Thursday, January 15, 2015
Posted by Thom Lambert
Why the Eleventh Circuit Should Reverse McWane
We know four key things about exclusive dealing.
First, exclusive dealing arrangements are ubiquitous. A modern American can hardly go a day without entering into a transaction with some business subject to exclusive dealing or a similar distribution restriction. Nearly every franchisee, for example, has agreed to purchase some product exclusively from its franchisor.
Second, exclusive dealing is a competitive “mixed bag.” Such arrangements may occasion anticompetitive harm (i.e., a reduction in overall market output) if they foreclose the exclusivity-demanding producer’s rivals from so many sales outlets that those rivals are driven below minimum efficient scale (MES) and therefore see their per-unit costs rise. On the other hand, exclusive dealing may secure a number of procompetitive (i.e., output-enhancing) benefits. They may: (1) encourage producers to invest in downstream buyers’ (e.g., retailers’) operations by preventing “interbrand free-riding” by competing producers; (2) reduce consumer prices by intensifying competition among producers for distribution; (3) enhance consumer welfare by reducing the costs associated with uncertain supply and demand; and (4) encourage the production of multi-component systems by protecting producers of complete systems from adverse “cherry picking” by producers of popular, high-margin individual components.
Third, for exclusive dealing to cause anticompetitive harm, several conditions must be satisfied. First, the degree of foreclosure occasioned by the perpetrator’s exclusive dealing must be substantial enough to drive (or hold) at least some rivals below MES. Second, it must be impracticable for foreclosed rivals to bypass the buyers subject to the exclusive dealing arrangements and sell to others by, say, integrating forward into distribution or selling through newly entering distributors. Finally, output-reducing exclusive dealing is unlikely absent significant barriers to entry in the producer market. If market power created by foreclosure-inducing exclusive dealing could be easily undermined by new firms entering the producer market in response to supracompetitive prices, producers (who generally have to “pay” something to induce exclusivity) would be unlikely to attempt monopolization via exclusive dealing, and even if they did so, consumer harm would be unlikely.
Given the degree to which the stars must align for exclusive dealing to occasion anticompetitive harm, the fourth thing we know about exclusive dealing should come as no surprise: Most instances of exclusive dealing enhance, rather than reduce, overall market output. See, e.g., Cooper et al., Vertical Antitrust Policy as a Problem of Inference, 23 Int’l J. Indus. Org. 639, 658 (2005) (observing that although “some studies find evidence consistent with both pro- and anticompetitive effects . . . virtually no studies claim to have identified instances where vertical practices were likely to have harmed competition”); Lafontaine & Slade, Exclusive Contracts and Vertical Restraints: Empirical Evidence and Public Policy in Handbook of Antitrust Economics 391 (Paolo Buccirossi ed., 2008) (“[I]t appears that when manufacturers choose to impose restraints, not only do they make themselves better off but they also typically allow consumers to benefit from higher quality products and better service provision”); O’Brien, The Antitrust Treatment of Vertical Restraints: Beyond the Possibility Theorems in The Pros and Cons of Vertical Restraints 40, 72-73 (2008) (observing that “with few exceptions, the literature does not support the view that [vertical restraints] are used for anticompetitive reasons”); Sass, The Competitive Effects of Exclusive Dealing: Evidence from the U.S. Beer Industry, 23 Int’l J. Indus. Org. 203 (2005) (concluding that exclusive dealing in the beer market increases market output).
The four things we know about exclusive dealing should influence the legal standard governing the practice. Since both theory and evidence suggest that most instances of exclusive dealing are output-enhancing, the presumption should be no liability, and the burden should rest squarely on plaintiffs to show actual or likely harm to competition itself (i.e., to overall market output), not merely injury to a competitor.
That is the approach the Supreme Court ultimately endorsed and the FTC once followed. Whereas the Supreme Court’s 1949 Standard Stations decision condemned exclusive dealing almost per se, the Court eventually instructed that courts should make a fuller inquiry into the actual competitive effect of the challenged exclusive dealing activity. See Tampa Electric Co. v. Nashville Coal Co., 365 U.S. 320, 329 (1961). In Beltone Electronics, 100 F.T.C. 68 (1982), the FTC followed Tampa Electric’s instruction, ultimately concluding that Complaint Counsel had not proven that the exclusive dealing at issues was anticompetitive. First, the Commission concluded that harm to competition had not been established, despite some apparent market foreclosure, because “other firms ha[d] recently entered the market or grown vigorously.” Id. at 210. Moreover, the defendant demonstrated a procompetitive benefit stemming from its exclusive dealing: the arrangement enabled a program in which the defendant producer would generate sales leads from potential customers and pass them along to its distributors without fear that the distributors would then direct the customers to different, higher margin brands. Id. at 216. In sum, the Commission reasoned that exclusive dealing should pass muster, despite some apparent foreclosure, when actual market experience shows continued competition through entry and the exclusive dealing facilitates some sort of output enhancement.
The FTC’s McWane decision substantially departed from the Beltone approach. In that case, the Commission condemned an exclusive dealing arrangement on the basis of a theoretical anticompetitive harm, even though the arrangement had been in place for long enough to generate an anticompetitive effect but had not apparently done so. Complaint counsel claimed that defendant McWane, the dominant producer of domestic iron pipe fittings, had monopolized the market by instituting a “full support policy” under which it would sell its products only to distributors that carried its fittings exclusively. The policy was subject to two exceptions: where McWane products were not available, and where a distributor purchased a McWane rival’s pipe in addition to its fittings. The FTC concluded that the policy caused anticompetitive harm by artificially holding McWane’s rivals below MES, thereby raising their costs and enhancing McWane’s ability to raise its own prices.
There are many reasons to question the Commission’s determination that McWane had engaged in anticompetitive exclusionary conduct. First, the evidence on what constituted MES in the relevant market was remarkably thin, consisting entirely of testimony by rival Star Pipe Products, Ltd. (“Star”) that it would face lower average costs if it owned a foundry but could not justify building one given its low (20 percent) market share. Countering that self-serving testimony were a couple of pieces of actual market evidence. First, the second-largest domestic seller of pipe fittings, Sigma Corp., somehow managed to enter the market and capture a 30% share (as opposed to Star’s 20%), without owning any of its own production facilities. Such success suggested that foundry ownership—and, thus, a level of sales sufficient to support foundry construction—may not be necessary for efficient scale in the domestic pipe fittings industry. So did Star’s ownsuccess. Star entered the domestic pipe fittings market in 2009, quickly grew to a 20% market share, and was on pace to continue growth when the McWane action commenced in January 2012. As dissenting Commissioner Josh Wright observed, “for [the Commission’s] view of MES to make sense on the facts that exist in the record, Star would have to be operating below MES, becoming less efficient over time as McWane’s Full Support Program further raised the costs of distribution, and yet remaining in the market and growing its business. Such a position strains credulity.”
In addition to failing to establish what constitutes MES in the domestic pipe fittings industry, the FTC never adequately established the degree of foreclosure occasioned by McWane’s full support program. In particular, the Commission made no effort to quantify the sales made to McWane’s rivals under the two exceptions to McWane’s full support policy. Such sales were obviously not foreclosed to McWane’s rivals, but the Commission essentially ignored them. Absent information on the volume of distributor purchases under exceptions to the full support program, it is simply impossible to assess the degree of foreclosure occasioned by the policy.
Not only did the Commission disregard deficiencies in the affirmative case against McWane, it also ignored several pieces of evidence suggesting that McWane’s exclusive dealing was not anticompetitive. First, the full support program did not require a commitment of exclusivity for any period of time; distributors purchasing from McWane could begin carrying rival brands at any point (though doing so might cause McWane to refuse to sell to them in the future). Courts have often held that short-duration exclusive dealing arrangements are less troubling than longer-term agreements; indeed, a number of courts presume the legality of exclusive dealing contracts of a year or less. See, e.g., Concord Boat Corp. v. Brunswick Corp., 207 F.3d 1039, 1059 (8th Cir. 2000).
Second, entry considerations suggested an absence of anticompetitive harm. If entry into a market is easy, there is little need to worry that exclusionary conduct will produce market power. Once the monopolist begins to exercise its power by reducing output and raising price, new entrants will appear on the scene, driving price and output back to competitive levels. The recent and successful entry of both Star and Sigma, who collectively gained about half the total market within a short period of time, suggested that entry into the domestic pipe fittings market is easy.
Finally, evidence of actual market performance indicated that McWane’s exclusive dealing policies did not generate anticompetitive effect. McWane enforced its full support program for the first year of Star’s participation in the domestic fittings market, but not thereafter. Star’s growth rate, however, was exactly the same during the enforcement of the policy as it was both before the policy was implemented and after it ended. That suggests that the program had no impact on rivals’ ability to enter the market and grow their business, in which case it could not be anticompetitive.
The Commission also virtually ignored a procompetitive justification for McWane’s full support policy. McWane produced a complete “system”—a full line of domestic pipe fittings—comprised of disparate but complementary parts. Both distributors and end-users have an interest in having ready access to all the parts in such a system, and a full support policy could help ensure such access. Because McWane’s fixed costs (e.g., the cost of casting a die) were similar for both rarely used and popular fittings, McWane’s average production cost for a rarely used fitting [i.e., (fixed costs + variable costs)/number of units produced] was higher than its average cost for an oft-used part. That meant that if McWane charged similar prices for technologically similar parts—a pricing practice purchasers often expect—it needed to “subsidize” production of rarely-used fittings with margins earned on popular parts. An equally efficient producer of only popular fittings would not have to engage in such “cross-subsidization” to finance the production of rarely used parts and would be able to sell its popular fittings at lower prices. But if too many buyers purchased their often-used fittings from the partial line producer, McWane could no longer afford to produce rarely used parts, and gaps in product availability would result.
McWane’s Full Support Program offered a solution to this problem. By requiring buyers of its fittings to refrain from handling those of other producers, McWane could prevent the sort of “cherry-picking” that would have rendered its production of obscure parts uneconomical. Because consumers, distributors, and even other producers of domestic iron pipe fittings all benefit from continued production of a full line of fittings, McWane’s full support program was far from an unreasonable form of competition. On the contrary, it was output-enhancing and thus procompetitive.
In the end, then, the FTC’s McWane decision failed to require adequate evidence in support of the articulated theory of anticompetitive harm, ignored actual market evidence suggesting an absence of such harm, and gave short shrift to an important procompetitive benefit of McWane’s exclusive dealing. The Commission’s apparent and unjustified hostility toward exclusive dealing arrangements is likely to discourage their use—despite their general efficiency—and thereby injure consumers. The Eleventh Circuit should reverse.
CORRECTION from Thom Lambert:
The ninth paragraph of this post refers to market shares of Sigma and Star (30% and 20%, respectively). An astute reader questioned whether the factual record supports those high market share figures. In searching for a record cite, I discovered that I had made a mistake on the actual market share figures. Commissioner Wright’s dissent, upon which I relied, redacts the precise market shares, and I now see from others’ posts that the market shares of Sigma and Star were significantly lower than I reported. In drafting the post, I relied on my own notes, which were obviously mistaken. I apologize for the inaccuracy.