Increasingly, the notion that updating antitrust policy with the insights of behavioral economics would significantly improve matters for consumers. Others have called for more major surgery, favoring an outright rejection of the current economic foundation of antitrust policy — and especially the portions of the foundation “Made in Chicago” — in favor of a new regime based on behavioral economics. There are plenty of antitrust scholars who’ve begun to make this case, with perhaps Professor Stucke having been the most prolific on this score. And behavioral economics has provided the intellectual support, or perhaps cover depending on who you ask, for the recent regulatory expansion involving consumer credit. The issue is also getting more and more attention. Competition Policy International recently published a symposium issue dedicated to the topic.
One might argue that the intellectual banter about the back and forth of “what school” reigns in antitrust is a waste of time. But that strikes me as inconsistent with the history of the development of antitrust doctrine, which has largely lagged the dominant economic paradigm a few decades. To the extent that the dominant economic paradigm changes, it is certainly quite plausible that we start to see changes in the doctrine. Though, for example, the Supreme Court has largely but not entirely resisted adopting Post-Chicago economics despite the fact that it has certainly established itself as the dominant economic paradigm in economics departments. Nonetheless, I think questions of competing economic models in antitrust and elsewhere can have real consequences — as can questions about how those battles should be resolved (e.g., I make the case here that that the Chicago School has defended its territory largely because the empirical evidence warrants that result).
All of that said, when the chatter about the appropriate intellectual foundation for antitrust shifts from the ivory tower to the regulatory agencies who can do something about it, the issue deserves more attention. Indeed, Commissioner Rosch has been making policy speeches about the value of incorporating behavioral economics into antitrust. The latest, this June 9th speech, gives the most detailed account of “the case” for antitrust law displacing conventional economic theory with the insights of the behavioral economic literature. It is not the first time Commissioner Rosch has sung the virtues of behavioral economics and antitrust. Nor, I suspect, will it be the last. But it is the most detailed intellectual case in support of “nudging” antitrust the Commissioner has offered thus far, and given both the source and level of detail, I think the claims are worthy of serious consideration.
In a series of posts beginning today, I am going to use Commissioner Rosch’s speech to analyze some of the issues involved, while reserving others for an article I am currently working on detailing the uses and abuses (especially the latter) of behavioral economics in antitrust. I’ve written quite skeptically about the incorporation of behavioral economics into antitrust previously. Some of this skepticism derives from documenting abuses of the behavioral economics literature in order to provide intellectual support for policy preferences; some from theoretical or empirical features of the behavioral theories that deserve more serious economic attention than can be supplied in a blog post. With respect to the former, please note that this criticism is one that I do not necessarily level at behavioral economists themselves, but instead those who would incorporate that literature into the law in a haphazard fashion.
In this blog series, I intend to highlight just some of the fundamental theoretical and empirical faultlines in the intellectual foundation of the behavioral antitrust enterprise, at least as it has been articulated thus far by Commissioner Rosch and others. I’ll save many of new arguments for that paper, which I hope will have some influence over the emerging debate over behavioral economics in antitrust and other consumer protection settings.
[Note to law review editors: look for the paper in the fall submission season! Feel free to contact me if you are interested]
But with the case being made strongly by influential antitrust policy makers at the Commission — there is no time like the present to point out some — but not all — of the existing logical and economic flaws in the hopes of starting a serious dialogue about what behavioral economics adds, if anything, to antitrust. And new arguments are not required to expose many of the flaws in Commissioner Rosch’s case for behavioral antitrust. Indeed, his claims fail largely on their own terms. Commissioner Rosch’s speech organizes the case in a reasonable three-part manner: starting with discussing insights from behavioral economics and their antitrust relevance as well as the so-called failure of conventional economics, then turning to and rejecting criticisms of behavioral antitrust, and finally concluding with some views on where all of the aforementioned takes antitrust.
I believe this issue deserves more serious attention than it appears to be getting. I will address Commissioner Rosch’s argument in a three-part series of posts responding to Commissioner’s Rosch’s case for “behavioral antitrust.” By this I mean an antitrust and competition policy regime that incorporates the insights of behavioral economics rather than mainstream economic theory, including neoclassical price theory and to some extent Post-Chicago models as well.
In Part I, below the fold, I’m going to focus on the Commissioner’s description of behavioral economics and what is has to offer antitrust.
I. What is Behavioral Economics? And What Isn’t Behavioral Economics?
Commissioner Rosch starts with a series of reasons that, “according to behavioral economics,” human beings do not maximize their own welfare and, accordingly behave irrationally in commercial decisions. It is worrisome, but evident nonetheless both from the series of reasons and the analysis that follows, that the Commissioner understands neither the behavioral economics literature nor the body of economic theory he criticizes.
First, Rosch points to “asymmetry in the information that is available to both buyers and sellers” as a reason that individuals depart from rational behavior. Rosch argues that:
“More generally speaking, behavioral economics seeks to identify similar instances of asymmetry which prevent perfect decision-making and then, if possible, adjust the default rules to eliminate as much of that asymmetry as possible.”
We’re off to a bad start. There is one condition which, quite frequently, prevents consumers from perfect-decision making: when information is costly to obtain. Perfection is costly. The efficient level of mistakes is not zero. Consumers or firms might – indeed, almost always do – have imperfect information. Buyers and sellers might also have asymmetrical information. This is quite a familiar problem in economics; so much so that it is addressed by any standard microeconomic textbook. If behavioral economics is just imperfect information — perhaps we’ve all been behavioral economists all along. But then what’s all the fuss about? Of course, behavioral economists themselves do not claim that problems of imperfect information require or are necessarily the product of irrationality. Note that this is not a claim that information asymmetries should not lead to regulation under any conditions or anything like that. Its just the simple point that the definition is wrong. Mistakes are taken to be a signal of market failure. Perhaps it is the behavioral economists who should be correcting Commissioner Rosch, but they do not appear interested in that task thus far. To repeat: perfect decision-making is costly. That individuals (or firms) economize on those costs need not have anything to do with irrationality.
This point is not novel. Economists have studied information in this context for a long time.
Consider the following passage:
Ignorance is like subzero weather: by a sufficient expenditure its effects upon people can be kept within tolerable or even comfortable bounds, but it would be wholly uneconomic entirely to eliminate all its effects. And, just as an analysis of man’s shelter and apparel would be somewhat incomplete if cold weather is ignored, so also our understanding of economic life will be incomplete if we do not systematically take account of the cold winds of ignorance.
The author of that passage? George Stigler in his classic paper on the economics of information and search costs that created price distributions. But note one doesn’t need to be a diehard Chicagoan to understand the economics of informational asymmetries. If you are doubtful that the rational choice economic paradigm has anything to say about economics outside of Chicago, how about Stiglitz? Or Akerlof — Lemons, anyone? Or how about Hirshleifer & Riley?
Note, to repeat, none of the above means that informational asymmetries cannot create problems that need solving, either through market mechanisms (Klein and Leffler), or the efficient regulation of consumer information (Beales, Craswell, and Salop). And none of the above is to say that behavioral economics doesn’t have anything useful to say about informational asymmetries. We’re talking about something much more simple here. The point is merely that Rosch’s characterization about what behavioral economics does — “seeks to identify similar instances of asymmetry which prevent perfect decision-making and then, if possible [sic] adjust the default rules to eliminate as much of that asymmetry as possible” — begins by misstating what behavioral economics does and seeks to do, turns to misstating what boring, old, conventional economics has to say about the problem, and in one fell swoop adopts a definition of market failure that appears to be “deviation from perfect competition.”
That’s enough on perfect decision-making as the benchmark for “rationality.” Commissioner Rosch offers three other contributions of behavioral economics. Let’s see if they fare any better.
Second, Commissioner Rosch argues that:
“behavioral economics recognizes that instant gratification is more important than long-run profit maximization for many human beings. This means that, among other things, we demand much more to give up or sell an object than we would be willing to pay to acquire that object.”
OK. This is an improvement. In fact, behavioral economics does have quite a bit to say about discount rates. The definition is a little bit loose, but it’s pretty plain that Rosch means to refer not to impatience generally (plain vanilla neoclassical economists DO talk about time preference), but to hyperbolic discounting, i.e. “present-bias” or the hypothesis that individuals discount the future at a greater rate in the short run than in the long run and therefore make systematically time-inconsistent choices. So let’s not quibble too much. There is a lot to say about the hyperbolic discounting literature, but we’ll leave that for another time and restrict ourselves to the nexus between the first and second sentences above.
Let’s assume Rosch meant hyperbolic discounting here and not just rational discount rates — but what gives with the assertion that hyperbolic discounting is what generates the gap between willingness to accept and willingness to pay? Now, there is some debate over whether a gap exists between WTP/WTA in the experimental literature as a result of subject misconceptions or preferences characterized by loss aversion or prospect theory — though the legal literature sometimes proceeds as if the question is settled. However, Commissioner Rosch has his cognitive quirks all mixed up here. The endowment effect and hyperbolic discounting are different things, and the former is not derived from the latter. Further, this sort of fast-and-loose play with behavioral economics concepts does not provide any comfort for those who note the “model selection problem:” that regulators are prone to picking whatever economic theory furthers pre-existing policy preferences without regard to concerns about scientific method, rigor, or empirically discernable results.
Third, Rosch argues that:
“behavioral economics has provided important insights that suggest the assumption that corporations, i.e. sellers — always behave rationally may not be correct. Neoclassical economics assumes that rational behavior cancels out irrational behavior, meaning that there is no need for economic analyses to account for irrational individual conduct in the analysis of firm behavior. As Commissioner Leary aptly noted as early as 2003, however, this assumption does not account for basic agency problems. Sellers, after all, no matter how large, are comprised of individuals who have “objectives of their own which do not necessarily coincide with those of the enterprise as a whole” and, as a result, the incentives of these “employee agents can prompt conduct that does not maximize the profits of their employer.”
Read that again. Behavioral economics may well have something to say about when firms do not behave rationally. But this isn’t it. Is it really possible that Rosch is attributing the economics of the theory of the firm and agency costs to behavioral economics? It is pretty hard to read the passage otherwise. What about Alchian, Demsetz, Coase, Demsetz, Jensen & Meckling, Williamson, Klein, Hart, Tirole, Holmstrom? Sigh. Again, if all behavioral economics amounts to is imperfect information and the principal-agent problem, we were all behavioral economists all along. But its worse than that.
Note, that a pattern is emerging that reveals the Commissioner’s underlying economic model. If consumers face information costs, and decision-making is imperfect because perfection is costly, the Commissioner concludes that consumers are irrational and regulation is required. If agency costs and principal-agent problems are present, and firms and individuals economize on those costs, irrationality is present, justifying yet further regulation. Behavioral economic concepts which are unrelated to one another are conflated to support interventionist goals; neoclassical economic staples are misquoted or ignored to advance the opposite of the propositions their authors directly assert. At best this can be labeled as yet another example of the Nirvana Fallacy.
There is yet another error here. It is an important one that suggests that traditional economics of the rational choice variety is getting a bad rap without its day in court, or Part III litigation, as the case may be. Rosch writes:
“Neoclassical economics assumes that rational behavior cancels out irrational behavior, meaning that there is no need for economic analyses to account for irrational individual conduct in the analysis of firm behavior.”
That is wrong at best and possibly much worse. While there is an argument that the presence of rational consumers will save irrational consumers from bad outcomes in the presence of competition, and, relatedly, that rational profit-maximizing firms will survive in equilibrium while irrational ones won’t, the conventional argument is that the rationality assumption works on average not because rational behavior cancels out the irrational, but because irrational behavior cancels out other irrational behavior. In other words, assuming irrationality produces no consistently testable implications. For example, if some irrationalities result in over-confident firms and others in firms with sub-optimal confidence, the rational assumption will work in practice in terms of predictive power. This is why the whole idea of behavioral economics is consistent, predictable irrationality. Where one can identify it, in theory, the models that incorporate insights about systematic deviations from rationality can outperform those hampered by the rationality assumption. Whether these models outperform “rational” models is an empirical question — not one that can be resolved by assertion (Friedman, 1953). It is important to get the question right.
Further, with respect to the claim that “need for economic analyses to account for irrational individual conduct in the analysis of firm behavior,” more discussion is warranted. I’ll rely on Gary Becker, an economist of some repute who was thinking about traditional economic theory and irrational behavior well ahead of the current rage, to set things straight:
Although it was long been agreed that traditional economic theory assumes rational behavior, at one time there was considerable disagreement over the meaning of the word “rational.” To many, the word suggested an outdated psychology, lightning-fast calculation, hedonistic motivation, and other presumably unrealistic behavior. As economic theory became more clearly and precisely formulated, controversy over the meaning of the assumptions greatly diminished, and now everyone more or less agrees that rational behavior simply implies consistent maximization of a well-ordered function, such as a utility or profit function.
Strong and even violent differences developed, however, at a different level. Critics claim that households and firms do not maximize, at least not consistently, that preferences are not well ordered, and that the theory is not useful in explaining behavior. Some theorists have replied that economic theory is valid only as a broad tendency, not in each specific instance; some noted that the “proof of pudding is in the eating,” and argued that this theory gives useful predictions even though decisions do not “seem” to be rational; still others claimed that only rational behavior has much chance of surviving a very harsh competitive world.
The purpose of this paper is not to contribute still another defense of economic rationality. Rather it is to show how the important theorems of modern economics result from a general principle which not only includes rational behavior and survivor arguments as special cases, but also much irrational behavior. … Since, however, [the theorems of modern economics] are shown to be consistent with a wide class of irrational behavior, a defense of them is not necessarily a defense of individual rational behavior. Indeed, perhaps the main conclusion of this study is that economic theory is much more compatible with irrational behavior than had been previously suspected.”
That is Gary Becker …. from 1962. Behavioral economists are generally working against a backdrop of economic theory that they themselves understand quite well. The mission of attempting to identify, document, and theorize about systematic deviations is one that has potential for economics. The test will be to compare predictive power of the models at the end of the day. That debate, I suspect, will be going on for some time. But it does not do any good to casually claim victory over a straw man in such an important setting. And the danger is from behavioral law and economists who play fast-and-loose with application of economic theory to law without understanding that backdrop.
These economic questions are important for firms, for antitrust, and, most importantly, for consumers. They deserve serious discussion and consideration. It is not productive to simply define behavioral economics as a “gap-filling” justification for interventionist antitrust policy whether or not the model fits leads to the worst sort of Nirvana Fallacy in which all deviations from market arrangements which the regulator both understands and deems reasonable correspond to market failures due to some behavioral bias.
The danger of this approach is that it will, inevitably, “write out” of antitrust the discipline that economics has provided it over the last several decades. It is that discipline which has led to the evolution of antitrust from the paradoxical doctrine of the 1960’s and prior to its current state of relative economic coherence. Consumers have benefited greatly from that evolution both because of the improvement in the quality of economics and because the economic insights had predictable testable implications that made them useful to judges. It is doubtful, thus far, that behavioral economics offers either of those benefits relative to the status quo. Chipping away at the methodological commitments of antitrust to economics threatens to reverse those benefits.
In Part II, I will discuss Commissioner Rosch’s analysis and breezy dismissal of some of the common critiques of behavioral antitrust and suggest that some of these critiques might have a bit more force than the Commissioner is willing to admit, and perhaps understands.