Ebook Fiscal Policy in the New Neoclassical Synthesis
What are the effects of changes in government expenditure on the business cycle? Historically, two different and mutually incompatible strands of theories have been advanced to answer this question. Broadly characterized, there is the Keynesian tradition, on the one hand, as captured in the familiar textbook IS-LM-Phillips-Curve model, with its focus on the relevance of aggregate demand disturbances on cyclical conditions. Expansionary fiscal policy, like any exogenous increase in aggregate demand, in this view allows demand constrained firms to sell more output, thus boosting income, employment, and, by the multiplier effect, consumption, because the inflexibility of goods prices that prevails in the short-run makes output demand determined; in other words, there is a non-vertical short-run aggregate supply curve which is upward sloping because prices are temporarily sticky and adjust only gradually, as typically captured in some version of a Phillips-Curve specification. In the following, we will call this mechanism the `aggregate demand effect' of fiscal policy for short; see Taylor (2000) for a recent discussion.
On the other hand, the effects of fiscal policy have been studied more recently in purely real dynamic general equilibrium models with optimizing agents and fully flexible prices, e.g. Baxter and King (1993). Here, the central mechanism by which fiscal policy influences the private economy is the negative wealth effect implied by the tax financing of rising government expenditure which, with standard preferences, induces an increase in labor supply, thus raising output and employment and depressing private consumption. This chain of events, which we will sometimes simply abbreviate as the `wealth effect' in what follows, is thoroughly different from the aggregate demand effect story, as any change in output and employment induced by fiscal policy is due to the optimal response of household labor supply. Consequently, the predictions of the neoclassical general equilibrium model are directly opposed to those of the Keynesian theory with respect to some important variables like wages and private consumption.
Both theories may be regarded as less than fully satisfactory due to debatable assumptions on which they are built. On the one hand, traditional IS-LM theorizing nowadays seems outdated to most researchers due to its lack of conventional microfoundations and rational expectations. On the other hand, the more recent real general equilibrium models may be inappropriate, too, because of their neglect of any frictions in the goods or labor market; particularly, there is a large and growing body of evidence in favor of temporary price stickiness at business cycle frequencies (see the survey in Taylor, 1999) which points to an important role for nominal variables to play in any business cycle theory. In related research, namely the literature on the transmission of monetary policy, a recent consensus model seems to have emerged, which is labelled either `New Neoclassical Synthesis' (by Goodfriend and King, 1997), or interchangeably `New-Keynesianism' (by, among others, Hairault and Portier, 1993), or `Neo-Monetarism' (by Kimball, 1995), or `Optimizing IS-LM' model (by McCallum and Nelson, 1999, and Casares and McCallum, 2000); henceforth, we prefer the term New Neoclassical Synthesis which we will abbreviate by NNS. The gist of this approach is the combination of an explicitly dynamic representative agent optimizing general equilibrium framework with short-run nominal frictions, particularly temporary stickiness of nominal goods prices, hence with a strong influence of monetary policy on real aggregates over the business cycle. Recent research in this area, particularly that by McCallum and Nelson (1999), has shown that much of the intuition gained from the simple IS-LM model carries over to the NNS model as far as the effects of monetary policy are concerned. Indeed, a (very) stylized version of this model has become the basic workhorse in the literature on monetary policy; see the survey by Clarida et al. (1999).
But if price stickiness is, as perceived in a large literature, the main reason why monetary policy matters, the question naturally arises whether this has any implications for the understanding of the way fiscal policy works. This is what the present paper intends to contribute. One of the questions that we pursue by this approach is in how far fiscal policy in a typical NNS model can serve as a microfoundation for the intuition conveyed by the IS-LM framework, thus complementing the work that McCallum and Nelson (1999) have done for the case of monetary policy. Specifically, we model a monetary economy with a cash-in-advance constraint where optimizing households accumulate capital subject to adjustment costs. Firms are assumed to be monopolistic competitors striving to realize a constant markup of product prices over marginal costs, which is hampered by the presence of the type of price stickiness originally presented in Calvo (1983), i.e. a stochastically arriving temporary inability on the part of a subset of all firms to change their output prices as they desire in any given period.
The government collects lump-sum taxes and seignorage and uses the receipts to purchase goods; in one variant of the model these may provide utility to the representative household, who finds government consumption to be an imperfect substitute for private consumption. Fiscal policy is modelled as a serially correlated exogenous increase in real government expenditures. In addition, we allow a monetary authority to endogenously react to the fiscal stance by adjusting nominal interest rates according to a feedback rule of the type advocated in Taylor (1993) and analyzed in a large monetary policy literature thereafter (see, e.g., the contributions in Taylor, 1999).
Download
PDF Ebook Fiscal Policy in the New Neoclassical Synthesis
Posted in :
