Evan Kwerel on the Origins of Spectrum Auctions

Tom Lenard:

Hello, and welcome back to the Technology Policy Institute’s podcast, Two Think Minimum. It’s Wednesday, April 20th, 2022. I’m Tom Lenard, President Emeritus and Senior Fellow with TPI. And I’m joined by my colleagues, Scott Wallsten, TPI’s President and Senior Fellow, and TPI Senior Fellow Sarah Oh Lam. 

Today, we are delighted to have as our guest, Evan Kwerel, who is Senior Economic Advisor at the Federal Communications Commission. The impact of Evan’s career at the FCC was recognized last year, when he was awarded the 2021 Paul Volcker Career Achievement Award for pioneering the use of spectrum auctions. To get an idea of what Evan has accomplished and to introduce the discussion, let me read the first couple of paragraphs from the citation. 

“During more than three decades as a Federal Communications Commission economist, Evan Kwerel has been a key driver of America’s wireless revolution, establishing the first-ever competitive auctions to allocate public airwaves for the transmission of sound, data, and video across the country while raising billions of dollars for the government.

The market-based FCC auctions of electromagnetic spectrum, the radio frequencies that carry voices between cell phones, television shows from broadcasters and online information from one computer to the next, were conceived and implemented by Kwerel based on many of the theories of 2020 Nobel Prize-winning economists Paul Milgrom and Bob Wilson.

Since the early 1990s, a total of 107 FCC spectrum auctions have generated more than $200 billion in revenue for the government.

After winning the Nobel Prize, Milgrom wrote that ‘Evan’s individual contributions were so major that it would have been appropriate for him to share this prize.’”

So, I think it’s to fair say that very few people working in government at any level have had the impact on public policy that Evan has had. So, I’d like to talk not only about what you did, Evan, but how you did it. The idea of spectrum auctions was introduced by Ronald Coase in 1959, and though it seems obvious now, it was not taken particularly seriously at the time, and it also usually takes more than a good idea in government. So, talk about how you and perhaps others introduced the idea at the FCC.

Evan Kwerel:

Thank you for inviting me to this podcast and giving me the opportunity to talk about it, because I do think it’s an interesting story, but before I start on that story, I want to put it in a slightly broader context, and the context is managing spectrum overall. And I think what I did was help implement Coase’s broader vision that markets could and should play a central role in managing spectrum. And as in that blurb you read, , I did it in collaboration with many others, , some remarkable people, , including Paul Milgrom and Bob Wilson and my fellow collaborator at the FCC, John Williams, an engineer, who was really instrumental in all of this. But I think, while spectrum auctions are the most well-known aspect of Coase’s vision, they’re not necessarily the most important. I mean the whole definition of property rights for spectrum is something that I spent a lot of time working with John Williams to try to move from a command-and-control approach to managing spectrum, to a market-based approach. And I think that’s not as well appreciated, but spectrum auctions are just one piece of the whole story of using markets to manage the spectrum. 

Well, now to the story. So, first of all, I want to start with something that Tom will appreciate, which is Bill Niskanen. When I was at the Council of Economic Advisors on the staff, I had left academia and I needed a regular job, a real job, and I had several offers, and one of the offers was from the FCC. I spoke to Bill about which job he thought would be the best, and he recommended the FCC. He then gave me one word of advice, just like in The Graduate. Tom is old enough to know this too. It’s always bad to give cultural references to young people who don’t know them, but… 

Tom Lenard:

I’m sure everybody knows this reference.

Evan Kwerel:

Well, this reference is when Benjamin is somewhat lost in terms of what he should be doing with the rest of his life. Mr. Robinson puts his arm around Benjamin, played by Dustin Hoffman, and says, “I have one word of advice for you: plastics.” 

Well, Bill Niskanen had one word of advice for me, which was spectrum. And, he was right. I mean, as Tom said in his introduction, having a good idea is not enough. You have to take that idea and sell it and implement it. So, I think that’s the story of my career at the FCC– taking the idea that market should play a central role in managing the spectrum and selling it and getting it implemented. Despite the fact I had no formal authority, and I still have no formal authority at the FCC, to direct anyone to do anything. I think that makes it a particularly interesting story and should give people hope that they too can make a difference within the federal bureaucracy, even if they’re not the head of an agency or in a senior management position. 

Tom Lenard:

So, were there some conditions that made this, the idea of property rights and markets for spectrum, ripe at that particular time?

Scott Wallsten:

Well, and also, I think the other side of that question is what kind opposition was there to the kinds of things you wanted to do at the time?

Evan Kwerel:

There was opposition and naysaying all along, but let me first try to address Tom’s point about why the time was ripe. Let’s just start with spectrum auctions and not the broader mission of creating flexible, clearly defined property rights. I think that the key thing with auctions was the introduction of cellular telephony, which affected both the definition of property rights and the attractiveness of auctions. The two things about cellular was first that it became apparent that spectrum was becoming much more valuable. I mean, before there were TV stations and other things that sold for money, but not the kind of value that was apparent in cellular. And so the value made the notion of spectrum auctions much more attractive, and there was gold there. The second piece on the spectrum auctions had to do with the budgetary structure at the time. I think in 1990, Congress passed PayGo (pay as you go) legislation, which Tom is probably familiar with, which required all new spending to be funded. And Tom probably also knows that the idea of giving the FCC auction authority would come up every once in a while, in OMB [inaudible], but it never got anywhere.

There was a lot of opposition, particularly by Democrats who had ties to broadcasters like Congressman Dingle. They were very concerned about the slippery slope, the camel’s nose under the tent, anything that could implicate charging fees to broadcasters. But then in 1992, as you will recall, Clinton won the White House, and he was very interested in spending money, and he needed sources of revenue. So, the Clinton administration proposed spectrum auctions. Now, I was involved with the promoting spectrum auctions going back many years before, doing staff work and preparing for congressional hearings and all that. But I don’t think that’s what… I mean that may have helped set the stage, but I don’t think that’s what tipped it over the line. I think the thing that tipped it over the line was they needed money. Spectrum now became a ripe target because of cellular, and Dingle went from absolute opposition to supporting it. Congress carved out broadcasters, it was to apply only to initial licenses with mutually exclusive applications. And it carved out broadcasting and public safety and various other things.

Scott Wallsten:

Let me ask you a question before you move on. So, the need for revenues was one of the reasons that were able to get political support for it. But the legislation, I believe, pretty explicitly said, “do not maximize revenues,” right? Or they didn’t say it that way, but they said, it’s not about the revenues. Why were they willing to kind of do the right thing and maximize some version of consumer welfare rather than revenues at the time, given that revenues were one of the reasons they became in favor of it?

Evan Kwerel:

Well, because there were conflicting interests. And so, what are you going to do? You’re going to say that we’re interested in all these different concerns, but by the way, it does raise a lot of revenue. When you’re dealing with democracy there is very few pure legislation that’s passed and that has a clearly find goal when there are people that have different objectives. So, they put them all in there, and they can be contradictory. And then the agencies are supposed to figure out how to balance these things.

Tom Lenard:

Would you agree, I’m asking this as a question, would you agree with the statement that this never would’ve happened, but for the revenue implications?

Evan Kwerel:

You’re asking for a yes or no answer. And if I were a good politician, I wouldn’t give a yes or no answer. I would qualify it. But I honestly believe that all the arguments that economists made about efficiency and dealing with inefficient rent-seeking were not the key drivers. I don’t think that was the sort of compelling argument for members of Congress. The compelling argument was they needed the money, and the PayGo restricted them. They needed the money, and they needed a new source of money. And this was a new source of money. I think without that, I don’t know that I would say it never would’ve happened. That’s an extreme statement, but it certainly would’ve taken longer.

Tom Lenard:

Well, and it also, I mean, it takes some power away from politicians. I mean, whether they’re in Congress, or before that it was more of a political exercise to allocate the spectrum. And then it became, obviously the politics are not out of it, but it’s more of a market-driven exercise.

Evan Kwerel:

I think that’s true. But I also think that’s overstated. I think that you know my friend and former colleague, Tom Hazlett’s view based on political economy of why did it take 34 years from Coase’s 1959 article on the FCC until we had spectrum authority. He emphasizes those kind of considerations. I think there’s something to it, but let me just say two things about it. One, some people have asserted that it was highly political. In my observation, and talking to people, it wasn’t highly political. The prior system was what the British called beauty contests, and what is formally known in the FCC as comparative hearings. It was a process that I think was not a highly political process. They set up criteria for choosing the best party to have the license, they had narrowly defined uses for what it could be used, because if you didn’t have that, how are you going to set up criteria for who’s the best user, if you don’t know what it’s going to be used for. And I don’t think it was a highly politicized process. 

Let make another, I don’t know if it’s just a point or just an amusing anecdote. So, I will share it with you. It’s one of my favorite anecdotes, Tom Hazlett organized a conference in the 1990s on spectrum policy. And among the people that he invited was Ronald Coase, which it was just a delight to have him there. Tom presented his paper on why it took so long to introduce the spectrum auctions, and then Coase gave some comments on that and broader things. His written comments are one of the most beautiful little pieces. If you’ve never seen it, you should read it. But one of the things that Coase says, he basically dismissed the political economy argument, and Coase said, “Never underestimate the stupidity of human behavior.” So anyway, I just thought coming from Ronald Coase, it was just beautiful, especially given the Stigler version of the Coase theorem.

That was a longwinded way of saying, I don’t think the politics of being able to control spectrum licenses was a big factor in terms of opposition. I think what was a big source of opposition was the fact that there were constituents, like the broadcasters, who didn’t want to pay money for spectrum. Other licensees also saw this as cost to them, and they didn’t want it. But I think the broadcasters… broadcasting was a lot more powerful, I think that was the biggest source of opposition. And it was a very simple argument. This could lead to government fees and costs of getting new licenses, and they didn’t want that. It’s just that simple. It’s not that they were going to give favors and they’re going to give you better coverage or something if you threaten to withhold their license. I don’t think that’s the story. That’s my opinion.

Scott Wallsten:

So, what about the other part of this, which is, as you mentioned earlier, the flexibility of licenses? Because that also kind of messes with the political economy of, you know, how the individual broadcasters might feel relative to their trade association.

Evan Kwerel:

Yeah, that wasn’t really high on the radar of anybody. The opposition to flexibility came mostly, as I recall, within FCC from the engineers. There was a traditional way of doing it, and they were very concerned about interference. They also thought they knew best with these things. Let me just tie this, flexibility, back to the right conditions for this whole revolution. I think cellular was also a piece of the right conditions for flexibility because it had to do with the architecture of cellular mobile systems. Until cellular, what we had were what the British called, “apparatus licenses.” They were point licenses pretty much. What we did was we licensed a transmitter at a specific location, and with a specific tower height, with a specific power. Just think of broadcasting or think about traditional land-mobile radio, like dispatch services or the original mobile telephony.

You had a big tower, with high power, at a specific location. And it was very much, all the parameters were managed by the FCC. Well, cellular doesn’t work like that. You can’t license a cellular system by just giving them, a single tower. What it is, is a system of multiple transmitters. The system operator has to have discretion as to where they put them and they might have to move them or further densify and add them all the time. And, and it provides a service over a wide area. So, it forced the FCC, in developing service rules, to come up with an area license instead of a point license. This is the area in which the cellular operators could provide service. Now originally, they stuck as much as they could to the traditional licensing, which meant that every tower had to be licensed individually. But this changed as part of the move to flexibility, which I give John Williams a lot of credit for. John doesn’t get enough credit for all this. I was the loudest voice <laugh>, but John was really the brains behind appreciating this.

Eventually, we moved to giving cellular operators the right to put their towers wherever they want, as long as they met interference constraints. Another thing that was done at the beginning, in line with making the fewest changes they could to accommodate cellular was to specify the technical standard used by cellular operators. Originally there was no technical flexibility. The FCC just blessed and then required AMPS, the 2G standard. Part of that was because that’s how they always did things. And part of it was because they wanted interoperability, including roaming, but cellular I think was the critical factor, both in terms of auctions, because there was value there. And in terms of flexibility, because it required a rethink of licensing. Now, it wasn’t quite as simple, but let me stop because I could go on about the PCS story and, and flexibility and the fight over that. But I’ll stop because you’ve got great questions, and I could talk forever about these things.

Tom Lenard:

It sounds to me like it must have been very difficult to design the first auction, particularly. 

Evan Kwerel:

Absolutely. A word that gets thrown around a lot, but that was central to making that a success and pretty much everything else I did a success, is collaboration. You need somebody to have a vision of where you’re going and get people aboard to work towards a common end. But when you’re dealing with a complicated problem such as this, to do it well, you need a tremendous amount of collaboration. And this ties into Tom’s question, how did I do this? How was this accomplished? I think that too often in government, agencies want to do everything themselves. If it’s not invented here, it’s not going to happen. I think one of my major contributions was to collaborate with leading academics, people that were basically, not basically, they were at the top of the field, people like Paul Milgrom and Bob Wilson. There’s a reason they got a Nobel Prize. These guys are geniuses, but having geniuses out there doesn’t mean that you’re going to be able to take advantage of their insights. So, I think part of collaboration, particularly for somebody inside government, is to recognize a good idea when you see one. And that’s, as Scott and Tom know, rarer in government than you might like.

Scott Wallsten:

It’s also, it wasn’t just collaboration. I mean, it was collaboration and persistence because it took so long. And also, you collaborated with kind of the same core group of people for decades who were all willing to kind of stick to it. That’s pretty rare.

Evan Kwerel:

Yeah, that’s absolutely true. It was mutual admiration which I think drove this. Let me back up a little bit, but you can push me back on the path if I get too far off. The short answer to that is that, from among lots of different proposals, I recognized, not immediately, but I recognized, the proposal that Milgrom and Wilson had was just brilliant. There was also an early proposal by Preston McAfee, working for AirTouch, that also proposed a simultaneous multiple round auction, but it didn’t have the elegant feature of a simultaneous closing rule. And when AirTouch first started coming around trying to sell that, I wasn’t in those original meetings, they just went to the OET or whatever the engineering group was called. And I heard it from the head of the OET, and I was skeptical because I didn’t understand it.

If you look back… I’m going to talk more about your question of how it was I stuck with a horse that was a winner, but I just wanted, to first address Tom’s point about figuring out what the right design was. When the FCC got auction authority in 1993, I was, since I had been writing about it, the man with one eye in the land of the blind. I knew more about spectrum auctions than anybody else. I didn’t know that much about it <laugh>, but I knew more about it than anyone else at the FCC. In the 1985 paper that I wrote with Lex Felker we said you should auction licenses simultaneously because of the interdependence of license values, you know, the complementarities as well as the substitutability, but I didn’t know how to do that.

And so, when I had to write the auction design section of the NPRM, notice of proposed rulemaking, I proposed a sequential auction. It was a simple form of a combinatorial auction that was based on auctions that they use in estate sales, where you bid on individual pieces, and then you can have a bid for the lot as a whole. Such auctions are also used for collections of art and stamps. This design addressed the question of whether you should have a nationwide license or individual licenses. And there was a lot of thought about what order the sequence should be, but that was the best I could come up with. It never occurred to me that we could do something simultaneously. It seemed like it was a complex integer programming problem that I had no idea how to do.

And we had one year to go from legislation to implementation, and we weren’t going to solve that in one year. So, I proposed something that was simple and doable. But then both McAfee and Milgrom and Wilson, came up with these totally novel approaches of doing a simultaneous auction that wasn’t perfect in terms of allowing for package bidding, but man, it went a lot of the way, really far. Let’s say 80% in terms of the benefits. And yet the mechanics of it were quite simple because it was a series of individual rounds. 

I’m trying to answer two questions at the same time. One, Scott’s question about how I stuck with the same core group of people. And Tom’s question of how I did this, because it wasn’t like there was one brilliant proposal and everybody agreed. There were a lot of economists, and they had all sorts of different views in representing their particular client’s interests. But any case, when Milgrom came in and explained it, I was just blown away. It just was such an elegant idea, simple and elegant. And Milgrom, he’s one of these people that not only is he a brilliant theorist, but he can explain things in English. He can explain the intuition without using a lot of economics jargon. So, it’s accessible to an intelligent, but not expert audience. So I was just impressed with him.

Tom Lenard:

Did you bring him into explain it to the commissioners, and who was the chair?

Evan Kwerel:

No. Reed Hundt. Actually, there was a period before Reed Hundt was the chair, when it was Quello. So, we were sort of in a holding pattern for a long time. I mean, we weren’t holding… things had to move forward because we had one year, but what I did do…. 

Okay. So, just in terms of institutional structure, Milgrom wasn’t working for us. McAfee wasn’t working for us. Ausubel wasn’t working for us. Cramton wasn’t working for us. They were all working for interested parties, which were the pretty much the telecoms, the Bells, and Cramton was working for MCI, and I don’t remember who Ausubel was working for. 

Okay. So, let me just answer Tom’s question. And so what I did do was get the FCC to hire John McMillan, who was a collaborator with Preston McAfee. Not only was he a great economist, he was extremely personable, and wrote very well and very quickly. There’s a Journal of Economic Perspectives piece that was originally his report that he did for the FCC. He was a great interlocutor, a great person to explain these things within the Commission. So, I think that that was really important.

Tom Lenard:

So was he the only the outside economist that was actually hired by the FCC?

Evan Kwerel:

Yes. And there’s an interesting story about that, which is related to what some people say is one of my important contributions to this. When I wrote the NPRM auction design section, there were a lot of references to economic papers on auctions. And what I heard from the industry was that they got the message that we were going to be taking economics seriously. So they went out and hired all the best economists who knew about auctions, including Paul Milgrom and Bob Wilson and Peter Cramton. And then … I was talking to somebody who, “Well, you know, so and so has hired Milgrom, and this one’s hired so and so, who’s your economist?” 

And I said, “I can’t tell you. It’s under negotiation, but I’ll let you know.” Well, we didn’t have a clue. I didn’t know that all these people… the private sector were hiring these people, but it was a great thing that they did.

So, I immediately went to my boss, Bob Pepper, and said, “We’ve got to hire somebody.” And he said “Who can we get?” So I called around. Everybody was hired. So, Preston McAfee said, look, call John McMillan, he’s great. He doesn’t do a lot of consulting. You will be very happy with him. I called McMillan and he was happy to do it. And we got him. But I think the fact that there was a strong economic orientation of the NPRM, sent the signal that led, at least I’ve been told, I’m not just making this up, to the deep-pocketed interested parties in hiring the best people around. And that’s how Milgrom ended up coming in doing an ex parte presentation that blew me away and started a beautiful relationship.

Scott Wallsten:

Is this summary accurate that sort of within a year, you went from thinking that it had to be a sequential auction to everyone realizing there was going to be auctions and finding economists to hire, Milgrom coming up with this simultaneous design that you saw as so elegant, and then figuring out ways to show other people who may not know economics so well why this was a great solution, and bringing it all to an actual product within a year. I mean, I have emails that I haven’t figured out how to answer in more time than that. I mean it’s pretty remarkable.

Evan Kwerel:

Scott, I’m glad you put it that way because people outside government have no idea what an accomplishment that was. From legislation… we actually went from legislation that was signed in September of 1993, to our first auction in August of 1994. Now this is perhaps of little interest to economists, but institutionally, the challenges, and I’ll get back to the economics challenges in your specific question, but the challenges of doing a rule making, which has specific amounts of time for comments and reply comments and reconsiderations, were incredible. It is just amazing that the FCC was able to compress the process into such a short time. And the second problem was contracting. The government contracting process is completely broken, and to do it that so quickly couldn’t have been done without Jerry Vaughn who led the implementation effort and was just brilliant in making all the pieces work.

But as far as convincing people, I give Reed Hundt and people on his team like Don Gips, an enormous amount of credit.  This ties in with Tom’s question, what did it take to implement this stuff? One of the things it takes is agency leadership who are willing to take risks. It’s not only Scott’s point, that this was complicated and novel and hard to do, but it was a very risky thing to do. You needed the agency leadership to make a considered  decision to take that risk and then be all in and give absolutely total resources to do that. Reed Hundt and Don Gips were willing to do this. It wasn’t an easy lift. When I explained this brilliant, never before attempted auction design, to the head of my office he said, “I don’t think so. I don’t want us to be a beta test site.” <laugh> 

To help decide which auction design to implement, Reed Hundt sent Don Gips around to ask staff two questions. One, what is the best auction design? That’s the question he asked me. And two, can we implement this? That’s the question he asked Karen Wrege, who the FCC had brought in to help implement the auction.  I said, I absolutely thought this was the best design, and explained why. And Karen, with whom I had discussed, in general terms, how the auction would work, said she was convinced that we could implement it. And then Reed Hundt was all in, but with his eyes wide open. He committed a huge amount of resources into this, including hiring experimental economists at Cal Tech to test different designs. I actually don’t remember at what point we hired them, and when they were testing spectrum auction designs using their own funds. 

I know we later hired Vernon Smith, but I’m not sure exactly when, but we did everything possible to test our auction design in experimental settings.  To reduce the risk, the FCC also had a backup plan in case there was a software failure in the first auction. At the auction, there was a Cal Tech student, who later became an economics professor, who developed in collaboration with, John Ledyard or one of the other Cal Tech professors, a manual backup method that could be run by hand if the electronic auction system failed. 

Scott Wallsten:

Did you ever try that? I mean, I know the FCC always runs mock auctions, particularly when it’s new. Did they try a mock auction with the ledger?

Evan Kwerel:

No. No. Well, I’m just saying that there was always a plan B, but it was still a very risky thing because, the software was developed on a very compressed schedule. So, let me just back up to talk more why the first auction worked. Another contribution that I made, which is underappreciated, except by me <laughs>, was not only advocating what I thought was the best design, but finding an application that was simple enough that we could do it, but important enough that if it succeeded, it would be considered a big success, but not so big that if it failed, it wouldn’t be a disaster. I spent a lot of time asking around about possible spectrum we could auction. And I came up… with these nationwide paging licenses that we call narrowband PCS.

The beauty of it was there were 10 identical licenses. So, it seemed to me that for 10 identical licenses, we should be able to develop the software and do this thing. Well, it turned out it was harder than you would think because among other things, we had this activity rule that was complicated. Incidentally, one of my major contributions to the design of the simultaneous multiple round auction, was asking Paul Milgrom a question that led to the development of the activity rule. And that ties into the beautiful relationship that I developed with Paul Milgrom. But in any case, to answer Scott’s question, yeah, convincing FCC leadership, was critical. But it couldn’t have happened without a receptive audience.  Chairman Reed Hundt deserves a lot of credit for that. He was willing to take the risk once convinced that this was the best thing. He wanted to do the best thing. He didn’t care that it had never been tried before, but he also devoted the resources to be sure that if it was possible to succeed, it would.

Tom Lenard:

Well, let me… this has been fascinating. I think we’re running out of time. I just want to maybe close with, I don’t know, exactly how to ask this, but in the alternative, if the FCC had not gone in the direction of auctions and more generally the property rights, a market approach for spectrum, it would’ve been pretty disastrous for the wireless revolution one would think, right? 

Evan Kwerel:

Absolutely. First let me just give a narrow response. I think because of the money, spectrum auctions would’ve happened without me.  But I also think they wouldn’t have been done as well. I expect we would’ve done spectrum auctions the way that government does other auctions, which is sequentially, like offshore oil and gas auctions. And, if my immediate boss had his way, we might have had sealed bids or maybe an open outcry auction sequentially.  We might have had the first auction with sealed bids because it was it clear that the FCC was capable of putting sealed bids in a box and picking the <laugh> the highest bids, even if that was not a very good approach. So yeah, I think it would’ve been a disaster. I don’t know how much of a disaster. A question that comes up is how important is good auction design in terms of efficient allocation?

And to what extent, would secondary markets achieve a comparable result? I’ve been back and forth, with Tom Hazlet on this point. I’m still convinced that auction design really matters, particularly when there’s high interdependence, like with the broadcast incentive auction. We needed not only economists, but computer scientists, Kevin Layton Brown, figure out how to repack TV stations on the fly during the auction, so we could find the least number of channels to provide a channel for every station that didn’t want to participate in the auction. And you had to do that nationwide simultaneously. It’s a very complicated inter-programming problem. If you just had these bilateral negations in the secondary market, I don’t know how you could have had that kind of coordination.

Scott Wallsten:

So, I think, are you saying then that you think the auctions are more important than secondary markets because it’s important to set the additional conditions properly and that the auctions are there for a necessary condition for secondary markets to work?

Evan Kwerel:

Well, I think property rights are necessary conditions for the secondary markets to work.  How property rights are defined matter.  The development of overlay licenses, which was also John Williams’s idea, helped make markets work better. So yes, I think you needed a smart regulator to set smart property rights to make secondary markets work better. But I also think that a lot of transactions that you would think would happen don’t happen in the secondary markets. Auctions do two things.  First, they create a thicker market with much more liquidity.  Second, when there are complex coordination problems, like with the Broadcast Incentive Auction or when you have relatively small licenses and you need to aggregate them in efficient ways, both in frequency and geography, then I think that auctions play an enormously valuable role.

Tom Lenard:

Well, I think we could go on with this for another hour, at least I could, but I think we’re basically out of time. So, but I really appreciate your taking the time, Evan, to explain all of this stuff for us. It’s fascinating.

Scott Wallsten:

Yeah, this is really interesting stuff.

Tom Lenard:

So thank you very much.

Evan Kwerel:

Well, you’re welcome. You know, this was fun. I prepared for some of this stuff and we didn’t get to all of it. There were all sorts of things that I could have talked about, but I thought this was great. 

Scott Wallsten:

I think we need to have part two here.

Evan Kwerel:

I thought you guys asked really good questions. I liked your questions. The fact is you have thought about these issues.

Sarah Oh Lam:

I have more questions. We could do a series of the history of spectrum auctions in that time period. The counterfactual without your work in it, we would be still doing beauty contests.

Evan Kwerel:

No, no. We had lotteries, which happened without me, but that was sort of a disaster itself. That’s a whole story. But in any case,I was a bit apprehensive about this, but I just really enjoyed it. In fact, I always enjoyed these things.

Tom Lenard:

Well, I have a whole bunch of [inaudible] questions which I didn’t ask.

Evan Kwerel:

Well, we could have a Zoom.

Scott Wallsten:

Wait, wait, before we go. One question I wanted to ask is something that you implied, but I didn’t know. Twice, you said that the British called them beauty contests. They also had another word for point licenses. I forget exactly what you said. 

Even Kwerel:

Apparatus license. 

Scott Wallsten:

Yeah. Apparatus license. Did the British have more to do with the initial way we used to allocate spectrum? I mean, were they the ones who came up with the way we used to allocate spectrum? Was it the whole British thing? 

Evan Kwerel:

I mean, I don’t know.

Tom Lenard:

Did we do auctions? Were we the first in the world to do auctions?

Evan Kwerel:

You know, this question has come up, and I’m not sure we were absolutely the first. I think it was in Australia or New Zealand. I have to check the dates that they may have done an auction before us. I think it’s fair to say that we were the most innovative in the design of auctions. The simultaneous multiple round auction became the workhorse auction design the world around. And I know this is not in the interview, but there’s a point I just wanted to make about what economists call path dependency, and this ties into Sarah’s counterfactual point. 

Okay. So, one of the things that I was well aware of having been in the government was that if you have something implemented and if it’s reasonably successful, you’re going to continue doing that for a very long time.  If you just look at the auctions for the offshore oil and gas, they may still use sealed bids. And I don’t know if they ever literally put them in drop boxes, but they may have in the beginning. In any case, I was of the opinion that it was really important to get the first auction designed right, because whatever we did, if it was successful, we would keep doing that. And if we did something like a sealed bid auction, we would keep doing that as long as there wasn’t some kind of scandal. And so, because of the way bureaucracy works, it wasn’t just getting that first auction right for its own sake. It was getting that first auction right because that was going to set the precedent for how every auction after that was going to work, and it turned out to be true. 

Once software was developed for doing simultaneous multiple round auction, I tried like heck to introduce other ideas including combinatorial auctions. I was even able to get consultants hired, to work on combinatorial auctions. But there was such resistance within to doing something new. Part of the reason the FCC was able to innovate in auction design was, because for the first auction there was no auction division. Everything was new. But once you have a bureaucracy doing things, they get really good at replicating what they did before. We were able to overcome this inertia with broadcast incentive auction by going to the outside to Larry Ausubel’s firm, Power Auctions, for the implementation instead of using our in-house contractors. As a result, even though we haven’t had additional two-sided auctions, we have had clock auctions, which was the design of the forward auction. 

Okay. Well, as I said, I could go on forever.

Tom Lenard:

<Laugh> Well, this is, as I said, really fascinating and we appreciate the time you’ve taken and appreciate your contribution in public policy. It’s really remarkable for those of us who spent some time in government. It’s very impressive to do all that. So, thank you very much. 

Evan Kwerel:

Thank You. Well, I appreciate that. I appreciate the stuff you guys do. I’ve seen more of your written… more of Scott’s stuff recently. I just think you’ve have a good sense for identifying important problems and supporting the basic economics with some empirical analysis. I think that’s really useful.

Scott Wallsten:

Well, thanks. I appreciate that. Of course, I don’t have the knack for solving important problems the way you do

Tom Lenard:

Well, thank you, Evan.

Scott Wallsten:

Talk to you soon, Evan.

Evan Kwerel:

Okay. Bye. Take care.