Rational Expectations and Inflation
Published December 1, 1997 | December 1997 issue
By Thomas J. Sargent
HarperCollins College Division
(Since acquired by Addison Wesley Longman)
This is the first time that I have been invited to contribute a book review to the type of publication that would run a transcript of a seance with a long-dead legislator as its lead article. I would like to thank the editor of The Region for giving me this unique opportunity.
Of course, this issue's "interview" with Carter Glass actually has a point that is more serious than its format. The point is that a central bank lives in a political environment, and that it must be designed with care to survive and operate effectively in that environment.
Indeed, designing the Federal Reserve to be such an effective central-banking system had been a major part of Carter Glass' life work.
The relationship of a central bank to the broad political system and its role within that system are delicate issues for central bankers. It is good public relations for them to portray themselves as acting in the public interest according to their own lights, but also as being duly subservient to the democratic process. The words "independence" and "accountability" figure heavily in the rhetoric of such discussions. The problem is that there is, if not an outright inconsistency, at least a strong tension between independence and accountability.
Consider what happened during the period between the late 1960s and early 1980s, for example. Congress, the president and much of the public initially desired the Fed to follow an expansionary monetary policy as a way to continue funding domestic social programs and also to fund the undeclared Vietnam War without raising taxes in the short run. The Fed did follow such a policy. It continued to do so for a decade during which that continued to be the widely preferred policy, but during which economists consistently warned that it was an unwise policy. Thus, the Fed was accountable to the political system but was not sufficiently independent of it. Finally, after inflation had risen high enough to be frightening, the Fed changed to a tight-money policy that had a foreseeable risk of leading to a recession.
When the switch was made, the president and most legislators and most voters viewed that risk as being warranted, and they seemed to understand that inflation could be controlled only if the Fed would persevere in its policy despite whatever problems it would cause in the short term. When a recession did come and proved to be serious, though, public opinion turned against perseverance but the Fed did not much change its course. The Fed thus demonstrated independence, but there is a clear sense in which it failed to be accountable. In fact, at one point in 1982, leaders of both political parties in Congress were calling for the removal of Chairman Paul Volcker from office, and the president and members of his administration were also making pointed criticisms of his conduct of policy. By any normal standard, such widespread condemnation of an agency's head is convincing evidence of the agency's lack of accountability. This history shows that, when push comes to shove, a central bank simply cannot be both completely independent and completely accountable at the same time.
In Rational Expectations and Inflation, Sargent provides a consistent way to think about the relationship between a government and its central bank, and he justifies that way of thinking via a number of historical case studies. Rather than emphasizing terms such as "independence" and "accountability," which suggest properties that the central bank has in isolation, Sargent emphasizes the notion of "strategic interdependence," which is a relationship between three types of "player." These are the central bank, the government and private economic agents. Each player has a strategy, which is a long-term, contingent plan for its conduct in the present and in the future in various eventualities. Each player understands the others' strategies, as well as its own. Moreover, each player's strategy is chosen to be optimal for that player, in view of the player's understanding of the other players' strategies.
Strategies have to be credible. At a minimum, credibility requires that, at a future date, a player must actually have incentive to take actions that its plan specifies to be taken at that date. (A plan that does not meet this test cannot be made credible simply by announcing it, because other players will expect the player in question to renege on it despite the announcement.) To say that the central bank, the government and the private sector are strategically interdependent means that their plans are interrelated in this way. I am afraid that this thumbnail sketch may be cryptic, but Sargent explains it clearly and nontechnically, even using a homespun analogy with a football game that runs through several chapters of the book.
With this concept, one can state clearly what is the role of a central bank. The existence of a central bank with appropriate powers can affect which policies of the government are credible, making some desirable plans possible to implement and making some potentially undesirable plans impossible to implement. The most striking example is the implementation of strategies designed to end hyperinflationsinflationary episodes so drastic that people's willingness to accept money in trade for goods is largely undermined. In his chapter "The Ends of Four Big Inflations," Sargent documents how this worked in the case of hyperinflations in four central European countries after World War I. He shows that hyperinflations were indeed harmful to the citizens of those countries, and that they were ended when two steps were taken. One was establishment of a central bank that was legally committed to refuse the government's demand for additional unsecured credit beyond what was already issued. The other was adoption of fiscal policies that realistically could fund retirement of the already-issued debt without resorting to reduction of its real value through inflation.
Sargent emphasizes that it was a change in institutions, and in their relationship to one another, that made the control of inflation possible. To make such a change in institutional design is a very different sort of policy than to take a one-time action such as redenominating the currency and contracting the money supply. Such one-shot reforms have been adopted, without much success on the whole, in some other attempts to bring hyperinflations to an end. The institutional-change approach is clearly distinguishable from the monetary-contraction approach in its effects. In cases where appropriate institutions have been adoptedeven at the point when their impending adoption has been credibly announcedinflation has declined precipitously. At the same time, contrary to what the one-shot approach to controlling inflation would have brought about, the money supply has risen in the short run. The reason is essentially that, once a credible program for long-run control of inflation has been adopted, people in the economy become willing once again to hold money. The effect of a hyperinflation is to drive the real value of the money stock (although not its nominal value) far below the level that is efficient for conducting economic activity, and the growth of the money stock following a successful attempt to control inflation simply reflects the return of the money stock's real value to its efficient level.
Sargent suggests that, in one respect, a hyperinflation is easier to stop than the sort of moderate inflation that numerous, relatively industrialized countries have experienced. The reason is that citizens in a country experiencing a hyperinflation are desperate to end it, and will thus give durable support to a plan that ends it with a reasonably equitable distribution of the costs of adjustment, while a reform designed to end a moderate inflation must be designed and implemented in a more politicized environment.
The problem in such an environment, which Sargent examines in the context of both the Reagan and Thatcher governments, is that the central bank's monetary-policy strategy and the government's fiscal-policy strategy can be mutually inconsistent. Sargent argues that the Reagan administration's dramatic expansion of the government deficit was plainly inconsistent with the Fed's monetary policy, and that the Thatcher government's gradualism in attempting to achieve fiscal restraint was not sufficiently credible (because of the British government's past failures to keep promises of restraint) for the Bank of England's policy to be tenable. He points to the French government's successful control of a moderate inflation in the mid-1920s as a model for how such a reform can be carried out more successfully. (Several subsequent researchers have taken issue with aspects of Sargent's view of the French case, but none has overturned the fundamental thesis that prompt, simultaneous establishment of credibility of the central bank and the fiscal authority is the essence of successful control of inflation. Sargent himself has included a new essay in the 1992 edition that somewhat revises the assessments of Reagan and Thatcher in the original essays, in light of developments subsequent to the original 1986 publication.)
Although it does not deal explicitly with the institutional features of the Federal Reserve System, Sargent's book provides a good understanding of why the institutional design of a central banksuch as the features of the Federal Reserve in relation to the U.S. financial system that figure so prominently in the "interview" with Sen. Glass elsewhere in this issueare important to the fulfillment of its mission. Although one of the book's chapters and sections of several other chapters are somewhat technical, it remains thoroughly intelligible if only its English-language text is read. Indeed, it is the best exposition of what monetary policy is all about, at this mostly nontechnical level, of which I know.
Although some of its material suffers from being too dated to retain its original journalistic aspect while not sufficiently distant in time to be fully satisfactory as economic history, Rational Expectations and Inflation on the whole remains fresh, stimulating and informative. I understand well that it is commercially impossible for publishers to keep books in print forever, but I am disappointed to learn that the publisher has decided to withdraw this one soon from publication. Readers of The Region should order it from their booksellers right away, while their orders can still be filled.