Interview with Nancy L. Stokey

When Nancy Stokey and her University of Chicago colleague and partner, Nobel Laureate Robert E. Lucas Jr., chose to spend a semester's sabbatical at the Minneapolis Fed this fall, staff and academic visitors alike were elated. It's a rare opportunity for economists to spend time with two such prominent economics thinkers in a relaxed, unstructured setting.

Arthur Rolnick, the Minneapolis Fed's director of Research, took this opportunity to discuss economic issues with Stokey.

In the following interview they talk about time-consistency theory, NAFTA, the new economy, women in economics and more.

What the Fed would like to do is anticipate inflation that might break out three months or six months from now, but it's hard to do that. And it's risky to pretend that you can do it when you can't.

ROLNICK: Fed Chairman Alan Greenspan (and former Fed Chairman Paul Volcker in a Region interview) made the case that the goal of price stability means that, in practice, the Fed will try to achieve an inflation rate that is low enough so that it doesn't change business or investment decisions. However, according to most models of money, the optimal policy is a slight rate of deflation. Are Greenspan and Volcker giving up too much by aiming too high?

STOKEY: I agree with Greenspan and Volcker. I think most of the gains come from having a low and very stable rate of inflation. While there might possibly be some incremental gains from making inflation exactly zero or slightly negative, they are probably very small, and there may be other risks from an extremely tight monetary policy. A predictable, stable environment allows people to make good investment decisions, good business decisions, so a low rate of inflation and stability are the most important goals for the Fed.

ROLNICK: Many at the Fed would argue that one of the reasons the U.S. economy has done so well over the last 18 years is that we brought about price stability. The Fed has brought inflation down from double-digit rates to well under 3 percent. But the recent experience in Japan appears to be a counterexample. Japan has had a low rate of inflation over the last 10 years, yet over this period their economy has stagnated. What lessons should we draw from these two experiences?

STOKEY: There are a lot of things that can go wrong with an economy, and some of them can't be fixed by monetary policy. What monetary policy is good for is controlling the rate of inflation, and a central banker should be content if he does his piece of the job right and keeps the inflation rate low. There may be other problems-with fiscal policy, with regulatory policy or with other things in the economic environment-that monetary policy just can't do much to correct. And central bankers can easily do more harm than good if they try to solve problems that monetary policy really isn't an effective tool for attacking.

ROLNICK: If we agree that a central bank's job is to keep inflation low, a key question needs to be addressed. Over what time period should the Fed aim to achieve this goal? Should it only worry about inflation over very long horizons, or do year-over-year, or quarter-over-quarter or even month-over-month changes matter?

STOKEY: Well, I suppose over a three- to six-month period if you see inflation picking up you have to start asking if you think that trend is going to continue. You probably know better than I do that trying to look into the future is very difficult. What the Fed would like to do is anticipate inflation that might break out three months or six months from now, but it's hard to do that. And it's risky to pretend that you can do it when you can't.

ROLNICK: Given that the outlook is so uncertain, how anticipatory should the Fed try to be? Maybe the Fed would be better off just waiting until it saw inflation before responding.

STOKEY: I guess the experience of the last decade makes that view look pretty good. There have been a number of people over the last 10 years who have said at various times that the expansion couldn't continue and that the Fed should start tightening. And in retrospect, tightening would have been a blunder. I think we honestly don't know at that level of refinement what the best policy is. Each side has its risk. Hitting the brakes too soon can derail an economy that is expanding and might continue to expand.

ROLNICK: And waiting too long has its risks?

STOKEY: That has its risks, too.

ROLNICK: That gets me into the next question-about the nonaccelerating inflation rate of unemployment (NAIRU), or the rate of unemployment below which inflation will begin to accelerate. To some extent the NAIRU has been a guide to monetary policy; the concept is built into most macroeconomic models. Yet, there are some who argue that the NAIRU has little, if any, theoretical support and that the empirical support is fairly weak. Thoughts on this concept?

STOKEY: I think it's not really such a good one. And I think that the empirical evidence points in that direction pretty strongly. In the 1960s people would have said the NAIRU was around 4 percent, then it crept up to around 8 percent, leveled off and remained there for a while. Now it's crept down, and I guess it's around 4 percent again. A number that is as variable as that over a few decades is probably not a good number to build policy around. It doesn't seem to be a stable number.

ROLNICK: But a similar criticism can be levied against monetarists: Which measure of the money supply should we look at and over what time horizon? It appears that the growth in the money supply, by any measure, over anything less than five years is a poor predictor of inflation.

STOKEY: It is always a question if you're going to follow a money supply rule, which monetary aggregate do you look at and how much smoothing do you do. As you suggest, over the short run the various aggregates may behave quite differently from one another. That's a problem. But something the monetarist's viewpoint has going for it is that there's a lot of evidence that over long periods the price level is going to move roughly one for one with any reasonable definition of the money supply. So in terms of the monetarist's rule as a guide to policymaking, the evidence is that over the long run you would not go very far wrong. You might make some short-run mistakes, but short-run mistakes are a problem for any type of policy rules.

ROLNICK: In other words, what the monetarist's view has going for it is a fairly stable long-run relationship between changes in the growth of money and the rate of inflation. The relationship between NAIRU and inflation has a lot of noise in the short run, and it doesn't even pretend to be a long-run relationship.

STOKEY: I don't know if anyone's defined what the NAIRU is for a very long time period.

ROLNICK: I think supporters of NAIRU would say the relationship disappears in the long run. Supposedly, if it has any value, it's in the short run.

STOKEY: The other thing I would say about the NAIRU is that it's completely a construct of economists' imaginations. It's not something you can go out and measure. By contrast, the various monetary aggregates, whether you like them or not, are well-defined objects. So we can agree about what they are, even if we disagree about which ones we should pay more attention to. The NAIRU is just a number that somebody makes up, and consequently it's a very slippery object.

ROLNICK: You've written quite a bit on the importance of consistency in policymaking, so-called time-consistent policies. This work seems to suggest that the best monetary policy is the most predictable policy. If that's true, why not make monetary policy as predictable as possible by following a Friedman-type rule of setting the money supply and putting it on cruise control?

STOKEY: Let me start by saying that probably wouldn't be a terrible policy. If you look around the world and over time, a lot of central bankers have done much worse than that, so it would be an improvement over a lot of policies that have actually been implemented. But the question is, given the information that a central banker has in front of him, can he do better than that? The debate now, at least among academics and I suppose among bankers as well, is whether you should pay more attention to interest rates or monetary aggregates. Obviously, bankers look at both, so the issue is how much weight do you put on the two? And there I think we don't know very much. In terms of what economists can offer by way of theoretical reasoning, in many simple models it doesn't make any difference, the two are equivalent.

In more complicated models two issues arise. The first is: What kind of shocks do we think the economy is going to have to cope with? A rule based on monetary aggregates lets you better cope with some types of shocks and an interest rate rule with other types. So one issue is the nature of the shocks one expects to face. Another issue is transparency. Interest rate rules have the advantage that it's easier for the public to monitor what the central bank is doing and keep it accountable.

ROLNICK: Let me go back to the importance of credibility and time-consistent policies. What exactly does that mean to the Fed? Does that mean that Congress should pass a law that says this is the rule that the Fed should follow? Or does it mean that the Fed should announce and make public its rule, make it more explicit?

STOKEY: That's a hard question. One way to achieve time consistency would be to make the Fed subject to strict operating rules, with little or no latitude to deviate from them. A regime of that sort has some benefits, and might be useful in some circumstances. For example, in an economy where inflation has been very high, and the government wants to bring it down, a regime that gives the central bank very little latitude might be quite useful. This is more or less what Argentina has done, by creating a currency board.

Now the question is: Can you mimic the positive features of that system, the discipline it imposes, in a system where the Fed, in fact, has a lot of latitude and can occasionally make use of that latitude? It seems that you can go quite a long way toward it with a system where the public can monitor what the Fed does, and as a consequence the Fed has an incentive to develop a reputation for conducting policy in a disciplined way. The key here is that people must be able to monitor what the Fed is doing, whether it is targeting interest rates or targeting the money growth rate, and they must be able to tell if the Fed is holding to its announced target in a consistent way.

ROLNICK: By consistent, do you also mean predictable?

STOKEY: Yes. The important thing is for the central bank to develop a reputation for refusing to use monetary policy to try to achieve other ends besides price stability. The potential danger in the second system arises when a central bank comes under pressure—at least for a short period of time—to ignore its mandate to keep prices stable and to do something else with monetary policy. The discipline comes if people have an expectation that the Fed will resist such pressure and stick with its mandate to focus on price stability, and the Fed has an incentive not to let them down. For this mechanism to work, however, the public has to be able to monitor the central bank, to identify situations where it is deviating from its mandated role.

ROLNICK: In light of that, do you think it is a good idea for the Fed to announce its decision right after each FOMC [Federal Open Market Committee] meeting?

STOKEY: Yes, absolutely. It's very hard for me to see any point in the Fed being secretive about its goals or what it's doing.

ROLNICK: Having said that, there are some who would push the Fed even further, to require the Fed to televise the FOMC meetings so the public could see how the discussions develop and how decisions are made. Should we push it that far?

STOKEY: No, I don't think so. Televising the meetings would hamper the discussion a great deal and that would be unfortunate. It's much easier for people to have a useful, serious exchange of ideas if the TV cameras aren't there.

ROLNICK: Getting back to credibility and accountability: Suppose the inflation rate started to go up, say to 5 percent. How would you discipline the central bankers in this case?

STOKEY: They're appointed for long terms, so the threat of not being reappointed is rather weak. The discipline of public pressure is more helpful. Suppose that the central banker loses credibility with the public if he deviates from his announced target, and suppose that the central banker realizes this is the case. He also knows that it is more difficult to conduct monetary policy in an environment where the public has lost faith in its central banker. Then the banker realizes that in the longer run he will pay for his lapse—he will be making his own job more difficult down the road if he deviates from his announced target.

ROLNICK: Some countries have contingent contracts with their central bankers and they can be fired if they do not achieve their inflation targets. Is that going too far?

STOKEY: That's an extreme measure and presumably would be carried out only under extreme circumstances. But under extreme circumstances it might be a good move. If the central banker wants to do well in his job, he needs the trust of the business community. If he loses that trust, if people feel the economic environment is more uncertain because of doubts about how monetary policy will be conducted, there will be unfortunate consequences. Firms will delay projects, decline to undertake investments, make the wrong investments and so on.

ROLNICK: Going back to another Region interview, this one with Professor Arnold Harberger who, when asked if the Chicago School still existed, said that "policy economics is very little represented at Chicago." He does add that you and Bob Lucas, along with Lars Hansen and a few others, make up a macroeconomic center. Do you think a definable Chicago School still exists and does it matter that a university economics department has such a philosophical identity?

STOKEY: No, I don't think there is a distinct Chicago School anymore. The Chicago School was really George Stigler and Milton Friedman. Both were very forceful and very articulate proponents of a conservative point of view on government policy at a time when that point of view was not so much in favor. They were exponents of the strengths of the free market at a time when government intervention was a very popular solution for practically every problem.

But in some sense those arguments have carried the day, and public views have changed. Now people are much more aware of the limits of what government can do and the extent to which free markets can effectively satisfy needs and wants. We've seen this change in viewpoint pretty clearly in the United States. After the fall of the Soviet Union and Eastern European communist states, one newspaper and magazine article after another appeared praising the success of the free market system. And some of the Western European countries that had rather heavy-handed government intervention have also started to back off that stance. So I think the rest of the world has moved much closer to the Chicago School point of view, and as a result Chicago is not so different from the rest of the world anymore. We are all Chicagoans now.

ROLNICK: You mentioned how free markets have won the day. I would argue that in one area it didn't, one area where Milton Friedman may have had it wrong. He argued that markets should determine exchange rates. Some academics, Professors Neil Wallace and John Kareken in particular, have questioned whether markets can price fiat monies. Indeed, over the last 28 years or so, under the floating exchange rate system, there's been an incredible amount of volatility in the markets for foreign exchange. Some of the best research done on these markets suggests that changes in fundamentals cannot explain exchange rate movements. It's just not in the data. Should we rethink the position that fiat monies should be priced in unfettered markets, and instead be priced by governments?

STOKEY: That's a hard one. As you say, we don't know very much about exchange rates. Pegged exchange rates have problems as well, so it's not obvious that a floating rate system is worse than one where rates are pegged.

ROLNICK: What about a credible peg?

STOKEY: I don't know if there is such a thing as a credible peg. Usually pegged has meant pegged for a while, followed eventually by a big adjustment.

ROLNICK: We peg currency prices in the United States. We have 12 different currencies, one for each district. We have a credible peg. I don't think you worry about whether you're using a Twelfth District Federal Reserve note or a Ninth District Federal Reserve note. I'd be surprised if you did. So there is a credible peg.

STOKEY: I'm not even sure where to look on the note to find out which district it comes from.

ROLNICK: Are you saying that it's just difficult in many cases to get credible policy?

STOKEY: Yes. For example, I think the reason the Europeans have adopted a common currency, the euro, is exactly because pegging the individual currencies to each other was not credible.

ROLNICK: So having a single currency makes the whole thing credible.

STOKEY: It certainly makes it more credible, by making it more difficult, more costly, to reverse. Once you have everyone using a common currency, it's quite costly and disruptive for one country to bail out and start issuing its own national currency again.

ROLNICK: Well, maybe. But why? Italy or one of these countries could decide one day to start up the printing presses. How difficult is that?

STOKEY: They could do it, but it would be much more difficult to do it after they've adopted the euro. If they have been using the Italian lira all along, and simply pegging to the Deutsche mark at such and such a rate, it requires just a stroke of the pen if they decide to change that rate. Reintroducing the lira would be much harder. Perhaps that is why it is easy to keep the various Fed district notes "pegged" to each other. As they stand, people can't tell them apart—at least not easily. So to do anything else, one of the district banks would have to invent a new currency, maybe here it could be the Minno, and peg it at a different rate. That would be hard to do.

ROLNICK: While I've got you on international topics, let's switch to NAFTA. It's been in effect for seven years. What criteria should we use to judge the success or failure of such agreements and, based on those criteria, is there any way we can say NAFTA is succeeding?

STOKEY: Although there are direct benefits to all the parties involved, I would look first at the Mexican economy. The United States is so much bigger and so much richer than Mexico, that opening up trade with them is not going to have a big direct impact on the U.S. economy. And in addition, one of the main benefits of NAFTA for the United States is to strengthen the domestic economy of Mexico. One problem the United States has had perennially is illegal immigration from Mexico, and to the extent we can do other things to make it more attractive for Mexicans to stay home is going to solve part of our problem of illegal immigrants. So I would first measure success in terms of better performance in Mexico.

I just looked at some numbers this morning [in mid-October], and they seem to be doing well. In the last four or five years the growth rate for GDP has been very good. And from the reports in the press, NAFTA does seem to have been successful in stimulating trade. And the booming U.S. economy must have been particularly good for Mexico under this agreement. Higher U.S. demand for manufactured goods produced in Mexico has to be good for them.

ROLNICK: But the media often portrays this international trade relationship as a zero sum game, that is, if Mexico is doing well, it must be because it is taking jobs from the United States. You're suggesting that we judge NAFTA's success by how well some other economy is doing. That way of assessing NAFTA appears inconsistent with the zero-sum view of trade.

STOKEY: I certainly don't think it's a zero-sum game. There have been gains all around—that's just the standard-gains-from-trade argument. My initial comment simply had to do with how easy it is to identify the gains. Because the U.S. economy is so much bigger in terms of the number of people and the size of the incomes, the gains here are much smaller relative to the size of the whole economy. It's like measuring the effect of adding a gallon of water to a bathtub vs. adding it to a swimming pool. It is simply harder to pick up in the second case. Now in terms of jobs, certainly we lose some jobs and we gain some. That's in some sense the whole point of trade—to change the mix of jobs in this economy. Presumably what we export are lower-paying jobs in manufacturing and what we gain are higher-paying jobs in more high-tech industries like software, pharmaceuticals and financial services.

ROLNICK: Let's move on to the Industrial Revolution because it has been a topic of conversation here at the Minneapolis Fed in trying to understand what's going on in our economy today. Some claim we're experiencing a new type of technology revolution and some economists here and elsewhere have tried to draw parallels between what went on in the Industrial Revolution and what's going on today. Are these useful comparisons? If they are, what have we learned?

STOKEY: It's an interesting comparison. The main idea is that occasionally you have some important technological change that has ramifications in many sectors of the economy. It's an interesting parallel, but it is not obvious what lessons can be learned from it.

ROLNICK: But you're saying there's a chance we're seeing something like that again. Presumably that's going to mean more economic growth, larger productivity gains. Is it clear who's going to benefit from this new technology? There used to be the fear that machines will come along and displace workers, for example, and there would be high rates of unemployment. Are you concerned?

STOKEY: No. There's always been a fear of machines taking jobs away from people, but in some sense the whole history of the economic progress of the human race has been finding ways to economize on labor. If I'm a farmer and it takes me all day to dig my field with a shovel, I'm going to be a poor farmer. If I can employ a tractor, I'm going to be a much richer farmer. Individuals have feared progress, and indeed there have been individuals who have lost their jobs to the tractor. But the upside is that they, or at least their children, become the ones who drive the tractors.

Is that happening now? Over the last decade labor markets have gotten very tight, so it seems that if people have lost jobs because of the new technologies, they're managing to find new ones without too much trouble. Maybe one lesson we can learn from previous episodes like this is that it's probably not going to last forever. Innovations that are set off by a fundamental technological change tend to come in waves. The first time around the key invention was the steam engine, the second time it was the gasoline engine and electricity, and this time it's the computer and the Internet. In each of the previous episodes, the key invention led to a wave of further innovations that exploited it and produced a substantial spurt in the growth rate for a while. But eventually, after the most important uses of the new technology have been exploited, the growth rate tapers off. Then you have to wait for another major invention. To the extent that the current expansion is driven by a wave of technological change based on computers and the Internet, we cannot expect it to last forever.

ROLNICK: I won't dispute that. But the one thing different about this new technology is that it's going to be very inexpensive for everybody to use. Which says that maybe there's a better chance that other economies that have, for whatever reason, lagged behind economic progress will have a better chance of catching up. If that were true, you would think that the spurt of growth might last a lot longer than other ones.

STOKEY: That might be. And it might be that the growth will be more widespread. Maybe this is an innovation that can benefit even relatively poor economies, because it helps them import ideas from their richer neighbors. For example, one problem in developing countries is training teachers, and the new technologies may be useful in easing that bottleneck. My nephew took a calculus course over the Internet when he was in high school, so the potential is certainly there.

ROLNICK: A question about the policy implications of globalization: It's not always clear what people have in mind when they use the term. Nevertheless, when we consider monetary policy, we realize that central banks have to consider coordinating policies more today than they ever have in the past. For example, given the rising number of international corporations, one can imagine a rising demand for an international currency. What do you think the implications of globalization are for central banking as well as economic policy generally?

STOKEY: For the main conduct of monetary policy, I don't think it should be much of a factor. As we started out saying, the main goal of monetary policy is domestic price stability, creating an environment that lets the domestic economy flourish. There's not much beyond that, that a central bank can do. Now there may be issues of exactly how policy is conducted, and it's certainly true that money moves very fast now around the globe and maybe bankers have to worry about that. I'm not enough of an expert to want to comment on that aspect.

ROLNICK: Let's go back to monetary policy. What if another currency, say the yen, becomes more stable than ours, and it becomes the international currency of choice; and suppose the yen starts circulating in this country. At a minimum, you might worry about seigniorage. Would such a development create other problems that we should worry about it?

STOKEY: The only places where foreign currencies have made big inroads are places where the domestic currency is a very unattractive alternative. It happens in countries that are experiencing high inflation, so people want to get out of the domestic currency. If the central bank keeps the inflation rate down, it is very unlikely that another currency is going to make inroads.

ROLNICK: Would there be any gains in coordinating monetary policies across countries to encourage a low inflation rate worldwide?

STOKEY: It would be great if inflation were low everywhere, but I'm not sure if it takes any particular coordination to achieve that. I would think the more important aspect of policy to coordinate is fiscal policy. Corporations can move capital around, they can move profits around. So coordinating fiscal policy may be much more important.

ROLNICK: You say coordinating fiscal policy is more important. I'm not sure why that's true. In the United States, for example, we don't worry about coordinating budget policies among states; states run their own independent fiscal policies, that doesn't seem to be an issue. Why should it be an issue across countries?

STOKEY: In the United States, some of the biggest taxes—the personal income tax, the Social Security tax and the corporate profits tax—are levied at the federal level, so coordination is not an issue. At the state and local level, it is true that state legislatures and city councils do not explicitly coordinate their policies when they set rates for state income taxes, wage taxes, excise taxes and so on. But neither do they operate in a vacuum. Cities and states are keenly aware that businesses and people choose where to locate, so fiscal authorities must compete to maintain a desirable environment. If their policies get very far out of line their tax base starts to erode. The key behind this discipline is the fact that individuals and businesses can move freely across city and state boundaries.

ROLNICK: I know the Europeans worry a lot about coordinating their fiscal policies. But I thought that was simply because it made the European bank's monetary policy more credible and the euro more stable.

STOKEY: There are two parts to fiscal policy. I agree that big budget deficits are probably inconsistent in the long run with a very stable monetary policy, so if you're going to join together with your neighbors in a common currency area you want to all agree to keep budgets more or less in balance. Note that most states in the United States run basically balanced budgets, so coordinating budget policies across states has never been an issue.

But the various national governments in Europe will also have to worry more and more about competing to attract businesses. In Europe it used to be much harder for businesses and people to move across national boundaries. That was true for reasons of language and culture, as well as legal restrictions. But as barriers have broken down, the European states will have to worry more about competing with their neighbors to keep old businesses and attract new ones.

ROLNICK: That makes sense if you're going for a common currency, but in this floating rate environment, is it necessary to worry about those countries that float against each other coordinating fiscal policy?

STOKEY: If one country tries to levy a higher tax on wages or a higher tax on capital, those factors can move to avoid taxation. So countries are going to have an incentive to harmonize the particulars of their tax policies.

ROLNICK: In an interview with the University of Chicago Chronicle, you spoke briefly about the effects of pollution, the ways society responds to pollution in terms of growth. What does history tell us about the way developing countries will respond to pollution over time? On the assumption that certain historical patterns exist, are there exceptions to this pattern?

STOKEY: My reading of the evidence is that there is a pattern, at least for some types of pollution. And it is that pollution levels vary systematically with per capita income levels. This is true both across countries and within a country over time. Pollution first rises with the level of development, then peaks at a critical level of per capita income, and then starts to decline with further increases in income. For a couple of major components of air pollution, the critical level is a per capita income of around $4,500.

My interpretation of this is that clean air is a luxury good, and below a per capita income of $4,500, people simply don't want to spend their income on cleaning up the air. They'd rather spend it on other things—food, shelter, education. At some point they become rich enough so that the pleasures of clean air and clean water start to become important, and collectively—at the voting booth or through some political action—they start to impose tighter regulations on industrial polluters, automobile emissions and so on.

ROLNICK: That gets me back NAFTA and Mexico. If NAFTA is successful and the Mexican economy is doing better, should we look forward to the Mexican economy generating less pollution?

STOKEY: I think so. Since they're just at the critical income level right now, I would expect to see some action taken in the next decade, not through any coercion on our part but simply in response to domestic pressure. I predict that very soon Mexicans will decide that the air quality in Mexico City is intolerable, and they'll do something about it.

ROLNICK: People are arguing that before we open up trade and provide all these benefits to these developing countries, we should impose certain standards on air quality. You're suggesting that that may not be necessary and may be counterproductive.

STOKEY: Exactly. I think the way to make the world cleaner is to encourage growth in less-developed countries. Once those economies get richer, they'll clean up on their own without any prodding from us.

ROLNICK: I'm not sure that advocates for NAFTA have made that argument, and they probably should.

STOKEY: I agree. It is clear that NAFTA is stimulating economic growth in Mexico, and faster growth will hasten the day when Mexicans take care of their pollution problems on their own.

ROLNICK: Let me turn to a topic that we've been interested in for a while. About a year and a half ago the Minneapolis Fed sponsored a symposium on economic literacy. A variety of issues came out of that symposium, and one was the question about why we haven't seen more women in economics programs, graduate programs in particular. While women have entered medicine and law in record numbers, there are relatively few women entering economics. I know you participated in a 1998 workshop related to this issue, Creating Career Opportunities for Female Economists. Any explanation for this observation, and what does it tell us about economics?

STOKEY: It's difficult to explain why women are not better represented in economics. It seems that balancing the demands of a career and the demands of a family is still a big issue for women. Investing in a Ph.D. requires quite a few years.

ROLNICK: More than medicine or law?

STOKEY: No, but what may be harder about economics, and this would be true in other sciences as well, is that it's hard to pursue a part-time career. It may be that as a lawyer or doctor or even as an academic in English or history, you can pursue a career part time for a while as you raise your family without hurting yourself too much. By contrast, in the sciences, including economic science, things change so quickly that it's very hard to keep up if you're not working full time.

ROLNICK: We've been grappling with this question now for a while. I don't know if your explanation will hold up to the data, but it's the best answer I've heard so far.

STOKEY: Relatively more women pursue Ph.D.s in the humanities, but I've never seen a systematic set of figures. It would be an interesting thing to look at.

ROLNICK: I understand you and Bob [Lucas] have a summer home in Door County, Wis. Besides the local fish boils and famous Door County cherry pies, which I've never had, what else do you like to do for relaxation when you're out of the city in this beautiful country?

STOKEY: We enjoy all sorts of outdoor activities—running, bicycling, hiking, bird watching. And sometimes it's nice to just sit back and relax—to watch the gulls dive and the waves roll in on Lake Michigan. It's beautiful there.

ROLNICK: I imagine you get there mostly in the summer.

STOKEY: Yes, that's when we can get away. We fold up our laptops and take our work along. Unfortunately it's too far away to get there much during the academic year. Although it's very lovely in the fall. There are quite a lot of sugar maples in Door County—maple syrup is a local product—so the foliage is very pretty right around this time of the year. In fact, it's a lot like Minneapolis.

ROLNICK: Thank you, Ms. Stokey.

STOKEY: It's been my pleasure.

Top