The following is a transcript of Eric Posner and Carl Shapiro’s debate on the proper role of economics in merger review at the 2024 Stigler Center Antitrust and Competition Conference.
Filippo Lancieri
Hi everyone. So, good afternoon again from Chicago, and thank you all for joining us. I’m Filippo, a post-doctoral fellow at ETH Zurich and a research fellow here at the Stigler Center. This afternoon we are really delighted to facilitate a debate between Eric Posner and Carl Shapiro, moderated by Deni Mantzari.
As usual, the views of our guests are their own and not those of the University of Chicago, and you can find this video later on YouTube. So keep in mind that you’re on the record and recorded.
We’re very happy to host this debate because, at the conference last year, we had a discussion on mergers and how to rethink the consumer welfare standard. And then Eric and Carl started a lively debate that we didn’t have much time to finish, and we were like. “Okay, why don’t we give them an hour to hash things out, you know?” And that’s basically what we’re going to have now.
So let me briefly introduce our speakers and moderator. So, in the middle, at 6’5… No, I’m joking. Eric Posner is the Kirkland & Ellis Distinguished Service Professor of Law and Arthur and Esther Kane Research Chair at the University of Chicago Law School. And before that, he served as Counsel to the Assistant Attorney General for the Antitrust Division of the Department of Justice from 2022 to 2023.
Carl Shapiro is the Distinguished Professor of the Graduate School at the University of California at Berkeley, and he had many positions in government as well, including as a member of the President’s Counsel of Economic Advisers, and also as Deputy Assistant Attorney General for Economics at the Antitrust Division.
And Deni is an Associate Professor in Competition Law and Policy at the University College London Faculty of Law. And I think it’s nice to say that Deni’s work really focuses on how courts comparatively employ economic evidence. And she has a book out that you should check out that’s called Courts, Regulators and the Scrutiny of Economic Evidence – Comparative Perspectives.
So, without further ado, I give it to you.
Eric Posner
Thank you so much, Filippo. I have one slide. Let’s see if it works. There it is, but it’s one unreadable slide. Can you read it? I can’t read it! [laughter]
It’s really nice to be here and to debate with Carl, who I greatly admire. I’ve read a huge amount of his work, so I’m prepared for this debate, I suppose. I think someone yesterday (was it Fiona?) said that law professors in America don’t actually talk about law. I’m going to have to break from that mold for the purpose of this debate because the question is the role of economics in merger review. And merger review is obviously a legal process. It’s effectively required by law. And the agencies are supposed to follow the law, although of course they have discretion as well. As enforcement agencies, they have a fair amount of discretion to choose how to allocate their time and which mergers to focus on, based on their assessment of the public interest.
But let me just start with the law. The operative language here (which I can barely read but I’ve memorized this) is that a merger is illegal if it may “substantially lessen competition or tend to create a monopoly.” So that’s the statute that we’re dealing with. It is, although a very short statute, in a way, fairly clear. Before I talk about how a lawyer would interpret it, how a judge or a lawyer would interpret the statute, let me start with what I understand the economic theory of antitrust law to be. And I use this very broadly because economists who write in industrial organization and write about antitrust obviously have disagreements and emphasize different aspects of this.
I think of the economic theory of antitrust law as going back to Oliver Williamson’s famous 1968 article, which proposed a very simple cost-benefit analysis for evaluating mergers. Williamson said later that, I think his boss Donald Turner had asked him to come up with some kind of test for determining which mergers the agencies should challenge. And Williamson said he talked to the lawyers and they just said, “Oh, you have to enforce the statute.” Which didn’t satisfy him, so he came up with this cost-benefit test. And his cost-benefit test is extremely simple. Famously simple. He even called it “the naive model” in his paper. He says that a merger should be blocked, basically, if it’s inefficient. If the costs are greater than the benefits. He identifies the benefits being efficiency, economies of scale, the benefits of a merger. And the costs are the deadweight cost that results from higher market power (if that should take place) and higher prices.
This view was adopted by Bork in his work. And while Williamson was very cautious, and most of his papers are actually qualifications and cautions about his test, Bork dismisses all of those qualifications (I’ll talk about some of them) and just says, “This is the test for merger challenges.” As far as I can tell (it’s a little unclear), this was the test that was adopted by the Reagan Administration. Possibly informally. I’m not sure. But they did stop challenging mergers in the late 1980s for a few years, and possibly because they applied this test.
Now, the odd thing about this test from a legal perspective, from the perspective of statutory interpretation, is that it basically says that the defendants, the merging parties, are allowed to merge as long as any harm that they impose on consumers, for example, is less than the benefits to them. Right? So, if their profits go up enough because of market power and reduced cost, the merger should be approved. That view didn’t stick. And I think it’s just so obviously in tension with the purpose of the law.
Ultimately, the view that replaced it (which I think is probably the economic view now, although economists disagree about it) is a consumer surplus view. A merger should be blocked if it will raise prices, basically. What it requires, this view, for a merger to go through is that the cost savings resulting from economies of scale are passed through to consumers, so that any increasing price pressure as a result of the greater market power is offset, and consumers are better off. I think of that as something like the economic theory. Certainly there are economists who make this argument. Maybe most economists. I believe Carl’s theory is something like this, based on his ProMarket piece in anticipation of this debate, but Carl can correct me if I’m wrong.
Let me turn to the to the law now. So, for the lawyers here, this is obvious. There’s a whole kind of infrastructure, or set of rules, or community understanding of how one interprets law. The rules here, the toolkit here, is not very complicated. It’s mostly common sense, but it’s worth going through it a little bit to understand what goes awry with the economic theory.
Courts, particularly these days, like to focus on the plain meaning of the statute. And this statute actually does have a discernable plain meaning, which is that we’re worried about competition going down. Competition means something like rivalry. It means you have a bunch of firms that are trying to poach each other’s customers, or other trading partners, by offering better terms to them or by engaging in marketing, or what have you. They’re fighting against each other for business. Right? That’s what rivalry means. You can think of a rivalry test if you want, or a competition test. But it doesn’t go beyond that. The term “substantially lessen competition” means that you have multiple firms that are doing this, that are competing with each other, and as a result of the merger, there’s less of that going on.
Now, there could be less of that going on simply because there are fewer firms. If one firm faces a lot of other firms, it may compete for business more vigorously than if it just faces one. Okay? Or it could happen because, let’s say (this has been discussed a lot), the firm that is acquired is a maverick. And so, firms that in the past might have coordinated or colluded or something like that, had not because of the behavior of the maverick. But after the acquisition of the maverick, that becomes easier or more likely to occur. Okay? So, the statute seems to be focused on that kind of behavior. And this is very much different from either form of cost benefit-analysis, the total surplus or the consumer surplus test.
The easiest way to see this is just to imagine a 2 to 1 merger. This is the beginning of our debate, I guess, a year ago. A 2 to 1 merger that could reduce prices because of big economies of scale that are passed through to consumers, but it seems pretty obviously to be prohibited by the statute, which says you’re not allowed to merge into a monopoly.
Well, what about a 3 to 2 merger? Or a 4 to 3 merger? If you just take “tend to create a monopoly,” which doesn’t just mean 2 to 1, it might mean 3 to 2 or 4 to 3, and you attach it to the “substantially lessen competition,” maybe that means 5 to 4, or 6 to 5. Who knows? That seems to be what the statute is prohibiting.
Now, you might ask, “Well, where’s the line for ‘substantially to lessen competition’? Is all rivalry, are all mergers prohibited because all mergers (well, as a general proposition, nearly all mergers) are going to reduce rivalry, at least between the merging firms?” And the answer is no. Congress has put in a di minimis test and left it to the courts to figure out what that should be.
The economic theory, as I understand it, claims that Congress’s purpose was to basically either help consumers, or promote economic growth, or push down prices, or promote output.
That may well have been one of Congress’s purposes. But Congress clearly had many other purposes when it enacted this statute, as had been talked about already at this conference, and people generally were worried about corporate power. In those days, powerful corporations would just bribe legislatures. They also were immune to regulatory enforcement because they had so much power and they were very nimble. They could move from state to state as people went after them. Corporations would hire thugs to crack down on striking workers. And so, there was a lot of concern about that, as well as the economic impact.
After this, there was a lot of additional legislation and opportunities for Congress to change these rules if it wanted to. It didn’t take that opportunity.
The mistake in the economic theory is to single out one of these multiple purposes and say, “That’s all that Section 7 is supposed to accomplish: increasing economic output,” while neglecting all these other purposes, which clearly motivated Congress and have remained a concern to the present day. This view of Section 7 is reflected in Supreme Court interpretations that remain valid law, even though they’re old. These are Supreme Court cases that have subsequently had a great deal of influence on outcomes.
Let me just answer the question, “What is the role of economics in merger review?” I do want to say, I’m not arguing there’s no role of economics in merger review. As Zephyr mentioned yesterday, whenever you’re trying to measure harm, and that’s a matter of concern, economics can play a role. It can play other roles as well, but on the crucial issue of what the liability rule is, I don’t think it’s correct to say that it is an economic rule.
Deni Mantzari
Thank you.
Carl? Your turn.
Carl Shapiro
Thanks.
So, last year we did have this interchange. At that time, Eric was arguing that mergers should be evaluated based on whether the margins of the merging firm would go up. If the margin went up, that was increased market power. That would be illegal. I then said, “Wait! What about if two small firms merge. They lower their costs. They pass through some of the cost savings to their customers, so prices go down. But they don’t pass through all of it, so their margins go up. You want to be against that merger?” Off we went. Eric seems to have wisely not pursued that line of argument further today, and he instead is going after Oliver Williamson, but okay.
The question posed was the proper role of economics in merger review. Since my mother taught me to follow instructions, I will start (rather than end) with the answer: to assist the legal system in enforcing the law. Okay? So, we better go to the law. I’m speaking now based on, not only my academic writings, but time in government serving as an expert witness in merger cases.
We understand the statute refers to the “may substantially lessen competition.” Let me just focus on that prong. We’ve also got the “may tend to create a monopoly.”
“May substantially lessen competition.” That’s all about the notion of “competition” is an economic concept or notion. We also have the “substantially.” So, we have to have some notion of magnitude or significance. This seems to bring in economics pretty centrally.
As an example, when I testified in the Staples/Office Depot case for the FTC, the market the FTC defined was “sales of office supplies to large companies,” essentially. We had very nice data about the two merging companies bidding for these contracts that these large companies would set to provide a whole bunch of office supplies and other stuff. Well, the merger would eliminate that bidding competition. That was very significant. That was about economics. It was about bidding. Who are the other bidders? How often did they win? These are all economic concepts. This was central to evaluating whether that merger would substantially lessen competition.
Another example. I testified for Steves Doors, the plaintiff in a successful private action in a merger case. They were challenging a merger of two of their suppliers of door skins, which are critical input into the doors. It was a consummated merger. And there I showed to the jury that the prices had gone up after the merger. That’s obviously economics, that Steves Doors was injured. They also had a damages claim. And further had to show the prices went up, not because costs went up, but rather because of the merger. And in fact, prices had gone up after the merger, and the costs of the suppliers had gone down.
All economics. Very central to the question, “Did those mergers substantially lessen competition?” I’m not sure what alternative method that Eric is suggesting, but judges and juries want to know, “Will people be harmed by the merger?” Harm will typically be, in these cases, customers. Of course, in other cases it could be suppliers or workers.
Okay, so that’s the role of economics, both actual and to some extent I would say proper, at least as I have used it.
But there’s a lot of misuse of economics, as well. This is a big problem in merger analysis. We had a long period of time where a whole phalanx of economists, paid by merging firms, were arguing a theory of critical loss that made no sense. And it got a lot of traction. Or they argued the only way merging could be a problem is if the two firms were the next-closest competitors to each other. Also not true. Okay? Just economically false. But these things got some traction.
So the actual role of economics is much more mixed because economics gets abused by defendants in these cases. It could be abused on either side, okay? But I’ve been testifying for 15 years, and all these testimonies have been for plaintiffs, mostly the government in these cases. So this is what I’m seeing.
To you give an example of that in the Steves Doors merger. This was a merger between two of these suppliers of these door skins, right during or right around the time of the Great Recession. As I said, I told the jury prices went up. The cost had actually gone down because input cost had gone down. And so the majority cost, the prices go up. The defendant, the economist testified for the defendants, which was Ted Snyder who at the time I believe was Dean of the Yale School of Management so I’ll have to get Fiona to straighten him out on this. He told the jury, “Oh, no problem because output went up after the merger.” That is, more doors and door skins were produced. And we know the Supreme Court likes to look at output, so he should look at output. Why did output go up at that period of time? Because we were coming out of the Great Recession, and housing starts were going way up, and so that had nothing to do with the price of door skins. It was not very plausible that, because the price of door skins and therefore doors was going up by $20 or $30 a door, that people would decide, “You know, I don’t think I need a door into that bedroom. I’ll just skip that.” So, that was nonsense. And the jury, I think, saw it as such.
So, there’s good and bad economics. I’m talking about the proper role of economics. There’s a separate issue, maybe we’ll talk about it, about how to discipline this type of really bad economics that gets paid for and put forward in some of these cases.
Further, how do we tell whether there’s a substantial lessening of competition? If you’re going to figure out there’s a substantial lessening, you better figure out what it is, what is “competition,” so you can see whether it went up or down, or significantly so. I like the definition of competition that the merger guidelines have. They say competition is “a process of rivalry that incentivizes businesses to offer lower prices, improve wages and working conditions, enhance quality and resiliency, innovate, and expand choice, among many other benefits.” I like that, and it fits with common sense and how economists think about competition. How would you tell if competition’s gone down? Well, a good way to diagnose that would be whether the benefits that competition generates are likely to go down.
Now, it’s all predictive in these cases. Almost all cases are not consummated mergers. How are you going to predict whether these benefits will go down and that, hence, we are led to something like a “will prices go up?” Okay. That’s a very sensible test diagnosis because, for one thing, it relates to harms, which judges and juries are going to care about, and they always have, going back to all those Supreme Court cases. In Philadelphia National Bank, the Supreme Court was concerned about whether business borrowers in the Philadelphia area would get less attractive terms or other conditions because of lessened rivalry among the banks. That’s consistent in the case law. So it’s a good diagnosis to look at whether prices will go up or quality go down, and so forth. That’s very much about economics and what economists can do.
The same thing would apply upstream. If we have a hospital merger, and we’re worried about whether this will reduce competition to hire nurses, we look at this and we say, “Okay. What are the choices the nurses have, in terms of where they can seek employment. Are these hospitals competing to get nurses?” We look at the geographic scope. Where can the nurses go? We might look at the share of employment of nurses by these hospitals. These are all economic concepts. Ultimately, we’re very concerned here about the nurses being harmed, and that would come in the form of lower wages. So again, this is what we do. The economics. I don’t know what the alternative is supposed to be.
All right, so then how has this evolved over time? There’s, I think, generally quite a productive interaction between the courts, the agencies, through the merger guidelines, and academic work, and other research in economics, that this diagnosis, how do we diagnose a substantial lessening of competition, we’ve changed our tools. We’ve gotten better, I would say, at how to diagnose. In the mid-20th century, the structure-conduct-performance paradigm was paramount. It reigned in industrial organization economics. That was reflected in the Philadelphia National Bank decision, which quoted a number of economists for the structural presumption, and quoted them on the shares that might be sufficient that led to the 30% share we have there.
Over decades, economists have learned. We have developed more game theoretic tools. We’ve understood many different types of markets and how different they are. We’ve developed unilateral effects. We have demand estimation. We have merger simulation. Bidding models. All these other tools. Those have been brought into practice. Merger guidelines have assisted with that. The latest iteration, we see that again in the form of talking about, not just labor markets, but platform markets as a leading example.
This process, I think, is quite healthy. It is how I would say economics has been used and has influenced the law. But it’s all about improving the way in which we interpret, implement this statute regarding substantial lessening of competition. I don’t really see an alternative.
Of course, none of this is to say that other goals of the antitrust laws are irrelevant. It’s just, when you go to court, and you’re looking at a merger between two suppliers of door skins, you’re talking about the effect on competition in door skins. You’re not talking about democracy or inequality, as important as those values may be and as important as they may have been to congress when they either passed the statute in 1914 or amended it in 1950. The statute speaks of economics. That’s what the courts want to know about. That’s what economists can help with.
Deni Mantzari
Thank you very much. I think we have two diametrically opposing opening statements that will help generate a lively debate.
We’ve heard (just to put the different pieces together) about whether we should follow a literal, a textual, or a purposive interpretation of the Clayton Act. We heard a lot about the misuse of economics and the use of economics in diagnosing the substantial lessening of competition. It’s under the Section 7 of the Clayton Act: “The merger is prohibited if the effect of such acquisition may be to substantially lessen competition or tend to create a monopoly.”
But are we looking only at price effects? Is this our only concern? Recent decision practice in Europe and in the U.S. has brought forward other concerns. We understand that economists are trying to fit these non-price effects conceptually into the existing framework of diagnosis. I will go a step further, and I will ask also whether they should have a freestanding basis, like the one enjoyed by price output reductions.
Eric Posner
Before getting into your question, I just want to renew my question to Carl from last time.
What I asked was, “If you have a 2 to 1 merger that results in lower prices is that a substantial lessening of competition?” Herb Hovenkamp jumped in and gave the example of, you know, let’s just imagine Lyft and Uber in a town. Suppose there’s no taxi service. It’s just the two of them. And they merge. Herb said that is not a substantial lessening of competition if prices go down. Okay? And I thought you disagreed, but what is your answer to that?
Would you say that’s a substantial lessening of competition, or not? A 2 to 1 merger. Prices go down.
Carl Shapiro
Well, 2 to 1 merger, by the way you’ve defined that, is a complete elimination of competition. So how could it not be a substantial lessening?
Eric Posner
Okay, but under your theory it would be lawful. So if it’s a 3 to 2…
Carl Shapiro
That’s such an easy one to diagnose. One thing, you already have the other prong that says, “tend to create a monopoly.” But the substantial, complete elimination of competition…
Eric Posner
Okay. So, 3 to 2 merger that results in lower prices. You approve it? Not substantial lessening of competition because it’s 3 to 2, rather than 2 to 1.
Carl Shapiro
I think it’s going to depend on the facts, or the way the case…
Eric Posner
Oh, that’s what lawyers say! [laughter]
Carl Shapiro
What? Because I actually look at the…? We learned from Jonathan Kanter last night we’re supposed to look at…
Eric Posner
I gave you the facts: You looked at the evidence. Prices go down.
Is it lawful or not lawful?
Carl Shapiro
So, wait. You’re asking a legal question, or are you asking an economic question?
Eric Posner
I’m asking both questions. Substantially lessening competition, which you interpreted economically.
Carl Shapiro
If the prices go down, that to me is an indicator. That’s my point. That’s a strong indicator that competition has gone up. There’s increased rivalry between those two firms.
Now, the only reason I said I want to know a little more is, I could be worried that there’d be coordinated effects and we haven’t seen them yet, so…
Eric Posner
Yeah, we won’t assume this. Try the other way.
Deni Mantzari
It’s by the fact that their market share…
Eric Posner
…because costs go down.
Carl Shapiro
Okay.
Eric Posner
Rivalry doesn’t have to… Rivalry obviously goes down.
Carl Shapiro
Well, you see, that’s your view. Okay? Rivalry goes down…
Eric Posner
But you think rivalry obviously goes down if we go from 2 to 1, but you don’t think it obviously goes down if we go to from 3 to 2. That’s just what the statute says.
Carl Shapiro.
This sounds like sort of a pointless scene, dancing on the head of a… you know? Like, seriously?
In practice, if you had a 3 to 2 merger and the parties showed the prices go down, I would say (as an expert, this my experience) that the parties would probably win that case. Okay?
Eric Posner
No. They would lose.
Carl Shapiro
Okay. Well, that’s because you don’t think there should be an efficiency defense.
Eric Posner
No, this is a prediction about the law. Well, yes, it’s a prediction about what a court would do interpreting the statute.
Carl Shapiro
All right, well, I don’t know. Just based on my experience with this… Fine, they might. They would do what they do. Okay? All I can do as an economist is say, from my point of view (let me state this as clear as I can), the fact that the price went down—and it’s going to stay down and all that—indicates to me that there is enhanced rivalry between these two firms. Probably because the cost went down. I think you said that.
That is a form of enhanced rivalry. You can have a lot of competition between two firms. We have models of that. We have examples of that. So, for you to say, “Oh, because there’s only two left, there must be reduced rivalry,” I don’t accept that in terms of what rivalry and competition means to me.
Who are you benefiting if you stop that merger, by the way?
Eric Posner
Well, this goes back to the statute. You’re just refusing to take into account these other purposes that Congress was concerned about at the time, and that people are still concerned…
Carl Shapiro
Wait. I thought we were talking about competition and rivalry. Now you trying to change the subject.
Eric Posner
No. It’s “substantially.” This is a statute that uses the words “substantially lessen competition.”
You insist over and over again that, when Congress says that, competition is meant in a purely economic sense to mean… Well, it’s more than an economic sense. It’s an economic sense of, not only rivalry, but rivalry that results in higher prices, but that’s not what… Congress passes this statute because, at the time, people thought there were too many monopolies. That it was too easy for companies to merge and become large monopolies. And that these monopolies were engaging in non-economic abuses of various types.
Carl Shapiro
So you have an interpretation of the statute…But tell me. If I go back to Philadelphia National Bank, which your father had a hand in, okay? A big hand, I understand…
Eric Posner
He didn’t sign it. [laughter]
Carl Shapiro
No, it’s a great opinion! I mean, don’t you think it’s a great opinion?
Eric Posner
No. [laughter]
Carl Shapiro
All right, fine. It was meant to be a compliment! You know? You have great genes.
Okay. What they’re saying is, this is a structural presumption, but it can be rebutted. So what’s it take to rebut it? If you read the decision, it’s very clear. You would rebut it by showing there was not a substantial lessening of competition. The harms they identify, “competition among banks exists at every level: price, variety of credit arrangements, convenience of location, attractiveness of credit,” all these things. Non-price things.
That’s what they’re looking for. If those effects did not happen, they would be rebutted. They consider some rebuttals, and they reject them because they didn’t meet this standard. So, by this very decision of the Supreme Court, they’re looking at these economic effects, in this market, in the banking market in Philadelphia. You’re making it seem like there’s no rebuttal.
Eric Posner
The court repudiates an efficiency defense following Brown Shoe, and the Supreme Court case is consistent about that.
Dani Mantzari
We don’t have a Supreme Court case, I understand, in this country, we don’t have one the last 50 years, right?
Eric Posner
We don’t have one in the last 50 years, but…
Deni Mantzari
Which is another question of how this way will…
Eric Posner
…these Supreme Court cases are relied on and cited by courts of appeals and district courts in every other case. And not only that…
Carl Shapiro
So you don’t think Philadelphia National Bank is good law? How do you rebut then?
Eric Posner
Pardon me?
Carl Shapiro
There’s no rebuttal? How do you rebut?
Eric Posner
Well, there has been movement among the courts of appeals, but the Supreme Court is the authoritative interpreter of the law…
Carl Shapiro
Yeah, that’s why I’m asking about Philadelphia National Bank.
Eric Posner
…so it’s good law.
Carl Shapiro
You read this decision, and you rebut by showing that these harms to the customers won’t happen. Okay?
Eric Posner
No, the court rebuts an efficiency defense. They bring up the possibility that there may be public interest reasons for allowing a merger to go through. For example, if you’re in the middle of a financial crisis and you need to merge your banks because one’s failing. That could be a justification. They’re not ruling it out. It is a structural presumption.
They’re not ruling out, by definition, that the defendants can rebut. Right? That’s the whole meaning of structural presumption. But it doesn’t say you can rebut on efficiency grounds.
Carl Shapiro
I didn’t say anything about efficiencies. I just said, you show that the prices go down instead of up. I don’t know why! I didn’t say the word efficiencies. I just said effects. I don’t know why happened.
Eric Posner
It’s not just not clear from the court whether you can do that. I do not think that the Supreme Court in the 1960s would approve a 2 to 1 merger because the prices went down.
Carl Shapiro
We’re not talking about 2 to 1 merger. We’re talking about Philadelphia National Bank.
Eric Posner
All right, a 5 to 4, or 4 to 3 merger. Brown Shoe was a few years earlier. They don’t repudiate Brown Shoe. They adopt Brown Shoe. They cite Brown Shoe as authority. They even mention Brown Shoe’s celebration of small businesses.
Philadelphia National Bank does not change the law. It’s simply extending the law. What Philadelphia National Bank establishes…
Carl Shapiro
Not going to change the law? I get that, actually.
Eric Posner
…what Philadelphia National Bank is. This is very common in judicial opinions. It becomes a more formal, it adopts a more formal stance than you had in Brown Shoe.
They’re saying, “Look. We’re going to create this presumption in order to give some people guidance.” This is, like, extremely typical for courts. “We don’t know all the consequences of adopting this general rule, so we’re not going to make it a bright line, irrebuttable rule. We’re going to allow defendants in future cases to try to come up with a reason for allowing the merger to go through. That’s totally fine.”
They didn’t say Brown Shoe is wrong. And in subsequent cases, the court relies on Brown Shoe, relies on Philadelphia National Bank, it’s all over the place.
Carl Shapiro
First off, I think one of the interesting things about Philadelphia National Bank is the court, the defendants say, “This is going to be really good for economic development in Philadelphia.” And they say, “That’s some sort of big, large cost-benefit analysis. That’s beyond what the courts can do. We’re not going to consider these broader things. We’re going to focus on whether it substantially lessens competition in this line of commerce.” They focus on the economic issues.
Brown Shoe they cite, as it always gets cited, for the proposition, the Clayton Act is about “competition, not competitors.” That’s predominantly, that’s overwhelmingly with Brown Shoe, and that’s what, that’s how economists think about it as well.
But, look. I am going to stop debating the law. Okay?
Eric Posner
Okay. Good.
Deni Mantzari
I fought the law, and the law won. Yeah?
Eric Posner
I was about to stop, also.
Carl Shapiro
But I beat you to it.
Eric Posner
Okay, so you get the last word.
I did want to respond to a couple other things you said. First of all I’m not repudiating my earlier view. I think of market power as just a way of measuring rivalry or competition. It’s just a statistic that could be useful.
I agree with the bad economics problem, and I do think this is worth talking about. What Carl is saying is that there’s good economics and bad economics. I agree with that completely. It can be done better. It can be done worse.
The real problem here is the courts. There are good judges and there are bad judges. The good judges, you can tell they’re really trying to understand the competing expert reports and figure out who’s right, but this is too complicated for these judges. This is not an argument for getting rid of economics. I agree with Carl’s first point, which is that economics exists to assist the agencies in enforcing the law. I do think that, if it plays too big a role in litigation, you get bad results.
The particular problem, if I can put this a little bit more strongly, is that economics puts a premium, as a matter of practice, on rigorous analysis. Good data. So the types of economic analysis that’s used in litigation, that can produce a prediction about price in a kind of a rigorous way, crowds out the other types of arguments that could be made, including economic arguments, because judges are impressed by this.
A good example, which I think you might agree with, is coordinated effects. Models aren’t as good. It’s much harder to make a prediction, traditionally, about the impact of a merger and coordinated effects. So, over a number of years, this concern about the effects of mergers, admittedly an economic effect of merger, gets crowded out by the focus on unilateral effects.
One of the advantages of having the simple rule that Congress, with great foresight, put together is that you keep things simple enough that courts can enforce the law.
Deni Mantzari
We are here not debating the role of economics, right? You do both agree that it has an important role to play. We are debating whether the economics we have is good, bad, or (may I add) very sophisticated. That is also a problem because law cares about administrability. Then this creates a huge information asymmetry between an expert economic expert witness and a generalist judge. What sort of procedural innovations can we adopt in order to overcome this information asymmetry. So I think the problem is not only good/bad, but also maybe very sophisticated economics. In the area of unilateral effects, for example.
Carl Shapiro
This has been a project of mine for at least 20 years, to try to identify relatively simple but robust economic concepts that can be used in mergers. And non-merger cases, too. There’s pay-for-delay cases. There’s more straightforward ways to go there, more plaintiff oriented the courts have been. Standard essential patents. And mergers.
In mergers, the notion of looking at diversion ratios. I gave a speech about that in 1995 or 1996 when I was Economics Deputy. That’s a way to look at the competition between the two merging firms. And the elimination of that competition alone can make for a substantial lessening of competition. I’ve been pushing that for almost 30 years. It’s hard because the courts are, you know, the case law is so rigid and set about, you have to define the relevance market and, typically, to measure market shares. Which economists know is not necessarily the most informative way to diagnose these things. I think the current, the latest merger guidelines make progress on that, with their guideline building on what we did in 2010. So there’s an example. You can look at diversion.
The example with Staples and Office Depot. We had bidding data. You could see how often they were bidding against each other. And if one were to raise their price, how likely it would be the other would get the business. That goes right to the heart of how economists think about this. I think it fits very well with the statute in terms of lessening of competition. You can add margins in there, too. Very important to the arithmetic of how that works. That’s the next level of sophistication.
Then when you go deeper to, “Wait, I want to do a merger simulation model,” that’s where things get, I think, basically too complicated for a lay judge to really evaluate very well. Much more trouble could be made by an economist who wants to muck things up, and you get more of the battle of experts that’s opaque to a non-expert. So I’ve not tended to want to go in that direction, unless I absolutely have to for some reason methodologically. But if you need to go to the next level there, I think what you need is some sort of special master or other court-appointed expert. I’ve pushed for that, too, but I think the defense bar doesn’t like it.
Eric Posner
Let me come back to this point about corporate power, which is actually in an interesting way connected to the neglect of coordinated effects.
The problem is that, as you have this better sort of more rigorous method, the courts are kind of drawn to it, in a kind of gravitational way. You’ve got these two experts with a lot of data, doing unilateral effects analysis. And then you have this bunch of cliches about coordinated effects, like, there’s a smaller number of firms and they can monitor each other. That stuff just doesn’t seem as impressive, even though in the real world it could really matter a lot.
But there’s also this point about corporate power. Carl said earlier, when he was working on the doors and door skins case. It has nothing to do with corporate power. These two manufacturers of doors and door skins get together. Like, are they going to have more power—I should say, political power—more power over, you know, to influence legislatures? Are they going to be harder to enforce the law against when they violate the law because they have more resources?
There’s no data out there that you can use to make a prediction about the impact of the merger on, the effect that this new firm is going to have on, legislative outcomes and regulatory outcomes. There just isn’t any data there. I think this was Tim [Wu]’s fish-breathing-in-water point. We know that that’s what’s going on. Even if some of us are skeptical about it, it’s clearly what motivated Congress and has repeatedly motivated legislatures over the years. Which is, by the way, why the Reagan Administration never tried to get Section 7 amended and convert it into a pure economic test. Because people are deeply concerned about it.
Where does that leave us? For an economist, an economist might say, “I don’t believe this stuff,” or, I think more plausibly, “Look, I don’t know how to measure it. I can measure prices. But I can’t measure impacts on political outcomes, so I’m going to disregard it.” But we can’t disregard a statute that is designed, in part, to prevent corporations from becoming so powerful that they can impact…
Carl Shapiro
See, but you just made that move. The statute talks about lessening of competition in the line of commerce in a section of the country. The courts have interest in that.
You can have this debate all you want about broader effects. It’s not what’s happening in the courtroom. It’s not what’s happening at the agencies. And I don’t think it can or should happen until the law gets changed because that’s what the statute says.
Eric Posner
No. It doesn’t.
The statute could have said…
Carl Shapiro
Wait. The statute doesn’t say lessen competition? I thought that was the one thing we agreed on.
Eric Posner
You’re defining competition in terms of its economic effects. But competition has other effects as well.
There are two things going on. Just think of… All right, I’m going to exploit my comparative disadvantage here. Think of the Cournot model. There’s the independent variable here. Number of firms. Okay. Then there’s the dependent variable. Right? The number of firms is rivalry. In that particular model it’s a measure of rivalry. And the ultimate price or output is an economic effect that we care about.
We could also do a political Cournot model. Political scientists have done this. There’s a second dependent variable that you’re looking at, which is political outcomes, which affect legislature, which might affect prices, or it might affect freedom, or it might affect something else. Congress wasn’t making this distinction. It was worried about trusts and big corporations because of a range of bad social effects. It is not clear… And I agree with you here. I definitely agree with you here. I do not think a court should, when it’s evaluating a merger, try to make a specific prediction about whether this new corporation is going to, you know, give donations to Republicans or Democrats.
What I do think is the case is that the argument for just a very strict “rivalry only” test, which is what is in the statute, is a way of preventing mergers that produce these other diffuse, unmeasurable, but important outcomes that Congress cares about.
Deni Mantzari [45:30]
I think, just to build up on this point: Is economics the only source of wisdom? Should economics be the only source of wisdom in merger control? With this question, I can open the floor to your questions.
We can take this first question over there.
Audience member 1
Directly on this question of the politics and the economics. I think one way to think of this is, the political market is not the same thing as the economic market. In other words, you could get a merger, which you thought had no economic impact at all, yet made that firm very large, and therefore gave it political strength.
I think when you’re worried about politics, it’s a completely legitimate thing to be worried about politics. But the measure of whether a company is strong politically may have nothing to do with its lessening competition of one market, but it’s overall size. If you want to say it’s bad because it’s big, you should go after that. That’s a totally different set of tools. You can tax those firms very heavily if they get large. You can put some cap on how big firms can get. But that’s a completely different thing from thinking about the antitrust question, which is about significant lessening of competition in an economic market. Not a political market. It’s a cognitive confusion, and I think it leads to bad law.
Eric Posner
I understand what you’re saying, but you’re assuming that there are different ways of modeling corporate influence on politics, and you’re saying, “But we might be concerned about just a really big company with, you know, tons of revenues and resources.”
But a really big company that has tons of resources and revenues, but is in a competitive market, or in multiple competitive markets, may not be able to have much influence on politics or the incentive to affect them because its competitors are also… If the firm tries to, let’s say, get the government to raise tariffs, it’s benefiting its rivals as well as itself. So there’s a free rider problem. And so another theory is that, as corporate power increases, and as market power increases…
Audience member 1
I think it just is a confusion between two things. One is the question about lessening of competition in a market. Another is the broad political thing.
Eric Posner
You’re redefining competition to mean price effects. Price effects are one effect of the reduction of competition. There are other effects.
Carl Shapiro
Let me put this in very practical terms. Okay. I hear you, Eric. You want to talk about… You want broader political factors to be… We should be thinking about those more because that was what Congress was thinking about in 1950 or 1914.
I guess I’m going to say to this group, as somebody’s been doing this for a while in the trenches: If we want to make progress with the current law, the statute and the case law, my view continues to be, I think we need all the more to make progress because horizontal merger enforcement is too lax. We have more, increasing evidence about that. The way to make progress is not to go off in some very different direction. It’s to discipline the use of economics in these cases, so we get more decisions by judges that are accurate, in terms of which mergers actually “may substantially lessen competition,” with the may being emphasized, as the current guidelines do.
I think bringing in all these other issues is not a helpful way to move forward.
Audience member 2 (Simcha Barkai)
I’m willing to accept that economic analysis can get the answer wrong in certain cases. I’m also willing to entertain a broader definition of competition, what sphere we’re talking about. Can you please provide a description of analysis and criteria for determining whether competition may be lessened by a merger, however you define it?
Analysis and criteria, yes or no, how do we decide?
Eric Posner
HHI thresholds are a good place to start. Low thresholds. In a simple horizontal merger, that’s in fact what courts do.
You need to define a market as firms that compete with each other. I’m not saying that economics plays no role here. I’m talking about the standard of liability that is used to evaluate whether a merger is lawful or not.
Deni Mantzari [50:25]
Tommaso and then John.
Audience member 3 (Tommaso Valletti)
Thanks. I just wanted to continue on what Simcha said. So you said, perhaps economics has not helped in reaching good decisions in some cases. You accepted that. I have in front of me an article, which is going to be published in a law and econ journal where Filippo and Deni are associate editors, so let’s assume it’s a very good journal, in fact.
This paper is interesting. It’s a meta study about consummated merger effects. A retrospective on retrospectives of horizontal mergers. What’s interesting about this paper is that the author is looking precisely at mergers which had been vetted by antitrust authorities. Precisely what we’re talking about here. Mergers where some antitrust investigation was had and where economic models had been used. They were approved—with remedies, without remedies, but were approved—because the enforcers thought that consumers would be better off, or at least not worse off. Otherwise they would not have been approved.
If our method (which we applied in the trenches, not about esoteric stuff) were correct, you would observe in practice when you do a retrospective, that prices either went down, either were efficiencies, or were constant and nothing had happened. Okay, perhaps these models were imperfect. Now, what this author finds is that in 52% of the cases (and she has hundreds and hundreds of these mergers) prices went up. It’s like tossing a coin. In 52% of the cases, prices went up. So it’s a bit of a stretch to say, “In some cases…”
[inaudible comment]
Well, remember these are approved mergers. They were approved because it was believed that consumers would not be harmed. Otherwise you cannot approve a merger if you believe that consumers are going to be harmed. You cannot, okay?
So it’s a bit of a stretch to say that there has been some good and some bad economics. Because on balance, we have some frameworks which are very coherent intellectually, but a bit detached from reality. Because our frameworks have been developed in the context of some models of horizontal product differentiation, static models, etc. If in practice our methods have little predictive power, what do you do? What do we do? How do we use good economics? So I want to go back to a challenge for the future. How do we make sure that this track record, which is very poor, doesn’t become even poorer, and hopefully improves?
Deni Mantzari
Can we also take the second question, and then address both questions?
Audience member 4
This is a fascinating discussion for me. I’m thinking about 1914. I’m enthusiastic about that. I’ve gone through all the JSTOR scholarship dealing with monopoly, which you can do now. You can figure out what was in the heads. This question about prediction, I think, is an interesting one. It’s of great concern to economics. But this concern about process, I think Professor Posner is absolutely right on the facts. Those were the issues that engaged the lawmakers and the academics. And the relative influence of, say, lawyer versus economists was very different in 1914 than it has been since, I don’t know, 1980.
So if we’re going to make an originalist argument about Clayton, as some have tried to make about the Sherman Act or the Antitrust Act of 1890 (which has been attributed to John Sherman mistakenly), then it seems to me that there’s a lot of material we could draw on. And this would be: jurists one, economists zero. That’s how I would look at it, if we really are concerned about intent.
Now, maybe we’re not concerned about intent. Maybe that intent is not helpful. But if we are, I think these issues concerning labor, issues concerning political influence, issues concerning all kinds of financial manipulation none of which have direct influence on price, were in the minds of lawmakers and jurists.
So, jurist one, economist zero. Do we care? Do we care what the lawmakers intended in 1914?
Deni Mantzari [55:40]
So, first Tommaso’s question.
Carl Shapiro
Well, I happen to be very familiar with the literature on merger retrospective, since I’ve just been finishing up a paper with Ali Yurukoglu at Stanford on that and some other related topics.
There’s a very mixed literature. There’s some areas, like hospital mergers, where it’s just really clear that there have been a lot of mergers that have caused prices to go up, been anticompetitive. In many other areas, it’s really mixed. It’s very hard to look at that literature and come up with, “Ah, I see what the diagnostic should be. How we could tell the good ones from the bad ones.” This is, of course, with the sample selection issue, only observing mergers that were allowed to go through.
I take your general point, Tommaso. I don’t think our tools are as bad as your question would suggest. I would dispute that. In general, I think our tools are pretty good. But it’s very hard to predict, generally. I’ve seen more than one judge go, “What? You’re telling me in this merger case I’m supposed to actually predict the future? I don’t have a crystal ball!” They’re simply startled that that’s what they’re supposed to do.
Now, simple rules, like Eric wants, are one way to go. We have some. We have a structural presumption. I have a paper with Herb Hovenkamp. I’ve been embracing that. It’s a critical way in which the government wins. Of course, then it turns the complexity into the market definition. There’s no getting around that. In some cases, it’s easy, maybe. In a lot of cases, it’s not. Particularly differentiated products. I just don’t think there’s a simple bright line rule. The people who say, “Oh, we’ll just do market shares.” Then they have market definition.
I think we’ve got to keep learning. I’m in favor of, in this area as in other areas, the way to proceed is to create rebuttable presumptions. Maybe it’s pretty hard to rebut sometimes. Here we have the rebuttable presumption based on market structure. I would go for a rebuttable presumption based on unilateral effects, without defining a market, with sufficiently important… And the merger guidelines are trying to go there. But they know they’re swimming upstream with the case law, so it’s a good next move to try to move the law there. I think those sort of diagnostic tools would help there, in terms of diversion, how much of the firms competing. Does it help us with coordinated effects? That particular one, that’s still a problem.
We’ve just got to keep working, but not get distracted by extraneous stuff that doesn’t have to do with competition in some line of commerce.
Deni Mantzari [58:35]
Eric, the other question maybe?
Eric Posner
Sure. That was a nice comment. Let me just try to clarify things here.
I do want to make it clear that I don’t think that price effects are irrelevant. This is one effect that should be a matter of concern when a merger takes place. It’s just that these other effects can’t be taken off the table, anymore than you can take, like, coordinated effects off the table in a world in which you care about prices.
Carl Shapiro
But you still haven’t given any constructive way of how we’re supposed to proceed that’s different.
Eric Posner
My question for you is, at what point do you decide that the price predictions are bad enough that you abandon the effort altogether? You have a number of theories, and you’ve done this. And Tommaso says to us, and the retrospective merger review literature says to us, “Well, it looks kind of random.” So, next to that…
Carl Shapiro
That’s not an accurate reading of the literature. Ashenfelter and Hosken, they have a very nice review of the Whirlpool-Maytag merger. The price went up. I was the DOJ’s expert in that case. We had all sorts of metrics about that. We predicted the price would go up, and we were right.
Eric Posner
But Tommaso said it was 50/50.
Carl Shapiro
No, that case was not 50/50.
Eric Posner
So you’re going to be able to find 50% of the cases where you’re right.
Carl Shapiro
I don’t think that follows, but go on.
Eric Posner
In any event, as I’ve said before, I’m a fan of the structural presumption. I think it should be strict, in the sense that the threshold should be low. And I think that there should be no efficiency defense. Now, on unilateral effects…
Carl Shapiro
Then what defense is there?
Eric Posner
Pardon me?
Carl Shapiro
How do you rebut the structural presumption?
Eric Posner
You could show, for example, that entry barriers are very low. You could show that the firms are stuck in long-term contracts. When the agencies are engaging in enforcement, they could use any kind of public interest consideration that is reasonable and consistent with the statute.
If you read the cases… You’ve read the cases, I’ve read the cases, and as you’ve said, the structural presumption is absolutely central in horizontal merger cases. It’s virtually always there because it’s something that the courts can understand.
If you have a strict standard, that’s a way of, not only (one hopes) preventing prices from increasing, but it also prevents corporations from becoming excessively powerful.
Deni Mantzari
We have two more minutes left, so that is your wrap-up comments.
Carl Shapiro
So you want to put all the weight onto that economic thing called the hypothetical monopolist test? Or do you want to use some other method for defining the market since you’re so big on the…
Eric Posner
I think economics is great for market definition.
Carl Shapiro
It is for that, but not the other stuff? Good to know.
Eric Posner
We agree on something.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
