Useful government spending, direct crowding-out and fiscal policy interdependence

Abstract

This paper introduces non-separability between private and public consumption in a dynamic, open economy with imperfect competition and nominal rigidities. This implies a direct crowding-out effect that, generalizing well known closed economy results, tends to reduce consumption following both domestic and foreign fiscal expansions. A less expected result is that substitutability has a positive effect on the short-run output spillover.

Introduction

Recent research in open economy macroeconomics, following the path-breaking contribution of Obstfeld and Rogoff (hereafter Obstfeld & Rogoff, 1995, Obstfeld & Rogoff, 1996), has seen the emergence of a new paradigm, that embodies market imperfections such as monopolistic competition and price-stickiness in an optimizing setup (see Lane, 2001 for a survey). In this new literature, unlike in the previously dominating Mundell–Fleming (MF) framework, the analysis of the international effects of macroeconomic policies is based on a rigorously microfounded framework. The representative agent approach also allows an explicit evaluation of the welfare effects of such policies. Although the OR framework is a flexible tool that can be used to address several issues in open economics macroeconomics, existing contributions tend to focus on the effects of money shocks, neglecting the role played by government spending. Even when government spending is introduced in the analysis, it is usually assumed to be pure waste (as in Obstfeld & Rogoff, 1995, Obstfeld & Rogoff, 1996) or to affect private utility only in an additively separable way (as in Corsetti and Pesenti, 2001).

We aim at filling in these gaps in the literature, developing a version of the OR model in which government expenditure affects utility in a non-separable fashion.¹ In doing this, we use a general preferences specification that nests both the cases of partial and perfect substitutability between private and public consumption.² The main purpose of this paper is that of studying the consequences of the introduction of non-separability on the short-run and long-run multipliers of domestic and foreign fiscal expansions. Although the total multipliers are different according to whether we have perfect or imperfect substitutability, the signs of the differential effects of non-separability are the same regardless of the degree of substitutability (see Table 2). Our specification of preferences rests on the idea that public expenditure can have a direct crowding-out effect on private consumption. The debate on crowding-out goes back to Keynes (1929), who referred to it as diversion; in more recent times, Bailey (1971) and Buiter (1977) have analyzed it in an IS/LM framework. This topic has also been addressed using modern, microfounded models, both for the closed and open economy case. Heijdra and Ligthart (1997), for example, introduce non-separability in a static, closed economy with imperfect competition. In the real business cycle tradition, the subject has been studied in perfect-competition, flexible-price models by Roche (1996) and Finn (1998). The main innovation of our paper is, therefore, that of studying direct crowding out in a model that contains all the elements of the new emerging paradigm in international macroeconomics: intertemporal optimization, imperfect competition and nominal rigidities.

We show that the introduction of useful government spending tends to reduce consumption in the short run following both domestic and foreign permanent expansions. It also has a negative impact on output, with the exception of the short run effect following a foreign expansion. Contrary to the MF case, in our model a short-run negative output spillover arises. The introduction of non-separability, having a positive differential effect, mitigates the negative spillover.³ Both the negative spillover and the positive differential effect of non-separability can be explained by the fact that our microfounded money demand equations are functions of consumption rather than gross income. The results relating to the role of non-separability are broadly consistent with those found by Roche (1996) and Finn (1998). Both these authors, for example, find that increasing the marginal rate of substitution between private and public consumption from zero to unity makes government spending acyclical.⁴ The same result can be derived in our model, both for the short and the long run.⁵ Further, in Roche’s model, the same sensitivity experiment increases the degree of negative correlation between government spending and private consumption (Roche, 1996). This is consistent with most of the effects of an increase in non-separability on consumption in our model.⁶

As our model nests the OR case of pure waste government expenditure, a second purpose of this paper is that of carrying out a detailed analysis of fiscal policy interdependence in that case. This is quite an interesting exercise, given that OR sketch a version of the model that encompasses government spending but they do not analyze the positive and normative implications of fiscal policy in any detail.

A limitation of our model is the assumption that public and private consumption have the same composition. The per-capita government real consumption indexes take the same form as the private sector ones. As a consequence, public spending falls equally on all differentiated goods, regardless of their place of production. It would not be difficult to consider a variant of the model with home bias in government spending. Some research has been done by the author in this direction but is not reported here for the sake of brevity.⁷

Another shortcoming of our treatment of fiscal policy is that of neglecting the role of the government as a producer. It is a common observation that a large part of government spending is for employment, but we do not distinguish between expenditure for consumption goods and expenditure for employment. Finn (1998), however, shows that failing to distinguish between these two different sub-aggregates of government expenditure can lead to overestimation of the government’s impact on the economic cycle. This is an important caveat to bear in mind in interpreting our results. Introducing similar issues in the model will undoubtedly be an interesting direction for future research.

The paper is organized as follows. Section 2 introduces the model. Section 3 discusses the multipliers of domestic and foreign fiscal expansions. Section 4 investigates some welfare implications, while Section 5 draws some conclusions.