The 2020 Nobel Prize in Economics goes to Paul Milgrom and Robert Wilson for auction theory and the improvement of auction designs. The Nobel Committee has a popular introduction and good scientific overview of auction theory. Billions of dollars of spectrum and other natural resources have been allocated using auctions designed by Milgrom and Wilson and their co-authors. The money won’t mean much to these winners, who have made plenty of money advising firms about how to bid in the auctions that they designed. Milgrom’s firm Auctionomics advertises its service and Milgrom notes: Milgrom has advised bidders in radio spectrum auctions, power auctions, and bankruptcy auctions. One advisee, Comcast and its consortium, SpectrumCo, followed the advice of a Milgrom’s team in FCC Auction 66 to
Alex Tabarrok considers the following as important: Uncategorized
This could be interesting, too:
Scott Sumner writes Is Covid’s impact semi-permanent?
Tyler Cowen writes Sunday assorted links
Alex Tabarrok writes Storyline–Abortion and Births
Tyler Cowen writes The class culture that is Britain
The 2020 Nobel Prize in Economics goes to Paul Milgrom and Robert Wilson for auction theory and the improvement of auction designs. The Nobel Committee has a popular introduction and good scientific overview of auction theory. Billions of dollars of spectrum and other natural resources have been allocated using auctions designed by Milgrom and Wilson and their co-authors.
The money won’t mean much to these winners, who have made plenty of money advising firms about how to bid in the auctions that they designed. Milgrom’s firm Auctionomics advertises its service and Milgrom notes:
Milgrom has advised bidders in radio spectrum auctions, power auctions, and bankruptcy auctions. One advisee, Comcast and its consortium, SpectrumCo, followed the advice of a Milgrom’s team in FCC Auction 66 to achieve the most exceptional performance in US spectrum auction history. SpectrumCo saved nearly $1.2 billion on its spectrum license purchases compared to the prices paid by other large bidders – such as T-Mobile and Verizon – for comparable spectrum acquired at the same time in the same auction. SpectrumCo’s tactics included a $750 million jump bid – the largest in the history of US spectrum auctions and a move that prompted the FCC to change the auction rules.
You can figure that Milgrom got a percentage of those savings! Milgrom also advised Yahoo and Google, among other tech firms, on their advertising auctions.
My post Mechanism Design for Grandma written for the Hurwicz, Maskin and Myerson Nobel, has some background on auctions.
Auction theory and auction practice arose together–this is not a case of theory being rediscovered decades later by practitioners but of the demands by practitioners leading to new theory and new theory leading to new institutions. The Nobel committee notes:
In the early 1990s, an explosion of the demand for mobile communication made the U.S. federal government decide to use an auction for allocating radio-spectrum licenses among telecommunication firms. Previously, the U.S. Federal Communications Commission (FCC) had only been allowed to rely either on administrative procedures—commonly referred to as “beauty contests”—or on lotteries. These methods had notably failed in a number of complex settings, at the expense of both taxpayers and end-users…The obvious alternative is to adopt an auction to as-sign licenses. In fact, as early as in the 1950s, the 1991 Laureate Ronald H. Coase argued that the basic principle should be to allocate objects, such as broadcasting licenses, to the firms who will make the most efficient use of them, and the best way to identify these firms is to assign the objects through a competitive price mechanism (Coase, 1959).
…Following the FCC policy shift, multi-object auctions turned from an esoteric topic at the fringe of microeconomic theory to a hot research topic almost overnight.
…For the 1994 FCC auction, the final version of the newly designed auction was the Simultaneous Multiple Round Auction (SMRA)…[which] raised some $20 billion for the U.S. federal government, twice the forecasted amount. This outcome attracted considerable media attention and led other governments to set up their own auctions. The U.K. 3G spectrum auction that concluded in 2000 raised about $34 billion for the British government (Binmore and Klemperer, 2002). The SMRA auction format became the dominant design for spectrum sales worldwide, and versions of it have been used in Canada, Finland, Germany, India, Norway, Poland, Spain, Sweden, theU.K., and the U.S. These auctions have generated hundreds of billions of dollars for governments worldwide.
Perhaps the most impressive culmination of this work was the 2017 incentive auctions which “simultaneously” bought licenses from over-the air broadcast television stations and resold them to modern cellular phone bidders while respecting constraints so that over-the-air frequencies could be repackaged in ways such that they would not interfere with one another. The auction bought licenses for $10 billion, sold them for $20 billion, generating $10 billion in profit and generating an even larger increase in consumer surplus.
The first is a reverse auction that determines a price at which the remaining over-the-air broadcasters voluntarily relinquish their existing spectrum-usage rights. The second is a forward auction of the freed-up spectrum. In 2017, the reverse auction removed 14 channels from broadcast use, at a cost of $10.1 billion. The forward auction sold 70 MHz of wireless internet licenses for $19.8 billion, and created 14 MHz of surplus spectrum. The two stages of the incentive auction thus generated just below $10 billion to U.S. taxpayers, freed up considerable spectrum for future use, and presumably raised the expected surpluses of sellers as well as buyers.
These auctions also brought home that economics is now tied to computer science. The complexity of the allocation process was so high that new algorithms had to be devised. In particular, repackaging of the frequencies involved solving hundreds of thousands of graph-coloring problems, an NP-hard problem. Computer scientist Kevin Leyton-Brown was brought in to design and optimize the necessary algorithms. At the same time, Milgrom and Segal had to prove that their auction could be characterized in such a way that it could be solved in reasonable time by known algorithms.
Computer scientist Tim Roughgarden has an excellent video lecture on how implementing the incentive auction required a combination of cutting-edge economics and computer science. More generally, mechanism design in the real world, including auction design, Uber’s supply and demand mechanism, blockchains like bitcoin and many other examples, requires both economists and computer scientists to devise institutions and algorithms that incentivize socially beneficial behavior and that can also be solved in real time for real populations.