Thom’s excellent post covers most of the important points in Leegin and offers a fairly comprehensive critique of what I deemed to be a surprisingly weak dissent from Justice Breyer. As we’ve noted over and over here at TOTM, the death of Dr. Miles is clearly the right outcome judged based upon the underlying antitrust fundamentals. As Thom and I have pointed out in various posts on RPM here at TOTM, the evidence overwhelmingly suggests that anticompetitive RPM is much talked about but rarely observed or documented. Given that the bulk of the contemporary evidence on RPM suggests that it is largely pro-competitive, I must admit that I was surprised by Tyler Cowen’s “casual guess” in a post at the VC that >50% of RPM are associated with attempts to collude.
In any event, I want to add a few of my own reactions to the Leegin decision to Thom’s analysis. I will primarily focus on the economic analysis in the majority opinion and some glaring deficiencies in the Breyer dissent and incorporate by reference Thom’s discussion of the stare decisis arguments which he discusses in great detail.
Kennedy’s Economic Analyis
Kennedy gets the economic analysis absolutely correct in large part because the opinion mirrors the structure and content of the Economists’ Brief. Part III.A of the opinion begins by recognizing that “the economics literature is replete with procompetitive justifications for RPM” and spelling out those justifications with citations to the literature. Part III.B goes on to note the possible anticompetitive effects of RPM and Part III.C follows with what has always been the punchline for the case against per se treatment of RPM: “it cannot be stated with any degree of confidence that RPM ‘always or almost always tends to restrict competition and decrease output.'”
The straightforward adoption of the Economists’ Brief analysis of RPM, in my view, is a good thing. For all of Justice Breyer’s complaints regarding the Court following the whims of economists who sometimes disagree with each other — there is virtually zero disagreement between economists concerning whether RPM always or almost always reduces output. Virtually every economist agrees that the answer is “NO.” Justice Breyer’s insinuation that economists disagree about RPM is misleading. The disagreement is over the weight that different explanations of RPM should be given. There simply is not disagreement over whether the collusion explanation meets the always or almost always standard. None.
The fundamental question which requires an answer in order to establish whether minimum RPM meets the “always or almost always” standard is “how often does RPM lead to anticompetitive effects in the real world?” The answer obviously requires reliance on empirical evidence. A fundamental divide between the majority and dissenting minority in Leegin is the view of the state of empirical evidence. I will return to this point below.
Recognition of the Role of Vertical Restraints in Facilitating Promotional Services Without Inter-Dealer Free-Riding
There is one particularly interesting point that I was thrilled to see recognized by the majority. As I have noted here before, the argument that “discount dealer” free-riding (a la Lester Telser) is the primary motivation for RPM has been a distraction. Since Klein and Murphy’s seminal 1988 JLE article on Vertical Restraints as Contract Enforcement Mechanisms, it has been accepted in the economics literature that RPM and other vertical restraints play a role in solving incentive incompatabilities between manufacturers and retailers over promotional services even in the absence of a inter-retailer free-riding problem. This point has long been ignored in antitrust cases, in my view, to the extreme detriment of the analytical coherence of our vertical restraints doctrine. Fortunately, again following the guidance of the Economists’ Brief, Part III.A notes that “RPM can also increase interbrand competition by encouraging retailer services that would not be provided even absent free-riding” (citing Klein & Murphy and Mathewson & Winter).
This point was not critical to reaching the correct decision in Leegin, but helps bring some economic clarity to vertical restraints doctrine that has been missing while the discussion incorrectly focuses on the famous “discount dealer” problem that probably describes a very small fraction of actual RPM usage. Discussion of RPM and other vertical restraints has been hampered by the inclination to try to fit a square peg in a round hole by forcing the “discount dealer” explanation into places where it simply doesn’t fit. Pitofsky noted this very early in his critique of this defense of RPM. Understanding the fundamental nature of the use of vertical restraints to enforce contracts over promotional services is critical to understanding a number of other contracting practices that attract antitrust scrutiny, e.g. slotting allowances, category management, and exclusive dealing. I hope that this portion of the opinion does not go unnoticed.
The Role of Empiricism
As I allluded to above, the critical question here is empirical and does not revolve around theoretical explanations of RPM. How often is RPM anticompetitive? Is it anticompetitive in the form of reducing output often enough to meet the “always or almost always” standard required to deviate from the rule of reason default? Kennedy’s majority opinion mentions the Overstreet and Ippolito studies which both suggest that anticompetitive RPM is rare. Breyer takes a shot at these studies in the dissent, noting that they focus on allegations of collusion rather than actual collusion, but this seems like a weak objection to me for several reasons. First, the burden of proof to show anticompetitive effects is on Breyer. Second, that allegations of collusion in antitrust cases rarely occurred in cases involving RPM may not be perfect evidence, but it is probative of the fundamental question. There are other more recent studies that support the pro-competitive RPM explanations that Kennedy does not cite, unfortunately.
In any event, it is worth noting that this attachment to empiricism in antitrust cases appears to be a trend in recent cases. The Court expressed a great desire to reconcile antitrust law with the empirical reality that patents do not confer market power in Independent Ink, and now in Leegin again affirmed its desire to square doctrine with empirical reality and economic theory.
Analytical Weaknesses in Breyer’s Dissent
I was truly surprised by what I take to be a rather weak dissent by Justice Breyer, a man who genuinely understands antitrust economics and principles. In fact, Justice Breyer begins his dissent by recognizing the “always or almost always” standard that must be satisfied in order to apply the per se rule (in the absence of overriding stare decisis concerns). Holding aside the stare decisis argument, Justice Breyer’s analysis of RPM and assessment of the evidence just doesn’t make economic sense. The dissent would have made a great deal more sense if Breyer would have just hung his hat on the stare decisis point and called it a day. He didn’t and it weakened the argument substantially.
Let’s start with the evidence. Justice Breyer’s discussion of the empirical evidence here strikes me as disingenuous. At page 5, Breyer discusses the incidence of anticompetitive RPM and hangs his hat on a thirty year old study that compared retail prices across states after the repeal of the Miller-Tydings Fair Trade Act which found that retail prices were higher by 19-27%. First, as discussed in the Economists Brief’ and elsewhere, these studies do not control for anything. And as discussed above, there is better evidence available.
Second, there is another glaring weakness in a study purporting to identify anticompetitive RPM that looks only at retail prices: BOTH procompetitive and anticompetitive theories of RPM predict higher prices!!! The key question is whether RPM reduces market output. A study that looks at retail prices simply cannot disentangle the competing theories. These are issues that were raised in oral argument and in the briefs and that Justice Breyer is well aware of and far too savvy an antitrust thinker to ignore.
Third, Breyer minimizes the import of the Overstreet and Ippolito studies because the identify only allegations of collusion & RPM rather than actual instances. Again, I note here that this is a rather weak objection. The burden is to present evidence that RPM always or almost always is anticompetitive. The argument that studies that show collusion is rarely alleged in antitrust cases involving RPM may reduce the weight one would like to give this evidence, but does nothing to further the case that RPM meets the per se standard.
The 1975 study that Breyer points to cannot alone meet this standard. As noted, a study only looking at retail prices that does not control for any differences between states is not capable of doing so. Unfortunately, Breyer treats the study as sufficient to meet this standard. This doesn’t meet the laugh test.
The second analytical weakness of Breyer’s dissent involves his economic analysis of RPM in the first instance. Breyer, starting at page 9 of the dissent, responds to the “discount dealer” free-riding argument by repeating the mantra that Pitofsky made famous: the free-riding justification is of limited utility because it doesn’t explain a number of the cases and we don’t know how often it occurs. This is well and good. But the justifications for RPM are not limited to that explanation, as noted in the majority opinion (and by extension, the FTC/ DOJ Brief and the Economists’ Brief). As noted above, a key explanation for the use of RPM is the Klein & Murphy explanation that RPM may be used to enforce efficient contracts involving promotional services or other non-contractible elements of performance. Here is Breyer’s response:
“The one arguable exception consists of the majority’s claim that ‘even absent free-riding,’ RPM ‘may be the most efficient way to expand the manufacturer’s market share by inducing the retailer’s performance and allowing it to use its own initiative and experience in providing valuable services.’ I cannot count this as an exception, however, because I do not understand how, in the absence of free-riding (and assuming competitiveness), an established producer would need RPM. Why … would a dealer not expand its market share as best that dealer sees fit, obtaining appropriate payment from consumers in the process? There may be an answer to this question. But I have not seen it.”
I don’t know whether this portion of the opinion is disingenuous or refreshingly honest. Maybe Justice Breyer doesn’t “get” the explanation offered in the Economists’ Brief and in the majority opinion articulating the use of RPM to induce promotional services. It’s complicated. But he hasn’t seen the explanation? I doubt it. The majority cites the Klein & Murphy and Mathewson and Winter articles as explaining the mechanism by which RPM induces promotional services without free-riding. The Economists’ Brief explains the basic economic mechanism at work in those articles in an accessible manner. That RPM and other vertical restraints can be used in this manner has been a well accepted portion of economic knowledge for 20 years! In any event, I am quite pleased that the majority did “get it” because this is a fundamental economic point. But I admit that I am surprised that Justice Breyer rejected an efficiency explanation for RPM that is well accepted in the economics literature on these grounds.
I see in Leegin a desire by the Court to square economic theory and empiricism with outdated antitrust doctrine. This is an excellent development and I am hopeful it will continue and spread to other incoherent areas of antitrust in the years to come. Finally, it will be interesting to watch the aftermath of Leegin in the states and perhaps in Congress. Dr. Miles may be dead. But it is a bit early to say goodbye to the per se rule against RPM altogether!