Notes
-
[*]
We would like to thank the anonymous referees for helpful comments.
-
[‡]
Galatasaray University, Department of Economics, Galatasaray Center for Economic Research (GIAM), 36 Çırağan avenue 34357, Istanbul, Turkey. Tel: +90 2122274480-590, fax: +90 2122285283. Courriel: izeyneloglu@gsu.edu.tr.
-
[✝]
Université de Strasbourg, Bureau d’Économie Théorique et Appliquée (BETA), UMR 7522 du CNRS, 61 Avenue de la Forêt Noire F 67085 Strasbourg Cedex, France. Tel: +33 (0) 3 68 85 20 69. Fax: +33 (0) 3 68 85 20 70. Courriel: koenig@unistra.fr. (Auteur correspondant)
-
[1]
In the initial steady-state, public spending is assumed to be equal to zero and the current account is balanced in both countries. The log deviation of any home variable x relative to x* is equal to the gap between the deviations of x and x* with respect to their steady-state values. Since the steady-state values of home and foreign variables are equal, the gap between the log-deviations of the two variables is equal to the gap between the log of those variables.
-
[2]
Coenen assumes that the ratio of taxes T to GDP responds automatically to public debt according to the following rule:
where Bt is public debt in period t, PY is the steady-state GDP and
is the target ratio of public debt-to-GDP.
-
[3]
This mechanism is represented by the following AR (1) process for public spending Gt augmented with a public debt feedback:
where a bar over a variable indicates the steady-state value and
is the output that would prevail in the absence of all nominal rigidities. εt is the public spending shock. A simple AR (1) process implies b = c = 0.
-
[4]
When public spending is assumed to be welfare enhancing as in Andersen and Spange [2006] and Lombardo and Sutherland [2004], the expected level and the variance of public spending also enter directly in the welfare equations.
-
[5]
For example, an increasing negative terms-of-trade gap which decreases the home competitive advantage combined with a higher relative foreign spending shifts work effort from home to foreign households.
1. Introduction
1 Several events explain the fairly recent revival of fiscal policy in the literature on policy efficiency, which has focused for a long time on the role of monetary policy. First, monetary policy is less efficient when the interest rates are close to zero, as in the case of USA and Europe, or absent as in the case of European countries which delegated monetary policy to the supranational European Central Bank (ECB). Therefore, active fiscal policy is the only instrument to regulate economic activity when monetary policy is inefficient or when there is an asymmetric shock in a monetary union. Second, the global economic crisis which is heavily felt during 2008-2009, has shown the necessity of resorting to fiscal policy as a stabilization tool as well as of controlling the evolution of public debt which has increased due to the crisis and the stimulus plans.
2 Because of their lack of microeconomic foundations, Mundell-Fleming type models do not allow to analyze fiscal policy in a satisfactory way. Similarly, real business cycle models with neoclassical features fail to offer a realistic view of fiscal policy issues since they neglect market imperfections and the empirically observed price and wage rigidity. Dynamic general equilibrium models with microeconomic foundations offer a renewal of the macroeconomic policy analysis by building a bridge between the two preceding approaches. Indeed, they adopt the intertemporal approach of the flexible-price models through optimizing behavior of agents and the price or wage stickiness of the aggregate Keynesian models. Moreover, they assume imperfect competition. This new framework appeared first in the new open economy macroeconomics (NOEM) literature initiated by Obstfeld and Rogoff [1995, 1996]. At the same time, it led to the extension of real business cycle models. These two streams of research then converged to the dynamic stochastic general equilibrium (DSGE) models following the methodological contributions of Schmitt-Grohe and Uribe [2004] and Lombardo and Sutherland [2007] among others.
3 This methodological evolution fostered new reflections on fiscal policies implemented in open economies that are in interaction with each other. Among these reflections, three themes arising from recent economic developments are especially important for open economies. The present paper aims to provide a survey of these three themes analyzed in a two-country framework.
4 First, fiscal policy is being implemented in increasingly interdependent economies characterized by a high degree of integration in financial and goods markets. While two-country general equilibrium models allow to take account of economic and financial interactions, they often assume perfect international integration of markets. However, the process of globalization has not yet achieved full integration of financial and goods markets, even in a monetary union where monetary integration does not necessarily imply full financial integration. Accordingly, several papers relax the assumptions of perfect international capital mobility and complete goods market integration in order to assess the impact of a higher economic integration on fiscal policy efficiency and to reconsider the conclusions of the Mundell-Fleming models.
5 Second, the public finance problem observed in most of the developed countries in the aftermath of the recent economic and financial crisis called for a better management of government spending and for a debt-financed fiscal policy that internalizes the necessity of a sustainable public deficit. This requires considering alternative fiscal policy designs in contrast to early models where fiscal policy design is limited to a tax-financed change in public consumption because of Ricardian households and the poor composition of public spending.
6 Third, the formation of the European Monetary Union (EMU) along with the recent global crisis revealed a role for fiscal policy as a stabilization tool and the need to set up a fiscal cooperation between interdependent countries to reduce undesired spillover effects. Several recent papers assess the potential gains from fiscal stabilization and cooperation in a monetary union within a stochastic general equilibrium framework using a second order approximation of social welfare functions. This approach allows a renewal of the traditional analysis on fiscal stabilization and cooperation carried out in open economy models under the assumption of certainty equivalence. This new approach is first used in a static setup and then extended to a dynamic framework. The dynamic version is mainly used to analyze the optimal combination of monetary and fiscal policy in a currency union governed by a common authority which ensures full cooperation between all stabilization policies instead of partial cooperation analyzed in the static version. Full cooperation between all policy makers can be considered as the future goal for the EMU members.
7 The paper is organized as follows: in section 2, we define the common features of the dynamic general equilibrium models which are used to analyze the fiscal policy in interdependent economies. Section 3 analyzes the effects of increasing integration in international financial and goods markets on fiscal policy efficiency. Section 4 shows how the fiscal authority can improve public spending management and how the necessity of restoring the public budget balance can be integrated in the stimulus plans in order to avoid permanent public deficits. Section 5 first studies the gains from monetary and fiscal stabilization and cooperation excluding coordination possibilities between the monetary and fiscal authority within a country. Then it discusses the optimal combination of monetary and fiscal policy assuming full cooperation among all policy-makers in dynamic stochastic two-country models. Section 6 offers possible extensions that might improve the literature’s explanatory capacity regarding the current economic problems such as unemployment, policy making under liquidity trap and demand shocks. Section 7 concludes.
2. Common features of dynamic general equilibrium models
8 Two-country general equilibrium models that analyze fiscal policy describe generally two identical interrelated economies, Home and Foreign, with imperfect competition in the goods or labor markets. There is a continuum of firms and households both of which are indexed on the intervals [0, n] in the domestic country and between (n, 1] abroad. The preferences of a representative home household j with rational expectations are defined by the following utility function:
∞ Mj
t = 0Pt
9
where β is the discount factor and E0 is the expectations operator. The utility
in period t depends on private consumption 
10 The period budget constraint of the home household can be written in nominal terms as follows:
11
where 
12
Although several papers define home private consumption 
jj j
C
t= [ω θ ( C H, t) θ + ( 1 − ω )θ C F
, t) θ ] θ − 1 [3]
13 where θ is the elasticity of substitution between home and foreign goods (θ > 1) and ω measures the share of home goods in the domestic consumption bundle.
14
The consumption indexes 
⎡⎣∫⎤⎦φ−1⎣⎡∫⎦⎤φ−1
n1
φ−1 φ−1
CjH, t = cjt (h) φ dh and CjF, t = cjt (f) φ df [4]
0n
15 The elasticity of substitution φ between two goods produced within one country may differ from the elasticity of substitution θ between home and foreign goods. However, the assumption φ = θ is widely used.
16
The overall home price index Pt corresponding to (3) is also a CES aggregation over home and foreign goods prices, PH, t and 
17 Household j determines his demand for a single home and foreign good by maximizing (3) under the constraint of minimum consumption expenditure which yields:
cjt (h) =tPH, t PHt, t ωCjt and cjt (f) =tPF, t PFt, t (1 − ω) Ctj [5]
18 where p (h) and p (f) denote respectively the prices of a single home and foreign good. All prices are expressed in the home currency.
19 Foreign equations are defined similarly to (1)- (5).
20
The law of one price implies that in equation (5), home and foreign prices
of any good are equal when expressed in the same currency:
21 The public sector in each country is composed of a central bank whose monetary policy can be active or passive and a fiscal authority whose policy can be discretionary or based on commitment. Public spending can be financed by taxes, public bonds and/or money. Most of the literature on two-country framework assumes that public consumption takes the same form as that of the private demand:
⎡⎣∫⎤⎦η−1⎣⎡∫⎦⎤η−1
n1
η−1 η−1
GH, t= gt (h) η dh andGF,t= gt(f) η df [6]
0n
22 where the elasticity of substitution between the goods produced within a country η can be equal to or different from the private consumption elasticity of substitution φ. Several recent papers consider other types of public spending such as public investment or public wage bill.
23
Aggregating the private and public home demand for a home good h
along with the foreign analogues gives the total demand
24 Following Obstfeld and Rogoff [1995], several papers assume that the prices of home and foreign goods expressed in local currency are fixed in the short run and perfectly flexible in the long run. The relation between short and long run results from the current account movements in each period. Indeed, a short run current account deficit/surplus deteriorates/ improves the foreign asset position. The resulting wealth effect alters the long run variables. This approach allows to obtain analytical solutions to the endogenous variables. In contrast, only a numerical solution can be obtained if one assumes sluggish price adjustment through Calvo [1983] pricing decisions, since this assumption implies a dynamic structure for the supply block.
25 In most of the two-country general equilibrium models, fiscal policy is analyzed under the assumption of perfect integration in goods or financial markets across countries. In the model above, the assumption of perfect international integration in goods market corresponds to a value of ½ for ω in equation (3) which implies that home and foreign goods have equal weight in the consumption bundle. Perfect international financial integration corresponds to assuming that the gap between home and foreign interest rates equals simply the expected fluctuations in the exchange rate. In reality, international market integration is not yet perfect. Therefore, it is interesting to relax these assumptions in order to evaluate the impact of higher integration on the effectiveness of fiscal policy.
26 The second class of simplifying assumptions concerns the nature of public spending which is one of the main instruments of fiscal policy. It is widely assumed that public spending is limited to public consumption which takes the same form as the private consumption. Government spending is assumed to be financed either by taxes or borrowing, without explicit considerations on the sustainability of public debt. Relaxing these assumptions allows to consider improvements in public spending management which has become an important goal for most of the developed countries, especially after the crisis in 2009. It also allows to take account of the impact of public spending on public debt.
27 Finally, the setup can be extended to a stochastic environment in order to evaluate the gains from fiscal stabilization policies based on fiscal rules as well as gains from fiscal cooperation at the international level. Moreover, this new setup allows to analyze the optimal combination of monetary and fiscal policy in a currency union.
3. Fiscal policy and the degree of international market integration
28 Financial market integration is often accompanied by higher integration in international goods markets encouraged by the World Trade Organization. The implications of increasing market integration for fiscal policy effectiveness is analyzed in Mundell-Fleming type aggregate models but are often neglected in micro-founded general equilibrium models. Several papers introduce imperfect international integration in financial and goods markets following Sutherland [1996] and Senay [1998] which are the first to analyze the impact of increasing market integration on the effects of a public spending shock in two-country dynamic general equilibrium models. The main results of the above-mentioned papers and of their extensions can be understood by analyzing a version of the general setup presented in section 2.
3.1. A model with imperfect market integration at the international level
29 The impact of the international integration process on the effectiveness of fiscal policy is analyzed in a version of the above model where the prices are fixed in the short run and flexible in the long run.
30 The representative home household j maximizes the following utility function:
∞ (Cj )1 − ρ Mj
t = 01 − ρ Ptμ
31 subject to the following budget constraint:
(1 + rt ) Pt Djt + Mjt − Mjt−1 [8]
32 In (8), D and F represent respectively the home and foreign bonds. The holdings of the foreign (home) bond by home (foreign) households imply transaction costs denoted by Z (Z*). The first order conditions imply:
P̄t()
t χ t+1
[9a]
= C−ρ it+1
Ct+1 = Ct β (1 + it+1 ) Pt+1 [9b]
C−ρW μ−1
tt
[9c]
Lt =
κPt
33 The above equations yield respectively a money demand function which depends on the consumption and the interest rate; a consumption Euler equation which gives the intertemporal consumption smoothing behavior; the optimal labor-leisure trade-off equation (labor supply) and the interest rate parity condition which defines the financial market equilibrium. Similar equations hold for the foreign households. Nominal wages, W and W*, are determined by the labor market clearing conditions at home and abroad.
34
Given the labor supply decisions, home and foreign firms produce differentiated goods according to the production function 
35 The home producer of a single good h will face the following demand function:
ytd (h) =t ω (Ct + Gt ) + [t ] (1 − ω) (Ct* + Gt* ) [10]
Pt P
t
36 The above equation is obtained by aggregating the demand function for household j given in (5) over all home households and combining with public demand, assuming θ = φ = η. A similar equation holds for the foreign producer of a single good f.
37 A lack of integration at the international level in the goods market may result from various imperfections which cause a deviation from the law of one price. According to Senay [1998], imperfect integration may result from the pricing decisions regarding internationally traded goods. Indeed, firms can choose to set their prices in the buyer’s currency (local currency pricing (LCP)) instead of the seller’s currency (producer currency pricing (PCP)) in order to avoid the effect of exchange rate fluctuations on prices. As pointed out by Warnock [1998], imperfect goods market integration may also result from idiosyncratic household preferences. Under that assumption, private and public consumption become biased towards domestically produced goods. The degree of bias is measured by the parameter ω in equation (3). For example, when ω >½, home consumer’s preferences are biased towards goods produced within the country implying a higher consumption of home goods with respect to foreign goods. In contrast, ω =½ implies that consumers in each country are indifferent to the origin of goods.
38 The assumption of biased preferences can also be used to model imperfect integration in financial markets. Indeed, despite the increasing international portfolio diversification in recent years, agents still prefer to hold domestic bonds (financial home bias) even though a higher share of foreign assets in the portfolio leads to a better risk diversification.
39
However most of the general equilibrium models on fiscal policy introduce imperfect financial integration through adjustment costs following
Sutherland [1996]. In this set up, it is assumed that home (foreign) agents
have free access to home (foreign) bonds market, but residents in each
country must bear a transaction cost when buying the bonds of the other
country. The real transaction costs incurred by home and foreign households, respectively denoted by Zt and 
2 γ (It ) and Zt* = 2 γ (It*)
[11]
Zt =
40
where the degree of financial integration is measured by the parameter γ > 0.
A fall in γ implies a higher financial market integration. The flow of transfers
from home bonds market to the foreign It and the foreign analogue 
41 where the domestic and foreign bonds, D and F, pay respectively the real return r and r*.
42 Given these financial transfers and the trade balance, the external equilibrium in each country is defined as:
(Ft + 1 − Ft ) − (Dt*+1 − Dt* ) = rt* Ft − rDt* +t
yy − Ct − Gt [13a]
Pt
(Dt*+1 − Dt* ) − (Ft + 1 − Ft ) = rDt* − rt* Ft +t* yt* − Ct* − Gt* [13b]
Pt
43 Money market equilibrium in each country is given by:
= and = [14]
PP
P* P*
44
where the consumer price index P and P* can be expressed as
45 As the model is non-linear, one has to take a first order approximation around the steady state in order to obtain a linear version which can be solved analytically. In the log-linear version all variables are expressed as percentage deviations from the steady state [1]. In addition, we assume that goods prices are fixed in the short run. This implies that the short run price deviation for each good is equal to zero. This yields two setups each of which will only be valid either in the short or the long run. In order to reduce the number of equations, one can subtract home equations from the foreign ones and then solve recursively for the relative deviations making use of the fact that short run consumption is related to the long run consumption by the Euler equation.
3.2. The impact of higher market integration on fiscal policy effectiveness
46 Using the above framework, Figure 1 below illustrates the effects of higher integration in both financial and goods markets on relative home and foreign consumption and output following a temporary one per cent increase in home tax-financed public spending. We retain the standard values for β = 0.99, μ = 2 and we impose κ = 0.9. Following Sutherland [1996] and Pierdzioch [2004a], we assume θ = φ = η = 6, ρ = 1.33 and χ = 1.
The effects of financial and goods market integration on fiscal policy impact on consumption and output
The effects of financial and goods market integration on fiscal policy impact on consumption and output
Note: Solid lines indicate the perfect financial integration case (γ = 0), while dashed lines correspond to a degree of imperfect financial integration (γ = 1). Abscise gives the degree of home-bias with ω ≥ 0,5.48 To sum up, a home fiscal expansion under imperfect integration increases home output more than the foreign output, decreases (increases) home (foreign) consumption, leads to a rise (fall) in the home (foreign) interest rate, causes a home trade deficit and a home currency depreciation.
49 A higher degree of financial integration which results from a decrease in the financial transaction costs (a lower γ) increases the capital movements. This reduces the impact of a fiscal expansion on the interest rates for a given degree of goods market integration. Therefore, home (foreign) consumption falls (rises) less. This, in turn, leads to a lower decrease (increase) in the home (foreign) money demand, which requires a lower home currency depreciation to restore the money market equilibrium. This mitigates the positive effect of the home currency depreciation on home trade deficit, which implies that the trade deficit is higher when financial integration increases. As home consumption falls less under strong financial integration, marginal utility of home consumption increases less which leads to a lower increase in labor supply and thereby in output.
50 To sum up, a higher financial integration mitigates the effects of a fiscal expansion on consumption and output in both countries, on home and foreign interest rates and on the exchange rate. The traditional Mundell-Fleming framework implies the same mitigating effects.
51 For a given degree of imperfect financial integration (γ = 1), a higher goods market integration (a lower value for ω) mitigates (amplifies), the effect of a fiscal expansion on home (foreign) output. Indeed, with a lower home bias, home demand increases less for home products and more for foreign goods, which implies a higher home trade deficit compared to the low integration case. As a result home country accumulates more debt vis-à-vis the foreign, which amplifies the increase (decrease) in the home (foreign) interest rate. Consequently, the higher goods market integration amplifies the fall (rise) in home (foreign) consumption resulting from a fiscal expansion. This amplifies the fall in home money demand, which requires a higher home exchange rate depreciation to restore the monetary equilibrium.
52 In sum, similarly to the case of higher financial integration, a higher goods market integration mitigates the effects of a fiscal expansion on home relative output. However, unlike the higher financial integration case, higher goods market integration amplifies the policy impact on the interest rate and thereby on consumption in both countries.
53 According to the panels (c) and (d) of Figure 1, in the long run where prices are flexible, a higher financial integration amplifies the fall (rise) in relative home consumption (output), following a fiscal expansion. The same result holds for the implications of higher goods market integration. Indeed, by increasing the debt accumulation at home in the short run, the fiscal expansion leads to a higher interest burden at home and therefore higher capital income abroad in the long run. The resulting fall in the long run home relative consumption determines an increase (decrease) in the marginal utility of home (foreign) consumption. This leads to a higher increase (decrease) in home (foreign) output according to the labor-leisure trade-off.
54 In a calibrated version of the above model with Calvo pricing decisions, Sutherland [1996] confirms the mitigation effect of higher financial integration on the response of home consumption and production in the short and medium run following a fiscal expansion. Moreover, using a panel VAR approach, Beetsma et al. [2008] show that the trade deficit resulting from a fiscal expansion increases with the degree of openness during the three years that follow the increase in public spending.
55 In contrast to Sutherland [1996], Pierdzioch [2004a] shows that in the presence of nominal income targeting monetary policy higher financial integration amplifies the impact of fiscal expansion on the exchange rate and home output in the medium run. This is due to the negative reaction of the monetary authorities to an increasing output. In another extension of Sutherland [1996], under labor market frictions in the form of efficiency wages, Çenesiz and Pierdzioch [2009] show that a higher financial integration amplify the negative (positive) impact of fiscal expansion on home relative consumption (production). Indeed, by increasing the gap between the home and foreign interest rate, it induces a higher currency depreciation due to lower money demand, which leads to a higher output at home. Senay [1998] provides a third extension of Sutherland’s model by introducing imperfect international goods market integration assuming local currency pricing (LCP). This setup confirms the results of the model presented above in section 3.
56 The analysis of the relation between higher financial and goods market integration and the fiscal policy seems particularly relevant for the European monetary union, whose integration process is not complete despite the existence of a common currency and where fiscal policy is the only stabilization tool from the members’ point-of-view. This question is analyzed by Koenig and Zeyneloglu [2010a] in a currency union version where prices are sticky in the short run and flexible in the long run. The results are similar regarding the mitigation effects of a higher financial and goods markets integration on consumption and output following a fiscal expansion. However, unlike Senay [1998], the degree of integration in each market influences the impact of a higher integration in the other. This difference comes from the fact that in Şenay [1998], imperfect integration in goods markets is measured by the imperfect pass-through of exchange rates while in Koenig and Zeyneloglu [2010a] it is measured by the home-good bias. However, numerical results in Pierdzioch [2004b] show that the impact of an increasing financial integration may be weak unless financial markets are complete, implying perfect international risk sharing.
57 The work listed above suggests a series of economic factors that are likely to affect the relation between financial integration and the impact of fiscal policy on national output. However, the results do not allow to arrive at a univocal conclusion. Such a relation also seems to be difficult to identify in empirical models. Indeed, by testing a model that extends Sutherland [1996] based on data from 24 OECD countries, Buch et al. [2005] cannot find a significant relation for the 1970s and 1980s. However, they find an inverse relation between the degree of financial integration and the output response to fiscal policy during the 90s, which confirms the results of the model presented above in section 3.
4. Public spending management and debt stabilization
58 There has been a change in the perception of fiscal policy since the mid 1990s which manifested itself by the efforts to improve the management of public spending rather than simply increasing public consumption in a discretionary way. Moreover, the recent crisis and the fiscal stimulus plans launched in 2008-2009 in the USA and Europe have raised concerns about the impact of fiscal policy on public debt accumulation. These concerns on the relation between fiscal policy and public debt stability led to a new concept of fiscal policy implementation in open economy general equilibrium models.
4.1. Public spending management
59 Most of the early open economy general equilibrium models assume the same form for public and private spending. This implies that goods consumed by private and public sectors have the same elasticity of substitution. This assumption rules out the possibility of public competition policies that are likely to increase fiscal policy effectiveness. Another common assumption is to consider public consumption as the only component of public spending neglecting different types of expenditures (e.g. on goods, on public employees’ wages, on infrastructures). Relaxing these assumptions allows to consider the effects of structural fiscal policy that lead to an improvement of public spending management. Moreover, those assumptions are likely to have an impact on the effectiveness as well as the international transmission mechanism of fiscal policy.
60 A public competition policy consists of increasing the price elasticity of public consumption which will reduce the monopoly power of private firms that sell goods to the government. In order to consider the effects of this type of policy, Ganelli [2008] uses a version of the model defined in section 3 assuming perfect international market integration (γ = 0 and ω = 0.5). The author allows for different values of substitution elasticity in public and private consumption (φ ≠ η). In this setup, increasing the price elasticity of public consumption mitigates the crowding-out effect of fiscal expansion on home private consumption. Indeed, this type of policy reduces the mark up rate of private firms and leads to a reallocation of resources in favor of private consumers. This implies that the initial fall in money demand and the resulting exchange rate depreciation are also mitigated. The exchange rate effect mitigates the increase in foreign private demand for home goods. On the other hand, a higher substitution elasticity implies that public demand is more sensitive to the fall in relative prices and thereby leads to a higher foreign public demand for home goods. This second effect dominates the first one, and the increase in home short run output is amplified. In contrast, the effect of a higher public substitution elasticity on home consumption and output following a fiscal expansion is the opposite in the long run. Therefore, fiscal policy has ambiguous effects on welfare in both countries but it is prosper-thyself and beggar-thy-neighbor for plausible values of the parameters.
61 In welfare-based evaluations of policy efficiency, it is often assumed that utility from public spending is low or zero, which is a crucial assumption since the presence of public spending in the individual utility function may mitigate or even reverse the negative effect of fiscal policy on consumption and leisure and hence on welfare. However, not all types of public spending generate a direct utility for households as shown in Ganelli and Tervala [2010]. The authors distinguish between public spending on goods which have a direct impact on household utility and spending on public infrastructure which can improve the productivity of private firms. In this setup, a permanent domestic shift in the composition of public spending toward public infrastructure increases home output and consumption. The effect on home welfare is positive provided that the public capital productivity is not too low and the utility of public consumption is not too high relative to the private consumption utility. The foreign welfare falls in the short run because foreign producers have to meet a higher demand with relatively less public infrastructure. In contrast, foreign welfare increases in the long run. This can lead to a virtuous technological cycle if the foreign country reacts to the fall in short run foreign welfare by increasing foreign public infrastructure expenditures.
62 To stimulate economic activity without increasing the budget deficit, governments in most countries decide to reduce the number of state employees in order to cut taxes. Distinguishing between spending for public employment and public consumption, Ganelli [2010] shows that a permanent tax cut financed by a lower wage bill leads to an increase in short run domestic consumption relative to foreign consumption because home agents anticipate a subsequent reduction in taxes. The resulting higher home consumption leads to an increase in money demand at home which causes an appreciation of the home currency. The tax cuts are not compensated with lower income in the long run since employment in the private sector increases thanks to wage flexibility. As the setup assumes that productivity is higher in the private sector with respect to the public sector, the fall in public employment increases the resources available for the more productive sector. Therefore long run consumption and output gaps across the countries increase.
4.2. Fiscal policy with a debt stabilization motive
63 The stimulus plans launched in Europe and U.S.A in 2008-2009 caused an increase in public debt, which forced governments to take certain measures in order to reduce the public debt to safer levels. In a setup with Ricardian equivalence, any public debt resulting from a fiscal expansion is gradually offset by an increase in taxes in the following periods. As for the empirical models in which dynamics of public spending is given by a simple AR (1) process allowing a gradual fall in public revenue and expenditures, they fail to evaluate correctly the dynamic effect of fiscal policy according to Favero and Giavazzi [2007]. Indeed, those models neglect the possible feedbacks from public debt to public spending and taxes.
64 Several recent papers introduce such feedbacks into the analysis of various fiscal measures. For example, Coenen et al. [2008] consider a positive feedback between taxes and the debt-to-output ratio implying that taxes will automatically increase whenever debt-to-output ratio increases [2].
65 In practice, governments advocate for reducing public spending rather than increasing taxes in order to stabilize public debt, especially because of the limited capacity of governments to levy taxes due to the reluctance of taxpayers. Accordingly, in a two-country setup, Corsetti et al. [2010] analyze the effects of an increase in public spending with a feedback from public debt to public spending including “hand-to-mouth” households who consume all their disposable income at each period. According to this feedback mechanism, an initial increase in public spending financed mostly by debt rather than taxes leads to an increase in public debt. Government starts reducing the public spending in response to higher debt. In this setup, public spending may even fall below its steady-state level in the medium run. The authors name this mechanism as spending reversals.
66 Under this mechanism, the long run interest rate, defined as the infinite sum of the future short run real interest rates, increases following the fiscal expansion as a reaction to the debt, but starts to fall after a while as public debt starts to stabilize. As a result, Ricardian households increase their consumption which enhances the positive short run effect of public spending on the consumption of hand-to-mouth households. In the absence of spending reversals, there would be a slight fall in the short run total private consumption because of the fall in Ricardian households’ consumption despite the increase in the consumption of hand-to-mouth households.
67 The positive short run effect of fiscal policy on output is reversed in the medium run as public spending falls in response to the increasing debt. The cross-border effects imply an increase in foreign consumption under debt stabilization whereas AR (1) type fiscal processes lead to a fall. However, since home consumption increases more than the foreign consumption, trade balance worsens (improves) more in the short run (medium run) with respect to the AR (1) type fiscal process. The immediate impact on foreign output is also positive but becomes negative in the medium term.
68 Although Corsetti et al. [2010] are not directly inspired from the fiscal stimulus plan implemented in Europe during the recent crisis, their feedback mechanism [3] may have important normative policy implications to the extent that they show why governments might want to accompany a fiscal stimulus plan with an announcement on future public spending cuts. An institutional constraint, such as the European Growth and Stability Pact guaranties the credibility of such an announcement.
69 Stimulated by the recent crisis, several other papers attempt to estimate the quantitative effect of the fiscal policy pursued during the global crisis. For example, Cogan et al. [2009] analyze the impact of the US fiscal policy as it is defined by the legislation in February 2009 taking into account the possible future variations of the interest rate. Similarly, Cwick and Wieland [2011] evaluate the impact of the European fiscal stimulus plan announced in 2009 using several models with different assumptions. Most of these models are applied by international institutions in Europe.
70 The majority of the papers that analyze the effectiveness of fiscal policy in light of the recent crisis consider a setup in which fiscal policy is the only shock. Therefore, those papers exclude an explicitly stochastic setup in which fiscal policy is not a source of shock but an instrument to respond to an exogenous disturbance. The stochastic setup with an exogenous disturbance is necessary to analyze the effectiveness of fiscal policy as a stabilization tool. To this end, the next section introduces the possibility of exogenous shocks and analyzes how a stabilization policy should react to those shocks.
5. Fiscal stabilization and cooperation in a two-country stochastic world
71 The literature on the stabilization role of fiscal policy has grown rapidly in recent years. It analyzes mainly the optimal fiscal reactions to asymmetric shocks and evaluates the welfare gains from such fiscal reactions. In this case, fiscal policy is no longer discretionary as in section 3 and 4. Instead, fiscal reactions are based on rules and commitment similarly to the stability and growth pact applied in Europe.
72 Obstfeld and Rogoff [2000, 2002] provide a first attempt to evaluate the gains from monetary stabilization and cooperation policies in a micro-founded static stochastic setup. That framework is then adapted to fiscal policy analysis. Others introduced a dynamic aspect in order to study the dynamics of optimal fiscal stabilization within a DSGE setup. As such, they extend the traditional analysis of optimal taxation to an open economy setup and combine this framework with that of optimal monetary policy. This extension was mainly used to analyze the optimal combination of monetary and fiscal policy in a currency union governed by a common authority which ensures full cooperation between all stabilization policies.
5.1. Welfare gains from international fiscal stabilization and coordination
73 The main structure of the static two-country stochastic models on fiscal stabilization is generally similar to that used by Obstfeld and Rogoff [2000, 2002] for monetary stabilization. These models are similar to those analyzed in the previous sections regarding household preferences and technology. However, the extension of the setup to a stochastic environment requires additional features. First, in contrast to the discretionary fiscal policy analyzed in the previous sections, fiscal action results endogenously from the maximization of a social welfare function. Therefore, fiscal policy must be defined in terms of a fiscal rule. Second, the social welfare function is derived endogenously from the individual utility function making use of the equilibrium conditions. It depends on the second moments of the relevant endogenous variables and of the exogenous shocks. Fiscal authority is able to affect social welfare by reacting to the exogenous shock according to the fiscal rule.
Social welfare
74 The expected utility of the representative home household j is defined by the following equation which is the static version of the relation (7):
)1 − ρ M
jj
κμ
EU=E +χlog +ηlogG−
(L)
where ρ, χ, η, μ, κ > 0 [15]
j 1−ρ Pt j μ j
75 In equation (15), ρ represents the coefficient of risk aversion and κ corresponds to a random productivity shock which affects the marginal utility of work effort. The preferences of the foreign household are similar.
76 The price and consumption indexes are of Cobb-Douglas type as in Coutinho [2008] and Koenig and Zeyneloglu [2010b] who assume a unit risk aversion coefficient. In a broader view, Lombardo and Sutherland [2004] assume consumption indexes of CES type. For simplicity, we follow Obstfeld and Rogoff [2002] and distinguish between traded and non-traded goods in order to allow for imperfect risk sharing. Cobb-Douglas type price and consumption indexes are given as follows:
C = CT CN and P = PT PN
77
where the traded goods consumption index CT is defined as 
78 Following Obstfeld and Rogoff [2000, 2002], most of the static-stochastic models assume that goods prices are flexible while wages are preset, arguing that this assumption is more realistic. Each home household j offers differentiated labor services Lj and each home firm k produces a differentiated good using the labor services supplied by the households according to the production function yk = Lk. The demand for labor input provided by a household j is given by the following equation for a constant elasticity of goods demand:
j
Lj,k = W̄ yk [16]
79 where ϕ and W denote respectively the elasticity of substitution between different types of labor and the wage index.
80 The representative home household j sets the nominal wage in advance of production and consumption and, ex post, supplies the labor services that firms demand at Wj. For this, he maximizes the expected utility function (15) taking into account the budget constraint (2) and the labor supply which is equal to demand given in (16). Remembering that total output is equal to total labor used for production, the first order condition with respect to W
81 j
82 yields the following equation for the optimal preset wage where the subscript j is dropped:
ϕ−1
E [(L/P) C− ρ]
83 The optimal wage depends on the expected marginal utility of consumption and on the expected future labor effort i.e. on the uncertainty implied by an eventual productivity shock which may trigger a policy reaction. The optimal preset foreign wage W* is defined similarly.
84 Contrary to the papers which assume predetermined wages, Lombardo and Sutherland [2004] assume predetermined prices where the optimal preset price includes a form of risk premium.
85 When prices are flexible, profit maximization in each country implies that prices are a markup over wages. Combined with the law of one price for traded goods, this relation between prices and wages in each country allows to express the terms of trade τ as follows:
86 Hence, the terms of trade depend on the variables that affect home and foreign wages given in (17) and its foreign analogue.
87 As in Obstfeld and Rogoff [2002], it will be convenient to express private consumption in terms of the total demand X measured in traded goods. In this case, the goods market equilibrium condition implies PT X = PH L. Assuming that public spending is equal to a fraction k of total spending, we have:
C = X (1 − k) [18]
P
88 Following Obstfeld and Rogoff [2002], we can omit the utility from real balances in (15). Using the goods market equilibrium condition given above and making use of the fact that prices are a markup over wages, it is possible to express expected utility given in (15) only in terms of private consumption. Then we can make use of (18) along with the definition of price indexes given further above to rewrite the expected utility in terms of the total spending and the terms of trade:
89 where c is some constant. Now we need to transform (19) so that it is expressed in terms of first and second moments of the key variables, which requires a second order approximation. For this, one first has to take the log of (19), then rewrite it in exponential form and put it into expectations operator. Assuming that all variables are log-normally distributed, it is possible to rewrite (19) as follows:
EU = cexp (1 − ρ) Ex + Eτ + σ2 +
2 2x
((1 − ρ) (1 − α) )22 (1 − ρ )2 (1 − α) }
8 σe+ 2 σex [20]
90 In (20), lowercase letters denote the log values and σyz denotes the covariance between any variable y and z. With a similar algebra, it is possible to replace the first moments in (20) with the second moments of x, τ and the productivity shock. Following the same steps for the foreign welfare, one can show that the second moments related to the terms of trade enter the foreign welfare with opposite sign. This implies that the terms of trade can be a potential source of conflict between home and foreign fiscal authorities which may result in a potential gain from cooperation strategy.
91 Social welfare function in Coutinho [2008] is similar to (20) due to its similarity with Obstfeld and Rogoff [2002]. In contrast, welfare functions defined in Lombardo and Sutherland [2004], Andersen and Spange [2006] and Spange [2007] depend on consumption, output, and public spending in addition to the terms of trade [4].
92 This setup allows a better analysis of stabilization and cooperation policy compared to the earlier certainty-equivalent models.
Fiscal policy
93 Fiscal policies considered in static stochastic models of stabilization and cooperation policy take the form of a state-contingent fiscal rule which implies that the fiscal reaction (a variation in the fiscal policy instrument) is conditional on the occurrence of a productivity shock and its magnitude is proportional to the innovation in the latter. The proportionality is determined by the policy parameter.
94 In order to study the effect of fiscal stabilization, the fiscal policy instrument must appear in household and firm decisions given by the first order conditions for maximum social welfare. There are several ways to achieve this. For example, the setup presented above introduces money through a cash-in-advance constraint which implies that public spending has a direct effect on money demand and thereby on the terms of trade (via the exchange rate). Another possibility is to consider that public spending is financed by distortionary taxes. The introduction of labor income taxes as in Andersen and Spange [2006] implies that the tax rate appears directly in the optimal wage and fiscal policy acts through its effect on labor supply. Indeed, when a shock occurs, the labor supply decision implied by the preset wage is no longer optimal for the household. In order to restore the optimality of labor leisure trade-off, fiscal authority can manipulate the labor supply decision by modifying the labor income tax rate. If public spending is financed by corporate income taxes as in Coutinho [2008], the tax rate appears in the pricing decisions of firms which determine the terms of trade. Hence, fiscal policy acts through its effect on goods prices in order to manipulate the latter.
95 Lombardo and Sutherland [2004] offer another way of introducing a link between private decisions and public spending by allowing asset trading among countries. Home asset yields a pay-off of one unit of home disposable income while the foreign asset yields one unit of foreign disposable income. International consumption risk arises from the volatility of relative real disposal income which, in each country, is related to that of relative prices of assets. Thus, fiscal authority influences the asset holding decisions and thereby the degree of international risk sharing through its effect on the disposable income.
96 In the Nash strategy fiscal authority maximizes the welfare in its own country whereas cooperation strategy requires maximizing a weighed average of home and foreign welfare. Welfare maximization by policy makers requires that fiscal policy instrument appear in the welfare. In our simple setup presented above, this is done by defining a simple linear public spending rule as a function of the exogenous shock. As such, one can express the variances/covariances of the endogenous variables in (20) as a function of the public spending which can then be chosen so as to maximize the welfare given in (20).
The welfare impact of fiscal stabilization and coordination
97 In the setup presented above, fiscal authorities aim at manipulating the terms of trade by increasing (decreasing) the home (foreign) public spending following an asymmetric productivity shock. Although active fiscal policy improves welfare, fiscal authorities cannot fully compensate the negative impact of the shock on welfare regardless of the chosen strategy. Whether cooperation is desirable compared to Nash strategy depends on the values of the parameters. This setup considers only fiscal policy and therefore is unable to analyze the question of cooperation between all policy makers. This question is especially important for the case of European Monetary Union. Indeed, by forming a currency union, Europe has achieved a situation that is equivalent to a perfect coordination of national monetary policies without, however, committing to coordinated fiscal policy. Since then, whether countries can benefit from fiscal coordination when monetary policy is already cooperative has been an appealing question for economists. This question is addressed by Andersen and Spange [2006] and Spange [2007] in similar setups. The former considers a currency union with generalized wage rigidity while the latter assumes wage rigidity only in one country, which may be justified by the disparity in wage setting behavior across European countries. In this setup, the impact of fiscal policy on welfare depends on the consumption and output volatility induced by the fiscal reaction. An increase in policy coefficients (a more aggressive fiscal reaction) affects the level of risk i.e. the volatility of consumption and output. Households cannot cover themselves against this risk because of the absence of financial markets.
98 In the case of generalized wage-rigidity, an increase in home (foreign) policy reaction in response to a productivity shock improves home (foreign) welfare by reducing consumption volatility despite higher output variance due to that of public spending. Under wage asymmetry, the impact of fiscal policy depends on the correlation of output with consumption. If the correlation is positive, a more aggressive fiscal policy improves welfare at home. Otherwise, fiscal policy may become counter-cyclical. In both cases, fiscal policy spillover effects are positive, implying potential gains from coordinated fiscal reactions. However, these gains are small according to the numerical analysis of Andersen and Spange [2006].
99 In contrast to Andersen and Spange [2006], Lombardo and Sutherland [2004] introduce monetary policy alongside fiscal policy. This allows the authors to consider the welfare gains from monetary and fiscal stabilization in countries with separate currencies as well as in a monetary union. Since the authors assume price rigidity, home policy makers have to choose between eliminating the distortion effect of price rigidity and reducing the deterioration of the international consumption risk sharing which arises from the volatility of relative real disposal income depending on the relative prices of financial assets. In order to restore the optimality of the preset prices disrupted by a lower disutility of work effort at home following a productivity shock, home policy makers are induced to reduce home output through their effect on aggregate demand while searching to offset the impact of the terms of trade on home output. Moreover, home fiscal authority reduces the variability of home disposable income in order to shift relative asset prices in favor of home households and thereby improve risk sharing. The trade-off between these two strategies will imply potential gains depending on the strategy chosen by monetary and fiscal authority.
100 Under preset prices and flexible exchange rates, international monetary cooperation alone is able to eliminate distortions from price rigidity regardless of the fiscal regime if productivity shocks are perfectly correlated across countries. This result applies also to a currency union where the central bank ensures the perfect monetary cooperation. Under both regimes, when monetary policy is coordinated, there are gains from fiscal cooperation if home and foreign shocks are asymmetric. Contrary to monetary cooperation, an international fiscal cooperation alone cannot achieve full stabilization under flexible exchange rates. However, the impact of fiscal cooperation on welfare is increased by a monetary cooperation. When monetary policy is uncoordinated, whether there are welfare gains from fiscal cooperation with respect to non-cooperative fiscal policy depends on the value of the substitution elasticity between home and foreign goods.
101 Some of the above-mentioned results are altered when one extends the analysis of Lombardo and Sutherland [2004] to a multi-sector economy. Indeed the number of sectors within a country offers another source of policy interaction if each sector is affected differently by the shock. In a multi-sector framework excluding fiscal policy, Canzoneri et al. [2005] show that monetary cooperation gains can be non-negligible. Koenig and Zeyneloglu [2010b] extend this analysis by introducing fiscal policy alongside the monetary policy. In this setup, policy makers do not seek to manipulate the terms of trade in contrast to other models mentioned above. Instead, they aim at restoring the optimality of labor-leisure trade-off. In the multi-sector setup presented in Koenig and Zeyneloglu [2010b], the trade-off between monetary and fiscal policy arises from two assumptions. First, although the fiscal authority can intervene separately in different sectors, complete home bias in public spending prevents it from reacting to shocks in domestically consumed foreign goods sector. Second, although monetary authority can intervene in all sectors, this intervention affects all sectors in the same way regardless of whether they are hit by the shock. Assuming three production sectors in each country (non-tradable goods, domestic and imported tradable goods), the authors show that welfare gains from fiscal and monetary stabilization and coordination depend not only on the degree of asymmetry among the shocks but also on the type of sector that is hit by the shock. For example, when the shock occurs only in the nontraded goods sector, active fiscal policy improves upon monetary cooperation regardless of the degree of correlation between home and foreign shocks.
102 In contrast to the analysis above, Coutinho [2008] abstracts from public spending to focus on the optimal design of tax rules that react to shocks assuming perfect consumption risk sharing. In this case, monetary policy alone is capable of eliminating the negative effect of the shock on welfare, implying that there are no interactions between monetary and fiscal policy. Nevertheless, it is possible for the fiscal policy to improve monetary policy efficiency when the two countries form a currency union since fiscal policy becomes the only instrument that allows to react to asymmetric shocks. Under Nash strategy each fiscal authority is induced to manipulate the terms of trade against the other. Hence there are gains from fiscal cooperation depending on the elasticity of labor supply.
103 The static stochastic models offer a tractable way of introducing new channels of interactions between different variables through covariances which did not exist in the first generation models that treated stabilization and cooperation issues. However, the analysis of the role for fiscal policy remains limited. Moreover, the static structure of the models leads to the omission of inflation in the welfare equation. Therefore, those models overlook the possible policy trade-offs between stabilizing output and inflation. The implications of such trade-offs are studied in dynamic stochastic setups analyzed in the next section.
5.2. Optimal monetary and fiscal policy combination in a currency union
104 The current coexistence of a common monetary authority and independent fiscal authorities in the European Monetary Union (EMU) gives rise to several criticisms on the fiscal policy coordination role for the Stability and Growth Pact as well as on the monetary policy management of the European Central Bank (ECB). The inconveniences implied by the absence of an explicit and full cooperation among all macroeconomic policies leads to the necessity of considering an evolution in favor of stronger cooperation between policy authorities to evaluate the potential additional gains. Several recent papers offer an analytical framework of a currency union where monetary and fiscal authorities fully cooperate in order to maximize welfare at the union level. Beetsma and Jensen [2005] and Gali and Monacelli [2008] provide such a framework by extending Benigno [2004] where the analysis is limited to optimal monetary policy. Ferrero [2009] extends Beetsma and Jensen [2005] and Gali and Monacelli [2008] by considering additional distortions on labor decisions through distortionary taxes. The inclusion of such taxes requires resorting to the approach proposed by Benigno and Woodford [2003] to define the social welfare function. Thus, it becomes possible to consider optimal stabilization policy in the absence of Ricardian equivalence.
Main features of dynamic stochastic general equilibrium models
105 DSGE models that analyze the optimal relation between the stabilization policies are characterized by several common assumptions. They assume imperfect competition on goods markets, complete financial markets at both national and international level (in contrast to the static-stochastic models) and staggered price setting à la Calvo [1983]. Moreover, in these models, money does not appear in the individual utility function which implies the absence of seigniorage revenues as means of financing public spending. The price-setting problem yields a forward looking Phillips curve which depends on the terms of trade, output and the fiscal policy instrument.
106 The currency union generally consists of two interdependent countries, except in Gali and Monacelli [2008] who argue that a multi-country framework including a continuum of small economies without interactions with each other, except the common monetary policy, is more realistic in a currency union consisting of 18 countries as the euro area.
107 While the monetary policy instrument is the interest rate without exception, the fiscal policy instrument can be public spending or tax rate. Several papers assume that public spending is welfare enhancing while others consider government expenditures as pure waste. However, a more important difference between these models stems from the assumptions concerning the way of financing public expenditures.
108 Several papers assume that public spending can be financed by either lump sum taxes or by public debt since Ricardian equivalence holds. Under that assumption, Beetsma and Jensen [2005] determine the optimal policy problem by taking a second order approximation of the individual utility function which is then maximized under technology and resource constraints. Instead of defining a welfare function, the authors define a social loss function which results from the deviations of the variables with respect to their target (flexible-price equilibrium) values. Optimal policies are obtained by minimizing the social loss function. The general expression of this loss function at the union level is given as follows:
l = ∑ βs − t Et (ls ) [21]
s=t
109 where the period loss function is given as:
~U ~U ~R
(
Cs ) (Gs ) − λτG (τ̃s ) (Gs
) [22]
110 where a tilde indicates the gap between current and benchmark values.
111
According to (22), the social loss depends on the quadratic values of the national inflation rates ((π )2, (π* )2), the terms-of-trade gap 
112 The social welfare function in Gali and Monacelli [2008] is somewhat simpler because the terms of trade can be expressed as a function of the gap between country inflation and union inflation. Moreover, the cross terms disappear since there are no interactions between member countries in their multi-country framework.
113 Assuming distortionary taxes as the fiscal policy instrument and government expenditures as pure waste, Ferrero [2009] derives a welfare function similar to the one defined above but without the variance as well as the cross-terms in public spending.
Optimal monetary and fiscal policy
114 The analysis of dynamic optimal monetary and fiscal policy in a monetary union consists first in determining a benchmark framework with flexible prices that corresponds to a dynamic optimal equilibrium yielding the maximum level of welfare at the union level. Then, a new equilibrium is defined under the assumption of sluggish prices. Taking a second-order approximation to the utility function making use of the second-order approximations of the equilibrium conditions yields a quadratic welfare function for both the flexible and the sticky-price versions. The gap between the welfare under flexible and sticky-price equilibriums can be interpreted as a social loss function due to price-rigidity.
115 While the monetary policy block is quite similar in all dynamic stochastic models of a currency union, the fiscal policy block may differ significantly. Indeed, in contrast to Beetsma and Jensen [2005] and Gali and Monacelli [2008], Ferrero [2009] analyzes the stabilization role for distortionary taxes and public debt rather than public spending financed by lump sum taxes.
116 Beetsma and Jensen [2005] solve for the optimal combination of national fiscal policies and common monetary policy under full cooperation among all policy makers with a commitment strategy characterized by inertial targeting rules.
117 This setup allows to define targeting rules for the monetary and fiscal authorities. Indeed, the optimal policy implies a relation between union-wide inflation rate and the union-wide consumption gap (the difference between the percentage deviation from the steady-state under sticky and flexible-price equilibriums). It is possible to interpret this relation as a monetary targeting rule which implies that in case of an increase in union-wide inflation monetary authority commits to increase the interest rate for a sufficient period of time in order to reduce the consumption gap. The commitment of monetary authority to a prolonged restrictive monetary policy helps reduce the inflationary expectations following a symmetric productivity shock, which stabilize consequently the current inflation rate. Contrary to the monetary policy, fiscal policy has no stabilization effect on union-wide variables and therefore its role is limited to the stabilization at the national level on which monetary authority has no direct effect. Indeed, the optimal fiscal rule implies a prolonged restrictive policy in case of an increase in relative inflation which reduces current inflation through its effect on inflation expectations. If there is a variation in the terms-of-trade gap implying an expenditure-switching effect and a variation in relative inflation, the effects on output can be mitigated by a relative fiscal policy. Therefore, in the case of equal price stickiness, the optimal policy combination consists for the central bank to stabilize the union wide inflation and consumption gap and for the national fiscal authorities to stabilize the relative inflation and the terms of trade. However, in the case of asymmetric price stickiness across countries, monetary policy affects not only the union-wide variables but also the terms of trade.
118 The stabilization effect of fiscal policy is lower under discretion with respect to commitment because the effect of discretionary fiscal policy on inflation expectations is lower since this type of policy is less credible than commitment policies. While the commitment policies do not imply an inconsistency problem under flexible prices since the economy returns quickly to the steady state, the problem may arise when prices are sticky as the economy returns to the steady state only gradually. Indeed, once expectations are formed, fiscal authority may have an incentive to deviate from the announced policy. This failure to commit to the stabilization policy can lead to a non-negligible welfare loss according to the numerical results of Beetsma and Jensen [2005].
119 In a similar setup Gali and Monacelli [2008] analyze the effects of monetary and fiscal reactions to an asymmetric productivity shock assuming full cooperation between the common monetary policy and national fiscal policies in each country. Under flexible prices, monetary authority sets union wide inflation to zero and fiscal authorities let country inflation absorb any variation in the productivity level without pushing output and public spending beyond their first best levels. Under staggered prices, the necessary fall in the price level to absorb an increase in productivity is only gradual. Therefore, a policy trade-off arises from the fact that monetary policy has to stabilize union-wide inflation and the fiscal authority has to increase public spending in order to reduce excess supply due to higher productivity. This implies a non-zero output and inflation gap. Imposing a strong external constraint, such as the European Stability and Growth Pact, to members of a currency union is not desirable if the fiscal authorities have to meet asymmetric shocks.
120 Assuming that public spending is financed by distortionary taxes and public debt, Ferrero [2009] introduces an additional interaction between monetary and fiscal authorities that is absent in the previous framework. Indeed, under this assumption, monetary authority can influence the real value of public debt through the choice of the interest rate and inflation. In turn, fiscal authority can affect the pricing decisions through the tax imposed on production. Therefore, in contrast to the previous models, optimal policy plan involves an active reaction of fiscal policy at the union level by responding with the tax instruments to the output gap and inflation deviations. Fiscal policy also reacts to relative inflation differential as in Beetsma and Jensen [2005] but can achieve only partial stabilization because it also needs to stabilize the time path of the distortions from taxes. Moreover, in this full cooperation setup, the terms of trade distortions must also be internalized in contrast to Benigno and De Paoli [2009] who show, in a small open economy setup, that fiscal authorities may want to exploit the terms of trade channel to improve welfare under flexible exchange rates in the case of asymmetric productivity shocks.
121 Besides optimal stabilization policies, Beetsma and Jensen [2005] and Ferrero [2009] analyze alternative monetary and fiscal rules since they have the advantage of being more transparent than discretionary stabilization policy. Indeed, optimal policy plan can be difficult to implement due to an excessively rigid character. In contrast, simple policy rules can be easy to implement by the monetary authority and to be understood by the private sector. Therefore, there may be rules that can potentially imply a welfare performance close or equal to the optimal policy plan.
122 Two combinations of monetary and fiscal policy rules are considered in Beetsma and Jensen [2005]. The first consists of a monetary policy which uses the interest rate to close the consumption gap and of national fiscal policies which stabilize only the relative inflation by acting on the terms of trade. The second combination consists of a Taylor rule for the monetary policy and of a public spending gap depending negatively on output deviation. These rules yield a higher welfare with respect to the optimal policy under discretion.
123 Ferrero [2009] considers two other rules which allow to compare the rigid stabilization policy implemented in the EMU to a more flexible policy. The current policy combination in the EMU is approximated by a strict zero-inflation targeting rule for the monetary policy and a constant public debt targeting rule for the fiscal policy. The alternative policy is the one that allows flexible targets for inflation and public debt. Numerical analysis shows that the welfare performance of a combination of flexible debt targeting fiscal policy with a strict inflation targeting monetary policy is close to that of the optimal policy plan and exceeds that of a strict debt targeting combined with flexible inflation targeting policy.
6. Possible extensions
124 It is possible to identify three common restrictions in most of the DSGE models. First, the definition of the shock to which fiscal policy reacts is limited to a productivity shock. Second, monetary authority is assumed to be able to choose any level of interest rate without any restrictions. Moreover, by evaluating fiscal policy effectiveness only through its effect on output and consumption, they overlook its impact on employment as well as possible policy trade-offs between inflation and unemployment. The observed effects of the recent financial and economic crises encourage relaxing those restrictions to allow for three categories of extensions. Until now, only a few of the two-country DSGE models have considered these extensions.
125 First, by limiting the exogenous disturbances to supply shocks, the DSGE models fail to provide a completely adequate framework to analyze fiscal stabilization policies in the spirit of the recent crisis triggered by demand and financial shocks. For example, in Beetsma and Jensen [2005], the loss function implies that demand shocks may play a role only if they can affect the terms of trade. However, similarly to Beetsma and Jensen [2005], most of the DSGE models assume complete financial markets and thereby perfect international consumption risk sharing. Under that assumption, the terms of trade are only affected by supply shocks. The terms of trade are independent of the demand shocks because changes in relative consumption are sufficient to absorb the impact of the shock. In contrast, when financial markets are incomplete and the marginal utility of consumption differs across countries, demand shocks affect inflation and thereby the gaps of variables in both countries.
126 The introduction of the portfolio choice analysis in a stochastic model following Devereux and Sutherland [2011] may allow for an adequate framework to analyze optimal fiscal stabilization policy against financial shocks. For this, one can assume an exogenous shock to the equilibrium portfolio choice so that the portfolio choice is no longer optimal. Then, fiscal authority can react to that shock through taxes or subsidies on asset holdings in order to stabilize the financial sector. Forlatti and Lambertini [2011] provide another way of linking the financial and real sectors by introducing housing sector in a stochastic general equilibrium model. Following an increase in default risk, households who have invested in housing through a borrowing (mortgage) contract experience a worsening of their financial situation and therefore have to cut on the consumption of non-durable goods while increasing their labor supply. However, the authors do not study the stabilization role of economic policy. The extension of this type of setup to the analysis of fiscal stabilization may help generalize the conclusions on the stabilization capacity of fiscal policy.
127 Following the recent crisis, the Fed and the ECB have sharply reduced their policy interest rates and committed to keep the interest rate at a low level for an indefinite period of time. This type of policy can be considered as an example of the second class of extensions mentioned above. Under these circumstances similar to a liquidity trap, monetary policy is no longer effective to stimulate the economy. Indeed, the nominal interest rate cannot be further reduced and any increase in the money supply takes the form of money balances held by households or firms. In this case, policy makers may consider resorting to fiscal policy in order to fight the fall in the economic activity and the deflationary trend. In a closed economy setup Christiano et al. [2011] show that fiscal policy seems to be more effective when interest rates are fixed, which justifies fiscal intervention when the interest rate is close to the zero lower bound. Indeed, an increase in public spending increases inflation/reduces deflation, which implies a fall in the real interest rate for a given nominal interest. This, in turn, leads to a higher demand for consumption and investment implying a further fall in the real interest rate through its impact on inflation. If this process is sufficiently persistent the deflationary crisis can be overcome. According to Eggertsson [2009], a taxcut policy, especially consumption taxes, may have a similar impact provided that it succeeds in stimulating demand. In contrast, tax-cuts aiming at increasing the supply (for example tax-cuts that induce a reduction in wage costs) may aggravate the recession by increasing the deflationary pressures. Some of these results may be altered in an open-economy context. For example, according to Fahri and Werning [2012], the increase in inflation and the resulting fall in the real interest rate following a fiscal expansion may be mitigated by the fall in foreign demand due to higher relative prices. Extending the analysis of fiscal multipliers to a two-country setup could allow to evaluate the effectiveness of restrictive fiscal policies implemented in Europe on fighting the symptoms of a liquidity trap.
128 However, the capacity to emerge from a liquidity trap does not depend solely on the impact of fiscal multipliers. Eggertsson and Woodford [2004] show that an optimal combination of monetary and fiscal policy aiming at eliminating deflationary expectations may also enhance the capacity to emerge from a liquidity trap. In order to reduce deflationary expectations, central bank should commit to a nearly zero interest rate for one period and pursue afterwards a more relaxed policy than the one required by the prevailing circumstances. Fiscal authority, in turn, should increase the taxes that affect supply during the liquidity trap situation while announcing a fall in future taxes with respect to the long run level. While this analysis is inspired by the case of Japan, it can be applied to the European Monetary Union in order to extend the analysis of the optimal combination of monetary and fiscal policy studied in section 5 above.
129 The third class of extensions requires introducing unemployment. By excluding labor market frictions, the DSGE models with Walrasian labor markets cannot explain the high unemployment rate and its significant fluctuations in the euro area and in other countries between 2008 and 2015. By introducing sluggish price adjustment and using utility functions that depend negatively on labor effort expressed in terms of hours worked, the New Keynesian DSGE models can explain how fiscal policy could influence the intensive margin of labor (number of hours per worker). Therefore, these models are capable of analyzing a fiscal action that consists of adjusting hours worked in order to avoid job losses similar to the one implemented in Germany during 2008-2009. However, they cannot explain the sharp rise in the number of unemployed workers during the recent economic crisis. Recently, a growing number of New Keynesian DSGE models introduced search and matching frictions in the labor market, following Diamond [1982] and Mortensen-Pissarides [1994], in order to analyze how the imperfect competition and the nominal rigidities can interact with these frictions to affect the size of the policy impact on unemployment defined at the extensive margin. In this setup, unemployment arises from the fact that matching a vacancy post and an unemployed individual is costly and takes time. Among others, Monacelli et al. [2010] use such a setup to evaluate empirically the effects of a fiscal expansion on employment as well as on the efficiency of the matching process measured by the ratio of vacancies to unemployment and the probability of finding a job. The authors show that an increase in public spending reduces unemployment. In contrast to Monacelli et al. [2010], Bruckner and Pappa [2012] consider a setup with endogenous labor force participation and assume the presence of workers who are excluded from the labor market. In this setup, the authors show empirically that in a number of OECD countries a fiscal expansion leads to an increase in the number of workers employed and those unemployed at the same time.
130 What are the possible channels through which fiscal policy is likely to have an impact on employment? First, fiscal policy can have an impact on the real interest rate via the inflation, which then will have an impact on the bargained wage. Second, fiscal policy can also affect the firm’s vacancy posting behavior through its impact on the marginal value of hiring one more worker. Third, if one assumes that, in the wage bargaining process, firm’s surplus is measured in terms of producer prices while the worker’s surplus is measured in consumer prices, then fiscal policy will affect the equilibrium wage through its impact on the terms of trade.
131 However, the majority of this literature considers the effects of public spending on goods and services in a closed or small open economy setup neglecting therefore the heterogeneity of the public spending and the fiscal spillover effects across countries. Recently, Stähler and Thomas [2012] extended that setup to the case of a two-country monetary union. The authors calibrate the model for Spain and the rest of the Euro Area to evaluate the effects of various fiscal consolidation measures implemented in Europe on employment and output. These measures consist in reducing public consumption and/or investment along with a cut in public sector wage burden through a fall in wages or in the number of state employees. They may also include a reduction of taxes on labor income and consumption. The results show that cutting the public investment is the worst measure because it leads to the largest decrease in the real output and the largest increase in the unemployment rate, while decreasing the public employment or wage bill has the least negative effects. The spillover effects of a home fiscal consolidation on the neighbor country through its effect on the terms of trade depend on the type of consolidation. As the model is calibrated for Spain, it is difficult to derive generalized conclusions given the heterogeneity of labor markets in the European Union. It would be interesting to extend that setup to the analysis of optimal fiscal policy taking into account of the different degree of labor market flexibility across countries. For this, one can extend the analysis of Blanchard and Gali [2010] who analyze the optimal monetary policy in fluid (USA) and sclerotic labor markets (Europe).
7. Conclusion
132 Several recent economic developments, such as the formation of the EMU and the global economic crisis which is heavily felt during 2008-2009, have shown the necessity of resorting to fiscal policy as a stabilization tool. However, the implementation of fiscal policy implies difficulties due especially to the increasing degree of market integration at the international level. Moreover, the use of fiscal policy may have negative effects on public budget balance. Finally, the increasing interdependence across the countries implies non-negligible spillover effects, which requires international coordination of fiscal stabilization policies, especially in monetary unions where fiscal policy is the main stabilization tool at the national level.
133 Some of the policy prescriptions of the literature have been adopted by policy makers. For example, by advocating for a stronger SGP in October 2010, the European Council seeks to force governments to take into account the necessity of a debt stabilization at low levels, when deciding their future fiscal behavior. Nevertheless, it may be difficult sometimes for policy makers to derive definite policy prescriptions regarding the impact of fiscal policy and the influence of the market integration on fiscal policy efficiency. Indeed, the literature has not achieved a solid conclusion on the impact of fiscal policy on consumption and thereby on welfare which is considered as a measure of policy efficiency. The ambiguity stems often from the sensitivity of the results to the specification of their micro-foundations such as the degree of Ricardian behavior, separability in preferences, etc. However, empirical results can provide guidance for the assumptions regarding the household behavior.
References
- ANDERSEN T. M., SPANGE M. [2006], ″International interdependencies in fiscal stabilization policies″, European Economic Review, Vol. 50 (5), 1169-1195.
- BEETSMA R. M. W. J., JENSEN H. [2005], ″Monetary and fiscal policy interactions in a micro founded model of monetary union″, Journal of International Economic, Vol. 67 (2), 320-352.
- BENIGNO P. [2004], ″Optimal monetary policy in a currency area″, Journal of International Economics, Vol. 63 (2), 293-320.
- BENIGNO P., WOODFORD M. [2003], ″Optimal monetary and fiscal policy: a linear-quadratic approach″, NBER Macroeconomics Annual 2003, Vol. 18, 271-364.
- BENIGNO G., DE PAOLI B. [2010], ″On the international dimension of fiscal policy″, Journal of Money, Credit and Banking, Vol. 42 (8), 1523-1542.
- BILBIIE F. [2011], ″Non-separable preferences, Frisch labor supply and the consumption multiplier of government spending: one solution to a fiscal policy puzzle″, Journal of Money, Credit and Banking, 43 (1), 221-251.
- BLANCHARD O., GALI J. [2010], ″Labor markets and monetary policy: a New-Keynesian model with unemployment″, American Economic Journal: Macroeconomics, Vol. 2 (2), 1-30.
- BRUCKNER M., PAPPA E. [2012], ″Fiscal expansions, unemployment, and labor force participation: theory and evidence″. International Economic Review 53 (4), 1205-1228.
- BUCH C. M., DOEPKE J., PIERDZIOCH C. [2005], ″Financial openness and business cycle volatility″, Journal of International Money and Finance, Vol. 24 (5), 744-765.
- CALVO G. A. [1983], ″Staggered contracts in a utility-maximizing framework″, Journal of Monetary Economics, Vol. 12 (3), 383-398.
- CANZONERI M. B., CUMBY R. E., DIBA B. T. [2005], ″The need for international policy coordination: What’s old, what’s new, what’s yet to come″, Journal of International Economic, Vol. 66 (2), 363-384.
- ÇENESIZ A., PIERDZIOCH C. [2009], ″Efficiency wages, financial market integration, and the fiscal multipliers″, Journal of International Money and Finance, Vol. 28 (5), 853-867.
- CHRISTIANO L., EICHENBAUM M., REBELO S. [2011], ″When is the government spending multiplier large?″, Journal of Political Economy, Vol. 119 (4), 78-121.
- COENEN G., MCADAM P., STRAUB R. [2008], ″Tax reform and labor-market performance in the euro area: A simulation-based analysis using the New Area-Wide Model″, Journal of Economic Dynamics and Control 32 (8), 2543-2583.
- COGAN J., CWICK T., TAYLO J. B., WIELAND V. [2010], ″New Keynesian versus old Keynesian government spending multipliers″, Journal of Economic Dynamics and Control, Vol. 34 (3), 281-295.
- CORSETTI G., MEIER A., MÜLLER J. [2010], ″Cross-border spillovers from fiscal stimulus″, International Journal of Central Banking, Vol. 6 (1), 5-37.
- COUTINHO L. [2008], ″Fiscal policy and macroeconomic stabilizations: What are the gains from cooperation?″, Open Economic Review. Vol. 19 (1), 81-120.
- CWICK T., WIELAND V. [2011], ″Keynesian government spending multipliers and spillovers in the Euro Area″, Economic Policy, Vol. 26 (67), 493-549.
- DEVEREUX M. B., SUTHERLAND A. [2011], ″Country portfolios in open economy macro-models.″, Journal of European Economic Association, Vol. 9 (2), 337-369.
- DIAMOND P. [1982], ″Aggregate demand management in search equilibrium″, Journal of Political Economy, Vol. 90, 882-902.
- EGGERTSSON G. B. [2009], ″What fiscal policy is effective at zero interest rates″, Federal Reserve Bank of New York, Staff Reports no 402.
- EGGERTSSON G. B., WOODFORD M. [2004], ″Optimal monetary and fiscal policy in a liquidity trap″, NBER Working Papers no 10840.
- FARHI E., WERNING I. [2012], ″Fiscal multiplier: liquidity traps and currency unions″, NBER Working Papers no 18381.
- FAVERO C., GIAVAZZI F. [2007], ″Debt and the effects of fiscal policy″, NBER Working Papers No12822.
- FERRERO A. [2009], ″Fiscal and monetary rules for a currency union″, Journal of International Economics, Vol. 77 (1), 1-10.
- FEVE P., SAHUC J.-G. [2015], ″On the size of government spending multipliers in the euro area″, Oxford Economic Papers, Vol. 67 (4), 865-885.
- FORLATI C., LAMBERTINI L. [2011], ″Risky mortgages in a DSGE model″, International Journal of Central Banking, Vol. 7 (1), 285-335.
- GALI J., MONACELLI T. [2008], ″Optimal monetary and fiscal policy in a currency union″, Journal of International Economics, Vol. 76 (1), 116-132.
- GANELLI G. [2008], ″Public spending management and macroeconomic interdependence″, Open Economic Review, Vol. 19 (2), 241-259.
- GANELLI G. [2010], ″The international effects of government spending composition″, Economic Modelling. Vol. 27 (3), 631-640.
- GANELLI G., TERVALA J. [2010], ″Public infrastructures, public consumption, and welfare in a New-Open Economy-Macro model″, Journal of Macroeconomics, Vol. 32 (3), 827-837.
- KOENIG G., ZEYNELOGLU I. [2010a], ″When does financial integration matter for fiscal policy in a currency union: A welfare-based approach″, Economic Modelling, Vol. 27 (3), 620-630.
- KOENIG G., ZEYNELOGLU I. [2010b], ″Monetary and fiscal policy efficiency and coordination in a multisector open-economy general equilibrium model″, Scottish Journal of Political Economy, Vol. 57 (4), 473-493.
- LOMBARDO G., SUTHERLAND A. [2004], ″Monetary and fiscal interactions in open Economies″, Journal of Macroeconomics, Vol. 26 (2), 319-347.
- LOMBARDO G., SUTHERLAND A. [2007], ″Computing second-order-accurate solutions for rational expectation models using linear solution methods″, Journal of Economic Dynamics & Control, Vol. 31 (2), 515-530.
- MONACELLI R., PEROTTI R., TRIGARI A. [2010], ″Unemployment fiscal multipliers″, Journal of Monetary Economics, Vol. 57, 531-553.
- MORTENSEN D., PISSARIDES C. [1994], ″Job creation and job destruction in the theory of unemployment″, Review of Economic Studies, Vol. 61 (3), 397-415.
- OBSTFELD M., ROGOFF K. [1995], ″Exchange rate dynamics redux″, Journal of Political Economy, Vol. 100 (3), 624-660.
- OBSTFELD M., ROGOFF K. [1996], Foundations of International Macroeconomics, Cambridge, MA: The MIT Press.
- OBSTFELD M., ROGOFF K. [2000], ″New directions for stochastic open economy models″, Journal of International Economics, Vol. 50 (1), 117-153.
- OBSTFELD M., ROGOFF K. [2002], ″Global implications of self-oriented national monetary rules″, Quarterly Journal of Economics, Vol. 117 (2), 503-536.
- PIERDZIOCH C. [2004a], ″Capital mobility and the effectiveness of fiscal policy in open economies″, Journal of Macroeconomics, Vol. 26 (3), 465-479.
- PIERDZIOCH C. [2004b], ″Financial integration and business cycle volatility in a monetary union″, Scottish Journal of Political Economy, Vol. 51, No. 3, 422-442.
- SCHMITT-GROHE S., URIBE M. [2004], ″Solving dynamic general equilibrium models using second order approximation to the policy function″, Journal of Economic Dynamics and Control, Vol. 28 (4), 755-775.
- SENAY Ö. [1998], ″The effects of goods and financial market integration on macro-economicvolatility″, The Manchester School Supplement, 39-61.
- SPANGE M. [2007], ″Do structural asymmetries increase the gains from international fiscal policy coordination?″, The Manchester school, Vol. 75 (1), 131-150.
- STÄHLER N., THOMAS C. [2012], ″FiMod-A DSGE model for fiscal policy simulation″, Economic Modelling, Vol. 29 (2), 239-261.
- SUTHERLAND A. [1996], ″Financial market integration and macroeconomic volatility″, Scandinavian Journal of Economics, Vol. 98 (4), 521-539.
- WARNOCK F. E. [1998], ″Idiosyncratic tastes in a two-country optimizing model: Implications of standard presumption″, FRB International Finance Discussion, No 631, Board of Governors of Federal Reserve System.
Mots-clés éditeurs : stabilisation et coordination, modèles d’équilibre général à deux pays, intégration économique, fiscal policy stabilization and coordination, economic integration, politique budgétaire, Open Economy Macroeconomics, Macroéconomie internationale, two-country general equilibrium models
Date de mise en ligne : 23/01/2017
https://doi.org/10.3917/redp.266.1023Notes
-
[*]
We would like to thank the anonymous referees for helpful comments.
-
[‡]
Galatasaray University, Department of Economics, Galatasaray Center for Economic Research (GIAM), 36 Çırağan avenue 34357, Istanbul, Turkey. Tel: +90 2122274480-590, fax: +90 2122285283. Courriel: izeyneloglu@gsu.edu.tr.
-
[✝]
Université de Strasbourg, Bureau d’Économie Théorique et Appliquée (BETA), UMR 7522 du CNRS, 61 Avenue de la Forêt Noire F 67085 Strasbourg Cedex, France. Tel: +33 (0) 3 68 85 20 69. Fax: +33 (0) 3 68 85 20 70. Courriel: koenig@unistra.fr. (Auteur correspondant)
-
[1]
In the initial steady-state, public spending is assumed to be equal to zero and the current account is balanced in both countries. The log deviation of any home variable x relative to x* is equal to the gap between the deviations of x and x* with respect to their steady-state values. Since the steady-state values of home and foreign variables are equal, the gap between the log-deviations of the two variables is equal to the gap between the log of those variables.
-
[2]
Coenen assumes that the ratio of taxes T to GDP responds automatically to public debt according to the following rule:
-
[3]
This mechanism is represented by the following AR (1) process for public spending Gt augmented with a public debt feedback:
-
[4]
When public spending is assumed to be welfare enhancing as in Andersen and Spange [2006] and Lombardo and Sutherland [2004], the expected level and the variance of public spending also enter directly in the welfare equations.
-
[5]
For example, an increasing negative terms-of-trade gap which decreases the home competitive advantage combined with a higher relative foreign spending shifts work effort from home to foreign households.
