Jean Tirole and Platform Markets

Today’s Nobel prize winner in economics, Jean Tirole (working with Rochet) is a pioneer in one of the most important new areas in the economy and economics, the study of platform markets. Platform markets, also called two-sided markets, are markets where a firm brings together two or more sides both of whom benefit by the existence of the platform and both of whom may (or may not) be charged. A trivial but telling example is the singles bar that brings together men and (usually) women. Other examples are the Xbox, a platform for game players and game developers, credit cards a platform for buyers and firms that accept that card, newspapers a platform for readers and advertisers, and malls a platform for customers and stores to meet. An important example for the internet age is that Google is a platform of search users and advertisers.

An key difficulty in these markets is that the price charged to one side of the market influences the demand on the other side of the market. The price a newspaper charges to readers, for example, influences the number of readers but that in turn influences the price that the advertisers, the other side of the market, are willing to pay to advertise in the newspaper. It further often happens that one side of the market is harder to “get” than the other and so the profit-maximizing prices on the two sides of the market are very different. One side of the market may even be “subsidized.” The price that newspapers charge readers, for example, is often much less than the cost of the newspaper. Or, to give another example, Microsoft makes money by selling its Xbox at close to cost or even below cost and charging game developers a fee for the right to write games for the Xbox and a royalty rate on their sales.¬†Google finds it optimal to give its services away for free and just charge one side, the advertisers, for being on the platform.

Antitrust and regulation of two-sided markets is challenging because the two sets of prices may look discriminatory or unfair even when they are welfare enhancing. In a mall, for example, it’s often the largest firm (the anchor) that gets the lowest price (sometimes even zero!). Does this represent an unfair advantage that a large firm has over smaller rivals or is it a rational consequence of the fact that the anchor store may bring the most customers to the other, smaller stores in the mall so that the total package is welfare maximizing? Is Microsoft engaging in predatory pricing if it prices the Xbox at or below cost? A singles bar may have “ladies are free night”. Sexist? or good economics? Platform markets mean that pricing at marginal cost can no longer be considered optimal in every market and pricing above marginal cost can no longer be considered as an indication of monopoly power. The analysis also impacts such issues as network neutrality. People worry, for example, that firms like Netflix may be the anchor stores of the internet and get better prices as a result. But the analysis of platform markets suggests that this isn’t necessarily welfare reducing. As these examples indicate is easy to go wrong regulation these markets and in fact Rochet and Tirole urge caution in regulating platform markets.

Rochet and Tirole provide one of earliest and most important analyses of pricing in these types of markets (see also here for an overview).

See Tyler’s post below for much more on Tirole.


Intuitively this ladies-night style discrimination doesn't worry me. The mutual benefits reaped that a mall and its biggest store seem (again intuitively only) like just rewards for success.

The bigger problem I see with platform markets is the network effect where customers/suppliers are attracted to a platform by each-other and not by the quality of price of the platform. We can have interesting arguments about whether Windows and Facebook are good at what they do; but it is hard to argue that they are much disciplined by the ordinary laws of competition.

'Network effects' - that point needs to be emphasized in any such discussions.

That is very chicken and the egg. How does a network get developed in the first place? crappy companies don't just start out with large networks, and companies with large networks like Microsoft often fail to dominate related markets (Bing, IE).

Somewhat weak companies can expand even by luck when network-effects are at work because once they have had their lucky break, they have a tailwind. That (sort of) happened for Microsoft, MS was probably very good at what it did back in the 80's, but what it did then was produce piece-of-sh** operating systems.

And even if a company needed to be good to grow, it needn't stay that way after it achieved a monopoly. The good news in operating systems is that Microsoft didn't do that. They dragged their feet in producing a decent OS, but over time they pulled it off; and they don't seem to be in a position to abuse pricing power.

I pay nothing for GMail, and I'm fine with that.

I assume the american tax payer pays, through the NSA budget.

Google isn't free. It is buying your valuable personal information in exchange for your "free" use of its products. Or to be more cynical you are not the consumer but rather the product that Google sells to advertisers.

In exchange for access to near limitless information outside your personal sphere. It actually is a rather fair trade in an objective sense. They have all your information and you gain all their information. It's just people are understandably more attached to "their" information.

As you note, the two-sided market that has been in the news a lot recently is the one between end-user facing ISPs and other networks, including Content Delivery Networks (CDNs.) In general bigger networks have long charged smaller networks for access, and similar networks have allowed free peering, based on all sorts of reasons and negotiations. However, a lot of people have an intuition that it's unfair for ISPs to collect money from both sides of the market, both CDNs and users, especially when there seems to be a monopoly with the users.

Yet some research implies that even (or especially) in the case of a monopoly, it is not beneficial for the end users to be the ones to pay all the costs while the other side pays nothing. In a monopoly, the total rent the monopolist extracts may be the same regardless of who directly pays, and it's possible that having the larger networks pay results in negotiating better prices.

AlexT's antitrust arguments were made by AT&T and its opponents back in the days of Ma Bell, as outlined in the book by an econ historian: Peter Temin "The Fall of the Bell System". Remember a (natural or almost natural) monopoly usually has marginal cost less than average cost for all or most of its curve, going from the left to right. So AT&T would argue before the FCC that the marginal cost should be used when pricing stuff (to keep out new entrants like MCI) while MCI and the anti-Bells would argue the opposite. However, during the rising cost portion, MC > average cost, and the arguments by each party are reversed. Finally, the commissioner of the FCC that voted in favor of MCI by a slim margin later got rich and/or worked for MCI, see here: (and this is what Temin says: "One of these commissioners, Kenneth Cox, participated in the panel discussion between Baxter and Trienens reported in the Introduction. He found his views so close to McGowan's that he left the Commission to work for MCI shortly after the vote. There is no evidence that Cox and McGowan discussed employment before the FCC's vote, but Cox could not have been unaware that his views made him an attractive potential officer of MCI.49 (AT&T, of course, never hired former regulators.) With only a hair less support, McGowan might never have gotten his fledgling enterprise off the ground. )

The old DC Beltway / K-street revolving door.

Also, competition is not only on price, but also variety and innovation. Ma Bell did all it could to prevent equipment not its own from connecting to the network.

Congratulations on taking the "Marginal Revolution's stupidest commentoer" title from mulp with the longivity thread, Ray.

@Careless--so careless of you, you misspelled 'commentator' in your one-liner re-buttal. I bet you are red-faced over that one. Also remember who you are talking to: a member of the 1%, an intellectual, a man twice your years with a girlfriend half your age. Dismissed.

I mean no insult, but I can always tell without looking whether it's Alex or Tyler writing the posts because the former makes a lot more typos. There are about a dozen missing commas and hyphens in this post, and few spelling errors. You can do better, Alex!

I disagree. Anything more than running a post once through spellcheck means an opportunity cost that would well exceed its value. This debate was settled in some long ago Yglesias thread.

Cowen has no opportunity cost since he is able to read and write an infinite amount in an arbitrarily-small unit of time, so he is a bad example.

I'm thinking about stock exchanges as platforms. Can we say that they are analogously two-sided, because liquidity-takers need liquidity makers? in much the same way that game developers need to have a lot of people playing the game? If so, then the natural result would seem to be the subsidization of the makers BY the takers.

Both systems exist of course: maker pays and taker pays. Does Tirole give us any light on why a particular exchange will try one rather than the other?

Fantastic goods from you, man. I have understand your stuff
previous to and you are just too fantastic.
I actually like what you've acquired here, certainly like
what you're saying and the way in which you say it.
You make it entertaining and you still care for to keep it sensible.
I cant wait to read far more from you. This is actually a great web site.

Comments for this post are closed