TAX REFORM IN TUNISIA: A SIMULATION BASED ANALYSIS USING THE NEW AREA-WIDE MODEL

Transcription

1 Inernaional Journal of Economics, Commerce and Managemen Unied Kingdom Vol. III, Issue 5, May 2015 hp://ijecm.co.uk/ ISSN TAX REFORM IN TUNISIA: A SIMULATION BASED ANALYSIS USING THE NEW AREA-WIDE MODEL ABDELLI Soulaima Universiy el Manar, Tunis, Tunisia selima11@yahoo.fr BELHADJ Besma Universiy of Carhage, Tunis, Tunisia besma.kaabi@isg.rnu.n Absrac This paper aims o modeling an emergen economy like Tunisia which is rarely used when calibraing he New Area-Wide Model. Indeed, some modificaions and feaures are inroduced for he calibraion of a small open economy of an emergen marke. For insance he presence of he non-ricardian agens as a financial fracion ha is very imporan when gauging he impac of ax reform, he wage heerogeneiy in he labor marke and he risk premia in he budge consrain. To perform a ax reform, his paper ess he long-run effecs of a number of fiscal consolidaion scenarios based on a revenue and expendiure side on he key macroeconomic aggregaes and simulaes he economic aciviy. We discover ha he fiscal consolidaion sraegy based on he revenue side when lowering disorionary axes has a posiive long-run effec on he oupu. Conrariwise, he consolidaion sraegies based on he expendiure side has a negaive long run impac on he oupu, bu replicaes an increase in he household s wages as well as he consumpion and he invesmen. The model is calibraed on he Tunisian economy using an annual daa over Keywords: Tax reform, fiscal policy, DSGE modeling, fiscal consolidaion, Tunisia, heerogeneous agen Licensed under Creaive Common Page 1

2 Soulaima & Besma INTRODUCTION The recen economic and financial crisis in Europe and in Tunisia urged he governmen o pu in place an ambiious fiscal simulus plans as well as many fiscal policy consolidaion sraegies. The fiscal sysem in Tunisia has suffered from many diseases. For example we noe he budge defici which is above 7% of GDP ha surpasses he inernaional norm (3%), he explosion of public deb which amoun o billion TND in 2013 so 47.2% of he GDP, he increase of he Sae s expendiure so ha he budge was billion TND in 2010 and rise o billion TND in 2014.Also we noe he fiscal injusice where he ax burden, suppored especially by employees is far superior o he oher classes of sociey. According o Jlassi (2013) here is no an ineresing measure o figh agains ax evasion and he informal marke. In his conex, Nabli (2013),he pas Governor of Cenral Bank of Tunisia,emphasized The fiscal policy is no aking robus decision concerning he public invesmen In fac he implemenaion of a largescale fiscal spending programs and a ax reform in Tunisia and around he world sparked a vigorous policy debae. Fiscal auhoriies mainly have he possibiliy of increasing or decreasing axes and/or cuing or increasing public spending. Wha are he expeced consequences from hese measures on he oupu, he consumpion and he invesmen in he shor and he long run? In his case,he choice of he New Area Wide Model (NAWM) presened by Coenen e al. (2008) and Chrisoffel e al. (2008) referred o as CMS and CCW respecively, is proposed as a new ool of modeling. The laer is a class of he dynamic sochasic general equilibrium model (DSGE) employed for developing micro-founded DSGE models. I is a wo-counry open-economy model developed a he European Cenral Bank (ECB s) for he fiscal policy simulaion. In fac, i is chosen because i included many characerisics. Firs i has a rich se of disorionary axes like axes on consumpion purchases, labor and capial income, as well as profis, so ha many fiscal policies insrumens can be analyzed. Second i is suiable for he quaniaive policy analysis, forecasing and assessing he macroeconomic effecs of he alernaive fiscal consolidaion sraegies as well as he disribuional effecs of many shocks in small or large counries. Third he NAWM enables us o beer undersand he impac of a ax reforms on he labour-marke and he economic performance. Fourh i incorporaes numerous nominal, real rigidiies and real fricions empirically significan like he sicky prices and wages, he habi persisence in consumpion, he invesmen adjusmen cos, he inernaional financial linkages, besides he financial fricions ha is he fracion of liquidiy consrained household. In his perspecive Merens and Ravn (2011) repored ha such fricions dampens he shor-run response of oupu, consumpion, invesmen and hours of work o unanicipaed ax changes and ameliorae he model s quaniaive performance. I is ineressan o compare beween Tunisia and Euro Area Licensed under Creaive Common Page 2

3 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom in erms of he naional ax srucures given is significan facor in explaining he differences in he labor uiliisaion and he economic performance. In fac he ax wedges in Tunisia is higher han in Euro Area. As he overall ax wedge in Tunisia amouns o roughly 74.3 % of he earnings of an average producion worker, while ha in he Euro Area is abou 64%. Thus Tunisia has an overall ax wedge abou 10% poins higher han ha of Euro Area. One of he imporan conribuions of his paper is o use his model in an emergen counries viz. Tunisia and o fi he model wih he daa. Indeed he model is frequenly used in Euro Area, rarely used in Emergen Asia and Brazil (see Jacquino and Sraub (2008); De Carvalho and Valli (2011) whereas i seems o be he firs one o my bes knowledge o be applied in Tunisia. I is worh noicing ha here are some differences from he CMS model considered as a conribuion o he lieraure of he medium-scale (DSGE) models for he fiscal policy analysis. Firs he labor marke is augmened by inroducing he heerogeneiy in labor skill o allow wage heerogeneiy amongs he households ha supply he same amoun of worked hour. In his perspecive, Valli and De Carvalho (2010) précised ha his heerogeneiy wihin he household generaes an imporan differences in he impulses responses of he model compared o CMS. Second he recursive form of he opimal wage seing as well as he wage dispersion indices s degree and he demand funcion for labor I and J, are correced relaing o he version of CMS. Third, he risk premia is added in he budge consrain. Ulimaely, he modeling of he domesic economy s moneary auhoriy wih he radiional Taylor forward looking rule includes an expecaions scheme of he inflaion o beer approximae he conduc of he policy o an inflaion argeing framework as picked up by Ben Chalbia (2010). Wih regards o he lieraure, here are an exensive exising lieraure exploring he effec of fiscal consolidaions on macroeconomic aggregaes, especially he consumpion and he oupu. For insance, Coenen e al. (2012) who examined seven srucural DSGE models used by policymaking insiuion exposed o a discreionary fiscal simulus shocks using seven differen fiscal insrumens. Also Coenen e al. (2008) who evaluaed he macroeconomic effecs of an alernaive fiscal consolidaion policies as a permanen reducion in he argeed governmen deb-o-oupu raio in he NAWM model. This srand of lieraure is also pursued by Kiao (2010) who described he emporary ax-cu and a rebae ransfer o simulae he economic aciviies. Evenually, De Carvalho and Valli (2011) who modeled a fiscal policy ha pursues primary surplus arges o sabilize he deb-o-gdp raio in Brazil and explored some of he implicaions of he fiscal policy wih he NAWM. I is worh noing ha he main objecive of his paper is o evaluae he benefis and he spillovers of reducing he axes using wo major governmen revenue componens which are he income axes, as well as he indirec axes on he consumpion goods. Beginning by he expendiure fiscal side, he ransiional effecs of he Licensed under Creaive Common Page 3

4 Soulaima & Besma moneary and governmen spending shock will drive down he oupu in he medium and he long run while he ransfer shock will push up he oupu in he medium run. For he fiscal revenue side, he sraegy based on lowering disorionary axes, leads o an expansion of he oupu in he long run.however, he resul differs in erms of consumpion as i deceleraes in he long run when reducing he consumpion axes and rises when reducing labor axes. Likewise Papageorgiou (2012), i is discovered ha he policy insrumen followed by he reducion in he consumpion ax rae causes a fall in he oupu which will recover in he long run. Furhermore, he exisence of heerogeneous household makes imporan he disribuional effecs of he fiscal consolidaion (Coenen e al. 2008).The paper is oulined as follows, secions 2 illusraes he specificaion of he NAWM model focusing on he exension proposed, secion 3 provides deails abou he calibraion sraegy and performs a dynamic simulaion.secion 4 experimens differen fiscal consolidaion sraegies and analyses he dynamic responses of he model as well as he disribuional effecs under many shocks. Finally, secion 5 concludes. THE MODEL The model following he spiri of (Coenen e al,2008),consiss of wo economies of a differen populaion size s and s-1. The novely in his conex is o inroduce he Tunisian economy, which is an emergen counry, proposed as a home economy and consider he Euro Area as a foreign economy. The foreign economy is modeled exacly in he same spiri as he CMS model. There is an inernaional linkage from he rade of goods and inernaional asses financial markes, enabling imperfec exchange-rae pass-hrough o consumer prices and financial inermediaion coss leading o he imperfec risk sharing. In every counry of he NAWM model, here are four ypes of he economic agens: he households, he firms, he fiscal and he moneary auhoriy. I is assumed ha he domesic economy is represened by a coninuum of households grouped ino wo ses I and J. The members of household I are indexed by i0, 1while he ohers are indexed by j 1,1. The firs ype of household J are called he consrained household or he non ricardian agens.they holds only heir labor income, canno rade in he financial and physical asses,and have no access o he capial markes and invesmen choices. The laer differ in erms of heir labor skills. Neverheless, he household I are called non consrained or ricardian households. They maximize heir lifeime ineremporal uiliy by choosing he consumpion good Ci, and he labor services Ni, so hey ren services o firms. They also choose he privae invesmen good I i, o purchase, which is linked o he nex period physical privae capial sock Ki, 1, he inensiy of capial Licensed under Creaive Common Page 4

5 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom uilizaion ui in producion. They accumulae physical capial; have access o financial markes, where hey rade domesic governmen bonds of he nex period Bi, 1 and inernaionally or privae foreign bonds BF i, 1. They also hold money for M i, for ransacion purposes, are wih beer labor skills and have full access o savings echnologies. Thus he uiliy funcion is defined as follows: E k 0 k C kc N i, k1 i, k1 1 i, k (1) 1) In his case denoes he ineremporal discoun facor, σ is he inverse of he ineremporal elasiciy of consumpion subsiuion, is he inverse of he labor effor s elasiciy relaive o he real wage or he Frisch elasiciy of labor supply, κ is he exernal habi persisence or he habi formaion in consumpion. This funcion is subjec o he period by period budge consrain : c (1 v )) P C P I W i, v i, k k i, R B K ( B ) rp R M F i, N (1 )( R u ( u ) P ) K P K (1 ) D TR T B S B M i, c, i, I, i, u 1 i i, 1 I, ((1 i, B F F I, i, 1 F F, ) S B i, 1 D i, i, i, i, i, i, i, N (1 ) W h i, 1 (2) The seady sae spread rp beween he ineres raes of foreign and domesic bonds is considered essenial in he emergen counries such as Brazil or Tunisia. In fac his risk premia,is included in he budge consrain of he household I,as As he spiri of De Carvalho and Valli (2011), in order o ake in consideraion he fac ha he real ineres raes in Tunisia are subsanially higher han in he developed world even in he seady sae. Then, S is considered as he nominal exchange rae expressed in erms of unis of he home currency per uni of foreign currency. R and RF, are respecively he riskless free reurns or he ineres rae on he domesic governmen bonds and he foreign bonds. (B F B F ) is he financial inermediaion premium ha he household member I should pay on heir inernaional deb when he exernal deb deviaes from he seady sae in he inernaional marke. i, is he incurred foreign risk inermediaion premium rebaed in a lump-sum manner in he negoiaion of he inernaional bonds. When γ B F >0, he domesic household members do no hold inernaionaly raded bonds and he ne foreign asse posiion is zero worldwide in he sochasic seady sae..this cos v ( v i, ) is he echnology ransacion cos on he purchase of he consumpion good is linked o he inverse of he household member's money-o-consumpion raio vi, and o γ v, 1, γ v, 2 > 0 (Schmi-Grohe and Uribe, 2007) and is idenical across boh ypes of Licensed under Creaive Common Page 5

6 Soulaima & Besma households. The household member I hold he sock of coningen securiies i,.this laer acs as an insurance agains he risks on he labor income or he wage income raded amongs all members of hese households. Concerning he earning side, W i, is he wage earned by household i for he uni of he labor services Ni,. Pc, and PI, are respecively he price of a uni privae final consumpion and he invesmen good. Rk, is he rae of reurn on he privae capial services rened o firms. TRi, is he ransfer s paymen from he governmen. Di, is he dividend paid by he household-member-owned firms. To finance is expendiure, he fiscal auhoriy akes par of he gross income of he household. I imposes he axes on differen sources of he income like he wage income W i, Ni,, he capial income R K K i,, he consumpion purchases and he dividend Di, income. Where c is he ax imposed on he consumpion purchase, N is he labor income axes, D is he axes on dividend, W h is he household s conribuion o social securiy imposed on he wage income, W h is he ax rae represening he firm s conribuion o social securiy, K for he foreign economy. This insrumen is is he capial income axes, T i, is a lump sum only paid chosen o sabilize he deb-o-gdp raio(de Carvalho and Valli,2011). I depends on he fiscal auhoriy s arge for he raio of governmen deb o oupu. The capial accumulaion owned by he household I depends on he generalized (, /, 1) adjusmen cos of invesmen Ii Ii. The laer is formulaed in erms of changes in he invesmen which hinge on he parameer of invesmen adjusmen cos funcion I and he depreciaion rae. For is par, he capial uilizaion cos of uilizing he physical capial sock from he seady sae u ( u i, ) relies on varying he inensiy u i, and he capial uilizaion cos funcion parameers u, 1, u, 2 >0. Also i is assumed ha boh ypes of he households ac as wage seers in monopolisic compeiive labor markes. As a consequence, hey supply sufficien differeniaed labor services o firms o saisfy he labor demand. The wages seing W i, are characerized by sicky nominal wages se à la Calvo (1983). The household can opimally rese heir nominal wage conrac wih probabiliy ~ laer choose he same wage rae W I, ~ W i, 1 of opimizing each period. The as well as he indexaion resuling in wo separae wages Phillips curves. W ~ I, is he opimal wage conrac chosen by he household I receiving he permission o rese heir wage in period. These households are supposed o maximize he lifeime uiliy, as defined by equaion (1).The oher households who canno opimize, readjus heir wages W i, based on a geomeric average of pas changes in he price of he privae consumpion good Pc, and he seady sae consumer inflaion rae c wih I he weigh of pas Licensed under Creaive Common Page 6

7 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom inflaion. The aggregae wage is based on he wo previous equaion of he wage seing so ha he wage index W I, evolves according o: W I, ~ 1 P, 1 1 (( 1 )( ) (( 1 I C ) ) ) 1, 1 1, I I I I W I I C W I P C, 2 I (3) The same reasoning holds for he household J. Turning o he firms, here are wo ypes of firms in he producive secor of he economy indexed by f 0,1.They operae under monopolisic compeiion wih Calvo-ype price rigidiies. The firs one produces he inermediae radable goods Y f, sold in he domesic marke. The second se, produces he non radable final goods like he privae consumpion QC, he privae invesmen Q I and he public consumpion good QG.The Final goods firms are assumed o operae under imperfec compeiion. They combine he impored wih he domesic inermediae goods in heir purchases. The inermediae good firm s f produces is inermediae good oupu Y f, in he compeiive markes, wih he increasing- reurns-o-scale Cobb-Douglas echnology, using he homogenous capial services, K f,, rened from he household I as inpu as well as he index of labor services N f, yielding a echnology consrain : Y f, max[ 1 z K f, N f,,0] (4) Where is he share of fixed cos in producion. z is he seady sae level of produciviy or he oal facor produciviy assumed o be idenical across firms. I evolves over ime according o serially correlaed process. In his case z is he auoregressive coefficien of produciviy and z, is a whie noise shock o produciviy. Since all he firms f face he same inpu prices and have access o he same producion echnology, he nominal marginal cos MCf, is idenical across firms ha is MCf, = MC. To adap his model o he emergen counry, we follow he approach of Fabia and De Carvalho (2010) and assume ha he firms ren labor wih unequal labor skills and ha he individual wih lower level of labor skills are more consrained on heir invesmen possibiliies and has lower level of educaion. This sraegy of modeling enable a seady sae where skillful workers earn more ye while working he same hours as he less skilled one. In Tunisia he labor conracs require 8 hour work day. The labor inpu used by firms f o produce he labor oupu N f, is given by: N (( 1 ) ( ) I 1 N ( ) ( N ) ) 1 J f, f, f, (5) Licensed under Creaive Common Page 7

8 Soulaima & Besma Where is he populaion size adjusmen o he firm s labor demand and he parameer [0,1 / ] is added o inroduce a bias in favor of a more skilled worker. When 1 his means ha here are equally skilled workers. The parameer > 1 denoes a price elasiciy o demand for a specific labor bundles beween wo households. N I f, and N J f, are expressed in erms of he inraemporal elasiciy of subsiuion beween he differeniaed labor services I, and J of he household I and J. Firms choose he quaniy required o hire in boh groups of households by aking average wages W I, and W J, and minimizing oal labor cos N I f, W I, + N J f, W J, subjec o N J f,.i yields from he firs order condiion he aggregae wage index given by: 1 (( 1 )( ) 1 W W ( W ) ) 1 I, J, 1 (6) In his case W I, and W J, are he nominal wage indexes for he household I and J. While he aggregae demand funcions for he labor of each group of households J and I are expressed as follow: N J f, W J, N W f, (7) N (1 )( W I I, f, ) N f W, (8) As for he households, when aking ino accoun he price seing, each inermediae firms charges differen prices a home and sell is differeniaed oupu Y f, in boh domesic and foreign marke under monopolisic compeiion wih calvo price rigidiies. Firm f receive permission o se opimally prices P opimizing H or X. H f,, and PX, f, in a given period wih a probabiliy of no However, he firms which are no allowed o opimize heir price, adjus heir price in he domesic and foreign marke such ha he price conracs P H f,, and PX, f, are indexed o a geomeric average of he pas inermediae good inflaion. Where PH, and P X, are he price indexes and H and X are he indexaion parameer deermining he weigh on he pas inflaion.more precisely H is he degree of domesic price of home good indexaion and X is he degree of foreign price of home good indexaion (expor good). H and X are he domesic and foreign inermediae goods seady sae inflaion raes. Licensed under Creaive Common Page 8

9 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom Afer aggregaing over firms he aggregae price index PH, se by he inermediae good firms f for he domesically sold produc evolves according o: P )( ~ 1 1, 1 1 (( 1 ) (( 1, PH, ) H H H P H H H PH, 1) ) 1 PH, 2 H (9) The same equaion holds for he aggregae index of he price conracs P X, se for he differeniaed sold abroad produc. As i is menioned previously for he hree ypes of final good firms, o produce he non-radable final privae consumpion good QC,he firms purchases he domesic and foreign impored inermediae goods respecively H C and IM C using a consanreurns-o-scale CES echnology: c 1 C 1 1 C (( (1 C ) (1 ) (1 IM IM C C Q C C H C C C ) ) 1 C IM C C Q C Where IM C ( IM C Q ) is he impor adjusmen cos faced by he consumpion-good firms when adjusing he use of impors i.e. is impored share of inpus in he producion of he final consumpion good. This cos is linked o he impor adjusmen cos parameer IM C. (10) of he consumpion c 1 is he inraemporal elasiciy of subsiuion beween he bundles of domesic and foreign inermediae goods. c defines he home bias in he producion of he final good s consumpion oward domesic inermediae goods. The weigh of he inpus in he consumpion baske are defined as ( 1 C ) and C.I is worh saing ha H C f, and IM C f * are respecively he use of he inermediae goods produced by he domesic and he foreign firm or foreign exporer f and f *. * 1 are he inraemporal elasiciies of subsiuion beween he domesic and foreign inermediae goods. Afer choosing he opimal use of each inermediae good f and f * by he consumpion good firms, he laer will combine he domesic and foreign C, inermediae good bundles H C and IM C ha minimize he oal inpu cos PH H PIM, IM C subjec o aggregaion consrain QC menioned above so ha he uni price of he privae consumpion PC, and he demand funcions for he inermediae good H C and IM C are obained respecively. The same reasoning holds for he modeling of he non radable final privae invesmen good QI.The firms combines is purchase of he domesically-produced inermediae goods and he impored foreign inermediae goods respecively H I and IM I, using a consanreurns-o-scale CES echnology.then, he final non radable public consumpion good QG is assumed o be only made of domesic inermediae good; ha is QG = H G.The opimal demand Licensed under Creaive Common Page 9

10 Soulaima & Besma for he domesic inermediae good f H G f, is linked on he uni price of good P G, PH, he public consumpion.the demand for he domesic and foreign inermediae goods f and f * are obained hrough aggregaing across he hree final-good firms where: H H C + H I + H G and IM = IM C + IM I. When i comes o he fiscal auhoriy, he CMS assumpion is considered. In his case, he fiscal domesic and foreign auhoriy purchases unis of he public consumpion good G wih price PG,,disribue ransfer TR o each household group in unequal amouns, issue bonds o refinance is deb B, earns seignorage on ousanding money holdings, M 1, levies axes on consumpion, labor, capial and dividends and adjuss expendiures and budge financing accordingly(see he period-by-period budge consrain of he fiscal auhoriy above).on he expendiure side, wo alernaive insrumens are considered: he real ransfer, he real governmen consumpion and he lump sum axes. Beginning wih he governmen consumpion consolidaion sraegy, he fiscal auhoriy adjuss he consumpion spending as a share of g Y seady-sae nominal oupu, PG, G / PY.This final public consumpion good, follows he auoregressive process. Thus ρ g is he auoregressive coefficien of governmen consumpion and ɛ g, is a whie noise shock o governmen expendiure. Turning o he ransfer, he fiscal policy adjus he ransfer as a fracion of he seady-sae nominal oupu assumed like Coenen e al.(2007) ha he fiscal policy via ransfers is unevenly disribued across he wo ypes of household, favoring he households J over he household I. Like (Coenen e al. 2008),i is assumed ha he loss in revenue is financed by a decrease in he fiscal auhoriy s ransfer paymens o households such ha he governmen spending and debo-oupu raios remain unchanged in he long run. The ransfer evolves according o he auoregressive process. This is explained by he auoregressive coefficien of he public r TR P Y Y.I is ransfers r.in fac he seady sae value of governmen ransfers is defined by r and he whie noise shock o governmen ransfers by r. These expendiures are financed by a differen ypes of disorionary axes, including axes on consumpion spending, labor and capial income. The oher disorionary axes are exogenously se by he fiscal auhoriy considered as consan X X wih X C, D, K, N, W h and W f. To model he Tunisian moneary auhoriy, we refer o he oucome of Ben Chalbia (2010) who displayed ha he Cenral Bank of Tunisia is direced more owards he fuure, precisely a ypical Taylor ineres rae rule under he forward-looking version feauring he ineres-rae smoohing. 4 PC, ( 1R)[ Y R ( ) ( g y) s ( S S1 R PC, 4 Y 1 R 4 4 R R ) 1 g y, (11) Licensed under Creaive Common Page 10

11 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom This equaion is deermined in erms of he consumer price inflaion and a quarerly oupu growh o provide an addiional informaion on he fuure pah of he inflaion and he oupu. Also i is crucial o menion ha he Taylor equaion is enriched wih he exchange rae as he Tunisian moneary auhoriy is assumed o follow a flexible exchange rae regime. Where is he annual inflaion arge of he moneary auhoriy, R4 = 4 is he annualized quarerly nominal equilibrium ineres rae, g y is he rae of he oupu growh in he seady sae(assumed o equal 1).The erm R, is he whie noise shock o he ineres rule or he moneary policy shock. I is worh o remember ha he foreign economy, follows he CMS represenaion. Also he same formulaion are considered for he marke clearing condiions holding in he equilibrium, excep some modificaions in he labor marke. These modificaions concern he degree of he wage dispersion indices saed in a recursive formulaion which are defined as follows: s ~ W (1 I)( W 1 I I, I C I I I, ) I ( c, 1 ) I, W I, s I, 1 (12) s ~ W J, W J, (1 J)( PY Y PYY 1 1 J I J C J J, ) J ( c, 1 ) W J, s J, 1 (13) Where W I, and WJ, are he wage inflaion raes of he household I and J. The aggregae demand for he labour services of variey i and j across he inermediae-good firms are respecively N I and N J. The same process is considered for he differeniaed labor services supplied by household J. CALIBRATION AND SIMULATIONS Calibraion There are differen mehods o esimae and evaluae DSGE models. In his case, he model is no esimaed bu insead calibraed on he Tunisian economy using an annual daa over Table 1 and 2 summarize he main differences beween he block of he wo regions. Table 1: Macroeconomic Aggregaes (seady sae raios) Raio Euro Area Tunisia PC C / PY Y PI I / PY Y PG G / PY Y PIM IM / PY Y M /( 1 C) Pc C c PIM IM / PY Y I PIM IM / PY Y B / P Y Y D / P Y 0 0 Y Licensed under Creaive Common Page 11

12 Soulaima & Besma The able shows he seady sae raios of he main expendiure caegories over nominal oupu. In order o obain an accurae resul, we ry o fi he NAWM o he sylized facs of he Tunisian economy daa which is no an easy ask for a model adaped o he Euro area. The calibraion for he Euro Area is similar o he one of (Coenen e al, 2008). However for Tunisia, we follow he radiional lieraure and he microeconomic evidence. In fac he naional accouns are considered from 1980 o 2011o se he seady-sae raios in he home counry. The laer is provided by he Inernaional Financial Saisics (IFS) and he World Bank daa. However concerning Tunisia,he parameer s value are aken similar o he one of emerging marke in he absence of reasonable proxies in he lieraure. This model involves 138 parameers comprised of aggregae share or raios, seady sae values, 231 endogenous variables and 21 shocks. The able 1 corresponds o he macroeconomic aggregaes or he seady sae raios (see he annex1 for more deails). The shares of he aggregae demand componens in GDP PC C / PY Y, PI I / PY Y, PGG / PY Y ) are based on he sample average calculaed from he naional accouns. Beginning wih he raio of he privae consumpion o oupu PC C / PYY, i is se o 0.63 in Tunisia. This value is no far from he one in Turkey which is calibraed o 0.7 by Çebi (2012). To obain a consisen specificaion of he seady sae rade linkage, i is elemenary o calibrae he impor o oupu raio in privae consumpion good PIM IM / PY Y or he degree of openness, which is ; he share of impor c PIM IM / PYY ha is 0.13 and he share of privae invesmen good I PIM IM / PY Y which is 0.09 in Tunisia. The money o consumpion raio M /( 1 C) Pc C is Moreover, o measure a counry s abiliy o make fuure paymens on is deb, i is crucial o deal wih he seady sae of he governmen deb o GDP raio B / P Y Y. The seady sae raios of governmen deb o oupu raio is fixed o 2.3. For is par, he share of governmen consumpion o oupu IFS Tunisian daa. PG G / PYY, is calibraed o is real value 0.16, on he basis of he annually Table 2: Calibraed Parameers and Seady Sae Variables (Household) Household Euro Area Tunisia s (relaive o Africa) I, J, I J Licensed under Creaive Common Page 12

13 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom The dividend income-o-oupu raio D / P Y Y is assumed o be zero in he seady sae. Then, he share of privae invesmen o oupu PI I / PYY is se o The calibraed parameers and he seady sae variables concerning he household s behavior is exhibied in he able 2 (see he annex 1 for more deails). In fac, he financial marke fricions w relaed o he size of he household J is close o 53 %. The rae of ime preference or he subjecive discoun rae is equal o 0.98.The annual depreciaion rae of he privae and public capial δ amouns o across all regions. This value is deeced by Ben Aissa and Rebei (2012) and Jouini and Rebei (2014) in Tunisia. I is sandard in he lieraure (see Cooley and Presco, 1995) or (Burda and Weder, 2002). The producion share of capial income in GDP, α=0.3 is oped for Tunisia which is very close o a ypical assumpion found in he lieraure usually around 1/3.Hence he esimae seems very reasonable. However, here is no agreemen in choosing he magniude of he inverse ineremporal elasiciy of subsiuion. Generally he esimaes are ranging from 1 o 4 in he lieraure (Lehmus, 2011). In Tunisia i amouns o 1 as he value chosen by Ben Aissa and Rebei (2012), Gouvea e al. (2008); Valli and De Carvalho (2010) in Brazil. The inverse labor supply Frisch elasiciy is picked o 2. The range commonly adoped in he business cycle lieraure is beween 1 and 2 especially in he emerging marke as deeced by Jacquino and Sraub (2008) as well as Gouvea e al. (2008) in Brazil. Our calibraion of he degree of habi formaion κ is fixed o 0.5, in line wih he research carried ou by Ben Aissa and Rebei (2012) in he Tunisian economy following he changes in he level of consumpion in developing counries. The calibraion of he Calvo parameer deermining he persisence of wage seing I, J, or he fracion of boh household members I and J no seing wages opimally each quarer corresponds o 0.75, consisenly wih Ben aissa and Rebei (2012) and Jacquino and Sraub (2008). The degree of wage indexaion I J is also seleced o Turning o he analysis of he able 3 relaive o he inermediae good firms. Firs, he share of he fixed cos in he producion is aken so ha o ensure zero profis in he seady sae, and discourage he oher firms o ener he marke in he long run, as documened by Vali and De Carvalho, (2010); Coenen e al. (2007); Chrisoffel e al. (2008). The value is picked o 0.143, similar o ha employed in Emerging Asia (see Jacquino and Sraub, 2008). Second, he seady sae produciviy level z is normalized o uniy (saionary). Third, he calvo probabiliy of he inermediae good firm s fracion selling heir differeniaed oupus in he domesic markes H or he fracion of home firms no seing heir prices opimally each quarer corresponds o The laer is consisen wih Medina and Soo (2007) respecively in Columbia and in Chile. For is pars, he degree of indexaion H is se o 0.5 (see Valli and de Carvelho, 2010). Licensed under Creaive Common Page 13

14 Soulaima & Besma The fracion of foreign firms no seing foreign prices opimally X is fixed o 0.3 and he corresponding degree of price indexaion X is fixed o 0.5. The inraemporal elasiciy of subsiuion I and J beween he household-specific bundles for he differeniaed labor services, are designed o 6. This value is also noiced in Brazil by De Carvalho and Valli (2011). On he oher side he informaion relaed o he final good firms are repored in he able 4(see he annex 1 for more deails). The inraemporal price elasiciies of subsiuion beween he home and foreign inermediae goods in forming he consumpion and invesmen bundles C, I is seleced o 1.5.This value is approximae o he one se by Çebi (2012) in Turkey. The home bias parameer, in paricular he one ha corresponds o he share of domesicallyproduced goods in he oal privae consumpion expendiure ν c, is seleced such ha i equals is empirical counerpar, precisely as 0.49 in Tunisia. This value is no very far from he one oped in Chile 0.65 by Medina and Soo (2007). Table 3: Inermediae good firms Inermediae good firms Euro area Tunisia z z I H X H X J Noe: The subscrip I represen he unconsrained households while J represen he liquidiy consrained household in he model. The same hing for H which sands for he radable secor while X represens he expor secor in he model Meanwhile he home bias in he producion of invesmen final goods ν I is calculaed o 0.5 in Tunisia see Medina and Soo (2007) and Burriel e al. (2010) in he Chilean economy and he Spanish economy. De Carvalho and Valli (2011) calibrae he home biases ν c and ν I from he demand equaions of impored goods using he seady sae value for he supply of consumpion and invesmen goods, and he impor quanum. On he oher hand, (Chrisoffel e al. 2008) chose seady-sae values for he home-bias parameers ν C and ν I ha allow he model o replicae he impor conen of consumpion and invesmen spending roughly 10 and 6 Licensed under Creaive Common Page 14

15 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom percen, expressed as shares of nominal GDP.Moreover, o ensure ha he households groups I and J work he same amoun of hours, he labor demand bias ν w is calibraed o Finally, he inraemporal elasiciy of subsiuion designed as he price elasiciy of demand beween he domesic and foreign inermediae goods for a specific inermediae-good variey is picked o 6, implying a seady-sae mark-up of 20%. A common value used in he lieraure, is confirmed by De Carvalho and Valli (2011) and Galí e al. (2005) for a small open economy in all. The calibraion is sill pursued wih he moneary auhoriy policy ha is observed in he able 5. For he parameer relaed o he ineres-rae response coefficiens on he annual inflaion in erm of deviaion from an inflaion arge, also called he ineres rae sensiiviy o inflaion gap,our reference is Çebi (2012),who ook a mean of 1.5 in Turkey, and Jouini and Rebei (2014) in Tunisia. Also, based on he calibraion of his laer and on Gouvea e al. (2008), he ineres rae sensiiviy o he oupu growh gap g y corresponds o 0.25 in Tunisia. The coefficien on he lagged ineres rae or he degree of he ineres rae ineria R, is designaed o 0.5 wih reference o he prior mean for he ineres rae smoohing parameer chosen by Jouini and Rebei (2014) and Çebi e al.(2012). On he oher side, he response of moneary auhoriy o exchange rae variaion s is se o 0.8.The seady sae of inflaion π is seleced o 1. In he emergen economies like Brazil, his value amouns o 1.04 (seevalli and De Carvalho (2010)). Afer calibraing he moneary auhoriy policy s parameres we now perform he fiscal auhoriy s parameers observed in he able 6 (see he annex 1).Firs, he sensiiviy of he aggregae lump-sum axes wih respec o he governmen deb-o-oupu raio BY is designaed o in Tunisia which is close o he value proposed by De Carvalho and Valli (2011) and Gouvea e al. (2008) in Brazil. Second, he Governmen deb-o-oupu raio BY is se o Third, he governmen ransfers-o-oupu raio or he governmen ransfer ( r ) is picked so ha he fiscal auhoriy's budge consrain is balanced a he neural deb-o-oupu raio or saisfied in he long run, is se o 0.05 in Tunisia. This raio is higher in he Euro Area due o a more generous unemploymen benefi sysem, a higher emphasis on pension schemes (Roeger and in Veld, 2009). The oher parameer which concerns he governmen consumpion-o- oupu raio ( g ) or he governmen expendiure, is based on he naional accouns daase seleced as Boh he governmen spending-o-oupu raio ρ g or he persisence of public consumpion and he public ransfer-o-oupu raio ρ r, have an auoregressive coefficien of Moving o he ax raes calibraion, some values are calculaed based on he curren ax law of Tunisia on average (see able 6 annex 1). They are presened as follow: 26.83% for he ax on income profis and capial ; c =29.5% for he consumpion axes, W h =7.64% and W f =17.21% for Licensed under Creaive Common Page 15

16 Soulaima & Besma subsequenly he social conribuion of he employees and employers, 0 for he ax rae on dividend τ D,as he dividend versed by he Tunisian socieies are no submied o ax and N =0.2 for he labor income ax rae. Evenually, able 7 summarize he adjusmen and ransacion cos. In fac 2 appearing in he funcional form of he consumpion ransacion for he domesic economy, is oped consisenly o Coenen e al. (2008) and Jacquino and Sraub (2008) equal o Table 4: Final good firms Final good firms Euro area Tunisia C I w C I Noe: The subscrip C represens consumpion goods secor while I sands for he invesmen goods secor. IM means he impored goods secor For he oher parameer 1, we refer o Jacquino and Sraub(2008) in emerging Asia and se i o in Tunisia. In addiion, he invesmen adjusmen cos funcion (he share of domesic good in he invesmen) I is calibraed o 4.58 in Tunisia based on IFS. In fac, based on he value of a prior disribuion se on Brazil o 4 by Gouvea e al. (2008) or 3 se by Valli and De Carvalho (2010). The capial adjusmen cos parameer value u1 is se o 0.05 following Valli and De Carvalho (2010). In addiion o u2 ha is picked o wih reference o Jacquino and Sraub (2008); Valli and De Carvalho (2010, 2011). For he parameers conrolling he adjusmen cos relaed o changing he impor share in consumpion IM C, hey are deeced o 4 in Tunisia based on he value se in Brazil, o dampen he sensiiviy of consumpion o he change of he erms of rade. Table 5: Moneary auhoriy Moneary auhoriy Euro Area Tunisia gy R s Licensed under Creaive Common Page 16

17 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom This parameer deermines he abiliy of household I o borrow from abroad in he shor o medium run. I influences he profile of is ineremporal consumpion and invesmen decision. Besides, he parameers of he impor adjusmen cos of invesmen IM I, akes he value of 2.5 (see Jacquino and Sraub, 2008). To end, he parameer governing he inermediaion cos funcion, also called he risk premium BF, is around 0.05 in Tunisia (see Jacquino and Sraub, 2008) so ha he evoluion of ne foreign asses has a small impac on he exchange rae and he rade in he shor run. I enables he ne foreign asse posiion o be sabilized o zero in he long run. Table 6: Fiscal policy Fiscal auhoriy Euro Area Tunisia B Y BY g g r r c D N W h W f Noe: The calibraion of he ax raes of he Euro Area, are similar o he one seleced by Chrisoffel e al.(2008) who provide he informaion from he OECD (2004) and Eurosa (2006). Simulaion and Policy analysis The simulaion in his paper concerns he dynamic responses effecs of moneary and fiscal shocks like he governmen spending and he ransfer o household shock as well as he dynamic properies of he NAWM illusraed in he figure 1. This figure depics he dynamic responses of he domesic variables o a moneary policy ighening represened by a black line, equal o a 1% increase in he annualized nominal ineres rae; o a persisen governmen spending shock, represened by a blue line, equal o a 1% increase in he seady- sae oupu and o a persisen ransfer shock, represened by a green dashed line, equal o 1% increase in he seady-sae oupu. I is crucial o noe ha all he dynamic responses are repored as percenage deviaions from he seady sae. Also, all he simulaions were achieved using he funcion soch_simul of DYNARE a MATLAB. Licensed under Creaive Common Page 17

18 Soulaima & Besma Table 7: Adjusmen and ransacion cos Adjusmen and ransacion coss Euro Area Tunisia u1 u2 I IM C IM I 0 5 BF The expendiure based consolidaion Afer he calibraion, we describe he dynamic response of he macroeconomic variables following he moneary, governmen expendiure and he ransfer shock. Moneary shock Afer a posiive ineres rae here is a fall in he oupu Y as emphasized by Çebi (2012); Rao e al. (2009); Chrisoffel e al. (2008); Fabia e Valli (2011); Burriel e al. (2010). This decline will pu pressure on he consumpion and invesmen and he labor demand N f,. Indeed Chrisoffel e al. (2008) elucidaed he drop of he invesmen by a emporally reducion in he domesic demand. I is crucial o remember he link beween he capial K i,, he ineres rae and he deb o oupu raio. Following his shock, i is noiced ha he ineres rae declines hen moves o zero which will boos he deb o oupu raio. De Casro e al. (2011) hough ha keeping he ineres rae fixed for a cerain period yields a fall in he inflaion. On he oher hand, here is a lower demand for he capial. For is pars, he inflaion especially he consumer s price inflaion c, slighly pushes up hen levels off. Chrisoffel e al. (2008) explicaed his decline by a fall in he domesic prices. The nex variable is he erms of rade ToT which is defined by he domesic impor price relaive o he expor price in domesic currency. Following he moneary shock, i falls hen will recover. Indeed, Chrisoffel e al. (2008) and Coenen e al. (2008) claimed ha he improved ToT lead o changes from domesic owards impored goods. More precisely, Coenen e al. (2008) jusified his improvemen by he appreciaion of he domesic currency. When i comes o he real wages W i,, i is noiced an upward rend in he firs 10 quarers ha will depreciaes in he medium run. Indeed, De Casro e al. (2011) and Fabia e al. (2011) elucidaed he delayed reducion in he wage by he curailed demand for he labor by he firms, due o he decline in Licensed under Creaive Common Page 18

19 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom he oupu. This fall pu downward pressure on he domesic price H.As far as he impor IM is concerned, here isa coninued fall of he laer. In his sense De Casro e al. (2011) poined ou ha his fall is caused by he decline in he oupu. Likewise, he impor prices PIM, dips.finally, we explain he relaionship beween he rade balance TB, he real exchange rae and he expor. As shown in figure 1, here is a negaive impac on he expor. Tha is why he rade balance sabilizes in he long run. Indeed Chrisoffel el al. (2008) and De Casro e al. (2011) inerpreed he reducion in he expors, by he improvemen of he ToT inducing a swiching effec o he impored goods. The real exchange rae, follows nearly he same rends of he ToT. Figure 1: Dynamic responses o a moneary, governmen and ransfer shock Transfer shock As regards he oupu, he laer hikes in shor run, pushing up he consumpion Ci, and he invesmen I i,, hen will drop in he long run. Precisely, he consumpion deceleraes by a small amoun a he beginning of he period hen will sharply recognizes an ascension in he medium and he long run.in fac, likewise many auhors Rao e al. (2009); Fabia and Valli, (2011); Forni e al. (2009),i is found ha he rise in he consumpion will pu upward pression on he wage in he long run.moreover as menioned before, he invesmen hikes. In his field, Coenen e al. (2008) revealed ha he unconsrained household amelioraes he invesmen purchases.as hey anicipae a rise in he opporuniy cos of capial formaion, encouraged by he decline Licensed under Creaive Common Page 19

20 Soulaima & Besma in he ineres rae R. For is pars, he ineres rae delicaely pushes up hen levels off promoing he accumulaion of he capial. Tha is why he deb falls. In addiion he wages declines in he medium run hen will pushes up for boh household. Anoher picures emerges wih he ToT, i rises in he firs 10 quarers hen will decline in he long run. Finally, we noe a decrease in he rade balance as well as he impor. Governmen shock When i comes o he governmen shock, one can see ha he decline of he oupu is moderae. This is explained by he decrease in he labor supply pushing down he expor of he economy. Also he consumpion slighly goes down, hen will sharply boos in he long run. This is induced by he rise of he employmen. A he same ime he invesmen is moving up because of he decline in he ineres rae. According o Papageorgiou (2012) and Coenen e al. (2008), he posiive impac of he invesmen is due o he expansion in hours worked ha will enhance he marginal produc of capial. On he oher hand, he ineres rae R goes up jus for a shor momen, puing downward pression on he deb. Then will fall in he medium and he long run puing upward pressure on he invesmen and he consumpion. Due o his posiive impac on he invesmen and he increase in hours worked, he capial rises, as discussed by Papageorgiou (2012) and Coenen e al. (2008). Conversely, he inflaion has a downward rend, because of he depreciaion in he marginal cos of he consumpion goods. Concerning he wages, here is an iniial decline which will sharply hikes in he long run. Moreover, here is a small decline in he impor a he beginning of he period, hen will peaks sharply in he medium and he long run. We also noe an improvemen of he ToT (decrease), which is inerpreed by he increase in he impors. In fac, he (TOT) deerioraes he rade balance. This rend can be clarified by he increase in he impor in he long run, even if here is a small developmen in he exchange rae. The disribuional effec The figure 2 depics he disribuional effecs of a gradually permanen increase in he moneary ineres rae (represened by a black solid line), he governmen spending shock (represened by a blue solid line) and he ransfer o household shock (represened by a green dashed line),equal o 1% increase in seady- sae oupu, on he domesic variables. The dynamic effecs are repored as percenage deviaions from he iniial seady sae. Saring wih he governmen consumpion shock, he liquidiy consrained households J cu back moderaely heir consumpion C J, due o he marginal drop in heir wage, hen will be recovered in he medium Licensed under Creaive Common Page 20

21 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom erm as heir wage hikes Unlike he consumpion CI, of he household I which is in expansion. In his field, Coenen e al. (2008) added ha he rise in governmen consumpion absorbs par of he economy's resources yielding a negaive wealh impac especially on he par of he household J. As hey feel poorer, hey will diminish heir consumpion and improve heir labour supply N J,. Following he sraegy of increasing he ransfer, he household J sill reduce heir consumpion purchase marginally, hen will srongly enhance i. Ye again, Coenen e al. (2008) poined ou ha here is a loss of income because of he unequal financing requiremen ha hey face. Figure 2: Disribuional effecs of permanen increase in a moneary ineres rae shock, governmen spending and ransfer shock o household The Revenue based consolidaion Concerning he dynamic effec on he revenue- based consolidaion, he figure 3 summarizes he impac of a permanen reducion in he consumpion and he labor income axes. Labor shock Saring by he labor shock, he laer pushes he oupu Y up. In his perspecive, Papageorgiou (2012) who evaluaed he permanen reducion in he labor income ax accompanied by a permanen increase in he consumpion axes, claimed ha he mos effecive policy o boos he oupu, promoe he welfare and simulae he economy is o reduce he labor income axes. He also highlighed ha his sraegy implies a rise in he labor supply of he household as well as he consumpion and he hours worked. Similarly Forni e al. (2009) inerpreed he rise in he oupu by he increase in he aggregae demand. In addiion o Kiao (2010) who jusified he higher oupu and he rapid economic growh by a posiive response of boh invesmen and consumpion. Moving o he effec on he consumpion variable, he laer iniially dampens hen Licensed under Creaive Common Page 21

22 Soulaima & Besma will boos sharply in he medium and he long run. When i comes o he real wages, W i, he laer dampens in he shor o medium run bu rises in he long run. In his conex, Sahler and Thomas (2011) clarified he depression of he wages by he imporance of he unemploymen. This laer is generaed by he drop of he oupu. On he oher hand, he middle lef panel of he Figure 3, shows ha he labor income axes induces an ascension of he capial K i,. In his conex Papageorgiou (2012) added ha he increases in hours worked has a posiive impac on he marginal produc of he privae capial. The laer encourages he rise of he fuure capial formaion which in reurn enhances he oupu in he long run. In he same way Kiao (2010) deeced a rise in capial when cuing he emporary income ax. Then, consisenly wih Sa hler and Thomas (2011), here is a developmen in he invesmen I i, due he fall in he ineres rae. In fac his fall is deeced in he medium and he long run hen will levels. The ( ToT ) slighly goes up in he firs en quarers hen will decline. Neverheless, he impor iniially cus down which will hen sharply jumps in he long and he medium run. Finally regarding he deb o oupu, we found ha he labor income axes drives down his deb in he shor o medium run. In his perspecive, Papageorgiou (2012) précised ha he reducion in labor axes financed by higher consumpion axes no only simulaes he economy bu also reduces he deb o oupu raio. Figure 3: Dynamic Responses o a consumpion and labor shock Licensed under Creaive Common Page 22

23 Inernaional Journal of Economics, Commerce and Managemen, Unied Kingdom Consumpion shock When i comes o he consumpion shock, he oupu sharply declines in he firs 10 quarers, hen will delicaely rises. Besides he consumpion iniially rises hen slighly dampens in he medium and he long run. Furhermore, he real wages iniially pushes up hen will decline in he medium and he long run.turning o he ( ToT ), is movemen shifs across he ime. I declines hen will moderaely rises. There is a fall in he capial as well as he invesmen. Indeed Forni e al. (2009) clarified ha he smooher increase in he Ricardian consumpion gradually shifs resources away from he invesmen. So when here is an income increases, he household I does no rely on he economy's capial sock. To end, we should no forge o describe he effec on he rade balance TB, he impor IM and he exchange rae S variables. In fac, here is sill a reducion in he impor yielding an improvemen of he rade balance. The laer declines marginally hen will picks and sabilize in he long run. Furhermore, he exchange rae goes down. This fall is due o he decline in he expor ha will in reurn enhance he deb. Moreover, wih he labor income and consumpion axes shocks, he consumer price index c slighly pushes up in he shor run hen will sabilize in he medium and he long run. This is also observed by Forni e al.(2009) and Sa hler and Thomas (2011). In fac Çebi (2012) elucidaed he iniial rise of he inflaion by he amelioraion of he marginal cos. However, he decline in he inflaion pus downward pressure on he ineres rae (Forni e al. 2009). The disribuional effec Figure 4: Disribuional effec of consumpion and labor income ax shock In an aemp o highligh he influences of he propagaion of consumpion and labor income shocks on he wo ypes of households s behavior, he figure 4 compares he dynamic responses of many seleced variables. I is observed ha he consumpion ax shock amelioraes he consumpion C J, of he consrained household J when heir wage W J, is improved, hen will reduce i as heir wage drops. The amelioraion of he household J s wages Licensed under Creaive Common Page 23