At a Glance
- Auction Design Innovator Created Platforms Adopted Across the Globe
- Stanford Economist Awarded CME Group-MSRI Prize
Paul Milgrom’s work in economics has made him a giant in the field. He’s developed world-changing ideas on game theory, auction theory and market design at places like Northwestern, Yale and Stanford. But when he graduated from the University of Michigan in 1970, he didn’t enter a PhD track, and had no plans to.
Instead, he moved to San Francisco where he worked for five years as an actuary before heading to Stanford to get an MBA. While there, his intellectual curiosity stood out, and staff recruited him for the PhD business program.
This set a course for an economist who helped change the way we understand modern economic theory. But in many ways, it was his early non-academic experience that informed his groundbreaking ideas and research.
“When I went from the working world back to graduate school, I was the student who was best grounded in real application,” says Milgrom, the Shirley and Leonard Ely professor of Humanities and Sciences in the department of economics at Stanford, where he’s worked for more than 30 years.
“I had practical experience that made my theory better grounded than some of the others.”
Forever Changing Economics
Milgrom’s commitment to real-world application led him to Northwestern University, where he was part of a group of economists who engineered a movement in economic theory.
For decades, economics was understood through general equilibrium theory. This states that somehow the market finds prices that balance supply and demand.
Milgrom and his Northwestern colleagues, which included Nobel laureates Roger Myerson and Bengt Holmstrom, began to ask different questions. In industrial competition, firms had market power and some had better information. This influenced their behavior in the market, which created a whole new set of problems that affected competition. The study of these problems among the young economists at Northwestern gave rise to game theory as a dominant economic principle.
“Game theory was ideally suited to ask all these additional questions because it had a language within which to describe, ‘Well, what are the moves that are available to players? What do the players know? How many competitors are there?’ You can vary all of that,” says Milgrom.
About his days at Northwestern he adds, “It was gathering many of the greatest young people who were doing things that were outside the mainstream in economics. And our understanding of the economic system was forever changed by the work of that group.”
Myerson says Milgrom stood out in the group due to his deep understanding of mathematical probability theory and his focus on auctions.
“Paul’s perspective was sharpened by his early focus on the study of auctions, which he recognized as an important application in which economists can begin to understand how prices are determined by competition among individuals who have different information,” says Myerson, now a professor of economics at the University of Chicago.
Paul Milgrom, second from left, with Alvin Roth, Robert Wilson and Bengt Holmstrom at the 2017 CME Group-MSRI Prize ceremony.
Shengwu Li, a junior fellow in Harvard’s economics department and former student of Milgrom’s at Stanford, says a lot of information about auctions was assumed before Milgrom released some of his early work on game theory.
“Before Paul’s work, we assumed that every bidder in an auction knows beforehand how valuable the object is. But real bidders might revise their estimates as the price changes. Paul’s contribution is studying what we can infer from prices, and what prices can infer from us.”
Auctioning Data Coverage
Milgrom has authored more than 100 influential academic papers, but his best known work might be his 1993 creation, with Robert Wilson, of an auction design that allowed companies to license the airwaves we now depend on for phone and data service.
The Federal Communications Commission (FCC) wanted a format that allowed both small and large telecommunications companies to participate in a process called spectrum auctions. In doing their research on the subject, regulators kept finding Milgrom and Wilson’s names cited in the footnotes of academic papers on auction theory. They eventually relied on their ideas to create the platform.
Milgrom and Wilson created an auction that allowed bidders to see prices as they developed and substitute among different kinds of licenses. Today, the FCC auctions have generated over $100 billion for the U.S. treasury and have been adopted by several other countries, including the U.K., Australia, Canada, Mexico, Germany and Spain.
Turning TV to Broadband
The average consumer may not be familiar with spectrum auctions, but they are, as Milgrom puts it “quite a big deal.” And none have been bigger than the one completed in 2017, the creation of which Milgrom calls his “single proudest professional accomplishment.”
In a highly complex auction designed by Milgrom, the FCC reallocated spectrum used for television broadcast to make it available to mobile carriers and broadband providers. It was more than five years in the making, and in the end T-Mobile, Dish Network and Comcast came away as the winners in a $19.8 billion auction.
Before that could happen, more than one million radio spectrum interference constraints had to be satisfied.
“It was an extremely complicated problem to figure out if you bought this station and didn’t buy that station, what kinds of rights would you be able to create to sell? And we needed to make the auction simple for the bidders so that the complexity didn’t overwhelm them,” says Milgrom.
Milgrom has seen no shortage of recognition for his contributions to economics. He’s a recipient of the Nemmers prize from Northwestern, the BBVA Frontiers of Knowledge Award in economics and, most recently, the CME Group-MSRI prize, which has seen five of its 10 recipients go on to win the Nobel Prize in Economic Sciences. Milgrom collaborators Wilson and Holmstrom have previously won the MSRI prize.
“Paul has been the preeminent leader in analytical approaches to the practical design of auctions, from the early spectrum auctions to today,” says Myerson. “But beyond the specific applications, he has also always thought deeply about how the ideas of auction theory can be extended to understand all kinds of markets.”
Looking for Complexity
Earlier in his career, Milgrom wrote a series of influential papers with John Roberts, now his colleague at Stanford, around industrial organizational economics. The first of these was on limit pricing, the practice of one firm lowering its prices enough to discourage entry by competition. It was his first look at game theory, and was influenced by a case he saw in his time as a consultant
He’s also partnered with Holmstrom several times for research on contract theory, which examined the optimal design for incentive structures within organizations, in particular why strong incentives can sometimes do more harm than good.
These days, Milgrom’s interests echo his recent work on the spectrum auction. He’s trying to understand markets where complex constraints must be respected – things like electricity markets, airport landing slots and other transportation markets where accidents occur if constraints are not honored correctly.
“You need a solid combination of prices and regulation where the prices guide the behavior, but the regulation assures that it stays within safe bounds, that you don’t have two planes trying to land at the same time. Or you don’t have blackouts in your electricity.”
Milgrom’s influence is experienced worldwide, but perhaps most felt by his students.
“Paul is really good at helping his students to see where the unexplored territory is,” says Li.
Li tells the story of proposing an issue for his Master’s thesis to his supervisor while at Oxford. His supervisor said it sounded like an issue that must have been studied already.
“And I said, ‘Well, Paul Milgrom tells me that as far as he knows it’s an open question,’ and I think my supervisor’s response was, “Well, if Paul Milgrom doesn’t know an answer to this question then it’s definitely an open question.’”