فشارهای سیاسی و انتخاب ابزار بهینه سیاست پولی

this paper extends Cukierman’s (1992) model of monetary policy discretion, private information, and credibility to an environment in which the quantity of money is endogenous and in which a monetary authority must choose between bank reserves or an interest rate as its instrument of monetary policy. This model is used to explore the determinants of credibility for the alternative policy instruments and to evaluate the manner in which political pressures on a monetary authority can influence the authority’s instrument choice. A key implication of the model is that such political pressures can influence the credibility of monetary policy differentially depending upon the choice of policy instrument, thereby inducing a monetary authority to choose an instrument that otherwise would fail to meet standard Poole (1970) criteria. Keywords: Optimal monetary policy; Monetary policy instruments; Monetary politics

Why do nearly all central banks use an interest rate as the monetary policy instrument? According to the instruments-targets literature, a central bank’s optimal instrument of monetary policy depends on the magnitudes of macroeconomic parameters and the relative sizes of the variances of structural disturbances. Specifically, since Poole’s (1970) original analysis, this literature [e.g., Friedman 1975 and McCallum and Hoehn 1983, and Benavie and Froyen (1983)] uniformly concludes that an interest-rate instrument is preferred to a bank-reserve (or monetary-aggregate) instrument if the variance of real expenditure disturbances is sufficiently small relative to the variance of financial-sector disturbances. Nevertheless, central banks in Europe and Asia consistently have used interest-rate instruments—even during periods in which Poole’s original condition probably has not been satisfied. In addition, even though Fair (1984) finds that simulations of Poole’s criteria for the late 1970s generally support the use of an interest-rate instrument by the Federal Reserve, in 1979 the Fed switched to a reserve instrument. Carlstrom and Fuerst (1995) propose a very different explanation that might help to explain the widespread use of an interest rate instrument even under circumstances in which the standard Poole rule is violated. They consider a model in which cash-in-advance constraints and portfolio rigidities produce distortions that can make a competitive equilibrium Pareto inefficient. Pegging the nominal interest rate eliminates these distortions and, therefore, is the optimal monetary policy. The reason is that an interest rate peg permits greater output responses to real shocks, thereby allowing agents greater flexibility to maximize their lifetime utility levels. This paper extends the literature about the widespread central bank use of an interest-rate instrument by presenting a model that shows that the variability of political pressures—or, more generally, the variability of any factors that influence central bank preferences—can influence instrument choice. In contrast to Carlstrom and Fuerst, the rationale we suggest for why central banks typically use an interest-rate instrument mirrors the traditional Poole approach in which a central bank seeks to stabilize inflation and real output in light of nominal shocks that have short-term real consequences. Nevertheless, the explanation we propose yields the standard Poole criterion only as a special case that emerges in the absence of variable political pressures. We are able to explain why an interest-rate instrument typically is more widely used when there are variable political pressures, and we can provide a rationale for why a central bank temporarily might switch to a reserve instrument, as the Fed did from 1979 until roughly 1982. Our explanation for these phenomena stems from developments in the recent literature on monetary rules, discretion, and credibility [e.g., Kydland and Prescott 1977, Barro and Gordon 1983a, Barro and Gordon 1983b and Persson and Tabellini 1990, and Cukierman (1992)]. It also draws on the political-monetary-economy literature as exemplified by the contributions of Havrilesky 1987 and Havrilesky 1993, who identifies structural elements, political linkages, and even personal characteristics that appear to have motivated Federal Reserve decision-making. The paper extends Cukierman’s (1992, Chap. 9) basic model of private information on changing objectives, discretion, and credibility. Following Friedman (1975), we consider a setting in which the monetary authority does not directly control a monetary aggregate. Hence, the model central bank, like those in the real world, must choose between either an interest-rate or a bank-reserve instrument of monetary policy. We demonstrate that whenever monetary policy is subject to control errors that real-world central banks also experience, the variability of political pressures can play a crucial role in influencing the choice of a monetary-policy instrument. Within the present model, as in Poole’s, the shapes and relative variations in the positions of the IS and LM schedules influence a monetary authority’s choice and day-to-day maintenance of a policy instrument. Nevertheless, the variability of political pressures can, because of its effects on central bank credibility, easily tilt the monetary authority toward the use of an interest-rate instrument instead of a reserve instrument. Below we show that the current widespread adoption of interest-rate instruments could arise because a conservative central bank, which leans toward achieving lower mean inflation and lower inflation variability even at the sacrifice of larger output gains, can enhance its credibility when it uses the interest rate as its policy instrument. The key factor that pushes a conservative central bank toward using an interest-rate instrument is the variability of political pressures combined with lower instrument control errors relative to the control errors associated with a reserve instrument. If the variance of political pressures is sufficiently high, whereas the variance of control errors under an interest-rate instrument is sufficiently low relative to that under a reserve instrument, then a conservative central bank may prefer an interest-rate instrument. Stated differently, if the variance of control errors with a reserve instrument is sufficiently large, then the central bank loses credibility in the face of variable political pressures. Such political pressures complicate private agent learning and thereby can make a reserve instrument less desirable than an interest-rate instrument even if the latter is preferred according to standard Poole criteria. Hence, this paper’s analysis implies that a relatively conservative central bank, such as the Federal Reserve, normally uses an interest-rate instrument in part because it gains credibility from using the instrument that can be implemented with greater precision. However, if there are no variable political pressures, the increased precision obtained by using an interest rate instrument is much less important.

The instrument choice model developed in this paper has two key implications. First, the standard Poole criterion for a central bank’s choice of the optimal monetary policy instrument represents only one aspect of a central bank’s policy problem when it faces political pressures. When the full effects of such external pressures are the central bank’s private information, a central bank also must take into account the credibility implications of its instrument choice. Second, to the extent that political-pressure and credibility considerations predominate in determining a central bank’s instrument choice, the model indicates that a conservative central bank may value the credibility-enhancing precision of a monetary policy instrument as well as its macroeconomic stabilization properties. To the extent that using an interest-rate instrument yields greater policy precision as compared with the implementation of a reserve-oriented policy procedure, this could produce a natural bias toward adopting an interest rate as the instrument of monetary policy. If the credibility gains from using an interest-rate instrument are sufficiently large, then a central bank would continue to use this instrument even if it might be able to achieve a stabilization gain by adopting a reserve-based approach to conducting monetary policy. At this point, we view the results of this paper primarily as suggestive of the considerable broadening of perspective on factors influencing the monetary policy instrument choice when political pressures are incorporated into a central bank’s policy problem. In work in progress we are considering the following issues. First, our comparison of losses above assumes that the central bank has not yet experienced political pressures when it makes its instrument choice. Yet such pressures could alter the central bank’s perception of output-inflation tradeoffs, thereby inducing the central bank to switch instruments more than once over time. Hence, further analysis using our basic model potentially could help to explain the factors that induced the Federal Reserve to switch policy instruments several times during the past 25 years. Nevertheless, the model in this paper at best provides a partial explanation for the Federal Reserve’s instrument switches in the 1970s and 1980s. One implication, for instance, is that the Federal Reserve may have switched to a reserve instrument in 1979 because the stabilization gain was sufficiently high to sacrifice the credibility-enhancing precision of an interest-rate instrument. Fair (1984), however, finds that simulations of Poole’s criteria for the late 1970s generally support the use of an interest-rate instrument by the Federal Reserve, which is inconsistent with this interpretation. This leads to the second issue that we are considering in further work. Our model in the present paper assumes that political-pressure shocks are incorporated directly into the central bank’s output-inflation preferences. If the central bank’s preferences are distinctive from those of the government officials who subject the central bank to such pressures and if the central bank possesses private information about its policy errors (the ζ disturbances in our model), then the potential exists for independent central bank policy strategies to partially offset the effects of such governmental pressures. This points to a potentially fruitful extension of our model, in which both government officials and central bank policy makers have their own private information sets. Government officials could use structural control over certain aspects of central-bank independence and an optimal degree of pressure to impinge on the central bank’s objectives, but in the absence of complete information about the precision with which the central bank can conduct policy and about the central bank’s true objectives. The central bank, in turn, would have the freedom to react to such governmental pressures by altering policy instrument settings, or even switching policy instruments. Finally, it would be interesting to extend this model to an infinite-horizon setting. Although we believe that our two-period model captures the essential intertemporal tradeoffs that the central bank faces, an infinite-horizon environment undoubtedly would yield more general, and richer, sets of conclusions about the central bank’s instrument-choice problem in a real-world setting with no “concluding” period. We leave these and other interesting issues for future research. and