Vertical mergers Double marginalization Foreclosure Empirical evidence Vertical mergers Industrial organization – lecture 5 Vertical mergers Double marginalization Foreclosure Empirical evidence Vertical mergers Vertical mergers join firms operating at different levels of production chain (e.g. producer and retailer). What are the effects of vertical merger compared to horizonal merger? Vertical mergers join firms producing complementary products. Internalization of this externality is Pareto improving. Case: In 2000 GE and Honeywell announced merger. GE produces jet engines and Honeywell produces electric and braking systems for aircraft engines. In July 2001 the merger was blocked by EC. Why? Vertical mergers Double marginalization Foreclosure Empirical evidence Double marginalization Pepall et al. (2010, pp. 325–328) What are the procompetitive effects of vertical mergers? • Each firm in a production chain provides essential input to other firms in the chain. • Firms on each level of the production chain have some market power. • Firms on each level of the production chain charges some mark-up above marginal costs. • The price for final consumer may be higher than tthe monopoly price. • This problem is called double marginalization. Vertical mergers Double marginalization Foreclosure Empirical evidence Double marginalization: Model Pepall et al. (2010, pp. 325–328) There is a single manufacturer m and single retailer r The producer produces the goodat constant unit cost c and sells it ot he retailer at a wholesale price w. The retailer resells the product to the final consumer at a final price P. The inverse demand function is linear P = A − BQ. Vertical mergers Double marginalization Foreclosure Empirical evidence Double marginalization: Solution Pepall et al. (2010, p. 301-304) Solution of the model is given by backward induction. Profit maximizing price and output of the retailer for given wholesale price w are Q(w) = A−w 2B and P(w) = A+w 2 Substituting the retailer’s output into the profit function of the manufacturer and maximiying with respect to w gives the optimal wholesale price w∗ = A+c 2 . The retailer’s equilibrium output is thus Q∗ = A−c 4B Vertical mergers Double marginalization Foreclosure Empirical evidence Double marginalization: Solution Pepall et al. (2010, p. 301-304) After merger the whole industry is monopolized. The profit maximizing output and price of the integrated firm are QI = A−c 2B and PI = A+c 2 . The merger results in lower price, greater quantity, higher profits and higher consumer surplus. Two assumptions are crucial for this analysis 1. Fixed proportion between inputs and outputs 2. Linear pricing Vertical mergers Double marginalization Foreclosure Empirical evidence Foreclosure Pepall et al. (2010, p. 332-333) There can be also anti-competitive effects of vertical mergers. The most important one is foreclosure. The integrated company may choose to deny a downstream competitior a source of inputs. Consider an industry with two independent manufacturers and two independent retailers. • Is vertical integration profitable? Yes • Can vertical integration harm the consumers? Yes We ilustrate the foreclosure logic in the model by Ordover, Saloner and Salop (1990). Vertical mergers Double marginalization Foreclosure Empirical evidence OSS model Pepall et al. (2010, p. 338-342) There are two manufacturer mi and two retailers ri . Prices are, again, denoted as pi and wi . The retailers’ demands are qr i = 1 − pi + βpj where parameter β ∈ measures the degree of product differentiation. Marginal cost are normalized to zero. We solve the model in three versions: 1. Without merger 2. With merger and foreclosure 3. With merger and without foreclosure Vertical mergers Double marginalization Foreclosure Empirical evidence OSS model: Retailer’s subgame Pepall et al. (2010, p. 338-342) Profit of retailer i given it pays wholesale price ci is Πr i = (pi − ci )(1 − pi + βpj ) We derive best-responses and solve for retailers’ prices. The equilibrium price is p∗ i (ci , cj ) = 2+β+2ci +βcj 4−β2 and the equilibrium quantity is q∗ i (ci , cj ) = 2+β−(2−β2 )ci +βcj 4−β2 You can check that the equilibrium profit is Πi = (q∗ i (ci , cj ))2 Vertical mergers Double marginalization Foreclosure Empirical evidence OSS model: Manufacturer’s subgame Pepall et al. (2010, p. 338-342) Version 1: Manufacturers charges w1 = w2 = 0 because of Bertrand competition. Version 2: Marginal cost of integrated firm is zero. Manufacturer 2 charges w2 = 2+β 2(2−β2) . Version 3: Manufacturers again behaves like Bertand competitors resulting in w1 = w2 = 0. Conclusion: The final prices increase if and only if the integrated firm can crediblly commit to not deliver to competing retailer. If such a commitment is available integration raises rival’s cost and final prices. However, existence of such a commitment is unlikely. Vertical mergers Double marginalization Foreclosure Empirical evidence Foreclosure: GE-Honeywell merger Pepall et al. (2010, p. 343-344) It is a pretty famous and a very controversial case. Citation of commission’s report (par. 355): Because of their inability to match the bundle offer ... independent suppliers will lose market shares to the benefit of the merged entity and experience an immediate damaging of profit shrinkage. As a result, the merger is likely to lead to market foreclosure ... and to the elimination of competition in these areas. Does it make sense? Vertical mergers Double marginalization Foreclosure Empirical evidence Empirical evidence: Concrete industry Pepall et al. (2010, p. 346-348) Hortacsu, Syverson (2007, JPE) study the effect of vertical integration on prices. Why concrete industry? • Fixed proportion between input and output • Variation in vertical integration • High transportation cost creates many local markets. Many markets mean many independent observations. Estimated equation Pit = α + βVIit + γXit where Pit is the average concrete price, VIit is the share of vertically integrated firms and Xit are control variables (market and year fixed effects, HHI, ...) Vertical mergers Double marginalization Foreclosure Empirical evidence Empirical evidence: Concrete industry Pepall et al. (2010, p. 346-348) Main results: Variable Market share of VI firms −0.09∗ −0.086∗ −0.043 Market share of multiple plant firms - −0.15 −0.001 Total factor productivity - - −0.293∗ R2 0.433 0.434 0.573 How do you interpret this?