Interview with Edward Green

"We start with the idea that, plausibly, what's going on is efficient."

These dozen words encapsulate a central theme in the lifework of Ed Green. While many economists seek to accurately portray the world as it is, complete with imperfections, and others describe an ideally efficient economy, Green's research presumes that the world and its institutions do, in fact, operate with rough efficiency—otherwise, participants would have worked out a better way. And that presumption calls for a broader perspective.

If cartels endure despite frequent price wars, for instance, perhaps price wars actually sustain cartels, argues Green. If state-owned enterprises have long lives, maybe they're valued for reasons not readily apparent to standard economic analysis. An institution that seems unstable and inefficient may not be so, he suggests, if viewed comprehensively.

Green's gift is to examine common problems from unique directions and to detect what escapes conventional scrutiny. His capacity for original, independent thought has been of particular value to the Federal Reserve, where he has served at the Board, the Minneapolis Fed and, most recently, the Chicago Fed, before joining the faculty at Pennsylvania State University. When the Fed needed a rigorous evaluation of its role in the payments system, for example, Green was chosen as chief economist.

Other themes underlie Green's work. The distributional implications of monetary policy are a key concern—one he feels is neglected often in policy and academic discussions. The need for humility in making policy recommendations is another. "The facts might one day overturn my advice," he cautions. "Contrary advice is ... also entitled to careful hearing and respectful treatment."

Such candor and generosity run throughout the following discussion—an unusually diverse yet cohesive Region interview ranging from the emergence of democracy in 13th century England to current controversies in central banking.

COLLUSION

Region: I'd like to begin with a question that goes back to one of your first papers: your 1984 Econometrica article with Robert Porter. It showed that under certain conditions, even among firms that are colluding in an optimal fashion, price wars can periodically occur. This was a major advance in game theory and theories of imperfect information.

Were there direct policy implications of this finding? And what advances have been made on the economics of collusion since your paper?

Green: We were influenced by a celebrated paper on cartels that [University of Chicago economist] George Stigler wrote two decades earlier. Stigler documented that cartels tend to break down within a year or two. He concluded that you don't need to prohibit collusion because it collapses on its own. In his view the social cost of collusion is temporary and low.

Rob and I looked at Stigler's evidence from a different perspective. Stigler documented that cartels in some industries broke down repeatedly, and that means they had to be resurrected repeatedly. If he was right that collusion was unprofitable, and if these attempts to restart cartels were costly, then the firms that kept trying to collude were repeating a mistake over and over again. Their executives should have learned after a few attempts that their short-lived cartels lost money, and then they should have said to themselves, "This is a nonstarter. Why bother?" Economists think that people making decisions typically get better at what they do in this way over time, but Stigler wasn't applying this insight to the cartel members.

So we turned Stigler's argument around and said collusion must be profitable under some circumstances since people were continuing to try to do it. Therefore, these apparent breakdowns must serve some role inside the framework of a larger, long-term, stable arrangement. And we found such a role: as an enforcement device. Whenever it looked to everyone in the cartel like cooperation among them was breaking down, they would suspend their cartel for a while and forgo profits in an episode that resembled a "price war." These episodes could sometimes be triggered by accident—those are the breakdowns that Stigler documented—but the knowledge that overproduction or price cutting would cause such an episode deterred cartel members from engaging in such behavior. In this way, a cartel agreement could be self-enforcing.

Game theorists had begun to model self-enforcing agreements sometime before Rob and I proposed a cartel theory. We did make one narrow advance: to explain how collusion could be enforced when there wasn't completely precise information about whether a breakdown of cooperation had happened and, if so, exactly who was to blame.

But I think the more important advance in the paper is our general approach to this type of problem. Our starting point is that we, as economists who study institutional arrangements from the outside, should give an observed arrangement the benefit of the doubt that it serves its participants well. The idea is to look at what happens, some of which may seem problematic—such as the breakdowns of collusion that we were trying to reconcile with cartel stability—and try to figure out what must the environment look like in order for those problematic aspects to be functional in the institution.

This is somewhat different from what used to be the typical way of thinking about economics: that there is positive economics that explains what goes on, and welfare economics that explains what's efficient, and that the two really don't meet until the end of the day. We start with the idea that, plausibly, what's going on is efficient. And we use that working hypothesis to guide our theorizing.

Regarding what's gone on in this field since our paper, first I'd say that Rob Porter has gone much further than I have. He has been a pioneer of what's now called structural econometric modeling: to take a theoretical model and actually estimate the parameters of the model rather than merely estimate some statistical features of the outcomes of the model. Therefore, he can address questions about when does collusion arise and when doesn't it? And what kinds of policies would create an environment in which collusion is less likely, or less socially costly?

In a companion paper to our joint paper, Rob wrote about a historical 19th-century railroad cartel, showing striking evidence that the kinds of breakdowns of cooperation that we had hypothesized did indeed take place in that industry. There's subsequently been a lot of work, some of which is Porter's, on collusion at auctions, including government auctions, such as sale of timber in the national forests and purchase of milk for school lunch programs. Significant public revenue and expenditure is at stake. So there is public benefit from research that helps keep these auctions—as well as many private-sector auctions—competitive.

There has also been further theoretical work on collusion. More broadly, I think we have a much better understanding today than we did 20 years ago of what determines whether or not an industry performs in the public interest. In my view the most striking contribution of this sort has been the development of "contestable markets theory."

According to this theory, the fact that only a single producer is active in a market doesn't necessarily mean that the producer charges a monopoly price. The theory argues that there might be another competitor waiting in the wings, who would undercut the incumbent if the incumbent did try to charge a monopoly price or anything above the competitive price. For instance, when there are economies of scale or scope in an industry, the socially efficient way to produce things may be to have a single active producer disciplined by the threat of entry of this kind.

There are specific conditions that have to be satisfied in order for such a market structure to lead to efficiency. Not all markets satisfy the conditions for contestability. But some important industries do satisfy those conditions. A number of them have been "deregulated" because this has been recognized. I've seen two separate studies that assess the effects of deregulation across the U.S. economy, and they both conclude that the overall effect is beneficial.

SECOND BANK

Region: In a recent article [PDF] in the Chicago Fed's Economic Perspectives, you argue—I think you say, tentatively—that conflict between creditors and debtors over price stability and credit may have been an important influence in the contentious history of the Second Bank of the United States, a predecessor to the Federal Reserve.

Was this creditor/debtor conflict more important than disputes over the expansion of federal powers or opposition to the monetary discipline that the Second Bank imposed on state-chartered banks? And does it have implications for banking today?

Green: I'd say it was two faces of the same conflict. Let me go back to what I was saying a moment ago about the methodology that Porter and I used in talking about collusion. It has to do with the structure of economic and social institutions, on the one hand, and with the equilibrium performance of those institutions, on the other.

This distinction is important because sometimes a single institution can have several equilibria that are very different from one another. The equilibria might have very different efficiency and distributional consequences. Basically, an equilibrium occurs when each participant in the institution thinks about what other participants are going to do and on the basis of that responds with the participant's own optimizing action, and the optimizing actions mesh.

Such a multiple-equilibrium situation is what I claimed may have happened in the case of the Second Bank of the United States. There was one equilibrium in which the Bank would impose monetary discipline on the banking sector, which at the time consisted of state-chartered banks. Many people thought that those banks made unsound loans and overissued money, leading to inflation. They favored the Second Bank because of the discipline they thought it would provide.

There was another equilibrium in which the Bank was going to be no more responsible than the state banks had been and was going to be essentially an engine for funding the state banks with inflationary finance. Some members of Congress who chartered the Second Bank may have been voting for the second equilibrium instead of the first because their constituents benefited from inflation that was mainly at the expense of other parts of the public.

With various supporters of the Second Bank having such radically different ideas of why it was good—well, there was going to be difficulty down the road. What happened is that for the first few years the Second Bank was run in a lax way and then was reformed and apparently did provide monetary discipline. That is, it flipped from one equilibrium to the other.

I drew two lessons from the episode. The first lesson is about the implications of what central banks do. The Second Bank, in many scholars' opinion, was close in a number of respects to a modern central bank. What modern central banks do has big distributive impacts as well as efficiency impacts. Central bankers today don't much like to talk about that. They like to talk about monetary discipline as being good for everyone, but at least in the short run it isn't. In fact, there are some equilibrium models in which, in the long run, if agents are heterogeneous in the economy, what's good for most of the agents may be bad for a minority. This issue does not receive as much attention in academic macroeconomics and monetary economics as it deserves.

The second point is that, if the vision of central banking I've just sketched is true, then it's important both for central bankers and for the political decision-makers who charter central banks to work as hard as they can to establish an institutional structure that avoids multiplicity of equilibria. I believe that it's much better for the question of what the central bank is going to do to be resolved irrevocably before the institution starts operating, rather than becoming a huge political problem for the institution, as it ultimately did for the Second Bank.

My model of good institutional design is the separation of powers in the U.S. Constitution. The founding fathers differed among themselves about what the powers of the president should be, for instance, but they forged a fully articulated compromise and invented institutional ways of enforcing it before sending the Constitution out to be ratified. A central bank charter needs to be carefully written in this way too.

Region: When I think about implications for today, I naturally think of the Fed having dual mandates: price stability and full employment. Is that a similar dilemma?

Green: Yes, it probably is. Let me first mention that there are many countries, so-called emerging market countries, that have much younger central banks than the United States has, and my impression is that this problem has recently been of the same magnitude in some of those countries as the problem of the Second Bank was when the United States was a young country and a young economy.

Region: Argentina?

Green: Actually, a few Argentina economists have told me that they see a parallel between their recent history and my reading of the Second Bank's history.

But getting back to the Federal Reserve, you're right about the dual mandate. In recent years, the Federal Reserve has made it very clear to the public that they think that what a central bank can do to foster full employment is to maintain a stable price environment. The formula they use is that they provide monetary stability directly and that, through doing so, they provide economic growth and full employment indirectly, so there's no conflict.

Now, of course, at times, that formula doesn't fly very well. When you get a situation with unusually high unemployment but low inflation, you're going to have people second-guessing the Fed, asking, Isn't the Fed making the wrong trade-off? Can't it relax its fixation on the goal of monetary stability a little bit in the short run in order to do something immediately about this terrible unemployment?

That sort of thing is one of the reasons people around the world have been talking about inflation targeting as a preferable operating policy for central banks. Some people in the United States would like to replace that legislative mandate dually to foster price stability and full employment with a more univocal direction to foster price stability.

INFLATION TARGETING

Region: And what is your view on inflation targeting?

Green: Because I view the central bank as being unavoidably something of a political institution, I'm not sure inflation targeting is very effective. To begin with, Congress can always undo a law, just as it would be undoing the current law by getting rid of the dual mandate. The politically more sophisticated way to ensure price stability would be to create an institutional structure for the Federal Reserve, such that the importance of price stability always gets full consideration internally and the Fed is able to communicate those considerations and their importance externally in a persuasive way.

In that regard I think that the Fed is an admirable institution, particularly because of the role of Reserve Bank presidents in the formation of monetary policy. The Fed is practically the only central bank where a near majority of the voting members of the monetary policy committee—the Federal Open Market Committee (FOMC) in the United States—each have an individual research staff that owes its allegiance to that individual committee member. The European Central Bank, which is also organized in this way, seems to have been modeled deliberately on the Fed.

In Britain, roughly a decade ago, they revised the institutions for forming monetary policy, and they have a monetary policy committee with several external members. They tend to be distinguished professors of monetary economics, for instance. A few years back, the Bank of England recognized that those members need to have such an independent research staff. They were given a small staff, but obviously the external members have far fewer resources than the staff that the Bank of England provides to the governor and the deputy governors. The Reserve Bank presidents are closer than that to parity with the Federal Reserve governors who sit on the FOMC.

I believe that this rough parity, and particularly the relative independence of Reserve Bank research departments from direct control by the Board of Governors and its staff, is worth its weight in gold for the credibility of the Fed. When the policy of the central bank, which usually is largely decided by the governor, goes away from the direction of the priority of monetary stability, then if you want to have effective high-level discipline (and if necessary at the end of the day, dissent), then the prospective whistle-blowers have to have the resources to make their case cogently. You can't have the governor of the central bank first denying access to economists, databases and other resources to dissenting members of the monetary policy committee and then turning around and saying that those members shouldn't try to second-guess him because he has heard from the experts and they haven't.

Each regional Reserve Bank has a research department that contributes to the collaborative national discussion of important issues and also provides somewhat independent review of the work by Board of Governor's staff that factors most directly into FOMC decisions. If economists at a Reserve Bank have reservations about work done by the Board's economists, and if the issue doesn't get resolved at a staff-to-staff level, then the president of that Reserve Bank can raise the issue in the FOMC itself and, if necessary, can raise it more publicly.

This arrangement is reminiscent of the constitutional separation of powers that I mentioned earlier. In both cases, the point of the arrangement is to provide incentives to public servants to avoid bad situations strenuously. Let me emphasize that, in the case of the Fed, the kind of last-ditch dissent by a large bloc of FOMC members that I was contemplating a moment ago rarely if ever comes to pass. Nevertheless, besides the day-to-day contributions that the Reserve Bank research departments make, their existence lends credibility to the idea that Reserve Bank presidents could function as effective and responsible dissenters if they had to.

The other thing I think the Fed has done wisely in the last few years is to reveal more quickly and more completely what the discussion in the FOMC has been. Within the last year or two, for instance, the FOMC has adopted a policy that minutes of meetings are released almost immediately and the identities of dissenting voting members are immediately made public.

I would guess that having this sort of transparency helps much more than having a targeting rule. If I had to depend on an institution where there were people well equipped to make the arguments in terms of price stability and where the arguments of those senior people would be transmitted almost instantly from an internal forum to a public forum, versus a system where a central bank governor gets up periodically and makes some formulaic statement about "we all expect inflation to remain low in coming months, so we don't have to take any contractionary actions now," I think the institutional commitment to transparency is a lot more trustworthy, when push comes to shove, than the formalistic inflation-targeting procedure solution would be.

WOODFORD'S INTEREST AND PRICES

Region: In your recent Journal of Economic Literature review of [Columbia University economist] Michael Woodford's Interest and Prices, you write, "It's likely to be a bible for central bank economists."

But you also have reservations, expressed at some length in the 14-page review. Can you give us the gist of your views about Woodford's book?

Green: Woodford has shown that the methodological revolution in macroeconomics 30 years ago doesn't necessarily imply some policy conclusions drawn from the specific models studied then. That's very important. I was reviewing the book in a journal for economics researchers, and in such a context, the point of the review is to give a counterpoint to the author and tell readers what things they ought to think about critically when they read the book. Simply rehashing the author's exposition of his or her own framework, and saying "oh, yeah, that's the natural way to think about it" doesn't accomplish anything. So the fact that I spent much of the review discussing alternative frameworks that might be adopted is by no means an indication of lack of respect for, or even necessarily of dissent from, what the author is saying.

Rather than talking about concerns about the book specifically, I'd like to mention two concerns I have about monetary economics today in general.

We've already discussed one of them. The point of view that's almost universally taken in monetary economics today is that, sometime around the early decades of the Fed, central bankers discovered or invented a set of monetary tools, such as open market operations, that were somehow nonpolitical. According to this view, as long as central bankers confine themselves to manipulating only a short-term interest rate—rather than, for instance, subsidizing certain sectors of the economy and withholding subsidy from other sectors—they are doing something that can solely have efficiency implications but not distributive implications.

I think it's true that what central bankers do today minimizes the distributive implications of monetary control, but my impression is that the distributive implications remain very significant. Think about the episode of moderate inflation in the United States in the late 1970s. After Chairman Volcker took decisive measures in 1979 to bring it under control, a lot of people were badly hurt.

The way economists talk about it, it was a short-term hurt. But it wasn't transient. The term "Rust Belt" originated to refer to what were basically abandoned industrial regions largely in the northeast United States. It may have been that longer-term trends were going to cause the decline of those industries anyway, but there's no question that monetary policy made the decline more sudden than it would have been otherwise. It caused layoffs of workers who, instead of gradually leaving those industries, were dumped en masse onto the labor market during a deep recession. It took most of them months or even years to find subsequent work, and perhaps they found less desirable subsequent work than they otherwise might have been able to had the Fed's disinflation policy been more gradual.

Now, most people think Volcker's policy was exactly the right one. But you shouldn't pretend that it didn't hurt some people pretty badly. As I point out in the review, the fact that it hurt some people badly and that they responded through the political process was an important constraint that the Fed had to deal with. A history of how the Fed dealt with that moderate inflation and its achievement in doing so would be incomplete if you abstracted from that distributional and political problem.

By the way, at this point, let me mention a review that I once wrote for the Region of a book on rational expectations and inflation by Tom Sargent [see the December 1997 Region]. Sargent has thought hard about the problems I've just been discussing. His book is a wonderful and readable discussion of the issue.

The second concern I have is the thinness of the data about macroeconomics. Basically, we deal with about half a century of quarterly data on various aggregates, so that's 200 observations. And we try to fit time series statistical models where 200 observations, spanning only a dozen or so business cycles, may not go very far. As a result, it's difficult to draw firm conclusions about some things. So there are disputes about whether some features of the data that seem to be picked up by certain statistical techniques are really there or not. And in various debates, both sides push pretty hard on their interpretations of the facts. In particular, there are issues about the response lags of employment to money- or interest-policy changes that are a matter of dispute between the economists who base their work on purely competitive models of the economy, as, for instance, Bob Lucas and Ed Prescott do, versus Woodford and others who base their modeling on imperfectly competitive representations of the economy.

What I was trying to do in my review of Woodford's book was to say, Let's keep in mind that we don't have enough data for anyone to have the last word on these things. Consequently, we have to accept the situation where economists, or for that matter citizens, who begin with strongly held views one way or another, simply may not be persuaded on the basis of the facts we now know or are likely to know in the near future by statistical work done by people on the other side of the question. People sometimes ask me if I am for or against these macroeconomic models with assumptions about imperfect competition. And my answer is, I'm neither for nor against. I think science progresses best when scientists study all of the models consistent with the firmly established data, and I don't regard the data as being ample enough to reject either sort of model out of hand.

Meanwhile, central banks have to make policy decisions about which those two types of model tend to have divergent recommendations. I can recommend a policy decision one way or another while acknowledging that the facts might one day overturn my advice. And I would also want to acknowledge in such a situation that the contrary advice is coming from people that I highly respect as economists. Their views are also entitled to careful hearing and respectful treatment. By the way, let me mention that one of the most careful and generous people with regard to making this sort of acknowledgment is Mike Woodford.

PAYMENTS SYSTEM

Region: You co-authored our 2000 Annual Report essay about the role of the Fed in the evolving payments system—suggesting that we should specialize in our comparative advantages of regulation, supervision, interbank settlements and research, and cede some of our role in providing direct payment services [see the 2000 Annual Report issue of the Region, April 2001].

What are your current thoughts about the Fed's evolution in payments?

Green: What Dick Todd and I were urging in the article is that the Fed evolve to the place where the central banks of most other industrialized countries already are. They have a policy role, in most cases an operational role in the backbone system for making large value payments (including the interbank payments that are made in the operation of monetary policy), but little or no role in so-called retail payments—the payments used to settle garden variety purchases of goods and services both by households and by firms outside the financial sector—manufacturing firms, for instance. Instead, those payments are handled by private-sector bank and nonbank entities that operate in an entrepreneurial way.

At this point, let me say that my further remarks past this summary of the essay I co-authored with Dick Todd reflect only my own views.

Earlier I mentioned the idea of contestability, that a single firm can be active in the market and nevertheless be disciplined by potential competition from other firms that have the capacity to enter the market. That situation can produce competitive outcomes despite the fact that we don't see those other potential competitors competing.

This applies to the payments field, where for some products like the Automatic Clearing House, there is a single non-Fed provider, and for some other products there are only a few large firms—for instance, in the nationwide correspondent banking business. So probably the most cogent argument for the Fed to be in this business is that it provides an alternative to what would otherwise be a monopolist. But if one takes the contestable market view of this industry, then those markets in the absence of the Fed would not be monopolistic.

In any event, someone who wants the Fed to be in this market bears the burden of proof to explain why we don't do this more generally: Why, for instance, don't we have a public-sector operating systems firm to compete with Microsoft? and so forth. If this is not good policy advice for the software industry and numerous other industries, then what's special about payments? I don't think you can make the argument that there's anything special. From a public policy viewpoint, the best course would be for the Fed simply to get out of the business.

Region: In your opening presentation at the Fed's 2004 conference on the economics of payments, you presented several challenges, regarding models, data and advice, among other things. Has progress been made in meeting those challenges?

Green: People have certainly been working on better models. One of the big issues about data that I mentioned in my talk is that the data are largely proprietary data of either private-sector entities or the Fed. The entity that has the public perspective from which it might see a gain to releasing some of that data is the Fed. Now, the Fed cites various issues, such as confidentiality, for not wanting to do so. But I think a better effort should be made.

One model for this effort is the establishment by the Census Bureau in the last decade or so of a number of census data centers for academic research. The Census actually owns the facilities and runs the computer programs that researchers submit, in order to keep the raw data under its own control, but enable researchers to use the data.

Something else that's done, and in fact a large number of private-sector firms do this, is allow data to be provided to researchers after masking those parts of the data—for instance, individuals' names—that would violate confidentiality.

But in payments we have a situation where issues of some significance for public policy are being debated, where the Fed has a position on the issues either as a regulator or as a participant in the market, and where outside researchers can't obtain the data that they need to draw an independent conclusion about the wisdom of what the Fed is doing. I would think that as a publicly motivated central bank, the Fed ought to be eager to enable such research. Citizens ought to be eager that the Fed do so. This goes along with what I said earlier in our interview about monetary policy, that the best incentive for good policy and guarantee that policy is being made well is openness and exposure to responsible criticism.

PRIVATE INFORMATION AND DEMOCRACY

Region: That might be a good segue to your historical analysis of the emergence of parliamentary systems in Europe back in the 13th century, in which you highlight the role of private information. Can you briefly explain why changes in private information may have been more influential in that transition toward democracy than shifts in the balance of military power? And does the same idea pertain to the spread of democracy in more modern times?

Green: I'm glad you ask about this—of all the research that I've done, this is what I'd most like people to read [see "On the Emergence of Parliamentary Government" in the Winter 1993 Quarterly Review]. I began in the 1980s to think about the spread of democracy. Around the world in the 1970s and 1980s, countries that had been dictatorships for decades made surprising political transitions. They moved into being democracies outright or went far in the direction of democratic reforms that looked likely to continue—and indeed, in many cases, have progressed.

One of the things that struck me is that this was a worldwide phenomenon that happened almost simultaneously, although without coordination, in about a dozen countries. Another thing that struck me was that the transformations that looked the most successful were the least violent.

In Iran, for instance, the shah's government used violence against its opponents and was nevertheless overthrown. But the result of the overthrow was a government that was not a great deal more democratic than the one it replaced. On the other hand, in Argentina, where the government previously had a horrible human rights record, there were nonviolent demonstrations that, for the most part, the government didn't respond to with heavy violence. At the end, the junta there resigned and a genuinely democratic government immediately took its place.

The accepted theory at the time about how countries move between dictatorship and democracy emphasized the ability of a dictator to get enough military power to bring everyone else under control and the ability of large masses of people at other times to oppose that force of arms successfully.

But that didn't seem to be what was going on. Indeed, in the situations where it looked like a contest of force-of-arms, you were seeing transitions from one dictatorial government to another. And where you saw genuine transitions to democracy, something else seemed to be happening.

About the same time, by coincidence, I was doing some leisure reading about the English history and the Magna Carta. I picked up a wonderful book about the topic by J. C. Holt [Magna Carta, 2nd ed., Cambridge University Press, 1992]. What I learned about that situation seemed to me very contemporary and also raised a puzzle.

The quarrel between the king and the barons—the barons having insisted on the Magna Carta—was about taxation, not about the level of taxation, though the level of taxation by King John was much higher than by his predecessor. The barons seemed to say that the high level of taxation on the whole was appropriate—the taxes were being levied to defend England from foreign enemies. Their concern, rather, was that they were not being duly consulted. This was a novel claim that did not have precedent in its favor. The puzzle is, Why did the barons focus on this narrow claim rather than relying on a traditional claim that had been successful in past episodes when English barons had resisted a king's taxes? Holt's conclusion was that the barons made a strategic mistake in casting the Magna Carta in those terms. But I thought that Holt, like Stigler, hadn't made the best posa king's taxes? Holt's conclusion was that the barons made a strategic mistake in casting the Magna Carta in those terms. But I thought that Holt, like Stigler, hadn't made the best possible effort to understand why the barons took the tack that they did.

If the barons' dissatisfaction had been about the level of taxation, it would have fit neatly into the balance of power framework. We could theorize that the barons got more powerful at the beginning of the 13th century than they'd been in the 12th century for one reason or another. And they were able to resist paying so much money to the king.

But their dissatisfaction wasn't about paying money at all, but rather about the procedures under which money should be paid. That calls for a different theory. The idea that there was a set of problems related to private information at the beginning of the 13th century that didn't exist in the middle of the 12th century was a theory that I could build a little bit of support for.

So this is another example of the methodology that I talked about at the beginning of our interview. I noticed a transition from one set of social or political institutions to another, and I operated with the presumption that each set of institutions was efficient in circumstances for which it was designed. Then my problem as an economic theorist was to ask what changed, such that the old institutions were efficient in the 12th century and the new institutions were efficient in the 13th century.

I'm pleased that I was able to write the paper without any mathematics whatsoever in the body of the text. This is a little paper that somebody who is not an economist can read and, I hope, learn something about how economists think.

Region: So, in a nutshell, how does private information explain the transformation of England's political structure?

Green: King John had previously been the king also of Normandy in France, but 10 years before the barons presented the king with the Magna Carta, he had lost Normandy when the French invaded. My idea was that Normandy previously had been the English radar screen for an invasion across the channel. The French would have had to invade Normandy before they could invade England. A number of the barons in England had also owned land in Normandy that the king had given them, so if the French had invaded Normandy, of course, they would have known about it. The king would have then said, "You barons see that your lands in Normandy have been invaded. Now you'd better cough up some money for me to raise an army or cough up some troops for me to defend your lands in England."

After the barons lost Normandy, they were in a position where the king plausibly had better information than they did about what was happening on the European continent. The king had diplomats there, but the barons didn't, for instance. So the king would come to the barons and say, "I need your resources to fight the French." The barons would be naturally suspicious and ask, "How do we know that the king wants our resources to fight the French rather than to buy himself some luxury goods?"

The institutions that the Magna Carta set up, particularly the parliament, were essentially a device for the barons to require the king to come and talk to them and lay out the case that there really was a threat and that he really needed to raise the level of resources he was asking for in order to defend them all against it.

Region: To justify his budget request, in a sense.

Green: Right, in modern equivalent.

Region: Why is this the paper you most want others to read?

Green: First, to be honest, because I think the worldwide transition from dictatorships to democracy inspires me more than other topics I've worked on. This is a truly important, fortunate development in our time. I suspect it's also related to another such development: the elevation of a substantial proportion of the world's population from extreme poverty to something approaching an industrialized-country standard of living. It's a privilege for me to have the opportunity to think about these things and perhaps to do some thinking that's useful in furthering them.

I hope that’s a fair summary. If not, please correct me, and then tell us how the post-Comecon economies have evolved vis-à-vis your predictions.

Green: Well, that’s a good summary. Let me mention that I received valuable help on this from two economists affiliated with the Minneapolis Fed, Stephen Parente and Jim Schmitz. Stephen helped me make the statistical arguments that the Soviet economy was not as bad as sometimes depicted. My thinking about how formerly planned economies can move forward drew heavily on Jim’s analysis of how small firms generate innovation in the U.S. economy.

Economists tended to underestimate the competence of Soviet industry because the Soviet economy was giving priority to producing military goods. Of course, the Soviet regime is widely viewed as having lost an arms race with the United States during the Reagan administration. That race was immensely costly. So the Soviets were producing the wrong goods for maximizing consumer wealth. In fact, central planning was an inept system for eliciting information from consumers about how they would like productive resources to be used and then giving incentives to the producers to produce those goods.

But if you ask, “Were Soviet industrial managers and workers competent at using their resources to produce the goods that the political system ordered them to produce?” The answer is, “Yes, they were pretty competent at it.” On the other hand, privatization of state enterprises was being widely urged at that time. If you started out with the assessment that Soviet managers were capable, and you added the premise that the Soviet leadership had been deposed and was being replaced by leaders who were going to give more priority to identifying consumer demand, then privatization didn’t make much sense. Once you had state-owned firms operating in some sort of a market environment where they were receiving price signals and where their performance could be evaluated—even if it would be partly a political evaluation rather than a narrowly economic evaluation based on how well they responded to those price signals—then you might have thought that turning over the firms to private ownership would not be the highest priority on the social agenda. But, in fact, it did have high priority.

And then, when you looked at how privatization was being done, you saw that lots of side conditions were being imposed. For instance, someone who bought a government-sponsored enterprise wouldn’t be allowed to lay off workers. So if you thought the goal of privatization was to enable a change of management toward people who would do what someone who acquired a failed enterprise in the United States would do, well, that was explicitly barred.

Actually, the governments doing privatization were trying to have their cake and eat it too. They were trying to force the new owners of the firms to continue the kind of social welfare programs administered through firms that made those firms unprofitable, while hoping that the mere transfer of ownership under those constraints would somehow miraculously make the firms profitable. It didn’t happen.

So the gist of my argument was that this was really a public-finance problem. The governments, as I just characterized them, were using state-owned enterprises essentially to administer social welfare—particularly, to take care of older workers who simply didn’t have the job skills to function in a capitalist economy. That was a cost that the government was going to have to pay one way or another. And if you privatized the firms as genuinely entrepreneurial firms without these social welfare responsibilities, that cost would have to be taken care of in some other way by tax funding.

You might believe for various reasons that the existing firms were well designed for playing that social welfare role. For example, their workers weren’t totally unproductive, although they were less productive than foreign workers who competed with them once the economy had opened up. Then it doesn’t seem like a bad idea just to keep the existing state-owned firms doing what they’ve been doing and subsidizing them to continue operating. Subsidizing those workers might be less costly than the alternative—putting them permanently on the welfare rolls—and the workers might be happier continuing to have dignified employment.

And instead of privatizing the old state-owned firms, let new small firms form to be truly entrepreneurial. Over time, let that sector grow. As the demographic bulge of the workers who, through no fault of their own, grew up in the communist regime and had to be subsidized for the rest of their life cycle drew down, you would then get your transition to an entrepreneurial private-ownership economy.

Region: It’s been 12 years since you developed that theory and, perhaps, prediction. Is that long enough to get a sense of what’s evolved?

Green: I think the observation that’s really in its favor is China. What’s happened in the former Soviet bloc area is a mix of results ranging from Russia, where my impression is things haven’t really gone well at all, to places like the Czech Republic that are quite prosperous on the whole and going down a path of integration into the European economy. But the most fortunate countries have had moderately high rates of growth, and the less fortunate countries have had disastrously low rates of growth.

In contrast, China has kept its existing state-owned enterprises while doing two things. One is to allow private ownership in various sectors, so that new small firms have formed. The other is they’ve written contracts with state-owned firms or divisions of state-owned firms that have basically said to the firm managers: You have to take care of your current workers, you have to pay us a fee for use of your capital, but you go out and do what you want. Whatever you have in excess of the fee is your profit.

Now, that’s been costly, and they’ve absorbed a lot of the cost in the banking system. They’ve made loans to state-owned enterprises that they knew to be insolvent. They’re insolvent because of these social welfare costs, and the Chinese have now reached a day of reckoning when they know that they have to make banking reforms. And they’re making some very intelligent reforms.

For instance, earlier this year China made injections of government capital into several of its state-owned banks and required a quid pro quo. The banks were required to adopt governance reforms that make them a lot more transparent. These reforms will put the banks on a course of shifting their lending from old-style disguised subsidy to genuinely profitable lending. Of course, the Chinese government and economy have huge resources now because of the trade surplus and so forth, so they’re in a fortunate position of being able to apply those resources to get the banking system into a state of health. The way I would tell the story is that the Chinese have made social welfare expenditures in order to take the pressure off their developing entrepreneurial system, and they’ve done some of that on credit—by credit, I mean what amounts to government borrowing through the banking system—and now it’s time to restore the long-run budget constraint.

CONTRACTS

Region: Your 1987 work on contracts as a means of smoothing consumption over time has stimulated a great deal of research on dynamic contracting and efficiency. What implications does your model have for trends in income distribution?

Green: I formulated this model in order to improve my understanding of Milton Friedman's permanent income hypothesis. Two other economists, Jonathan Thomas and Tim Worrall, independently studied virtually the same model.

Friedman said something along the lines of, Households have a view of what their average income is going to be over the next half dozen years or so, and when there are year-to-year fluctuations around that, they either borrow in low-income years or save in high-income years. It's very much a theory of short- and intermediate-range decision-making. Subsequent economists formalized Friedman's intuitive analysis by assuming either that there is no insurance at all against these random, year-to-year fluctuations—so that people have to insure themselves by holding "precautionary—wealth to spend on a rainy day—or else that people can get full insurance.

It seemed to me that the extent of insurance ought to be something that a good theory predicts, rather than something that is assumed as a starting point. Moreover, it seemed to me that partial insurance, rather than either of the extreme cases of no insurance and full insurance, is what exists in practice. I modeled a simple economy for which I could study what the efficient allocation would look like, without making any confining assumptions about its features or about what type of contract would implement it. A kind of partial insurance turned out to be the answer.

Because of various technical features, my model had an implication that the inequality of consumption goes to infinity as time approaches infinity. Robert Lucas and Andrew Atkeson, in particular, emphasized that long-run feature of the model. They used it as the basis for a theory of income distribution and a welfare theory of optimal income distribution. That's a substantially different application than I was making of the model. And I have some reservations about that application on a couple of scores.

One is that the model in the form that I analyzed it, and that Atkeson and Lucas have analyzed it, doesn't have human capital. Particularly when we think about policy to ameliorate inequalities of income distribution, policies regarding human capital are what come to mind. For example, the work that the Minneapolis Fed has been doing to acquaint citizens with the benefits of early childhood education is in that sphere, but it couldn't be framed in the simple version of my model.

The other respect in which I have some qualms involves the way this model gets interpreted, as a model of what is called the dynastic family. There's an initial set of parents, Adam and Eve, who have children, and the parents get utility from the child's utility, so ultimately it's just Adam and Eve whose utility is optimized. The welfare lesson drawn from this dynastic-family analysis is that in order to make Adam and Eve as happy as possible, we have to tolerate a world in which in future generations some of their children are driven into extreme poverty in order that other children can enjoy the benefits of high consumption.

Well, as a social welfare criterion, I'm unhappy with the dynastic family. Adam and Eve's great-great-grandchildren receive zero weight in the social welfare decision except to the extent that Adam and Eve feel sympathy for them or altruism toward them. The ethical position that I'd prefer to take is something like the position that John Rawls took in A Theory of Justice, where he said society ought to maximize the welfare of the worst-off person. And the worst-off person is one of Adam and Eve's great-great-grandchildren, not Adam and Eve, whose welfare is being maximized in the dynastic family.

So I think the model is very successful for the uses that I made of it. And by the way, Chris Phelan at the Minneapolis Fed has done a lot of research showing that this asymptotic spreading apart of consumption levels and welfare levels is a fragile conclusion from my model [see "Class Action" in the September 2005 Region]. Different assumptions about the shape of utility functions and other technical things will give different conclusions.

On the basis of Chris' work, I feel comfortable in saying that model of what I called uninsurable income is a good model for studying relatively near-term issues but is not the model of choice for studying long-term welfare distribution. One doesn't logically have to accept the long-run implications of my particular model in order to use the model to analyze short-term issues. The assumptions that have extreme implications about the long run are made just because they make the model easier to analyze. We know that they can be changed in sensible ways that get rid of those extreme implications but don't affect the implications pertaining to Friedman's ideas.

A NEW JOURNAL

Region: You are co-editor of Theoretical Economics, a new journal. Its inaugural issue is in March, I believe.

Green: That's correct. Next spring anyway.

Region: Can you tell us about it? What are its goals? What will make it different from other journals?

Green: Well, this is one of a number of journals starting in various academic disciplines that are being called open-access journals. Like other journals, they are peer-reviewed—that is, a researcher sends an article and it's sent to other researchers who verify that it's correct, significant and worthy of publication.

But unlike virtually all other journals—the Minneapolis Fed's Quarterly Review being a notable exception—Theoretical Economics is not published under a copyright. Or more technically, it's published under an extremely permissive copyright called the Creative Commons copyright. So while today commercial publishers are, for instance, charging $25 to download from the Web an article published in a recent issue of an economics journal, our journal is going to be up on the Web for free, with no subscription required.

Those of us on the editorial board of this journal take to heart the view emphasized by [Stanford University economist] Kenneth Arrow that information is a public good. The efficient price for readers to pay is zero. By the way, we are running the journal economically and charging authors a $75 submission fee that's going to cover our operating costs. Many other journals also charge submission fees in this range in addition to their subscription fees. We've recruited an enthusiastic and experienced editorial board that includes, among others, half a dozen former editors-in-chief of first-rank economics journals.

I think this journal is going to be a focal point for economic theorists. A journal survives and prospers first, by persuading the authors of the best papers to submit them to that journal, and second, by handling those papers professionally. Our editorial board has the credibility with leading economists to do the former and the intellectual depth and breadth of experience to do the latter.

Region: We look forward to seeing it. And thanks so much for your time.

Green: It's been a real pleasure.