Economic theory is essential to antitrust law. It is economic analysis that constrains antitrust law and harnesses it so that it is used to protect consumers rather than competitors. And the relationship between economics and antitrust is responsible for the successful evolution of antitrust from its economically incoherent origins to its present state. In my view, which I’ve expressed in greater detail elsewhere, the fundamental challenge for antitrust is one that is created by having “too many theories” without methodological commitments from regulators and courts on how to select between them. The proliferation of economic models that came along with the rise of Post-Chicago economics and integration of game theory into industrial organization has led to a state of affairs where a regulator or court has a broad spectrum of models to choose from when analyzing an antitrust issue.
This may have been a positive development for economic science. This is not the place to have that debate. But for law and economics the proliferation of theoretical models without attention paid to empirical testing of models, combined with the above-mentioned “model selection” issue allow regulators and courts significant degrees of freedom to select any model they like — why not the one that matches their ideological prior beliefs or policy preferences? Taken to the extreme, this model selection problem threatens to strip the disciplining force that economics has placed on antitrust law that was a key part of the successful evolution of that body of law over the last fifty years. The stakes involved in appropriately specifying the link between economic theory and antitrust law, and understanding the role of empirical evidence in that relationship, are high.
The importance of the issue is why I have criticized both what I view to be a trend toward reducing the role of economics and economists in antitrust, as well as those who turn the issue of model selection into an ideological debate rather than one driven by economic theory and evidence. Readers of TOTM will know that I have, on occasion, criticized Commissioner Rosch on these grounds. Well — apparently he was reading (see n.3) and has devoted some time in a recent speech to “dispelling a misconception about how I view economics and economists, attributable I fear to not making my views clear.”
I applaud Commissioner Rosch for recognizing the confusion that some of his views on economics may have caused and further applaud him for attempting to set them straight in a public forum. Transparency is useful in contexts such as these precisely because it allows these views to be strained and tested by those who disagree (and also because it puts the antitrust and business communities on notice). Here is what the Commissioner had to say about his critics:
More specifically, critics – overwhelmingly Chicago School apologists – have suggested that I am anti-economics and anti-economist and, indeed, that I doubt that economics should play a role in antitrust law enforcement. Those are not my views. My views are as follows….
If one follows n.3 in that speech, despite the use of the plural, the reference is apparently one that is exclusive to me and this series of posts on Commissioner Rosch’s treatment of economics, economists, and the role of economic theory in antitrust:
Commissioner Rosch on the Smaller Role of Economists in Antitrust Litigation
Inter-Agency Scuffling Over Section 2: What Role for Economists and Economics at the FTC and DOJ?
Commissioner Rosch v. Economics Again
And here are a few that the Commissioner didn’t cite for the sake of completeness:
No Ovation for the FTCs Latest Enforcement Theory
Is the Chicago School Really Dead? How Do You Know?
In a later speech on the “Redemption of a Republican,” in which the punchline is that the “confessions of sin” from Posner and George Osborne in the UK redeem Rosch’s views and might even save us from clinging to “bankrupt” economics, the Commissioner refers to his “orthodox Chicago School critics” without identification or citation, writing that:
The orthodox Chicago School economist community has been especially dumbfounded. Some economists have denounced my remarks questioning the use of economic formulae as reflecting a general bias against economists. With specific reference to my January remarks, they have both asserted, on the one hand, that the current economic crisis says nothing about microeconomics as opposed to macroeconomics and at the same time have denied that any Chicago School economist has ever asserted that markets are perfect or self-correcting or that businesspeople are rational. They have also asserted that most of the decent post-Chicago School economics thinking has come from orthodox Chicago School economists. After all of this criticism, I was starting to question whether I really was a loyal Republican.
I suspect I’m also the dumbfounded denouncer — though I refer readers to this post to see what I actually said (see also here) because it differs substantially from Rosch’s characterization (if indeed it purports to summarize my positions). Suffice it so say that one can read very closely and there is no claim there about what Chicago School economists have said or not said about businesspeople being rational.
As for the “assertion” that most of the decent post-Chicago thinking has come from orthodox Chicago School economists — again, this is simply untrue and misleading. Instead, the claim is that folks like Aaron Director anticipated many of the economic insights of the raising rivals’ cost models, that Chicagoans such as Ben Klein and Tom Hazlett are responsible for some of the leading empirical examples of Post-Chicago phenomena, and that Dennis Carlton (Chicago GSB) extended some of the theoretical work (for more on this, see here). Don’t, however, take it from a Chicago School apologist. Instead, here it is straight from the leading thinker of the Post-Chicago economic movement: “it is important to recognize that [the Post-Chicago] approach has its root in the economic analysis of Chicago School commentators.” See Steven C. Salop, Economic Analysis of Exclusionary Vertical Conduct: Where Chicago Has Overshot the Mark, in OVERSHOT THE MARK, at 144. Similarly, try to talk about coordinated effects without invoking Stigler (1968).
This is why the prefix “orthodox” in front of Chicago School is misleading — though perhaps an effective rhetorical device. Chicagoans have done important empirical work on market failures and helped to develop and test Post-Chicago models. By what meaningful definition does this constitute “orthodoxy”? Nevermind the persistent use of the term orthodox to imply a static, monolithic body of economic knowledge. This term by colloquial implication glosses over important differences between the work and approaches of, for example, Alchian, Klein and Oliver Williamson and in understanding how asset specifity and opportunism influence firm behavior, or between Klein, Telser and Marvel on vertical restraints, or between Demsetz and Coase on the theory of the firm. The label both misleads, errs as a matter of economic history, and favors ideology and shorthand labels over substance. None of this, of course, is new to political rhetoric. To some, to invoke the term Chicago School is not to reference a broad intellectual movement but to utilize a heuristic to describe a reflexively anti-interventionist position that typically involves (in the eyes of the evoker) irrational disdain for government regulation. While current financial times make it somewhat fashionable for journalists and casual observers to toss around without regard to accuracy or evidence caricatured versions of entire schools of economic thought (to which serious scholars have devoted decades of intellectual energy), antitrust experts have generally carefully avoided such style of commentary in favor of careful analysis of competing theories and evidence.
One more quick — but related — about misrepresenting views. Rosch describes George Stigler’s work on oligopoly theory as follows:
Nobel Prize winning economist George Stigler’s 1964 article “A Theory of Oligopoly” in which he explained that it was improper to assume that firms in an oligopolistic market would find a way to agree to raise prices above competitive levels.
This is wrong. Stigler said it was wrong to assume that firms would inevitably collude. Rather, he proposed a framework for analyzing the factors that would affect their ability to do so. That framework provides the basis for much of our modern antitrust understanding on collusion and coordinated effects. It is simply wrong to insinuate that Stigler thought it was impossible for oligopolists to collude successfully.
Ultimately, I’m far more concerned with the misrepresentation of ideas, theories and evidence than Commissioner Rosch’s description of me as a “Chicago apologist.” The label Commissioner Rosch assigns to me or others does not resolve the substantive merits of the “model selection” problem or the fundamental question here of whether Rosch’s ideas on the relationship between economics and antitrust are sensible. It’s true that I’ve written as a positive matter that the Roberts Court’s antitrust jurisprudence is more Chicago than Harvard. I’ve criticized those who have caricatured and Chicago School scholars and ideas and avoided taking on the substantive merits of those economic ideas in favor of ideological labels. I’ve also argued that the appropriate way to settle intellectual battles in antitrust between competing models is with reference to the empirical evidence — and that attempts to explain the persistence of Chicago School economics in antitrust jurisprudence without rejecting the hypothesis that it is the body of economic theory that is most consistent with the available empirical evidence is unlikely to lead to good antitrust policy.
In short, an analysis of the existing theory and evidence that suggests that the Chicago School’s contributions to antitrust economics hold up quite well relative to competing theories when the contest is run with reference to empirical data rather than who shouts the loudest is not an apology as I understand the word. I therefore conclude that I am no Chicago School apologist. Commissioner Rosch obviously disagrees. In this context, this is a rhetorical and political term not an economic one, so I don’t have much else to say about it — well, maybe one thing. Apologist is not a neutral word. I point this out because the use of the term “apologist” contradicts the Commissioner’s anecdotal introduction to his speech wherein he analogizes the intellectual debates in antitrust to the scientific debates over quantum mechanics and general relativity in characterizing the relationship between atoms and sub-atomic particles. The analogy appears to hint that the Commissioner endorses an approach grounded in the scientific method to settle these disputes. But the use of rhetoric like “apologists” to describe critics that have taken on the substance of ideas and evaluated the evidence in a manner consistent with social science methodology dispels that notion and suggests mostly interest in superficial labels and t-shirt slogans rather than serious engagement with ideas and evidence. Its use does not imply that Commissioner Rosch and I simply have different views on the relevant economics that we can debate on the substantive merits. Rather, it implies that I know that the competing economic approaches he advocates are superior, but I am in a state of denial or perhaps that I am just being intellectually dishonest about the Chicagoans losing the war of ideas in antitrust economics. I do not read the term as a neutral one. It’s important, however, not to get too distracted by name calling. As name calling is a tactic typically relied upon to avoid delving deeply into the substantive merits of an issue, its use signals that it is especially important to ignore it here. But I will say that, given its use, I’ve granted myself permission in this post to treat the Commissioner as a hostile witness, as it were.
The relevant question is not whether Commissioner Rosch’s views on economics and antitrust law and whether they do in fact, or do not, have those qualities that I’ve criticized in previous posts. I’ll focus on that issue in this post, and will argue that Rosch’s attempt to clarify his views on the role of economics and economists on antitrust was unsuccessful in the sense that 1) those views remain unclear in many instances, 2) are a primary example of the “model selection” problem described above, 3) will lead to overly aggressive antitrust enforcement, and 4) thus I conclude are likely to result in significant potential to harm consumers if they express agency enforcement priorities or policies.
Let’s begin with Rosch’s basic argument — though readers should take a look at the entire speech — which makes three major moves. The first is to discredit Chicago School economics, the second is to link the current financial crisis to Chicago School economics, and the third is to argue that a compilation of alternative theories does a better job of explaining the real world than the Chicago theories and therefore should form the basis of antitrust policy. In my discussion, I’ll collapse the first two moves into a single argument that the financial crisis proves the Chicago School is dead or discredited. At each of these stages, Rosch makes critical errors of commission or omission that either undermine the argument or create significant confusion. We’ll discuss some of these here.
DOES THE FINANCIAL CRISIS PROVE THAT THE CHICAGO SCHOOL IS REALLY DEAD?
Let’s begin with the obvious point: the appropriate way to judge Chicago School antitrust economic theories is by how well they predict firm behavior relative to alternatives. I understand the rhetorical attraction of pointing to folks like Judge Posner and Alan Greenspan to argue that presumptions that markets work are misplaced. How much of the current financial crisis to blame on markets versus government is a bit outside the scope of this post, but given the attention that both Greenspan and Posner got in the media for these statements, I certainly understand that they can be effective. But these statements are not evidence. Nor is a headcount of public figures and intellectuals the best approach to assessing evidence. I imagine this approach was not how the physicists settled debates about quantum mechanics and relativity.
What’s a better and more scientific approach? There is a substantial body of empirical evidence on business conduct that is antitrust relevant. Economists have gone out and studied how these contracts work in the market and their competitive consequences. Let me take this point a little bit further. Assume that Commissioner Rosch is right that what we’ve learned from the financial crisis and confessions of the likes of Posner, Osborne, Greenspan, Paulsen and whomever else is that markets are fragile, entry solves fewer problems than the Chicagoan models predicted, and vertical contracts exclude more often. That’s testable. A proper economic approach would be to test that belief against the existing data. Let’s take vertical integration and contractual restraints as an example. If Rosch is right about the newly understood fragility of markets to exclusionary practices, we ought to see them out there, right? Empirical evidence ought to be easy to find. But what does the data say?
Recent comprehensive literature surveys from very well respected industrial organization economists, including some who have spent significant time at the enforcement agencies, conclude that the best available empirical evidence is that most forms of single firm conduct are predominantly pro-competitive in practice (See great slides here from Lafontaine & Slade, Dan O’Brien or the paper from Cooper, et al.). Lafontaine & Slade conclude, after synthesizing and summarizing the evidence in dozens of studies, that:
the empirical evidence leads one to conclude that consumer well being tends to be congruent with manufacturer profits, at least with respect to the voluntary adoption of vertical restraints. When the government intervenes and forces firms to adopt (or discontinue the use of) vertical restraints, in contrast, it tends to make consumers worse off.
Cooper, et al. (a paper written by a team of agency economists, not an orthodox Chicagoan among the bunch) evaluate the existing theory and evidence and conclude similarly:
Empirical analyses of vertical integration and control have failed to find compelling evidence that these practices have harmed competition, and numerous studies find otherwise. Some studies find evidence consistent with both pro- and anticompetitive effects; but virtually no studies can claim to have identified instances where vertical practices were likely to have harmed competition.
At the very minimum, this state of evidence should give Rosch significant pause with respect to his being “certain that the Chicago School does not even accurately portray how buyers or sellers behave.” The financial crisis does not change the state of evidence or relative empirical support for Chicago versus Post-Chicago, behavioral or any other theories of vertical restraints or other forms of conduct. One must ask the question: if Commissioner Rosch is right that the financial markets have taught us that there is more anticompetitive conduct out there than we’ve recognized because we’ve been blinded by orthodox Chicago theories, where — and what — is the evidence?
I must mention in passing one other thing. The specific markets that failed in Greenspan and Posner’s telling happen to be extraordinarily-regulated, heavily politicized industries. In Posner’s telling in particular, the failure of capitalism that he describes is at least as much a failure of government as of firms. Perhaps it is the case that the particular market failures that have led to the financial crisis are the same as the sorts of failures relevant to antitrust theory. I doubt it (for example, it’s hard to see how excessive executive compensation would lead to less rational corporate players with respect to entry, but executive compensation has, in many tellings, taken on a central role in the market failures attributed to the financial crisis. Moreover, if we are going to take the market failures of the crisis to tear down the Chicago School edifice, shouldn’t we also at least nod in the direction of the government failures as well? Perhaps, again, the parallels are weak, but immense amounts of government oversight of the financial sector failed to prevent-and may have exacerbated-the crisis. Isn’t it possible that increased government oversight in the antitrust arena might also be less than perfect? In short, whatever the crisis says about “Orthodox” Chicago School economics as applied to financial market regulation, it’s not at all clear that it says anything about its application to antitrust law.
DO ALTERNATIVE ECONOMIC THEORIES DO A BETTER JOB OF EXPLAINING ANTITRUST RELEVANT BUSINESS ACTIVITY?
The second major move in Commissioner Rosch’s argument is to present alternative theoretical frameworks that do not suffer from the (alleged) manifest errors of the Chicago School. He lists three: (1) Post-Chicago economics involving game theoretic models, (2) what he describes as “experimentation” theories, citing Farrell (2006) and (3) behavioral economics. In Rosch’s view, the Post-Chicago models have in common with Chicago School economics that they assume profit-maximization by firms and rational behavior by consumers whereas the “experimentation” and behavioral theories deviate from the rationality presumption in one way or another.
Again, an evidence based and social scientific approach to evaluating competing theories is to compare the predictive power of those theories with respect to the phenomena studied. In other words, one can derive testable implications and evaluate the theories against the existing evidence. I incorporate by reference here the discussion above with respect to the state of evidence on vertical restraints and single firm conduct as an example that defies Rosch’s declaration of the Chicago School’s death with respect to antitrust economics. Tellingly, Commissioner Rosch doesn’t cite to a single empirical study to defend the proposition that these theories outperform the Chicago School theories by this measure. Indeed, empirical evidence appears not to play a significant role in selecting the best model upon which to base antitrust policy. One thing is certain, though, we don’t need evidence to know that the Chicago School is dead because Richard Posner wrote a book that said something about market failure.
QED. Redemption complete.
Commissioner Rosch’s attempt to clarify his views on the role of economists and economics in antitrust, particularly in the discussion about alternatives to the Chicago School, leaves me with more questions than answers. I’ll share some of those questions here to give a flavor, but encourage readers to take a look at the entire speech.
Rosch and Post-Chicago Economics
Let’s begin with the obvious tension between the Commissioner’s glib dismissal of the Chicago School with: (1) the fact that the founding father of Post-Chicago economics has recognized in published work that the roots of the raising rivals’ cost theories exist in Aaron Director’s work (which obviously doesn’t diminish importance of the RRC contributions), (2) the empirical and theoretical contributions that Chicagoans have made to the literature studying exclusion and predation, and (3) the empirical evidence suggesting that whatever one’s priors about the relative explanatory power of Chicago and Post-Chicago models. With the concession that much more empirical work must be done in this area, it is virtually impossible to conclude after a serious analysis of the literature that the “Chicago School is … dead.”
I’ve written extensively on the relative predictive power of Chicago and Post-Chicago theories of vertical restraints here and here, and refer readers to those papers for more detailed discussion. There is also the inconsistency of the expression of inability to “understand complex economic formulae” with the adulation of Post-Chicago game theoretic models involving very sophisticated models, assumptions, and mathematics. Is the Commissioner saying that sophisticated economic formula are too complicated to be useful, but that he understands the mathematics in these models well enough to know what their conditions and assumptions are, when they fit the real world data in a given case, and the impact on consumer welfare? And, of course, to know without a single reference to the empirical literature that these models have greater predictive power? These are rhetorical questions, of course. My tentative hypothesis is that Commissioner Rosch’s attachment to these economic models is out of a desire to select a model to match a policy preference and not the model that best fits the data.
Meanwhile, one can most certainly be a defender of Chicago School antitrust economics without being an inveterate adherent to the use of economic formulae in court. In fact, the heart of the Chicago School, as Rosch himself hints at by pointing out that some if its most important texts, those by Posner and Bork, contain no economic formulae, is not mathematical but analytical. Indeed, economic analysis can lead to a prescription for favoring bright line safe harbors that do not place large burdens on courts to engage in economic analysis. It incorporates a way of thinking, a manner of understanding behavior and responses to incentives. These can, sometimes helpfully, be reduced to mathematics. But that is not the core of the discipline, and certainly not when compared to the essence of Post-Chicago economics.
Commissioner Rosch Goes Austrian (But Doesn’t Know It?)
Perhaps the most confusing portion of Rosch’s speech is the discussion of the second altenative economic theory to the Chicago School. Rosch describes this theory as the “experimentation” theory, and credits it to a paper David Teece wrote on dynamic competition, innovation and antitrust for the GMU/Microsoft Conference on the Law and Economics of Innovation (that Geoff and I organized) along with another article by FTC Bureau of Economics Director Joe Farrell in the Antitrust Bulletin in 2006 entitled: Complexity, Diversity, and Antitrust (which is really a short article based on a lunchtime keynote at an AAI event). Rosch describes experimentation theory as deviating from the stringent assumption of Chicago School and even Post-Chicago economics that firms and consumers are maximizers. Here is Commissioner Rosch’s description:
I call his theory “experimentation” theory. Under that theory, which focuses on the sell-side of markets, most business firms do not engage in behavior that they consider profit-maximizing from the get-go. Instead, Farrell argues, firms engage in a trial-and-error process to identify which conduct will be profit-maximizing for themselves over the long run. Sometimes their experiments are successful, and sometimes they are not.
David Teece has also set forth a similar theory in the context of discussing the economics that underlie innovation. Teece has argued that the concept of static competition, which looks only at price competition by rational agents for existing products, “reflects an intellectual framework” and “not a state of the world.” In contrast, he argues, dynamic competition is driven by the trial and error efforts of innovators and institutional structures that support innovation. Again, speaking frankly, the theories that Farrell and Teece have posited are most consistent with my own real world experience. Perhaps that is because the behavior of most firms is determined by the decisions of middle mangers, not senior executives, much less economists.
Sound familiar? If I understand this correctly, it’s almost as if firms act as if … competition is something like a discovery procedure. Or that competition is an evolutionary process driven by experimentation and trial-and-error. I wonder if anybody has ever said that before? Maybe our friends at Organization and Markets or Austrian Economists can help out the Commissioner with some Hayek and Alchian references. While I’m heartened that the Commissioner Rosch describes these theories as most consistent with his understanding of the way that the world works, its a bit troublesome that he views the policy prescriptions of these models as favoring more rather than less intervention or as favoring policies meaningfully different than those of the Chicago School economics. The parallels between recognition of the value of this sort of experimentation and Austrian economics are clear — but also clearly not what the Commissioner had in mind.
Behavioral Economics and Antitrust
The third set of models comes from behavioral economics, which Rosch argues might be particularly useful in giving us a more robust understanding of consumer behavior but also improve our predictions about firm behavior. Again, this portion of Rosch’s speech leaves more questions than answers. As with the other sections, there is little discussion of empirical evidence that behavioral economic models will improve our predictive understanding of antitrust behavior (see my work here on the relative predictive power of behavioral versus neoclassical economic models in some markets).
Other than the empirical basis for selecting behavioral models rather than others in particular situations, or how to deal with the behavioral economics problem of “competing” behavioral quirks which make prediction difficult, a key question is why the Commissioners appears to believe that a “behavioral approach” or an approach that incorporates the observation that firms might act irrationally in markets leads to a conclusion that more interventionist antitrust is appropriate. Perhaps entry will not occur when it would otherwise be profitable and so a monopolist can exercise monopoly power longer than if one supposed free entry. That’s not, however, the only story one can tell about behavioral quirks. What about the firm that irrationally chooses not to exercise monopoly power when it would be profitable to do so? This is not to say that behavioral economic theories of firm behavior should be irrelevant. Like the other theories, whether Chicago, Post-Chicago, behavioral or otherwise, the key question is one of model selection — which theories have the best empirical support and should therefore guide policy decisions.
There is another issue. One of the fundamental and oft-discussed criticisms of the connection between behavioral economics and paternalism in the literature is the Nirvana Fallacy observation. That is, accepting that it is true that consumers are systematically irrational in identifiable ways that lead to market failures, it still does not follow that regulators will systematically improve outcomes if they too (and judges, and lawyers, and juries…) are also subject to the same irrationalities. It is not enough to justify paternalistic intervention (soft, hard, libertarian, or otherwise) simply to show that consumers make mistakes. The burden is to demonstrate that the government can make better choices for the individual than can the individual can make themselves. In accounting for the long run costs of paternalism, we must also be mindful of dynamic effects that are likely to follow from paternalistic decision-making before intervening (on this last point, see Klick and Mitchell in the Minnesota L. Rev., or more recently Ed Glaeser’s essay on Paternalism and Psychology). Roschs’ inadvertent fondness for Austrian Economics ought to help him to see that it is not immediately apparent that the constrained interventions of regulators will be superior to dynamic, decentralized-even imperfect-markets.
Rosch apparently has conflicting views, or at least views that are not clear, on this aggregation problem of how these biases play out within groups. More specifically, apparently these biases infect profit-maximizing decisions and rationality within firms but apparently not within regulatory agencies:
[Dennis] Carlton also has suggested to me that firms who behave irrationally will end up losing out to profit-maximizing firms in the long run. But that may take a very long time, and consumer welfare may suffer grievously in the meantime. And, if that is so, the complexity and uncertainty associated with identifying and regulating irrational conduct is of no comfort to the Commission as an antitrust law enforcement agency. To the contrary, it raises a number of questions about antitrust law enforcement. The most fundamental of those questions, however, is whether the complexity and uncertainty in this respect mean that the Commission should eschew antitrust enforcement for fear of making a mistake or should make a threshold decision about what kind of behavior by market participants is at issue and then analyze those practices or transactions differently, depending on that threshold determination.
Rosch clearly believes that these behavioral anomalies will result in inefficient firm behavior, but implicitly assumes that regulators are immune or at least less susceptible to succumbing to the systematic biases that these anomalies create within firms. Again, a more rigorous analysis of the lessons for the behavioral economics literature for antitrust is more sophisticated and nuanced than Rosch lets on. Not only might behavioral anomalies that exist in laboratories and market studies lead firms to exclude or engage in anticompetitive behavior less often than they might otherwise, but also one would think that to the extent one is willing to take the lessons of this literature seriously that must include an extension of the same principles to regulators. In fact, there is a literature suggesting that exposure to market competition is likely to mitigate these biases. To the extent that regulators are not exposed to that sort of competition inherently, there is reason to believe that behavioral quirks and biases are more likely to infect decision-making of regulators than firms. Perhaps we’ve inadvertently stumbled on a behavioral economic defense of the existence of multiple antitrust agencies in the US.
Note that this is not an argument about regulator or judicial incompetence. Commissioner Rosch and I actually agree here that antitrust analysis is extremely complex and that there are gains to simplifying economic tests. In fact, I’ve taken great pains and quite a bit of time and energy to work on a project (with former FTC BE Director Mike Baye) that provides some of the first systematic evidence that some antitrust cases are indeed too complex for generalist judges. This is an argument that regulators applying very complex economic concepts will sometimes err and that the right question is not whether regulators should give up because the task is complex. The right question is what the right approach is to balance the social costs of both types of error: false positives and false negatives, as well as administrative costs. The proper role of economics in antitrust is to help determine which approach minimizes those costs and harnesses the power of antitrust to give the greatest rate of return for consumers? Understanding the implications of the economic theory for consumer welfare and the empirical evidence and carefully weighing the empirical evidence is crucial to answering that question, yet each of these play an inappropriately small role in Rosch’s views on the link between economics and antitrust.
FINAL THOUGHTS
Rosch’s discussion of the Chicago School as well as various competing models exposes some inherent tensions and confusion in his views. However, there is a much deeper problem. Recall the fundamental problem facing antitrust in light of the proliferation of theoretical models is that regulators and judges are free to select the model that matches ideological or naive priors without some methodological commitments to evidence based antitrust policy aimed at selecting the models with superior predictive power. Commisisoner Rosch’s discussion of the relationship between economics and antitrust, in my view, exemplify this phenomenon.
For instance, take the Commission’s actions in N-Data, or Ovation. Commissioner Rosch has endorsed an “evasion of pricing constraint” theory of antitrust harm wherein a firm that evades a constraint that influences prices violates the antitrust law. The theory in Ovation, for example, was that Merck was a multi-product monopolist who was constrained in its pricing of Indocin because it was concerned about a reduction in demand for its other products because consumers would be upset if it priced this life-saving drug at “too high” a level. Under the theory, Merck is setting a profit-maximizing price and figures out that Indocin’s profit-maximizing price would be higher because it is unconstrained by these reputational considerations. This “evasion of constraint” theory is gaining popularity at the Commission. Let’s follow the pattern. First, Rambus is based on the concept that evasion of patent disclosure rules in the standard setting context is a violation of Section 2 and Section 5. Second, N-Data is based on the concept that evasion of a contractual pricing constraint in the form of a RAND commitment is a violation of at least section 5 even when the monopoly power is lawfully acquired. Third, Ovation now adds to the list the evasion of reputational constraints on pricing as the genesis of actionable antitrust conduct. As I’ve discussed on the blog previously, despite claims to the contrary, this approach does not have meaningful limiting principles:
Here are a few examples of conduct the FTC could go after that satisfy the “evading a constraint” theory. Why not a monopolist whose pricing is constrained by current demand. That is, the profit-maximizing monopoly price is $20 but the monopolist would REALLY like to charge $25. It is only the fact that current demand is not high enough to support that price that prevents the monopolist from charging it. So, our monopolist comes up with a plan (lets call it a scheme, it sounds worse) to evade the pricing constraint created by current demand by advertising its product to consumers and touting its virtues. Or perhaps its going to invest in the quality of the product. In either event, the purpose of the advertising is to shift the demand to the right and result in higher prices. No matter that output increases, it doesn’t matter because the monopolist is evading a pricing constraint and presumably has violated Section 2. Evading reputational constraints on demand for X is not analytically any different evasion of the constraints imposed on demand by consumer preferences… .
What if Merck evaluated its product line and decided that it would be better off by dropping some of its product portfolio so that it could increase the price of Indocin? Merck’s decision to drop products from its portfolio, or even the design of those product offerings, is surely an evasion of pricing constraints and a violation of Section 2 if it has monopoly power – and perhaps even if it doesn’t under Section 5. Perhaps the Commissioner’s commitments to competition as a discovery process are equivocal? Or what if Merck decided to create a subsidiary to sell Indocin under a different brand name and trade dress to mitigate the reputational costs it would bear from charging a higher price? Or fired the CEO who decided that charging the monopoly price in the first instance would be a bad idea in favor of a new manager who reached a different conclusion and wanted to increase the price? This evasion theory fairly quickly evaporates to the notion that the antitrust law governs prices that are determined to be unreasonable, and not the competitive process itself.
This is what happens when antitrust law unhinges itself from economics, which I fear is a trend at the Commission if N-Data and Ovation are representative.
Do Post-Chicago economic models support this model of competitive harm endorsed in Ovation? Behavioral models? Experimentation? Is this discussion really about selecting the most robust and empirically grounded models to increase predictive power and harness the power of antitrust to maximize consumer welfare or is something else going on: selecting economic models that match priors or policy desires and simultaneously give a “cover” of economic sophistication and rigor? In my view, the evidence is most consistent with the latter possibility.
Predictive power of our economic models matters. Antitrust is, as Dan O’Brien has recently noted, badly in need of an applicability discussion. “Conservative economists” have always highly valued empirical evidence and predictive power. Post-Chicago economists who have contributed game theoretic models of unilateral conduct, collusion, and mergers generally motivate these models with the notion that they will improve predictive power relative to the simple Chicago approach – much like the behavioral models now popular in the L&E literature. Each of these approaches seeks to explain real world phenomena. Our discussions of the “right” economic models should be one that is grounded in taking seriously the testable implications and conditions underlying competing models and testing them against the best available empirical evidence to inform our approach. Progress will not be made in solving the model selection problem without methodological commitments such as these.
In short, there is a significant need for evidence based antitrust. Ideological and political appeals, t-shirt slogans, and adding up statements from public intellectuals like Greenspan and Posner might be sufficient for selling books, political rhetoric and redemption but they are not now, nor will they ever be, a substitute for the more analytically rigorous approach of rolling up one’s sleeves and doing the work necessary to test these competing models to improve antitrust enforcement.. This is what rigorous economics and economists have to offer antitrust.