Fiscal Policy under Alternative Fiscal Discipline Regimes in a Currency Union

Transcription

1 Ekonomický časopis, 66, 208, č. 2, s Fiscal Policy under Alernaive Fiscal Discipline Regimes in a Currency Union Irem ZEYNELOGLU* Absrac The presen paper uses a wo-counry overlapping generaions framework in order o assess he implicaions of he degree of fiscal discipline on he fiscal policy effeciveness in a currency union. The resuls show ha, iniially, a fiscal simulus implemened under he condiion of reurning o a balanced budge leads o a higher increase in per capia oupu and consumpion compared o a fiscal expansion wih permanenly higher public deb. However, in he medium run, he sric fiscal discipline case leads o an oupu recession despie he increase in privae per capia consumpion whereas a loosening of he fiscal discipline helps avoid he recession a he cos of higher public deb. The overlapping-generaions framework shows also he demographic impac on he fiscal policy effeciveness under differen degrees of fiscal discipline. Keywords: fiscal policy, fiscal discipline, currency union JEL Classificaion: E62, F4, H62 Inroducion Following he economic sagnaion in Europe due o he global financial crisis, he European Commission offered in November 2008 a fiscal simulus sraegy for he member counries. In order o faciliae he implemenaion of naional measures, he Commission allowed counries o exceed emporarily he limis for he public deb (60% of GDP) indicaed in he Sabiliy and Growh Pac (SGP). Consequenly, public deb increased rapidly during he wo years of fiscal simulus, which led he union members o carry ou public spending cus in order o reduce public deb o he level se by he SGP. This ype of feedback from public deb o public spending is also observed for several counries in empirical sudies * Irem ZEYNELOGLU, Galaasaray Universiy, Faculy of Economic and Adminisraive Sciences, Deparmen of Economics, 36 Çırağan Avenue, Isanbul, Turkey; izeyneloglu@gsu.edu.r

2 60 such as Gali and Peroi (2003). However, he sandard heoreical models ha analyse he effec of a deb-financed public spending shock generally overlook he possibiliy of a procedure ha allows o improve public finance following a deerioraion due o he fiscal expansion. In many of he new open economy macroeconomics models inspired by Obsfeld and Rogoff (995), he Ricardian equivalence assumpion implies ha he deb burden is me enirely by an increase in axes. Similarly, dynamic sochasic general equilibrium models consider public spending pah generally as an AR() process in which increases in public spending and axes fade away gradually. However, as shown by Favero and Giavazzi (2007) using US daa, a simple AR() process does no allow o esimae correcly he dynamic effecs of a fiscal policy because i neglecs he possible feedbacks from public deb owards public spending and axes. In order o reconcile heoreical analyses and he empirical observaions, Corsei, Meier and Müller (200; 202) inroduce a fiscal rule wih feedbacks from public deb o boh public spending and axes ino a wo-counry general equilibrium model wih flexible exchange raes. This leads o a wo-phase ransiion dynamics for public spending. In he firs phase public spending increases and is mosly deb-financed. In he second phase, public spending decreases in order o miigae he negaive impac on public deb observed in he firs phase. Corsei, Meier and Müller (200) find evidence of spending reversals for fiscal ransmission using US daa in a VAR model. The presen seup is similar o Corsei, Meier and Müller (200; 202) regarding he law of moion for public spending bu differs from heir analysis by he choice of is framework and aim. Firs, in conras o Corsei, Meier and Müller (200; 202), he presen framework uses a wo-counry overlapping generaions (OLG) framework inspired by Weil (989). Second, in conras o he flexible exchange rae assumpion in Corsei, Meier and Müller (200; 202), he presen OLG seup is applied o a currency union. The OLG seup is paricularly appropriae o analyse he relaion beween public deb and public spending since i allows o eliminae public deb neuraliy. However, as indicaed by Weil (989; 2008), he finie lives assumpion in wo-period OLG models, such as Blanchard (985), is no necessary o eliminae public deb neuraliy. In conras, he assumpion of new generaions arriving a each period is crucial. The presen seup keeps he assumpion of infiniely-lived households, which allows o analyse he impac of populaion growh on fiscal policy effeciveness. Moreover, infinie-life assumpion enables o compare he resuls of he presen seup o hose of he dynamic sochasic general equilibrium (DSGE) framework which generally assumes infiniely-lived households. The discree-ime srucure of he presen seup also helps compare he implicaions of he presen seup wih he DSGE framework.

3 6 This seup is used o analyse he shor and medium run effecs of a home fiscal expansion incorporaing a feedback mechanism ha would miigae he negaive effecs of fiscal expansion on public finance in he medium run. This can be inerpreed as a fiscal expansion under srong budge discipline, as inended by he new SGP. The dynamic effecs of his ype of policy is compared o hose of a fiscal expansion ha is concerned only wih he sabilizaion of public deb wihou being concerned abou he level a which he public deb is sabilized. This second ype of policy can be inerpreed as a fiscal expansion wih a weak fiscal discipline, similar o he one ha is being implemened in he USA. Indeed, in conras o Europe, he USA chose o mainain fiscal expansion in 20 despie he worsening of public defici, hoping ha he oupu growh resuling from acive fiscal policy will help sabilize public defici and deb a some level in he medium run and reduce in he long-run. The resuls show ha a domesic fiscal expansion under srong discipline increases he shor run oupu and consumpion in boh counries and allow o resore he budge balance in he medium run and hen o achieve a budge surplus leading o a reducion in public deb. This improvemen in public finance is accompanied by a reducion in foreign deb. However, his ype of policy leads o an oupu recession and o deflaion in he medium run. In conras, a fiscal expansion under weak fiscal discipline avoids he recession and deflaion bu leads o a higher level of public and foreign deb. The resuls also show ha increasing fiscal discipline leads o higher volailiy in shor run real ineres rae, inflaion and axes. Finally, he OLG seup shows o wha exen he impac of fiscal expansion on per capia variables under various degrees of fiscal discipline is influenced by he populaion growh rae which specifies he degree of deviaion from he Ricardian equivalence. The paper is organized as follows: secion describes he heoreical seup whereas secion 2 calibraes he model and gives he dynamics of he fiscal policy impac on main variables under alernaive fiscal discipline regimes. Secion 3 considers alernaive calibraions for various parameers. Las secion concludes.. The Seup There are wo counries forming a currency union, called home and foreign. A any period, populaion in each counry consiss of a combinaion of differen generaions of households: households born in previous periods and hose who ener he economy in period (new-borns). The rae of arrival of new generaions is denoed by n which is also he populaion growh rae. Since all households wihin each generaion live infiniely, he period populaion of he home counry

4 62 is defined as N = N0 ( + n). Foreign counry populaion N grows a he same consan rae. Iniial union populaion is normalized o. Following Weil (989), inergeneraional bequess are excluded. Iniially a households reside in home counry while he foreign populaion is equal o ( a). Home and foreign goods varieies are also indexed respecively on he arrays [0, a) and (a, ]. Populaion increases in ime while he number of goods varieies remains unchanged... Households There are wo ypes of households in each counry: he firs ype has access o financial markes and is called asse holders. The second ype of households do no have access o financial markes, consume all heir disposable income and herefore are called hand-o-mouh households. This ype of consumer behaviour is empirically observed in various indusrialized counries as poined ou by Campbell and Mankiw (989). The share of he asse holders (non-asse holders) is equal o λ ( λ) in each counry. Moreover, following Weil (989), i is assumed ha he newborns do no own financial asses. However, hey own he presen discouned value of heir human wealh defined as he labour income and share of profis ne of axes. Household Behaviour In any period, a represenaive home asse holder from generaion υ 0, maximizes he following uiliy funcion wih respec o consumpion [ ] C υ and leisure ( L υ ) where is normalized o : L υ is he labour supply. Time endowmen ( ) s C υ ρ υ s υ U = E β + ηlog( Ls ) 0 < β <, η > 0, ρ > () s= ρ where β and ρ are respecively he subecive discoun facor and he coefficien of risk aversion. The real consumpion index C υ is a CES ype funcion defined as: θ θ θ θ θ θ θ θ υ υ υ C = a ( CH ) + ( a) ( CF ) (2) where θ > gives he elasiciy of subsiuion beween home (H) and foreign (F) goods.

5 63 In (2), he consumpion sub-indexes over he goods produced a home (h) and abroad (f) and consumed by home household from generaion υ are defined as: σ a σ σ σ υ σ and υ CH = ( ) c ( h) dh a 0 σ σ σ σ υ σ (3) C υ F = ( ) c ( f ) df a a where σ > is he elasiciy of subsiuion beween goods produced wihin a counry. The home consumpion price indexes corresponding o C υ F are defined as: C υ H and and a σ PH, = p ( h) dh σ a 0 P p f df σ * * σ = ( ) F, a a (4) The overall price index is given as P ap a P θ θ θ = H, + ( ) F, The preferences of foreign households are similar wih aserisks denoing foreign variables. Idenical preferences and he law of one price imply ha overall prices expressed in he common currency are equal across counries. In addiion, he erms of rade is defined as S = PF, / PH,. I is assumed ha home and foreign households do no have biased preferences oward he goods produced in heir counry. υ 0, holds privae In period, home asse-holder from generaion [ ] bonds which are raded inernaionally along wih public bonds which are raded a he naional level. These bonds are denoed by B υ + and correspond o asses purchased a he beginning of period, arriving a mauriy a he beginning of +. The asse holder receives profis income by supplying labour a wage rae marke and pays lump sum axes follows: a i Π di from home firms, earns wage 0 W in he perfecly compeiive labour PT. Hence, he budge consrain is given as a υ υ υ i υ + = ( + ) + + Π 0 PB P r B W L di PC PT (5) where r denoes he real rae of reurn on boh bonds beween periods and. The budge consrain of a foreign asse holder is similar o (5). The individual demand of he represenaive home asse holder from generaion υ for a ypical home and foreign good is given as follows:.

6 64 σ θ p ( h ) υ PH, c ( h) = C PH, P υ σ θ * * p ( f ) P υ F, and c ( f ) = C * * PF, P υ (6) Similar equaions can be defined abroad. Maximizing uiliy () under he budge consrain (5) wih respec o C υ,, υ υ + L leads o he following firs order condiions which give respecively he B ineremporal consumpion pah, and he opimal labour supply of he asse holders: ( C ) E ( r )( C ) υ ρ υ ρ = β (7) υ L υ ηp ( C ) = (8) W ρ The uiliy maximizaion of a foreign asse holder leads o similar equaions. In conras o an asse holder, a home hand-o-mouh household born in period υ consumes his/her enire disposable income a each period. Thus, he consumpion of a non-asse holder is given as follows: W C = L T (9) υ υ υ P Equaions (2) (4), (6), (8), (9) and heir foreign analogues hold also for hand-o-mouh households. Aggregae per capia Values As i is radiional in OLG models, one needs o express all variables in per capia erms (indicaed by he subscrip PC). This consiss in aggregaing he relevan variables across all generaions and hen dividing by he number of households. In wha follows, a variable wihou any reference o a specific generaion will denoe he aggregae level, while he subscrips A and HM refer respecively o asse holders and hand-o-mouh households. Given ha here are a( + n) asse holders in he home counry in period, he opimal per capia labour supply derived from (8) and is counerpar for non- -asse holders are given as follows: A,PC ρ A,PC L ηp ( C ) = and W HM,PC ρ HM,PC L ηp ( C ) = (0) W

7 65 The per capia consumpions are differen for he wo ypes of households. The following per capia consumpion of a home hand-o-mouh household is derived from (9): W C = L T () HM,PC HM,PC HM,PC P For he asse holders, he per capia consumpion is derived from (7). Aggregaing equaion (7) across all generaions of asse holders alive a period gives. A, υ = 0 ρ A, υ = ρ A, υ = ρ a( C ) n( C )... n( n) ( C ) = ae C + n C n + n C β E ( + r ) + A, υ = 0 ρ A, υ = ρ A, υ = ρ ( + ) ( + ) ( ) ( + ) The lef-hand side of he expression above gives he oal consumpion of all A asse holders alive a dae (( C ) ρ ) while he righ-hand side is differen from he consumpion of all households alive a + ( ( C + ) ) since i does no include A, he period + consumpion of individuals born in + ( an( + n) ( C υ = + ) ρ ). However, his equaion can be rewrien as follows so as o include his consumpion: ( C ) E ( C ) an( n) ( C ) A ρ A ρ A, υ= + ρ = β E ( + r + ) A ρ + (2) Dividing boh sides by he number of households alive a period yields he per capia consumpion Euler equaion as follows: ( C ) E ( C ) an( n) ( C ) A,PC ρ A,PC ρ A,PC, υ= + ρ = β E ( + r + ) This ineremporal consumpion equaion includes he period + consumpion of he new-borns ( C + A,PC, υ = + ρ ) which is one of he fundamenal differences wih he represenaive household models. The arrival of new generaions ils he ineremporal consumpion pah by increasing curren consumpion. Indeed, wih posiive populaion growh (n > 0), households anicipae ha he fuure ax burden due o a deb-financed increase in curren public spending will be shared among a greaer number of households. This implies ha he discouned per capia value of fuure axes is lower han he curren per capia public spending. (3) If he populaion grows a rae n and iniial populaion is equal o a, he generaion υ = has an members, generaion υ = 2 has a( + n) 2 a( + n) = an( + n) members, generaion 3 has a( + n) 3 a( + n) 2 = an( + n) 2 members and so on.

8 66 The difference is perceived as a ne increase in wealh. Thus, asse holders respond by increasing heir curren consumpion..2. Firms Each monopolisically compeiive home firm i (owned enirely by home i households) produces a quaniy Y of a single good according o he following producion funcion: Y i i = L. The same is rue for he foreign counry. Firms se heir prices according o he price adusmen mechanism defined by Calvo (983) who assumes ha, in each period, only a randomly seleced fracion α of firms ges he chance o rese heir prices by maximizing heir marke value. The new opimal price chosen by a home firm i in period is he same for all adusing firms. The price index for home goods is hen an average of he opimal price p o charged by he adusing firms and he average of he price charged by he remaining firms: σ o σ σ H, ( α )( ) α H, P = p + P (4) A similar expression can be derived for he aggregae price index for foreign goods..3. Fiscal Auhoriy In he presen seup public spending is resriced o governmen consumpion expendiures. Oher ypes of public spending, such as spending on infrasrucure or he governmen wage bill, are excluded for simplificaion purposes. I is assumed ha governmen consumes from each available good in he same way as households so ha here is no home bias for domesically produced goods. This implies ha public consumpion indexes are similar o equaions (2). Governmen expendiures G are financed by lump sum axes T and by public deb D. This gives he following public budge consrain for period : G + ( + r ) D = T + D + (5) In he above equaion, D + denoes public bonds issued in period arriving a mauriy in period +. The evoluion of public spending and axes are defined similarly o Corsei, Meier and Müller (200; 202) in aggregae erms. I is possible o express public spending and axes in per capia erms as follows where a bar over any variable denoes he seady-sae level of ha variable:

9 67 PC PC PC PC + ( φgg ) φgg φgd ( ) + ε+ G = G + G + + n D + (6) φ PC g PC PC G PC = φ PC + d T G D G Equaion (6) gives he law of moion for public spending, where ε is he oneime sochasic i.i.d. shock. In conras o he ypical AR() process, he evoluion of governmen expendiure depends also on he level of public deb. Indeed, assuming ha φ gd < 0, following an iniial increase in public spending, fiscal auhoriy will adus governmen expendiure endogenously hrough fuure public spending cus as he ousanding public deb increases. This implies ha he fiscal auhoriy has o reduce public spending and deb someime afer he fiscal expansion when faced o an upper limi for public deb and defici. If such an upper limi is no required ( φ gd = 0 ), public spending will follow an exogenous (7) pah owards is seady-sae level following he shock. Similarly, equaion (7) ranslaes he effors of reducing public defici hrough higher axes in response o an increase in public spending and deb assuming ha φ d > 0. The parameer φ is he feedback from public spending o axes and defines he responsiveness g of axes o public spending..4. Moneary Auhoriy The common cenral bank is assumed o pursue an ineres rae policy ha CPI responds o he deviaions of union-wide consumer price inflaion π and oupu from heir seady-sae levels. Following Roemberg and Woodford (999), he ineres rule is given as: 2 CPI u u + φi φi φpπ+ φy i = ( ) i + i + + ( Y Y ) (8) where he superscrip u indicaes union-wide levels. The assumpion ha period + nominal ineres rae i + depends on is lagged value allows for ineres rae smoohing so ha susained changes in oupu and inflaion lead only o gradual changes in he nominal ineres rae..5. General Equilibrium General equilibrium is a sequence of prices, wages, consumpion and producion levels as well as policy insrumens and financial asses which mee he following condiions: 2 Seady-sae inflaion is zero by definiion.

10 68 (i) Aggregae oal per capia labour demand mus be equal o he sum of labour supplied by asse holders and hand-o-mouh consumers in boh counries, which implies he following expression a home and a similar equaion for he foreign counry: L di = L = ( λ) L + λl (9) 0 a i,pc PC A,PC HM,PC (ii) Goods marke mus be in equilibrium in each counry where goods supply is given by he home and foreign producion funcions. Toal demand for a single home good is given as follows: σ θ d,pc p ( h) PH, u,pc u,pc y ( h) = ( C + G ) PH, P (20) Equaion (20) is obained by combining he per capia versions of aggregae home and foreign privae demand for good h (derived from equaion (6)) along wih home and foreign public demand which are expressed similarly. (iii) Inernaional bond marke mus be in equilibrium which requires ha global ne privae asses be zero: af PC *PC + a F + + ( ) = 0 (2) where F and F* denoe he privae bonds which are inernaionally raded beween home and foreign counries such ha a home F = B D and abroad * * * F = B D (iv) he following ransversaliy condiion has o be me boh in aggregae and per capia erms: M lim q T, T B T + = 0 T P + T (22) where q, + is defined as he marke discoun facor for dae + T on dae. T To solve he model one mus log-linearize he relevan equaions in per capia erms around he seady-sae. The seady-sae is defined as he special equilibrium wih zero inflaion and perfecly flexible prices where inernaional rade is in balance so ha privae deb is nil. Moreover, public budge is also balanced in he seady-sae so ha axes are equal o public spending. The log-linearizaion allows o express all per capia variables in erms of heir percenage deviaion from he seady-sae level. The log-linear model is solved numerically under he assumpion of raional expecaions.

11 69 2. Calibraion and Resuls This secion compares he impac of wo ypes of fiscal policy. Firs, i considers an increase in home public spending assuming ha fiscal auhoriy is no concerned abou he budge defici as long as public deb is no explosive (weak fiscal discipline). Second, he same policy exercise is reconsidered assuming ha fiscal policy is designed o ensure a balanced budge in he medium run (srong fiscal discipline). Foreign public spending is assumed o remain unchanged hroughou he res. 2.. Parameerisaion Table below summarizes he values and he descripions of various parameers used for calibraion on quarerly basis. T a b l e Calibraed Parameers Parameers Value Parameers Value Subecive discoun facor β 0.99 Weigh of leisure in uiliy η Calib. Subsiuion elasiciy beween home σ 6 Moneary rule coefficien on φ 0. y domesic goods oupu Subsiuion elasiciy beween home θ 6 Moneary rule coefficien on φ.5 p and foreign goods inflaion Measure of price rigidiy α 0.75 Ineres rae smoohing φ i 0.9 Coefficien of risk aversion ρ.35 Public spending persisence φ gg 0.8 Growh rae of counry populaion n Tax response o public deb φ 0.04 d Share of public spending financed by φ 0.3 Public spending response o g φ 0.04 gd axes public deb Share of hand-o-mouh consumer λ 0.3 Home counry size a 0.3 Source: See he ex below. The discoun facor β and he risk aversion coefficien ρ in equaion () are respecively se o 0.99 for each quarer and o.35 according o he esimaions of Smes and Wooers (2003). The value of η in he same equaion is calibraed o have a seady sae labour supply equal o one hird of he available ime. The size of he home counry a and he rae of populaion growh n are respecively se o 0.3 which corresponds roughly o he share of Germany in he European GDP and o 0.25% which is equal o he average populaion growh rae in he Economic and Moneary Union (EMU) in 204. The share of hand-omouh consumers λ is se o 0.30 which is he average range esimaed in he lieraure (Campbell and Mankiw, 989; Gali, Lopez-Salido and Valles, 2007). Inra-emporal subsiuion elasiciy wihin a counry σ is se o 6. Given he high inegraion of he goods markes across EMU members, he goods subsiuion

12 70 elasiciy across counries θ is assumed o be equal o he subsiuion elasiciy wihin he counry. The measure of price rigidiy α in equaion (4) is equal o 0.75 which, on a quarerly basis, implies ha prices remain fixed during one year on average. The ineres rae smoohing parameer φ i in equaion (8) is se o 0.9 which is in line wih he esimaions in he recen lieraure (e.g. Fendel and Frenkel, 2006). The responsiveness of he ineres rae o he inflaion deviaion φp is radiionally se o.5 while he coefficien on oupu φy is assumed o be equal o 0.. Regarding he fiscal policy parameers, he presen paper follows Gali and Peroi (2003) who esimae a fiscal policy rule for several members of he EU. Following hese esimaes, he presen paper assumes φ gd = and ses accordingly φ d = 0.04 o indicae he srong fiscal discipline case. For he weak fiscal discipline case, he las wo parameers are calibraed o have deb sabiliy a an endogenously deermined poin in ime. Following Gali and Peroi (2003), he presen paper ses he persisence coefficien of fiscal policy φ gg o 0.8. The reacion of axes o public spending φg is arbirarily se o 0.3. I is assumed ha he share of public spending in GDP is equal o 20% in he seady-sae Fiscal Policy Effeciveness under Weak and Srong Fiscal Discipline Figure below shows he impac of a public spending increase by one per cen of he seady sae real GDP on he main variables under weak fiscal discipline (solid lines) and under srong fiscal discipline (dashed lines) where he x-axis measures ime in quarers and y-axis gives he percenage deviaion of he relevan variable from is seady sae level. 3 Effecs on Public Finance In he weak fiscal discipline case, home fiscal auhoriy increases public spending wihou worrying abou he public deb persisence as long as i is no explosive. According o (7) and (8), per capia public spending increases less a each period compared o he seady sae level and reurns gradually o he iniial level (firs panel of he firs row), while per capia axes increase less han he public spending (no shown in he figure). Governmen will issue new public bonds in order o finance he resuling budge defici as well as he service of he previous deb. This will lead o an increase in per capia public deb which slows down progressively and leads o a medium run sabilizaion a a level higher han he seady-sae (as shown in he firs panel of he las row). 3 Quaniy variables are expressed in aggregae per capia erms.

13 7 F i g u r e The Effec of Fiscal Policy on Main Variables Noe: Solid line corresponds o he weak discipline case while he dashed line corresponds o he srong discipline case. Horizonal axis measures ime in quarers and y-axis measures he percenage deviaion of he relevan variable from he seady-sae. Source: Auhor s own calculaions. In he srong fiscal discipline case, he fiscal auhoriy seeks o resore he budge balance in he medium run and o reduce progressively he public deb. Therefore, public spending increases by a lower amoun compared o he weak discipline case as a reacion o he higher public deb. I falls below he seady- -sae level in he medium run, before reurning o is iniial long-erm level. As a resul, afer an iniial increase, boh aggregae and per capia public deb sar a decreasing rend in he medium run and reurn o he seady-sae level in he long-run. Effecs on Prices and Real Variables The iniial effecs of a home fiscal expansion on prices and real variables are similar under boh ypes of discipline, bu he ransiion effecs in he medium run depend on he degree of fiscal discipline.

14 72 Iniially, he increase in public spending simulaes he aggregae public demand in he same way in boh counries regardless of he degree of fiscal discipline since he absence of home bias implies equal expendiure-shifing effec of public spending a home and abroad. A he aggregae level, foreign aggregae demand increases more han he home aggregae demand especially due o he fac ha foreign households benefi from he home fiscal simulus wihou having o pay for he cos in he form of higher axes in conras o home households. As a resul, foreign prices increase more han home prices implying an increase in he erms of rade (las panel of he las row in Figure ), which reallocaes demand owards home goods. However his expendiure-swiching effec, which increases he demand for home goods, is dominaed by he expendiure-shifing effec which increases he demand for foreign goods. The resuling home rade defici causes an increase in privae deb vis-à-vis he foreign counry in he shor run under boh fiscal discipline cases (middle panel of he las row in Figure ). The upward pressure on prices in boh counries is limied as prices are sicky in he shor run. This implies ha oupu is parly demand driven. Therefore aggregae oupu increases more in he home counry compared o he foreign oupu. However, since he increase in he aggregae oupu is shared among a greaer number of households in he foreign counry, he foreign per capia oupu increases less han a home as shown in he middle panel of he firs row in Figure. In order o mee he increase in he demand, firms increase heir labour demand. This leads o an increase in real wages (as seen in he las panel of he second row in Figure ) o he exen ha nominal wage increases more han he prices. Following he increase in oupu a he union level, he cenral bank reacs iniially by increasing he nominal ineres rae which decreases he asse holders consumpion. This miigaes he consumpion increase due o he hand-o-mouh households, bu he ne effec is posiive. However, as he reacion of he union oupu fades, he gap beween he curren and saionary ineres rae falls progressively under he weak fiscal discipline. Hence he negaive impac of he ineres rae on asse holders consumpion is miigaed in he medium run. Under boh ypes of fiscal discipline, consumpion of asse holders and hand-o-mouh households evolve in he opposie way. Hence, oal per capia privae consumpion reurns o is iniial value in he medium run (he firs panel of he second row). This implies ha, oal demand is mainly driven by public spending especially in he medium run. Therefore, he evoluion of per capia oupu is similar o ha of per capia public spending. Thus, as can be seen in he second panel of he firs row in Figure, per capia oupu increases less a each period in boh counries. However, under weak fiscal discipline oupu remains above he iniial

15 73 level in he medium run, whereas i falls below he seady-sae level under srong fiscal discipline. The reducion in hours worked, due o he fall in oupu, brings abou a decrease in real wages in boh counries. As boh home and foreign demand fall below he iniial level in he medium erm under srong discipline, goods prices also sar o fall. Because of he ineres rae smoohing parameer, large umps in he ineres rae are ruled ou even in he presence of deflaion observed in he srong discipline case. Therefore, he real ineres rae remains below he iniial level even in he long-run where prices reurn o heir seady-sae level (las panel of he hird row in Figure ). In conras, under weak discipline, he real ineres rae remains above he iniial level since he evoluion of oupu excludes he possibiliy of a deflaion. The evoluion of foreign deb is also relaed o ha of goods demand. As he home demand for foreign goods falls below he iniial level under srong discipline case in he medium run, rade defici and foreign deb sar o fall before reurning o he iniial zero-deb equilibrium in he long-run. In conras, due o higher levels of foreign goods demand, foreign deb sabilizes a a higher level in he medium run under weak fiscal discipline. Some of he resuls menioned above are similar o hose repored by Corsei, Meier and Müller e al. (200) wih whom he presen paper shares he design of fiscal policy. Specifically, public spending, home oupu and inflaion responses are similar under boh fiscal cases. However, home consumpion response is posiive in he presen seup in boh cases whereas Corsei, Meier and Müller (200) find a negaive consumpion response in he absence of fiscal feedback parameers. The auhors explain he evoluion of consumpion by ha of he long-run real ineres rae which hey define as he infinie sum of he expeced shor run real ineres raes. Indeed, he laer falls below is seady-sae level in he presence of spending reversals whereas wihou he feedback mechanism long-run real ineres rae remains above he seady-sae level. This posiive response of he long-run real ineres rae brings abou a fall in consumpion. In conras, he presen seup implies a posiive consumpion response even when he real ineres rae remains high compared o he seady-sae level. This can be explained by he fac ha wih posiive populaion growh, households anicipae a fall in per capia fuure axes when public spending is deb-financed (weak discipline). Expeced axes affec posiively he consumpion of all households whereas he negaive effec of he ineres rae concerns only he consumpion of asse holders. Neverheless, in he presen seup consumpion increases more under srong discipline as in Corsei, Meier and Müller (200). Regarding he spill-over effecs, he resuls of he presen paper differs from hose of Corsei, Meier and Müller (200) in he weak discipline case. Under weak

16 74 discipline, spill-over effecs remain posiive in he presen seup while Corsei, Meier and Müller (200) find negaive spill-over effecs in he absence of he feedback mechanism. This is due o he fac ha Corsei, Meier and Müller (200) assume home good bias boh in privae and public demand whereas he presen seup absracs from home bias. Moreover, Corsei, Meier and Müller (200) consider flexible exchange rae regime while he presen seup is applied o a currency union. The overlapping-generaions framework gives some insigh on he inergeneraional effecs of a fiscal expansion. In he weak fiscal discipline case, governmen pays only he ineress by issuing new bonds in he following period. Therefore, new generaions do no bear he burden of former deb in he form of higher axes. In conras, under high fiscal discipline, governmen resors o ax increases in order o reduce public deb. Therefore newborns bear he ax burden wihou receiving any ineres income since hey do no hold asses by assumpion. Moreover, under srong fiscal discipline he iniial impac of fiscal expansion is higher compared o he weak discipline case bu he medium run effec is a recession since oupu falls below he seady-sae in he medium run. This implies ha he srong discipline regime favours curren generaions a he expense of fuure generaions. 3. Sensiiviy Analysis This secion considers alernaive values for he populaion growh rae, for he parameers ha affec fiscal policy ransmission and for he policy parameers in order o assess heir impac on he implicaions of public spending under srong and weak fiscal discipline regimes. The parameer n will be allowed o vary in his secion for wo reasons. Firs, he populaion growh differs across members of he euro zone, suggesing ha he ransiion dynamics may differ from one counry o anoher. Second, populaion growh may play an imporan role in an overlapping generaions model since i is one of he facors ha deermine he degree of deviaion from he Ricardian equivalence. In a wo counry seup, he size of he policy implemening counry is imporan since i affecs he inernaional policy ransmission and hereby he effeciveness of he fiscal simulus. As he counry size differs significanly from one member o he oher in he euro zone, i will be convenien o consider alernaive values for his parameer. The elasiciy of subsiuion beween home and foreign goods (θ) also affecs he size of he spill-over effecs of a home fiscal expansion. Therefore, alernaive values for θ will be considered.

17 75 Among policy parameers, he moneary rule coefficien on oupu ( φ y ) plays an imporan role on he fiscal policy ransmission boh a home and in he foreign counry. The ax response coefficien φ g is also an imporan policy parameer. While some of he members of he euro zone rely relaively more on axes, ohers prefer spending cus o reduce public defici raher han increasing axes. Therefore, alernaive values of he ax response coefficien φ g will be considered in his secion. 3.. Populaion Growh and Fiscal Policy Effeciveness In he presen seup, he deviaion from he Ricardian equivalence depends on he populaion growh. Therefore populaion growh will influence he impac of public spending on consumpion and oupu in boh counries. Table 2 gives he firs year per capia oupu and consumpion mulipliers in he wo counries for various values of he populaion growh. 4 A row-by-row comparison of Table 2 allows o see how mulipliers vary wih he populaion growh rae whereas a column-by-column inspecion allows o compare he performances of he wo fiscal regimes. T a b l e 2 Oupu and Consumpion per capia Mulipliers and Populaion Growh n PC *PC Y Y PC C Weak fiscal discipline *PC C Srong fiscal discipline PC *PC Y Y PC C *PC C ϕ g = ϕ g = Source: Auhor s own calculaions. As Table 2 shows, under weak fiscal discipline, he effec of populaion growh on consumpion and oupu mulipliers depends on he iniial increase in axes given by φ. This suggess ha populaion growh works hrough wo g differen mechanisms wih opposie effecs. The firs effec is on aggregae values and works hrough he ypical OLG channel, which implies ha he discouned value of per capia fuure axes is lower han he discouned value of public spending financed wih public deb. Indeed, he fiscal auhoriy has o increase fuure axes in order o pay he public deb conraced a any period. However, wih posiive populaion growh, households anicipae ha he increased fuure 4 Specifically, he able gives he cumulaive four-quarer response of he relevan variable wih respec o he cumulaive four-quarer change in public spending.

18 76 burden of axes due o higher public deb will be shared among a greaer number of home households. Hence, hey perceive public spending as an increase in heir ne wealh and hereby raise heir curren consumpion (as implied by equaion (2) a he aggregae level and by equaion (3) in per capia erms). This is he ypical resul in OLG models where Ricardian equivalence does no hold. In he presen seup, he ypical OLG resul holds when fiscal discipline is low and when iniially all public spending is deb-financed ( φ g = 0). The second effec of populaion growh works hrough is mechanical effec on per capia variables since populaion growh leads o a redisribuion of aggregae variables among a larger number of households. For a given level of aggregae consumpion, dividing by a larger populaion will decrease he per capia level. As he presen seup shows, when some of he iniial increase in public spending is ax-financed ( φ g > 0 ) or when fiscal discipline is srong, he ypical OLG resul does no hold any more suggesing ha he aggregae consumpion does no increase enough o compensae for he higher populaion. This implies ha he second effec of populaion growh menioned above is higher han he firs effec. As a resul, per capia consumpion and oupu mulipliers in boh counries fall as populaion growh increases. Given φ g > 0, under boh fiscal regimes, a higher n leads o a higher increase in he consumpion of asse holders boh a he aggregae level as implied by equaion (2) and in per capia erms as implied by equaion (3). The increase in aggregae oupu simulaes he consumpion of hand-o-mouh households, which amplifies he impac of n on he response of aggregae oupu and consumpion following he fiscal expansion. The resuling spill-over effec on he aggregae foreign oupu and consumpion is posiive, bu he effec in per capia erms on all variables is negaive due o he high populaion increase. Table 2 gives also he effec of φ g on he fiscal mulipliers. Inspecion of he firs and hird rows of Table 2 (keeping n consan a 0.25%) shows ha under weak fiscal discipline, a fall in ax response o public spending ( φ g ), leads o a fall in he oupu muliplier whereas he effec on all oher mulipliers is posiive. Indeed, he negaive wealh effec of axes on home consumpion is lower when φ g = 0, which increases he home consumpion muliplier. Despie higher privae demand, oupu increases less when φ g = 0. This is due o he fac ha he erms of rade increase less when public deb is enirely deb-financed. Therefore, world demand swiches o home producers bu less han when iniial public spending is parially ax-financed ( φ g = 0. 3 ). Hence, home oupu muliplier falls while he foreign oupu muliplier increases. In conras, under srong

19 77 discipline, a fall in ax response o public spending ( φ g ) increases he expansionary effec of fiscal policy on all variables. Indeed, he negaive effec of he erms of rade on home oupu is more han compensaed by he increase in home privae demand which is higher under srong discipline. As a resul, home oupu muliplier slighly increases following he fall in φ g under srong discipline in conras o he weak discipline case. This resul suggess ha in imes of fiscal expansion, i is beer o finance exra public spending wih deb (raher han axes) provided ha fiscal discipline is high The Effec of a Change in Policy Parameers and Fiscal Transmission Table 3 gives he firs year per capia oupu and consumpion mulipliers in he wo counries for various values of he elasiciy of subsiuion beween home and foreign goods θ, he size of he policy-implemening counry a and he moneary policy response o oupu gap φ y. 5 A row-by-row comparison of Table 3 allows o see how mulipliers vary wih he relevan parameer whereas a column-by-column inspecion allows o compare he performances of he wo fiscal regimes. The firs row of each panel gives he mulipliers of he benchmark calibraion. T a b l e 3 Sensiiviy Analysis for Various Parameers Weak fiscal discipline PC *PC Y Y PC C *PC C Srong fiscal discipline PC *PC Y Y PC C *PC C (a) θ = θ = θ = (b) a = a = a = (c) φ y = φ y = φ y = Source: Auhor s own calculaions. As θ decreases, home and foreign goods become poor subsiues and demand swiches less easily from one counry o he oher following relaive price movemens. Therefore, he erms of rade increase more following he fiscal expansion wih a low elasiciy of subsiuion. When θ decreases, he advanage of he erms of rade lass longer which implies higher oupu a home under boh fiscal 5 Specifically, he able gives he cumulaive four-quarer response of he relevan variable wih respec o he cumulaive four-quarer change in public spending.

20 78 regimes. Accordingly, foreign oupu falls wih θ. Lower income in he foreign counry brings abou a fall in foreign consumpion. Panel (a) in Table 3 shows ha home and foreign variables reac similarly o a fall in θ regardless of he degree of fiscal discipline. Table 3 shows ha counry size is imporan for he fiscal mulipliers (panel b). As counry size increases, oupu and consumpion mulipliers in boh counries increase regardless of he fiscal regime. Indeed, a bigger counry size leads o a higher expansion of world demand following he fiscal shock which, hen, simulaes foreign variables hrough spillover effecs. This resul suggess ha he effeciveness of fiscal simulus plans is likely o be increased when hose plans are generalized in a currency union. Moreover, he posiive spillover effecs across he counries sugges ha here may be gains from fiscal policy cooperaion among he implemening counries. Panel (c) in Table 3 shows ha a higher reacion coefficien of oupu in he moneary policy causes a fall in consumpion and oupu mulipliers in boh counries regardless of he fiscal regime. When moneary auhoriy reacs more aggressively o oupu deviaions, any expansionary effec of public spending is couner balanced by a higher increase in he nominal ineres rae. This implies ha, some kind of cooperaion beween fiscal and moneary auhoriies, inducing a less acive moneary behaviour, may increase he effeciveness of fiscal simulus plans. Conclusion The presen paper compares he dynamic effecs of wo ypes of fiscal policies based on differen degrees of fiscal discipline. Shor run consumpion and oupu mulipliers of fiscal policy are lower under weak discipline compared o he srong discipline case. However, in he medium run fiscal policy under weak discipline simulaes oupu beer han he srong discipline case since i helps avoid oupu recession and deflaion as long as a minimum of budge discipline is respeced so ha neiher public deb nor foreign privae deb are explosive. The European fiscal simulus plan implemened afer he 2009 crisis in order o simulae economic growh while avoiding he undesired effecs on public finance seems similar o he fiscal simulus plan under srong fiscal discipline presened in he previous secion. Alhough he European simulus plan does no imply he endogenous feedback mechanism, here seems o be an exogenous feedback from public deb o public spending. Indeed, in 20, he European Commission asked he member counries o end he fiscal simulus plans afer wo years of implemenaion and o ake resricive precauions in order o reduce public defici. This fiscal consolidaion policy implemened in 20 allowed he

21 79 members o reduce average euro zone fiscal defici from 4.% in 20 o 2.% in 205 and o slow down he increase in public deb albei a a higher level han he upper limi indicaed by he SGP. A he same ime, economic recession increased unemploymen beyond % in 202 and 203 and he economic growh in 205 remained below he pre-crisis level (2.6% in 2007). Moreover, inflaion is around 0% in 205 which remains well below he ECB s arge of 2%, implying a risk of deflaion. In conras o he European reacion o he crisis, he American fiscal simulus plan implemened in 2009 was subec o a lower fiscal discipline and hence coninued in 20. The US GDP growh raes in 202 and 203 were above 2% and well above hose of he euro zone in 204 and 205 wih an unemploymen rae equal o he half of he European unemploymen. Public deb in he US as a percenage of GDP ends o sabilize wih a lower growh rae beween 202 and 204 compared o he previous period. The OLG srucure of he presen seup allows o see how he conribuion of he fuure generaions o he financing of he fiscal simulus affecs he consumpion of curren generaions. I also shows ha, regarding oupu growh, srong fiscal discipline favours curren generaion agains he fuure generaions since he srong fiscal discipline case causes a sligh oupu recession in he medium run. Finally, his OLG srucure shows he effec of populaion growh on he fiscal mulipliers under various degrees of fiscal discipline. The presen seup neglecs he possibiliy of muliple equilibria and herefore does no accoun for governmen solvency and sovereign deb issues. Indeed, in a muliple equilibrium seup, households may refuse o lend governmen even when he permanen public deb is sable, which may cause governmen insolvency and hereby render undesirable a high level of public deb. Incorporaing a risk premium on he public bond of he counry wih he highes ousanding public deb would work similarly. The risk premium would induce he public auhoriies o refrain from excessive deb even under weak discipline case and hereby reduce he gap beween he resuls of he wo fiscal regimes. Inroducing hese issues ino he presen seup may be ineresing for fuure research. References BLANCHARD, O. J. (985): Deb, Deficis, and Finie Horizons. Journal of Poliical Economy, 93, No. 2, pp CALVO, G. A. (983): Saggered Prices in a Uiliy-maximizing Framework. Journal of Moneary Economics, 2, No. 3, pp CAMPBELL, J. Y. MANKIW, G. (989): Consumpion, Income, and Ineres Raes: Reinerpreing he Time Series Evidence. In: BLANCHARD, O. J. and FISCHER, P. J. (eds): NBER Macroeconomics Annual 989. Cambridge, MA: MIT Press, pp

22 80 CORSETTI, G. MEIER, A. MÜLLER, G. (200): Cross-border Spillovers from Fiscal Simulus. Inernaional Journal of Cenral Banking, 6, No., pp CORSETTI, G. MEIER, A. MÜLLER, G. (202): Fiscal Simulus wih Spending Reversals. The Review of Economics and Saisics, 94, No. 4, pp FENDEL, R. M. FRENKEL, M. R. (2006): Five Years of Single European Moneary Policy in Pracice: Is he ECB Rule-based? Conemporary Economic Policy, 24, No., pp FAVERO, C. GIAVAZZI, F. (2007): Deb and he Effecs of Fiscal Policy. [NBER Working Paper 2833.] Cambridge: NBER. GALI, J. PEROTTI, R. (2003): Fiscal Policy and Moneary Inegraion in Europe. Economic Policy, 8, No. 37, pp GALI, J. LOPEZ-SALIDO, J. D. VALLES, J. (2007): Undersanding he Effecs of Governmen Spending on Consumpion. Journal of he European Economic Associaion, 5, No., pp OBSTFELD, M. ROGOFF, K. (995): Exchange Rae Dynamics Redux. Journal of Poliical Economy, 03, No. 3, pp ROTEMBERG, J. J. WOODFORD, M. (999): Ineres Rae Rules in an Esimaed Sicky Price Model. In: TAYLOR, J. B. (ed.): Moneary Policy Rules. Chicago: The Universiy of Chicago Press. SMETS, F. WOOTERS, R. (2003): An Esimaed Sochasic Dynamic General Equilibrium Model of he Euro Area. Journal of he European Economic Associaion,, No. 5, pp WEIL, P. (989): Overlapping Families of Infiniely-lived Agen. Journal of Public Economics, 38, No. 2, pp WEIL, P. (2008): Overlapping Generaions: The Firs Jubilee. Journal of Economic Perspecives, 22, No. 4, pp