Academic journal article Federal Reserve Bank of New York Economic Policy Review

Policy Rules and Targets: Framing the Central Banker's Problem

Academic journal article Federal Reserve Bank of New York Economic Policy Review

Policy Rules and Targets: Framing the Central Banker's Problem

Article excerpt

Central bank policymakers are not primarily random number generators.(1) Reading both the financial press and the work of academics, however, one might get the opposite impression. Reporters (and the readers of their stories) seem to attach considerable importance to each Federal Open Market Committee policy decision. Academic work on the impact of central bank policy gives a similar impression, as statistical procedures produce a time series of pure white noise innovations that are labeled "policy shocks."(2) But central bankers expend substantial energy attempting to tailor their actions to current economic conditions. In other words, policymakers are reacting to the environment, not injecting noise.

But what is central bank policy anyway? The policymaker's problem can be characterized in the following way. Using an instrument such as an interest rate, together with knowledge of the evolution of the economy (aggregate output and the price level), the policymaker seeks to stabilize output and prices about some path that is thought to be optimal. In carrying out this goal, the policymaker must often trade off variability in output for variability in prices because it is generally not possible to stabilize both. This process yields what most people would call a policy rule, that is, a systematic rule for adjusting the quantity that the central bank controls as the state of the economy fluctuates. In other words, the study of policy should focus on the systematic portion of policymakers' actions, not the shocks.

In this essay, I discuss a number of conceptual and practical issues associated with viewing policymaking in this analytical framework. These issues include the implications for policymaking of the slope of the output-inflation variability trade-off, the influence of various types of uncertainty on the policymaker's problem, the consequences of the fact that the nominal interest rate cannot fall below zero, and possible justifications for interest rate smoothing.

Although my intention is to raise, rather than resolve, key questions concerning the formulation of a policy rule, I do offer important new evidence on one point. This concerns the potential consequences of the move by many central banks to adopt some form of price-level or inflation targeting. In taking this approach, central banks are implicitly altering the relative importance of inflation and output variability in their objectives, increasing the weight they attach to the former relative to the latter. But the data suggest that the output-inflation variability tradeoff is extremely steep, implying that an effort to decrease inflation variability modestly could lead to a significant increase in output variability. Thus, policymakers considering pure inflation targeting should be aware that their change in emphasis could have undesirable side effects.


As I suggested in the introduction, central bank policy can be thought of as the solution to a problem in which the policymaker uses an interest rate to stabilize the variability of output and prices about some path. A truly complete description of the policymaker's problem begins with an intertemporal general equilibrium model based on a social welfare function (tastes), production functions (technology), and market imperfections that cause nominal shocks to have real effects (nominal rigidities). The goal would be welfare maximization.

I do not propose to delineate the fully specified problem. Instead, I begin with a commonly used quadratic loss function that might be a second-order approximation to the objective function in this more detailed problem.(3) The policymaker seeks to minimize the discounted sum of squared deviations of output and prices from their target paths. The general form of such a loss function can be


where [p. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.