An Interview with Michael Spence
Steve Lin
Issue date: 10/15/01 Section: News
|
October 11th – Barely a day after receiving word that he had been awarded the 2001 Nobel Memorial Prize in Economics Sciences, Michael Spence is back in Palo Alto, weathering a media onslaught. Over the past 24 hours, he has talked 12 hours straight, flown halfway across the Pacific, and gotten an hour-and-a-half’s worth of sleep.
Yet pulling up to the steps of the Knight Building on his motorcycle for his latest interview, the ex-GSB Dean bears the undeniable look of excitement and exhilaration on his face.
Reporter: How does it feel to be mentioned in the same breath as William Sharpe and Myron Scholes—all three of you Nobel Prize winners from Stanford GSB?
Dean Spence: It’s very humbling. It’s a great honor. And it’s a very hard thing to describe. I don’t think any one sets out to win the Nobel prize, but it just sort of happens, or if you’re lucky it happens. And I think there is a lot of luck. It’s a huge honor to be included in a group with those two [Sharpe and Scholes] and the two that got the Nobel Prize with me, George Akerlof and Joseph Stiglitz, who are very, very impressive.
Reporter: Do you think this feeling will sink in after a month or so?
Dean Spence: I think so. The other thing is that you don’t know what your life will be like for a while. They say it will change for a little bit.
Reporter: How much has your life changed lately?
Dean Spence: Well, I’ll tell you what happened yesterday. The people from the Royal Swedish Academy of Sciences phoned me here in California, but I was in Hawaii at the time. So Barbara Buell, who works here in GSB News and Publications, saw it on the website, knew where I was, and phoned me at 3:30 am. Later, a colleague at Oak Hill Capital phoned me and told me, “You better hit the deck running because you’re going to be talking for the next twelve hours.” Sure enough, I was on the phone. And a few Hawaii-based interviewers and camera folks came over. So I talked steadily for 12 hours. Then I got dressed, got on a plane, and flew here overnight. I got about an hour and a half sleep and have been talking ever since.
Reporter: Can you explain your Nobel Prize winning research in the context of E200 [the core MBA microeconomics course at Stanford GSB]?
Dean Spence: (Laughs) Yeah, right. This all got started in the 1970s. Basically, when you learn microeconomics for the first time, the implicit assumption is that everyone knows what they need to know and they know all the same stuff—what the products are, what the characteristics are, where the markets are, etc. There are some markets that come close to satisfying all those conditions. Go down to the supermarket: You’re staring at a head of lettuce—there ain’t a hell of a lot that you don’t know. But there are some markets where the sellers have more information and the products are more complicated. They have attributes and they vary one from another.
It wasn’t as if we discovered this fact—everybody knew it. All we set out to do was to build an addition to economic theory that captures those markets. And there are a lot of them. Most financial markets have characteristics like that [asymmetric information]. One of the most common markets is labor markets. This also happens a lot in health insurance markets, where the buyers know a lot more than the sellers do, such as their personal health. So really bad things can happen in this market. When the buyers can’t distinguish among the products, then the market price reflects the average quality. So when a buyer has a product with better quality, he says, “I’m not going to sell it for that!”
This example was the famous [George Akerlof] paper “The Market for Lemons” in the market for used cars: The guys with good used cars say, “I’ve got the high quality car— I’m not selling it. I’m going to keep it.” So that drops the average quality and the [market] price goes down. The next guy in the spectrum says, “I’m even worse off.” So he exits too. In the extreme case, a market like that can collapse and unravel from the top down, until the only thing left are lemons. That process of the high quality products exiting, at least in insurance markets, is now called adverse selection. It is now very well understood.
So then I came along. I thought Akerlof’s paper was brilliant. I read it when I was a graduate student. I thought that if the people with high quality products can signal it to buyers at reasonable costs, then they will certainly do it. Then the price will reflect the quality of their products. So that led me to signaling, which is what I’m know for.
You can’t just say it [signal a higher quality product], since the low quality people will just imitate the signal, since they don’t want the signal to be distinguished. So that doesn’t do it—those things don’t carry information in equilibrium. So the question is: what are the characteristics of things [signals] that survive in the market? The answer is: the signal has to be something visible to buyers that is less costly to undertake the higher the quality of the product. So what you want is for the high quality people to afford it [to signal] and for the low quality people not to be able to.
For example, warranty: If I have a very high quality, low-breakdown product, I can offer a very elaborate warranty and it won’t ruin me financially. The low quality people can’t do that. So those signals persist and don’t get destroyed by imitation. The reason they self-persist is that it is more expensive for the low quality product [to offer the same signal].
And that’s a general principle: the things that survive in equilibrium and carry information—we call them signals—are products with the property that the costs are negatively correlated with the quality that the buyers care about, so that the suppliers who are sending the signal invest in more of whatever the signal is the higher the quality [of the product].
And that’s how education gets to be. Generally, there are classes of jobs that you associate with being smart, like software engineers or management solutions. But you can’t see that. I could look at you until hell freezes over and I wouldn’t know whether you had those characteristics at all. So I would have to look at other things such as your job history and how many times you changed jobs. I can look at references and I can look at what kind of educational institutions you went into and how well you did.
Education turns out to be one of the most common signals in the economy. But it’s not just a signal, as people invest in education because it changes their productivity— adding to the human capital component of education. It also distinguishes you. If you went to Stanford or Harvard, and I went to some place no one’s ever heard of, that will show up in the job offers you get. And the reason that education survives as a market signal is that experience confirms that there is a positive correlation between those snazzy institutions, your level of education, and your capabilities as observed. So the next guy comes along, and they’re thinking, “This guy went to the same school, he must also be capable.”
This is basically a rational expectations market. The experience of a buyer in the market has to match the expectations they have. If there’s a mismatch, they will change their expectations. So when you build a model that is in equilibrium, there’s a return. People will invest in signals given the return, and the return has to match the inflowing data as the market goes on. So it is not completely different from the rational expectations model in macroeconomics. The expectations have to be right, so that the return to the seller—or what the buyer will pay--as function of the level of the signal must be in equilibrium. They are first-order differential equations. What’s interesting about that is that the equations have multiple solutions, so the markets can have multiple equilibrium.
Joe Stiglitz was doing work on screening mechanisms. Instead of signaling to distinguish yourself, the buyer sides set up screening mechanisms.
Reporter: To what extent is the value of our education at the GSB derived from the signaling effect? When I step out into industry I’m going to have gained some value from classroom education, the network of contacts—but I’m also going to gain some value solely from the fact that there’s a Stanford MBA attached to my resume…
Dean Spence: And that there’s an admissions process at the GSB.
Reporter: Right.
Dean Spence: Well not quite. The signal could be false. But, in this case the screening [admissions] process is real. First, it’s an empirical question—I’m not sure anybody knows the answer, and second, it’s a hard empirical question to answer. Like any joint product, if you require two things to get the output and you take away either one, you lose the output
What we have at Stanford and other top business schools is very good education and very smart people. You put them together and you have very effective people going out the other end. People keep wanting to ask, “If you take away the education or if you took away the smarts, what would happen?” The answer is, probably in both cases, if you took away either one you’d get a much worse outcome. So it’s really the combination. If it turns out that over 10 years that employees find out that Stanford and Harvard Business school graduates are no different than anyone else, than that’ll be the end of that as a signal.
So the truth is that it [your question] is empirically complex to measure because there is an identification problem and things are highly correlated. It can be done, but you’d need an experimental method. If you’re really smart we’d have to send you to a lower-branded school but then you’d say, “I’m not doing that!”
Reporter: Have you contacted your colleagues since you’ve gotten back?
Dean Spence: I had a meeting yesterday. They phoned me in Hawaii. So we chatted briefly—all 45 of us, and they clapped. I’m coming over tomorrow afternoon for an event, so it’ll be the first time I see a bunch of them.
Reporter: So their reaction to all this is…
Dean Spence: They seemed thrilled for me. I have 400 hundred some-odd emails in my inbox from all over the world and all my colleagues here. I’m going to need to talk to Bill Sharpe and Myron Scholes, since they’ve been through all this and they’ll have a lot say—maybe even how to survive the next week. I’ll talk to their spouses, as you have to buy a lot of fancy clothes for this. These [Nobel ceremonies] are really big deals. Very formal—white tie and tails. They do all this in Stockholm in Sweden. I don’t know a lot about it yet, but the actual awards ceremony is extremely fancy: dazzling ball gowns, white ties. All the Swedish royalty is there apparently.
Reporter: Have you conferred with Joe Stiglitz and George Akerlof yet?
Dean Spence: Oh no, I haven’t talked to George and Joe. I know them very well. Joe and I worked together here in the Economics department in Stanford. And George I’ve known for years. We [three] used to talk all the time. I’d read their papers and we’d go to seminars together. I’ll probably give them a call tonight when the dust dies down. (Laughs.)
Reporter: When you guys were working on this, did you have any idea as to its significance?
Dean Spence: I guess so. I thought, and I think George and Joe thought, that the information characteristics of the markets were genuinely important—that they needed to become a part of the ordinary economic theory and not as an exceptional case. The more I worked on this the more I realized it was ubiquitous inside organizations and inside markets. What I really thought was that if enough young, really smart people—smarter than I—in 10 years picked it up and said, “This interesting. We can do more with this,” it would be embedded into economics. And I think that’s what happened. It’s partly luck and partly because it needed to be done.
Reporter: Do you see some of your work integrating its way into GSB courses, either as part of the core or as an elective?
Dean Spence: Oh, yeah. Bits and pieces of this are now taught as a regular part of economic theory. This is going to appear everywhere. Agency theory is essentially the same model. It’ll appear in courses such as organizational behavior and incentives. It’s a screening model. It’ll structure a schedule to cause people who are different to sort themselves out into different buckets. The Internet is not, if you look at it, just a band of information delivered at high speeds. It’s a lot about transaction costs. These [subjects] are also about transaction costs. So I would say it [my work] is here to stay, but it’s not going to appear as a separate subject. It’ll appear in Marketing, OB, wherever it belongs in a slightly different form.
Reporter: Speaking of the Internet I’ve read that your theory helps explain the burst of the high-tech bubble.
Dean Spence: Yeah, somebody thought that was true. I’m not sure that’s actually true. The part of it that’s probably true is that this hi-tech bubble was caused by the potential arrival of a revolutionary set of circumstances. The Internet infrastructure put us into a range where we didn’t have data with relevance, so we were unconstrained in forming expectations by any data. People were free to make up virtually whatever they wanted, at least in the initial stages. And they sort of suspended the ordinary rules and said, “I don’t know how these people are going to make any money, but anyway they are.” And then the data started to roll in. There actually was data to confront the beliefs—and they didn’t match. That’s what started the end and the crash down towards more realistic valuations. It doesn’t change the fact that the Internet is probably revolutionary with respects to economies and markets and a lot of other things. But it was a period in which when there was literally a vacuum with respect to data and information to rely on in order to understand what the business model would be and where the revenues would come from. So people said, “Well, we’ve never been here before. We’ll just make up whatever.” (Laughs.) I’m being a little facetious, but that’s kind of what’s been going on.
Reporter: Do you see any other research in the pipeline at the GSB that might be recognized in the same manner as yours?
Dean Spence: There’s a lot. I think the work that Roberts and Milgram and others did on organizational design and incentives is being built on and will very likely at some point be recognized. You never know what’s going to show up. There’s game theory, which hasn’t really been fully recognized in the Nobel process yet. Game theory has many, many branches, parts, and applications. I don’t know if the networking parts of organizational theory are sufficiently far from economics that they wouldn’t count [in the Nobel process], but I think there is some very important research being done there by people like Joel Podolny. There’s quite a lot of interesting work on human resources—Eddie Lazear’s personnel management. So who knows? It’s hard to tell. Nobody I know ever set out to win a Nobel Prize. They set out to enjoy teaching and research and being with smart students. And if it works out, it works out.
Reporter: What are your future plans with the GSB? You recently co-developed and co-taught a course.
Dean Spence: Yeah, and then I took a year off. Now I’m starting the third year [as ex-Dean] and I haven’t really figured it out. I’m retired, but all that means is that I tend to be part time. I’m not coming back fulltime for the simple reason that it’s not appropriate for a former dean to be hanging around school all the time in the face of his or her successor. And I’m a partner in Oak Hill Venture Partners. I have this interest in the Internet and the informational structure of the economy. It’s sort of like revisiting all of this stuff now that the world’s changed a fair amount. And I can imagine finding a teaching pattern that was good for the students and at the same time didn’t pin me down too much. So we’ll see. I’m going to kind of experiment. I’m absolutely not leaving the area or losing interest in the school—I’ll just do it at some distance
Reporter: You three [Spence, Stiglitz, and Akerlof] just won a million dollars. What are you going to do with it?
Dean Spence: (Laughs.) Well it depends, but the government is going get take some of it. We’re going to spend some of it getting over to Sweden. I’m going to take the kids because I think it’s one of those once-in-a-lifetime opportunities to see all this. This time [the 2001 Nobel awards ceremony] is the 100th anniversary, so they have invited all the living Nobel laureates in all the fields—physics, medicine, etc.— to come back for this 100th anniversary. So it’ll be quite a party. It’s in early December. I haven’t got the letter yet, so I don’t know what it’s going to be like.
Reporter: On behalf of all of the students at the GSB, congratulations on your award and best of luck in the coming weeks.
Dean Spence: Thank you. It’s a great honor.