Introduction Directory UMM :Data Elmu:jurnal:J-a:Journal of Economics and Business:Vol50.Issue4.July1998:

Suitable Policy Instruments for Monetary Rules Saranna R. Thornton The performance of different policy instruments is examined in counterfactual simulations using McCallum’s adaptive rule. Three different macroeconomic models are used, re- spectively, to simulate values of nominal GDP from 1964:Q1–1995:Q4. In comparison with historical, discretionary monetary policy, rules using M2 as the policy instrument produce substantial reductions in the variability of GDP around a fixed trend—even with money control error. M2 performs well, despite the presence of money control error, because the feedback term in McCallum’s rule, in combination with the relative stability of the linkage between M2 and GDP, successfully counteracts most of the impacts of undesired quarterly changes in money growth. © 1998 Elsevier Science Inc. Keywords: Monetary policy; Monetary rule JEL classification: E42, E52

I. Introduction

The introduction of bills in Congress that mandate price stability as the Fed’s primary policy objective, and the support these bills received from Federal Reserve Officials and other economists, suggests a growing consensus on the primary goal of monetary policy. Yet, debate continues regarding the appropriate means, to achieve this goal. Some propose using discretionary monetary policy, while others favor monetary rules. Supporters of rules argue that discretionary policy will exhibit an inflationary bias when the central bank’s objective function negatively weights inflation and positively weights levels of output above the full-employment level. 1 Proponents of rules argue that inflationary bias can be eliminated if the central bank adopts a policy rule capable of Department of Economics, Hampden-Sydney College, Hampden-Sydney, Virginia. Address correspondence to: Dr. S. R. Thornton, Department of Economics, Hampden-Sydney College, Hampden-Sydney, VA 23943. 1 Recent conduct of monetary policy does not contradict theory. Discretionary policy has been successful in achieving inflation stability—not price stability. Furthermore, recent inflation stability is serendipitous and not invariant to the composition of the FOMC. Journal of Economics and Business 1998; 50:379 –397 0148-6195 98 19.00 © 1998 Elsevier Science Inc., New York, New York PII S0148-61959800010-1 ensuring long-run price stability. Even central bankers acknowledge that policy rules can provide information useful in the formulation of discretionary policy. 2 Interest in adaptive monetary rules has risen 3 as economists search for a rule that: 1 is operational; 2 performs well i.e., ensures long-run price stability in a variety of plausible macroeconomic models; 3 performs well when financial innovations or other shocks alter the linkages between the policy instrument and the intermediate target e.g., velocity, and 4 promotes greater levels of price stability than historical discretionary policies have. In an attempt to isolate a rule that best meets these criteria, the performance of three versions of McCallum’s adaptive rule was evaluated from 1964:Q1–1995:Q4. Rules utilize M2, the St. Louis and the Federal Reserve Board Monetary Base, respec- tively, as policy instruments. 4 Rules utilizing M2 and the base were examined because research by McCallum 1987, 1988, 1990, Judd and Motley 1991, 1992, Feldstein and Stock 1993, Thornton 1993, among others, suggests these aggregates may meet the four key criteria listed above. The performance of rules utilizing other commonly proposed policy instruments was not examined because research indicates that they don’t meet all criteria listed above. For example, Friedman 1988 found that rules utilizing measures of reserves are unlikely to ensure long-run price stability. Thornton 1993 found that M1 did not perform as well as M2 or the St. Louis monetary base in variations of McCallum’s rule. Judd and Motley 1992 found that interest rates did not perform well as a policy instrument in a rule which targeted a stable price level, and Clark 1994 found that a rule utilizing interest rates to target a stable inflation rate failed to reduce the variability of real GDP growth or inflation. 5 . Section II reviews some issues regarding the choice of policy instrument and concludes that both M2 and the monetary base exhibit some deficiencies. Section III explains the GDP simulation procedure utilized to evaluate rule performance and develops baseline performance measures utilizing a St. Louis monetary base rule, a Federal Reserve Board monetary base rule, and an M2 rule 6 —the latter under an initial assumption of no money control error. Different measures of the base were used in order to consider the hypothesis that specification of the monetary base affects rule performance. Results were also derived utilizing an M2 rule which incorporates likely degrees of money control error. The performance of the policy rules is compared to assess their effectiveness. Section IV 2 Statement by Governor Lawrence Meyer, January 5, 1997, AEA meetings, New Orleans, LA. 3 See, for example, McCallum 1987, 1988, 1990; Judd and Motley 1991, 1992, 1993; Feldstein and Stock 1993; Dotsey and Otrok 1994; Thornton 1993. 4 M2 doesn’t meet the strict definition of a policy instrument as a variable under the direct control of the Fed. M2 is considered an instrument for achieving a nominal income target, assuming there is an intermediate policy instrument utilized to target M2. 5 The performance of interest rate rules in Clark 1994 may seem inconsistent with the Fed’s recent experience using the Federal Funds rate to target a constant inflation rate. The inconsistency may be due to the specification of interest-rate rules. They typically include one feedback term which alters the policy instrument in response to a deviation from the rule’s nominal income target. In contrast, McCallum’s rule includes a velocity moving average term along with the feedback term. Both terms alter the policy instrument when the velocity growth rate i.e., the linkage between the policy instrument and the target variable changes. Proposed interest-rate rules can be thought of as analogous to McCallum’s rule, minus the velocity term. And, in fact, in GDP simulations, McCallum’s rule did not perform as well, with only the feedback term. Better results might be obtained with interest-rate rules by rewriting them to incorporate a term which explicitly adjusts the interest-rate response when there is a change in the relationship linking interest rates to income growth. 6 All simulations utilize rules that target nominal income. Henceforth, the term ‘M2 rule’ or ‘monetary base rule’ refers not to the rule’s target, but to its policy instrument. 380 S. R. Thornton reviews some caveats bearing on any final determination of the success of a particular rule. Results suggest that monetary base rules do not consistently produce greater levels of price stability—performance is model dependent. In contrast, the M2 rule consistently met the four criteria listed above during the sample period 1964:Q1–1995:Q4.

II. Choosing a Policy Instrument