EconTalk |
Michael Munger on Antitrust
Apr 25 2022

bad-amazon-300x300.jpg Are tech giants such as Google, Amazon, or Facebook dangerous? Do they have too much power? Dive into the murky waters of antitrust as Michael Munger of Duke University talks with EconTalk host Russ Roberts about monopoly, antitrust policy, and competition in the 21st century.

Don Boudreaux on Market Failure, Government Failure and the Economics of Antitrust Regulation
Don Boudreaux of George Mason University talks with EconTalk host Russ Roberts about when market failure can be improved by government intervention. After discussing the evolution of economic thinking about externalities and public goods, the conversation turns to the case...
Richard Epstein on Classical Liberalism, Libertarianism, and Lochner
Richard Epstein, of New York University and Stanford University's Hoover Institution, talks with EconTalk host Russ Roberts about the differences between classical liberalism and hard-line libertarianism. What is the proper role of the state? Topics discussed include the Constitution, prudent...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


Apr 25 2022 at 8:25pm

I think Visa is a better example than UL. The bank or credit card company gathers a bunch of data about me. Visa and Starbucks don’t have all the data the bank has. Visa basically just tells Starbucks that someone else can vouch for me to have enough credit to complete the transaction. But really Visa is providing trust from my bank to Starbucks without Starbucks having to know anything about my credit score or who I bank with.

Mike’s solution to Russ’ concern (Essentially Web 3.0) is something I have thought a lot about. I only know as much about it as the next internet weirdo that claims to know everything, but I won’t let that stop me from proselytizing:

Moxie Marlinspike (the founder of Signal and owner of a cool name) has recently been publicly critical of the Web 3.0 model. His argument is essentially that in order for us to have personal and portable data (e.g. a reputation score) we need a computer somewhere, owned by someone to maintain this data. Most people opt to allow someone like google to maintain their data on a server that they update. It takes thousands of developers and programmers, thousands of hours a week to maintain, update, and improve their wonderful and frankly unbelievably useful software that is google. If you are running your own search engine locally (e.g. Searx), yes you own all the data, but you have to maintain the server that it sits on. You have to run updates and do other software things that I don’t understand. If I outsource this, the data is sitting on someone else’s computer. They have data about me and they will require resources to maintain the server. If I continue to maintain it locally, I am fighting Adam Smith and the division of labor.

More importantly, if using blockchain software, how will this system be maintained, updated, improved? Blockchain suffers from collective action problems. Bitcoin improves in fits and starts with workarounds and in-fighting. Google improves with a push of a button and we all automatically have new and improved software. It is hard to compete with the ease of use, despite privacy concerns.

The power does not concern me much. If a company was to abuse you too much you would leave. Banks are very sticky, and most people bank where their parents did. But better banks come along, and people do move. It may take fraud, but competition lives on in the market for social media and software.

I think I know how Mike would respond; the solution to this problem is unknowable today, but presumably there is a solution. There is a very large denomination bill sitting on the street that some entrepreneur can pick up if he solves the problem, and my statement is essentially that all the bills have been picked up.

Michael C. Munger
Apr 26 2022 at 9:13am

Thanks, Ethan!

And the “I don’t know how it will be solved, but…” answer is, as you say, unsatisfying. Some answers are better than others, but there will be lock-in protecting whatever answer is selected first.

Thanks for listening, and commenting!


Earl Rodd
Apr 26 2022 at 4:21pm

Ethan’s comment said most of what I wanted to say better than I would have. I do have some small comments:
1. I think the “barriers to entry” problem deserves more attention. Mike certainly made the case that applying traditional anti-trust thinking doesn’t help out much with the big tech social media situation. Two barriers to entry with social media platforms are: 1) the up-front cost of having to have human/machine systems to censor “trash” and 2) the perhaps bigger up-front cost of having to have full scale automated systems to cope with automated copyright infringement notices.

2. Mike and Russ discussed how traditional “price” doesn’t help us in trying to match anti-trust to social media. One “cost” that I think is big, but very hard to even measure, is lack of choice. I could call it quality, but it’s a kind of quality different for different people. I can think of many features that would make me urge all my (few) friends I care about to move to a different platform. Sadly, competitors, few as they are, tend to clone. For example, an upstart called actually is in business. Yet I was horrified when I saw their main page looked just like Facebook with all the “junk” I would like to see gone and nothing extra I would like.

3. Small point on “portable reputation.” I think there would need to be multiple competing “bankers” of such reputations. Also, use needs to be voluntary to have any hope of its not devolving into a social credit system. Even in the example mentioned, porting one’s Uber reputation to AirB&B, I might not want to do that – preferring to build AirB&B as a separate business. I wouldn’t want some “teething” problems in AirB&B to destroy my established Uber reputation! That’s not because I want be a crook on AirB&B – they are just separate businesses.

Apr 27 2022 at 7:54am

Great episode, and on things that I have been thinking about a lot from a public policy perspective.

There are a range of, so called, harms that these digital platforms can be accused of causing.  And, as you rightly discussed the problem might be whether the traditional tools of anti-trust are fit for purpose (assuming you agree with the harms).

To my mind the most pressing challenge is the exercise of power with respect to control of bottlenecks – this might be control of the operating system, control of the device, control of the payment system.

Near monopoly position for the ‘core’ business where there are positive network effects doesn’t cause great concern.  However when the core service has become so ubiquitous that it is almost a utility service, or ‘essential service, for the operation of parts of the economy, then bottlenecks (and then access regulation) come into play.  Facebook as a tool for groups of friends or family or community to communicate on should be allowed to flourish (or not) according to the rules and interests of the private business running the platform.  However when that platform becomes a primary means of communication, and thus is used by all political parties for communication during an election, it seems that suddenly leaving complete control of access to the rules set by the private business is problematic.  Similarly, when innovation in markets that depend upon a platform is dependant on access to the relevant App Store, then allowing access on reasonable terms and conditions might be relevant.

Ex post anti trust type regulation can address specific harms caused to specific firms who have the wherewithal to bring the case.  Ex ante access type regulations that allow emerging firms or business to have a chance can mitigate a generic class of problems before the harm is caused.

Any reaction to this would be interesting.

Keep up the fabulous work.


Apr 27 2022 at 12:22pm

Users are not the customers Facebook or Google. Advertisers are. Contrary to Russ and Michael’s quick dismissal of advertisers, there are not substitutes for many companies to working with Facebook or Google. Depending on your product, you are either dependent on Facebook or on Google or on Amazon. My last business received 70% of its customers from Google. Amazon would not let us advertise on their platform because they saw us as competitive. Facebook did nothing for us. When I started with my business, Google was an important, but minor source of traffic. By the time I left, our engineering teams was driven by changes in googles algorithm. Some of this dominance came from googles superior product. However, that is not the whole story. They were allowed to acquire competitors. They were allowed to vertically integrate and acquire the ad distribution ecosystem. By the time I left, Google knew how many clicks I got, how much I paid for them, and how much money I made on each click. Not surprisingly, every year I paid more for each click. Google was a giant margin sucking machine. Munger’s reputation ideas do nothing to address the stranglehold that publishers and small retailers find themselves in. It is worth noting that there is credible evidence that Google, Amazon and Facebook Have all committed fraud against their publishing and advertising partners. In the early days, Amazons fraudulent behavior to partners was dismissed as something in the market would take care of. Now that the competition is obliterated, their impact is discounted. These companies need to be broken up.

Brett Reynolds
Apr 30 2022 at 11:28am

Joe Heath argues that what the libertarians miss, when they say that they have a right to contract and so they should be able to contract for anti-competitive practices is that the whole reason for the contract is efficiency: the whole society is made better of because contracts allow for various efficiencies (e.g., returns to scale, etc.). As a result, contracts to engage in rent seeking undermine the very reason for the existence of contracts. This is why its unethical.

May 3 2022 at 5:26am

I really liked where this conversation ended up and good job bringing attention to the non-portability of reputation as a barrier to entry for competitors.

At one point I was thinking about the giant downside of having one portable reputation is that if my reputation is incorrectly harmed due to error I may be effectively locked out of many important services.  Similar to what happens when people have errors on their credit score, but much worse.  Ditto for cases of identity theft, someone might be able to hijack my identity and trash it.

Then I realized that with the type of system Mike outlined, there is no reason I could not maintain multiple reputations.

I might have a great Uber reputation and I want to try my hand at AirBnB.  I can take the risk and use the same ‘persona’ or I could create a brand new ‘persona’ .    Both strategies have obvious advantages and disadvantages but that decision is entirely under my control.

Being able to create and maintain multiple persona does not remove any of the benefits of the 3rd party portal reputation system and allows additional benefits.

And on a light hearted note, those wife shopping jokes really show your ages LOL.  I don’t know anyone under the age of 60 that would relate to them.

Andrew Manton
May 5 2022 at 6:46pm

Much traditional thinking is poorly suited to thinking about IT issues because the cost curves of IT “production” are so unusual.  The marginal cost for most users is almost zero, maybe a little bandwidth and some storage.  Hence the rewards to scale are so great.  Perhaps this offers some hope and alternatives to policy makers.

Comments are closed.


EconTalk Extra, conversation starters for this podcast episode:

Watch this podcast episode on YouTube:

This week's guest:

This week's focus:

Additional ideas and people mentioned in this podcast episode:

A few more readings and background resources:

A few more EconTalk podcast episodes:

* As an Amazon Associate, Econlib earns from qualifying purchases.

TimePodcast Episode Highlights

Intro. [Recording date: March 15, 2022.]

Russ Roberts: Today is March 15th, 2022 and my guest is Mike Munger of Duke University. This is Mike's 41st appearance on EconTalk--last year in December of 2021 to discuss constitutions. Our topic for today is antitrust. Mike, welcome back to EconTalk.

Michael Munger: It is a pleasure as always. Thank you, Russ.


Russ Roberts: So, this is based on an article you've written recently, that we will post link to, trying to see the role of antitrust in today's world as possibly very different from in the past and how it might relate to, actually, our previous conversation on constitutions.

But, let's start with talking about what we might call--and you call--traditional antitrust policy. What was traditional antitrust policy?

Michael Munger: We have a very good podcast on EconTalk about this where you talked to Richard Epstein. And, as Richard, not surprisingly correctly, describes it, traditionally contracts and restraint of trade at common law were not enforceable.

And, if you're sort of outside of being trained in economics, you might say, 'They can still make the contract. I just can't go to court to enforce them.' Well, that's often enough.

So, you and I are competitors, and we write a contract saying we promise not to cut price below a certain level. That's called price fixing.

And then--but you do it--because the competitive pressures to do that are--that's why we rely on competition to discipline the activities of firms in industries.

So, you have strong incentives to cut your price. If you think I'm not going to cut mine, you have strong incentives to cut yours; and I take you to court. And, the court says, 'Well, this is not enforceable.'

And, just the fact that it's not enforceable means that we can rely on competition and the temptation to defect to prevent raising price and restricting output.

Now, there's an alternative contract we could write and that's called a merger or acquisition. Instead of having a contract saying, 'I can't cut your price. We can't cut our prices,' I can buy you.

And then, this is the Coasean solution. So, Ronald Coase would say that if one firm buys the other, then they can unify their pricing and output decisions. They could increase their profits by doing that.

So, in some ways the origin of antitrust was that firms seem to be moving in the direction of collecting into larger and larger firms, rather than trying to do the old thing, which was to have contracts and restraint of trade.

The common law restriction on non-enforceability was sufficient to discipline firms for the most part, but in the late part of the 19th century firms recognize we can solve that problem by merging.

And, those kinds of mergers have an intermediate stage. It doesn't have to be a full merger. We can create a larger umbrella legal organization called a trust, and all of the firms enter their asset into the trust. They agree to have those assets managed by a board of the trust, and then they get dividends.

So, antitrust is literally trying to break up this intermediate form called a trust--that's not a full merger or acquisition--that is a way of achieving the objectives of those contracts and restraint of trade that under the common law we're not enforceable.

Russ Roberts: Let me try to summarize that. That was pretty intense, man.

Michael Munger: There was a lot there.

Russ Roberts: There was a lot there. So, the way I understood what you said is the following. We want to protect consumers. We're going to come back later and talk about whether that's the right measure of what we should be trying to do in public policy in this area.

But, we want to protect consumers. We're worried about firms raising price and selling less than they might otherwise sell to the consumer. And, in the--

Michael Munger: Can I interrupt for a second? All firms always want to raise price. They're always trying to do that, but competition prevents it.

So, the reason we get low prices is that it is in the firm's self-interest, in a competitive system. They do not love consumers.

Russ Roberts: No, they don't.

Michael Munger: They want to raise price. But the reason they charge a low price is that it is in their self-interest.

Russ Roberts: Right. And, that is the case--just the first point you made was that you'd think we could avoid that self-interest of the firm's desire for a high price by saying, 'Hey, you're making the same thing I'm making, the same product, let's make an agreement to fix price.'

But, once the law--the common law--refused to enforce that kind of contract, that meant that firms' natural impulse to compete is going to be unleashed because they cannot create an agreement enforced by the state to not compete.

So, just the nature of common law and what contracts would be enforced versus not enforced at least had the potential to unleash competition. So, firms had to find a different way to make higher profits than they otherwise would in competition with each other.

Michael Munger: They didn't have to find a different way, but competitive pressures for profits meant that entrepreneurs sought new ways.

And, entrepreneurship can take both competitive and non-competitive avenues. The question is whether the state grants this.

Russ Roberts: So, if I can't make a legal agreement with you that the state will enforce to not lower our price, I could buy you, you could buy me, and then we could charge a higher price because there wouldn't be anyone else to undercut us.

And, then now you've talked about an in-between solution from the late 19th century called a trust, where we wouldn't literally be the same company, but our assets would be jointly managed by some umbrella organization called the trust.

And, that would effectively achieve the same end as the merger. Is that good?

Michael Munger: Yes. And, trusts, unlike contracts and restraint of trade, are enforceable in court.

Russ Roberts: Or they were.

Michael Munger: Yes.

Russ Roberts: So, the anti--

Michael Munger: So, at the time, trusts became the answer. So that, the argument was that trusts were not inherently an improvement in efficiency. There were an instrumental way around the common-law restriction on enforceability of contracts and restraint of trade.


Russ Roberts: Okay. Now, for economists listening and other people who might be thinking about this, obviously how competitive an industry might be is often a opaque question.

It's not clear what level of competition exists in an industry. There's a big debate going back many, many decades about whether sometimes two firms might be enough for them to compete as long as they can't legally agree to fix prices.

And, they can illegally agree. They could have an implicit or explicit agreement. It wouldn't be enforced by the courts--this is before antitrust law came along.

But, the other possibility would be, 'Okay, so it can't be enforced by the court, but you and I will sit down, we'll say: look, this seems like a good price for us; let's use it.'

Now, antitrust policy comes along, and that's called price fixing. That's against the law. There's an interesting debate among some of our friends as to whether you really need that.

But, let's go back to that point. So, we're in a world now where firms in their urge for higher profits have created these umbrella trusts; and their prices presumably are higher, their profits are higher. They like it a lot. But, what happens?

Michael Munger: Well, let's add a parenthesis. You use the word 'competition,' and we haven't defined it. And it is a very difficult term to define.

So usually, economists, when they say competition, they mean that every firm is so small that it's a price taker--which means that the price is given by the collective interaction of demands and supplies. And so the result is that everyone's a single price.

So, if I grow wheat and my neighbor grows wheat, we are competitors. But, if you ask them, they say, 'Well, we go to the same church. We're not really competitors.'

What the world thinks about competition is Toyota and Ford. Two companies that actually do set their own prices. And, they're conscious of the interaction between their price-and-output decisions and the price-and-output decisions of other companies.


Russ Roberts: I got to stop you there, though. Sorry, sorry. This is a really important point for listeners. It's one of the deepest things that economics understands; and I'm stopping you, Mike, because I think you may have confused some folks who aren't used to the way we use words.

So, every firm that sells stuff gets to write down somewhere on a sign or in a brochure or online what its prices are. So, in that sense, every firm sets its own price.

Now, I've been talking on the program a couple the times about how in Israel and in other places and sometimes outside the United States, the price is not explicitly portrayed somewhere, so you have to kind of negotiate.

But, forget that part. That's a subtlety. In America, when you walk into a grocery store, every price in that store has been physically posted by the grocery store.

When you go to Lands' End or Amazon or any online retailer, their prices, they're sitting there. That can lead people to falsely believe that the firm, by writing those down, gets to choose them.

Sometimes that's true. Most of the time it's not, in many situations at least. It's true you can write down anything you want, but if you pick a number that's too high, no one will shop at your store. They will shop at your cheaper competitor's.

If you pick a price that's too low, you'll have so much demand that you will have a very unpleasant experience trying to figure out who should get the scarce stuff that you have.

When we talk about a competitive price in economics--or a better way to say it, when we talk about a price taker, the phrase we used a minute ago, that is a firm that takes price as given. So, the number it writes down in its catalog or on the sign in the store, it does get to write down whatever number they want. But generally they are writing down what they perceive to be the price that other people are charging, more or less, for a good of the quality that they're selling.

And so, 'setting price' is a very challenging phrase because in economics it means something very narrow. Whereas for everyday listeners, it means something much broader. So, sorry about that, but that's an important footnote.

Michael Munger: No, though--the reason that I wanted to make the stark distinction between the two endpoints was to talk about a point in the middle.

So, often we think that firms that have discretion about what price to charge, if they all charge the same price, that must be the result of collusion. That is, we have some kind of price-fixing contract.

Well, that's not true in wheat. All farmers get the same price for wheat. But that's a competitive market.

So, I go to an intersection and there's three gas stations. Right now in the United States, in North Carolina at least, the price of gas is $4.31.

Russ Roberts: Per gallon.

Michael Munger: Per gallon. All three of them have $4.31 per gallon. Is that competition or is it collusion? Many people say, 'Oh, it must be collusion. Look, they're all charging the same price.'

And they can change the numbers. They chose the same price.

It's probably competition. If they raise their price, nobody will buy any gas. And, if they cut their price, the other firms will match them.

Russ Roberts: And we have to add--or I want to add; you can agree if you want--that if you're at a gas station, if you're at an intersection and you see that one of the gas stations is a little more expensive than the others, it could be they have some locational advantage--the way the traffic flow is. It could be it's a more pleasant experience. It could be stuff works better at their place, or it could be a short-run blip that hasn't been fixed yet, but probably will. That's the way we look at the world.

Michael Munger: There are differences in quality, but what I'm saying is suppose you see all three happen to have the same price. Is that competition or collusion? We actually can't say.

Russ Roberts: Yeah.

Michael Munger: So, if Ford and Toyota call each other and agree on a price--'You charge this, I'll charge this'--or if they call each other and they agree on output quotas--that, 'You only build this many cars, and I will only build this many cars,' and it's fewer than they could build--that seems like anti-competitive behavior. That seems like an agreement in restraint of trade. 'I promise not to cut my price,' or 'I promise not to produce more than a certain amount.'

So, those are the things that for the first time, the Sherman Act, which was passed in 1890, it's often called the Sherman Antitrust Act.

The Sherman Act has two clauses.

One has to do with the structure of the industry. That is: Are there so few firms in the industry that the firm has excessive market power? And, 'excessive' might be in the eye of the beholder.

The second is about behavior. Is there an activity that would tend to monopolize the industry in the form of restrictions on price or restrictions on output?

So, for the first time, in the United States at least--and really pretty much in the world; the United States was one of the world leaders here--we moved from the common law restriction on the enforceability of contracts and restraint of trade to criminalizing, actually making it a prosecutable offense.

So, different from, 'I don't get to enforce the contract,' to 'If I engage in these activities, that's per se illegal.' It is, on its face, illegal. All I have to do is show that you negotiated an agreement to restrict output to a certain level or to keep price at a certain level and you will be guilty of a violation of the Sherman Antitrust Act.

Russ Roberts: I probably used this example a long time ago, maybe even when we talked about it, but oh maybe 20 something years ago, they tapped the phones of airline executives. And they caught a major American airline CEO [Chief Executive Officer] calling another major American airline CEO saying, 'You know that route that we're killing ourselves on, not making any money, or maybe losing money, because we're so zealous in our competitiveness? Let's stop that. Could we just charge X?'

And, the other executive said--there was a silence, and he says, 'We can't talk about that.' And the other guy says, 'We can talk about whatever we want.'

Yeah, you can, but you can go to jail.


Russ Roberts: Now, before we go on, you and I have a lot of friends who think that we don't need a law against that. That: We could agree to charge a certain price, but we will have trouble sustaining that--for reasons we've talked about earlier--or we may have to hide our ability to not keep the deal.

There's a lot of cartels that promise a certain--one of the members of the cartel promise to restrain output at a certain level. But, so, they sell the oil from a different flag at a different kind of carrier. And, they figure out a way to violate the agreement. Because, when you restrict output as a group--and by doing so you drive up the price. Everybody then has an incentive--this is a very important point--to cut the price on the side as long as everybody else keeps the agreement.

So, that's why our friends tend to think that those kind of agreements are pretty unsustainable. Because there's no legal way to sustain them. And the natural impulses to cheat on the deal are always going to be there. What are your thoughts on that?

Michael Munger: I have a lot of thoughts on that. I'll try to be brief.

First, you and I have said twice now if everybody keeps the same price I have an incentive to cut the price. That's true, but it's also true if everybody else cuts their price.

So, regardless of whether everybody else keeps the same price or cuts their price, I have an incentive to cut my price. Knowing that, I will cut my price first; and that's what undercuts the cartel agreement.

So, it makes it--I don't want to be the patsy who waits last and loses market share. 'Dammit, they lowered their price. Who would've thought?'

Russ Roberts: Yeah: I'm sitting there, I've got all this limited amount I agreed to produce so that we can all agree to keep the price high. I don't have any customers.

Michael Munger: Yeah. 'I'm a loser. I can't believe I believed them. I never will again.'

Russ Roberts: Yeah.

Michael Munger: So, this is a fundamental disagreement. And, again, I want to credit Richard Epstein for having pointed this out. There's a fundamental difference, even between economic conservatives and libertarians, on this point.

Economic libertarians tend to think: This is my property, that's your property. If you and I want to write a contract, the state has no right to tell--because it's a natural-right-sort-of-interpretation of contracts. So, my right of contract means that you and I get to write contracts in restraint of trade; who is the state not to say we can't do that?

Economic conservatives--and Richard Epstein is one; and George Stigler was one. George Stigler was a big fan of the Sherman Act. He thought that the Sherman Act was exactly the right level of enforcement. And so, let me say again, because I said it quickly last time: What differentiates many of our friends--you said many of our friends think it's not necessary. Many of our friends think it is necessary. The standard Chicago view accepts the Sherman Act as being right.

Russ Roberts: Well, we have a lot of friends, so it's complicated.

Michael Munger: And, I have a lot of libertarian friends. But, what's interesting is that this divides--this question of the Sherman Act divides economic conservatives and libertarians.

So, the Sherman Act says: After a proposed merger, acquisition, or some other contractual change, is the structure of the industry conducive to competition?

And, if it's not, then the merger or contract provision is illegal. It should be turned down by the antitrust authorities.

The second question is behavior. And, that might include forming conglomerates, resale price maintenance, tying arrangements. All of those are behaviors that might lead toward market power.

So, what's important about that distinction between economic conservatives and libertarians is the claim that the antitrust authorities might very well act, not just as a law-enforcement agency, but as a regulatory authority.

And so, that sort of entering wedge--there's a difference between law enforcement. So, 'Here's the law. It's clear. If you violate it, you will be punished.'

That's not what we do in antitrust. What happens is: If you and another company want to merge, you have to go and ask the permission of the antitrust authority. So, the Antitrust Division of the Justice Department and Antitrust of the Federal Trade Commission. You have to get the permission of those authorities.

And it's not clear what the standards are. It's going to take months or years. You have to submit a huge number of documents. And, since the law--the standards--are not clear, they're acting much more like a regulatory agency than a law enforcement agency.

And, I think that would have surprised the people who passed a Sherman Antitrust Act. Because, they were thinking of it more as: Here's the law. If you violate it, you go to jail. That's what George Stigler said: good. That's what economic conservatives say: good.

I think most economic conservatives would agree with libertarians that what antitrust law had become by the post-World War II period was wrong, incorrect, a mistake.


Russ Roberts: Well, let's turn to that, because we haven't used the word much so far, which is: monopoly. But 'monopoly' literally means one seller, a single seller.

And obviously, if you're the only seller, you have some control over price. You don't take price as given. You, in order to sell at a higher price, are going to produce less than you otherwise would.

And most--I'd say everybody--would say, 'Well, that's going to be kind of hard on consumers,' relative to at least an imaginable alternative.

Michael Munger: We haven't said why.

Russ Roberts: Right. But, we would generally say--let me say it differently. Public policy--among economists; I'm not going to say among politicians--public policy among economists has historically, until recently, focused on the fact that consumers are made worse off if there's less competition in an industry.

And, I would argue that you could pretend that's the ideal by which the Sherman Act and other interventions in market behavior were passed.

But, it's not obvious to me that that is the effect or the actual reason. A lot of times, it seems to me--you can agree or disagree, let me know--it seems to me a lot of times the actual bureaucrats don't care so much about consumers. They respond to political pressure, all kinds of things that have nothing to do with the economists' blackboard idea that fewer competitors, say, or market concentration, or if the extreme monopoly are bad for consumers. They care about lots of other things. Do you agree with that or not?

Michael Munger: Again, gosh, that's a big grab bag of stuff.

Russ Roberts: It's hard being a guest on the program. I don't know why you would come back for punishment 41 times.

Michael Munger: This is it. Forget it.

Russ Roberts: This is the last one. Oh, no.

Michael Munger: Yeah. I've had enough.

The argument against monopoly is not so much that the firm makes excessive profits. Although, there's some of that.

The argument against monopoly is that the restriction on output and raising price means that there are many consumers who would be willing to pay more than it costs to make the product, who are not allowed to buy it because of the artificial restriction on price.

So, economists have a term of art: it's called a 'deadweight loss.' And, Arnold Harberger, at the University of Chicago, famously estimated the size of the loss that results from all of those--they turn out to be triangles in the demand and supply diagram, or something close to triangles.

And so, the cost of monopoly is foregone alternatives for consumers that they would have purchased, and the difficulty of innovation.

So again, the foregone options that consumers would have because in a competitive market you have more reason to innovate and improve quality and produce more things. So, there's a static and dynamic component to the argument against monopoly.

So, the question is: Is that really a problem? How would we tell that there's enough market power to constitute something like monopoly?

Now, when I was at the Federal Trade Commission [FTC]--I was there with my good friend Bill Duggan, from the University of Chicago. He was at Dartmouth then, but he's trained at the University of Chicago. And he proposed a test for monopoly.

Now, the point is: it's not monopoly before the proposed merger. The question is, what is the market structure after the proposed merger? So, post-merger, what is the market structure?

So, what Bill Duggan proposed was we rank the firms by size from largest to smallest, post merger. And, then after the first firm, we draw a line. If there are none, it's a monopoly. Yes, that was an economics joke. The point is--yeah, it was a shaggy dog story.

Russ Roberts: So bad, so bad.

Michael Munger: I'm not kidding, I'm not kidding. That is totally the test we proposed. Because you need to do this in analytic terms. So, you have a set of steps. You draw a line after the first firm. If there are no more, then it's a monopoly. If there are more, it's not a monopoly and it should go through.

Now, you might want to say something about market shares, that it needs to be large enough that it actually competes, but if it's literally not a monopoly then it doesn't violate the Antitrust Act.

So, this raises the question, the heinous criminals of the tried to change antitrust during the 1970s and 1980s, like Robert Bork, his book The Antitrust Paradox--what was his standard for the post-merger industry structure? He said: Three. There had to be at least three firms after the merger not to violate.

Now, these are rules of thumb. But notice the advantage. Instead of, 'I have to ask the Justice Department for permission,' I can just perform the test that I just proposed; and if there are two below the line after the first firm, that would mean that the merger would be okay. We can go through with the merger and the authorities would sign off. It would take no time.

As it stands, we don't have a determinate rule. So, you and I can laugh about my shaggy dog story. But if the rule is three, that would be a clear rule. And, maybe it's not perfect, but having a clear rule would be better than the current system where it's not clear whether a merger is going to be allowed or not.


Russ Roberts: Yeah. So, this is just--it's horrifying to me. Because, here's the way I see economists looking at these issues. When you have one firm, you have a monopoly and it's bad for consumers.

If you have an infinite number of firms--which I just love--if you have an infinite number of firms and they're producing a homogeneous product, you have perfect competition and it's fabulous for the consumer.

But, everything in between is what we would call--is it Greek?--oligopoly? Some--a few--sellers. So you either have one, infinite, or in between.

Now, by definition, almost always in the real world, it's an oligopoly. But, we know there are many oligopolies that are really, really competitive, meaning: It's true it's not an infinite number, but it's close enough.

And we have some oligopolies where it's not one--it's more than one, it's two or more--but it acts a lot like a monopoly would or at least it's not really that other ideal of very competitive.

So, the intermediate case is kind of like the whole thing and we don't have a measure. Now, for a long time, there were these, quote, "measures of market concentration" where a proportion of the output is produced by the three largest firms, say, or Bork's rule of thumb, or anything like that.

But, just to take a silly example, there's a search engine that you may use now and then called Google. They're not a monopoly. People call them a monopoly all the time. They do not have a monopoly. They don't sell a much either, by the way--that's a whole other issue we're going to get to. And we're almost half an hour into the talk and we better get to it soon. But Google's got an enormous market share of the market we would call the search market. We're going to debate perhaps what market we should be looking at.

But, if we call that--the issue we're worried about--and say, 'Gee, they have a lot of control over the search market': They don't have all of it. There's a lot of search engines out there still. There were a lot when they started. There's still a lot. There's DuckDuckGo; I'm sure there's five others I can't think of.

So, they don't have a monopoly. It's just that they have an enormous share of the market, and they kind of throw their weight around a lot, because they're important, really important.

So, it's not like a perfect competitive market, even though there's lots of firms.

So, that whole thing, I just think is--you know, Harold Demsetz I think wrote an article on this: that we don't really have a good way of categorizing the structure of an industry. And I think he wrote it in the 1970s. I'm not sure we've made a lot of progress.

Michael Munger: One of the problems that the antitrust authorities face is the first step when they get a petition for a merger or contract provision for them to sign off on.

First thing they have to do is to define the relevant industry or market. And, that's a very difficult problem. Because, if you define the market as being small enough, then 7UP has the monopoly on selling 7UP. It's called a trademark. So, we have an enforceable monopoly.

Companies that make drugs have an enforceable monopoly called a patent that was given to them by the state. So, the market has to be something bigger than just the individual firm because all firms by definition have a monopoly on their own product. And, that's just the result of trademark, copyright, and patent block.

Fair enough. So, how would I define the market for search engines?

Well, what I might do is look at all the companies that have search engines that have some non-trivial market share and then look at their relative market shares, and then try to conclude whether one of these companies has excessive market power. And, the question is: How much would be excessive?

What I think is interesting about the point you raised before is that having perfect competition as our benchmark means that anything that is closer to perfect competition is not only desirable, but an improvement over a situation that is not perfect competition.

So, if we were to break Google up into 10 smaller companies, that would be a prima facie improvement because it is closer to perfect competition.

That seems like a very strange claim in settings where it is the very--well, there's two things. One is that sometimes the size of the firm is a reflection of network economies. Sometimes the size of the firm is a reflection of relative excellence. So, famously, and if you're really good, more people use your product.

So, there's a famous antitrust case in 1945, the Alcoa case. So, US versus Aluminum Company of America, in which Judge Learned Hand--by far the best name any judge has ever had. 'Learned Hand' is an excellent name for a judge.

So, the question was presented to Learned Hand, that this large aluminum company has a very large market share. And there are some constituent firms: it's close to a trust but it's really a larger firm. They had all bought each other and they had 90% or more of the domestic virgin aluminum market.

Now, let's ask our two questions. Would it be better if it were broken up into smaller firms? No, because there's enormous--enormous--economies of scale in the mining, the refining, and the production of aluminum. It's extremely energy intensive. So, there obvious production benefits in terms of lower cost of having a relatively larger firm.

The second question is: what's the nature of the market? Well, aluminum, perhaps uniquely, is--the competition for virgin aluminum is recycled aluminum.

We could recycle aluminum way back in the 1930s. By the 1940s we were reusing large amounts of aluminum. And, recycled aluminum is much cheaper than virgin aluminum because it's already been refined. All you have to do is melt it down. That's expensive. It takes a lot of energy, but it's much cheaper than refining virgin aluminum.

And, by that standard Alcoa had about half of the market. And there were many recycling firms--that's pretty competitive. So, that's not even close to being a monopoly.

Learned Hand made two judgments. First: I'm going to define the market--the relevant market--as the virgin aluminum market. That's what's relevant, and there Alcoa has a gigantic monopoly.

Second, how did they acquire this? And, he found in the record memos that Alcoa was saying, 'If we can produce more, we'll be able to cut our price and increase our market share.' And, he found that to be a violation of the Sherman Act.

That's just competition. The fact that this company was trying to cut--because they were cutting their price. Remember what's alleged is: These--I started to say a B-word--these bad people kept cutting their price. That's completely anti-competitive because it drove their competitors out of business.

So, that may be the single worst antitrust case ever written by Judge 'Un-Learned' Hand.

Russ Roberts: Yeah. Oh, and that's one of the best dad jokes, or maybe jokes generally, ever told on EconTalk on Learned Hand.


Russ Roberts: I have to make an allusion--an a-llusion--to the diamond market. Because, diamonds are even better than aluminum: because they're indestructible, more or less. And so, you could--

Michael Munger: We recycle a lot of them--

Russ Roberts: Tons of them. But there's a little catch. So, De Beers is the world's largest miner and purveyor of diamonds. You'd think they'd have to compete with the ever-growing market of used diamonds, just like Alcoa had to compete with the ever-growing market of used, recycled aluminum.

But, I think you and I have discussed on this program before: One of the most remarkable things about an engagement ring and other things like it is that people want their own. They don't want someone else's. [Note: The prior EconTalk conversation to which Russ is referring was probably a conversation with Bryan Caplan on signalling, to which we have linked.--Econlib Ed.]

Now, occasionally through a family, a grandmother's ring will get passed down, a diamond ring, but in general De Beers has--eagerly I'm sure, but maybe it's not even their doing--created a cultural bias against a used diamond. Even though it's undetectable to anyone except the person who bought it.

It's really a phenomenal thing when you think about it. People, when they go shopping for an engagement ring, overwhelmingly shop for a new diamond, rather than one on eBay.

There are a lot of reasons for it. Cultural, whatever. And of course, cubic zirconium, which is an imitation, is a competitor for De Beers; but it's more easily detected. And so, De Beers has a lot of, I'm guessing, some serious market power. Just guessing.

Michael Munger: They are one of the only--part of the reason, of course, is that they have not a monopoly but a substantial market share of the original mining. And, they have to be very careful to control that. And, they are able to raise the price far above--well, I started to say diamonds are not a necessity, but my wife is home. I don't want her to hear.

Russ Roberts: Yeah, that's a mistake.

Michael Munger: Diamonds may be a necessity in a certain sense--

Russ Roberts: Some situations--

Michael Munger: This is actually one of the cases where you might concede the claim often made by people on the Left that advertising creates a market. De Beers has spent an awful lot of money shifting the demand for new diamonds out.

Russ Roberts: Trying, trying.

Michael Munger: I think they have. I think they have. I think it has had that effect.

Russ Roberts: So, if you get married, how many months' salary should you--? Well, I think maybe, I don't know, maybe 10 years' of salary would be the appropriate amount to spend on an engagement ring if you love the person.

Michael Munger: If that's the debate you're having, De Beers has already won.

Russ Roberts: I think you're on to something there. It's a good point. It's a fascinating, I think understudied phenomenon.

But I want to move on.


Russ Roberts: So, I want to try to come back to this issue of market and price, which I alluded to with search engines. So, Google has a tremendous proportion of the market share of the search engine market.

Their share of the advertising market is dramatically smaller. Their share of the online advertising market is smaller still--is also small relative to its share of the search engine market.

And, a lot of people would say--and I'm curious if you agree: Google is not an antitrust problem. They do not hurt consumers. In fact, they help consumers. They give away their product. How could you possibly think that it would be good for consumers to break up Google or to dismantle Facebook, another free product, or Twitter, another free product. These are free products. This is good for consumers. It's free. Well, they could pay us to use it but other than that, what's the harm?

Michael Munger: Some of my friends on the Left, hearing that, reply scornfully, 'Yes, and heroin dealers give away their product for free at first, also, in order to get consumers hooked. And then of course they have them forever.' So, once you're part of the Google universe, then you become part of their data network. And, these data are sold to all sorts of other companies that are then able to exploit you the way De Beers--we all know--exploits people.

Well, I don't know that. It's not so clear that having this free product makes me, net, worse off. So, I have two thoughts on that.

One is if you ask people, 'Do you wish that these online platforms'--a term we'll define soon--'that these online platforms use less of your data?' People say, 'Yes. I wish that the online platforms harvested less of my data.'

If you ask them, 'Suppose that there were an identical service that didn't harvest your data and it cost you $5 a month?' 'No.' Maybe 10%. Ninety percent of people would happily in effect sell their data for $5 a month. So, that's not so bad.

The second point is my good friend, Tom Hazlett, has a recent paper on Facebook where he looks at the antitrust case against Facebook. We'll post a link. It's fascinating what the allegation is.

So, the claim is that Facebook has dramatically raised its price. Now, you might say, 'Well, but wait, the price of Facebook is still zero.'

Well, yes, they would concede that, as Tom points out. He actually gives detailed discussion of this. But, they have increased the amount of data that they use and they have increasingly targeted ads that you are more likely to be interested in.

So, the result is that they are able to get more money from you for products--other products--that you actually want to buy. And so, as a result, the price has gone up.

Now, Tom, of course like me, has a certain bias against these arguments and would say that doesn't pass the laugh test. But, that is, verbatim--that is the argument that has been written. The price was zero, it is zero, and it has gone up dramatically.


Russ Roberts: Let me give the other side its due, which comes up now and then here, and then we'll take it into a more Mungerian direction.

So, first of all, it's not free. The price of all these platforms are embedded in the prices of the stuff I buy, and that's hidden from me in complicated ways. It's a little bit like the old days when they used to post taxes, state and federal taxes, for gasoline at the pump.

It used to be that when you buy gasoline in the United States, it would tell you the price and it would tell you what portion of that price came from federal tax and what portion came from state tax.

Now, it doesn't. And so, most people--except economists--tend to not think about the fact that--you said gas is $4-something in North Carolina. It's different in lots of different states, mainly because states differ by their state tax on gasoline.

So, it's not exactly free. The real cost of these services we get from Google and elsewhere are embedded in the prices of those goods who advertise on those sites.

Then there's two other worries. And then I think we're going to move to your one, which I think is the right point, which is about power.

So, the other point is that: It's true they send me stuff that I think I'd like to buy, and I really like that. But maybe they don't send me stuff I'd like to buy, but only stuff that people who've paid them even more money to make sure I don't see the stuff I really want, and they've changed their algorithm so that I buy the stuff that the people who pay them a lot want me to see.

So, that's alarming.

And then there's the next point, which is much more alarming to me, which is: We get our news from these places. And we get our information from these places.

People lie at these places all the time. That creates a call for truth-telling, or monitoring of claims on these sites. A horrible idea in my view, because most things are not as discernible--and we can tell the version of the Dave Schmidtz story in a minute, if you'd like, Mike. But, they are powerful, because they survive and thrive by having massive amounts of people on the platform.

And, because of that it's really hard to find an alternative--there are alternatives. They're quieter. They're not as exciting; and that gives them the freedom to do stuff to me. And, it's true: they might be restrained by doing that from all kinds of different ways. But, the fact that there's this wiggle room for taking advantage of me: I don't like that. I understand why people don't like it. I'm not sure that there's an easy solution to that, but I understand why people are upset.

And, now to keep us focused, since I've taken you off down 10 different lanes--you can go any way you want--my point is that traditional antitrust is obsessed with the lack of competition leading to higher prices and less output.

The current giant players of the internet: that's not the issue. At least, discernible issue. Most of, a lot of what they give away is free. They do a great job. Most of us like them.

The issue isn't, 'Ooh, I paid a lot for that. We need to do something at about that.'

It's about power.

And it's not market power. It's related to it, but it's not the traditional antitrust argument that: Because you have market power, you can raise the price. This is: Because you have market power, they can do things to you that are not even so transparent.

Michael Munger: That's an excellent segue. So, let's take Facebook as the example. So, the usual problems that we talk about in antitrust are trusts, horizontal mergers, vertical integration, or contract provisions.

What we're seeing now with these platforms is conglomerate acquisitions. So, if Facebook buys TikTok, that's a different thing. If Facebook buys WhatsApp, that's a different thing.

And so, a conglomerate aggregation is when a firm with a monopoly in one part of the market--that it is the--because Facebook has a monopoly on Facebook. Unsurprisingly. And, it provides such good service that many of us want to be able to watch each other's cat videos without switching--or baseball videos--without switching platforms.

And so, of course it got larger. There's going to be a single large firm when there are substantial network economies like that, where all of us being on the same platform has a lot of advantages.

Now, a conglomerate merger is when, for example, Facebook buys TikTok, and is therefore able to exploit synergies. It extends its monopoly into this other market, or at least that's what's alleged.

So, what's interesting about Facebook is Facebook bought a bunch of other companies. Each of those mergers was approved by the antitrust authority.

It was years later that the antitrust authority is now saying, 'Wait, not so fast, you have power.' Well, what was it that they could have done that they would have known?

So, you can find memos saying, 'We would like to increase the number of consumers on Facebook who use this company we're buying called TikTok.'

Evil! Clearly evil personified!

Well, no, probably not.

And so, the question is: What actually is the standard?

So, we have a set of laws. The firm applies to the Federal Trade Commission in good faith and the merger is approved. Five years later they are prosecuted and fined billions of dollars, if the current suit goes through, for having done--what? It seems like an ex post facto law.

So, the Dave Schmidtz story, the reason why this is so difficult, is Dave Schmidtz has a parable he the calls Desert Town.

And, desert is D-E-S-E-R-T, meaning deserve. So, the hero of the story is driving through a town, is pulled over by a policeman. The policeman says, 'Do you know what you did wrong?'

And the hero says, 'I was driving on this main street and there's a bunch of cars stopped at the cross streets. There's no stop sign. There's no stop lights but everybody was blowing their horn at me and shaking their fists. I really don't know what I did wrong. I just kept going because I thought I would have the right of way. It's just a convention: If you're on the main street, then the cross streets stop. So, I didn't see that there was a problem.'

And, the cops: 'Yeah, I figured you must be new in town. In this town we allocate according to desert. That is, you have to deserve things. So, if you come to an intersection, everybody has to get out of their car, go to the center of the intersection and discuss who has the most urgent or morally significant errand. And, if you have the more urgent or morally significant errand, when we go back to our cars, you get to go first. Then the others go. And then the cars next in line, they get out of their cars and they come to the middle of the intersection: They discuss it and they do the same thing.'

And, the hero stares at the cop and says, 'Okay. You've got to be joking. I don't understand what the joke is.' And, the cop says, 'Justice is not a joke, sir. I was going to let you off with a warning until you said that.'

So, consider two alternatives. One, manifestly unjust institution called the stoplight. You may have had this experience, Russ. I know I have.

I have been really late. My urgent errand was pressing, but there was a red light and I had to wait two or three minutes while a bunch of schmoes--and they weren't going anywhere, they could have let me by. But, the light was holding me up.

If you take a snapshot of that system, look how unfair it is. These cars are being stopped and those cars are being allowed to go. That's completely unfair.

However, dynamically, that system over time means that at most I have to wait two minutes. Now, it has nothing to do with the merit of the urgency of my errand. It is arbitrary. But, I have to wait at most two minutes.

The system where we get out of our cars and go to the center--imagine this in Jerusalem, in a major street in Jerusalem or Tel Aviv. Plus, I don't mean to--but, the length of the argument might invoke many parts of Hebrew Bible doctrine. There might be a bunch of claims.

Russ Roberts: Yeah.

Michael Munger: So, the level of argumentation would be excellent. It could be 20 minutes. But, probably it would be a minimum of seven minutes.

But, it is true that there would be justice in the sense that--suppose people are honest, because you know your errand is more urgent.

So, after seven minutes, the more urgent errand would get to go. So, there's two systems. One: maximum wait of two minutes but unjust.

Two: minimum wait of seven minutes, but just. It is not hard to imagine that we would unanimously pick the stoplight over the everybody gets to argue.

Because this is a Welleslian[?] principle. The least well off are better off with the stoplight than they are--so, justice is a fine thing, but the costs of achieving justice are very high.

Now, in the antitrust case, if I don't know whether something that I've done is illegal, and I won't know until the authorities tell me years later, that means that the things that I would do, even the things that improve market share and improve consumers' welfare, I'll decide not to do, because I can't tell whether or not it's going to be allowed. And it may cost me billions of dollars if I lose the suit. And, it will certainly cost me hundreds of millions of dollars if I win the suit, because the litigation itself is incredibly expensive.


Russ Roberts: So, you and I might look at some acquisition by Facebook or Google or Amazon and say, 'Well, that obviously isn't good for consumers. It's good for them.' And, therefore we should probably just not allow it. And, if we did allow it in the past, we should now change our mind. The argument would go and say, we should break them up, split it up, divest them--they should be divested from some of these acquisitions.

Of course, some of those acquisitions are, strangely enough, the acquiring company knows more about it than we do. They have a plan for how they might use them.

We could, at the stoplight, require them to make that case; and on a case-by-case basis decide what's good or what isn't.

I think the real question is: Is there an alternative? Wouldn't that be better?

The things on the table, especially the things on the table from traditional antitrust justifications, seem like a really bad idea.

So, the question for me--and I'm probably kind of a broken record here over many, many episodes, and you've probably heard them all too, Mike, I know you're a pretty religious listener--

Michael Munger: I'm a big fan of EconTalk.

Russ Roberts: And, you should be. That's why he's on 41 times.

Michael Munger: I've got a sweatshirt. I have a lot of merch. I should have worn the sweatshirt. I should have modeled it.

Russ Roberts: Yeah. You're looking at kind of dapper in the tie. For those only listening on audio--which is most of you by far who are hearing this--Mike not only is wearing a tie, but there's a unicorn in the background, as there always is on video with Mike. And, he's got a unicorn cup also, which I love that. Just Google unicorn Mike Munger, and you'll find all the insights Mike has into all the life cycle habits of your and my favorite corned animal.

Okay. So, my claim would be that the traditional remedies of antitrust--the prevention of mergers, dealing with price fixing, putting the consumer's welfare as the single standard--are really inappropriate for these kind of firms.

They don't work. It's not intellectually sound to even think about it this way. The alternative that some would propose: Well, let's not worry just about the consumer. Let's worry about whether it's fair to other competitors. Let's worry about whether it's good for the environment. Let's worry about whether it's good for--when we start getting into not just talking in the intersection for seven minutes, everything's going to pretty much come to a standstill. Or worse, a set of criteria will be invoked that are really probably not so good for the world. Potentially, maybe it'll be good for the world: I should be open to that.

But, I think the real issue is these issues of power over what I can see, what I discover, the things that are hidden from me in the search engine and the algorithm.

And, I'll take one example--I've mentioned it before. Yeah. A lot of people said, 'Hey, they use these cookies and they can then learn what I'm doing.' So, I assume there's some legislation and I don't pay enough attention to it that said you have to give people a choice.

I've gotten into the habit of, quote, "accepting all cookies" most of the time. I think most people do. So, all we've done really in that case is fooled ourselves into thinking we've given consumer control.

And, maybe we have in the sense that if it really got horrible we would stop checking that box every time. But, most people either don't care or don't see any harm from it.

And so, the real question is: What can policy do? And, a side note, what could economists contribute to the design of the policy that might be helpful in thinking about whether there should be any constraints on these kind of firms?

I've suggested it and others I think are out there doing things that are somewhat like this, that if I could have a little more control over my data, including things like my followers on Twitter or my friends on Facebook, my posts on Twitter--if I could port those to a competitive--a competing alternative, that would restrain these platforms, because right now I'd have to start from scratch.

And so, I have an enormous bias toward the status quo. That has made no headway. I don't see--a little headway. People have worked a little bit in this area. But I don't see a lot of people being successful in getting people to find these kind of alternatives.

Do you see anything on the horizon that might be good, better than, either A, applying the really bad paradigm of traditional antitrust to these firms; or B, worse to my taste, trying to find some new goals other than the consumer paradigm?

Michael Munger: One of the rules of antitrust, as far as I can tell, is that by the time a firm gets antitrust attention for being large, it's already done.

So, Facebook is basically done. Facebook is shrinking fast. The stock is plummeting. By the time it gets the attention of the regulator it doesn't really matter.

But, still I have a four-part answer to your question, and I'll try to get through each of them quickly. We've mentioned but we haven't defined the consumer welfare standard.

So, in the late 1970s, the Sylvania television case in the Supreme Court, the decision was there's this whole dog's [dog barking] breakfast of different motivations for whether a merger's going to be allowed.

Russ Roberts: Was that your dog, Mike?

Michael Munger: Yes. He wants breakfast.

Russ Roberts: That's like the S-I-R-I word, would you say that thing or A-L-E-X-A, all of a sudden you get attention. Is it true that in your house if you say, 'Dog's breakfast,' your dog barks?

For a minute I thought you pushed a button to play a recording of a barking dog. Can you do that again? Can we try that trick?

Michael Munger: I'll try it, but it won't work because the person he's barking at already passed. 'Dog's breakfast!' No, it's not like Siri. That would've been great.

Russ Roberts: Okay--

Michael Munger: I'm sorry--

Russ Roberts: Carry on. I apologize.

Michael Munger: I'll use a different phrase. The whole array of different arguments that are used about why we should try to solve things in antitrust, it means that we're in the middle of the intersection and argue about it because there's always going to be trade-offs among them.

So there's a certain set of an arrangement of these--'I want to protect labor. I want to get rid of inequality. I want protect consumers. I want to protect competitors. I want to increase the number of choices that are available to consumers.'

The Sylvania case--it was actually about a resale price maintenance provision--but the claim was that the Sherman Act should be interpreted as encouraging inter-brand competition, in the sense that consumers are benefited from the kind of competition that lowers price, increases output, and increases quality.

So, that's the only thing that matters for antitrust.

And so, from 1977 until very recently, that one standard has been our stoplight. There's other things that matter, but those don't count.

And, the antitrust authorities have tried a number of times to bring alternatives but generally it has become clear that by the time they bring these other kinds of cases the company is already done. IBM [International Business Machines], Microsoft, and now Facebook: By the time they've brought it, this is just recreation to allow the attorneys at the Justice Department to make their bones. You don't get a job in a private firm after 10 years by not bringing cases.

So, if you're at the Justice Department and you work for 10 years on a case, you'd get a much better private sector job when you leave the Justice Department. So, there's a sort of bureaucratic imperative--that that's what we're trying to accomplish.

Now, second, we have come to the point where an alternative view of antitrust, bringing back the dog's breakfast of other goals, is we're trying to move back to those. And that's a terrible idea for platforms, which, by their nature, benefit consumers by being large--by having network economies.

So, the idea that breaking up Facebook would help consumers is nuts. You may want to prevent conglomerate mergers, but that's not what basic antitrust is allowing.

So, if you want to disallow them buying TikTok, knock yourself out, but breaking up Facebook is just lunatic. That doesn't make any sense.

On the other hand, these firms are very large and they may have--not consumer welfare problems--but they may, third, have power.

And the question is: how should the antitrust policy be addressed to problems of power that are really social? I don't know if they're political or economic.

So the third point is: Antitrust, it's not just that we should stick with consumer welfare. Antitrust generally is a terrible way to try to address the problem of power.

Which brings me to the fourth point. And, I actually want to concede that some people on the Left have a point in worrying about power. So: All my unwritten work is brilliant. And so, right now I have a brilliant book on antitrust, because I haven't written it yet.

Russ Roberts: Oh, so good.

Michael Munger: This is a great chapter, because I haven't even started it. The best chapter in the book is about Foucault. Michel Foucault.


Russ Roberts: Yeah. Give us some background on Michel. I always thought it was Mike, but he's French.

Michael Munger: "Everybody Loves Mikey."

Russ Roberts: Yeah, they do. And, if you haven't read that article by Mike Munger, it's one of my favorites. It's really actually a profoundly deep article that its title may mislead as to its significance. But, we'll put a link up to it.

Give us some background on Foucault for people who are not in the academy, which is where I think Foucault mainly lives.

Michael Munger: Sure. Foucault is, I think, often mischaracterized as being a leftist postmodernist, and in fact he was an intellectual chameleon. At the end of his life he became very interested in Friedrich Hayek.

His main concern was about power, different manifestations of power.

And, one of his most famous writings on power was on something he called panopticon. And, panopticon was the concept he took from Jeremy Bentham.

Jeremy Bentham had written about the way to have a prison was to have a central tower where the windows were dark, but looked outward. And so, the guards could constantly observe the prisoners, and all of the cells were floor to ceiling windows looking out and bars looking in so that the light shone in from outside.

So, like, two or three people could constantly keep under surveillance all of the prisoners. But the windows are dark, so the prisoners could never tell if they were being observed or not. So, they assumed that they were always being observed.

That, to me, is the problem of Facebook and Google. We have created a private panopticon where these large firms are at least potentially aware of almost everything we say or do. Our email, our social media, Twitter--

Russ Roberts: Our purchases--

Michael Munger: Absolutely.

Russ Roberts: Travel. Where I am at any point in time.

Michael Munger: So, Foucault thought that panopticon was a direction that society was taking. He used the metaphor of the prison panopticon as a way to illustrate the sort of society that--and actually a number of authors, the book 1984 is clearly that sort of constant surveillance of everyone.

Does it matter if the firms are private? Well, the distinction between being private and public starts to disappear if the scale is so large.

Now, you can say I can stop using it. But, if there's almost no way to avoid my data being used by Google, Facebook--and the knowledge of my purchases--is it a problem of concentrated political and social power?

And, I don't know the answer to that; but I think it is right to ask the question. I just want to make the point that antitrust is not only the solution: It will make the problem worse.

What we need to do is to begin to think about the problem of political and social power. And Foucault's metaphor of panopticon is the right way to think about it. Constant surveillance of almost every aspect of our lives.


Russ Roberts: Okay. So, let's talk about my personal metamorphosis, which I know is utterly fascinating to you, Mike. And maybe to our listeners--probably not. But it might be helpful in understanding this.

So, I used to argue it's private: it's totally different. It's true there aren't a lot of competitors, but there are competitors. And, therefore it's a different thing when the government censors. It's a different thing when the government takes advantage of you, because you don't have any place to go--other than moving, which is enormously expensive.

But, now I started to think: Well, if you have to leave Google, if you have to leave Gmail, if you have to leave Facebook, for some people that's kind of like leaving the country. It's very costly. That does give them power to abuse you. So, I started to think maybe I was overly simplistic about that.

But: Let me ask you this. Suppose they exploited that power in a really ugly way. Not the way they currently use it. The way they currently use it is they send you things you tend to like--at least I think that's probably true.

But, they could abuse it and they have the power to abuse it. So, if they really did abuse it--that is, if they looked at your purchase and they sent your wife an email--or better yet they sent you the email--and said, 'This is the email we've composed to your wife about some of your recent purchases, Mike. Would you like us to send it for a small fee? We'll, be happy to throw this out.'

Now, that's--

Michael Munger: Benefit me. Right? That would be a mutually beneficial transaction. I would pay for her not to know about those purchases.

Russ Roberts: So, let's say you did. And eventually it got out that this is what one of these companies was actually doing. I think that would be a little different than that poll you mentioned earlier. I think the poll would come out a little differently about the private data.

So, one could argue--I'm not ready to argue this yet--but one could argue that the potential competition from new entrants--which is legal, but incredibly hard and it always is; but, somehow often it still happens--but, that potential competition keeps these firms in line. And they are private. It does matter, because they're not the government. They can't throw you in jail if you don't like their email and their price and all that.

And, you could say, 'I am going to--.' I mean, there's an amazing memo from Jeff Bezos--I don't know if we'll link to it or not; and it's at least PG 13 rated, but people can look it up--where he was blackmailed, effectively--essentially was blackmailed by a media company that said, 'We've got some pictures. We're going to publish them.' And, he basically said, 'Go ahead, publish them.'

And he then talked about--they had a claim as to why, that Amazon stockholders deserved to know about his bad judgment. And I think it was photographs. I might be misremembering it. But basically he said, 'I think I'm pretty good at business. I got some other evidence that might offset that. So, if that's your strategy, it's not going to win.'

But, I think that's a thin reed to lean on, the one I'm suggesting. That, if they did abuse their power in really evil and dark ways, that then the new alternatives would come along.

You could argue they've kind of done that on information. The ability of large platforms [?to?] control the information we see is, I think, an incredibly dangerous thing for democracy, both from the Left and the Right. I don't see how we're going to solve that.

Michael Munger: I have a solution.

Russ Roberts: Great. I'm so relieved.[?]

Michael Munger: It is almost as optimistic and perhaps naive and retrospect as your brief--not that brief: you had an autonomous cars period where the autonomous cars[crosstalk 01:06:12]. Historians--

Russ Roberts: I went through a phase.

Michael Munger: So, in the future, at the University of Chicago they will have a Robertseania[Robertsiania?] room. And so, in the Robertseania Museum, they'll talk about the autonomous cars phase; and that one, it didn't turn out as well.

Russ Roberts: Oh, hang on. I also had a Theranos phase, too, where I thought, 'This is incredible.'

Michael Munger: Yes, yeah.

Russ Roberts: All these tests were one drop of blood, actually it's only an atom. Actually only need to look at your blood, just look at your skin.

Michael Munger: One should not make predictions, especially about the future.

Russ Roberts: Yeah.

Michael Munger: But, I'm going to do that. And, I believe that it is possible to have different levels of blockchain identities, where you only reveal your identity to the extent that it's necessary to assure the other party of whatever transaction you're engaging in, that there's a third party that has guaranteed that I'm reliable.

Russ Roberts: Hmm.

Michael Munger: And so, I don't need to sell my data in most instances. I can own my data, and another company that is paying for access to it can then sell that assurance to the counterparty that wants to buy it.

So, the way that that might work is: Suppose that I am on Airbnb and I have a great reputation. No--suppose that I am on Uber and I have a great reputation as a driver. But I want to open an Airbnb apartment. But I have no reviews on that--which means it's going to be hard.

If I could port my reputation from one application to another--make reputations portable and free existing, and have me own them-- but, in my blockchain identity--blockchain gives me a chance to be pseudonymous, neither anonymous nor fully known. And so, the pseudonymous identity that I have would allow me to have all of these reputation scores, and it would be a general reputation score.

In China now, they call it a Social Credit Score. So, you can see the direction I'm going, where there's a potential problem. This Social Credit Score might contain where you were dissident? are there political things that we disagree with?

So, we need some third party. We need an Underwriters' Laboratories, a private nonprofit organization that takes care of these reputations.

If we could do that, then Facebook could not sell my data. All they could do--other companies could ask for access to the set of purchases that I have made associated with some 64-character pseudonymous identity--which I know is me and the third party knows is me, but which the purchaser has no way of discovering who this actually is.

We could do all of the assurance that we do now with actual identities--my name, maybe social security number--but we could do it with a pseudonymous blockchain identity. Reputations need to be portable.

Russ Roberts: I'm glad it's not complicated, that solution. [?] It's simple. It's just a blockchain thing.

Michael Munger: Russ, I just had heart surgery, and I wrote an article about how Adam Smith saved my life.

Russ Roberts: Yes.

Michael Munger: In the sense that division of labor creates emergent orders that are very complicated that we can use without understanding them.

Russ Roberts: Yep. Beautiful piece.

Michael Munger: So, I don't need to understand any of this.

Russ Roberts: True. Well said.

Michael Munger: All I need to be able to do--a third party--entrepreneurs--any problem that we can raise is a profit opportunity for entrepreneurs to solve. This is a huge one. If we could make reputation portable, and if we could make identity pseudonymous, so that people have--they're able to sell, they own it.

Right now, we serf the internet--S-E-R-F [as opposed to 'surf' the Internet--Econlib Ed.]. We throw all off value to the Lords of the Manor. So, the Facebook Manor, I'm a serf on the Facebook Manor and so I serf the Internet.

So, instead of that, if I owned my own identity--and if you want access to the pseudonymous components of my reputation, that I'll give you assurance that I'm a reliable counterparty in this transaction, knock yourself out, pay for it.

So, this is no different than other forms of contract enforcement. This is an easily-solved problem conceptually. I don't understand it, but I wasn't awake during my heart surgery, either.

Russ Roberts: Achm, achm, achm, ach. I'll just say that the serf joke was even better than the Learned Hand. It's really good. Did you make that up?

Michael Munger: It's not original.

Russ Roberts: Okay. But, it's really good.

Michael Munger: It's not original.

Russ Roberts: It's really good though.

Michael Munger: That's from the double Tapscott book [Don Tapscott and Alex Tapscott, Blockchain Revolution--Econlib Ed.].


Russ Roberts: Okay. The problem with that--it strikes me, at first glance--is that: I'm that Uber driver, I'm fantastic, I've got a 4.99, I serve great coffee, I give people water bottles, and I let them charge their phones. I'm beloved by my Uber passengers.

I open an Airbnb, I put up horrible photos that are dishonest, that don't capture where the place really is. And actually--I worry about this now, I won't go into it--but I think there's some challenges now with this model that are a little bit troubling.

But, let's say I take advantage of the fact that I have this great reputation on Uber to build a really dishonest Airbnb. That would have to--

Michael Munger: Wait. That hurts your Uber score--

Russ Roberts: It would have to rebound. Does it, in your world?

Michael Munger: You have one reputation score; and it's across the board. So, if you cheat on any margin, your reputation score is harmed on all dimensions.

Russ Roberts: Okay. Well, that might--that's it.

Michael Munger: The whole point of having one reputation score is that you cannot use it to cheat on some dimension, because it'll harm you on all the others.

Russ Roberts: I thought it was more like a silo thing. I misunderstood--

Michael Munger: It is portable across all, so you have a single reputation score. But, that's what China's using for Social Credit, also.

Russ Roberts: Which is frightening.

Michael Munger: And so, how could we keep it so that this third-party, Underwriters Laboratories for reputation, nonprofit--the government says, 'That's a lot of really interesting information. We'd like to look at that, please.'

Russ Roberts: Yeah.

Michael Munger: 'Um, well, no, you can't.' 'Actually, we have a letter from the President saying that the National Security Administration is giving us an advisory. We will look at it now. Thank you. Or we will come with guns.'

Russ Roberts: Yeah. It's got some challenges, but everything is imperfect. We know that.


Russ Roberts: I want to close with a name that we have not mentioned, which is kind of interesting, and that name is Joseph Schumpeter.

So, in Capitalism, Socialism and Democracy, Book Two, which I have not read in forever--I think it's Book Two, which is fantastic and I strongly recommend it--it's about this, if I remember correctly: You probably remember better than I do. But, the idea of it, if I remember correctly, is that small firms come along and they do their job really well.

And, they become large firms. They do their job really well, meaning that they serve customers at reasonable prices, and they innovate and everybody likes them, and they grow really big.

And that, of course, creates an opportunity for them to take advantage of consumers. And, sometimes they do. And they try. But that creates an opportunity for new firms to come along.

The fundamental question for me is whether that is the world we still live in. 
And, it's a brick-and-mortar world. You gave a bunch of examples: IBM [International Business Machines], Microsoft; there were others before them that everyone agreed, 'They're horrible, they're taking advantage of folks, we got to be protected from them.'

By the time the case finished, they were already irrelevant because new competitors had come along and taken it and done a better job of serving consumers. Large firms aren't very nimble.

Are we in a different world? That, to me, is the fundamental question. Do we need to revise our priors about the power of entry, innovation, and competition to protect consumers in this current world? Your thoughts? And then take us out.

Michael Munger: In the short run, yes. In the long run, no. And the reason is that the problem that you have raised is so severe that a lot of entrepreneurs are going to try and at first fail, but then later succeed in solving it.

So, the problem that you've raised, the reason that there are such large firms--let's say Uber. You design a new ride-sharing app and it's way better than Uber.

And so, you announce to the world, 'My app is available, if you download it to your phone.' So, I do. I'm an early innovator. Early adopter.

So, I look at my phone and I think, 'This is a great app.' And so, I look and there's no drivers in my area. Zero. But, then later that day there's one, but he has no reputation. I don't know if he's any good or not.

And so, I say, 'Well, this is a fine app but the barriers to entry from not having portable reputation means that Uber has an insuperable advantage.' The barriers to entry from having these portfolios of reputations that are non-transferrable: That's the reason you can't enter the industry.

Why can't you enter the Facebook industry? Facebook's not an especially good way of sharing information and cat pictures--or in my case, dog pictures. But, it would be very difficult for anyone to enter it because I cannot transfer somewhere else and take all my pictures, all my interactions. I don't own those things. I'm a serf. Facebook owns them. The Lord of the Manor owns them.

So, if I could--if, on Spotify, I cannot take with me the set of preferences that have been stored about music that I like--I can't go to another music platform--so, it's a barrier to entry.

A better music platform would not be able to attract me from Spotify, even with better quality and lower price, because I have all of this sunk costs that I have to start over, for the other company.

If those things were portable--if we can figure out a way to make all of the infrastructure that I have created on one of these platforms, portable to another, I own them--the serf gets property rights and becomes a yeoman.

Then--I've probably pushed this metaphor too far--but I'm no longer a serf.

I have property rights to the value that I've created because my Spotify record, that's value. Why does Spotify own that? I should be able to have that information.

And, in fact, this is like Buchanan's order emerges in the process of its development.

I didn't look and rank all the songs on Spotify. Over time, I created a set of binary comparisons: Like this; Don't like this. But that ends up being a good map of what music I like.

That should be mine. Or it could--it's not morally should be. It could be mine if we had a system where the things that I create are associated with a pseudonymous identity, that I could then port over to--I could just then take it with me to the new and better music platform. That would create competition.

The only reason there's no competition among these platforms is that I don't own my own information. And a blockchain app 20 years from now may be will solve it.

Russ Roberts: I really like Spotify. And don't say anything negative about them. And I'm--in all seriousness, what we call in economics, the consumer surplus, meaning the value I get from the product above and beyond what I have to pay, from Spotify is so enormous. The amount of pleasure I have gotten from access to that music, and the--I think I've said it before, I think it's one of the most glorious things on the internet that people have created and we don't appreciate it enough I don't think. We take it for granted now.

But, it's un-be-lievably great. If you love music. And the complaint, to bring us full circle, the complaint that I think people would say about Spotify: 'Well, it's good for people who like to listen to music, but what about people who like to make music?'

And it's true. It's a little bit harder now in certain ways for them to make a large amount of money. I'm not sure it's really destroyed or harmed innovation.

And, you could argue that the aluminum in the ground, to go really off the reservation here from the songs that have already been made, it's pretty fabulous for consumer wellbeing for another few hundred years.

Tastes do change, culture changes. But, I see a lot of people with an urge to create, and it's true, it might not be as lucrative in the future as it has been in the past, but I'm not sure that's really a decisive argument.

But, it highlights this issue of whether government policy in terms of interaction between consumers and sellers should be designed around what consumers want.

Should it include whether it's fair to producers? Inequality? Global climate issues? There are many, many things we could add to the conversation at the intersection. Your turn.

Michael Munger: It seems to me that the current sort of--Well, I have said that the Left has rediscovered originalism. So, originalism is usually a conservative doctrine. But the Left now has gone back to the legislative debates in the passage of the Sherman Act.

And, there was all this dog's breakfast of things that the legislators were concerned about. And, they're saying, 'This is clearly what the Sherman Act should still be about. We should bring it back. That was the original intent of the Sherman Act.'

So, I think the Left has correctly identified a problem, but egregiously misidentified the solution to such an extent that it's going to just lead us in the wrong direction.

There is a problem, and I think the problem will more and more, unless we solve it, approximate the sort of hellscape of a universal panopticon.

The solution is to be able to have people--it's a property right solution. It's a standard economic idea. It's just intellectual property. I own the data and the value that I create on these platforms.

And, if somebody wants to buy it, they can go nuts. Because, I would probably sell it for not very much. But I have of control over it, so we can solve the growing panopticon problem by being able to draw the blinds.

And, if you want to look, I'll raise the blind. You get to see the part of me that I'm willing to share.

However, I can wear a disguise. I can wear a unicorn head because the level of identity that I need to reveal need be no greater than that required to give you assurance of my value as a counterparty. You don't get access to all my data. You just get access to the fact that my reputation score is 98.

That means that I am trustworthy--98 in 5,000 transactions. That means that, like any reputation, there's a capital asset that is depreciable. If I ruin my reputation, I will ruin it not only in this setting, but in all the others.

So, everyone will have their own reputation.

So, reputation we usually think is a way of solving this problem for firms. It can be consumers also because everyone will be their own brand.

I am participating in a large-scale experiment where these two-sided markets are matching people on platforms. So, I think platforms are both the problem and ultimately the solution.

Russ Roberts: My guest today has been Mike Munger. I'm really glad he's better after his heart surgery. Thank you, Mike.

Michael Munger: Thank you.

More EconTalk Episodes