Habit formation, adjustment costs, and international transmission of fiscal policy

Journal of International Money and Finance 32 (2013) 341–359

Contents lists available at SciVerse ScienceDirect

Journal of International Money and Finance journal homepage: www.elsevier.com/locate/jimf

Habit formation, adjustment costs, and international transmission of fiscal policy Ramon A. Gonzalez-Hernandez, Cem Karayalcin* Central Bank of the Dominican Republic and Lake Forest College, Department of Economics, Florida International University, DM Building, 3rd. Floor, 11200 S.W. 8th Street, Miami, FL 33199, USA

a b s t r a c t JEL classification: E13 E20 E32 Keywords: Habit formation Adjustment costs Fiscal policy International correlations

The paper studies the effects of fiscal policy in an integrated world economy. The setup is one with habit-forming endogenous rates of time preference and adjustment costs in investment. Most of the predictions of the model are in line with the recent empirical literature on fiscal policy. For instance, in response to a balanced fiscal expansion, we obtain positive long-run output multipliers, long-run increases in employment, short- and medium-run increases in wages and decreases in investment. Our results suggest that short-run government spending multipliers are smaller than tax multipliers. Most importantly, we show that the model can generate positive short- and medium-run consumption responses to a positive fiscal shock. This is relevant as negative consumption responses are considered to be one of the main challenges facing neo-classical models of fiscal policy.
2012 Published by Elsevier Ltd.

1. Introduction The ineffectiveness of conventional monetary policy instruments in dealing with the recession that started in late 2007 has once again pushed questions concerning the role and effectiveness fiscal policy to the forefront of the economic agenda. Furthermore, as evidenced by intense discussions concerning fiscal policy coordination among the G-20 countries, there is heightened awareness that the transmission of fiscal shocks across countries is a top policy issue. Thus, it is interesting to note that, unlike in the case of monetary policy, there does not seem to be an agreement among economists with regard to

* Corresponding author. Tel.: þ1 305 348 3285; fax: þ1 305 348 1524. E-mail address: karayalc@fiu.edu (C. Karayalcin). 0261-5606/$ – see front matter
2012 Published by Elsevier Ltd. http://dx.doi.org/10.1016/j.jimonfin.2012.04.008

342

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

the channels of transmission or the effects of fiscal shocks. As Perotti (2008) puts it, “perfectly reasonable economists can and do disagree even on the basic effects of a shock to government spending on goods and services: neo-classical models predict that private consumption and the real wage will fall, while some neo-Keynesian models predict the opposite”. Disagreements concerning the effects of fiscal policy have given rise to a large and growing empirical literature that uses increasingly more sophisticated tools and there appears to be some stylized facts that emerge from the more careful studies. These can be summarized as follows:  Government spending shocks have a positive effect on output, while positive tax shocks have negative effects.1 Government spending multipliers tend to be small.2 There is no consistent evidence that spending multipliers exceed tax multipliers (Blanchard and Perotti, 2002).  Private consumption increases in response to an increase in government spending and decreases in response to a rise in taxes.3  Increases in government spending and in taxes crowd out private investment (Alesina et al., 2002; Blanchard and Perotti, 2002; Mountford and Uhlig, 2008.)  Long-run real interest rates rise when government spending is increased (Fatas and Mihov, 2001; Gale and Orszag, 2004; Perotti, 2005; Dai and Philippon, 2004; Favero and Giavazzi, 2007)  Higher government spending leads to increased manufacturing wages (Fatas and Mihov, 2001; Perotti, 2008) and higher employment (Fatas and Mihov, 2001).  Stock prices fall in response to positive government spending shocks (Afonso and Sousa, 2009). Some of these stylized facts appear to contradict either the traditional Keynesian theory or the neoclassical approach. For instance, Blanchard and Perotti (2002) emphasize the lack of consistent evidence for government spending multipliers to exceed tax multipliers and note that “the response of investment, which decreases in response to both increases in taxes and increases in spending, is hard to reconcile with the Keynesian approach”. On the other hand, the neo-classical model is commonly believed to fall short in predicting the response of consumption. As Fatas and Mihov (2001) put it, “the biggest challenge to the [RBC model] is its inability to predict the response of consumption to shocks to government expenditures”. In what follows we construct an equilibrium model that tries to address the challenges raised by the empirical literature. Our construct extends the standard setup by incorporating habit formation and endogenous rates of time preference in an open economy environment with two-large economies. The endogeneity of time preferences is consistent with the assertion of Hicks (1965), who points out that the independence of consumption levels between successive periods implied by conventional timeadditive preferences is counter-intuitive and normally one should expect complementarity between them. Hicks’ argument has been corroborated by empirical findings that have generated strong rejections of time-additive preferences.4 We calibrate and simulate this model and show that it predicts, most importantly, that a balanced fiscal expansion (i) lowers output in the short-run and raises it in the medium and long-run, generating a balanced-budget multiplier of about 0.8; (ii) increases private consumption in the short- and medium-run, while reducing it in the long-run; (iii) reduces prices of equity and private investment; (iv) raises long-run real interest rates; (v) raises wages

1 Blanchard and Perotti (2002), Perotti (2005), and Fatas and Mihov (2001). However, Giavazzi and Pagano (1990) and Alesina and Ardagna (1998) have provided evidence of negative spending multipliers during “large fiscal consolidations”, while Perotti (1999) finds a similar outcome only in circumstances of “fiscal stress” (unusually high debt-to-GDP ratios). 2 Perotti (2005) estimates them to be larger than 1 only in the US and in the pre-1980 period. Though Fatas and Mihov (2001) write that “[t]here is a strong and persistent reaction of private output to a fiscal shock”, they find that the “maximum effect of an approximately 1% increase in spending is attained about two years after the shock with private output increasing by 0.3%”. 3 Blanchard and Perotti (2002), Perotti (2005, 2008), and Fatas and Mihov (2001). The “dummy variable” approach of Ramey and Shapiro (1998) and its extensions to VAR by Edelberg et al. (1999) and Burnside et al. (2004) typically find that during episodes of large, exogenous increases in defense spending private consumption falls. However, Perotti (2008) shows that once the restrictions imposed by these approaches are removed, private consumption increases in response to the fiscal shocks of the Ramey–Shapiro episodes. 4 See Obstfeld (1990) on the relevant literature.

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

343

in the short- and medium-run, lowering them in the long-run; and (vi) lowers employment in the short-run, but raises it in the long-run. In short, the model is able to replicate the most significant stylized facts concerning fiscal policy. The model we present below also predicts cross-country correlations in consumption, investment, employment and output that are all positive. Moreover, the predicted positive cross-country correlation in output is higher than the one in consumption. These predictions are in agreement with the observed empirical evidence and constitute progress over standard RBC models, despite the fact that the predicted correlations remain quantitatively higher than the ones reported in both Backus et al. (1995) and Ambler et al. (2004).5 Another important improvement is that the ratio of home-toforeign consumption exhibits cyclical adjustment. This stands in contrast with the typical result found in most of the two-country models in the literature where this ratio remains constant along the transition paths. Finally, the setup generates cyclical adjustment paths that are persistent, addressing another major shortcoming of real business cycle models, which need to rely on ad hoc shocks to generate persistence. In terms of the questions posed, there are three theoretical papers most closely related to the one we present here: Deveraux and Shi (1991), Corsetti and Pesenti (2001), and Karayalcin (2003). The first one develops a two-country model where rates of time preference are endogenous as in here (though not of the habit-forming variety). A positive fiscal shock in this first model is predicted to have consequences that run counter to the stylized facts outlined above: consumption and investment fall both in the short- and long-run, with the long-run real interest rates also experiencing a decline. In a two-country model with nominal rigidities, Corsetti and Pesenti (2001) predict that while there will be no change in consumption in the short-run, higher government spending will reduce consumption globally, while increasing income and wages in the long-run. Karayalcin (2003) builds a model that combines habit formation with adjustment costs to analyze the effects of fiscal policy in a small open economy. The incorporation of adjustment costs in investment tends to lower the degree of consumption smoothing as agents are no longer able to undertake frictionless adjustments in the capital stock. Used together with habit-forming preferences, adjustments costs give rise to cyclical and persistent transition paths. Here we differ from Karayalcin (2003) in extending the model to the case of two-large economies enabling us to study the effects of fiscal shocks on real interest rates and their international transmission. The basic changes we introduce to the Deveraux and Shi (1991) environment (and that account for the differences in predictions) are habit formation and adjustment costs in investment. Habit formation (i.e., the existence of “adjacent complementarity” between an individual’s current and past consumption or felicity levels) has recently been the subject of intensive investigation. The influence of such behavior on asset pricing was analyzed and empirically tested by, most prominently, Abel (1990) and Constantinides (1990). Ferson and Constantinides (1991), Heien and Durham (1991), and Heaton (1995) test habit persistence using consumption data and found the interaction between these two to have important effects. Carroll and Weil (1994) argue that the observed relationship between growth and saving may be explained by a model of consumption with habit formation.6 Mansoorian (1993, 1996) and Obstfeld (1992) fruitfully apply this idea to the analysis of open economies. Fuhrer and Klein (2006) lends support to the hypothesis that habit formation characterizes consumption behavior among most of the G-7 countries. Sommer (2007) shows that habit formation in consumer preferences can explain two well-known failures of the permanent income hypothesis. Jermann (1998) develops a closed economy model that uses habit formation and adjustment costs in investment and shows that

5 As is well-known, a comparison of the predictions of the calibration exercise in Backus et al. (1995) with both the observed correlations in their sample and with the more robust cross-country correlations in the Ambler et al. (2004) study (See Table 1) yields three important puzzles: (i) the observed low positive cross-country correlation in consumption is inconsistent with the large positive correlation predicted by theory, (ii) the stylized facts show positive cross-country output correlations that exceed cross-country consumption correlations while the baseline model points in the opposite direction (with the additional flaw that the predicted cross-country correlation in output is negative), and (iii) cross-country correlations of investment, employment and output are positive rather than negative as suggested by the baseline real business cycle model. However, Chari et al. (2002) present models which go a long way in solving puzzles (ii) and (iii). 6 Campbell (1994) stresses the need to explore models of habit formation.

344

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

this combination explains the historical equity premium and the average risk-free return observed in the data, while replicating many of the salient business cycle properties.7 The rest of the paper is organized as follows. Section 2 sets up the model of the integrated world economy consisting of two-large countries with fully integrated stock, output and capital markets. Section 3 analyzes the international transmission of a balanced-budget fiscal expansion that originates in one of the countries, and presents the elasticities and international cross-country correlations that arise from the calibration of our model. Section 4 provides some concluding remarks. 2. The model Consider and integrated world economy consisting of two-large countries, labeled “home” and “foreign”, with both countries having a similar structure in terms of preferences and technology. The number of households is normalized to one in each country. Firms within each country employ capital and labor to generate the same single traded good that can be used for consumption and investment. Governments collect lump-sum taxes in units of the single good being produced and use the entire tax proceeds to finance their expenditures. Government spending in both countries positively affects household utility and influences consumption and labor supply decisions directly. All agents operate under perfect foresight and there is no labor mobility8 among the two countries in the setup. The world markets for stock, output and capital are fully integrated in the framework. We now proceed to describe the behavior of households and firms in some detail. 2.1. Households Households derive felicity from consumption of private (C) and public (G) goods, and leisure (1  L).9 Saving takes the form of accumulation of equities issued by home and foreign firms. Equities from both countries are perfect substitutes in the portfolios of agents, thus they must yield the same rate of return (R ¼ R*).10 The representative household in the home country maximizes lifetime utility by choosing private consumption (C) and leisure (1  L) optimally, treating the level of government spending (G) as given at each point in time. In what follows, we adopt the habit-forming recursive time preference structure proposed by Epstein and Shi (1993), which is a more tractable extension of the utility function considered in Ryder and Heal (1973). In this setting consumption habits are modeled through the representative agent’s endogenous rate of time preference rather than through the agent’s instantaneous utility function. The lifetime utility takes the following form:

Z

N 0

0 VðC; L; GÞexp@ 

Z

t 0

1

bðZðsÞÞdsAdt

(1)

where V($,$,$) is the instantaneous utility function and b(Z) is an endogenous discount rate that depends on an index of past utility (stock of habits) denoted by Z. It is assumed that b(Z) satisfies b > 0, b0 > 0, b00 ¼ 0.

7 The idea that habit formation might help explain the international consumption correlations puzzle was suggested by Shi (1999), who conjectures that by habit-forming preferences would help reduce the correlations generated by the standard model in a two-country setup. 8 A central assumption in most models of international trade and finance is that labor is much less mobile internationally than either commodities or capital. Language and cultural barriers, family and ethnic ties, and political barriers all work to make international migration difficult. 9 Leisure is defined as total time available to households minus the time dedicated to labor activities. We set total time equal to one without loss of generality. 10 An asterisk denotes that a variable pertains to the foreign country.

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

345

We further set V($,$,$) h 1 which yields particularly simple solutions while preserving all the Z t bðZðsÞÞdsÞ the repreessential properties of a more general function. Now defining aðtÞhexpð 0

sentative home household’s optimization problem can be expressed as follows:

Z

N

max 0

ðaðtÞÞdt

(2)

subject to

A_ ¼ RA þ WL  C  T;

A0 > 0 given;

Z_ ¼ s½UðC; L; GÞ  Z;

a_ ¼ bðZÞa;

(3)

Z0 > 0 given;

(4)

a0 ¼ 1;

(5)

C; L; G; A  0 where s > 0 is the speed of adjustment in the stock of “habits” and A, W, T, U(C,L,G) denote the level of non-human (financial) wealth, wage rate, lump-sum tax receipts, and an aggregator function. To see what is involved in the optimization problem, it is useful to integrate (4)

Z ZðtÞ ¼ s

t N

UðC; L; GÞexpðsðs  tÞÞds

(6)

showing that Z(t) is a weighted average of past felicity levels with weights declining exponentially into the past at the rate s. If s ¼ 0 the conventional time-additive utility function with a constant rate of time preference is obtained. If, on the other hand, s ¼ N we get the Uzawa (1968) endogenous rate of time preference. We specialize the aggregator function to

h UðC; L; GÞ ¼ for w > 1;

C z G1z

g

z˛ð0; 1Þ;

ð1  LÞð1gÞ 1w g˛ð0; 1Þ;

i1w

1

þ X;

(7)

X>0

where we use a standard affine transformation of a typical nested CRRA utility functional. An important implication of this specification is that government spending affects the marginal utilities consumption and leisure, leading to direct interactive effects with private consumption and labor supply decisions. Another consequence of (7) is that U(C,L,G) satisfies UC > 0, UL < 0, UCC < 0, UCL > 0, and ULL < 0, as is supported by the conventional economic theory. The discounted Hamiltonian for the problem in (2)–(5) is

~ ½UðC; L; GÞ  Z  Fb ~ ðZÞ ~ H ¼ a þ MðRA þ WL  C  TÞ þ Js

(8)

with

~ aM; Mh

~ haJ; J

~ haF F

The standard solution technique yields, in addition to the constraints (3) and (4), the first-order necessary conditions

UC ¼ M=sJ;

(9)

UL =UC ¼ W;

(10)

346

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

_ ¼ M½bðZÞ  R M

(11)

J_ ¼ Jðs þ bðZÞÞ þ Fb0 ðZÞ

(12)

F_ ¼ 1 þ bðZÞF

(13)

where F is the shadow value of the auxiliary variable a and J and M represent the shadow values of the stock of habits and the non-human household wealth, respectively. Note that Ft by definition also measures the maximized lifetime utility starting from an arbitrary point in time t. First-order condition (9) reveals that an optimizing household will compare the marginal benefit of an extra unit of consumption with the effective or normalized marginal cost of increasing non-human wealth. Equation (9) differs from those obtained in time-additive frameworks in that it also incorporates the shadow value of the stock of habits as well. Equation (10) states the well-known intratemporal condition that the marginal rate of substitution of leisure for private consumption equals the opportunity cost represented by the real wage rate. Conditions (11)–(13) present the optimal rules of motion for the shadow values of non-human household wealth, the stock of habits and of the shadow value of the auxiliary variable a. In addition the following transversality conditions must be satisfied:

lim ai ðtÞMi ðtÞAi ðtÞ ¼ lim ai ðtÞJi ðtÞZi ðtÞ ¼ lim ai ðtÞFi ðtÞ ¼ 0:

t/N

t/N

t/N

It can be shown, as in Epstein and Shi (1993) that this setup gives rise to an endogenous rate of time preference r ¼ b(Z)  [J(s þ b(Z)) þ Fb0 (Z)]/J. The assumption b0 ($) > 0 implies that as households get wealthier they become more impatient to consume. This notion has struck some as counter-intuitive. However, those who defend the notion have offered a number of justifications of such “increasing marginal impatience” (see Lucas and Stokey, 1984). First, such preferences are necessary for dynamic stability. Second, the alternative, b0 ($) < 0, which implies that as households get wealthier their desire to accumulate wealth increases, is also counter-intuitive. Third, one could justify increasing impatience as pertaining to economies with relatively higher levels of consumption, while decreasing impatience would apply to low levels of consumption and wealth.11 Finally, there is empirical evidence supporting this last notion in the form of a non-linear relationship between income levels and savings rates, showing that savings rates first rise and then fall with increases in income levels.12 Moving now to a description of foreign households, it should suffice to point out that equations (9)– (13) have their foreign counterparts, with foreign variables denoted by asterisks.

2.2. Firms Identical, competitive firms in each country employ capital, K, and labor, L, to produce a single good that is used for both consumption and investment under constant returns to scale. Here we formalize firm behavior in the home country and leave it to the reader to extend this to the foreign firms. We assume that the production function is of the conventional constant elasticity of substitution (CES) type: Y ¼ F(K,L) ¼ L[dKh þ (1  d)Lh]1/h. For analytical convenience, the rate of depreciation of capital is set equal to zero. We suppose that there are adjustment costs in investment, so that it takes I [1 þ G(I/K)] units of output to increase the capital stock by I units. The installation cost function is specialized to G(I/K) ¼ (1/2c)(I/K). Firms choose the time path of investment to maximize the present discounted value of net profits _ 13 The solution of this problem yields P ¼ F(K,L)  I(1 þ G)  WL subject to the constraint I ¼ K.

Q ¼ 1 þ GðI=KÞ þ ðI=KÞG0 ðI=KÞ; 11 12 13

See Epstein and Shi (1993), Epstein (1987), and Obstfeld (1990) for more on increasing impatience. See, for instance, Sahota (1993), Masson et al. (1995), and Ogaki et al. (1996). For notational convenience it is assumed that investment is exclusively financed by retained earnings.

(14)

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

347

Q_ ¼ RQ  FK ðK; LÞ  ðI=KÞ2 G0 ðI=KÞ;

(15)

W ¼ FL ðK; LÞ:

(16)

where (14) implies that the shadow value Q of capital (Tobin’s Q) is equal to one plus the marginal cost of investment. The law of motion for Q is given by (15). (16) is the usual intratemporal equilibrium condition of marginal productivity of labor being equal to the wage rate. Equation (14) can be used to express investment as the following function of Q:

I ¼ K_ ¼ K4ðQ Þ;

40 ðQ Þ ¼ c > 0;

4ð1Þ ¼ 0:

(17)

That is, investment is an increasing function of the value of capital; and when the value of capital equals its unitary replacement cost (Q ¼ 1) investment is zero. Analogous optimality conditions hold in the foreign country. 2.3. The government The government is modeled in the most simple way here. As Ricardian Equivalence holds in our setting, we assume that all government spending is financed by lump-sum taxes. Thus, G ¼ T and G* ¼ T*. Examples of previous studies with household utility depending positively on the level government spending can be found in Bailey (1962) and Baxter and King (1993). 2.4. Markets and prices At a given point in time for households to be satisfied with the composition of their portfolios the rates of return on home and foreign equities, R and R), should be identical

 Rh

P þ QI

 þ

QK

    P þ Q  I Q_ Q_ þ  hR ¼   Q K Q Q

(18)

 where the terms in parenthesis denote an equity’s current yield, and Q_ =Q and Q_ =Q  represent capital gains. The current yield terms consist of “cash” dividends, P and P*, and equity dividends, QI and Q)I). For modeling purposes it helps to define Q h Q)/Q as the relative price of foreign equity, and rewrite condition (18) as follows:

Q_ ¼

 

Q

P þ QI

Q

K



 

P þ QQI QK 

 (19)

Clearance of the world output market requires that world supply equal world demand

 FðK; LÞ þ F  ðK  ; L Þ ¼ C þ C  þ Ið1 þ Gð$ÞÞ þ I  1 þ G ð$Þ þ G þ G :

(20)

Finally, world capital market equilibrium determines the rate at which home households accumulate foreign equity. This is the market where flows of equity of uniform yield are traded using current output as the means of payment. Defining the stocks of non-human or financial wealth in the home and foreign country as A h QK þ Q)B and A) h Q)K)  Q)B, where B represents the net holdings of foreign equity by domestic households, and using (3), (18) and G ¼ T, capital market equilibrium can be represented by the following equation:

B_ ¼



1 Q

     B FðK; LÞ  Ið1 þ GðI=KÞÞ þ P þ Q  I   C  G K

which yields the accumulation of foreign assets by domestic households.

(21)

348

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

2.5. Characterization of the equilibrium To characterize the equilibrium behavior in this economy we begin by describing its steady state. The long-run relations that must be satisfied come from the market clearance and optimality conditions.14 From (11) to (15), their foreign counterparts, and specializing b(Z) ¼ xZ as well as b(Z)) ¼ x)Z), it is straightforward to see that15 







 Q ¼ 1; Q ¼ 1; R ¼ bðZÞhxZ; R ¼ bðZ Þhx Z ; F ¼ R

J¼

x Rðs þ RÞ



;

J ¼ 

1



; F ¼ R

 1

;

x

  ; R ðs þ R Þ 

(I)



R ¼ FK ðK; LÞ; R ¼ FK  ðK ; L Þ Similarly from (4), (9), (10) together with (16), their foreign counterparts, and equilibrium conditions (18), (20) and (21), the following steady-state relations are derived: 



M ¼ sJUC ðC; G; LÞ; 

UL ðC; G; LÞ UC ðC; G; LÞ





U  ðC ; G ; L Þ ¼ Z ;

UðC; G; LÞ ¼ Z;















U  ðC ; G ; L Þ     L    ¼ W ¼ FL ðK ; L Þ; UC  ðC ; G ; L Þ

¼ W ¼ FL ðK; LÞ ; 



M ¼ s J UC  ðC ; G ; L Þ;







(II)



FðK; LÞ þ F  ðK ; L Þ ¼ C þ G þ C þ G ;

RhR ;

RB þ FðK; LÞ ¼ C þ G 



Note that Q ¼ 1 holds by definition of Q, and that from RhR , it follows that F ¼ F . The steadystate values of J()), F()), Z()), M()), C()), L()), K()), W()), R()), B are obtained from the nineteen equations shown in blocks (I) and (II). Government spending in both countries, G and G), are given parameters. With the steady-state values of all the variables at hand, it is convenient to follow Campbell (1994) and loglinearize the optimality conditions and dynamic equations around the steady state. We start by loglinearizing the expressions for consumption and labor (in what follows lowercase letters denote log deviations from the steady state, with, for instance c ¼ ln C  ln C, denoting the log deviation of consumption):

c ¼ hc;j j  hc;m m þ hc;k k

(22)

 l ¼ y εw;k k  c

(23)

where



sc yl  εw;l

hc;j h

ε1 l

 εw;l

yl hL=ð1  LÞ ;



;

yh

hc;m ¼ hc;j ; 1

yl  εw;l

;

sc ðw  1Þð1  gÞyl εw;k   hc;k h ε1  εw;l l

sc ¼

1

zgðw  1Þ þ 1

h

εw;k ¼

14

dL ð1 þ hÞ h; dL þ ð1  dÞK h

εw;l ¼ εw;k ;

εl h

1

yl ðsc ðw  1Þð1  gÞ þ 1Þ

In what follows steady-state values of the variables are represented with an upperbar. Note that with endogenous rates of time preference we do not have to impose, as in the standard model, the strong and arbitrary condition that both countries have the same exogenous rate of time preference. One consequence of this is that the real interest rate will freely adjust in the long-run. Another consequence is that policies will not be hysteretic; on this see Karayalcin (1999). 15

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

349

Coefficients hc,k, h, and hc,m denote the elasticities of consumption with respect to the capital stock and the shadow prices of the stock of habits (j) and wealth (m). The intertemporal elasticity of substitution of consumption is denoted by sc, while εl stands for the wage elasticity of labor supply. yl represents the steady-state ratio of labor to leisure, whereas εw,l and εw,k stand for the elasticity of the wage rate with respect to the labor and capital inputs. Finally, y measures the effect of changes in consumption on labor supply taking into account the negative effect of an increase in employment on the real wage. Notice that in this setting, the fact that government spending is interacting directly with private consumption in the utility function, magnifies the intertemporal elasticity of substitution of consumption. The expression for hc,k shows that a rise in the capital stock will increase consumption if w > 1, given the fact εw,k > 0 and εw,l < 0. Intuitively, an increase in the capital stock raises the real wage rate leading to an increase in labor supply16; given uCL > 0 (as would be the case when w > 1) this pushes consumption up. If preferences are separable in consumption and leisure, as in the logarithmic utility case with w ¼ 1, accumulation of capital would, ceteris paribus, have no effect on consumption.17 An increase in m or a decrease in j raises the marginal utility of consumption (equation (9)), changing the consumption-leisure trade-off against consumption and in favor of labor services supplied.18 The effect of changes in the capital stock and the relevant shadow prices on labor supply can also be seen from the following expression:

εw;k sc sc j þ  m þ  k ¼ hl;j j þ hl;m m þ hl;k k l ¼  1 1 1 εl  εw;l εl  εw;l εl  εw;l

(23.1)

As the economy accumulates capital, the real wage rises by a factor of εw,k, inducing an increase in labor supply. Yet, this accumulation of capital, by stimulating consumption (when w > 1) also lowers labor supply to a smaller extent. Loglinearizing (18) and (20), the following expressions for r ¼ ln R  ln R and q ¼ ln Q  ln Q are obtained:

r ¼ hr;k k þ hr;l l þ hr;q q

(24)

q ¼ hq;k k þ hq;l l þ hq;k k þ hq;l l þ hq;c c þ hq;c c þ hq;q q

(25)

With these expressions for r and q, and using the conditions for optimal behavior of households and firms in both countries, as well as the equilibrium relations from stock, output and capital markets, we obtain a system of twelve differential equations characterizing the evolution of the integrated world economy over time

x_ T ¼ UxT

(26)

where x ¼ (m, m*, j, j*, f, f*,q, k, k*, z, z*, b), x_ is a vector containing the derivatives of the variables in x with respect to time, and U is a 12  12 Jacobian matrix of partial derivatives with its ij elements given by h_i; j ¼ vx_ i =vxj ; ci˛x_ and cj˛x.19 Since the system has five predetermined variables (k, k*, z, z*, b) and seven jumping variables (m, m*, j, j*, f, f*,q), for it to be locally saddlepath stable it must be the case that five of the eigenvalues must be either real and negative or have negative real parts. Our simulations show that this is the case here. Also, under the plausible benchmark parameters we use, the variables in the system exhibit cyclical transient paths.20

16

Here we assume that the real wage is not sufficiently high to induce a backward bending section in the labor supply curve. If w < 1, increases in income will tend to reduce consumption. This contradicts empirical evidence. Again, the results for the home country apply mutatis mutandis to the foreign economy. 19 See the appendix for the coefficients that appear in (24), (25) and in U of the system given by (26). 20 On this, Epstein and Shi (1993) show that cycles are more likely to happen if habits adjust at a pace slower than a critical level, with this critical level increasing with the value of the parameter that governs the response of the endogenous discount rate to changes in the stock of habits (x here). 17

18

350

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

3. Transmission of fiscal shocks in the world economy Here we study the international transmission of a permanent domestic fiscal expansion financed by nondistortionary taxes dT ¼ dG > 0. To calibrate the model we choose the parameters so that we have initial values of L z 1/3,21 R z 0.01 (which is the standard per quarter value in real business cycles models), C/Y z 0.7, G/ Y z 0.3, and (R*K)/Y z 1/3 in both economies. Also we allow for the home economy be the net creditor country in the setting.22 Parameters in the production function are almost symmetrical. Regarding the utility aggregator, our choice of parameter values reflect the following considerations: (i) we assume w > 1 for reasons discussed above; (ii) as suggested in Guvenen (2006) the elasticity of intertemporal substitution of consumption, sc, should be in the range (0,1). Given the parameters that determine sc (see p. 20), we choose w, z, and g so that sc ¼ 0.64 and sc ¼ 0:88. We also set symmetrical values of z and g in both economies, so that we confer a fair amount of relative importance to government spending and consumption relative to leisure in the utility function. As the title of this paper suggests, differences in the speed of adjustment of the stock of habits and adjustment costs in investment are relevant for the international transmission of fiscal policy, leading to initial responses in private consumption and cross-country correlations that are qualitatively consistent with the empirical evidence. The initial values of the parameters and variables are displayed at the bottom of Table 2, and to the best of our knowledge are chosen in conformance to the best practice in the literature where available. We show the short-run (impact) and long-run (across steady states) elasticities in both countries with respect to a 1% balanced-budget expansion in the home economy in Table 1 and then illustrate graphically the corresponding adjustment paths in Fig. 1. We explain the underlying economic forces that generate these paths, highlighting both the initial reaction of each relevant variable to the shock and the subsequent convergence to the new steady state. Table 3 presents the crosscountry correlations that summarize the predicted international comovements that arise from our model. 3.1. The long-run 3.1.1. Home economy In the long-run a 1% balanced fiscal expansion leads to the typical crowding-out effect in private spending. Both consumption and the capital stock fall with respect to the original steady state. The percentage reductions are 0.36% and 0.12%, respectively. The increase in government spending also leads to higher employment levels in the long-run, 0.39% in the simulation. Intuitively, the higher level of taxes needed to finance the increase in government spending, G, lowers household wealth, triggering a fall in both consumption and leisure. The rise in long-run employment (work effort) increases the marginal productivity of capital (the real interest rate goes up by 0.35%) and reduces the marginal productivity of labor (the wage rate goes down by 0.17%). The fall in the capital stock and the rise in employment have opposing effects on output in the long-run. Under the parameters chosen here, output rises in the long-run by 0.22%, indicating a long-run multiplier of 0.81 for the balanced-budget fiscal expansion. The increase in home government spending lowers home holdings of net foreign assets in the long-run. 3.1.2. Foreign economy The crowding-out effect of the domestic fiscal expansion is “exported” to its trading partner: foreign consumption (0.11%), capital (0.63%), output (0.29%), labor (0.12%) and its marginal productivity (0.17%) all decline in the long-run. This result is a consequence of the increase in the long-run real interest rate (0.35%) induced by the demand shock in the home country.

21 This is consistent with the finding by Ghez and Becker (1975) that households allocate approximately one-third of their productive time to market activities. 22 There are no significant qualitative or quantitative differences if we assume the opposite.

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

351

Table 1 International transmission of the fiscal policy shock. Short-run (impact) and Long-run (across steady states) elasticities. Shock: 1% balanced-budget expansion in the home economy Variable

Impact (home)

Long-run (home)

Impact (foreign)

Long-run (foreign)

Consumption (C) Labor (L) Capital stock (K) Investment (I)a Output (Y) FK(K,L) FL(K,L)

10.41 11.31 0.00 0.43 7.76 7.79 4.03

0.36 0.39 0.12 0.00 0.22 0.35 0.17

1.73 2.07 0.00 0.07 1.39 1.40 0.69

0.11 0.12 0.63 0.00 0.29 0.35 0.17

Notes: Benchmark initial parameters used in the calibration: -Production function: L z 0.151; d ¼ 0.33; h ¼ 0.005 L) z Z0.154; d) ¼ 0.33; h) ¼ 0.005 -Utility function: g ¼ 0.48; q ¼ 3.0; X z 13.54; G ¼ 0.027; z ¼ 0.58 g) ¼ 0.48; q) ¼ 1.5; X) z 2.59; G) ¼ 0.027; z) ¼ 0.58 -Other key parameters: s ¼ 0.5; x ¼ 15; c ¼ 0.9 s* ¼ 0.1; x* ¼ 11; c* ¼ 0.1 This set of parameters is consistent with the following initial values: C/Y z 0.73; G/Y z 0.27; RK/Y z 0.33; L ¼ 0.33 C)/Y) z 0.73; G)/Y) z 0.27; R)K)/Y) z 0.33; L) ¼ 0.33 R ¼ R) z 0.01. a The elasticities of investment on impact correspond to the period immediately after the shock.

3.2. Transitional dynamics To analyze how the home economy adjusts in response to the government shock and how this fiscal expansion is internationally transmitted observe Fig. 1. The figure shows the adjustment paths of output, consumption, the capital stock and labor and their marginal productivities, the shadow price of capital (Tobin’s Q) and investment in both countries, as well as the evolution of the world’s return on equity (equal in both economies), and the current account as a percentage of output in the home economy (the net creditor) in our setting. 3.2.1. Home economy Initially, given the predetermined capital stock, increased provision of government services financed by higher taxes has two opposing effects on consumption and leisure. On impact, the marginal utility of wealth M drops, leading, ceteris paribus, to a decrease in the marginal utility of consumption.

Table 2 Cross-country correlations: stylized facts. From Ambler et al. (2004)

From Backus et al. (1995)

Variable

Sample averagea 1960:1–2000:4 all countries

Backus et al. sampleb 1970:1–1990:2 Europe vs. U.S.

Observed 1970:1–1990:2 Europe vs. U.S.

Baseline model prediction

Output Consumption Investment Employment

0.22 0.14 0.18 0.20

0.59 0.51 0.53 0.61

0.66 0.51 0.53 0.33

0.21 0.88 0.94 0.78

(0.03) (0.02) (0.04) (0.03)

(0.02) (0.02) (0.02) (0.01)

First line: cross-country correlation. Second line: standard deviation. a This column correspond to the averages from 190 pairwise cross-country correlations in Ambler et al. (2004). b This column correspond to the correlation of each variable with respect to the same U.S. variable using the countries and the period considered by Backus et al. (1995).

352

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

1.0%

12.0%

0.0%

10.0%

-1.0%

8.0%

-2.0% -3.0%

6.0%

-4.0%

4.0%

-5.0%

2.0%

-6.0% -7.0%

0.0%

-8.0%

-2.0%

5

10 15 20 25 30 35 40 45 50 Time

10 15 20 25 30 35 40 45 50 Time

5

10 15 20 25 30 35 40 45 50 Time

5

10 15 20 25 30 35 40 45 50 Time

5

10 15 20 25 30 35 40 45 50 Time

2.0%

0.0% -0.1%

0.0%

-0.2%

-2.0%

-0.3%

-4.0%

-0.4%

-6.0%

-0.5%

-8.0%

-0.6%

-10.0%

-0.7%

5

5

10 15 20 25 30 35 40 45 50 Time

-12.0%

2.0%

5.0%

0.0%

4.0% 3.0%

-2.0%

2.0% -4.0%

1.0%

-6.0%

0.0%

-8.0%

-1.0%

5

10 15 20 25 30 35 40 45 50 Time

0.1%

0.1%

0.0%

0.0%

-0.1%

-0.1%

-0.2%

-0.2%

-0.3%

-0.3%

-0.4%

-0.4%

-0.5%

5

10 15 20 25 30 35 40 45 50 Time

-0.5%

R = R*

CA/Y=[B-B(-1)]/Y

2.0%

3.0%

0.0%

2.0%

-2.0%

1.0%

-4.0%

0.0%

-6.0%

-1.0%

-8.0%

5

10 15 20 25 30 35 40 45 50 Time

-2.0%

5

10 15 20 25 30 35 40 45 50 Time

Fig. 1. Transient paths after a 1% balanced-budget fiscal expansion in the home economy.

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359 Table 3 Business cycle correlations.

comovement

353

cross-country

Shock: 1% balanced-budget expansion in the home economy Corr(C, C*) Corr(Y, Y*) Corr(L, L*) Corr(I, I*) Corr(Q, Q*)

0.76 0.78 0.77 0.85 0.94

Note: The values in the table are the Pearson’s correlation coefficients calculated from the values of the variables along the entire transient paths.

In addition, the shadow price J of the level of habits Z falls on impact, reflecting the direct positive effect of the increase in government spending on felicity, and leading to an increase in the marginal utility of consumption. The first effect dominates in our simulation and consumption increases (by 10.41%) on impact. However, with higher leisure and, thus, lower labor supply (employment), on impact output falls as well (by 7.76%). This result suggests that the tax multiplier exceeds the government spending multiplier in the very short-run. As a result of the decrease in labor supply (by 11.31%), the wage rate raises (by 4.03%) and the marginal productivity of capital falls (by 7.79%) leading to an immediate decrease in profits in home country firms. One consequence of this is a decline in domestic equity prices on impact (by 0.30%), triggering an immediate fall in investment (by 0.43%). Despite the rise in consumption and government spending, the fall in national income, coupled with the reduction in investment results in current account surpluses for the initial periods after the shock. In the medium-run, the home economy experiences current account deficits that lead to a lower level of its holdings of foreign assets in the final equilibrium. As Fig. 1 shows, after its initial decline, employment immediately starts to recover and so does output. Several periods after the shock, both employment and output overshoot their higher post-shock long-run equilibrium levels.23 On the other hand, following the decline in the capital stock, the marginal productivity of capital starts rising, leading to an increase in equity prices. As the stock market recovers and domestic equity prices rise, Q goes above the critical level of 1 when firms again find it profitable to undertake investments. Home wages initially remain above their pre-shock levels, but then follow a downward adjustment path, reflecting the fall in the home capital stock and, thus, the decline in the productivity of labor. The rate of return on home and foreign equities displays the same overshooting pattern common to other variables. Note that the non-monotonic transient paths in the figures are consistent with the cyclical dynamics generated by the complex characteristic roots of the system. 3.2.2. Foreign economy The reduction in the world’s return on equity generated by the fiscal expansion in the home country initially translates into lower interest payments on its existing debt in the foreign economy (net debtor in the setting). Consumption increases (by 1.73%) and leisure raises as well. The increase in leisure is reflected in a reduction in labor supply and employment on impact (by 2.07%). The lower level of employment induces both a decline in the marginal productivity of capital (by 1.40%) and an increase in the real wage (by 0.69%). As foreign labor supply and employment falls, output decreases (by 1.39%) immediately after the shock. Profits follow suit and equity prices drop on impact (by 0.45%). A disinvestment process starts in the foreign economy. As is the case of the home country, both employment and output start their recovery after the initial decline. However, unlike in the home economy, output ends up 0.29% lower that its initial pre-shock

23 About a year and a half after the initial fiscal expansion, output peaks at 0.39% above its initial steady-state level, implying a maximum multiplier along the transition of about 1.5. Fatas and Mihov (2001) obtain a maximum multiplier close to 1, two years after the spending increase in their simulation.

354

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

1.06 1.04 1.02 1.00 0.98 0.96 0.94

5

10

15

20

25

30

35

40

45

50 Time

Fig. 2. Transient path of the ratio of home-to-foreign consumption levels after the 1% balanced-budget fiscal expansion in the home economy.

level. As in its trading partner, capital decumulation in the foreign economy raises the marginal productivity of capital, starting a recovery process in the stock market. Yet, this process never pushes equity prices above the replacement cost of capital, with the consequence that the foreign capital stock converges monotonically to its lower post-shock level. Foreign wages follow suit and fall monotonically along the adjustment path. 3.3. Cross-country correlations The main point to highlight here is that, on impact the ratio of home-to-foreign consumption levels, C/C), jumps up, and then falls, starting a cyclical adjustment path which ends at a lower steady-state level (Fig. 2). This behavior contrasts with the standard result in most of the two-country models in the literature, where the ratio remains constant along the transition path, giving rise to a correlation coefficient of unity. We should also note that the cross-country consumption correlation we obtain, though higher than the one suggested by stylized facts, is lower than the value of 0.88 predicted by the baseline Backus et al. model. Moreover, the predicted cross-country consumption correlation in our model is lower than the positive cross-country output correlation in line with the observed ranking in the data. Given the evolution of labor discussed above, we also obtain positive cross-country employment correlations, albeit higher than the ones suggested by stylized facts. Furthermore, the model here consistently yields relatively high positive correlations in investment for reasons discussed above (see Table 3). Finally, given the link between investment and the stock market prices, the model predicts positive cross-country correlations in equity prices as well. 4. Conclusion The paper studies the effects of fiscal policy in an integrated world economy and the associated international cross-country correlations. The setup adopts habit-forming endogenous rates of time preference and adjustment costs in investment. Most of the predictions of the model are in line with the recent empirical literature on fiscal policy. For instance, in response to a balanced fiscal expansion we obtain a positive long-run output multiplier, long-run increases in employment, short-run increases in wages and decreases in investment. Our results also suggest that the short-run

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

355

government spending multiplier is smaller than the tax multiplier, given the initial contractionary effect on output originated by the balanced-budget fiscal expansion. This and the fact that we obtain expansionary effects on output in the medium and long-run are in line with the absence of consistent empirical evidence that one multiplier or the other dominates. Most importantly, we show that the model can generate positive short- and medium-run consumption responses to a positive fiscal shock. This is important as negative consumption responses are considered to be one of the main challenges facing neo-classical models of fiscal policy. We also find that the setup generates an increase in longrun real interest rates, a result most models relying on fixed rates of time preference cannot produce, as their long-run interest rates are tied to the constant time preference rates. In addition, the model predicts positive cross-country correlations in output, investment (and the stock market) and employment. Finally, we show that the model generates persistent responses to shocks.

Appendix. Coefficients



hr;k ¼

hq;k ¼



LFK L



R

RK

 < 0;



ðK c þ K c Þ

hr;l ¼

> 0;

hq;l ¼

2



L FL L

WL

> 0;

 ðK c þ K c Þ

C

hq;l ¼

W L

> 0;

ðK c þ K c Þ

 < 0;



ðK c þ K c Þ

 



ðK c þ K c Þ

hr;q ¼ 1 < 0;

 

R K



hq;c ¼

> 0;

RK

 

hq;k ¼



hq;c ¼

C

> 0;



 

ðK c þ K c Þ

 < 0;

hq;q ¼



K c 



ðK c þ K c Þ

<0

The matrix U in (26) is the following:

2

hm_ ;m hm_ ;m hm_ ;j hm_ ;j hm_ ;q hm_ ;k hm_ ;k hm_ ;z 0 0 6h  hm_  ;q hm_  ;k hm_  ;k 0 0 0 6 m_ ;m hm_  ;m hm_  ;j hm_  ;j 6 0 hj_ ;j hj_ ;f hj_ ;z 0 0 0 0 0 0 6 6     6 0 h h 0 0 0 0 0 0 0 _ _ j ;j j ;f 6 6 0 h h 0 0 0 0 0 0 0 _ 6 f;f f_ ;z 6 hf_  ;f 0 0 0 0 0 0 0 0 6 0 U¼6 6 h_ hq_ ;j hq_ ;j hq_ ;q hq_ ;k hq_ ;k 0 0 0 6 q;m hq_ ;m 6h hk;_ q hk;k hk;k 0 0 0 _ _  6 k;_ m hk;_ m hk;_ j hk;_ j 6    6 hk_  ;m hk_  ;m hk_  ;j hk_  ;j h h h 0 0 0 _ _ _  k ;q k ;k k ;k 6 6 h_ h h h 0 0 0 0 0 0 _ _ m j z ; z ; z ;k z_ ;z 6 6 0 hz_  ;m hz_  ;j hz_  ;k 0 0 0 0 0 0 4 hb;_ m hb;_ m hb;_ j hb;_ j hb;_ q hb;k hb;k 0 0 0 _ _  hk;_ m ¼ hk;_ j ; hk;_ j ¼ 

hk;_ m ¼ hk;_ j ;

cðC þ yW LÞhc;j ;  ðK c þ K c Þ



hk;_ j ¼ 

 

cðC þ y W L Þhc ;j ;  ðK c þ K c Þ

0

hm_  ;z 0

hj_  ;z 0

hf_  ;z 0 0 0 0

hz_  ;z 0

0 0 0 0 0 0 0 0 0 0 0

hb;b _

3 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 5

356

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

cc K ;  ðK c þ K c Þ 







hk;_ q ¼ 

 

 









hk;k _  ¼

c R K þ y W L εw ;k  hc ;k  C hc ;k ;  ðK c þ K c Þ

hk_  ;m ¼

c hk;_ m ; c

hk_  ;m ¼



hk_  ;q

hj_ ;j ¼ ðs þ RÞ; 

hm_ ;m ¼ hm_ ;j ; 

hk_  ;j ¼

c hk;k _ ; c

hk_  ;k ¼

hj_ ;f ¼ ðs þ RÞ;

hj_  ;j ¼ ðs þ R Þ;

hm_ ;j ¼ R

c hk;_ m ; c



c c þ hk;_ q ¼ ; c



c R K þ yW L εw;k  hc;k  C hc;k ;  ðK c þ K c Þ

hk;k ¼ _

c hk;_ j ; c

hk_  ;k ¼

hk_  ;j ¼

c hk;_ j ; c

c hk;k _  ; c

hj_ ;z ¼ R; 

hj_  ;f ¼ ðs þ R Þ;



hj_  ;z ¼ R ;

hm_ ;m ¼ hm_ ;j ;

ðC þ yW LÞhc;j  ðK c þ K c Þ

2

þ

yL FL L hc;j RK

 ;

hm_ ;j ¼ R

     ðC þ y W L Þhc ;j ;   ðK c þ K c Þ



hm_ ;q ¼ 

c R K ;  ðK c þ K c Þ 

hm_ ;k ¼ R

 C hc;k  R K  yW L εw;k  hc;k 

ðK c þ K c Þ



LFKL R

2





yL FL L εw;k  hc;k

      C hc ;k  R K  y W L εw ;k  hc ;k ; ¼ R  ðK c þ K c Þ

RK



hm_ ;k

hm_ ;z ¼ R;



since R ¼ R , we have

hm_  ;m ¼ hm_ ;m ; hm_  ;m ¼ hm_ ;m ; hm_  ;j ¼ hm_ ;j ; hm_  ;j ¼ hm_ ;j ; hm_  ;q ¼ hm_ ;q ; hm_  ;k ¼ hm_ ;k ; hm_  ;k ¼ hm_ ;k ; hm_  ;z ¼ hm_ ;z hf_ ;f ¼ R;

hf_ ;z ¼ R;

hf_  ;f ¼ hf_ ;f ;

hf_  ;z ¼ hf_ ;z ;

 ;

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

hq_ ;m ¼ hq_ ;j ;

hq_ ;m ¼ hq_ ;j ; hq_ ;j ¼ yLFK L hc;j ;



hq_ ;j ¼ y L FK  L hc ;j ; 

hq_ ;q ¼ R;



hq_ ;k ¼ FK K K þ yLFK L εw;k  hc;k ; 





hq_ ;k ¼ FK  K  K  y L FK  L εw ;k  hc ;k

hz_ ;m ¼ hz_ ;j ;

hz_ ;k

hz_ ;j ¼

  M ðC þ yWLÞhc;j ; Z J

   M C hc;k  yW L εw;k  hc;k ; ¼ Z J

hz_  ;m ¼ hz_  ;j ;

hz_ ;k ¼ 

M

J

 

hz_  ;j ¼

M

J

 



Z

Z



hb;_ m ¼ hb;_ j ;

K c ðC þ yW LÞhc;j 

BðK c þ K c Þ 

;

 

K cðC þ y W L Þhc ;j 

BðK c þ K c Þ



K cK c

hb;_ q ¼

;  BðK c þ K c Þ

hb;k ¼ _

c K h_ ; c B k;k



K B

hb;k _  ¼  hk_  ;k  hq_ ;k ; hb;b ¼ R _

 

ðC þ y W L Þhc ;j



hb;_ j ¼

hz_ ;z ¼ s;



     C hc ;k  y W L εw ;k  hc ;k

hb;_ m ¼ hb;_ j ;

hb;_ j ¼ 



 hq_ ;j ;

;

 ;

hz_  ;z ¼ s ;

357

358

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

References Abel, A., 1990. Asset prices under habit formation and catching up with the Joneses. American Economic Review 80 (2), 38–42. Afonso, A., Sousa, R., 2009. The Macroeconomic Effects of Fiscal Policy. European Central Bank. Working Paper Series 991. Alesina, A., Ardagna, S., 1998. Tales of fiscal adjustment. Economic Policy 27, 489–545. Alesina, A., Ardagna, S., Perotti, R., Schiantarelli, F., June 2002. Fiscal policy, profits, and investment. American Economic Review 92 (3), 571–589. Ambler, S., Cardia, E., Zimmermann, C., 2004. International business cycles: what are the facts? Journal of Monetary Economics 51 (2), 257–276. Backus, D., Kehoe, P., Kydland, F., 1995. International real business cycles: theory and evidence. In: Cooley, T. (Ed.), Frontiers of Business Cycle Research. Princeton University Press, Princeton, NJ, pp. 331–356. Bailey, M., 1962. National Income and the Price Level: A Study in Macroeconomic Theory. McGraw Hill, New York. Baxter, M., King, R., 1993. Fiscal policy in general equilibrium. American Economic Review 83, 315–334. Blanchard, O., Perotti, R., 2002. An empirical characterization of the dynamic effects of changes in government spending and taxes on output. Quarterly Journal of Economics 177, 1329–1368. Burnside, C., Eichenbaum, M., Fisher, J., March 2004. Fiscal shocks and their consequences. Journal of Economic Theory 115 (1), 89–117. Carroll, C., Weil, D., 1994. Saving and growth: a reinterpretation. Carnegie Rochester Conference Series on Public Policy 40, 133–192. Campbell, J., 1994. Inspecting the mechanism: an analytical approach to the stochastic growth model. Journal of Monetary Economics 33, 463–506. Chari, V., Kehoe, P., McGrattan, E., 2002. Can sticky price models generate volatile and persistent real exchange rates? Review of Economic Studies 69 (3), 533–563. Constantinides, G., 1990. Habit formation: a resolution of the equity premium puzzle. Journal of Political Economy 98 (3), 519–543. Corsetti, G., Pesenti, P., 2001. Welfare and macroeconomic interdependence. Quarterly Journal of Economics 116 (2), 421–445. Dai, Q., Philippon, T., March 14, 2004. Government Deficits and Interest Rates: A No-arbitrage Structural VAR Approach. Mimeo, New York University. Deveraux, N., Shi, S., 1991. Capital accumulation and the current account in a two-country model. Journal of International Economics 30, 1–25. Edelberg, W., Eichenbaum, M., Fisher, J., 1999. Understanding the effects of a shock to government purchases. Review of Economics Dynamics 2, 166–206. Epstein, L., 1987. A simple dynamic general equilibrium model. Journal of Economic Theory 31, 133–152. Epstein, L., Shi, S., 1993. Habits and time preference. International Economic Review 34, 61–84. Favero, C., Giavazzi, F., 2007. Debt and the Effects of Fiscal Policy. Federal Reserve Bank of Boston. Working Paper No. 07-4. Fatas, A., Mihov, I., 2001. The Effects of Fiscal Policy on Consumption and Employment: Theory and Evidence. Mimeo, INSEAD and CEPR. Ferson, W., Constantinides, G., 1991. Habit persistence and durability in aggregate consumption: empirical tests. Journal of Financial Economics 29 (2), 199–240. Fuhrer, J., Klein, M., 2006. Risky habits: on risk sharing, habit formation, and the interpretation of international consumption correlations. Review of International Economics 14 (4), 722–740. Gale, W., Orszag, P., 2004. Budget deficits, national saving, and interest rates. Brookings Papers on Economic Activity 2, 101–210. Giavazzi, F., Pagano, M., 1990. Can severe fiscal contractions be expansionary? Tales of two small European countries. In: Blanchard, O., Fischer, S. (Eds.), NBER Macroeconomics Annual. MIT Press, pp. 75–110. Ghez, G., Becker, G., 1975. The Allocation of Time and Goods over the Life Cycle. National Bureau of Economic Research, New York. Guvenen, F., 2006. Reconciling conflicting evidence on the elasticity of inter temporal substitution: a macroeconomic perspective. Journal of Monetary Economics 53, 1451–1472. Heaton, J., 1995. An empirical investigation of asset pricing with temporally dependent preference specifications. Econometrica 63, 681–717. Heien, D., Durham, C., 1991. A test of the habit formation hypothesis using household data. Review of Economics and Statistics 73, 189–199. Hicks, J., 1965. Capital and Growth. Clarendon Press, Oxford. Jermann, U., 1998. Asset pricing in production economies. Journal of Monetary Economics 41, 257–275. Karayalcin, C., 1999. Temporary and permanent government spending in a small open economy. Journal of Monetary Economics 43, 125–141. Karayalcin, C., 2003. Habit formation and government spending in a small open economy. Macroeconomic Dynamics 7, 407–423. Lucas, R., Stokey, N., 1984. Optimal growth with many consumers. Journal of Economic Theory 32, 139–171. Mansoorian, A., 1993. Habit persistence and the Harberger-Laursen-Metzler effect in an infinite horizon model. Journal of International Economics 34, 153–166. Mansoorian, A., 1996. On the macroeconomic policy implications of habit persistence. Journal of Money, Credit, and Banking 28, 119–129. Masson, P., Bayoumi, T., Samiei, H., September 1995. Saving Behavior in Industrial and Developing Countries. International Monetary Fund, Staff Studies for the World Economic Outlook. Mountford, A., Uhlig, H., 2008. What Are the Effects of Fiscal Policy Shocks? NBER. Working Paper 14551. Obstfeld, M., 1990. Intertemporal dependence, impatience, and dynamics. Journal of Monetary Economics 26, 45–75. Obstfeld, M., 1992. International adjustment with habit-forming consumption: a diagrammatic exposition. Review of International Economics 1, 32–48.

R.A. Gonzalez-Hernandez, C. Karayalcin / Journal of International Money and Finance 32 (2013) 341–359

359

Ogaki, M., Ostry, J., Reinhart, C., 1996. Saving behavior in low and middle-income developing countries. A comparison. IMF Staff Papers 43 (1), 38–71. Perotti, R., 1999. Fiscal policy in good times and bad. Quarterly Journal of Economics 114, 1399–1436. Perotti, R., 2005. Estimating the Effects of the Fiscal Policy in OECD Countries. CEPR. Discussion Paper 4842. Perotti, R., 2008. In: Acemoglu, Daron, Rogoff, Kenneth, Woodford, Michael (Eds.), 2008. Search of the Transmission Mechanism of Fiscal Policy. Chapter in NBER Book NBER Macroeconomics Annual 2007, vol. 22, pp. 169–226. Ramey, V., Shapiro, M., 1998. Costly capital reallocation and the effects of government spending. Carnegie Rochester Conference Series on Public Policy 48, 145–194. Ryder, H., Heal, G., 1973. Optimum growth with intertemporally dependent preferences. Review of Economic Studies 40, 1–33. Sahota, G., 1993. Saving and distribution. In: Gapinski, J.H. (Ed.), The Economics of Saving. Kluwer Academic Publishers, Boston. Shi, S., 1999. Fashion and wealth accumulation. Economic Theory 14, 439–461. Sommer, M., 2007. Habit formation and aggregate consumption dynamics. The B.E. Journal of Macroeconomics 7 (1) (Advances), Article 21. http://www.bepress.com/bejm/vol7/iss1/art21. Uzawa, H., 1968. Time preference, the consumption function, and optimum asset holdings. In: Wolfe, J.N. (Ed.), Value, Capital and Growth: Papers in Honor of Sir John Hicks. Aldine, Chicago.