I am an academic economist, with some limited experience in actual financial markets. My knowledge of central banking comes mostly from occasions such as these – hardly a qualification to comment on monetary policy. I have done research on many different areas of economics, but no one would think of me as a monetary economist. However I spent most of my career at the University of Chicago where it is a given that all economic problems are amenable to treatment by a common set of tools, and for this reason my lack of credentials will not prevent me from speaking today.

Professor Issing talks about three levels of uncertainty. The first is uncertainty with respect to the state of the economy, which is by no means trivial, but relatively amenable to be treated in the classical framework of decision-making under uncertainty. The second is uncertainty about the structure of the economy – uncertainty about the suitability of different models to describe the functioning of the economy and about the relevant values of the parameters in the models. Finally Professor Issing considers the uncertainty that central bankers face about the reaction of economic agents as well as the uncertainty that the policies themselves generate among private agents. From a purely theoretical viewpoint the distinction between these two last categories may seem unnecessary. After all, at least since the rational expectations revolution, economists have understood that any reasonable economic model must consider explicitly the reaction of private agents to policies. A related point is that the economic agents have the same concerns about model misspecification that policy makers have. [2] However, I sympathize with Prof. Issing’s view that, in practice, because of the difficulty in understanding the formation of expectations, this third category represents a special challenge to policy makers. For this same reason, much of the current academic literature emphasizes the management of expectations at least as much as the direct consequences of any policy moves.

The focus on expectations increased the recognition of the need for predictable and systematic monetary rules. This is however not a new preoccupation. As my colleague Michael Woodford has stated “just a century ago, no one had any idea how to establish a reasonably predictable monetary standard except by guaranteeing the convertibility of money into a precious metal such as gold.” [3] Milton Friedman’s suggestion of a fixed growth rate for a monetary aggregate was motivated by the same quest – that of a monetary policy that produces a stable currency. The modern research indicates that monetary rules should define clearly the target variables, such as the rate of inflation, and the levels for these variables at which the rule would aim. In addition the monetary rule should delineate, as much as possible, how the central bank would use the instruments at its disposal to bring the target variables to the desired levels.

The target variables need not be in all cases reduced to the inflation rate, but there is no point in choosing targets that cannot be achieved with the instruments at the disposal of the central bank. In spite of the hot air produced at conferences like this one, I do not believe that the ECB can turn Paris into a tropical paradise. I also do not believe it can prevent all asset price bubbles while preserving the allocative function of capital markets.

The target levels signal to agents the policies that will be followed in the different states of the world. A Central Bank that announces “inflation will be maintained under 2%” signals a tough stand on inflation but little concern about possible deflation. On the other hand an announcement of a target rate “between 1 and 2%” shows a symmetric concern for the dangers of deflation.

Professor Issing is absolutely correct when he points out that the connection between actions on instruments and targets is model dependent and that central bankers need not base their policy moves on prescriptions of a single – necessarily wrong – model of the economy. A commitment to rules does not stop a Central Bank from using several models that produce different statistics on the “true” state of the economy and that suggest distinct policies at any point in time, provided it clarifies how these distinct models affect policy moves. The success of the policy rule in managing expectations will in any case depend on how well the Central Bank communicates its objectives and the models it uses to choose policies, and on the perceptions that market participants have of the quality of these models. These perceptions no doubt change with the evolution of the theoretical and empirical knowledge in economics. A policy maker may even, in some cases, override the model suggestions with judgment, but should bear in mind that every such move, even if rationalized ex-post, affects his ability to influence agents’ expectations.

At the end of his lecture, Professor Issing discusses the role of predictability in central banks’ policy in reducing uncertainty and volatility in financial markets. Recent research I have conducted with my colleague Wei Xiong at Princeton [4] builds theoretical models in which divergence of opinions generates asset price bubbles. In these models, I am willing to pay in excess of my own valuation for an asset because ownership of the asset gives me the option to sell that asset in the future to someone else that would value it more than me. Divergence of opinions may result from asymmetric information or because agents place diverse weights on the different signals that they observe. In any case, this divergence of opinions increases when signals that everyone observes and deem relevant become noisier. Under this view, a more volatile policy by the Central Bank, by increasing the divergence of opinions among agents would increase asset price bubbles. Central banks may not be able to stop all asset price bubbles, but should refrain from contributing to them.

At a recent event at the University of Chicago, commemorating Milton Friedman’s 90^th birthday, Federal Reserve Governor Ben Bernanke said to Milton, and Anna Schwartz who was also present, “Regarding the great depression. You’re right, we [the Federal Reserve] did it. We’re very sorry. But thanks to you, we won’t do it again.” Progress in monetary economics has given central bankers a much greater ability to achieve price stability and to avoid policies that increase macroeconomic fluctuations. Policy makers should all, as Professor Issing, follow the developments in the economic literature, for as Baudelaire said in his “Counterfeit Money,” another topic of interest to central bankers, “le plus irréparable des vices est de faire le mal par bêtise.”

J.A. S.