At the heart of economics, there is a tension between two very different tasks, often leading to altogether different viewpoints. On the one hand, we try to understand the economy as a system, that is, we strive to understand how markets relate to one another, how information flows through markets and other institutions to achieve the apparent miracle of coordination of the myriad activities that occur in the economy. On the other hand, economists try to understand and help with practical, real-world concerns, such as determining the best rules to use for physical or online markets, the best procedures for buying or selling objects, and sensible principles for corporate governance.
The research of Stanford University’s Robert Wilson, who along with his student and Stanford colleague Paul Milgrom won the Nobel Prize in economics earlier this week, is motivated for the most part by the second task—helping with practical concerns, specifically designing and facilitating exchanges. Tellingly, Wilson’s master’s degree and doctorate are in business administration, not economics. His earliest article on auctions, published (in 1967!) in the journal Management Science, finds inspiration in a thesis on bidding for mineral rights via sealed tender. Some of his latest work has been inspired by issues related to transmission and generation in electricity markets. In the span of half a century, Wilson’s work and that of an impressive cadre of former students, including Milgrom, and fellow Nobel laureates Bengt Holmstrom and Al Roth, from MIT and Stanford respectively, has contributed to create market design as a field.
The Nobel announcement singles out Wilson and Milgrom’s contributions to the theory of auctions and to the design of new auction formats. In 1969, Wilson provided the first rigorous account of bidding in an auction with common values. An auction has common values when what a bidder is willing to pay for the object being sold can be affected by how much other bidders are ready to pay. Suppose that the bidders are oil companies and what is at stake is an offshore parcel. Oil companies may have different estimates of the worth of the parcel because they have access to different sources of information. For instance, they may own adjacent parcels. If a company knows that the others value the parcel highly, it may revise what it is willing to pay upwards. Likewise, if the company knows that others do not value the parcel, it may revise what it is willing to pay downwards.
The so-called “winner’s curse” occurs if a bidder wins the parcel at auction but comes to regret it after discovering that others did not value it as much. Rational bidders will avoid this fate by shading their bids downwards. Subsequent work by Milgrom and by Milgrom and Robert Weber, from Northwestern University, showed that an ascending auction allows for the gradual release of information, thereby eliminating the need for bidders to be overly cautious, and generating as a consequence larger expected revenue for the seller.
Milgrom and Wilson are also pioneers in the design of auctions of multiple objects with interconnected values. This research line was spurred by the decision to auction the radio frequencies in the 1990s, switching away from an earlier reliance on less competitive methods for the assignment of frequencies. The format adopted, as a consequence of the research of Milgrom and Wilson and others, used multiple rounds of auctions. Multiple rounds allow bidders to gradually discover the best frequency package for them given others’ willingness to pay.
Using the auction format to find an efficient allocation of frequency packages raised expected revenue, since the spectrum ended up in the hands of those who value it the most and were willing to pay for it. This benefited taxpayers but also consumers because those who value the spectrum the most also generally are better at delivering content to consumers.
One could argue that getting the initial allocation of the spectrum right the first time is not very important because companies can trade among themselves after the auction. However, trading in such a decentralized way with interconnected values is complicated. The initial auction, if properly designed, may help reduce the costs of bargaining between spectrum holders.
More recently, Milgrom and others have advised the Federal Communications Commission in the repackaging of the radio spectrum from broadcast television to wireless broadband services, via a reverse auction to buy back the remaining broadcaster spectrum and a forward auction to allocate the freed-up space to wireless providers.
Real-world applications of market design like those described above follow certain patterns. First, technological change opens up possibilities for a successful exchange. For instance, in the case of electricity markets, because of technological change, at some point economies of scale in generation diminished enough to make markets for generation a feasible alternative to regulation.
Competition in these imperfect circumstances, however, can be helped by the “architecture of the market,” to borrow a phrase from Wilson. Economic research helps in imagining alternative designs and anticipating their consequences in the classroom, but also, in learning and experimenting pragmatically and in adapting market institutions to different places and circumstances. In a way, economic research that motivates and helps along in the expansion of the role of markets is also part of the technological change.
I began by mentioning that the research of Milgrom and Wilson has largely focused on practical concerns rather than the loftier view of the economy as a system. Yet their successful and enormously influential work also has implications for the way we understand the economy as a whole. Markets are always a work in progress; they differ in lesser or greater extent from the textbook parable of perfect competition. They are being designed by economic agents, most often in a decentralized way, and sometimes requiring some explicit coordination to set up the rules for exchange. In a liberal society, that coordination can be carried out by the institutions we set up to govern ourselves, informed by research in economics, management, etc. that is itself the result of a marketplace of ideas.
There is, in the current intellectual climate, some disenchantment with market institutions. By looking at cases in which competition falls far from textbook examples, like bidding with common values or like competition in market supply schedules in electrical markets, the work that was awarded this year’s economics Nobel reminds us that the scope and rules of exchange can accommodate the challenges posed by technological innovation. Liberal society not only fosters technological innovation, it fosters the intellectual innovation to promote exchange under evolving opportunities.