Steven J. Cernak[1]

For this panel, we have been asked to revisit now-Judge Frank Easterbrook’s seminal 1984 article, “Limits of Antitrust”. Is it still an appropriate guide to antitrust enforcement, both for the FTC and U.S. courts? Or, like many of the rest of us 35 years later, is it a little creaky and, perhaps, ready to be thanked for its fine service and then retired for something else shiny and new?
In my view, the underlying motivating factor for Easterbrook’s “limits” remains at least as true today as back in 1984 and I think something at least like its focus on “the costs of action and information” should continue to drive antitrust enforcement and litigation. Perhaps there can be a more nuanced view of the Type I and Type II errors for particular situations given the development of our learning, as Alan, one of my fellow panelists, suggested a few years ago.[2] But the underlying rationale for those heuristics – that antitrust enforcers and courts should show some humility about why, when and how often they intrude into the market and take steps to “first, do no harm”[3]– is just as true and important today as it was 35 years ago.

Humility in “Limits”

Let’s take a closer look at “Limits” to try to make explicit the underlying admonition that antitrust courts and enforcers should be humble about the good they can accomplish. In the first sentence of the paper, Easterbrook says that antitrust’s goal “is to perfect the operation of competitive markets.”[4] The problem, he says, is that in the real world, competition is messy. It is not like the atomistic competition of an econ textbook. There is plenty of cooperation in various forms, much of which nobody would call anti-competitive, such as all the cooperation that goes on within a firm. I would also add as examples many of the joint ventures that have gone on and continue now within an industry where I am still active, the automotive industry. But if cooperation within a firm seems benign at worst and an agreement on future prices is definitely bad, what about all the combinations of cooperation and competition in between? As Easterbrook describes it:

Are 10-year exclusive dealing contracts between oil companies and service stations too long? Too short? Just right? Does it matter whether there are two oil companies or twenty? 200 stations or 20,000?[5]:

To make matters worse for the poor judge or enforcer (although perhaps providing some comfort that others are equally confused), it is not like the actions of each market participant are always well thought-out or straight out of an MBA business strategy class. As Easterbrook says, “Firms try dozens of practices. Most of them are flops, and the firms must try something else or disappear.”[6] I can distinctly remember early in my career sitting in a meeting where the division decided it needed more revenue this quarter and so needed to raise prices. I expected to see some elasticity estimate next but instead the assumption was that the market was perfectly inelastic, at least over this time range – no sales would be lost and the entire price increase would be paid by the customers. When I asked how this could be, the manager acknowledged the probability that some sales would be lost – but he had no idea how many, no time to figure it out and so this estimate was the best he could do with the limited information and time available.

So if the competitive process that the antitrust judge or enforcer is meant to perfect is complex and if even market participants can’t always figure it out, what’s a judge or enforcer to do? As Easterbrook points out in the context of antitrust litigation, “the judge knows even less about the business than the lawyers”[7] hired by the companies and yet has to make a decision. Easterbrook implicitly suggests that the judge or enforcement leader should have the humility to admit that she might not be able to divine the perfectly correct answer and, instead, “employ some presumptions and filters that will help separate pro- and anti-competitive explanations”[8] and reduce the costs of the decision process and of any mistakes.

Three Varieties of Humility

When I speak of the humility that underlies Easterbrook’s “Limits”, I think there are at least three strains or varieties of humility to keep in mind. The first two have been covered extensively elsewhere, including by some great thinkers, so I intend to focus more on the third.

First, there is the humility to accept that it can be impossible to gather all the knowledge necessary to fully understand the complex markets involved in any antitrust question as well as to confidently predict all the primary, secondary and important tertiary effects, intended or unintended, of any intervention into that market. I think Easterbrook makes this clear in “Limits” – “the judge knows even less about the business than the lawyers” – and others, like F. A. Hayek in his 1974 “Pretense of Knowledge” speech upon winning the Nobel Prize in Economics, have covered this ground extensively. I will just note that Hayek also provides helpful advice to governments that sounds remarkably similar to Easterbrook’s when, near the end of the speech, he suggests:

If man is not to do more harm than good in his efforts to improve the social order, he will have to learn that in this, as in all other fields where essential complexity of an organized kind prevails, he cannot acquire the full knowledge which would make mastery of the events possible. He will therefore have to use what knowledge he can achieve, not to shape the results as the craftsman shapes his handiwork, but rather to cultivate a growth by providing the appropriate environment, in the manner in which the gardener does this for his plants.[9]

Second, there is the humility to recognize that any judge or enforcer, like any other human being, is subject to her own biases and predilections, whether based on experience or the institutional framework within which she works. Yes, markets and their participants might not always act in ways we like but enforcers are not perfect either. Again, this idea is not new, dating all the way back to at least Madison’s remark in Federalist 51 — “If men were angels, no government would be necessary. If angels were to govern men, neither external nor internal controls on government would be necessary”[10]– and up through Bill Kovacic’s application to antitrust agencies in a 2016 article – “Agency leaders are not angels.”[11] A new article from Thibault Schrepel, Antitrust without Romance, more than ably covers this concept and the application of the public choice thinking of James Buchanan and others to antitrust enforcement.[12]

I want to spend a little more time on the third type of humility – the recognition that we are not the first ones to face some of these questions and, in fact, though the particulars might be a little different, we might be able to learn something from those that came before us. This humility is implicit in “Limits” when you see the presumptions, filters and focus on error costs as distillations of learning from past experiences.

Now, I often see this failure to appreciate history in my clients, like those manufacturers who are convinced that the issue of poor customer service and other brand-destroying actions by distributors – oh yeah, and low resale prices – began with the Internet. I’m sure the heads of the Dr. Miles Company would have something to say about it.[13]

I think we can also see this failure, this lack of humility, in some, but not all, of the reactions to the currently wildly successful companies that have built up huge market shares and seem to be indestructible. Is the right action a breakup of a successful company or drastic changes to how investors invest in companies in the same industry? Or might a better action, with fewer negative unintended consequences, be to ensure that competitors of these behemoths are able to compete to better serve customers and try to wrest away any market power? I was reminded of this lesson the other day as I drove past a Babies R Us store – not too close, however, as the parking lot was walled off while the bankrupt former category-killer Toys R Us sells off the land and buildings.

A favorite historical example in response to current fears of unbeatable alleged monopolists is A&P or the Great Atlantic & Pacific Tea Company. The original supermarket, huge market share, vertically integrated, “sapping the civic life of local communities” (to quote Rep. Wright Patman, as he was inspired to draft the Robinson-Patman Act) – it went from 16,000 stores in 1930 to a quarter of that number 20 years later to competitive irrelevance several years later to out of business. Former FTC Chairman Tim Muris has a great article recounting the tale.[14]

Given my background, I want to relate a couple other examples. First, consider this quote from a U.S. Senator:

It is evident that business has grown to such an extent, mergers have been taking place with such rapidity, and economic power is being concentrated in fewer and fewer hands to such a degree, that the legislative and executive power of this Nation should come quickly to an understanding as to a formula for clarifying the antitrust laws by which we can stabilize our economy.”

A quote from someone running for or supporting a candidate for President in 2020? No, Senator Joseph O’Mahoney (D-WY) on November 8, 1955 as he was kicking off 18 days of hearings into the antitrust issues raised by the operations of General Motors Corporation. GM, the vertically-integrated company with over 50% of the light duty vehicle market at the time. GM, with a dominant share of the refrigerator business through its Frigidaire subsidiary at the time. GM, with its Electro-Motive Division subsidiary having sold more than 60% of the locomotives operating at that time. GM, the company which hasn’t made a refrigerator or locomotive in decades and that declared bankruptcy in 2009. There was plenty of talk at the hearings, from both Senators and experts alike, about GM’s dominance and ability to head off meaningful entry. One expert predicted that “It may turn out that Chrysler Corp.’s entry in 1923 is the last successful one.” Almost two years to the day later, a foreign entrant established a small US sales subsidiary in California. You may have heard of them – Toyota. (As an aside, keep this prediction from the 1955 hearings in mind as different hearings get underway soon.)[15]

Just one more example from automotive history, but one very appropriate for an FTC hearing looking into actions from 35 years ago. At the beginning of my talk, I mentioned joint ventures in the automotive industry. One of the biggest JVs started business 35 years ago. In 1984, NUMMI, New United Motor Manufacturing, Inc., started production of small cars in Fremont, CA. NUMMI was a production joint venture between GM and Toyota designed to produce small cars, help GM learn the mysteries of Toyota’s high-quality, low-cost production methods, and convince Toyota that such methods could be implemented by U.S. workers. It almost didn’t happen. The FTC had barely approved the joint venture the prior year with one of the dissenting commissioners asking “if this joint venture between the first and third largest automobile companies does not violate the antitrust laws, what does the Commission think will?”[16]

But approve it the FTC did, although with some conditions, including an ongoing requirement to annually share with Commission compliance staff certain documents regarding the interaction among the companies. (I know, I collected and shared those documents with FTC staff, who always seemed to spend less time with me in suburban Detroit than he did with Toyota and NUMMI staff in northern California.)

In many ways, NUMMI was a great success: GM did learn much about the Toyota Production System and those efficiencies have now spread throughout the company and industry. Toyota was convinced that its methods could work in the U.S. and now makes over a million vehicles here. In other ways, NUMMI was a failure: It never made much money, if any, in any given year for its parents and the vehicles that it produced for GM were never great sellers by industry standards. But perhaps most pertinent for antitrust purposes is what didn’t happen: The cooperation of the two companies did not bring about decreased output or increased prices or any other negative effect on competition. As Kathy Fenton said in a 2005 Antitrust Law Journal article, a whole new generation of antitrust lawyers by that time could ask, “What was the big deal?”[17]

Now, all this talk about history and humility doesn’t mean that the FTC or our hypothetical judge should be frightened into inaction – after all, even “Limits” recognizes legitimate antitrust actions. Nor do I think the exact rules described in Limits cannot be adjusted – it was published in the Texas Law Review after all, not inscribed on stone tablets.

And while I think much can and should be learned from history, I don’t think it is sufficient for me to yell “But, but, but, A&P” and consider the argument done. As Jonathan Baker has pointed out, the high market shares of such past giants as GM, RCA and Xerox persisted for quite some time.[18] Antitrust law should consider if it knows whether the persistence of such high shares shows the willingness and ability of these successful competitors to continue to meet the desires of consumers or, instead, the blocking of the rise of effective new competitors. In doing so, antitrust scholars should rely not just on theories of potential anti-competitive conduct but on empirical work, like the early Chicago School did, to give us the confidence to implement the theories.[19]

So in the end, I think the right approach to both adjustments to a “Limits” approach and to antitrust enforcement itself is, to echo former Acting FTC Chairman Maureen Ohlhausen, a respectful “regulatory humility” to what we know and can improve by intervention in a market.[20] If antitrust law is meant to ensure that the market aspects of democratic capitalism – a system, that some have said has “lengthened the life span, made the elimination of poverty and famine thinkable, enlarged the range of human choice” [21]– then we should be confident that we have learned something in the intervening 35 years before we make any changes. What do we know now that Easterbrook didn’t know then? About the complex interactions of customers, suppliers and competitors? About human beings, whether actual or potential buyers, sellers, investors or enforcers, and how they react to various incentives? About technology diffusion, or when R&D is successful? The enduring legacy of “Limits of Antitrust” should not be its answers to questions like these but ensuring that we – practitioners, judges and enforcement agencies – ask those questions anew to see if we can now come up with better answers.


[1] Of Counsel at Schiff Hardin LLP. This submission is a lightly-revised version of the remarks presented at the June 12 FTC Hearing on Competition and Consumer Protection in the 21st Century on the panel “Revisiting ‘The Limits of Antitrust’”. The author wishes to thank Thibault Schrepel, Tad Lipsky and Alec Kraus for their helpful comments on an earlier version of the speaker notes that is the basis for this submission. Also, Alec Kraus provided helpful research assistance. Those remarks and this paper are the views of the author alone and not necessarily the views of any past, present or future employer or client.

[2]  Alan Devlin and Michael Jacobs, Antitrust Error, 52 Wm. & Mary L. Rev. 75 (2010).

[3] See Philip Marsden, Who Should Trust-Bust? Hippocrates, not Hipsters

[4] Easterbrook, Limits of Antitrust, 63 Univ. of Texas Law Review 1, 1984, p. 1. (hereinafter, “Limits”).

[5] Limits, p. 2.

[6] Limits, p. 5.

[7] Limits, p. 5.

[8] Limits, p. 9.

[9] F.A. Hayek, Pretense of Knowledge, available at

[10] Available at

[11] William E. Kovacic & David A. Hyman, Consume or Invest: What Do/Should Agency Leaders Maximize, 91 WASH. L. REV. 295 (2016)

[12] Available at

[13] See Dr. Miles Medical Co. v. John D. Park & Sons Co., 220 U.S. 373 (1911).

[14] Timothy J. Muris & Jonathan E. Neuchterlein, Antitrust in the Internet Era: The Legacy of United States v. A&P, available at

[15] For more details, see Cernak, Maybe History Can Provide Useful Perspective, AntitrustConnect, posted August 1, 2017 and available at

[16] Gen. Motors Corp., 48 Fed. Reg. 57,257 (Dec. 28, 1983) (Bailey, Comm’r, dissenting).

[17] GM/Toyota: Twenty Years Later, 72 Antitrust Law Journal 1013 (2005).

[18] Jonathan Baker, Taking the Error Out of “Error Cost” Analysis: What’s Wrong with Antitrust’s Right, 80 Antitrust L.J. 1 (2015).

[19] See Bruce H. Kobayashi & Timothy J. Muris, Chicago, Post-Chicago, and Beyond: Time to Let Go of the 20th Century, 78 Antitrust L.J. 147 (2012)

[20] See, for example, Regulatory Humility in Practice, available at

[21] Michael Novak, The Spirit of Democratic Capitalism, 1982.

This page as PDF