturers are not totally uninformed in our framework: with some probability, they observe the state of demand (i.e., high or low willingness to pay) and can, therefore, condition the degree of price authority granted to their retailers on this information. We study the incentives of these manufacturers to exchange demand information and examine the competitive and welfare effects of this agreement — i.e., its impact on prices, profits and consumer surplus. We show the following results.
First, we establish that full delegation never occurs in equilibrium regardless of whether manufacturers share demand information or not. That is, as long as the distribution cost is positive (even negligible), manufacturers will never allow retailers to set prices in all demand states freely. Specifically, manufacturers always have an incentive to constrain their retailers’ pricing decisions (in at least one state of nature) to prevent them from charging excessive prices compared to what upstream profit maximization requires.
Second, given that full delegation is never an equilibrium, we characterize equilibria with partial delegation in both information-sharing regimes. In these equilibria, retailers are entitled to choose their preferred price only when demand is low and are, instead, constrained to charge a price lower than what they would like in the high-demand state. Hence, partial delegation equilibria feature a price cap or, equivalently, a list price: a result broadly consistent with customized pricing with discretion (see, e.g., Phillips et al., 2021). As intuition suggests, partial delegation equilibria exist when distribution costs are not too high, so that the conflict of interest between manufacturers and retailers is not too pronounced. The region of parameters where partial delegation occurs in equilibrium shrinks when products become closer substitutes because retailers are more incentivized to pass on their distribution costs to consumers to shield against increased competition. On the contrary, this region of parameters expands as manufacturers’ information accuracy and the significance of demand uncertainty increase. The higher the probability that manufacturers are informed, the easier for them to sustain an equilibrium with partial delegation. Specifically, a manufacturer that expects its rival to be informed with a higher probability will be keener to delegate because it expects a lower opponent’s price, which, by strategic complementarity, also makes his retailer willing to charge a lower price, endogenously reducing its bias towards an excessive price. The opponent will, in turn, expect a lower price and accordingly reduce its price as well. Moreover, the greater the significance of demand uncertainty, the higher the cost for the manufacturers to give up flexibility and implement a rigid pricing rule (pooling) that makes prices unresponsive to demand shocks — i.e., the greater the cost of not allowing the retailers to engage in price tailoring.
Third, we show that manufacturers mutually benefit from exchanging their demand information only when products are sufficiently differentiated, when their information accuracy is high, and distribution costs are not too low. The following effects determine this result. When manufacturers share information, each learns the demand shock with greater probability than without information sharing (because under the information-sharing agreement an uninformed manufacturer learns the state of demand when the rival is informed). Hence, other things being equal, the information exchange benefits manufacturers because it mitigates the conflict of interest with their retailers. Yet, when both manufacturers are uninformed and know this is the case because the information-sharing agreement is in place, retailers charge higher prices in the low-demand state. In this state of nature, each retailer is sure that the opponent will pass on its cost to consumers and will be keener to raise its price, as opposed to the no information sharing regime where there is uncertainty on whether the rival manufacturer is informed or not. This price-increasing effect has an ambiguous impact on the manufacturers’ profits: it benefits manufacturers since it softens competition; but, when distribution costs are excessively high, there is excessive pass-through, reducing sales and thus profits. Overall, exchanging demand information benefits manufacturers if retailers’ distribution costs are not too small and products are sufficiently differentiated. In this case, solving internal agency conflicts is relatively more important than softening competition. We also show that expected prices are lower with than without information sharing. Essentially, by aligning incentives within organizations, information sharing reduces the pass-through rate according to which distribution costs are passed on to consumers. Hence, from an ex ante point of view, the exchange of information unambiguously benefits consumers.
Finally, to discuss the robustness and the limitations of these results, in the online Appendix, we also examine (pooling) equilibria in which, when uninformed, manufacturers retain full price authority — i.e., they force a singleton price irrespective of the demand state. In these equilibria, results align with the existing literature (see, e.g., Kühn and Vives (1995) and Vives, 2006) since manufacturers de facto behave as uninformed, vertically integrated oligopolists. When a pooling equilibrium arises with and without information sharing, the exchange of information has a neutral impact on expected prices and consumer surplus. However, sharing information unambiguously benefits manufacturers because it allows them to tailor prices to demand, while increasing the extent of vertical control. By contrast, when the equilibrium features partial delegation under information sharing and pooling without information sharing, we find that consumers are hurt by information sharing. Yet, in this hybrid scenario, manufacturers will not share information when products are sufficiently differentiated and the significance of demand uncertainty is not too high. We also argue why our results remain qualitatively valid under alternative information structures, demand specifications, and organizational structures (i.e., inter-brand competition), identifying new avenues for future research whenever possible.
The rest of the paper is organized as follows. Section 2 lays down the baseline model. In Section 3, we develop two useful benchmarks and show that full delegation cannot occur in equilibrium irrespective of the information sharing regime. In Section 4, we characterize and show the existence of partial delegation equilibria with and without information sharing and then study the effect of these agreements on equilibrium prices, profits and consumer surplus. In Section 5, we discuss robustness and relate our results to the existing work. Section 6 concludes by discussing possible avenues for future research. Proofs are in the Appendix. In the online Appendix, we provide additional material and robustness checks.
This paper was originally published for International Journal of Industrial Organization here. 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.