/cdn.vox-cdn.com/uploads/chorus_image/image/41751110/Tirole_jean.0.0.jpeg)
Jean Tirole, the French economist who won the Nobel Prize today, is known for the breadth and scope of his work across a number of fields. The Royal Swedish Academy specifically cited his work on competition and anti-trust issues as prizeworthy.
But I think many people will be most interested in his 2002 paper "Platform Competition in Two-Sided Markets," co-authored with Jean-Charles Rochet. Among other things, the paper offers a powerful explanation of why so many leading internet companies — most prominently Google and Facebook — don't charge for their products.
In the most simplified thinking about business, a company has suppliers and then it has customers. But a platform market is two-sided. Apple sells iPhones to customers, but it also collects 30 percent of the gross sale price of apps in the app store.
(Ggparker)
In a naive pricing scheme, the platform owner simply charges both sides of the market the profit-maximizing price. But this is much too simple. The number of people who own iPhones is a factor that developers need to consider when contemplating how interested they are in making iPhone apps. And the number of apps available is a factor customers need to consider when contemplating how interested they are in buying iPhones.
(ggparker)
If giving up revenue on one side of the market drastically shifts the demand curve on the other side of the market, then it can make sense to discount.
The correct pricing calculation can be very difficult to make, and one of the points of the paper is simply to underscore the great diversity of strategies that firms in two-sided markets undertake. But the Internet offers an example of a particular case of a two-sided market, namely one in which the marginal cost of serving an additional customer on one side of the market is extremely low.
In other words, while it costs a lot of money to run Facebook it costs very little money to serve one more Facebook customer. The same is true for Google. Indexing the web is expensive. Paying engineers to work on the search algorithm is expensive. But serving one additional customer costs basically nothing.
Consequently, the dominant strategy on the web has proven to be charging $0 to the customers on one side of the equation in order to maximize the size of your user base. Then you make your money on the second side of the market. These days you often hear people say things like "if you're not paying for the product, you are the product."
There is something to this, but the Tirole-Rochet paper shows that this same dynamic exists in a variety of industries. Any time the market is two-sided you are both the customer and the product simultaneously. That's true whether the platform owner is charging you or not. What makes the ad-supported web services special is that it's economically feasible for them to offer a product that's free to one side of the market, in a way that isn't true for companies who are dealing with higher variable costs. What is true is that because of the interactions between the two sides of the market, these platform industries can lend themselves to total domination by a single firm.
And an important question for the world going forward is to how to think about regulating that kind of market. Since Google doesn't charge its mass customer base anything at all, you're never going to find evidence of monopoly power in price hikes.
Instead, Tirole said at today's brief press event that regulators need to be watchful that the incumbent isn't erecting barriers to entry that might prevent its replacement by a new more dynamic firm. This is the part where it would be ideal to present a one-paragraph account of how you do that, but as Tyler Cowen writes "many of [Tirole's] papers show 'it's complicated,' rather than presenting easily summarizable, intuitive solutions which make for good blog posts." So you're going to have to accept a mediocre blog post.
One takeaway from Tirole's work on regulation in general is that to a greater extent than people appreciate, policy problems sometimes exist because the questions are genuinely difficult rather than because policymakers are corrupt or feckless.