IN 1991 ALVIN ROTH, who in 2012 would share the Nobel prize in economics, was asked to speculate about how the discipline might change over the century to come. “In the long term,” he wrote, “the real test of our success will be not merely how well we understand the general principles which govern economic interactions, but how well we can bring this knowledge to bear on practical questions of microeconomic engineering.” Mr Roth did his bit to move the field in that direction: among other things, he helped design market mechanisms to match sick patients with kidney donors. The Swedish Riksbank, which awards what is officially known as the Prize in Economic Sciences in Memory of Alfred Nobel, seems to agree. On October 12th it gave this year’s prize to Paul Milgrom and Robert Wilson, both of Stanford University, for their work on auction theory and design. Their work epitomises economics as engineering.
Auctions are an ancient mechanism for selling valuable commodities, from fine art to a fisherman’s catch to government bonds. Historically, a few forms of auction have been dominant. In an English auction, ascending bids are made until a sole winner remains; in the Dutch variety, a high opening price is set and is slowly reduced until a bidder is found. In the 1960s William Vickrey, who shared the Nobel in 1996, developed what became known as auction theory. He applied game theory to assess bidders’ optimal strategies and studied the revenue and efficiency properties of different auction formats. But Vickrey concentrated on a relatively narrow set of cases, in which each bidder’s valuation of the good being sold is unrelated to those of all the other bidders. In practice, however, what one person believes an auctioned item to be worth often depends on the valuations of other bidders or the seller, because they may have access to private information about its value, clues to which are revealed in the course of the auction.
In the 1960s and 1970s, Mr Wilson began working out how to analyse these other scenarios. He first tackled cases in which the item for sale has some uncertain common value—for example, a plot of land with oil beneath it. Different participants may have different information about its precise value (eg, the quantity of oil). In such cases, the winning bidder often discovers that the information others had about the common value led them to place a lower value on the auctioned item. This may mean that the winning bidder overestimated its value and paid too much, a phenomenon known as the winner’s curse.
Mr Wilson’s work in this vein laid the groundwork for analysis of still more complicated scenarios, in which bidders’ unique private valuations and an item’s common value are both taken into consideration. The value of an oilfield, for instance, might depend on both the quantity of oil in the ground and on the cost-effectiveness of different bidders’ extraction technologies. Mr Milgrom (whose doctoral thesis was supervised by Mr Wilson) took the lead in studying these sorts of cases, and derived a number of important lessons from his analyses. Auction structures that elicit more private information from bidders (such as English auctions, where every participant can observe who bids what and who drops out) reduce the winner’s curse problem compared with formats where very little private information is divulged. In some cases, it may also be to the seller’s advantage to provide more information to bidders about the item to be sold.
Much like Mr Roth, Messrs Milgrom and Wilson eventually put the knowledge gained from their theoretical work to practical use. Before the early 1990s, America’s government allocated portions of radio spectrum through unwieldy means: either through what economists call a “beauty contest”, in which bidders essentially explained why they deserved a slice of spectrum more than others, or by lottery. In 1993 Congress permitted the Federal Communications Commission to allocate spectrum using an auction instead. Yet it was not clear how an auction might work, given the complexities involved. Different bidders from different industries had wildly varying assessments of how slices of spectrum might be used, and the value of one piece of spectrum often depended critically on what other slices an owner also controlled. The laureates worked with another economist, Preston McAfee, now at Google, to invent an entirely new format, known as the “simultaneous multiple round auction (SMRA)”. When first used in 1994, this format raised $617m for an American government that had previously earned almost nothing from its distributions of spectrum rights.
SMRA-style auctions are now used routinely in many countries and in contexts other than spectrum sales—in distributing natural resource rights, for instance, or selling electricity. Practical problems like that of spectrum distribution have also continued to motivate the laureates’ research and led to the development of still other specialised auction formats. Mr Milgrom and Mr Wilson became the embodiment of the economist as engineer: applying theory in order to devise functional solutions to such problems.
As a result the researchers are more enmeshed in real-world problems than the typical prizewinner. Mr Milgrom advised Time Warner and Comcast, telecommunications firms, on their participation in radio-spectrum auctions in 2006; the work helped save his clients more than $1bn. In 2009 he co-founded a firm, Auctionomics, that provides consulting services to those looking to operate and to bid in auctions.
It is a different sort of work from that which many aspiring academics imagine themselves pursuing. But if that is not inducement enough for young scholars to think about the returns to also donning an engineer’s cap, the Riksbank’s admiration might well be. This year’s award is the third since 2007 to honour “mechanism design”, or the use of economic principles to design markets in order to solve real-world problems.
A Global Asset Management Seoul Korea Magazine