Skip Main Navigation
Page Content
This event has ended

Save This Event

Event Saved

CPI Audio Interviews with MIT Professor Schmalensee

Competition Policy International + CCIA

Friday, November 9, 2018 from 8:30 AM to 6:30 PM (EST)

CPI Audio Interviews with MIT Professor Schmalensee

Ticket Information

Type End
Standard Registration   more info Ended Free  

Share CPI Audio Interviews with MIT Professor Schmalensee

Event Details




Professor of Management and Economics 

Dean Emeritus of the MIT Sloan School of Management


Ahead of the upcoming CPI & CCIA Conference at Harvard Law School on November 9th, Challenges to Antitrust in a Changing Economy, CPI’s Managing Editor Sam Sadden has interviewed one of the key-note speakers Richard Schmalensee.


In this audio interview,  MIT Professor Schmalensee anticipates some key points that he plans to cover during his speech. Specifically, the five red herrings that have been floated in the debate leading up to and following the recent Supreme Court’s decision in American Express.


This conference is co-organized by CPI and CCIA.

See the full program of the conference and register free here



Listen to the full interview here


Interview Transcript 


Sadden: Ahead of the upcoming Competition Policy International Conference at Harvard University on November 9th, Challenges to Antitrust in a Changing Economy, we wanted to take the opportunity to talk with one of our keynote speakers about the topics he plans to cover. Today, we have the great pleasure to speak with MIT Professor Dick Schmalensee.

Professor Schmalensee, thank you very much for taking time to sit down with us today at CPI. You recently co-authored an article, along with David Evans, in the CPI Antitrust Chronicle, entitled Two-Sided Red Herrings. You refer to the debate leading up to and following the Supreme Court's recent decision in American Express, and you raised and then attempt to debunk some red herrings that have been floated around in these debates.

First off, could you please add a little context to our non-US residents? What's so important about the Amex case and, more generally, two-sided and multi-sided platforms?


Schmalensee: Well, two-side and multi-sided platforms are an old business model. Think newspapers. Newspapers exist to connect advertisers to readers. Credit cards, of course, are pretty old. They act to connect merchants to customers, the shoppers. Generally, what a two-sided or multi-sided platform does is it facilitates interactions between members of two distinct groups, two or more distinct groups. Those groups value each other, and as a result, their demands for the services of the platform are interdependent. Merchants care about how many consumers use a particular card. Consumers care about how many merchants accept it.

This isn't the first time that US courts have dealt with multi-sided platforms. There are newspaper cases going back decades, but it's the first time that the Supreme Court talked about multi-sided platforms. People from outside the US will recognize that multi-sided platforms, online travel agents, travel recommendations services, have attracted a lot of attention outside the US, and these are, of course, multi-sided platforms that connect hotels or airlines to travelers, and at least one of their business practices, the so-called Most Favored Consumer clause, has attracted a lot of attention outside the US, a little less attention here.


Sadden: All right. Moving on to a couple of the red herrings that you bring up in your recent paper and that you will be covering in your keynote speech at the upcoming conference, CPI Conference on November 9th, the first red herring that you discuss is the idea that the two sides are simply complements, and so to the idea of the case of tennis balls and tennis rackets, what did Justice Breyer put it in his Amex dissent? Was it gasoline and tires?


Schmalensee: Gasoline and tires. He talked about screws and nuts in the oral argument and then went to gasoline and tires in the dissent.


Sadden: Why are these questions even being raised, and what are people missing here on the idea of complements?


Schmalensee: Well, this is strange. I have to say, when I first read it, the literature on multi-sided platforms is not 100 years old, but it's not that young. It's been around among economists since the turn of the century, and it's been in leading journals since about 2003, so 15 years.

David Evans and I, in a paper some time ago, managed to find several hundred papers in the literature that dealt with this basic business model, and nowhere in all of that writing, in any major journal, does anybody say, “Oh, no, this is nothing new. This is just complements.” When Jean Tirole received his Nobel Prize and the citation mentioned his work on two-sided platforms, his seminal work on two-sided platforms, nobody said, “Oh, but those are just complements,” so hearing it in an antitrust context is just very strange.

Look, products which are complements are typically sold to the same consumers, gasoline and tires, tennis rackets and tennis balls, the demand, coffee and cream to go way back, the demands are interdependent in the sense that the price of one will affect the demand for the other, but it's the same customer. The reason is a customer cares about both coffee and cream or tennis rackets and tennis balls, and so the price of one affects that customer's demand for the other.

Two-sided platforms deal with different customers. It's not one customer who cares about prices of two different things. It's two different customer groups that care about price and who's on the other side of the business. If you just think about the two-sided platforms, they're just completely different from tennis rackets and tennis balls.

The other absolutely clear difference is that a lot of people sell gasoline and don't sell tires, but for American Express just to come to the case, it must deal with both merchants and shoppers. If it doesn't have both groups connected, it doesn't have a business. It doesn't have the option that people selling gasoline and tires or tennis rackets and tennis balls have of just selling one. Multi-sided platforms need both or all, if more than two, all groups to engage, or they don't have a business. 


Sadden: I think in the same sort of vein as that is another red herring that you bring up in the paper, is the argument that markets must be one-sided since the services to the two sides aren't interchangeable. What is the overriding issue with this interchangeability argument?


Schmalensee: Well, as the people who made it in the Amex case point out, this isn't where you normally approach thinking about market boundaries. You say, “Well, what's reasonably interchangeable with a tennis racket? Is a squash racket reasonably interchangeable?” No. Then you look at the supply side, and you say, "Well, people who make squash rackets, maybe they can switch into making tennis rackets," and you have this question, but it doesn't make any sense in the case of two-sided platforms. It is true that what credit cards do for merchants is different than for what credit cards do for customers, but, look, it's also true that automobile tires and automobile engines are not reasonably interchangeable, but nobody says, “It makes no sense to think about a market for new cars.”

When objects or services or, for that matter, goods are consumed in fixed proportions, inevitably, then it makes sense to consider the aggregate, and, here, what David Evans and I argued, and a number of other people have argued, is in the case of credit cards, the market is linking merchants and customers. It is processing the transactions. That's how people in the business track market share. That's how all the services that deal with this industry track activity. They don't look at one side or look at the other side. They look at the activity of linking, and if you want to do market share, you can look at transactions volume, which is what is typically done.

It is really strange. As I say, there's nothing, nothing in the economic literature that says, “You can’t have a market that is linking two different groups,” just like there’s nobody in the economics literature that says, “You can't have a market for your new automobiles because the parts in the new automobile aren't reasonably interchangeable.”

Competition in credit cards, competition in person-to-person payment systems, occur at the system level, and so that's the market. It’s competing systems. The fact that they do different things is no more relevant than the fact that automobile engines and tires are not reasonably interchangeable.

Again, what’s strange about this is, after all these years in the literature and after hundreds of papers, this argument comes out of left field, and, again, in the context of automobiles, you say, “Well, this is ... There's nothing new here.” Automobiles consist of tires and engines and fixed proportions, and so you buy a car. You don't say, “Oh, no, we can only think about tires because ...” Sometimes, it makes sense to think about tires, but, sometimes, it makes sense to think about the whole thing.


Sadden: I think what's interesting here is we can also then tie this into things that are happening now. With keeping these red herrings that you brought up in mind, what does this mean for potentially for Europe and some of the new antitrust probes that are going on there related to Google, potentially Amazon, and the like?


Schmalensee: Well, one of the things that’s interesting, I find, is that you don't see these red herrings much outside the United States. There seems to be a much, maybe because Jean Tirole is French, but there seems to be a much greater acceptance that the economics literature is not crazy outside the US. In a sense, there's nothing revolutionary here in the Amex case in terms of the majority's focus on a market that really has to do with the system, has to do with both sides, but an awful lot of these platforms originated in the US, so it's funny that the economic analysis of them is being sort of trashed in the antitrust context.

I don't know that the Amex case has much by way of analytical implications for Europe. What I think it means in the US, perhaps because it's not as clear as many people would like it to be, (a) we can't ignore ... Courts and lawyers can't ignore two-sidedness when it is present and important; but, (b) it will be a controversial subject until some of the issues related to it get sorted out.


Sadden: Nor can they probably try and create the image of their being two-sidedness when it's not there, either.


Schmalensee: I think the majority's definition in the Amex case was sort of over-broad. It suggested that, well, if you serve two different groups, then maybe you're two-sided, but any business has to deal with its suppliers and has to deal with its customers, so the notion that every business is really two-sided, and so we've mostly ignored this fact for hundreds of years, let's just continue to ignore it, that's another red herring. It suggests that whole literature was about nothing. It suggests that there's no difference between American Express and the merchants it serves, that the fact that it basically sets prices to two very distinct groups and it's in the business of linking them, is just like the supermarket down the street that buys groceries from its wholesaler, posts the price and sells them to me.

Well, if you look at those businesses, they don't feel the same. The grocery store down the street can get supplies by calling up its wholesaler, and the wholesaler may give it a better deal if it's bigger, so you can sing a song about how well that's two-sided because the wholesaler cares about how many people shop at this store. Yeah, maybe, but what doesn't happen is the grocery store doesn't facilitate an interaction, a transaction, between the wholesaler and the customer, and it doesn't usually have to attract the wholesaler. It places an order, and assuming its credit is good, it gets the goods it needs to resell.

There's no business of attracting customers and wholesalers, whereas Amex, any payment system, really does have to attract merchants and does have to attract people to carry and use the card.

I think we may see attempts made at that. I don't think it's happened in Europe or outside the US, that people have tried to claim two-sidedness for some purpose when it clearly isn't relevant. I expect it'll happen in the wake of the Amex decision in the US. I also expect the courts will figure it out in short order and clarify. I'm not too concerned. Some commentators have said, "Well, the Amex case guts antitrust law." It doesn't. It just doesn't. The notion that you should, which the dissent seem to advocate, that you always should only look at one side of the market, the side of the market where the complaint was made. Then, all newspapers are predatory because newspapers are sold below cost, and it's impossible to predate by charging low advertising rates in newspapers even though, by doing that, you can become unprofitable as a paper and maybe drive somebody, some other newspaper, out of business.

It's going to take a while. I hope not too long a while.


Sadden: I think newspapers have enough problems on their hands right now as it is.


Schmalensee: There was a great newspaper predation case in 1953 with the Supreme Court. I guess 1950 may have been the district court, where the district court did just that. It said, "Well, if the newspaper's profitable, you can't be doing predation, even if you're selling the papers below cost. Thank you very much." Well, that's a two-sided analysis done in the early '50s. We can do it again.


Sadden: Definitely. All right. Well, thank you so much, Professor Schmalensee, for taking time to speak with us today. I would love to keep talking, but we'll save it for your keynote speech coming up on November 9th and for the other things down the road, as well. We look forward to seeing you in Cambridge on the 9th at the CPI Conference, Challenges to Antitrust in a Changing Economy, and we also look forward to seeing many of our listeners and subscribers, as well. Thank you, again. We really appreciate it.


Schmalensee: Thank you very much. I look forward to seeing everyone on the 9th.


Sadden: Thanks.

Have questions about CPI Audio Interviews with MIT Professor Schmalensee? Contact Competition Policy International + CCIA

When & Where

Harvard Law School
1585 Massachusetts Avenue
Cambridge, MA 02138

Friday, November 9, 2018 from 8:30 AM to 6:30 PM (EST)

  Add to my calendar


Competition Policy International + CCIA

This year's 2nd Annual conference, co-organized by CPI and CCIA, aims to pick up where we left off last year with the debate on competition law and economics in the tech industry. Doubtless, the last decade has seen a growing thirst for innovation in many industries worldwide. Innovation certainly makes economies more dynamic and competitive, but it can pose challenges for legislative and regulatory bodies trying to keep pace with rapidly evolving businesses. This conference will address key issues affecting the tech industry in this climate of constant changes. Competition, Concentration & Common Ownership, Competition in Digital Advertising, The Economics Behind Digital Services, FTC Hearings and other Ongoing Consultations, are some of the topics that will be discussed by leading antitrust academics, enforcers, and private practitioners.

Registration includes continental breakfast, coffee, lunch, and a closing reception.

CPI is a leading platform that promotes antitrust debates via publications and live events worldwide. CPI releases daily newsletters, bi-monthly Antitrust Chronicles,annual special edition Chronicles, and publishes antitrust books. CPI also organizes roundtables and conferences globally.

CCIA is an international not-for-profit organization dedicated to innovation. It promotes open markets, systems, networks and full, fair and open competition in the computer, telecommunications and Internet industries.

  Contact the Organizer

Interested in hosting your own event?

Join millions of people on Eventbrite.

Please log in or sign up

In order to purchase these tickets in installments, you'll need an Eventbrite account. Log in or sign up for a free account to continue.