Ebook Productive Government Expenditure in Monetary Business Cycle Models
The recent literature on the effects of public policies on short-run macroeconomic dynamics shows two remarkable trends. On the one hand, a rather widespread consensus has emerged on the question how monetary policy is transmitted to the economy. The New Keynesian paradigm of intertemporally optimizing agents who are subject to temporary price stickiness has proved to be a workhorse capable of capturing the essential relation between monetary variables and the real economy in a transparent and empirically successful way. The monetary policy literature has consequently moved onwards to study the question of how precisely monetary policy should be conducted in this framework (e.g. Clarida et al., 1999, or Wood ford, 2003). On the other hand, there is no such unanimity with respect to the other branch of macroeconomic policies, namely fiscal policy. In particular, the question how the relation between government expenditures and private activity over the business cycle should be modeled is currently open to debate.
The reason is that, unlike with monetary policy, in the case of fiscal policy there is no generally accepted model that is able to capture the way in which variations in government expenditures seem to affect the business cycle. Empirically, positive government demand shocks appear to set forth positive temporary responses of employment, wages, and private consumption (e.g. Blanchard and Perotti, 2002, Fatas and Mihov, 2002, Gali et al., 2004, Canzoneri et al., 2003). This evidence is not easily squared with the notion of intertemporally optimizing household behavior under rational expectations. With optimizing households, the increased public command over available resources that is associated with higher government demand implies an incentive to work and save more to offset the negative consequences to household wealth. Thus, increased fiscal spending should be expected to lead to a reduction in private consumption demand, increased labor supply and thus lowered real wages. In fact, this is the counterfactual theoretical prediction of the flexible price models in e.g. Baxter and King (1993) or Campbell (1994), as well as of the sticky price models in Linnemann and Schabert (2003) or Canzoneri et al. (2003).
The natural conclusion to this mismatch between theory and evidence is that wealth effects can hardly be the only channel through which fiscal policy affects the business cycle. Consequently, the literature has produced a number of ways in which wealth effects are mitigated or compensated by other mechanisms. Ravn et al. (2004a,b) suggest that wage and consumption responses to demand shocks are less likely to be negative in models where there are countercyclical markups, either as a result from habit formation with respect to individual goods or from subsistence points in household’s utility functions. Ludvigson (1996) shows that with distortionary income taxation higher government expenditures can temporarily raise consumption through an intertemporal substitution effect if they are accompanied by a large enough increase in debt to let the tax rate decline initially, which would however imply a strong negative effect on real wages. An alternative explanation is that households, or a part of them at least, do not optimize intertemporally at all, and therefore do not cut back their consumption in response to higher government expenditures, but instead determine their consumption as a simple reaction function to current disposable income. This assumption has been used by Mankiw (2000) in a growth model, and has recently been applied to business cycle models with monetary and fiscal policy by Gali et al. (2004). They show that if a substantial fraction of households follow the ‘rule of thumb’ of choosing their consumption as a fraction of current disposable income, then higher fiscal spending can be associated with increased private consumption and real wages. However, recent results by Coenen and Straub (2005), who estimate a dynamic general equilibrium model with ‘rule of thumb’-consumers with Euro area data cast doubt on the empirical relevance of this explanation.
The present paper examines another possibility, which allows to retain the assumption of an optimizing representative agent, to argue that the actual reduction in household wealth implied by increased government expenditures may be not that large in fact. In particular, if government spending is not completely wasteful, but rather enhances the productivity of the private production sector, then it can be expected that the reduction of private wealth through taxation can be mitigated by the improvement in production possibilities. We present a sticky price model with government spending entering the private firms’ production functions with an elasticity equal to or larger than the share of government expenditures in output. The central bank’s interest rate policy is assumed to follow an active inflation feedback rule, such that the real interest rate, and thus private consumption, is closely related to inflation. Since in the case of Calvo (1983) style staggered price adjustment inflation depends on real marginal costs, which are in turn influenced by the productive contribution of government spending, the possibility of positive effects of fiscal policy shocks on consumption, wages, and employment arises. The results are shown to depend on the fraction of government expenditures that are financed by distortionary (income) taxation and, correspondingly, on the degree of debt financing. It is further demonstrated that the stance of monetary policy is not irrelevant for the effects of fiscal policy. In particular, a passive interest rate policy can revert the results, while the interaction of monetary and fiscal policy is decisive for equilibrium stability and uniqueness.
The idea that government spending is a productive input to firms’ production functions has been expounded in a large empirical and theoretical literature. Empirically, a positive output elasticity of public investment or capital has been found in many studies (e.g. Aschauer, 1989; a survey is provided by Gramlich, 1994). The theoretical consequences of this observation have been mainly discussed in the context of growth models, following the seminal contribution of Barro (1990); see Turnovsky (2000) for a survey. From the viewpoint of business cycle theory, the closest predecessor of the current paper is Turnovsky and Fisher (1995), who study the short-run effects of productive government spending in a model with endogenous labor supply and lump-sum taxation. Fisher and Turnovsky (1998) analyze the role of distortionary taxes in a model with inelastic labor supply. None of the earlier studies considers sticky prices.
The results of this paper can be summarized as follows. First, a moderate production elasticity of government expenditures is sufficient, in a model with lump-sum taxes, to generate effects that are qualitatively consistent with the empirically observed pattern of positive employment, wage and consumption responses to a fiscal shock. The result with respect to consumption depends on the relative sizes of the output elasticity of government spending and its share in output, the effect being stronger if the share is smaller than in a long-run optimum, in the sense of Barro (1990). Second, if taxes are distortionary, there is of course a trade-off between the expansionary effects of productive spending and the contractionary ones of taxation. With a continuously balanced budget, the conditions for the spending effect to prevail are extremely restrictive. Third, however, if the empirically plausible assumption of low variations of tax rates at business cycle frequencies is adopted, and fiscal spending changes are in the short-run to a considerable degree reflected in temporary debt accumulation, it appears that the results found for the lump-sum tax case go through for a large set of reasonable parameter values. Fourth, since the mechanism by which these effects occur are tightly linked to the effect of fiscal policy on costs and prices, the results depend crucially on whether the central bank sets interest rates actively or passively in reaction to inflation.
The rest of the paper is organized as follows. Section 2 presents the model. In section 3 we examine the macroeconomic responses to expansionary government expenditure shocks. Therein, we consider first, for analytical convenience, the stylized cases where government expenditures are entirely financed by either lump-sum or labor income taxation. In the final part of section 3 we then turn to the more realistic case where government receipts are jointly raised by income taxation and the issuance of interest bearing debt. In section 4 we briefly assess to role of monetary policy for the transmission of fiscal policy shocks. Section 5 concludes.
Download
PDF Ebook Productive Government Expenditure in Monetary Business Cycle Models
Posted in :
