Edward C. Prescott
|Edward C. Prescott is an economics professor at Arizona State University, and for 23 years has been a research adviser at the Federal Reserve Bank of Minneapolis. He and Finn Kydland, a professor at Carnegie Mellon University and at the University of California, Santa Barbara, won the 2004 Nobel Prize in Economic Sciences for their contributions to dynamic macroeconomics namely, their work on the theory of economic policy and their research on the causes of business cycles. Their papers on these subjects, grounded in the dynamic general discipline, helped move the economics profession decisively away from traditional Keynesian macroeconomics, which saw inflation and unemployment as a trade-off, the two ends of a seesaw that could be controlled by monetary policies. Prescott received his Ph.D. from Carnegie Mellon in 1967. This interview was conducted in October, shortly after the Nobel Prize announcement was made.|
Prescott and Kydlands 1977 paper, Rules Rather Than Discretion: The Inconsistency of Optimal Plans, found that since individuals and firms make decisions based on their expectations decisions about future policy decisions, selecting the policy action that is best in the current situation does not result in best outcomes. This research has led central banks around the world to strengthen their institutions and their independence in order to pursue optimal economic policies policies that are not time-inconsistent, to use the terminology the two researchers used.
Prescott and Kydlands path-breaking 1982 paper on business cycles, Time to Build and Aggregate Fluctuations, changed the way economists look at business cycles. Their research established that productivity and other real shocks account for business cycle variations. Given that business cycle fluctuations are optimal responses to persistent changes in real factors, they found that stabilization policy is misguided. In their models, once these misguided policies were largely abandoned, the magnitude of business cycle fluctuations declined dramatically; in addition, the focus of fiscal policy shifted to important long-term considerations such as the framework of a good tax system and the amount of money that should be allocated to human and infrastructure investments and to welfare. This paper also pioneered a methodology for using the dynamic general equilibrium to quantitatively address business cycle and other macro questions.
you and Finn Kydland set out to write what became Rules Rather Than
Discretion: The Inconsistency of Optimal Plans, did you plan to write
the paper you wound up writing?
Prescott: No, just the opposite. My dissertation was about control theory how to use control theory and macro econometric models to better stabilize the economy. Lucas in 1974 wrote his paper Econometric Policy Evaluation: A Critique [the paper was published in 1976] that found that these macro econometric policy evaluation procedures are inconsistent with everything we knew about dynamic economic theory.
I accepted the critique and set to work developing a control procedure that
is consistent with dynamic economic theory. First we used recursive methods
to find time-consistent policies. Essentially, the action taken at each
point in time was best, given that actions would be similarly selected in
the future. Then the insight hit us when we realized, Oh, its
time-consistent but not optimal. That was the punch line. It had major
implications because policy is a game where you have to talk about rules.
We wound up saying that setting rules is better than having discretion.
Being able to tie your hands so you cant make changes to try to achieve
short-term goals makes it possible for you to better serve as an agent of
FEN: I know the paper had a significant impact on the economics profession pretty quickly, but how long did it take before the broader policy implications exerted a tug on monetary policies?
Prescott: Needless to say, the Federal Reserve System likes independence and wanted independence. And they loved having theoretical support for it. One of my former students, Ed Green, was an economist at the Board of Governors. Ed introduced Ken Rogoff, who was also there and who is now a Harvard professor, to the time-inconsistency problem. Both of them had an important impact educating people as to the time-inconsistency problem and the value of having an independent central bank to mitigate the problem.
The Fed had lost its independence in the 1970s, or maybe earlier. Monetary policy had very undesirable outcomes in the late 1970s, with high inflation and high unemployment. There was no trade-off there, as Lucas said there wouldnt be. Fed Chairman Volcker deserves a huge amount of credit for bringing back the Feds credibility, and I guess President Reagan as well, for not undercutting Volcker. Since then, the independence of the Fed has been maintained.
FEN: Have there been hiccups or times since then when the Fed was seen as less credible?
Prescott: The independence of the Fed was attacked by President Clinton in 1996 or 1997. He wanted to take away all of the Feds regulatory responsibilities. A president not worried about getting reelected benefits from an independent Fed. Its a way to tie the presidents hands. There can be all sorts of rationalizations in the other direction, but in the end it results in bad outcomes. When it gets really extreme, you find yourself in the Argentine case. There are lots of examples of hyperinflations and defaults in Europe and elsewhere. When we get a couple of generations removed from all this, will criticism of the Feds independence raise its ugly head again? I dont know.
FEN: One of the Federal Reserves goals is to maintain price stability. Over the last 15 years, a number of countries, including New Zealand, Sweden and the United Kingdom, have instituted a policy of inflation targeting. What would be gained or lost if the United States did that?
Prescott: The Federal Reserve doesnt decide what the objectives are. The elected representatives decide that. If these elected representatives said the responsibility of the Fed is to maintain a low, stable, predictable inflation rate, I would be happy. I think the research Finn and I have done and hes the expert on the monetary side shows that except for Lucas-type monetary surprises, monetary policy has little effect. These surprises cant be used to stabilize the economy. You can surprise people but you cant predictably surprise people. If you try, you get into situations where people are trying to second-guess each other, and then you get into a game with bad outcomes for all. If you have a good explicit or implicit rule and I think the current implicit rule is a good one there is reasonable price stability.
FEN: If Congress did as you suggested and specified that the Fed should maintain a low, stable, predictable inflation rate, what would the benefits be?
Prescott: Price-level uncertainties lead to redistributions of wealth between borrowers and lenders. If you happen to be lucky, youre happy, but if theres unanticipated inflation, money goes from the lenders to borrowers, from the old to the young. This uncertainty leads to some mutually beneficial borrowing and lending contracts not being entered into.
FEN: How does money shift from the old to the young?
Prescott: Theres a large amount of borrowing and lending. People take out 30-year mortgages. If theres a modest amount of unanticipated inflation, the real value of their liabilities goes down, so theyve won. If they win on this, someone else has to lose. Who loses? Probably some old people who have money in pension funds, because its the pension funds that hold these mortgages.
FEN: Does the huge growth of the fixed-income markets, and the mortgage markets in particular, have any impact on the Feds decisions about interest rates?
Prescott: Thats not relevant. Whats important is productivity. If productivity grows dramatically, as was almost surely the case in the high-tech boom of the late 1990s, the real interest rate will be high. Evidence of this is that the return of inflation-adjusted bonds was high in this period. But this was due to real, not monetary, factors. All that the Fed affects is the short-term interest rate relative to the long-term rate, and, effectively, the inflation rate. If inflation becomes a problem, the Fed increases the interest rate on short-term debt relative to the long-term interest rate.
FEN: Were you concerned when the federal funds rate was recently at its lowest level since World War II that the Federal Reserve might have a hard time fighting deflation, if that became necessary?
Prescott: No. Some people worried about deflation and how the Fed could lower the rate further if the price levels started to fall a little bit. As long as the deflation is not a big surprise, real consequences are minimal. The U.S. had a period of rapid growth in the late nineteenth century when there was deflation. Norway had spectacular growth and rapidly falling prices over much of the interwar period.
Japan has been growing rapidly in the last two years, after losing a decade of growth. Japan lost that decade of growth because productivity did not grow. Fumio Hayashi and I found that that was the case we said that once Japans productivity started growing again and the government moved away from anti-growth policies, Japans productivity would pick up. This is what happened over the last two years there.
FEN: Anti-growth policies like regulatory barriers?
Prescott: I dont have good quantitative numbers on this, but it appears that they did less subsidizing of inefficient producers. If youre a business and youre thinking of entering an industry, and you know your inefficient competitors will be subsidized, why enter? You cant make a profit. On the other hand, if theyre not going to be subsidized, then you enter and youre the more efficient one. You make money and the existing firms say, Uh oh, if we want to survive we have to be more productive, and they get more productive and then everybody is better off.
FEN: How did all the bad bank loans fit into this?
Prescott: They had a major financial crisis associated with the collapse of their stock market and the collapse of the price of their land. Their banks became insolvent. Their liabilities exceeded their assets. My bet and conjecture is that the government effectively subsidized the inefficient producers because the banks didnt want to write off the loans. And this was done with the encouragement of the bureaucracies. Banks were protected from competition from new entrants, in particular from foreign banks. The industrial banking and bureaucratic complex is the reason Japan has significantly less output per hour than America and western Europe. In industries where the Japanese export, they produce more per hour; but in their service industries, Japans output per hour is half the American level. A little foreign competition would be a great thing for the Japanese people.
FEN: How did you and Kydland start working on your 1982 paper on the causes of business cycles?
Prescott: In the 1970s everybody thought that monetary shocks were the reason the economy fluctuates. That includes Finn and me. But we were influenced by Ragnar Frisch, who stressed the importance of a propagation mechanism for cyclical fluctuations, and there was not a monetary shock propagation mechanism that resulted in the economy displaying the business cycle facts.
[Frisch and Jan Tinbergen won the first Nobel Prize in Economic Sciences in 1969 for developing dynamic models to study the economy and for launching the field of econometrics. A propagation mechanism is something that specifies how the consequences of a shock to an economy gradually unfold or propagate through time.]
We had informational distortions in our model to mimic monetary surprises in the economy signal-extraction problems, I guess, like in Lucass Expectations and the Neutrality of Money. We had Frischs time to build the time it takes to plan and build a new factory or office building and hoped it would be the propagation mechanism. We found out it didnt work. But it turned out that highly persistent changes in real factors mostly productivity shocks gave rise to economic fluctuations of the business cycle variety.
By the way, it turns out our real-shock story is an old one Knut Wicksell and Arthur Cecil Pigou were famous economists who adhered to that view. I just got a nice little note from Paul Samuelson telling me to look at Wicksell, so I got Wicksells book out of the library. [Wicksell was an eccentric Swedish monetary theorist and Malthusian whose influential 1898 book Interest and Prices contributed to the emerging field of macroeconomics.]
Just about everybody thought that economic fluctuations were inconsistent with the real-shock view. The following question and answer based on static partial-equilibrium-type reasoning is the reason: Why should productivity be high when the labor input is high? Production theory says that when more labor is employed, labor productivity that is, output per hour is lower. The real wage doesnt move around much over the cycle; therefore, we should see leisure and consumption moving in the same direction, or consumption and labor supply moving in opposite directions. The facts are that when labor supply is high, consumption is high. Here I mean high relative to trend. But the business cycles turned out to be just what dynamic economic theory predicted. It was a total surprise.
FEN: What was your and Kydlands reaction when your model wasnt what you had anticipated?
Prescott: Initially we were disappointed when we found that labor supply errors along with time-to-build did not give rise to business cycles. Then we varied the variances of some of the shocks, and business cycles appeared. And then the light got turned on. Persistent real shock gave rise to business cycle fluctuations.
We knew what the business cycle facts were for the U.S. economy. For each model economy studied, we computed its equilibrium process and used that process to generate time series of the economic variables. We then checked to see whether that economy displayed the business cycle facts. We found that if there were persistent real shocks and the labor supply was sensitive to the real wage, a model economy displayed the business cycle facts. Later we found that productivity shocks were persistent and of the magnitude that gave fluctuations of the size observed.
In 1984, there was an important dissertation by Richard Rogerson; I was his adviser. He did some aggregations where people had the choice of working or not working. As the result of aggregation, he found that aggregate labor supply is highly sensitive to the real wage even though individual labor supplies are not. In 1985 Gary Hansen put the Rogerson aggregation into the business cycle model that Finn and I did, stripped away the unimportant stuff, and wrote a beautiful, clean paper that made everyone understand the business cycle theory that Finn and I developed.
FEN: Your 1982 paper on business cycles led to a bumper crop of papers on the same topic. But theres still controversy about whether it was supply-side shocks that really caused particular business cycle changes. If you were writing the paper now, would you accommodate any of the criticisms, or are they missing the point?
Prescott: Those people ceased to be economists. We dont use supply and demand in macroeconomics anymore. They are partial-equilibrium concepts. If you see a purported macro economist using those terms, ignore him. We use peoples ability and willingness to substitute to organize our empirical knowledge. Economists should organize their empirical knowledge around these data whether they are macro economists, students of public finance, students of international economics, or growth theorists. Given policies and the ability and willingness to substitute, economists predict what will happen.
FEN: Heres a broad, overarching question. What do the issues youve focused on have in common? Or, to flip the question around, how would you describe your approach to problems across a range of different areas?
Prescott: The magic thing is general equilibrium. The models always start off with people with preferences that specify what they will choose when given a choice and technologies specifying the ability to transform inputs into outputs. The structure that I use heavily is the growth model with its decision about how to split output between current consumption and investment, and its decision about how to split time between market activities and non-market activities. It is amazing how useful this simple abstraction is.
Ive come to love the national accounts. Without these economic statistics there would be no macroeconomics. I love modeling and coming up with explicit structures that I can play with and use to build economic intuition. I worry about the interaction between theory and the reported statistics. Another issue is what model economy to use when addressing a given question. People want a mechanical procedure for model economy selection. Im pleased that theres not such a procedure, because if there were it would put me out of business.