London School of Economics

March 25, 2013

It is a privilege and an honor to be here today. I met Mervyn King for the first time in 1975.  I was struck at that time that he had the highest product of intellectual acuity and elegance of accent of anyone I had ever met. And that remains the case today.

Mervyn is a person whose thought and effort has ranged widely, and that is surely appropriate. There is an important difference between physical science and economic science. Physical science tries to understand a constant world better and better, a world that is not affected by the product of that science. Economic science, on the other hand, seeks to understand a changing world, a world whose functioning is importantly affected by economic thought. That is surely the arc of Mervyn King’s career as a central banker – early on, playing a seminal role in the design of a highly influential inflation targeting regime, and then, as conditions warranted, playing a critical role in thinking about how to respond to profound financial crisis.

I want to reflect on two broad subjects today. First, the ways in which I believe this crisis will force a substantial reconstruction of macroeconomics. Second, and overlapping somewhat with what Olivier said, the ways in which I believe this crisis will, over time, redefine the role of central banks.

A premise of virtually everything I studied in graduate school, and virtually everything I taught as a professor of macroeconomics for some years, was that a coherent model had an equilibrium, and that if conditions changed, it would move to a new equilibrium.

Surely the events of the last few years call that proposition into question.  If the world had unfolded forward from 2007 with no policy actions taken, no lending of last resort, no expansionary monetary policy, no expansionary fiscal policy, I would suggest to you that there is a real possibility that the right approximation would have been an unbounded downward spiral, a possibility ruled out in any textbook model or almost any model published in a journal in the last several decades.

Related, almost every treatment of macroeconomic fluctuations over the last 30 years, whether of a so-called new Keynesian variety or of a so-called classical variety, has presumed that output fluctuated. Specifically, they presume that it fluctuated around a trend that was determined from somewhere. If one managed policy badly, there would be more volatility. If one managed policy well, there would be less volatility. But the average level of output would not be affected by stabilization policy. That led many to believe that stabilization policy was a second order question.

The events of the last six years surely demonstrate that the more important view is that macroeconomics was about avoiding the waste associated with recessions. For no one could seriously suppose that if more enlightened regulatory and macroeconomic policy had mitigated what we have experienced for the last six years, the resulting lost output and lost employment would somehow have been shaved off some future growth. And so the significance of stabilization policy becomes that much greater. The type of model that is then necessary, a model of periodic malfunction, becomes quite different from a model that presumes a constant mean, and discusses only the magnitude of cyclical fluctuations.

In the same way, the events of the last years suggest that the traditional breakdown between the cyclical and structural that is central to much of macroeconomic thinking has become highly problematic.

If there’s anyone in this room who believes that a reasonable forecast for the US or British economy at any future date would be represented by a trend line constructed through 2007 from any previous point, I have a bridge that I would like to sell you. Again, we are learning that cyclical developments can have very long term structural implications.

And of course, there is the observation that Olivier stressed, that the financial sector is central to understanding at least the most consequential macroeconomic fluctuations, the kinds we have just observed.

It almost makes me wonder whether Keynes badly mistitled his book.  Perhaps instead of titling it A General Theory of Employment, Interest and Money, he might have titled it, A Specific Theory of Employment, Interest and Money in Conditions of Depression and Liquidity Trap. Had he done so, there might have been two important benefits – the avoidance of the substantial errors of overexpansion and inflation carried out in the 1960s and the 1970s in the name of Keynesian policy, and a greater willingness to heed some of the lessons of Keynes during this most recent, difficult period. This would be a very different macroeconomics than the macroeconomics I studied, or the macroeconomics that has been taught in the past.

What about central banking?  Central banking, as it was understood during my time in the Treasury, was complex in its implementation, but was not so complex in its basic vision.  It was understood that surprising people with a little inflation was always a temptation, because it would generate substantial output, but that it was a fool’s game, because when anticipated, you get the inflation and not the output. Therefore, several measures should be taken to assure credible low inflation while allowing some room to contribute to improved economic stability, perhaps along the lines of a Taylor Rule or some such device. Responsibility for that function was to be assigned to society’s most rectitudinous figures, such as Mervyn. These figures were to be insulated from political pressure and made accountable. It was a system that was implemented in different ways in different countries, and there was much discussion in the central banking community of different mandates.  But the idea was always the same.

But there is no role in that thinking for most of what has preoccupied us for the last five years, the provision of finance in extraordinary ways. Make no mistake – if a loan is absolutely riskless, with absolute rigid certainty that it will be repaid, someone in the private sector will make it.

And so when last resort financing is provided because no one other than the central bank will provide it, it is because of some possibility in some contingency that it will not come back. That means taking credit risk. That is okay, but it’s not quite the classic central banking theory.

And yet, what central banks did in the 1930s, what central banks have done in the last six years, has probably been more consequential than everything they did in between.

Therefore, when central banks are most important, they are doing something for which their complete insulation raises more questions.  Similarly, there are questions surrounding the interaction between monetary and fiscal policies.  These arise when central banks pursue schemes directed at stimulating investment in particular sectors, or stimulating lending in particular areas.  An area that has always seemed to me most puzzling is the question of the maturity of a country’s debt. It has been the case in more than one country in recent years that the central bank commits itself to shortening the maturity of the public sector’s outstanding debt in the name of economic expansion. Simultaneously, Finance Ministries commit themselves to lengthening the maturity of the public sector’s outstanding debt in the name of assuring low, long-term funding costs. And one is left to wonder how these decisions are being made, and whether they should be made in some consolidated way.

There is also another question that Olivier referred to, of prudential macroeconomic supervision.  Surely, whether you need to have as many targets as instruments, it is a reasonable judgment that if you have more targets, it is good if you can think of more instruments.  And as the aspirations for the kinds of problems that central banks are going to avoid proliferate, there’s an inevitable desire for them to have more instruments.

And yet, those instruments – changing the down payments that people pay when they buy a house, for example – inevitably touch on issues of concern to voters and to the political process.

There is also the set of issues involved with exchange rates. I find myself very much in agreement with the very important distinctions that Ben Bernanke drew a few moments ago between easing monetary policies in ways that lead a currency to depreciate and the imposition of a tariff in terms of its impact on the global system.  I think those distinctions are entirely correct. Nonetheless, the question of exchange rates is a question of international significance, and in some cases, of foreign policy importance.

I do not doubt for a moment the importance of independent central banking.  And I think it is very important to remember that just as the woes of the last five or six years reflect, in part, the fact that the lessons of the distant past about financial stability and the need to maintain demand were forgotten, there is the risk that in the aftermath of this episode the lessons of the 1980s and 1990s about the importance of credibility in resisting inflation will be forgotten as well.

Nonetheless, I believe there is much that will need to be charted with respect to the role of central banks in free society.  It has been my observation that one can speak freely – more freely, at least – and think more widely outside of public office than in public office.  Or academic office, for that matter.

I think, therefore, if there is a silver lining in the fact that Mervyn King will no longer be in an official position of public service, it is that, as a free man, he will be able to devote his formidable intellect to these very, very important questions.

Thank you very much.