On Monday, Leonid Hurwicz, Eric Maskin and Roger Myerson were awarded the 2007 Nobel Prize in Economics for their contributions to mechanism design theory. To the non-economist, mechanism design theory seems to be a highly abstract construct: technically impressive, perhaps, but with little relevance to policy and everyday life. In fact, despite its complexity and mathematical character, mechanism design is deeply embedded in policy.
Leonid Hurwicz started the field in the 1960s when he considered a very policy-oriented problem. How should a planner reach a decision when the quality of the decision relies on information spread among a number of people? The mathematical formulation of this problem is at the heart of mechanism design theory. Among his key insights is the idea that any solution should take into account the incentives of self-interested agents, i.e., the people on whose information the decision relies must find it in their interest to reveal that information.
Hurwicz’ analysis spoke to the great intellectual debate of the time. Was Capitalism or Socialism the better system? His results pointed out to a major weakness of Socialism – the lack of proper individual incentives. His results did not, however, let Capitalism off lightly, because individual incentives are not always aligned with social incentives. It did, however, help governments think about how best to regulate a capitalist economy.
Key contributions from Maskin and Myerson developed the theory further in the 1970s and 1980s. In its current form, mechanism design theory provides a general framework to study any collective decision problem. A mechanism design problem has three key inputs:
· A collective decision problem, such as the allocation of work in a team, the allocation of spectrum for mobile telephony or funding for public schools;
· A measure of quality to evaluate any candidate solution, for example efficiency, profits or social welfare;
· A description of the resources – informational or otherwise – held by the participants.
A mechanism specifies the set of messages that participants can use to transmit information and the decision that will be taken conditional on the messages that are sent. Once a mechanism is in place, participants effectively “play a game” where they send messages (e.g., a bid in an auction) as a function of their information. The goal is to find a mechanism with an equilibrium decision outcome (sometimes required to be unique) that is best according to the given measure of quality. The strength of mechanism design lies in its generality: any procedure, market-based or not, can be evaluated within a unified framework.
How mechanism design is used in economic policy-making
Today, mechanism design theory is part of the standard toolkit of every economist, and every economist uses it – consciously or unconsciously – almost daily. Mechanism design theory has affected virtually all areas of policy. Its policy implications lie at two levels. First, mechanism design theory tells us when markets or market-based institutions are likely to yield desirable outcomes (remember Adam Smith’s Invisible Hand?) and when other institutions will be better at achieving the desired goals. Second, mechanism design theory gives us guidance to design such alternative institutions when markets fail. In the rest of this article, we describe some of the policy areas affected by mechanism design.
Auctions are one of the early applications of mechanism design theory and one area to which both Maskin and Myerson have heavily contributed. The problem here is to decide how to allocate some object(s) among potential bidders when the value of the object(s) to a bidder is only known to that bidder. The objective can be to maximize the revenue raised from the sale, or to ensure that the objects go to those who value them the most (achieving the goal of efficiency). Today, governments use auctions to allocate spectrum, airport slots, oil drilling rights, timber or land. They also use reverse auctions to procure goods and services from the private sector. In each case, mechanism design theory has successfully been applied to give recommendations on how to best set the rules of the auction as a function of the government’s objective and the nature of private information.
Regulation too has been revolutionized by mechanism design. In work with David Baron, Myerson was the first to abandon the conventional assumption that regulators “know everything” and to use mechanism design to derive optimal regulatory schemes insuring the provision of public services at least cost. Other researchers later showed that, if cost realizations are observable, simple schemes like cost-plus contracts (where the firm is paid a fixed reimbursement plus the difference between a cost target and the cost realization) can achieve this objective. Such results have had a profound influence on actual regulatory schemes and on the design of contracts between international institutions and infrastructure providers in developing countries.
Environmental policy has also benefited from the insights of mechanism design theory. The theory tells us that cap-and-trade systems, such as the system put in place by the Kyoto Protocol, are more cost-effective ways to reduce carbon dioxide emissions than command-and-control mechanisms. Mechanism design theory also tells us how to design environmentally sustainable fishing and hunting schemes in natural reserves. A few years ago, the UK government called on Maskin to help design a scheme that would maximize the corporate sector’s commitment to carbon dioxide emissions reduction for a given budget. Mechanism design theory was again put at work to provide a solution.
Mechanism design theory has also had a big effect on our thinking about development problems in poor countries. Today’s focus is on fostering institutions and individual initiative. Traditional solutions to community problems such as access to credit, land sharing arrangements and natural resources management have been revisited and improved in light of mechanism design theory, and new solutions have been proposed. For example, the current success of microfinance is due in large part to a good understanding not only of the needs and objectives of small borrowers but also of the severe incentive problems faced by these borrowers. Mechanism design helps evaluate the relative performance of different microfinance arrangements.
These are just a few examples of mechanism design in action. They show that the reach of mechanism design theory is very broad. The theory represents a major breakthrough in the modern economic analysis of institutions and markets and will have a lasting influence for the design of economic policies.