Is There a Stabilizing Role for Fiscal Policy in the New Consensus?

Abstract

This paper considers the possibility of using fiscal rather than monetary policy as the instrument of stabilization policy in a new consensus framework. Describing the conduct of fiscal policy in terms of a ‘pseudo Taylor rule’, it is shown that fiscal policy is as, if not more, effective than monetary policy as a tool for macroeconomic stabilization. The conclusion reached is that the comparative neglect of fiscal policy as an instrument of stabilization policy in new consensus macroeconomics is unwarranted.

An earlier version of this paper was presented at the Meetings of the Eastern Economic Association, New York City, March 2005. I would like to thank conference participants for their helpful comments. Any remaining errors are my own.

Notes

¹Absent supply shocks, from which we abstract in equation (2) for the sake of simplicity, this is the form and interpretation of the Phillips curve found in, for example, Gordon (1997) and Staiger et al. (1997a, 1997b). Alternatively, equation (2) can be interpreted as an expectation-augmented Phillips curve of the form:

where p ^e = p _{− 1} and this latter assumption is justified by appeal to the notion that time series inflation data are characterized by a unit root (see, for example, Ball & Mankiw, 2002, pp. 118–119).

Note also that it is conventional to assume that the Phillips curve displays dynamic homogeneity i.e. that the long run values of real variables are homogeneous of degree zero in inflation, so that φ = 1.

²Taylor rules usually specify the nominal interest rate as being the instrument of monetary policy. But since the central bank must be able to change the real interest rate in order for the equilibrium of the new consensus model to be stable (on which see below), and since it is usually assumed in the new consensus literature that central banks can, in fact, do this, specifying the Taylor rule in terms of the real interest rate involves no great aberration.

³See Woodford (2001, p. 233) for the analogous condition when the Taylor rule describes the adjustment of the nominal interest rate.

⁴A similar stability problem arises if the IS curve is specified so that real output is responsive to both the real interest rate and the rate of profit (see Alonso-Gonzàlez & Palacio-Vera, 2002). This problem is not discussed in detail above because unlike other stability problems discussed in the text, it need not be specific to the use of monetary policy as the instrument of stabilization policy.

⁵We return to discuss the empirical significance of these observations regarding the interest (in) elasticity of aggregate expenditures in Section 3.3 below.

As might be expected, the potential problems for stabilization policy posed by the zero lower bound on nominal interest rates have not failed to escape the attention of advocates of the new consensus (see, for example, Woodford, 2003, pp. 251–252). A variety of policy responses to these problems have been proposed, including varying the central bank's inflation target (Eggertsson & Woodford, 2003) and changing the instrument of stabilization policy to a weighted average of the interest rate and nominal exchange rate (McCallum, 2005). Eggertsson & Woodford (2004) even contemplate using fiscal policy to stabilize the economy when the zero lower bound binds, although this is the only case in which they deem it appropriate to use fiscal policy as an instrument of stabilization policy.

⁶Note that this ‘passive’ monetary policy involves the central bank both accurately estimating and keeping track of the natural rate of interest as it changes over time. Monetary policy is passive, therefore, only in the sense that the interest rate is not the main instrument of stabilization policy: estimating and keeping track of the natural rate of interest are far from trivial tasks for the central bank to perform.

⁷Taylor (2000) has already dubbed such rules ‘fiscal policy rules,’ in order to effect comparison with his original monetary policy rule. But since Taylor's monetary policy rule has become universally known as the Taylor rule, the term ‘pseudo Taylor rule’ is preferred in what follows to describe the conduct of discretionary fiscal policy aimed at stabilization of the economy.

⁸It should be noted that this role is ‘implicit’ in the expression for the equilibrium real interest rate derived in the previous section, given the definition of y ₀. According to Arestis & Sawyer (2003, p. 7), an equilibrium interest rate that depends on the conduct of fiscal policy in this fashion does not merit description as a ‘natural’ rate of interest at all.

⁹This result dovetails with the more general discussion of Arestis & Sawyer (2003, pp. 8–14), who are skeptical of claims that crowding out limits the efficacy of fiscal policy.

¹⁰See, for example, Barro (1974).

¹¹For example, Arestis & Sawyer (2006) calculate that a 5 percentage point change in the real interest rate would be necessary to offset a 1% change in GDP. See also Arestis & Sawyer (2002) for a survey of evidence suggesting that the impact of changes in the interest rate on inflation – another important aspect of the stabilization process in new consensus models - is modest.

¹²See also Arestis & Sawyer (2003, pp. 9, 13, 21–28) for discussion of empirical evidence regarding the interest inelasticity of aggregate spending and Ricardian equivalence effects.