اثرات تعادل عمومی از سیاست های مالی: برآورد برای منطقه یورو
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|28828||2009||27 صفحه PDF||سفارش دهید||17028 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Public Economics, Volume 93, Issues 3–4, April 2009, Pages 559–585
This paper describes a dynamic stochastic general equilibrium model featuring a fraction of non-Ricardian agents in order to estimate the effects of fiscal policy in the Euro area. The model takes into account distortionary taxation on labor and capital income and on consumption, while expenditures are broken down into purchases of goods and services, compensation of public employees and transfers to households. A newly computed quarterly data set of fiscal variables is used. Our results point to the prevalence of mild Keynesian effects of public expenditures. In particular, although innovations in fiscal policy variables tend to be rather persistent, government purchases of goods and services and compensations for public employees have small and short-lived expansionary effects on private consumption, while innovations in transfers to households show a slightly more sizeable and lasting effect. The effects are more significant on the revenue side: decreases in labor income and consumption tax rates have sizeable effects on consumption and output, while a reduction in capital income tax favors investment and output in the medium run. Finally our estimates suggest that fiscal policy variables contribute little to the cyclical variability of the main macro variables.
This paper reconsiders the economic effects of fiscal policy using an estimated dynamic stochastic general equilibrium model for the Euro area. We try to better understand how these effects depend on the composition of expenditures and revenues, as well as on the interaction with monetary policy. Recent years have witnessed significant changes in the fiscal position of both the United States and the Euro area. In many circumstances the main motivation behind these shifts has been related to cyclical considerations as policy makers have tried to support economic activity through fiscal stimulus. However, most of the discretionary measures undertaken, both on the spending and on the revenue side, were backed by little consensus among economists on their short to medium run effects. This lack of consensus stems from the difficulty economists have in building models able to replicate the main empirical regularities concerning fiscal variables. Frictionless models with optimizing forward-looking agents, as RBC models, for example, seem to be ill-suited to study the effects of government spending. In this context, Baxter and King (1993) have shown that any increase in expenditures brings about – as the government intertemporal budget constraint has to be satisfied – an increase in the discounted value of future taxes. This amounts to a negative wealth effect on households that induces a decrease in their private consumption, a contemporaneous increase in labor supply and, therefore, a decrease in the marginal productivity of labor and in real wages; as in the model the steady state capital labor ratio does not change, investment will increase. These theoretical correlations do not square with the empirical evidence coming from applied research. The debate on the empirical effects of fiscal policy shocks (in particular on the effects of government expenditure shocks on private consumption) is still unsettled. However, the disagreement mainly concerns the effects of increases in expenditures related to military buildups in the US — Perotti (2007) argues that the response of private consumption is positive, while Ramey (2008) that it is negative. For our purposes military buildups are somehow special events that do not really apply to the European case, as there is no relevant example of such events in European countries over our sample period (1980–2005). The literature on the effects of fiscal policy in “normal times” – that is abstracting from military buildups – mainly finds a moderately positive (or a non negative) response of private consumption to government expenditure shocks;1 also employment and real wages tend to grow, while the response of private investment is generally negative.2 The new-Keynesian paradigm, which mainly adds real frictions and nominal rigidities to an RBC framework, displays the same wealth-effect mechanism that entails a reduction in private consumption and an expansion in labor supply following a government spending shock.3 In this context, however, real wages may increase as a result of an outward shift of the labor demand induced by the expanding demand in the presence of sticky prices (with a reduction in price markups). In order to fill the gap with the evidence, the literature has recently moved away from the representative infinitely-lived rational agent. In particular Mankiw (2000) has argued that a model where Ricardian and non-Ricardian agents (that cannot save or borrow and therefore consume their income period by period) coexist is better suited for fiscal policy analysis with respect to both neoclassical and overlapping generation models.4 Building on this framework, Galí, López-Salído and Vallés (2007, henceforth GLSV) add rule-of-thumb agents to a standard new-Keynesian model. They show that both price stickiness and the presence of rule-of-thumb consumers are necessary elements in order to have a positive response of private consumption for reasonable calibrations of the parameters. “Rule-of-thumb consumers partly insulate aggregate demand from the negative wealth effects generated by the higher levels of (current and future) taxes needed to finance the fiscal expansion, while making it more sensitive to current disposable income. Sticky prices make it possible for real wages to increase (or, at least, to decline by a smaller amount) even in the face of a drop in the marginal product of labor, as the price markup may adjust sufficiently downward to absorb the resulting gap. The combined effect of a higher real wage and higher employment raises current labor income and hence stimulates the consumption of rule-of-thumb households”.5,6 In this paper we contribute to the debate on the macroeconomic effects of fiscal policy by estimating on Euro area data a DSGE model which puts the idea of GLSV into the framework of Christiano, Eichenbaum and Evans (2005). The latter includes a number of frictions proved to be useful for estimation purposes, as shown in particular on Euro area data by Smets and Wouters (2003, henceforth SW).7 We extend this framework with a relatively rich description of the fiscal policy side. In particular, for government revenues we consider and estimate fiscal policy rules defined on distortionary tax rates, while previous literature (GLSV, and Coenen and Straub, 2005, henceforth CS) had essentially focused on lump-sum taxes. In order to do so, we compute quarterly average effective tax rates on labor income, capital income and consumption for the Euro area following the methodology of Mendoza, Razin and Tesar (1994).8 On the expenditure side, we take into consideration the fact that the variable generally used in the literature as a proxy for government purchases of goods and services, that is government consumption from National Accounts data, includes both purchases of goods and services and compensations for government employees, as recognized earlier by Rotemberg and Woodford (1992) and later by Finn (1998). In fact, in the case of the Euro area over the last twenty five years (the sample period we consider), the employee's compensations share of government expenditures averaged 60% approximately. While government purchases of goods and services is a direct component of aggregate demand, compensations of government employees affect the economy mainly through their effects on employment and wages. We therefore define government consumption excluding compensations for public employees (an aggregate that we will label government purchases) and model public employment separately. The model is estimated using Bayesian inference methods on the Euro area data from 1980 to 2005. Bayesian technique – as forcefully claimed by Fernandez-Villaverde and Rubio-Ramirez (2006) – is now the standard tool for the estimation of DSGE models. Fernandez-Villaverde and Rubio-Ramirez (2004) show how, in practical applications, the Bayesian approach delivers very strong performance, especially on small samples. Our paper tries to assess the response of the main macro variables to a wide range of fiscal shocks, including revenue ones. We are not concerned only with the effects of government spending shocks on private consumption, although the issue has recently attracted considerable attention. In relation to this latter issue, we would like to stress that our estimation strategy allows also for a negative response, as we discuss in some detail in Section 5.4 below. Although fiscal policy is run at national level, the focus of this paper is on the Euro area. As cross country spillover effects from fiscal policy shocks tend to be limited,9 shocks in one single country mainly affect that country's variables, so that the effect on Euro area variables mainly depends on the weight of the country within the Euro area. Our estimated effects should therefore be interpreted as the (weighted) average effects of fiscal shocks across Euro area countries. On the other hand, focusing on the Euro area as a whole has advantages: first, we can easily take into account the role of the common monetary policy; second, we can keep the model relatively simple and disregard all the theoretical and empirical issues related to the analysis of a single country in a monetary union; third, we can build on a model specification, the one proposed by SW, proven to match the Euro area data quite well; last, there is not an obvious candidate country for our study, as there are no official quarterly data available for any of the Euro area countries previous to 1999. To our knowledge, this is the first paper that estimates a medium scale DSGE model with a detailed role for fiscal policy (featuring both distortionary taxes and detailing expenditures in its main components) on Euro area data. We use both state of the art econometric techniques for the estimation and a newly computed quarterly data set for fiscal policy variables (as government sector data in the Euro area are mainly available on an annual basis). We believe that the use of a rich set of data (especially for the government sector that is the focus of our paper) is necessary for a proper identification of parameters and shocks. In particular, having data on distortionary taxes can potentially improve the estimates of certain shocks: for example, it might help disentangling the effects of a shock affecting the consumption-leisure intratemporal trade-off from those of a change in the labor income tax rate, or the effects of an investment efficiency shock from those of a capital income tax rate shock. Our approach, therefore, overcomes some of the weaknesses related to the interpretation of shocks pointed out by Chari, Kehoe and McGrattan (2008). Our results point to a significant share of non-Ricardian agents (between 30 and 40%) and to the prevalence of mild Keynesian effects of public expenditures. In particular, although innovations in fiscal policy variables tend to be rather persistent, government purchases of goods and services and compensations for public employees have small and short lived expansionary effects on private consumption, while innovations in transfers to households show a slightly more sizeable and lasting effect. The effects are more significant on the revenue side: decreases in labor income and consumption tax rates have sizeable effects on consumption and output, while a reduction in the capital income tax favors investment and output in the medium run. Finally our estimates suggest that fiscal policy variables contribute little to the cyclical variability of the main macro variables. The paper is organized as follows. Section 2 describes in detail the model and our assumptions regarding policies. Section 3 sketches the techniques we use to solve and estimate the model, and describes the data and the assumptions regarding prior distributions. Section 4 presents our estimated parameter distributions, that are then used in Section 5 to discuss the effects of fiscal policy innovations. In Section 6 we summarize our results.
نتیجه گیری انگلیسی
In this paper we have presented new evidence regarding the macroeconomic effects of fiscal policy in the Euro area. To this end, we have developed a general equilibrium model and estimated its structural parameters through Bayesian techniques. As most of the Euro area official data on government accounts are available only at an annual frequency and given the importance for our purposes of including detailed information on government variables, we have also computed quarterly data for important fiscal policy series. Our results point to a significant share of non-Ricardian agents and to the prevalence of mild Keynesian effects of fiscal policy. In particular, although innovations in fiscal policy variables tend to be rather persistent, government purchases of goods and services and compensations for public employees have small and short lived expansionary effects on private consumption, while innovations to transfers to households show a slightly more sizeable and lasting effect. The effects tend to be more significant and lasting on the revenue side: decreases in labor income and consumption tax rates have sizeable effects on consumption and output, while a reduction in capital income tax favors investment and output in the medium run. Moreover, our results suggest that fiscal policy variables contribute little to the cyclical variability of the main macro variables. The reported evidence seems to favor the new-Keynesian framework. The estimated impact increases in private aggregate consumption and real wages after shocks to government spending items, and in private sector employment after a government employment shock, are not consistent with standard RBC models. While our model is rather general, we have restricted our focus to a closed economy setup. Although we believe this is a good approximation for an economic area as the Euro area, as SW have shown, we might still be missing some effects coming from the external channel. This, however, is a topic for future research.