Making Auctions Work in Practice Is Worth a Nobel

(Bloomberg Opinion) -- Quite literally since the days of Adam Smith, economists have been great believers in the invisible hand that guides markets. But sometimes, we give the invisible hand a little help.

When coordination problems or other complexities stop the market from finding the right price on its own, we often bring people together and discover prices through auctions. That’s the specialty of Paul Milgrom and Robert B. Wilson, two Stanford economists — and it’s the reason they received the Nobel Memorial Prize in Economic Sciences on Monday.

Auctions are everywhere, of course. People use them to sell art, antiques and houses; to find the right price for fresh fish and used cars; and to raise money for nonprofits. Auctions also hide in the backdrop of daily life, controlling the advertisements we see on the internet — and maybe even the page you’re looking at right now.

In many settings, the best auction to use is one in which bidders take turns offering higher and higher prices until all but one drop out. That’s an “English auction” in economese, and it’s great because people drop out exactly when the bidding exceeds their values for the good that’s being sold. This guarantees that the good goes to the bidder who values it the most; moreover the public process ensures that auctioneers can’t cheat by quoting an inflated price.(1)

Sometimes, though, other factors can interfere. Imagine an auction of an unusual Persian rug in a room with some expert Persian rug collectors. Amateur bidders may not be guided by their own estimates of the rug’s value, but rather by the actions of the experts. In such settings the ideal auction format is unclear.

We might think, for example, that the Persian rug sellers would prefer to hide people’s bids, so that inexperienced bidders can’t see when the experts bid low. But that also hides when the experts bid a lot — so maybe reducing transparency isn’t such a good idea after all. And the seller isn’t necessarily restricted to simple auction formats — in principle, he or she could use something strange like an all-pay auction, in which the good goes to the highest bidder, but everyone has to pay their bids.

So which auction format should rug sellers choose? Before Milgrom’s work, economists had no general framework for thinking through this question. They tended to rely on models that delivered extraordinary (even Nobel-winning) insights into how auctions might work in theory, but provided little guidance as to which auctions we should use in practice.

In joint work with Robert Weber, Milgrom developed and studied new, more nuanced models of bidders’ values and behavior. They found that transparency is often the way to go irrespective of whether there are experts in the room; indeed, auctioneers can even increase revenue by sharing information about the goods with bidders. This explains the detailed descriptions of item quality and provenance that we see at auction houses like Christie’s and on sites like eBay.

Wilson started out doing research in game theory. Along the way, he realized that early auction models relied on two assumptions: clarity about the rules of the game, and understanding of overarching features of the world, such as agents’ preferences. To think about real-world markets, Wilson reasoned, economists would have to do away with the latter type of assumption, since they treat agents as having information nobody could ever possibly have.

It sounds almost obvious when stated like that — but at the time, this observation was revolutionary. It was so influential that it became known as the “Wilson doctrine.”(2)And it led Wilson to be one of the first to build auction models that could map directly into real-world use.

Wilson and Milgrom’s auction models soon led them into real-world auction design. In 1993, the Federal Communications Commission invited them — along with Preston McAfee and John McMillan — to develop the first auction to allocate electromagnetic spectrum usage rights in the U.S.

Spectrum rights are a lot more complicated to auction than Persian rugs. Some companies want to build national networks, and only want to buy a license for one region if they can buy other regional licenses at the same time.

Moreover, when the government is the auctioneer, transparency and fairness are paramount, as is avoiding accidentally giving out licenses in ways that would create monopolies. And on top of that, the bidders in spectrum auctions are sophisticated and well-resourced telecoms; if there’s some way to game the system to get a better price, they’ll find it.

No auction mechanism in economists’ toolbox came close to solving all these problems at once. So Milgrom, Wilson and their collaborators developed a new type of auction that builds on the ideas of the classic English auction but sells multiple goods at once and lets bidders make simultaneous offers on packages of licenses. The approach was so successful that it has become a template for spectrum allocation worldwide.

Since then, Milgrom and Wilson have worked on auction designs in a range of industries. Wilson, for example, has been intensely involved in the design of the dynamic auctions used in electricity markets. Milgrom, meanwhile, advised the U.S. Treasury Department on the auctions used in the bank bailouts during the financial crisis, and recently wrapped up a new, even more complex spectrum auction that transferred licenses from television broadcasters to wireless providers.

All this work has helped start a revolution in market design, a new field in which economists develop practical theory that can be put to work in engineering (or reengineering) real-world marketplaces. And Milgrom and Wilson haven’t just done market design themselves — they’ve trained many extraordinary students in the field, too.

These days, thanks to Milgrom and Wilson, we help the invisible hand out all the time. They’re well deserving of the Nobel Prize — and I’d put in a bid for them any day.

(1) That said, participants still have to watch out for shill bidding.

(2) Not to be confused with this one.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Scott Duke Kominers is the MBA Class of 1960 Associate Professor of Business Administration at Harvard Business School, and a faculty affiliate of the Harvard Department of Economics. Previously, he was a junior fellow at the Harvard Society of Fellows and the inaugural research scholar at the Becker Friedman Institute for Research in Economics at the University of Chicago.

For more articles like this, please visit us at bloomberg.com/opinion

Subscribe now to stay ahead with the most trusted business news source.

©2020 Bloomberg L.P.