Why Nobel Laureate Paul Milgrom believes auctions can help create a more efficient and sustainable economy.
"Going once, going twice, sold!" You're unlikely to hear this iconic sentence in many of today's auctions - especially when public goods are at stake. A new kind of auction that was held in the US in 1994 turned out to be groundbreaking in this respect: instead of expensive art, radio frequency bands for mobile phones were sold off to the highest bidders.
The US government tasked two Stanford University economists, Paul Milgrom and his mentor Robert Wilson, with devising rules for the complicated sale, as both had studied auctions for many years. Milgrom and Wilson developed a novel concept for the bidding process, which proved to be a huge success. The first auction brought in 617 million US dollars - considerably more than many had expected. More radio spectrum auctions followed and by 1997, buyers had paid a total of 23 billion US dollars for wireless and satellite licenses.
As a result of this success, the new auction model was soon copied around the world. In 2020, Sweden's central bank awarded Milgrom and Wilson the Prize in Economic Sciences in Memory of Alfred Nobel for their pioneering work. Through his company Auctionomics, Milgrom now advises governments around the globe who are looking to use auctions to distribute public goods fairly and at the same time address issues such as overfishing, drought and air pollution.
In an interview with MAG/NET, the 74-year-old explains why auctions play an important role in making the economy more sustainable and help address very complex matters such as the trading of water use rights. The latter is a topic that Milgrom is currently focusing on, and one he explored in his lecture at this year's Lindau Nobel Laureate Meeting.
Professor Milgrom, what are the advantages selling public goods by means of an auction?
When people think about Sotheby's and Christie's - these are auctions of unique items. But we also have to figure out what we're going to pay for carbon emissions, electricity or water rights. Auctions are well suited for the competition between buyers and sellers because fundamentally, what auctions are about is discovering the best price for a transaction.
How do auctions for public goods differ from those used to sell other items, say, on eBay?
Private auction houses and eBay apply a relatively simple bidding principle. But with interrelated items, getting an auction to be fair and efficient can raise complications. When we’re reassigning television channels, for example, we have to make sure that the broadcasts don't interfere with each other.
Another example is the allocation of water rights. Not just because the quantities have to add up but because some of the water that you use returns to the river, and the use of water by one farm can affect the availability of water to other farms.
Far beyond selling famous paintings, auctions are now also used to sell government bonds and emissions certificates. Competitive bidding is omnipresent in the economy and is being applied in a growing number of areas.
Paul Milgrom and Robert Wilson, two economists from Stanford University, have been studying auctions for decades. In 2020, they received the Nobel Prize in Economics for their efforts. One of the aspects they have examined is the "winner's curse". This phenomenon occurs when auction participants fear that they may overpay with the winning bid, thus making it more likely that they will hold back. To avoid this, auctions should be made transparent and a realistic price should be communicated to the bidders from the outset, Milgrom and Wilson argue.
Drawing from their research, the two scientists developed the concept of simultaneous multiple-round auctions, a groundbreaking new model for auctioning radio spectrum licenses in the US, beginning in 1994. As the name would suggest, all items are auctioned off simultaneously at these auctions. Bidders learn about competing bids at the end of each round. They can then decide whether to keep bidding or drop out. The auction ends when no further bids are submitted.
This concept has since become well established and is also used to auction other public goods, such as fishing rights or electricity.
These things are related and you can't just sell things one at a time. You need to take a look at how the whole system interacts. That requires a special type of auction, and those are the ones I tend to work on most often.
What are some of the most important aspects to consider in the design of an auction?
You want the rules to be simple so that it is easy for participants to figure out how to bid. At the same time, you want flexibility so that someone can come in and think, "I really want these rights here - but if that doesn't work, I'll try over there." And you want an auction that makes it possible for participants to express that preference, or bid for that outcome - and if they’re willing to pay a sufficient price, to be able to get what they want without being overcharged.
How do you prevent participants from waiting until the last second to place their bids in order to catch the biggest fish at the lowest price?
That is called auction sniping, and it was an important factor in the development of the 1994 radio spectrum auction in the US - my first big practical application in this field. At a charity auction I had observed that people looked around but didn't place their bids until close to the end, so there was no time for others to make a counterbid. This was on my mind when my colleague Robert Wilson and I started designing the radio spectrum auction for the US government. It was clear to us that we had to prevent sniping.
How did you do that?
We came up with a set of new and unusual rules that made the auction very successful and was copied all over the world. One rule was that the auction would take place in a series of rounds. And if there was a bid during one round, then there was always another round afterwards. So you could never put in a bid and prevent people from responding. To encourage active participation, we also created a rule that all bidders had to be equally active in previous rounds. You couldn’t save all of your activity until the end of the auction and outbid the others. That forced the price discovery process to happen, and it worked wonderfully.
Governments are certainly grateful to you. The auctions for mobile communications licenses alone have contributed billions to state coffers. But telecom companies often complain that they ultimately lack the money to expand network coverage because they were forced to spend so much on the spectrum licenses.
That's a political argument and I don't think it can be taken very seriously. If you have an asset that belongs to a private company, that company wouldn’t sell it to you for less than it's worth. They would have an auction for it. And all of a sudden, these companies that believe in capitalism think that it's a bad idea for the government to behave in the same way a private company would behave? If a technology is valuable, companies can finance their investments.
With interrelated items, getting an auction to be fair and efficient can raise complications.
Even in quickly evolving segments like the internet and mobile communications?
I've heard that argument before, but even if it's true: companies don't have to pay more than it's worth. If they're bidding for it, it's because they expect to be able to profit from it. And if it's valuable spectrum they can use it as security when they’re borrowing. I really think those arguments are just lobbying, they're intended to make the companies pay less to the government.
Can the mobile spectrum auction model serve as a blueprint for other public goods?
The way you design an auction depends very much on the context. For example, the allocation of water rights is not just about the amount of water each individual farm receives, because a significant amount of the water returns to nature - be it in streams or groundwater; and what one farmer does impacts the surrounding farms. So these things are related and you can’t just sell things one at a time. You need to take a look at how the whole system interacts. Those kinds of auctions really call for a closed design, where you take into account all the effects that are going on.
What does it take for that to work?
Technology plays a huge role. Satellite imagery is an example: When you take an acre foot of water out of a stream and use it on a field, not all of it is actually used by the field. Some of it evaporates. Some of it goes into the plants. And those things can be detected by satellite measurements these days. That’s an important part of the solution because if we want to ensure these rights can be traded fairly, we first need to understand how the system works and what impact trading has on other stakeholders.
We are facing a similar challenge at the global level: there is only one planet, but everyone lays claim to its resources.
With regard to climate change it's clear that we need to limit carbon emissions, and we need to ensure that those limited emissions are used correctly. There are some very valuable things that cause carbon to be emitted. And we want to direct our limited emissions to these more valuable uses. And when things can be done in ways that emit less carbon, we would like to promote that as well. Carbon markets, which set a price for CO2 emissions, help us achieve both of these goals.
If investors support the development of new solutions, that's a step in the right direction.
Emissions trading is criticized for not being effective enough.
It's been very controversial, true. And like in the case of water rights, it requires measurement. We need to know how much carbon is actually being emitted and who is causing the emissions - or who they should be attributed to. Take gasoline as an example: who is responsible for the emissions it causes - the refinery that produces the gasoline? Or the car owners who burn it? And maybe you can't measure the final uses, maybe it's easier to measure the amount of gasoline that’s produced. Those are going to be among the complexities that we need to address in order for the system to work.
Complex usually means slow. Can we speed this process up?
I'm pessimistic about that. We saw that in the United States. One alternative to emissions trading is carbon taxation, but people were against huge new taxes. The Biden administration decided to go the other way, with rewards for reducing carbon use instead of penalties for continuing carbon use. We’re still struggling to find things that will actually work, and I’m not sure we’re making progress fast enough.
What can investors do to help?
We've already talked about the important role that technology plays. If investors support the development of new solutions, that's a step in the right direction. Water consumption is a good example: We need better ways to measure the effects of water usage to create effective water markets. Once we have effective water markets, it will become economical for people to reduce their water use. And then the technologies that help the farmers reduce their water use will become valuable. Right now, desalination is energy intensive and expensive - but technologies along those lines could be developed that will help address these problems.
Can market mechanisms help solve these problems - or do they have limits?
What markets can do that governments can't is they can be responsive to lots of individual variations. They take advantage of decentralized knowledge, decentralized initiative. Governments can’t do that. But markets can also fail in many ways. It's not always straightforward to create markets that work well. And when it can't be done, sometimes the only solution is government intervention. My goal is to lubricate these markets and help them function better in places where I think they do have potential.
Header Visual © Keystone/ AP Photo / Seth Wenig.