The Macroeconomic Effects of Migration from the New European Union Member States to the United Kingdom

Transcription

1 WP/07/6 The Macroeconomic Effecs of Migraion from he New European Union Member Saes o he Unied Kingdom Dora Iakova

2

3 007 Inernaional Moneary Fund WP/07/6 IMF Working Paper European Deparmen The Macroeconomic Effecs of Migraion from he New European Union Member Saes o he Unied Kingdom Prepared by Dora Iakova Auhorized for disribuion by James Morsink March 007 Absrac This Working Paper should no be repored as represening he views of he IMF. The views expressed in his Working Paper are hose of he auhor(s) and do no necessarily represen hose of he IMF or IMF policy. Working Papers describe research in progress by he auhor(s) and are published o elici commens and o furher debae. The Unied Kingdom allowed workers from he en new European Union member counries immediae access o is labor marke afer he accession in 004. This paper uses a general equilibrium framework o explore he dynamic adjusmen of he UK economy o he posaccession surge in immigraion. Simulaions show ha immigraion is likely o have posiive effecs on economic growh, capial accumulaion, consumpion, and he public finances. JEL Classificaion Numbers: E, E, E7, J6, H55, F Keywords: Immigraion, new EU member sae globalizaion, free labor movemen Auhor s Address: diakova@imf.org I am graeful o Hamid Faruqee and Naalia Tamirisa for sharing he Mulimod code and o Alina Carare for useful discussions. I would also like o hank James Morsink and paricipans in a seminar a he HM Treasury for helpful commens.

4 Conens Page I. Inroducion...3 II. Mehodology...5 III. Simulaions...7 IV. Conclusions...0 Appendix I. Descripion of he Model... References...7

5 3 I. INTRODUCTION The Unied Kingdom was one of he firs counries o allow workers from he new European Union members immediae access o is labor markes in May 004. The ransiional arrangemens se ou in he Accession Treay allow counries o choose o limi free labor movemen for a period of up o seven year wih reviews of he policy afer he second and fifh years. Only Ireland, he U.K., and Sweden chose no o impose resricions in 004. In he summer of 006, Greece, Porugal, Finland, and Spain also decided o lif all resricions for he en accession counrie and six oher member saes adoped a more liberal policy. Saring in 007, he U.K. will also accep workers from Bulgaria and Romania in cerain secor alhough subjec o igh quaniy resricions. Alhough he exising saisics are imperfec, hey sugges ha a subsanial number of workers from he new member saes have arrived in he U.K. since 004. Daa sources on immigraion include he Worker Regisraion Scheme, he Inernaional Passenger Survey, he Labor Force Survey, and Naional Insurance numbers daa. All employees from he eigh new member saes (EU8) are required o regiser under he Worker Regisraion Scheme (WRS). Beween May 004 and Sepember 006, 50,000 people have regisered, equivalen o an annual increase of approximaely 0.7 percen of he work force. The WRS is he mos frequenly cied daa source, however i is subjec o wo major caveas. Firs, i measures only he gross inflow of workers. There is no de-regisraion requiremen, so people who work in he U.K. for a shor ime and hen reurn o heir counry would sill be couned in he saisics. Second, he self-employed are no required o regiser in he WRS. The Inernaional Passenger Survey (IPS), he official daa source on migraion, couns as immigrans hose who sae an inenion o reside in he U.K. for a leas a year upon enering he counry. Based on ha, he ne flow of immigrans from he EU8 was 49,000 in 004, and 64,000 in 005. The survey is volunary and does no capure people who change heir inenions once in he counry, even hough i aemps o adjus for ha error. In addiion, only passengers coming hrough he principal air and sea roues are inerviewed, while a subsanial number of EU8 workers come in hrough he smaller airpors. Overall, he survey is likely o undersae ne EU8 migraion. The Labor Force Survey (LFS) provides an esimae of he sock of EU8-born residens in he Unied Kingdom. Based on ha measure, beween May 004 and Ciizens of Cyprus and Mala were allowed o work in he U.K. prior o 004, so he change in he regime applied only o he remaining eigh counries (Esonia, Czech Republic, Hungary, Poland, Lihuania, Lavia, Slovenia, and Slovakia). They will be referred o as EU8 in he res of he ex.

6 4 Sepember 006, he number of EU8-born residens increased by 65,000 (Blanchflower e al, Table E), equivalen o an annual inflow of abou 0.35 percen of he labor force. However, only households residing a heir curren address for a leas six monhs are covered by he survey (people living in communiy esablishmens are no covered), so recen migrans and mos emporary workers are excluded. Naional Insurance (NI) numbers are issued o all employed and self-employed individuals for ax and benefi purposes. Beween April 004 and March 006, approximaely 380,000 numbers were issued o EU8 naionals. Similar o he WRS, his saisic measures gross flows (unlike he WRS, i includes self-employed people). Discussions of he macroeconomic effecs of immigraion have become a focus of policy and media aenion, following his surge in labor flows from he new member saes. The effecs on growh, wage and public finances are widely debaed, even hough he exising empirical evidence is sill limied. Saleheen and Shadforh (006) describe he characerisics of he recen EU8 migrans. Riley and Weale (006) esimae ha he inflow from he new member saes in has conribued 0. percen o GDP growh. Using regional daa, Gilpin e al. (006) and Blanchflower e al. (007) find ha immigraion is no relaed o he rise in unemploymen in 005/6. Mos of he exising heoreical and empirical lieraure on immigraion is based on parial equilibrium, saic models. 3 These model a bes, provide guidance only o he expeced shor run effecs on he labor marke. Saic analysis predics ha an increase in one facor of producion reduces is relaive price, herefore wages would fall afer an increase in labor supply (or unemploymen would rise in he presence of wage rigidiies). However, open economies adjus dynamically, and he long-erm effecs on he labor marke and he economy more generally are likely o differ from he shor-erm effecs. Using a general equilibrium dynamic model, his paper explores he likely evoluion of he economy afer a rise in accession-relaed immigraion. Given he change in regulaions governing labor mobiliy, pos-accession migraion represens a ruly exogenous shock o labor supply, which makes i an ideal case for economic analysis. The model has an overlapping-generaions srucure, calibraed o reflec he demographic profile of he Unied Kingdom. The following quesions are addressed. Wha is he impac of immigraion on economic growh, capial accumulaion, consumpion, public finance and he curren accoun? Do he shor-erm macroeconomic effecs of an immigraion shock differ from he long-run effecs? Can immigraion help miigae he adverse effecs of populaion aging? 3 A recen excepion is Barrell e al. (006). Tha sudy uses a general equilibrium macroeconomic model o illusrae he effecs of changing ne migraion on oupu and unemploymen in he U.K. and Germany.

7 5 II. METHODOLOGY The analysis is based on Mulimod, a dynamic general equilibrium model wih demographic feaure developed a he IMF. 4 Consumers are assumed o have a finie planning horizon, which affecs heir consumpion/savings decisions. The producion funcion is Cobb- Dougla wih capial and homogeneous labor as he facors of producion. Invesmen behavior is based on he Tobin s Q-heory invesmen growh acceleraes when he expeced marginal produc of capial is greaer han is cos. Perfec capial markes are assumed, allowing firms o borrow freely a he world ineres rae. Governmen spending on pensions depends on he share of he populaion above reiremen age (he pension paymen per reiree is fixed as a share of GDP). An endogenous ax rae ensures ha he raio of governmen deb o GDP converges o a arge level. Demographic developmens affec boh he supply and he demand sides of he economy. On he supply side, earnings are presumed o be a good indicaor of changes in relaive produciviy and labor supply ha occur over an individual s working life. The paern of earnings over ime is hump-shaped as workers accumulae experience, heir produciviy and earnings increase. Afer peaking in middle age, earnings gradually decline ino reiremen. The rae of paricipaion in he labor marke follows a similar paern. Changes in he age srucure of he populaion will hus affec he effecive labor supply. On he demand side, individuals are assumed o smooh consumpion based on heir anicipaed life-cycle income. Younger individuals end o be ne borrower since heir curren income is below heir permanen income. Middle-age individual whose relaive earning are a heir peak, save for reiremen. Therefore, changes in he age profile of he populaion will also have an impac on aggregae consumpion and savings. The age-earnings profiles for he Unied Kingdom are presened in Figure. They are esimaed using daa on acual hourly earnings from he Annual Survey of Hours and Earnings (ASHE) for he period The earnings are adjused for he labor force paricipaion raes of he differen age cohors (obained from he Labor Force Survey), and normalized relaive o he per capia earnings of he younges cohor. The earnings profile hus represens he average earnings per person Relaive Income Figure. Age-Earnings Disribuion Unied Kingdom, ` Age See Laxon e al. (998) for a descripion of Mulimod. Faruqee and Laxon (000) and Faruque (00) conain a deailed discussion of he feaures of he model wih demographics. Annex conains a shor descripion of he mos relevan equaions in he model.

8 6 (no per worker) wihin each age caegory. The esimaed earnings are shown a he approximae mid-poin of he age ranges used in he ASHE. The relaive earnings profiles have been relaively sable over he sample period. An exponenial funcion is fied hrough he age-earnings profiles o approximae he ime-series paern of a person s lifeime earnings: y b ( ) ( ) ( (, ) ( ) ) s b s b3 s s = a e + a e + a a e, where y is relaive labor income, is he curren year, and s is he year in which he individual enered he labor marke. The b parameers are esimaed by non-linear leas square using a grid search over he a parameers. 5 The esimaed values are he following: b=0.08, b=0., b3=0.09, all significan a he 0 percen level. The baseline scenario is calibraed o reflec he main feaures of he UK economy. The labor share of oupu is 69 percen. The seady sae capial-oupu raio is.3, similar he hisorical average. The rae of capial depreciaion is 5 percen. The rae of oal facor produciviy growh is assumed o be consan a.3 percen. The long-run real ineres rae is assumed o be 4 percen. The iniial level of he public deb is se a 40 percen of GDP, and he arge deb level is also se a 40 percen (so here are no ransiional dynamics if he pension sysem is kep in balance a all imes). Since he resuls are presened in deviaions from he baseline, hey are quie robus o changes in he calibraion parameers. An immigraion shock is inroduced ino he model as an increase in he number of young people in he economy. This is based on exising empirical evidence for he U.K., which suggess ha more han 80 percen of he new EU-8 immigrans are beween 8 and 34 years of age. Based on he laes Home Office Accession Monioring Repor (November 006), 43 percen of all WRS applicans were aged beween 8 and 4, and furher 39 percen were beween 5 and 34. Virually all are working 98 percen of applicaions for Naional Insurance Numbers made by accession counry naionals in he wo years afer accession were for employmen purposes. 6 The size of he immigraion shock is chosen o broadly mach he recen U.K. experience and academic esimaes of he oal migraion poenial from he new member counries. Specifically, i is assumed ha 00,000 young workers from he new member saes enered permanenly he U.K. labor marke in 005 and 006 respecively. 7 A gradual reducion in 5 For a more deailed explanaion of his esimaion, see Faruqee and Laxon (000). I liss he various parameer resricions ha need o be saisfied, such as a non-negaiviy consrain on earnings. 6 To preven welfare ourism, he U.K. has resriced access o social benefis for new EU-8 immigrans during he firs year of employmen. Tha may accoun for he fac ha pracically all immigrans hold jobs and he percenage receiving any social benefis is much smaller han ha of he naive populaion. 7 The assumed size of ne migraion during is broadly based on he LFS saisics. I is higher han suggesed by IPS (as discussed, his saisic may undersae acual ne migraion from he EU8), bu (coninued )

9 7 ne migraion is assumed over he nex en years as oher counries open heir labor markes and he number of poenial immigrans declines (see Figure ). Esimaes by Boeri e al. (005) and oher sugges ha he ne emigraion poenial of workers from he new member saes (including Bulgaria and Romania) is equivalen o abou 3 percen of heir respecive populaion wih mos of hem expeced o migrae wihin he firs en years afer accession. The oal ne immigraion o he U.K. assumed in his analysis amouns o abou a quarer of ha, reflecing he srong diversion of flows o he U.K. in he iniial years of accession. Esimaes of boh acual and poenial migraion are highly uncerain, so his scenario should be aken only as an illusraion and should no be inerpreed as a forecas. Figure. Assumed Ne Migraion Thousands III. SIMULATIONS The simulaions are performed in wo seps. Firs, a baseline scenario based on he official projeced populaion growh is esimaed. The scenario uses hisorical daa and he 004 Governmen Acuary Deparmen populaion projecions hrough 050. A gradual reurn o seady populaion growh is assumed hereafer (o ensure convergence in he model). Nex, an immigraion shock is inroduced as an exogenous increase of he number of young individuals in he economy. 8 The resuls are compared o he simulaion based on he laes significanly lower han he WRS and NI numbers (which measure gross inflows). The assumpions are made purely for illusraive purposes and do no represen a judgmen on he acual ne inflows. 8 I is appropriae o make his assumpion here since EU8 migraion represens a ruly exogenous shock, and is magniude was largely unanicipaed. Ne migraion from oher counries is relaively seady and predicable, and is incorporaed ino he populaion and labor force projecions made by he Office of Naional Saisics.

10 8 populaion projecions. The effecs of he shock on he main macroeconomic variables are presened in Figure 3 (expressed in percenage poin deviaion from a no-immigraion baseline). Immigraion booss invesmen growh and he capial sock. Following a posiive labor supply shock, he marginal produc of capial goes up, while he marginal produc of labor (and real wages) decline. Tha induces accumulaion of capial o he poin where is marginal produc and marginal cos (deermined by he world ineres rae) are equalized. As capial increase he marginal produc of labor increase reversing he iniial downward pressure on real wages. In seady sae, wage growh would be he same as before he immigraion shock (assuming consan produciviy). Figure 3. Macroeconomic Effecs of Immigraion (Deviaions from a no-immigraion baseline level) Real GDP Real GDP Real GDP per capia (percen change) Curren Accoun Balance and Savings CA Savings (in percen of GDP) Invesmen and Capial Sock Invesmen Capial sock (percen change) Consumpion - - Consumpion - Consumpion per capia (in percen of GDP) Pension Expendiure (in percen of GDP) Elderly Dependency Raio (percen of adul populaion) Boh oal and per capia oupu and consumpion increase in he medium erm. Toal oupu and consumpion are higher a all horizons as he populaion increases permanenly. Per capia oupu declines for he firs few years as he young immigrans are iniially less producive han he average person (remember he hump-shaped profile of he age-earnings

11 9 disribuion). 9 As hey accumulae experience, produciviy increases (above ha of he average worker), raising per capia oupu. In he medium erm, when immigrans reach he peak of heir earnings poenial, hey save a subsanial share of heir earnings. Therefore he increase in consumpion is less pronounced ha he increase of oupu. As immigrans sar o reire and consume heir asse per capia consumpion becomes higher han per capia income. In he very long run, per capia real GDP and consumpion reurn o baseline. The curren accoun iniially deerioraes and hen improves relaive o he baseline. The deerioraion in he iniial years mainly reflecs he rise in invesmen. The endency of young workers o be ne borrowers (or, in he case of immigran o send remiances abroad) also conribues o he iniial deerioraion of he curren accoun. Laer, when he immigrans savings rae rise he curren accoun improves. As he EU8 workers reach reiremen age, he curren accoun reurns o is iniial value. As he elderly dependency raio decline pension expendiure in percen of GDP declines. Tha would allow he governmen o eiher reduce deb or cu axes. The simulaion assumes a cu in axes which booss disposable income and consumpion. This resul holds as long as he average produciviy of immigrans is similar o ha of he oal populaion, and heir propensiy o use publicly-provided benefis is also similar o he average. So far, very few EU8 immigrans have applied for social benefi bu one may expec ha once hey become eligible for all benefi hey will use hem wih roughly he same inensiy as he naive populaion. Can immigraion solve he fiscal problems relaed o aging? The EU8 migraion wave represens a one-off posiive demographic shock. Therefore, he posiive fiscal effecs would be emporary since he immigrans would also age. The ne raes of migraion necessary o offse he effec of aging in he long run are very high, so permanen produciviy improvemens and a well-designed pension sysem will coninue o be imporan for ensuring fiscal susainabiliy. The resuls are sensiive o some of he assumpions in he model. Firs, immigrans are assumed o have exacly he same skill disribuion and produciviy as he naive workers. While his is broadly rue for he sock of immigrans in he U.K. (see Sal and Millar, 006), he secoral disribuion of immigrans from he new member saes sugges ha hey may be 9 This resul assumes ha he new immigrans have he same labor paricipaion profile, and he same produciviy as an average person of ha age. In fac, immigrans may have higher labor paricipaion rae (a leas iniially). Taking his ino accoun could reduce he magniude of he iniial mild decline in per capia oupu. However, i is reasonable o assume ha, over ime, he labor force paricipaion rae of he immigrans will converge o ha of he naive populaion (ha has been rue for he older immigran cohor see Sal and Millar (006)).

12 0 less skilled han he average worker. 0 Tha would reduce he esimaed effecs. Second, he model assumes perfec flexibiliy in he labor and goods markes: here is no unemploymen and economic oupu is coninuously a poenial. Relaxaion of hese assumpions could influence he adjusmen dynamics. For example, he minimum wage ses a floor on he adjusmen of real wages in some unskilled occupaions. Tha could lead o a rise in unemploymen. Third, he producion funcion is assumed o be Cobb-Douglas. Changing he form of he producion funcion (for example allowing for worker complimenariy) could produce higher posiive effecs. Finally, he magniude of he macroeconomics effecs is proporional o assumed size of he migraion shock. If he model is exended o include a fixed facor of producion (land), he effecs would vary wih he size of he shock. IV. CONCLUSIONS The resuls of his analysis sugges ha he macroeconomic effecs of immigraion from he EU8 are broadly favorable. Boh oal and per capia oupu and consumpion increase in he medium erm (as immigrans become more producive and invesmen growh acceleraes). The curren accoun iniially deeriorae reflecing he higher invesmen growh. Thereafer, he curren accoun improves relaive o he no-immigraion baseline, before reurning o seady sae. Any iniial downward pressure on wages would be emporary. As capial accumulaes o mee he needs of he new worker wages would grow in line wih he rising marginal produciviy of labor. The raio of fiscal expendiure o GDP falls as he dependency raio decline allowing he governmen o reduce deb or emporarily cu axes. 0 However, i is possible ha many new immigrans ake jobs ha do no correspond fully o heir educaional qualificaions or innae skills. Over ime, hey may experience upward occupaional mobiliy and become more producive as hey accumulae locaion-specific human capial (see Barre (007) for evidence of ha for Ireland). Boh academic sudies and he press ofen sugges ha immigrans ake jobs ha are no aracive for he locals (or would no even exis if i was no for he immigrans). Tha increases he efficiency of he labor marke and may induce an increase in he labor force paricipaion of he local workers. Tha effec could be modeled as labor complimenariy in he producion funcion. Borjas (00) suggess anoher channel hrough which immigrans may improve economic efficiency by locaing in regions wih he highes marginal produc of labor.

13 APPENDIX I. DESCRIPTION OF THE MODEL Demographics are modeled hrough cohor-specific birh raes and ime-varying populaion growh. Life-cycle dynamics are generaed using age-earnings profiles. Populaion Growh The basic equaion for populaion dynamics is given by: N ( = n( = b( p(, N( () where N is he populaion level and n is he growh rae equal o he difference beween he "birh" rae b and he deah rae p; dos denoe derivaives wih respec o ime. Inegraingup equaion () over ime yields an expression (up o a consan of inegraion) for he size of he oal populaion a any momen in ime: N( = e - n(v)dv - = e. () Equaion () shows ha populaion size evolves according o he accumulaion of pas changes o is growh rae--i.e., he pas difference beween birh and deah rae which deermines he size of he curren populaion as growh facor imes he size of he iniial populaion. Dependency Raio To faciliae analysis of demographic issue measures ha characerize he age disribuion of he populaion are defined. Noe ha he size of he populaion, which unil now has been defined in relaive erms vis-à-vis a reference populaion, can also be defined in aggregaive erms as he sum of exising individuals across all generaions (indexed by s): - - p( s) N( = n( ds = b( s) N( s) e ds. (3) This appendix follows Faruqee (00), which conains a more deailed presenaion. Laxon e al. (998) has a full descripion of Mulimod. The birh rae b is defined as he arrival rae of new aduls and N is he adul populaion. Wih consan birh and deah rae he number of agens belonging o a generaion s a ime s (i.e., a he ime hey are born), as a proporion of he conemporaneous populaion, is given by n ( s) = bn( s) ; he number survivors from ha cohor a ime is hen given by [b(v)- p(v)]dv -p(- ( = bn( s) e, where p is he common deah rae facing all agens. n s)

14 Similarly, an elderly dependency raio is defined as he proporion of oal populaion above a cerain age: j( n( φ( = ds ; 0< φ <, (4) - N( where φ measures he relaive cohor size, as a share of he oal populaion, of individuals older han some hreshold age level--indexed by j(. Assuming ha his age definiion does no change, he index j( moves wih ime o include new dependen who have jus reached he hreshold age a each momen in ime (i.e., j ( = ). In he case where birh raes are consan, i can be shown ha he dependency raio φ would also be consan. 3 Oherwise, he dependency raio evolves over ime according o he ime derivaive of equaion (4) above: n( j(, φ ( ) = - [ p + n( ] φ(. (5) N( A each momen in ime, he change in he dependency raio is deermined by he relaive size of new dependens aaining he hreshold age, less he proporion of he elderly pφ who die each period and accouning for growh in he populaion base nφ i.e., he scaling variable. Age-earnings Profiles To incorporae he life-cycle dimension in he analysi age-earnings profiles are consruced. Labor income iniially rises wih age and experience, before evenually declining wih reiremen. The individual labor inpu (which is inelasically supplied) varies in effecive erms across agens from differen generaions. The effecive labor supply (measured in efficiency unis) is assumed o have he following ime series paern: α ( ) [ a e ] ( s) α ( s) b s + a e + ( a a ) e μ( ; s,a >0 > a, α > α > 0. 3 l( = ) (6) The exponenial form is chosen for analyical racabiliy, and he specific choice of hree inflecion poins is chosen o provide reasonable fi of he daa. The las erm μ in = b 3 Wih seady populaion growh, he dependency raio would sele down o is long-run value: φ e - Δ, where Δ - j( and b are consans.

15 3 equaion (6), which is no age-specific, represens general labor produciviy growh ha depends on economy-wide consideraions (e.g., echnology). 4 Assuming equaion (6) summarizes he age-specific feaures underlying differences in labor earnings across cohor he individual labor income can be expressed as he produc of he aggregae wage rae and individual labor supply: y ( = w( l(. (7) Thu for a given wage rae, he relaionship beween age and earnings will exhibi a similar (hump-shaped) ime profile as beween age and labor supply; his profile has been documened empirically by age-earnings disribuions in a broad cross-secion of counries. 5 Aggregaing over all individuals renders oal labor income: Y ( = - w( l( n( ds (8) = w( L(, where L is aggregae labor inpu, also measured in efficiency unis. Using he definiion of individual labor supply in (6), one can also wrie L as he sum of wo componens L + L -- represening aggregaes of he facors underlying he ageearnings funcion in (7). 6 In he presence of demographic dynamic he law of moion governing he behavior of aggregae labor inpu can hen be wrien as: L ( = L ( + L (, = b( N( e μ + ( μ - p -α ) L ( + ( μ - p -α ) L (. (9) Inuiively, equaion (9) defines changes in L as depending on he effecive labor supply of new enrans o he workforce and he deah and produciviy changes among exising workers. μ = e μ ( is assumed, where he aggregae produciviy index a 4 Henceforh, consan produciviy growh: ime zero is normalized o uniy. 5 See for example Jappelli and Pagano (989). Reorganizing erms implici in (7), one could also express labor income as a funcion of a cohor-specific wage w( --commensurae wih individual produciviy. 6 α k ( s) μ Specifically, Lk ( = - lk ( n( ds, where lk ( = ak e e L μ ( = a b( N( e ( α p) L (. k k k + k for k=, and, hu

16 4 Consumpion Agens are assumed o maximize expeced uiliy over heir lifeimes subjec o a dynamic budge consrain. Specifically, he evoluion of financial asses a ( for an individual or household is deermined by is saving, equal o he difference beween income and consumpion: a ( = ( r + p) a( + y( -τ ( - c( (0) where r is he ineres rae, y - τ is disposable labor income, and c is consumpion, all expressed in real erms (unis of consumpion). 7 Ignoring any capial marke imperfecion consumpion is based on an agen's permanen income. Explicily, opimal consumpion (wih log uiliy) is given by: 8 c ( = ( θ + p) [ a( + h( ]. () where θ is he rae of ime preference and h( is a measure of an agen's human wealh-- equal o he presen value of fuure labor income. 9 Because labor income and human wealh evenually decline over a person's lifeime wih reiremen, he saving behavior implied by equaion () suggess ha agens evenually build up financial wealh a( o ensure a cerain level of reiremen consumpion (Faruqee and Laxon, 000). Aggregaing again, oal consumpion as a funcion of (financial and human) wealh can be expressed as follows: C ( = ( θ + p) [ A( + H ( ]. () where A is aggregae financial wealh and H is aggregae human wealh. Financial wealh consiss of domesic equiy and bond holdings and, in he open-economy case, holdings of 7 The erm pw( in he dynamic budge consrain reflecs he efficien operaion of he life insurance or annuiies marke. 8 For he simulaion consan relaive risk aversion (CRRA) uiliy is assumed (see Blanchard, 985). 9 For a fixed real ineres rae, individual human wealh can be wrien as: -( r+p)( v- h( [ y( v) -τ ( v)] e dv. Correspondingly, he dynamic equaion for an individual s human wealh is given by: h ( = ( r + p) h( -[ y( -τ ( ]. Given he shape of he age-earnings profile and assuming proporional labor income axes --i.e., τ ( = τy(, noe ha human wealh can also be wrien as: h = h ( + h ( ), where ( = k=,. ( h k - ( τ ) w( v)l ( v) e k ( r+p)( v- dv for

17 5 ne foreign asses; A/K+B+F. As for aggregae human wealh--reflecing he presen value of economy-wide labor income sream is behavior can be characerized as follows: d H ( = h s N s ds - d (, ) (, ) = h(, b( N( + r( H ( - [ Y ( - T ( ], (3) where T is oal labor axes and Y - T is oal disposable labor income. Equaion (3) shows ha he incremenal change in he sock of aggregae human wealh is influenced by he addiional human wealh of he newes generaion. 0 Equaions (9) and (3) summarize he role of demographic dynamics for economic behavior hrough boh supply-side and demand-side channels. On he supply side, changes in he demographic profile of he economy impac he supply of labor, given he differences across age groups summarized in he age-earnings profile. On he demand side, aggregae consumpion and saving behavior will also be affeced hrough aggregae human wealh dynamics in he face of life-cycle income and demographic change. Pension Sysem A pension sysem can be inroduced ino he framework as follows. Consider firs he simple case of a lump-sum ransfer scheme: α( ; s > r( = + β ( ; s j( j( (4) where he ransfers r( paid or received by individual depending on heir age. Younger generaions ( s > j( ) pay ino he sysem while older agens or pensioners ( s j( ) receive a benefi. Redefining he scheme in erms of payroll ax financing is sraighforward. In ha case, individual conribuions would be proporional o wage income: α ( = τ y(. For any ransfer scheme, a full-financing condiion can be wrien as follows: ss r( N( ds = 0. (5) 0 Given he shape of he labor income profile, he evoluion of individual human wealh for he newes generaion h(, a each momen in ime is: h μ (, = ( r + p + α ) h (, + ( r + p + α ) h (, - ( τ ) w(e

18 6 This general condiion mus hold for he ransfer scheme o be deemed fully financed (i.e., no unfunded liabiliies from he ransfer sysem). Full-financing in he specific case of lumpsum ransfer is hen given by: β ( α( φ( =. (6) φ( Equaion (6) shows he well known condiion ha he benefi-o-conribuion raio mus equal he suppor raio, defined as he number of working-age persons relaive o elderly dependens. If full-financing is no saisfied, here would exis a financing gap (posiive or negaive) reflecing he degree of over- or underfunding. In he case of a shorfall, he social securiy defici would have o be closed hrough increase in ax revenues or governmen borrowing. In he case of payroll axe individual conribuions would be age-dependen, deermined by social securiy axes paid on individual labor income. In he simulaion individual social securiy benefis (in percen of GDP) are aken as exogenous policy parameers.

19 7 REFERENCES Barrell, R., C. Guillemineau, and I. Liadze, 006, Migraion in Europe, Naional Insiue Economic Review, Vol. 98 (Ocober), pp Barre, A. and D. Duffy, 007, Immigran Year of Arrival and Occupaional Aainmen: Looking for Evidence of Assimilaion in Ireland, ESRI, Ireland, mimeo. Blanchard, O., 985, Deb, Deficis and Finie Horizon Journal of Poliical Economy, Vol. 93, pp Blanchflower, D., J. Saleheen, and C. Shadforh, 007, The Impac of he Recen Migraion from Easern Europe on he UK Economy, IZA Discussion Paper No. 65. Boeri, T. and H. Brucker, 005, Migraion, Co-ordinaion Failures and EU Enlargemen, IZA Discussion Paper 600. Borja G., 00, Does Immigraion Grease he Wheels of he Labor Marke? Brookings Papers on Economic Aciviy, Brookings Insiuion, pp Faruqee, H. and D. Laxon, 000, "Life-Cycle Dynasie Saving: Implicaions for Closed and Small Open Economie" IMF Working Paper 00/6 (Washingon: Inernaional Moneary Fund). Faruqee, H., 00, Populaion Aging and Is Macroeconomic Implicaions: A Framework for Analysi IMF Working Paper 0/6. Gilpin, N., Heny, M., Lemo S., Pores J. and C. Bullen, 006, The Impac of Free Movemen of Workers from Cenral and Easern Europe on he UK Labor Marke, Working Paper No 9, Deparmen of Work and Pension London, 006 Jappelli, T. and M. Pagano, 989, Consumpion and Capial Marke Imperfecions An Inernaional Comparison, American Economic Review, Vol. 79, pp Laxon, D., and oher 998, MULTIMOD Mark III: The Core Dynamic and Seady-Sae Model IMF Occasional Paper No. 64 (Washingon: Inernaional Moneary Fund). Riley, R. and M. Weale, 006, Immigraion and is Effec Naional Insiue Economic Review, Vol. 98 (Ocober), pp Saleheen, J. and C. Shadforh, 006, The Characerisics of Immigrans: A View from he Daa, Bank of England Quarerly Bullein, 46(4), pp Sal, J. and J. Millar, 006, Foreign Labor in he U.K.: Curren Paerns and Trend Office of Naional Saisic Labor Marke Trend Ocober 006.