Part IV Theory of competition policy Chapter 17

Part IV. Theory of competition policy Chapter 17. Vertically related markets Slides Industrial Organization: Markets and Strategies Paul Belleflamme and Martin Peitz © Cambridge University Press 2009

Chapter 17 - Objectives Chapter 17. Learning objectives • Double marginalization problem within a monopoly context • Contracts between the upstream and downstream firm that differ from linear pricing • Role of resale-price maintenance and exclusive territories • Role of exclusive dealing contracts • Oligopolistic industry upstream and downstream • Effects of vertical mergers © Cambridge University Press 2009 2

Chapter 17 – The double-marginalization problem • Pricing inefficiency • In a market with firms operating only at one level of a vertical supply chain, retail prices are higher than in a market with vertically integrated firms • A downstream firm applies a margin to the wholesale price which includes the margin of an upstream firm • Retailer does not take into account the externality exerted on the upstream firm by changing the retail price © Cambridge University Press 2009 3

Chapter 17 – The double-marginalization problem Linear pricing and double marginalization • Market demand Q(p) = a – bp • Marginal costs c < a/b • Upstream firm called manufacturer, downstream firm called retailer • The producer sets the wholesale price w at stage 1 • At stage 2 the retailer (assumed not to incur any costs) observes the wholesale price and sets the retail price p © Cambridge University Press 2009 4

Chapter 17 – The double-marginalization problem Linear pricing and double marginalization (cont’d) • By backward induction • • First-order condition equivalent to • Retail price © Cambridge University Press 2009 5

Chapter 17 – The double-marginalization problem Linear pricing and double marginalization (cont’d) • A vertically integrated firm would • pm = a/(2 b) + c/2 • pm < p* • Lesson: In a vertically related industry with an upstream and a downstream monopolist in which each firm maintains the price-setting power of its product, the retail price is above the monopoly price set by a vertically integrated firm. • Double marginalization is most pronounced in successive monopoly • Insight remains relevant under imperfect competition • As one layer of the market loses its market power, double marginalization becomes less pronounced © Cambridge University Press 2009 6

Chapter 17 – The double-marginalization problem Contractual solutions to the doublemarginalization problem • Mimic the monopoly solution of the vertically integrated firm by a profit-sharing agreement • But • Retailer’s price not easily observable • Retailer has nonobservable marginal costs • Thus informational problems between the two parties would lead to divergence from the vertically integrated solution © Cambridge University Press 2009 7

Chapter 17 – The double-marginalization problem Contractual solutions to the double-marginalization problem (cont’d) • Result under the assumptions that there is a single retailer and manufacturer is fully informed • Two-part tariff consisting of a wholesale price per unit and a fixed • • • fee • Wholesale price equal to marginal manufacturer costs • Manufacturer makes profits from a fixed fee Profit-maximizing level equal to the profit of the retailer gross of this fee Upstream firm absorbs all profits Fee can be understood as a franchise fee • Several retailers • Retailer has to set the fixed fee equal to the gross profit of the • smallest retailer to ensure that all retailers participate Manufacturer optimally set its unit price above marginal costs to extract profits from all retailers © Cambridge University Press 2009 8

Chapter 17 – The double-marginalization problem Contractual solutions to the doublemarginalization problem (cont’d) • Lesson: Nonlinear pricing and other contracts can solve the double-marginalization problem. • Resale-price maintenance • Upstream firm mandates prices downstream • Presence of local monopolists in the downstream market who can become active in resale • may make it optimal for manufacturer not to use nonlinear pricing © Cambridge University Press 2009 9

Chapter 17 – The double-marginalization problem Double marginalization and retail services • In retailing downstream firms often provide complementary services • Explain to consumers the functioning of a product • Thus increase the consumer’s willingness-to-pay or • Hold excess sales staff to keep lines short • Moral hazard problem • manufacturer would like to compensate the retailer for such efforts but does not observe them • Market demand Q(p, s) depends on • The retail price • Service level or quality provided by the retailer • Retailer incurs a cost ψ(s) per unit of output • Under vertical integration maximizes • Solution of this vertically integrated structure pm and sm © Cambridge University Press 2009 10

Chapter 17 – The double-marginalization problem Double marginalization and retail services (cont’d) • Linear wholesale price • Manufacturer maximizes with respect to w at stage 1 • Retailer maximizes with respect to p and s at stage 2 • With the retail price we have again a doublemarginalization problem • Retailer distorts its service provision • Does not take into account that an increase in services also increases the manufacturer’s profit © Cambridge University Press 2009 11

Chapter 17 – Resale-price maintenance and exclusive territories Resale-price maintenance • In markets in which retailer’s investments are of little relevance • Wonder whether resale-price maintenance inflicts any social harm • Manufacturers have an incentive to keep retailer margins small • Competition at the manufacturer level may be sufficiently strong to keep retail prices down • Arguably a much bigger role for retailers beyond the reduction of transaction costs • Retailers engage in costly service provision © Cambridge University Press 2009 12

Chapter 17 – Resale-price maintenance and exclusive territories Resale-price maintenance (cont’d) • Single manufacturer and two competing retailers • Retailers are horizontally differentiated on the Hotelling line • Compete in prices and service • Retailer profit • Manufacturer’s profit • Without RPM linear non-discriminatory wholesale price w • Total profit of the industry © Cambridge University Press 2009 13

Chapter 17 – Resale-price maintenance and exclusive territories Resale-price maintenance (cont’d) • Total industry profits are maximized only if and • Retailer i’s profit can be written as • First order conditions © Cambridge University Press 2009 14

Chapter 17 – Resale-price maintenance and exclusive territories Resale-price maintenance (cont’d) • Second term: externality with respect to price and service • • level Third term: horizontal externality Retailers implement the solution that maximizes total profits of the full vertical structure if the last two terms in both first-order conditions cancel out to zero • Lesson: The use of resale-price maintenance by a manufacturer leads to higher retail prices and more retail services if consumers are more sensitive to price competition than to service competition. Conversely, this leads to lower prices and fewer retail services if consumers are less sensitive to price competition than to service competition. © Cambridge University Press 2009 15

Chapter 17 – Resale-price maintenance and exclusive territories Resale-price maintenance (cont’d) • In the Hotelling model outlined above • Increase in services by the retailer is less effective in stealing business from the competitor than a reduction in price • Retailers are biased towards price competition • Manufacturer can improve by setting a price floor that is binding in the equilibrium • Manufacturer can use two-part tariffs to obtain the full profit of the vertically integrated solution • Demonstrates the use of RPM to implement the vertically integrated solution © Cambridge University Press 2009 16

Chapter 17 – Resale-price maintenance and exclusive territories Resale-price maintenance (cont’d) • If one Retailer is engaged in promotional efforts other retailers may free-ride on its activities • Additional horizontal externality • If consumers first seek advice about the usefulness of product and then shop around for the best price • Retailer has little incentive to invest in services • Manufacturer can use resale-price maintenance to avoid destructive price competition • RPM protects the profit margin of the retailer • Since the retailer cannot be undercut, consumers have no reason to become ‘disloyal’ © Cambridge University Press 2009 17

Chapter 17 – Resale-price maintenance and exclusive territories Resale-price maintenance (cont’d) • Possibly anticompetitive effect of RPM • RPM can affect the sustainability of a cartel in the upstream market • Cartels become less stable if wholesale prices cannot be observed by other cartel members • Difficult to distinguish between retail price changes due to cost changes in the downstream market and those due to individual deviations by cartel members • RPM eliminates retail price variation • Price deviations by cartel members are easier to detect © Cambridge University Press 2009 18

Chapter 17 – Resale-price maintenance and exclusive territories Exclusive territories • Increase market power of downstream firms • Why could upstream firms be interested in granting market power to downstream firms? • Exclusive territories may increase downstream investment in services • Protects part of the rents that are generated through the investment • For instance if a car dealer invests in its sales staff to provide important pre-purchase information, other car dealers who sell products without these services but at a lower price may undermine the investment incentives • Why have upstream firms have an incentive to offer rents downstream if we abstract from investment decisions downstream? • May be in the interest of upstream firms to offer exclusive territories • Makes upstream demand less elastic • Thus reduces competition • Even though exclusive territories create market power downstream and thus lead to reduced output in the industry © Cambridge University Press 2009 19

Chapter 17 – Resale-price maintenance and exclusive territories Exclusive territories (cont’d) • Two region model with exclusive territories Region a Retailer of product 1 in region a Retailer of product 2 in region a Manufacturer 1 Manufacturer 2 Retailer of product 1 in region b Retailer of product 2 in region b Region b © Cambridge University Press 2009 20

Chapter 17 – Resale-price maintenance and exclusive territories Exclusive territories (cont’d) • Two-products, two-region price competition model with • • linear market demand in each region Absent exclusive territories Pure Bertrand competition between retailers of the same product • pia = pib = wi • Wholesale price wi plus unit cost of retailing, which, for simplicity • • • is set equal to zero Perfectly competitive retailing sector Two upstream firms set their price as in standard duopoly model Demand in region k: αk(1 – pik + dpjk), k = a, b , αa+ αb = 1 and 0 ≤ d < 1 Upstream firm maximizes wi (1 – wi + dwj) p* = w* = 1/(2 -d) upstream firm makes profit πc = 1/(2 -d)² © Cambridge University Press 2009 21

Chapter 17 – Resale-price maintenance and exclusive territories Exclusive territories (cont’d) • With exclusive territories • Retailers obtain strictly positive profits under this arrrangement • Willing to sign contracts which give them exclusivity in their • region (under the condition that they do not sell in the other region) Timing • Stage 1: manufacturers set their wholesale price • Stage 2: after learning the wholesale price of their own supplier, retailers set retail prices simultaneously • in region k: • • retailer of product i maximizes αk(pik – wi) (1 – pik + dpjk) First-order condition of profit maximization 1 – 2 pik + dpjk + wi = 0 Retailer i does not observe the competing retailer j’s input price • Has to form an expectation about that price • Each retailer believes that the competing retailer within the same region faces the symmetric equilibrium wholesale price w* © Cambridge University Press 2009 22

Chapter 17 – Resale-price maintenance and exclusive territories Exclusive territories (cont’d) • If wi = w* • Manufacturer i maximizes wi[1 – pik(wi , w*)+ dpjk(wi , w*)] • First order condition • Due to double marginalization, p > pc • Profit with exclusive territories π* is larger than profit πc without exclusive territories as long as d is large enough i. e. , the two products are close enough substitutes • Lesson: Manufacturers may make higher profits if they sell through exclusive territories than if they do not. Retailers are also better off. However, consumers suffer and total surplus is reduced. © Cambridge University Press 2009 23

Chapter 17 – Exclusive dealing Anticompetitive effects of exclusive dealing contracts? The Chicago critique • Exclusive dealing clauses were largely seen as • • anticompetitve according to the antitrust doctrine that prevailed in the first half of the twentieth century in the US View was challenged by Chicago Law School Argument • Downstream firm which foresees that more attractive terms are • available under competition will demand a compensation from the proposing upstream firm for signing a long-term contract with an exclusivity clause Exclusive dealing, when observed in practice, due to efficiency gains • Correct point: incentives of the downstream firm have to be considered © Cambridge University Press 2009 24

Chapter 17 – Exclusive dealing Anticompetitive effects of exclusive dealing contracts? The Chicago critique (cont’d) • Two sellers, an incumbent and an entrant offer purchasing contracts for homogeneous product • Buyer interpreted as a retailer that operates as a monopolist in its markets with demand Q(p) • Incumbent faces constant marginal cost of production c. I • entrant more efficient c. E < c. I but also has to pay entry cost e • Assume that entry occurs if no exclusive dealing clause is signed (c. I – c. E) Q(c. I) > e • Entry is efficient • Payment m in return for signing a legally binding exclusive dealing contract • Buyer decides whether to accept or reject • After observing whether exclusive dealing will prevail • Potential entrant decides whether to enter • Firms in the market set prices simultaneously © Cambridge University Press 2009 25

Chapter 17 – Exclusive dealing Anticompetitive effects of exclusive dealing contracts? The chicago critique (cont’d) • If the potential entrant did not enter • Incumbent would obtain monopoly profit • If the entrant does enter • Asymmetric Bertrand competition: price is equal to c. I • Incumbent makes zero sales • For obtaining exclusivity, the incumbent is willing to pay up to to the buyer up-front πm • Buyer suffers a loss from accepting exclusivity: has to pay the price pm instead of c. I • Loss • ΔCS > πm buyer cannot be compensated by the incumbent to m accept the exclusive dealing clause at payment m≤π • Not profitable for the incumbent to induce the buyer to accept exclusive dealing • Exclusive dealing cannot be anticompetitive © Cambridge University Press 2009 26

Chapter 17 – Exclusive dealing Vertical integration and long-term contracts as partial deterrence devices • Modifying previous model to show that vertical integration • leads to socially insufficient entry Show that long-term exclusive dealing contracts can achieve the same outcome • Wilingness-to-pay for all consumers to be identical and equal to 1 • Mass of consumers with unit demand • c. I = ½, • c. E takes a realization between 0 and 1, is uniformly distributed • between 0 and 1 Socially optimal allocation • Buyer obtains the good from the incumbent if c. I < c. E • Initially incumbent does not know the entrant’s cost parameter © Cambridge University Press 2009 27

Chapter 17 – Exclusive dealing Vertical integration and long-term contracts as partial deterrence devices (cont’d) • First: model with entry in the absence of vertical integration or long-term contracting • Equilibrium price is ½ if c. I > c. E and c. E if c. I < c. E • Incumbent’s expected profit is • Entrant’s expected profit • Expected price • Expected net buyer surplus 3/8 © Cambridge University Press 2009 28

Chapter 17 – Exclusive dealing Vertical integration and long-term contracts as partial deterrence devices (cont’d) • Second: model with vertical integration • Vertically integrated firm minimizes its expected cost of obtaining the product • Internally at cost c. I = 1/2 • Externally through the entrant • Since c. E is uniform on [0, 1] • Incumbent firm offers p = 1/4 to the entrant • • Insufficient entry: if c. E ∊ (1/4, 1/2), entry does not take place but would be socially efficient Lesson: Vertical integration between an upstream and a downstream seller may lead to an inefficient allocation because the integrated firm uses its market power to offer too low a price for the product of the outside seller. © Cambridge University Press 2009 29

Chapter 17 – Exclusive dealing Vertical integration and long-term contracts as partial deterrence devices (cont’d) • Third: long term exclusive dealing contracts • • At t = 1 the seller proposes a contract to the buyer At t = 2 the buyer accepts or rejects the contract At t = 3 the entrant observes its costs and then decides whether to enter At t = 4 • Consider the following contract • Suppose the buyer accepts this contract • Price competition between sellers if the contract was rejected at t = 2 • The contract applies to the incumbent whereas the entrant sets its price • Incumbent offers the product at price 3/4 • Sets a penalty for a breach of contract equal to 1/2 • Entrant makes an offer to the buyer • Buyer accepts any price • At stage 4 entrant sets its price equal to 1/4 provided that it covers cost • At stage 3 the entrant becomes active if its costs are weakly less • than 1/4 Allocation is the same as under vertical integration © Cambridge University Press 2009 30

Chapter 17 – Exclusive dealing Vertical integration and long-term contracts as partial deterrence devices (cont’d) • To be checked 1. Buyer must have an incentive to sign the contract 2. Must be profitable for the incumbent seller to propose such a contract • On 1 • If the buyer rejects the contract • Entry takes place if c. E ≤ 1/2 • Price equal to c. I = 1/2 • The incumbent’s monopoly position allows him to extract the full surplus • Entry does not take place if c. E > 1/2 • Expected price to be paid by the buyer is • Buyer has no incentive to reject the contract position by the incumbent © Cambridge University Press 2009 31

Chapter 17 – Exclusive dealing Vertical integration and long-term contracts as partial deterrence devices (cont’d) • On 2 • Show that incumbent is indeed better off with the long-term contract • Incumbent seller makes expected profit © Cambridge University Press 2009 32

Chapter 17 – Exclusive dealing Vertical integration and long-term contracts as partial deterrence devices (cont’d) • If the incumbent seller did not propose any contract at stage 1, it would sell at c. E whenever c. I ≤ c. E • The expected profit which is clearly less than 5/16 • Lesson: Under imperfect competition, exclusive dealing contracts that are signed before entry takes place can constitute a barrier of entry. In effect, there is too little entry from a welfare perspective. • General lesson: exclusive dealing can be anticompetitive © Cambridge University Press 2009 33

Chapter 17 – Exclusive dealing Full exclusion and multiple buyers • Potential entrant has known costs c. E < c. I • Suppose entrant enjoys increasing returns thus average cost is decreasing • If a sufficiently large number of buyers sign up an exclusive dealing contract with the incumbent, entrant cannot offer attractive terms to the remaining buyers • By signing up with the incumbent, buyers exert a negative externality on other buyers • Incumbent thus avoids entry if it manages to convince a sufficiently large number of buyers to sign the exclusive dealing contract © Cambridge University Press 2009 34

Chapter 17 – Exclusive dealing Buyer (mis-)coordination • Two buyers instead of one • Buyers simultaneously decide whether to sign the exclusive dealing contract • Incumbent is assumed to offer the same contract to both buyers • In the absence of exclusive dealing contracts, entry by the more efficient firm takes place: 2(c. I – c. E) Q(c. I) > f • Assume that an entrant who sells to one buyer only cannot recover its entry costs at price c. I , (c. I – c. E) Q(c. I) < f • If the incumbent can sign the exclusive dealing contract with one buyer, entry is not viable • Incumbent is willing to pay up to πm to each buyer • Suppose that buyer 2 signs • Entry will not take place • Incumbent will set the monopoly price pm independent of whether buyer 1 signs • For any positive payment buyer 1 has an incentive to sign (Applies symmetrically to buyer 2) © Cambridge University Press 2009 35

Chapter 17 – Exclusive dealing Buyer (mis-)coordination (cont’d) • Lesson: Due to buyer miscoordination, an incumbent firm can possibly make buyers sign exclusive dealing clauses. Here, the incumbent firm is better off with these clauses in place and the more efficient rival firm is excluded from the market. • Another equilibrium in which buyer do coordinate their decisions • In this case: back to the original setting in which exclusive dealing clauses are not anticompetitive © Cambridge University Press 2009 36

Chapter 17 – Exclusive dealing Sequential contract proposal • Incumbent can propose contracts sequentially • Such sequential contracting is more profitable for the incumbent than a divide-and-conquer strategy • Incumbent first offers a contract to buyer 1, then to buyer 2 • Under the assumption 2πm – ΔCS > 0 • Profitable to offer a contract to buyer 2 that it prefers over rejecting the contract in case buyer 1 rejects the contract • Buyer 1 foresees if it rejects the contract, the incumbent will make sure that buyer 2 signs it • Buyer 1 is willing to sign any contract that leaves a nonnegative net surplus for itself • Once buyer 1 has signed, buyer 2 also signs any contract that does not make it worse off than not buying at all • Incumbent can exclude the entrant at zero cost © Cambridge University Press 2009 37

Chapter 17 – Exclusive dealing Exclusive contracts and investment incentives • Investments at the retailer level that affect the value of a product • A seller may train the sales staff of a retailer • In the absence of exclusive dealing the other seller may reap some of the benefit from the investment • Sales staff may use their newly acquired skills to increase the sales of this alternative seller • Incumbent seller I, buyer B, entrant E (external source) • Incumbent seller or the buyer decides about an • • investment which is not contractible In the absence of exclusive dealing, buyer is free to buy from an external source With exclusive dealing, buyer needs to obtain the agreement of the incumbent seller to use the external source instead © Cambridge University Press 2009 38

Chapter 17 – Exclusive dealing Exclusive contracts and investment incentives (cont’d) • Suppose that the entrant seller does not have any bargaining power • Concerning incumbent seller and buyer: the two parties evenly split the surplus that is achieved on top of the disagreement value • First understand • Exclusive dealing contract does not affect investment levels if the investment does not affect the surplus obtained by contracting with the external source, i. e. , if it leaves disagreement profits unchanged • Total surplus © Cambridge University Press 2009 39

Chapter 17 – Exclusive dealing Exclusive contracts and investment incentives (cont’d) • Nonexclusive contracts • Buyer’s profit is • Incumbent seller’s profit is • The profit maximizing investment level AI(c. I) determined by the solution of TS'(c. I)/2= AI'(c. I) © Cambridge University Press 2009 40

Chapter 17 – Exclusive dealing Exclusive contracts and investment incentives (cont’d) • Under exclusion • Incumbent seller and buyer obtain • Disagreement levels are independent of the incumbent seller’s investment • Its investment choice is not affected by the presence of exclusive dealing • Lesson: In a bargaining environment in which incumbent seller and buyer share any surplus above their disagreement profits and in which profits of any entrant seller are held down to zero, exclusive dealing does not affect the surplus that can be generated by contracting with the entrant seller instead. © Cambridge University Press 2009 41

Chapter 17 – Exclusive dealing Exclusive contracts and investment incentives (cont’d) • Show that exclusive dealing can be efficiency-enhancing • Suppose that there are positive spillovers • Also c. E depends negatively on the investment level: When reducing c. I the incumbent seller also reduces c. E • Under an exclusive contract • With the nonexclusive contract • Incumbent seller’s profit is • Incumbent seller now obtains • Profit-maximizing investment level determined by solving in c. I • Left-hand side is greater than TS'(c. I)/2= – 1/2 • Incumbent seller has a stronger incentive to invest under exclusion © Cambridge University Press 2009 42

Chapter 17 – Vertical oligopoly and vertical mergers Exclusionary effects of vertical mergers • Input foreclosure: vertical integration may lead to higher input prices for competitors • Higher price may be due to vertically integrated firms • • not selling inputs on the market or, at least, restricting their supply Vertical integration can be used as a tool to increase rival’s costs Lesson: Vertical integration may raise the costs of nonintegrated downstream rivals. A higher wholesale price may or may not lead to higher retail prices. • Such mergers may not hurt consumers (since retail prices may fall) and often are not profitable for the firms involved in the merger © Cambridge University Press 2009 43

Chapter 17 – Vertical oligopoly and vertical mergers Coordinated effects of vertical mergers • Vertical integration may improve the viability of collusion among competing firms • Key reason for vertical integration facilitating collusion: outlets effect • Vertical integration by an upstream firm reduces the number of outlets through which its rivals can sell when deviating • Reduces their profit from cheating and thus facilitates collusion • Counteracting: punishment effect • If an upstream firm integrates with a downstream firm these profits now become part of the merged entity • The merged entity can expect to make more profits in the non-cooperative punishment phase than the upstream firm would make alone © Cambridge University Press 2009 44

Chapter 17 – Vertical oligopoly and vertical mergers Coordinated effects of vertical mergers (cont’d) • Merged entity suffers less than a stand-alone upstream firm from a switch from collusive to punishment phases • outlets effect outweighs the punishment effect so that the net effect of a vertical merger is to facilitate collusion • Lesson: Downstream vertical integration reduces the number of outlets through which upstream rivals can sell and thus reduces profits if deviating from a collusive outcome. Vertical integration may then facilitate collusion. © Cambridge University Press 2009 45

Chapter 17 - Review questions • Suppose that an industry consists of two upstream monopolists who exclusively sell at a linear price to one downstream firm each. What would be the effect of vertical integration (so that each upstream monopolist owns its retail outlet) on the final good price? • What are possible efficiency-defences of the use of resale-price maintenance? • For which reasons can it be profitable for manufacturers to grant exclusive territories to their retailers? • Provide two reasons why the Chicago school argument on exclusive dealing (namely that, whenever exclusive dealing is observed, it must be welfare improving) is wrong. • Should competition authorities prohibit vertical mergers that lead to higher input prices? • What are possible coordinated effects of vertical mergers? © Cambridge University Press 2009 46