How technology must serve to correct the problems it created is part of the reflections of this Stanford University professor, an expert in complex mathematics and algorithms, awarded the Nobel Prize "for improvements in auction theories and inventions of new auction formats."
He helped shape the online advertising industry, collaborating with dot-com companies to build markets, which eventually became the flourishing ad tech industry. He designed the auction protocol used by the US Federal Communications Commission to determine which telephone company gets which cellular frequencies. And he led the team that designed the incentive auction to reallocate radio frequencies from TV broadcasting to wireless broadband uses, what we now know as "streaming." He is the author of several books and has held editorial positions at the "American Economic Review" and the "Journal of Economic Theory and Econometrics." He is a member of the Econometric Society and the American Academy of Arts and Sciences. He is also a member of the US National Academy of Sciences and vice president of the American Economic Association.
In the interview, he reflects on the role of economists in designing emerging markets for their proper functioning, the use of technology, and the development of artificial intelligence to solve the most urgent problems that threaten humanity.
Jorge Fontevecchia Co-founder of Editorial Perfil - CEO of Perfil Network.
@fontevecchia
@Fontevecchia 05-04-2025 23:55 —To introduce the audience to the subject of your specialty, what are auctions, what are they for, and what types of auctions are the most common?
—I use auctions to describe any resource allocation process that involves prices. Auctions are used in multiple areas, such as in the purchase and sale of consumer goods on sites like eBay. There are auctions that companies use to acquire goods and services in which they try to negotiate quality and price, and these auctions are more complicated because it is not just about price, but can also involve quality or other attributes. There are auctions that involve the allocation of groups of goods, so that, for example, in radio spectrum auctions in the United States, a company might or would want to acquire licenses that cover several different geographical areas that allow it to develop a business plan. That is a different kind of auction where combinations of things matter. There are really two ways to classify them. One is to distinguish between sealed-bid auctions, where you make an offer or a list of offers only once and the result is calculated from that, and dynamic auctions that take place in a series of stages. Dynamic auctions can involve prices going up, as in ordinary consumer auctions, or going down over time, as in Dutch flower auctions. There is really a wide variety of these methods that are in use.
"Currently, the first version of technology has led us astray."
This is not to the liking of authoritarians
The exercise of professional and critical journalism is a fundamental pillar of democracy. That is why it bothers those who believe they are the owners of the truth.
—There was a kind of two phases. From the founding of the Federal Communications Commission, sometime in the late 1920s to 1993, there were no auctions. And at the beginning, most of that period of time, there were comparative hearings, which meant that if you had a use you proposed for the spectrum, and I had a use I proposed for the spectrum, there would be a hearing in which the Commission would weigh which application best served the public interest. For example, they would say, "You know, you can start a TV channel in Buenos Aires, and I can't." I do not think that was the decision that was made. But what happened was that in the 80s, we had the introduction of mobile phones and we had small telephone companies throughout the country that wanted to start providing mobile phone services. And over time there were thousands of applications that far exceeded the capacity of the Federal Communications Commission to use comparative hearings. They wanted to use auctions at that time, but there was a political struggle about it. And as a result, they ended up using a lottery for a few years, where if you applied for a license, they would simply draw a name out of a hat and decide, "Well, we are going to accept your application and not Joe's." It was something random and not very efficient, which led some people to make a lot of money in a lottery run, essentially, by the government. So, in 1993, Congress authorized the Federal Communications Commission to use auctions to sell spectrum rights to the bidder who was willing to offer the highest value for the spectrum. And it also raised money for the government. But when they decided to do that, there were some complications that led to my original auction design. There were large companies that wanted to create national networks. There were a lot of existing small companies that only wanted to provide local service in their areas. And we did not yet know that mobile phones would work globally, that they could be taken everywhere. At first, it was thought that it would be something local. And it was very difficult to decide what to sell. Should a license be sold for the whole country, for half of the country, for the states, or for the cities? So the government decided to sell everything together so that bidders could bid for whatever they wanted. And they were not sure how to do it. That is where the simultaneous multi-round auction came into play. People could bid for everything simultaneously, but then they could know what was expensive and what was cheap, and they could continue bidding to acquire the packages of licenses that were most valuable to them.
CONCERNED ABOUT TRUMP. "I am very concerned about how climate change can worsen problems for everyone. We will need to rely on science to find solutions to big problems."
—In the case of this incentive auction that the FCC applied in 2017, it transformed the media by transferring the ownership of waves from television stations to data providers, paving the way for the streaming industry. Did you imagine at that time the impact it would have on how the media changed and how we get informed today?
—I mentioned a lot of that. I was not the only one who imagined it. The reason for the incentive auction, when Barack Obama was elected president of the United States, was that one of the two main technical priorities was to increase the availability of radio spectrum for these emerging new services. There was already enormous growth in wireless and data communications when Obama was elected. Everyone predicted the development of technology and believed that there would be much more data communications. There is no spectrum available for it. We needed to free up part of the television spectrum, and the television spectrum is because what had been happening before the incentive auction was that more people were using cable or satellites instead of local over-the-air broadcasts, and it was felt that the value of the spectrum and over-the-air television broadcasts was falling. The value of data was increasing. And if we wanted to facilitate this type of transition, we needed a process. And the problem was: what should the process be? It was incredibly complicated.
"I hope the left improves, because I think the far right is very dangerous."
—Is it common for new problems to arise and that requires adjustments between theory and practice to generate new auction formats?
—What happens in economic theory, and I think in the theory of many fields, is that it is abstract. It is supposed to be general, that it offers a broad view of what is happening. But when you are solving a specific problem, you have to pay attention to the details. In the case of the Federal Communications Commission, radio spectrum interference, no one had taken that into account in auction theory. I cannot connect your station and a neighboring station to the same channel without generating interference. We had to consider the interference that could be generated when moving. We had to take the free spectrum. We had to move some people who were broadcasting on certain channels to others. We had to do it without generating interference or buying space at the same time. It was a very complex problem, and no one had analyzed restrictions of such complexity before.
—Given that technology has exacerbated political polarization through social networks and the trends towards ideological extremes on the part of the population who largely receive news that comforts them and pronounces their confirmation bias, can technology with new tools solve the problems it creates in this area that weaken democracy and peaceful coexistence among citizens? Do you think technology can solve the problem of technology?
—I am waiting for ideas to do it. That is, I think many of us understand exactly what you say. People live in what we call echo chambers. They only listen to reinforcements for the opinions they already have. I know that, for my part, I make a great effort to subscribe to both left-wing and right-wing newspapers to try, at least, to be aware of the most intelligent opinions on both sides. There are intelligent people on both sides, but we need to find a way to encourage more communication of that type. I think that, currently, the first version of technology has led us astray, but changes are taking place. That is, we know that AI is changing the way people get information. And I hope that some of my colleagues and some of my brilliant Stanford students have ideas to promote a more comprehensive vision that weakens the resonance chamber, or even eliminates it.
UNDERSTANDING THE WORLD OF AUCTIONS. "The reason for the incentive auction was that one of the two main technical priorities was to increase the availability of radio spectrum for emerging new services."
—I ask you to explain to our audience why the role of auctions is important in the management of public resources.
—From the point of view of public resources, there are generally two main objectives. If it is sold, one of the objectives is to generate income. I remember that when I was working at the FCC, the Republicans, in particular, were very interested in how much money would be generated. And the other objective is to promote a better allocation of resources, and again, that specific case related to clearing television broadcasts. But if you are acquiring covid-19 vaccines, and you want to do it at the lowest possible cost and have them distributed correctly to the right groups of people, you design your resource allocation process to achieve the objectives you set for yourself. So, as an advisor, I first talk to people about what they are trying to achieve and then I try to design the process that promotes that objective, that reduces the costs of what they buy, increases the income of what they sell, or improves the allocation to benefit people, or favored groups. I could tell you more about this, but I will stop here.
"People live in echo chambers, they only listen to reinforcements for the opinions they already have."
—The market is a communication system. In the time of the former Soviet Union, without computers, the Ministry of Planning, Gozplan in Russian, manually calculated the preferences of the population simulating a market combined with the supply capacity of its production system, but they could never be as perfect as the real market, they were always late and ended up allocating more resources and being inefficient. With current computer and computational resources, could incentives be correctly allocated to simulate a virtual communication tool between supply and demand as effective as the market? Or will this be possible in the near future?
—I do not think so. All this is speculative. But there are two important things that are happening technologically. One is computing power, as you mentioned, and the other is communication, which is going in exactly the opposite direction. And I think that is the most powerful force. The truth is that a huge amount of information is needed for everything to work well. That is, in your own business, if someone is sick and you need to replace them, the central planner does not know who is able to take one position or another. If a machine breaks down, the central planner does not know who can fix it or what can be done in the meantime. All that information is local. And what is really needed is to allow local people to make decisions effectively so that they fit into the plan as a whole. Communication is improving, which makes it easier and easier to effectively manage a system by providing the necessary information. In fact, I can sit here if I need to know something, write a question, and get the information I think I need to answer my questions. I think technology is moving in the opposite direction to the one you mention.
—Could it be said that bank runs are a form of auction, I do not know if illegal but that end up generating crises? Could your theory help solve the problem of bank runs?
—That is a different problem. In fact, recently Phil Dibbig and Doug Diamond won the Nobel Prize for their work on banks, and with Ben Bernanke and bank regulation. And I think we have a different understanding, quite different, that bank runs are a coordination problem. Once a run starts, you have an incentive to participate because you do not want to be the last to try to withdraw your money, that is what happens in bank runs. And the government helps to prevent them by insuring deposits and giving confidence to depositors that they do not need to participate in the run. That seems to be the most sensible way to try to prevent them.
"It is necessary for companies to succeed and grow, even if they approach monopolies."
—One of your specialties is the financial market. What is your vision of the world of cryptocurrencies within the financial market?
—I must say that cryptocurrencies ended up surprising me. At first, the demand for cryptocurrencies came from the criminal sector, from people who wanted to keep their financial transactions private. And later, it turned out that distrust of the government, the desire for privacy to have transactions that could not be monitored, the desire to have money in a form that was not subject to the government printing money and inflating its value. Those were the things that promoted the development of cryptocurrency. But it is very expensive. We know the energy consumption that Bitcoin demands, it uses an extremely large amount of electricity. Just like in big cities, the amount of electricity that is consumed does not generate similar value. In addition, they are vulnerable to attacks. If they become very valuable, they can become more vulnerable to attacks. And it is really not yet clear what the limits are. Cryptocurrencies will have to be better protected and governments could decide that they need to be regulated because of the criminal part of it. I am not sure how the balance is found in cryptocurrencies. I am a bit skeptical. I myself have a very small investment in cryptocurrencies.
—Tell us about the field of market design, which encompasses this vision of allowing economists to work as "engineers" when design can improve the functioning of new or existing markets that do not function properly with their current operating rules.
—It is not just about them not working properly, by the way, but about new markets for things that have never been traded before. In market design, we interpret the market as something more than what your readers will probably anticipate, by allocating organs, children to classrooms, or students to residences. Those are also resource allocation problems. We do not use money for those problems, but that is largely part of market design. We try to figure out how to do a good job allocating those resources as well. You may know, for example, that when you browse the web, every time you see an ad on a website, an auction has taken place in a matter of milliseconds that places the most valuable ad on that site, the advertiser who was willing to pay more for that impression. We have new markets that are proposed for 'cap and trade' to limit environmental damage and allocate it. We have new markets for labor allocation where skilled people can advertise their skills. If you need someone to do repairs at home, you can find someone online. We have dating platforms that also seek to connect people properly. They are all new markets, it is not about deficient markets that need to be improved. Some of these markets did not even exist before. There were no centralized or organized methods to allocate these resources. And it works much better when we can organize them.
EDUCATION IN CRISIS. "If Trump damages universities, and I think it is not yet clear how much damage he will cause, but he is certainly a threat, that worries me."
—What do you think about regulations in the markets? Do they work, or are they useful in some markets, and not in others?
—The types of regulation that are useful definitely vary from market to market. And they are regulations that are not only imposed by the government. I can give an example in internet advertising. Publishers like people to bid to display ads. That is how they raise money. But they do not like to see spam, malware, or fraudulent ads on their websites. That hurts them. That is why they create regulations, create filters, and create rules to try to block them. So, we see extensive regulation even in private markets. We try to eliminate fraud in financial markets. The vulnerability of markets to these problems varies. Some require more regulation, others less. I guess the main question is not whether markets require regulation, but whether governments should do it, or is it better to have private markets that respond to the specific problems and challenges that those individual markets face.
—There is a conflict between two economic theories. Libertarians and neoliberals. Libertarians, at least in Argentina, President Javier Milei, says that there are no market failures. Is that a mistaken view of neoliberals, that the market is always perfect? Do you think the market has failures, or not?
—We certainly have examples of markets that failed. You mentioned the previous example of bank runs. That is a problem. In those markets, transactions were carried out freely without government guarantees. We used to have problems with bank runs. And, in general, those problems were solved with the Silicon Valley banking crisis here in the United States a couple of years ago. It did not spread. No other bank was harmed because the government intervened, and even increased the guarantees beyond what was promised, to say: we want people in the United States to consider banks safe. There are different types of failures in different markets. Some markets work quite well without regulation. I think the problem that worries or should worry libertarians is real. Excessive regulation can be harmful, it can prevent innovation or people from participating in really beneficial transactions through bureaucrats who have no idea what is happening between the people involved. Therefore, we need a flexible policy, but we need regulation. Even when the government does not intervene, we have private regulation, as I have already described, in relation to internet advertising.
—And another question related to this is whether monopolies are always bad, and whether they have a solution over time, or is regulation necessary when there are monopolies in a situation of abuse and market control?
—Monopolies are more complicated than you just described. Sometimes what we do not want is to prevent competition from companies that want to grow. Right now, for example, in the United States, we have antitrust actions against Google, because it has had a dominant position in search. It got a dominant position in search, when it started as one of the small search companies. It had a much better search technology and grew. And if there were a law that prevented it from growing, it would have no incentive to innovate. It is necessary for companies to succeed and grow, even if they approach monopolies. And it must be taken into account that what breaks those monopolies is innovation. For example, today, with the emergence of AI, many people are starting to use Anthropic, the chat GPT, or other sites to search, because they believe that new technologies are better. What keeps these companies, what keeps monopolists alert, is the possibility of competition. What must be ruled out, what must be regulated is what is called monopolization. And I understand that the language can confuse people, because it seems that it only means creating a monopoly. But what it really means legally, at least here in the United States, are the actions that are taken to gain control of the market, which do not refer to setting lower prices or better quality, but to blocking or creating barriers, actions that harm competition. So, we have to make sure that competition is fair, that competition is not harmed. And then we have to allow people to compete, even if their companies grow as a result of it.
"What must be ruled out, what must be regulated, is what is called monopolization."
—You are one of the founders of the company Auctionomics and chairman of the Board of Directors, you invented the "Milgrom allocation auction," the patented design that makes Auctionomics unique, could you give us an example of a practical application of this type of auction?
—Yes, in fact, the Milgrom allocation auction, to which you refer, is very similar to Paul Klemperer's product mix auction. Paul is an economist at Oxford, and in the United Kingdom, his auction was used for Treasury bonds, just when I was trying to get mine used for US Treasury bonds during the financial crisis. He got the Bank of England to use his auction design for it. And I have also advocated the use of my auction for certain agricultural commodities. For the most part, they have resisted. I have not had much success in adopting that design.
—Is it true that the idea for Auctionomics arose as a result of the 2008 financial crisis, when you spoke with the US Treasury Department about designing an auction to sell distressed assets?
—Yes, that is where the assignment auction came from; it was my design, which is basically the same as had been proposed in the United Kingdom. The United Kingdom adopted it, the United States did not. But it was to sell distressed assets. And yes, that was the first commercial auction design we had. We have worked on many other commercial designs subsequently, although it is largely confidential and I will not be able to tell you much more about it.
—Did winning the Nobel Prize change your life?
—You know, I did not think it would change my life at all. I did not plan to win a prize, just to continue doing the same work. But it had a much greater impact than I thought. It is a very famous prize. And it changes the way people receive my ideas. In my professional life, people are more willing to listen. Especially outside the Academy, people think, he won a Nobel Prize, maybe we should listen to what he has to say. And when I want to do something new, different from what people believed, they are much more likely to listen and adopt my ideas. So, I found that, professionally, this has a great impact on how I am perceived, especially outside of academia.
—You mentioned in another interview that having a daughter helped you realize the gender discrimination that can exist, even in the academic world, and that this served as a basis for you to start dedicating yourself to training more women economists. Do you think there is a backlash against diversity, equity, and inclusion policies?
—Well, that is two parts. Let me answer them separately. The first part is not so much that I became more aware of discrimination, but that I realized that teaching had to be different. That the kind of things that would encourage the participation of my male and female students were different. I want to quickly move on to the second part of your question, but I started to change the way I taught to make it more effective for both women and men. And that was important. You asked about diversity, equity, and inclusion. As we know, and I am sure many of your readers know, in the United States, Donald Trump has been very hard on that. That is not what I saw myself doing. What I do is try to create teaching methods that work for everyone. I want a diverse group of people to succeed, but I do not want to do it by discriminating, making others perform worse. I want to do it by improving everyone. And for me, what I learned from teaching my daughter is that the things that worked for her and that worked for my son were different things. And I needed to have a wider set of tools to help everyone achieve what they were capable of achieving. I had a very different goal from the DEI goals.
"I use auctions to describe any resource allocation process that involves prices."
—You work with artificial intelligence, use machine learning algorithms to run auction simulations and determine the best strategy. What is your vision on the power of artificial intelligence? What do you think of the use of artificial intelligence as a tool for the efficiency of public administration?
—Artificial intelligence, as we have it today, is quite new. It is only a few years since we have had these very powerful AI algorithms. And I have found effective ways to use it in my work, both to research, find information, and summarize it, when there are millions of pages of evidence or information available. To get summaries and find out what is happening, artificial intelligence can do it in ways that were not possible before. As for its other uses, it depends on how it continues to develop. Right now, I think it is a valuable tool for some people. It helps to automate some activities that we all do. It does not replace me. I do not think it will replace you, but it could help both of us to collect and refine information. And we will see what capabilities it brings in the future. I do not think it is yet; well, there are some jobs that will certainly be threatened in the very near future. There are many other jobs that I consider to be well protected.
—Do you think artificial intelligence can be a useful tool to achieve greater social equity?
—That is an interesting question that invites reflection. I think it will first create more inequality. That is, those who benefit from its use and those who do not, will probably be different groups, and they could be even more distant. In the long term, it will provide a valuable tool for everyone. It will benefit a wide range of people, but it could benefit some more than others. I would be surprised if it did not benefit some more than others. So, it can have the effect of improving the well-being of some and, at the same time, increasing inequality. I think both things are possible.
"The types of regulation that are useful definitely vary from market to market."
—Are you concerned about climate change and the climatic catastrophes that seem to be accelerating more and more, as a consequence of climate change?
—Absolutely. I am very concerned. About the loss of species, the loss of habitats. Right now, we already have tremendous tensions in the world derived from the migrations of people, due to wars, etc. But when we have droughts or changes in the climate and areas that become less productive and people want to move, that is only going to get worse. I am very concerned about how climate change can worsen the problems for all kinds of people, both for the people directly affected, and for the migrants who arrive in places where there is resistance to migration.
—Is it true what I heard about you organizing academic barbecues in the garden of your own house, together with your university students and colleagues?
—Yes, it is absolutely true. And it is one of my secrets, when we were talking about my daughter, I think it is one of the secrets of my success as a student advisor. I have tried to create a community. The reason I do it is that I want students to interact more with each other academically. And I want them to see me as someone who encourages them. One of the things that happens during research is that there are periods of failure, when the research does not go well, and there is a risk of depression. This is well known among graduate students, the risk of students becoming depressed. When we have a community, there are other people to talk to, other ideas, support. And there is also competition where people set standards and no one wants to be the worst in the group. All my students at Stanford aspire to be excellent and see what others do, which inspires them to do more. I have had great success with students who come and there are both things, there are barbecues, it is true, but there are also
