Who is Jean Tirole?

To be the sole recipient of a Nobel Prize in Economics is a rare honour. But Jean Tirole’s work covers an exceptional range—from natural monopolies to open source software.

The Nobel Prize in Economics 2014, more formally known as Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, has been awarded to Jean Tirole. In awarding the prize the committee has recognized how his ‘rigour has facilitated realism’.

Indeed, Tirole is unique among the Nobel laureates in the sense that despite his formidable technical expertise, his research work is very much focused on ‘relevant’ issues, of the kind that dominate policy discussions. Crystallizing essential features of different market-structures in crisp models and drawing non-obvious policy inferences therefrom is his research style.

In what follows I discuss some of his works cited by the committee, but do bear in mind that Tirole is a prolific and eclectic scholar, and a summary like this can at best give a flavour of his work and be only indicative.

Regulation of Natural Monopolies

Regular monopolies are defined as industries with high initial cost of production and low marginal cost. Put differently, average cost keeps declining with scale of production. Examples include railways, gas and water utilities, telephone networks etc. Regulation of such industries is challenging as with ‘marginal cost’ pricing—something competition naturally converges to— firms will not even breakeven. But that implies that such a cost structure inevitably leads to a monopolistic outcome and letting firms decide their own price may result in exorbitant prices, lower level of production and super-normal profits.

Typically, regulators try to impose ‘cost plus’ pricing in such situations. Tirole’s major contribution has been an analysis of the informational requirements of such practices. In particular, he examined a typical scenario.

A regulator is trying to compensate a firm for undertaking a project (or trying to procure a single good, such details are irrelevant to the point). The firm has an internal estimate of the fixed cost, this is private information. There are four stages in the process.

1. The regulator can ask the firm to report its estimate, though the firm’s report may not be truthful.
2. Based on the reported estimate, the regulator signs a contract with the firm whereby the regulator commits to paying a lump-sum amount and also bear a fraction of cost overruns (ie, excess of actual cost over reported cost)
3. Armed with the contract, the firm implements the project. More crucially, in this stage, it can take a number of measuresto reduce costs. These efforts are costly for the firm to undertake and again the regulator has no means of  verifying them
4. Finally the project is completed, the regulator can verify the actual cost, compute cost overrun and contract is implemented, payment is made and the game terminates.

The question Tirole asked is what kind of contract should be signed in Stage 2. Suppose the regulator tries to entice the firm into undertaking maximum cost-reducing effort in Stage 3, then in compensating the firm it should give no weightage to cost overruns at all.

Then, the costs resulting from not undertaking preventive measures are fully internalized by the firm. But payment will be made solely on the basis of reported cost and the firm develops an incentive to submit exaggerated cost estimates in the first stage of the game. Now assume the polar opposite situation where all the weight is placed on actual realized cost and none on reported cost. A different problem arises now: the firm has no incentive to exaggerate its costestimate, but then it has no incentive to undertake costreducing measure in stage 3 too! After all costs will be fully reimbursed by the regulator anyway!

Tirole’s solution is not to specify a single contract but to offer a ‘menu of contracts’ in Stage 2. Specifically, such contracts have following form: “If you submit the lowest possible estimate, then the entire estimate will be paid in a lump-sum form.  As your reported estimate increases, only a fraction of the estimated cost will be paid in lump-sum form and the rest of the compensation will be in form of fraction of cost-overrun.”

Tirole shows that such contracts can be designed to have a ‘self-selection’ property: firms will report their estimated costs truthfully in Stage 1. Moreover, this is the bestpossible contract that regulators can hope to implement with given informational constraints.

Dynamic Industrial Organization

Industrial Organization is a subfield of economics that investigates market structures with game-theoretical tools. Though the field originated with the brilliant 19^thcentury French economist Augustin Cournot, it came of age following the gametheory revolution in microeconomics. Industrial Organization analysis, based on the Nash Equilibrium, suffers from problems of plenty in dynamic, multi-stage settings though. In dynamic games, there are just way too much equilibria.

Economists have developed various ‘refinements’ to weed-out uninteresting equilibria that are ‘fragile’ in some sense. Tirole, together with 2007 Nobel laureate Eric Maskin, proposed a refinement called Markov Perfection. In it, strategies are required to be stochastic processes satisfying the Markov Property—that is, it is memory less: the future probability distribution depends only on the present state, not what preceded it. This idea has been used in studying patent races, dynamic formation of collusion, entry deterrence etc.

Two-sided markets and Open Source Software

Many businesses act as platforms that bring together two different sets of users. An obvious example is a newspaper that brings together readers and advertisers. Other examples include operating systems (which bring together users and application developers), debit/credit card companies (merchants and users), web browsers (users and web-servers), portals/TV channels (eyeballs/viewers to advertisers), and shopping malls (consumers and shops).

Revenue models in these industries vary considerably though. Video game developers for instance make money off the gameproducers (by imposing a fixed cost in the form of development kits and royalties), By contrast, Windows had chosen an opposite revenue models. Why these differences? Tirole analyzed this empirically important though neglected market structure. Tirole showed that such markets have a natural ‘multiplicative demand’ structure. Given this particular structure, profit-maximizing division of prices depends on relative elasticity in both sides and most of the idiosyncratic features can be explained.

One particular work of Tirole that I find particularly interesting is Open Source Software. From an economic point of view, OSS presents a puzzle: after all, time is a resource and a precious one, it has alternative uses and thus has opportunity cost.

Why do so many programmers invest their time in various projects and then release their products free of cost? One set of explanation has been offered in form of non-economic motivations like altruism. But that raises the question why programmers are more inclined to be altruistic than say a typical dentist? Tirole, based on the casestudy of a number of well-known open source projects—Apache, Sendmail, Linux and Perl—offered an alternative explanation based on signaling theory.

Programmers vary in their ability to execute large projects, and this ability is hidden and costly to verify. By participating in successful completion of popular open source projects, elite programmers signal their ability. Though OSS projects may be costly to implement, their successful completion brings enormous prestige and  ‘star’ status to programmers (think of Linus Torvalds, Larry Wall or Guido van Rossum), which in turn has enormous impact on their career trajectory. This also explains a number of norms and conventions in the Open Source community, for example a fierce insistence on scrupulous attribution (because a signal must be visible). Personally, I believe there is more to the OSS story than mere career considerations, but Tirole’s theory already explains a lot.

Given the pragmatic and prescriptive orientation of much of Tirole’s work and also his sheer range, it is tempting to view his achievements as a cookbook of unrelated recipes. However, this impression is totally unwarranted. As the Nobel Committee noted, the total of his work exceeds the sum of its parts.

The Sveriges Bank Prize is not awarded in economics per se, but in Economic Sciences. What is science is debatable, but at least in part it has something to do with reductionism. One takes numerous complex phenomena and explains them in terms of something more fundamental and less complex.

Hence the clichés that all biology is chemistry, all chemistry is quantum mechanics and so on. Tirole’s work shows that a whole range of markets and institutional diversity emerges from, and in turn can be explained by, a handful of primitives: taste, technology and most crucially informational asymmetry. Given his body of work, a Nobel Prize, and a still rarer honour of being a sole recipient, is richly deserved.