Nobel Prizewinner Kenneth Arrow on Economic Thought and Academic Freedom

Roger Bowen

On August 16, 2005, AAUP general secretary Roger Bowen visited Stanford University to interview economist Kenneth Arrow. Arrow is probably known best for his “impossibility theorem,” which he proved in his PhD dissertation and described in his 1951 book Social Choice and Individual Values. The theorem proves that under certain assumptions about people’s preferences among options, it is always impossible to find a rule under which one option emerges as the most preferred. In 1972, Arrow was awarded the Nobel Memorial Prize in Economics with British economist John Hicks, for “pioneering contributions to general equilibrium theory and welfare theory.” Arrow was a faculty member in economics at Stanford University from 1949 to 1968, after which he joined the Harvard University economics faculty. In 1980, he returned to Stanford, where he remains. The transcript of his conversation with Roger Bowen follows.

Bowen: You’ve been a member of the AAUP for a good number of years.

Arrow: I can’t remember how many exactly. But my guess is I joined soon after I came to Stanford in 1949.

Bowen: And your doctoral dissertation was completed in the early 1950s, was it not?

Arrow: It was published in 1951. It was pretty much completed by winter 1949.

Bowen: Why did you choose the title Social Choice and Individual Value? And what was it that took you in that direction?

Arrow: Well, chance. It was Louis Pasteur who said that “chance favors the prepared mind.” I’m a great believer that events happen by chance, and this choice of topics certainly exemplifies that. I never would have dreamed it was a serious topic. I was regarded as a very good student. But I was rather less taken with myself than other people were. I knew I could learn anything, but had no inkling I could have an original idea. I didn’t seem to have any good ideas. When I returned to graduate study after the war, I went back to Columbia, and I tried to write a thesis. It was a very ambitious thesis, which was going to challenge the then-leading theorists, one of whom was John Hicks. Later, I was awarded the Nobel Prize jointly with him. I was very impressed with his book Value and Capital.

As a graduate student, I was critical. I could see the errors in his book but questioned whether I could improve on his work. There was one particular point that bothered me, which was that he essentially assumed that each firm was really owned by one person. Although a firm is a unitary object, if it is making investments, then a number of owners—stockholders—all want to maximize profits. There’s no divergence of interest in that sense. But the owners may have different ideas of what the future will bring. I asked what determines what a firm does. I’m thinking, that’s simply what the majority of shares votes. So it’s very simple: investment project A is better than investment project B if more shares vote for A than for B. I wanted to understand what “rationality” meant. People were trying to maximize something.

Bowen: What earlier influences took you in this direction?

Arrow: Bertrand Russell, and this is again another instance of chance. I was very interested in logic as an undergraduate. In fact, even as a high school student, I’d read books on logic.

In my last year of college, which was 1939–40, war broke out. There was a very famous Polish logician, Alfred Tarski, who was caught by war in the United States. He had come for a meeting at the outbreak of the war. So City College, where I was a student, hired him to fill some vacancy in the philosophy department. I knew Tarski was one of the great names. He gave two courses. One looked pretty elementary; even though I hadn’t taken any such course, I had read all this stuff. I took the more difficult of his courses. It was on relations. You know, like X is greater than Y, or X is bigger than Y, or X is similar to Y. And I particularly liked the idea of an order. You know, X is better than Y, and Y is better than Z, then X has got to be better than Z. So when I started really getting serious about economics, I realized immediately that’s what they’re talking about. This ordering concept that the logicians have is exactly what economists are talking about. It was just something I stored in my mind. I didn’t develop the idea immediately, because it didn’t seem worth thinking about.

Years later, I came back to the problem about the multi-owner firm. I thought, well, this defines what I mean by investment project A being preferred or chosen rather than B. Well, that relation had better be transitive. Now, supposing project A is better than project B, and project B is better than project C. I wanted to make sure that project A was better than project C. So in about half an hour of scribbling, I realized that it was not necessarily so. You could have a majority of the people favor A over B, a majority prefer B over C, and yet a majority prefer C to A. This logic seemed applicable to the theory of the multi-owner firm. I wondered how you’d get positive answers. And I sort of finessed it by inventing some other hypothesis. My own views became clearer in contrast to Hicks’s.

Bowen: Who else influenced the development of your theory?

Arrow: Tarski had asked me if I would proofread his book, an elementary book on logic. It was written in Polish, was translated into German, then into English. The translator was philosopher Olaf Helmer. I went to RAND for the summer of 1948. RAND had hired Helmer to ponder what warfare was going to be like in the future. He told me that one thing was bothering him: RAND was devoted to trying to apply game theory to warfare, with the players being the United States and the Soviet Union. He wondered how one could apply notions of rationality to this topic. The United States wasn’t a person, nor was the Soviet Union, but both were made up of many individuals, each with his or her own preferences. He wondered how you put this all together.

There was a paper that had developed a formal way of treating that problem. What it did was correct, but it didn’t ask the right questions; it really was a voting question. And within a week, I had an argument. Then two weeks later, I made a better argument. I thought, you know, this is interesting. This is worth a dissertation. If Olaf Helmer had not asked me the question about game theory, I would never have used my background in logic and developed my theory.

Bowen: You also had a background in mathematics.

Arrow: Well, I had a background. I was an undergraduate major in mathematics.

Bowen: And this was at City College. So your interest in corporations and aggregation of individuals’ interests, logic, and philosophy all come together in a rather fascinating way.

Arrow: Yes. At the same time, with the economic ideas and rationality. So, yes, these came together, so I wrote to my previous thesis adviser at Columbia to say that I had an entirely different thesis.

Bowen: Did you finish your doctorate at Columbia?

Arrow: Yes.

Bowen: And is your dissertation your single most important scholarly accomplishment?

Arrow: I’ve done some very important things—four or five things that I think are important. Certainly, the dissertation was the most original, but I’m not sure it was the most important.

Bowen: Let’s turn to higher education. You’ve been in higher education most of your life. You’ve helped society understand the market and how it works in terms of rational choice theory. You’ve made enormous contributions. These days, more and more, we hear about the higher education “industry,” about the bottom-line mentality. We have talked about the hollowing out of the professoriate, as there are fewer and fewer full-time, continuing faculty and tenure-track lines and more reliance on part-time faculty—academic gypsies who go from campus to campus and sell their labor for not much, given the number of years they’ve invested in earning their PhDs. This may be “rational” in economic terms, but are we in trouble because the higher education industry is working according to market principles?

Arrow: You know, it has bothered me. But there are several levels here. For example, universities themselves try to make money by licensing the products of research. That bothers me. But it would not necessarily be good for the professor to have complete control over his research findings either, which may be more distorting than the university’s having it. It’s hard to see when value is created by the university or by somebody outside the university. It’s hard to see that they shouldn’t get some part of it. And these labs are very costly. The costs of production, to use economic language, are going up. The laboratory is the big thing. I see what happens in the buildings of this campus. You know, its expansion is largely for scientific laboratories. You’ve got to pay for that somehow.

We try to get government grants. But the government is limited in meeting demand. The faculty want the grants. So we get other forms. Part of it consists of license fees; part of it consists of research contracts with interested parties. These come with strings attached, sometimes restricting what can be made public.

Even a small-scale research area like economics, for example, that is devoted to economic policy research, relies on federal funds, albeit the amount is trivial compared with what the sciences receive. We get grants from foundations, but these foundations frequently are created by business institutions. And they have an ideological agenda.

On the whole, we at Stanford have had leaders who have fended off these problems. But it’s easy to imagine ethical problems arising when accepting grants from industrial corporations. The potential for conflict of interest exists, although I’ve not really seen any good examples personally.

But it seems to me the potential exists.

Bowen: There was a study done recently by an economist at Santa Clara University, Daniel Klein, showing disproportionate numbers of registered Democrats versus registered Republicans in various departments at the University of California, Berkeley, and Stanford. [The study’s findings are available at http://www.ratio.se/pdf/wp/dk_aw_voter.pdf.] He has concluded that this kind of ideological imbalance has a negative impact on the education of students. He implies that there is a temptation to hire one’s own. Conservative activist David Horowitz has made much the same kind of statement, saying that faculty are preaching rather than teaching. And why? Because there’s a gross imbalance between liberals and conservatives in the professoriate. Effectively, they are calling for government regulation of the academy. Do you worry about calls for legislation at the state level to correct this situation? Does this worry you as an economist or as a professor?

Arrow: You know, it certainly does worry me. It would worry me a lot if legislation passed. There was a concern at one time that there would be repression of the left. And now there are concerns that the left is taking over. It’s hard for me to judge, of course, but I must say that my department contains a number of Republicans. And they were appointed by a democratic group, whose members said these guys are good, and we’ve got to hire them. And so far, I have not seen it work the other way, but I’m a little concerned about where it could swing. In this case, the criticism seems to be just wrong, because I think the departments hire on the basis of merit. And I think it’s nonsense to say that we’re discriminating against Republicans. We hire them all the time. On the other hand, there was a department here that until the 1960s would not appoint a Jew. And, finally, the university did interfere, you see, in that case. The dean took over the department. He took away the power to appoint from the department and changed its composition in three or four years. In fact, I was amazed how rapidly he was able to turn things around to strengthen an already very good department. To defend the autonomy of that department would not have been something I would have been very happy to do.

Bowen: The economics department at the University of Chicago has had a reputation for many years for being quite conservative. Do you think that’s the exception that proves the rule that you hire, as you said earlier, on the basis of merit, not on the basis of party identification of ideology?

Arrow: There are people in that department who are not conservative. It’s a very good department. Most of the conservatives are really quite outstanding. I think they flock together. I don’t think it’s entirely the case where you pick your own kind. I don’t think the economics department here is reproducing itself. They’re different politically and methodologically. I think methodological problems have been bigger more often than political issues. I do not believ the university, the central administration, should be totally unconcerned about appointments. I used to believe that the department had to be completely autonomous. It took me a couple of years to realize that that was not right.

I can remember an instance at Chicago in which there was an incident involving a professor of economics who was sort of a village atheist type. He was a very good economist, but a little eccentric. He believed that religion was one of the big oppressive things in this world. This fellow saw a priest in class. He went and gave his whole lecture on the evils of the Catholic church. The next time the priest came, he gives another lecture. The priest finally quit the class, and the professor said that he could finally go on with the course.

Well, you know, the priest went to the chair of the department who had a very good record on academic freedom at the university. And the chair said it was a question of academic freedom. He wouldn’t interfere with this professor.

Bowen: The AAUP would disagree. Our point, which the 1940 Statement of Principles on Academic Freedom and Tenure says very clearly, is that you’re not to introduce into your classroom subject matter that is extraneous or has no relation to the discipline. So if I’m a political scientist, and I spend all my time talking about the gospel, one religion or the other, that is irresponsible on my part. And I think it is also a violation of academic freedom, because with freedom comes the responsibility to stay true to your discipline. Speak from your area of expertise, not outside of it, and I think you’re on safe grounds.

Arrow: The name of the course was “The History of Economic Thought.” The Catholic church was relevant. But the professor went way beyond. Yet it was not completely irrelevant, because of Catholicism’s role in the emergence of economic doctrine.

Bowen: We have students today across America who are complaining because they think the classroom is too political. That happens especially during an election year. But when faculty disclose their voting preference in the classroom, students complain. I have no trouble at all if faculty members in political science talk about American politics, the role of the presidency in American history, or the sociology of power. That kind of discussion can happen so long as it’s well managed by the faculty member in a professional way. The intent should never be to offend students. Rather it should be to educate them.

Arrow: That’s right. Well, suppose you’re a professor and the student disagrees with your views or raises another point. And he says the stupidest thing possible, and the professor replies that the student doesn’t really belong in this class. Now, it’s clear that’s bad pedagogy. What you do about such a situation is another question. But it seems to me there are regulations and procedures for dealing with such an event.

Bowen: Do you believe that the climate of academic freedom permits scholars like you to think and teach creatively without worrying about consequences? Do we have a marketplace of ideas, where the very best ideas are embraced and celebrated, the worst ideas are cast aside?

Arrow: This is absolutely essential from an instrumental point of view, and mainly for the development of knowledge. You know, John Stuart Mill said, “Error is also an important contributor to knowledge.” I’ve not seen in recent years, at least at Stanford, any real threats to academic freedom. But I think it’s essential to maintain it. We had these problems in the 1950s, which in the end worked out. But they were serious issues of repression. I think at the moment the climate is pretty good, but one has to keep vigilant about academic freedom. It never is without danger.

Bowen: What’s your next book? Are you working on something now?

Arrow: Well, I’m working on articles. I’m not really a big book writer. I’m working on environmental questions— the proper measurement of environmental problems and how this modifies our measures of economic well-being. There’s a whole group of us who are working together on that. I also have a revived interest in health matters, especially in the economics of anti-malarial drugs.

Bowen: I want to thank you on behalf of the AAUP for being such a loyal member for so long. And for being, I think, one of eighteen Nobel Prizewinners we have in the membership—but I believe the only Nobel Prizewinner in economics. We’re honored to have you as a member and as a stalwart supporter of academic freedom.

Arrow: Well, I appreciate your saying that. And I must say the American Association of University Professors has been really one of the great institutions in this country.

Kenneth Arrow, an economist at Stanford University, was awarded the Nobel Memorial Prize in Economics with British economist John Hicks, for “pioneering contributions to general equilibrium theory and welfare theory.”

 Printer-Friendly Page