Yesterday, in my contribution to the Antitrust & Competition Policy Blog’s Section 5 symposium, I discussed the FTC’s use of Section 5 to evade the tough standards facing plaintiffs bringing Section 2 claims and how that evasion was likely to cost consumers by stripping out the error-cost protections embedded in modern monopolization law. I also argued that the Commission’s various justifications for bringing the case under Section 5 were both unpersuasive and unprincipled. Some of the justifications are to do with the general trend towards favoring Section 5 as a stand alone authority, others rely on the institutional expertise of the Commission relative to judges in federal district court, and still others on the nature of competition in the microprocessor market, e.g. Commissioner Rosch’s claim that the difficulty in distinguishing harm to competitors from harm to competition in this setting supports a Section 5 case.
These purported justifications got me thinking. There are competing theories of the Commission’s reliance on Section 5. One is that the choice of statute is driven by dissatisfaction with the error cost protections afforded consumers by demanding rigorous proof of consumer harm under Section 2, the appeal of the evading those requirements, and the various procedural benefits of keeping things “in house.” The other school of thought is that the Commission, in cases like Intel, is taking a principled approach wherein some features of the Intel case (as Commissioner Rosch argues) favor application of Section 5. I’ve already tipped my hand here somewhat that I think the proffered justifications don’t carry much water. But what would be really nice support for the Commission’s argument is that in similar cases involving vigorous competition between a dominant player and challenging incumbent, involving the same sort of loyalty discounts and market share discounts as the contracts forming the basis of the Intel complaint, the Commission invoked Section 5 on similar reasoning.
So when else has the Commission, faced with business practices involving aggressive competition for distribution, including market share discounts, turned to a less stringent statutory authority than Section 2?
FTC v. McCormick. The similarities between McCormick and Intel are interesting. Both involve competition for distribution. Both suppliers arguably have monopoly power in plausible relevant antitrust markets. Further, the conduct in each case involves “market share discounts,” or contracts that conditioned the granting of discounts on retail distributors agreeing to commit high percentages of their shelf space to McCormick spice products. In McCormick’s case, in response to a vigorous price war with its leading rival Burns, McCormick offered significant discounts to its retailers in exchange for shelf space commitments approaching 90 percent in some cases. Like Intel, the contracts at issue appear to be just a small fraction of the total market. In McCormick, for example, out of 2000 or so contracts with retailers, the Commission focused on a handful: “in no fewer than five instances . .. by providing different deal rates consisting of preferential upfront ‘slotting’-type payments or allowances, discounts, rebates, deductions, free goods, or other financial benefits to some purchasers of McCormick products including, but not limited to, McCormick’s spice line.” The FTC’s Complaint against McCormick, like the Complaint against Intel, reads like a conventional monopolization claim:
McCormick has commonly included provisions that, much as is sometimes seen with slotting allowances, restrict the ability of customers to deal in the products of competing spice suppliers. Such provisions typically demand that the customer allocate the large majority of the space devoted to spice products – in some case 90% of all shelf space devoted to packaged spices, herbs, seasonings and flavorings of the kinds offered by McCormick – to McCormick.
So what did the Commission do in McCormick? Did it bring the case under Section 2? Section 5? Or even under a primary-line Robinson-Patman Act claim that would have also required the Commission to meet the stringent Brooke Group requirements? Nope. Instead, over the dissent of two Commissioners, McCormick ultimately agreed to a consent order with the FTC preventing its use of discriminatory prices between retailers, i.e. secondary-line injury under the Robinson-Patman Act which required no actual showing of loss of consumer welfare.
In my paper on Antitrust Law and Competition for Distribution, I criticized the FTC for bringing (and extracting the settlement) McCormick under the discredited Robinson-Patman Act at all, but even more worthy of criticism was the decision to evade the traditional consumer welfare standards in favor of the secondary-line injury theory which required no such proof:
The availability of the Act’s weaker standard suggests a backdoor for plaintiffs unable to meet the more stringent burden of proving competitive injury in a monopolization or primary-line claim. While it is doubtful that the FTC’s prosecution of McCormick represents a revitalization of the Act, the prosecution does create uncertainty for manufacturers in the growing number of industries relying on retailer promotional effort and product placement for sales.
Sound familiar? Here is what I wrote about the Commission’s more recent invocation of a weaker substantive standard when it is unlikely to prevail under Section 2:
The Section 5 enforcement action will cost consumers (win or lose) in at least two ways. The first is that Commission will expend significant resources litigating a case with Intel involving conduct that has already been limited by a private settlement, exploiting resources that could be used to tackle other (error-cost justified) problems. The second is that the Commission’s invocation of (and awkward justification for) Section 5 will result in uncertainty which will chill some pro-competitive conduct, including discounting behavior by firms in high-tech industries and across the economy.
Whatever one thinks about the competitive merits of Intel’s underlying conduct, the Commission’s use of Section 5 should be seen for what it is: an attempt to evade requirements to demonstrate consumer harm under Section 2 that exist to protect consumers from the social costs of false positives. Such an approach is bound to harm competition and consumers in the long run because it gives the Commission the option to apply its “watered down” standard to whatever business conduct it views as potentially problematic. This approach is a recipe for Type I error and should be rejected by fans of consumer-welfare based antitrust policy.
The justifications offered for bringing McCormick under the Robinson-Patman Act, like the justifications offered for bringing Intel under Section 5, are not persuasive. And in my view, the fact that the FTC brought claims in these two cases involving very similar conduct under two different statutes over a ten year span suggests to me that the Commission does not have a coherent analytical theory of its approach to competition for distribution cases. So why the RP Act in McCormick and Section 5 in Intel? My suspicion is that on the one hand the current debate surrounding the appropriate scope of Section 5 had the Commission looking for a good, high profile case to test its Section 5 authority, write an “expert” Commission opinion on the subject and attempt to persuade the federal court of appeals that eventually hears the case about the right approach to loyalty discounts under Section 2 and 5. On the other hand, the Robinson-Patman Act has only become increasingly discredited since McCormick in 2000 with the AMC repeal recommendation. But I remain unpersuaded that the Commission’s choice to bring the case under its Section 5 authority is a principled one.
To borrow and put a twist a line from Justice Potter Stewart, the only consistency I can find in the Commission’s approach to competition for distribution and market share discounts is that it chooses to avoid Section 2!