Global Aging Pressures: Impact of Fiscal Adjustment, Policy Cooperation, and Structural Reforms

Transcription

1 WP/07/196 Global Aging Pressures: Impac of Fiscal Adjusmen, Policy Cooperaion, and Srucural Reforms Dennis Boman and Manmohan Kumar

2

3 2007 Inernaional Moneary Fund WP/07/196 IMF Working Paper Fiscal Affairs Deparmen Global Aging Pressures: Impac of Fiscal Adjusmen, Policy Cooperaion, and Srucural Reforms Prepared by Dennis Boman and Manmohan S. Kumar 1 Augus 2007 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. Demographic pressures will maerialize in many economies over he nex few decades. We examine he macroeconomic impac of alernaive fiscal adjusmen and srucural reform sraegies o address hese global aging pressures using he IMF s Global Fiscal Model (GFM). The resuls sugges subsanial spillover effecs of aging hrough inernaional financial channels. To mainain susainabiliy, fiscal adjusmen needs o be broad-based, while avoiding increases in direc axes. There are subsanial benefis from fiscal cooperaion, while negaive growh effecs can be offse by complemenary srucural reforms in produc and labor markes wih he benefis accruing early and o all incomegroups. JEL Classificaion Numbers: E62, F41, F42, H30, H55, H62 Keywords: Global aging, fiscal adjusmen, fiscal cooperaion, Lisbon Agenda, IMF s Global Fiscal Mode Auhor s Address: dboman@imf.org, mkumar@imf.org 1 We are graeful o Teresa Ter-Minassian, Bob Traa, Mark De Broeck, Sephan Danninger, Jorg Decressin, Luc Everaer, seminar paricipans, and in paricular our discussan Robin Boadway, a FAD s Academic Panel meeing on April 18 in Washingon DC, and paricipans a he conference Fiscal Policy Challenges in Europe organized by he German Federal Minisry of Finance and he Cener for European Economic Research (ZEW), Berlin, March 22 and 23 for many helpful commens and suggesions. The views expressed in he paper are hose of he auhors and do no necessarily represen hose of he IMF or IMF policy.

4 2 CONTENTS PAGE Inroducion... 4 I. Demographic Trends and Fiscal pressures... 5 II. Key Feaures of he IMF s Global Fiscal Model... 7 III. Calibraing The Model... 9 IV. Deb Dynamics And Fiscal Adjusmen In Germany A. Deb Dynamics: Populaion Aging in Germany and in Trading Parners B. Fiscal Adjusmen in Germany V. Global Aging and Fiscal Adjusmen: Inernaional Cooperaion? VI. Reform Complemenary? A. Increasing Labor Paricipaion B. Higher R&D Spending C. Higher Produc Marke Compeiion D. Implemenaion of Lisbon Agenda in Germany and he Euro Area VII. Summary and Conclusions Appendix A. Specificaion of GFM References Tables 1. Key Macroeconomic and Fiscal Variables in he Iniial Seady Sae 1/ Germany: Fiscal Variables and Key Parameer Values Germany: Alernaive Fiscal Adjusmen Sraegies: / Adjusmen Sraegies in he euro area and he U.S.: Ne Presen Value of GDP of Fiscal Cooperaion 1/ Sensiiviy Analysis of Ne Presen Value of GDP of Fiscal Cooperaion 1/...33 Figures 1. Germany: Primary Expendiure Effecs on German Deb Dynamics of Populaion Aging in Germany 1/ : The Effecs of Alernaive Aging Cos Projecions on Deb Dynamics : Effecs on German Deb Dynamics of Global Populaion Aging 1/ Macroeconomic Effecs of Alernaive Fiscal Adjusmen Sraegies 1/ Deb Dynamics in Germany Wih Fiscal Adjusmen : Macroeconomic Effecs on Germany of Adjusmen Package 1/ Effecs on Real GDP in Germany of Fiscal Cooperaion 1/...29

5 3 8. con. Effecs on Real GDP in he euro area of Fiscal Cooperaion con. Effecs on Real GDP in he Unied Saes of Fiscal Cooperaion con. Effecs on Real GDP of Sequenial Sop-and-Go Policies 1/ : Macroeconomic Effecs of Fiscal Cooperaion on Germany 1/ : Fiscal Cooperaion and Higher Labor Paricipaion in Germany 1/ : Fiscal Cooperaion and Higher R&D Spending in Germany 1/ : Fiscal Cooperaion and Produc Marke Reform in Germany 1/ : Fiscal Cooperaion and he Lisbon Agenda: Effecs on Germany 1/ : Fiscal Cooperaion and he Lisbon Agenda: Effecs on he euro area 1/ : Fiscal Cooperaion and Delayed and Parial Implemenaion of he : Fiscal Cooperaion and Delayed and Parial Implemenaion of he...48

6 4 INTRODUCTION To address global aging pressures, his paper underakes a rigorous analysis of he macroeconomic effecs of fiscal adjusmen in Europe pursued in combinaion wih ax, labor and produc marke reforms using he IMF s Global Fiscal Model (GFM). We calibrae he model o a large open European economy (Germany), as well o he res of he euro area, Unied Saes, and he res of he world. GFM is a mulicounry dynamic general equilibrium model specifically designed o explore fiscal policy issues, wih srong micro foundaions, a wide menu of axes, and a sylized financial secor block. A key feaure of he model is ha Ricardian equivalence does no hold owing o overlapping generaions in he spiri of Blanchard-Weil, limied access of some consumers o financial markes, and disorionary axaion. Wihin his framework, he paper explores he following four issues: Firs, European counries are facing subsanial medium- and long-erm fiscal pressures from populaion aging. Even wih he imporan pension reforms and recen improvemens in srucural deficis ha have been underaken in a number of counries, mainaining deb susainabiliy will require subsanial fiscal adjusmen (including furher enilemen reforms), and we assess he relaive pros and cons of prefunding fuure healh and pension spending. Such fiscal adjusmen could be implemened in a variey of ways, and we quanify he effecs of alernaive revenue and/or expendiure based consolidaions on oupu, invesmen, and employmen. A second issue recognizes ha fiscal pressures from aging are no limied o any one counry or region, bu insead will be occurring in many advanced and emerging marke counries a broadly he same ime, albei wih differen magniudes and saring fiscal posiions. We explore wha his implies for he efficacy of fiscal adjusmen, aking ino accoun he effecs on inernaional capial flows. Imporanly, we address he role policy cooperaion can play in his environmen. Third, as fiscal consolidaions are generally conracionary in he shor erm, a key quesion arises, from boh an economic and a poliical economy perspecive, as o wheher complimenary policies exis o reduce he adverse effecs and, herefore, help he poliical feasibiliy of reform. One area of complimenary measures, cenral in he renewed Lisbon Sraegy, is srucural reform, in paricular increasing labor paricipaion (parly hrough ax reform, in paricular moving from direc o indirec axaion), produc marke liberalizaion, and enhancing incenives for innovaion and produciviy growh. We assess o wha exen hese policies can help ameliorae he near-erm conracionary effecs of fiscal consolidaion, for boh oupu and consumpion. Fourh, we analyze he sensiiviy of our resuls o he behavioral assumpions underlying he response of firms and consumers o changes in fiscal policy and srucural reform. Specifically, in he non-ricardian seing of GFM, we explore he

7 5 effecs of changes in hree of he fundamenal deerminans of he effecs of fiscal policy: (i) he planning horizon of consumers; (ii) he fracion of liquidiy-consrained consumers; and (iii) he elasiciy of labor supply. Furhermore, as GFM is rooed in consumer and producer opimizaion, we examine he exen o which he ineremporal elasiciy of subsiuion deermines he impac of fiscal policy and srucural reform. The res of he paper is organized as follows: Secion II noes he medium and longer-erm fiscal pressures facing he European economies, paricularly resuling from he projeced demographic challenges, and he magniude of hese challenges elsewhere; Secion III discusses key feaures of he IMF s Global Fiscal Model, in paricular he exension of he New Open Economy Macroeconomics framework o incorporae sufficien non-ricardian feaures o allow for an analysis of he effecs of fiscal policy; an annex provides echnical deails of he specificaion of households, firms, governmen and exernal secor; Secion IV calibraes he model for a large open European economy like Germany, as well as he res of he euro area, and he Unied Saes; Secion V provides an analysis of he impac of a range of ax and expendiure measures, boh individually and as par of comprehensive sraegy in Germany for aaining deb susainabiliy, while he following secion examines he implicaions of cooperaive fiscal consolidaion in parner counries facing similar aging pressures. Secion VII examines he effec of implemening he Lisbon Agenda, and a las secion provides summary and conclusions. I. DEMOGRAPHIC TRENDS AND FISCAL PRESSURES The budgeary challenges facing he EU counries arising from he demographic rends have been analyzed exensively by he European Commission, counry auhoriies, as well as a wide range of observers (see EC 2006a, EC 2006b; Hauner e al 2007). The assessmen of he budgeary challenges is based on he projecions by Eurosa ha indicae on average a doubling of he old-age dependency raio (populaion older han 65 relaive o working-age populaion) from 2005 o 2050 in he EU25 counries. Over his period, he modal age-cohors move from mid-30s o lae 50s. These changes are projeced o exer significan upward pressure on age-relaed expendiures, primarily relaing o pensions and healh care, by he public secor: for he EU25 as a whole, he average increase (weighed) in age relaed expendiures is projeced a 3.4 percen of GDP; for he EU15, i is 3.7 percen of GDP and for he new EU member saes (EU10) i is 0.2 percen of GDP. However, amongs he EU10, Poland has implemened a major pension reform ha implies a marked decline in is pension expendiures amouning o 5.9 percen of GDP. Excluding Poland, he average age relaed expendiures for EU10 are likely o increase by 4½ percen of GDP (see EC 2006b). If one considers separaely he age relaed expendiures emanaing from pensions, and from healh and long-erm care, he following picure emerges: pension expendiures are projeced o increase by 2.3 percen of GDP on average in EU15 and by 0.3 percen of GDP in EU10 (excluding Poland he increase amouns o 4.8 percen of GDP). Healh and long-erm care

8 6 spending is envisaged o increase by 2.3 percen of GDP in EU15 and by 1½ percen of GDP in EU10. The base line scenario assumes ha he increase in life-expecancy will lead o some posponemen of he need for addiional care. The healh care projecion, somewha conservaively, assumes an elasiciy of demand higher han uniy (1.1) in he shor-erm, bu his is expeced o gradually decline o uniy over he projecion period. There is a considerable degree of uncerainy relaing o hese projecions, no leas arising from he impac of echnology and he evoluion of healh care coss, as well as demographics. Moreover, he impac of aging on he poenial growh of he economy which would have an imporan bearing on susainabiliy is also subjec o wide confidence inervals. These uncerainies are amply illusraed for each of he Euro area counries, and Germany is no excepion. In he case of Germany, according o EU esimaes, he projeced increase in age relaed spending is somewha below he EU average, rising by abou 2 ¾ percen of GDP beween 2004 and This is due o an increase in pension expendiures of around 1.7, which reflecs he far-reaching reforms ha have been enaced (including Agenda 2010 and Harz reforms). The increase in healh-care expendiure is projeced o be 1.2 percen of GDP, lower han he EU average (see EC 2006b). Noneheless, even hese low esimaes, given Germany s curren budgeary posiion lead o a susainabiliy gap of 3½ percen o 4 percen of GDP. However, as he discussion in Secion IV below indicaes, here are oher esimaes of aging coss for Germany ha are subsanially higher. The projeced increases in age relaed spending and he susainabiliy gaps in oher major counries and regions are also sriking: in he case of he Unied Saes, in paricular, he increase in age-relaed spending over is projeced a 5.9 percen of GDP. Given he srucural primary balance of around -1.8 percen (2005), he U.S. faces a susainabiliy gap of around 7 percen of GDP (OECD, 2001; and Hauner, Leigh, and Skaarup, 2007). 3 There are marked aging relaed budgeary pressures also in Japan, where he iniial condiions, boh in erms of he large primary defici and high deb levels, also warran subsanial adjusmen (Hauner, Leigh, and Skaarup, 2007, and Boman, Edison, and N Diaye, 2007). 4 I is worh noing ha he aging issue is of course no limied o indusrial counries: here are marked pressures in many major emerging markes also. Among he larges nine emerging markes, 5 while he populaion in some is projeced o grow subsanially over he nex four o 3 The esimaed adjusmen need is very sensiive o he assumpions regarding he ineres rae-growh differenial. In he case of he Unied Saes, a 1 percen ineres rae/growh differenial suggess a primary gap of around 7.2 percen, wih he age relaed spending accouning for he bulk of ha. 4 A 1 percen ineres rae/growh differenial suggess a primary gap of around 5½ percen, wih aging coss accouning for around a hird of his, and he res due o he high primary defici and deb. 5 These include Argenina, Brazil, China, India, Indonesia, Korea, Mexico, Russia, and Turkey.

9 7 five decades, he 65+ par of he populaions is projeced o increase significanly in all counries (beween 150 and 400 percen), and he old-age dependency raio is expeced o, on average, riple by 2050 (see Kumar and Skaarup, 2007). (Korea faces he seepes increase (almos a five-fold increase, albei from a favorable saring period wih a sizable srucural surplus), bu here are also significan pressures in China and Russia). The projeced increase of he old-age dependency raio in hese counries is also markedly higher han in he G-7 counries, as are he projeced age-relaed expendiures (Kumar and Skaarup op. ci). II. KEY FEATURES OF THE IMF S GLOBAL FISCAL MODEL The IMF's Global Fiscal Model (GFM) exends he NOEM framework o incorporae sufficien non-ricardianness o allow for an analysis of he effecs of fiscal policy and of inerdependence. 6 The radiional New Open Economy Models (NOEM) do no depar from he Ricardian equivalence hypohesis enough o allow deailed consideraion of fiscal policy issues. 7 Insead, since hese models feaure a represenaive agen framework wih lump-sum axaion, he analysis is resriced o he effecs of balanced budge fiscal policies. If he Ricardian equivalence hypohesis holds fully, many of he crucial fiscal policy quesions posed in his paper and in he real world would be virually irrelevan. Generally speaking, complee Ricardian equivalence, on which here is scan empirical evidence, will hold in case consumers are homogenous and have an infinie planning horizon, if axaion is lump sum, if access o financial markes by all agens is complee, and if governmen deb is riskless. In such a seing, emporary changes in policy ha increases governmen deb will affec he composiion of naional saving, bu no is level. Any increase in he governmen defici will be mached by higher privae savings as agens anicipae higher fuure ax conribuions, wih no effec on ineres raes, consumpion, invesmen incenives, or oupu. There are hree key reasons why full Ricardian equivalence does no hold in GFM. Firs, he model feaures overlapping generaions in he spiri of Blanchard-Weil. The use of overlapping generaions allows he assumpion of Ricardian equivalence o be relaxed, implying ha governmen deb is perceived as ne wealh. Essenially, consumers have a shor, and more realisic, planning horizon, which implies ha even emporary changes in fiscal policy affec heir incenives o consume and work as hey discoun any fuure fiscal policy reacion. Second, GFM incorporaes he assumpion ha some consumers do no have 6 GFM is described in more deail in Boman and ohers (2006). 7 See Obsfeld and Rogoff (1995, 1996), Bes and Devereux (2001), Caselli (2001), Corsei and Peseni (2001) and Ganelli (2003a). In a recen paper, Erceg, Guerrieri, and Gus (2005) add rule-of-humb consumers o a model based on he represenaive agen paradigm and hen use he model o sudy he effecs of recen U.S. fiscal deficis on he curren accoun defici. No surprisingly, hey find much smaller effecs han in models ha allow for he possibiliy ha permanen increases in governmen deb can have permanen consequences for he sock of ne foreign liabiliies and he world real ineres rae.

10 8 sufficien access o financial markes o smooh heir consumpion over ime. This is consisen wih evidence ha even in he advanced economies up o a hird of he consumers are liquidiy consrained. Liquidiy-consrained agens consume heir enire disposable income every period, and, herefore, any change in fiscal policy ha affecs his disposable income will have real effecs. Third, GFM allows labor supply and capial accumulaion o be endogenous and respond o changes in incenives relaed o he afer-ax real wage or he afer-ax rae of reurn of capial. This in urn allows he model o incorporae he assumpion of disorionary axes, and analyze he consequences of changes in hese axes. In addiion, one furher difference beween radiional NOEM models and GFM is he absence of nominal rigidiies in he laer. In he curren seup, i is assumed ha wages and prices are fully flexible. This assumpion implies ha he cenral bank follows money argeing, which limis he analysis of he ineracion beween moneary and fiscal policy. Also, shor-erm mulipliers will be smaller han is he case for models wih nominal rigidiies. In his conex, i should also be noed ha capial mobiliy in GFM is perfec implying ha ineres raes are se in world markes. As a resul, especially for small open economies, he crowding-ou effecs of governmen deb via higher ineres raes will end o be smaller han would be he case if here were impedimens o capial flows and inernaional rade. These feaures noneheless provide a useful benchmark for he analysis, especially regarding he medium- and long-erm effecs of fiscal policy. Wih regard o inerdependence, in he version of GFM used in his paper, he world consiss of four blocks (Germany, res of euro area, Unied Saes, and he res of he world). Assuming ha all consumers in each of he regions face idenical survival probabiliies, he relaive size of he populaions remain consan and his essenially fixes he relaive size of he home economy (Germany). NOEM models have been exended by ohers o allow for an analysis of fiscal policy issues. An overlapping generaions seing has been brough ino NOEM framework by Ghironi (2003a and 2003b), and by Ganelli (2003a and 2003b). 8 The former does no consider he effecs of governmen deb, bu shows ha an overlapping generaions srucure following Blanchard (1985) and Weil (1989) ensures he exisence of a well-defined seady sae for ne foreign asse holdings (for an early analysis of his, see Buier, 1981). Ghironi, Iscan, and Rebucci (2005) describe how differences in agens discoun raes across counries gives rise o nonzero ne foreign asse posiions in he long run. 8 See Frenkel and Razin (1992) for a diagrammaic exposiion of a wo-counry overlapping-generaions model wihou disorionary axaion.

11 9 Ganelli (2003b) is he firs aemp o analyze alernaive fiscal policies in a NOEM model wih finie lives. Apar from including endogenous labor supply and liquidiy-consrained consumers, GFM exends his approach in four oher major direcions: The uiliy funcion is less resricive, permiing he analysis of alernaive values for he ineremporal elasiciy of subsiuion. This parameer affecs he sensiiviy of consumers o changes in he real ineres rae. As shown below, i has imporan implicaions for an assessmen of he impac of fiscal policy. A he same ime, he producion srucure is exended o include endogenous capial formaion, which provides an addiional channel hrough which governmen deb can poenially crowd ou economic aciviy and allows for he consideraion of corporae and personal income axaion. In GFM, invesmen is driven by a Tobin s Q relaionship, wih firms responding sluggishly o differences beween he fuure discouned value of profis and he marke value of he capial sock. In addiion, he supply of labor is made endogenous and consequenly labor income axes will be disorionary. The model feaures boh raded and nonraded goods, which allows us o consider he erms of rade effecs of changes in fiscal policy and he implicaions of various degrees in home bias in eiher privae and, especially, governmen consumpion. Compared o oher fiscal models, GFM feaures a richer menu of axaion. The axes included are social securiy conribuions paid by workers and employers, a corporae income ax levied on accouning profis of firms, and a personal income ax levied on labor income, accouning profis, and ineres income (on governmen bonds and ne foreign asses), and a value added ax. Excep for income axaion, each of hese axes has a single, albei differen, marginal rae, which coincides wih he average ax rae. GFM also has a sylized financial secor block, wih wo kinds of asses, namely governmen deb (which can be raded inernaionally) and equiy (which is held domesically). Changes in he ousanding sock of deb have direc implicaions for long-erm ineres raes, which as a resul of he perfec capial mobiliy assumpion are shared inernaionally. III. CALIBRATING THE MODEL The model is parameerized o reflec key macroeconomic feaures of Germany, as well as he res of he euro area, he Unied Saes, and res of he world (Tables 1 and 2). In paricular, he raios of consumpion, invesmen, wage income, and income from capial relaive o GDP are se o heir values in Similarly, key fiscal variables revenue-o- GDP raios from axaion of corporae and personal income, from consumpion ax, and from social securiy conribuions by workers and employers, as well as curren governmen spending have been calibraed o he fiscal srucure of Germany, and ha of he oher

12 10 regions. We presen he macroeconomic effecs of he simulaions relaive o he baseline, which is he iniial calibraed seady sae. Table 1. Key Macroeconomic and Fiscal Variables in he Iniial Seady Sae 1/ (percen of GDP) Germany Euro Area (excl. Germany) Unied Saes Res of he World Size (percen of world GDP) Expendiure Raios Privae consumpion Governmen expendiure 2/ Invesmen Expors Facor Incomes Capial Labor Governmen Deb Trade Flow Marix Toal expors o: Germany Euro Area (excl. Germany) Unied Saes Res of he World / Naional expendiure accouns a marke prices 2/ Ne of governmen ransfers Table 2. Germany: Fiscal Variables and Key Parameer Values Rule-of-Thumb Versus Forward-Looking Consumers (percen share of variable) Type of consumers Rule of humb 40 Ineres Raes Forward looking 60 Nominal shor erm 5.1 Consumpion Real shor erm 3.0 Rule of humb 15.6 Forward looking 84.4 Price markups over marginal coss (in percen) Tradables 14.3 Ineremporal Elasiciy of Subsiuion 0.33 Nonradables 26.7 Elasiciy of labor supply 0.92 Probabiliy of survival 0.90 Elasiciy of subsiuion beween capial and labor 0.75 Home bias in governmen expendiure (excl. ransfers) yes Nominal rigidiies no Effecive Tax Raes, Governmen Revenue, and Transfers Social securiy conribuions workers Personal income ax 11.0 Rae 27.2 Rae 8.9 Revenue (o GDP) 8.7 Revenue (o GDP) Social securiy conribuions employers Consumpion 10.1 Rae 22.8 Rae 6.2 Revenue (o GDP) 8.7 Revenue (o GDP) Corporae income ax Transfers Rae 29.0 As a percen of GDP 19.7 Revenue (o GDP) 8.1

13 11 The key behavioral parameers are based on microeconomic evidence. These include parameers characerizing real rigidiies in invesmen, markups for firms and workers, he elasiciy of labor supply o afer-ax wages, he elasiciy of subsiuion beween labor and capial, he elasiciy of ineremporal subsiuion, and he rae of ime preference. Simulaions examine he impac of changing he values of he following key parameers: 9 The wedge beween he rae of ime preference and he yield on governmen bonds. This parameer, which deermines consumers degree of impaience, has no been subjec o exensive microeconomic analysis. We se he baseline value of he wedge o 10 percen (corresponding o a planning horizon of 10 years), wih an alernaive simulaion using 1 percen (corresponding o a planning horizon of 100 years). The fracion of liquidiy-consrained consumers. The baseline assumes ha 40 percen of consumers experience liquidiy consrains. These consumers have no wealh and consume one-fourh of aggregae consumpion. An alernaive simulaion assumes ha 10 percen of individuals are liquidiy consrained. As noed in Table 2, alhough accouning for 40 percen of consumers, his group only accouns for 15 percen of aggregae consumpion. Moreover, he assumpion of 40 percen liquidiy-consrained consumers is in he mid-range of exising esimaes for advanced economies (See Campbell and Mankiw (1991) for an early esimae of percen of consumers for he U.S. and Iacoviello (2004) for a more recen esimae for he U.S. suggesing ha percen of consumpion is subjec o liquidiy consrains). Noe also ha he presence of opimizing consumers which have shor planning horizons also leads o a high marginal propensiy o consume. The sensiiviy of labor supply o he real afer-ax wage (Frisch elasiciy). The baseline value (-8) is a he low end of hose found by microeconomic sudies given ha he specific German microevidence poins o a more inelasic relaionship (Evers, de Mooij, and van Vuuren, 2005). Alernaive simulaions assume almos compleely inelasic labor supply (-1). The elasiciy of ineremporal subsiuion. The baseline value for his parameer, which describes he sensiiviy of consumpion o changes in he real ineres rae, is The parameer value in he alernaive simulaion (-0.29) is consisen wih he lower end of microeconomic esimaes. 9 Oher srucural parameers have been calibraed using evidence from Laxon and Peseni (2003) and Baini, N Diaye, and Rebucci (2005).

14 12 IV. DEBT DYNAMICS AND FISCAL ADJUSTMENT IN GERMANY A. Deb Dynamics: Populaion Aging in Germany and in Trading Parners To evaluae he deb dynamics in Germany we assume aging-relaed expendiure pressures of 4 percen of GDP by 2050 in our simulaions (his is in he middle range of exising esimaes; see discussion below). Figure 1 shows he ime profile of hese spending pressures. We also facor in he hree-percenage poin increase in he sandard VAT rae from 16 percen o 19 percen in 2007 esimaed o generae addiional revenue of 1 percen of GDP. To ake ino accoun VAT exempions, we have mapped his ino a corresponding effecive VAT rae of 10.1 percen of oal consumpion in 2006, which increases by 1.9 percenage poins from 2007 onward. In addiion, we incorporae he payroll-ax relief equivalen of 0.4 percen of GDP, effecive January 2007 and a reducion in he CIT rae a a revenue loss of ¼ percen of GDP from 2008 onward. The proposed reform would reduce he marginal CIT rae from an average of 39 percen o less han 30 percen, parly financed hrough base-broadening measures. Since he governmen s plan wih regard o he CIT reform involves measures in several areas, for racabiliy, he simulaion posis an equivalen ax relief hrough a CIT rae reducion only, yielding a revenue loss of ¼ percen of GDP wihou offseing base broadening (see Boman and Danninger (2007) for a more deailed descripion and assessmen of hese ax reform proposals and is impac on deb dynamics, oupu, and consumpion) Overall, however, he reforms are likely o achieve a more efficien ax sysem by shifing from direc o indirec axaion. Over he long run, shifing revenue from direc axaion o less disorionary indirec axes increases growh hrough higher employmen and invesmen growh. This is relevan in an aging sociey where he direc ax base could conrac, while he indirec ax base is more sable (see Boman and Danninger, 2007, for a deailed evaluaion of hese ax policy changes).

15 13 Figure 1. Germany: Primary Expendiure (In percen of GDP) 49.0% 48.0% 47.0% 46.0% 4% 44.0% 43.0% 42.0% 41.0% Primary expendiure 4% Figure 2. Effecs on German Deb Dynamics of Populaion Aging in Germany 1/ (percen of GDP) Source: GFM simulaions. 1/ Includes he esimaed revenue from higher VAT raes in 2007 and revenue loss from lower social securiy conribuions by workers and employers in 2007 and lower corporae income axaion in 2008; aging-relaed expendiure coss are 4 percen of GDP by 2050.

16 14 In GFM, responses of he economy o hese spending pressures and changes in ax policy are deermined endogenously. The srucurally adjused defici a end-2006 is included, as well as he effec of inflaion and growh on deb dynamics. Simulaions projec a srucural defici of around 2 percen of GDP by end-2008 afer he effecs of he ax reform have fully worked hemselves ou. This figure is consisen wih projecions in IMF (2006), alhough he defici and, herefore, he required furher adjusmen o achieve srucural balance could be lower if, oher hings given, he laes projecions for he srucural defici in 2006 (somewha below 2 percen of GDP) are susained. 11 Figure 2 repors he resuling deb dynamics relaive o he iniial seady sae level in Germany, holding aging-relaed expendiure in he res of he euro area or he Unied Saes consan. Alhough deb declines moderaely in he shor-erm, following lower aging coss and he higher VAT in 2007, i is clearly unsusainable in he long erm, and never declines below he Maasrich limi. This deb pah will maerialize under curren policies and highlighs he magniude of he fiscal burden facing policymakers, alhough in pracice remedial measures will mos cerainly be aken a some poin. The deb dynamics depend crucially on projecions relaed o aging coss. The analysis in Figure 2 assumes ha, even afer far-reaching reforms (Agenda 2010 and Harz reforms), aging-relaed expendiures are projeced o increase by some 4 percen of GDP by This fiscal-aging cos profile is aken from a long-erm fiscal scenario developed in Braumann and ohers (2006), and is in beween a more opimisic scenario by he auhoriies (Federal Minisry of Finance, 2005; and Werding and Kalschüz, 2005) and he EU s Aging Working Group (2¾ percen), and a more pessimisic view expressed by he IFO insiue (7¾ percen). Figure 3 illusraes he sensiiviy of he deb dynamics in Germany o hese alernaive esimaes. Despie he sensiiviy of he deb dynamics o he uncerainy of aging coss, deb is unsusainable under curren policies even if we ignore he effec of populaion aging in oher counries and he resuling pressure on real ineres raes in Germany. 11 There are considerable uncerainies wih regard o he evenual cos of he CIT reform, and he ineremporal allocaion of consumpion and revenues in anicipaion of he VAT increase. In addiion, he revenue buoyancy may have exceeded uniy due o progressive axaion, and would work he oher way as growh slows. Finally, as discussed furher below, here are subsanial risks ha he esimaes of aging coss may in fac be undersaed.

17 15 Figure 3: The Effecs of Alernaive Aging Cos Projecions on Deb Dynamics in Germany 1/ (percen of GDP) Baseline: 4 percen of GDP aging coss by 2050 Maasrich deb limi High aging coss (7.75 percen of GDP by 2050) Low aging coss (2.75 percen by 2050) Source: GFM simulaions 1/ Includes he esimaed revenue from higher VAT raes in 2007 and revenue loss from lower social securiy conribuions by workers and employers in 2007 and lower corporae income axaion in Taking ino accoun aging pressures in he res of he euro area and he deb oulook for he Unied Saes, adds a dramaic dimension o he deb dynamics due o Germany s own aging profile. Figure 4 illusraes he effecs of global aging, in addiion o fiscal pressures from populaion aging in Germany, on deb dynamics in Germany, under curren policies. As he euro area and he U.S. face significan spending pressures from aging and, given weak saring fiscal posiions, fiscal deficis increase, and curren accoun deficis emerge or increase leading o higher real ineres raes. Germany is an open economy implying ha hese higher raes will increase borrowing coss for he governmen on newly issued deb and makes he deb oulook furher unsusainable. This link beween counries hrough inernaional financial markes is he key spillover effec of global aging and ignoring he inerdependencies beween counries will lead o a misleading judgmen abou he macroeconomic oulook of a counry, and he efficacy of a policy response.

18 16 As we discuss in more deail in Secion B below, his pressure on global ineres raes may be miigaed o some exen if enilemen reform occurs or is expeced. In ha case, opimizing, forward-looking, consumers depending on he lengh of heir planning horizons and on he expeced magniude of he reform will save more oday in anicipaion of lower sae pension benefis in he fuure. Also, in an aging sociey, he capial-labor raio will likely increase causing downward pressure on real ineres raes. On he oher side, however, he deb rajecory above does no assume a posiive risk premium for example as a funcion of he level of governmen deb, or ne capial inflows undersaing he likely effecs of aging on real ineres raes in he absence of measures. B. Fiscal Adjusmen in Germany As deb is unsusainable, we explore alernaive adjusmen sraegies o achieve deb susainabiliy defined here as a governmen deb-o-gdp raio of below 60 percen by Thus, he size and composiion of he fiscal adjusmen are deermined exogenously, such ha he resuling deb rajecory which is deermined endogenously in GFM remains below he Maasrich Limi.

19 17 Figure 4: Effecs on German Deb Dynamics of Global Populaion Aging 1/ (In percen of GDP) Aging in Germany alone Maasrich deb limi Aging in Germany, euro area, and U.S Source: GFM simulaions and IMF long-erm public finances projecions for Germany 1/ Includes he esimaed revenue from higher VAT raes in 2007 and revenue loss from lower social securiy conribuions by workers and employers in 2007 and lower corporae income axaion in 2008; aging-relaed expendiure coss are 4 percen of GDP by Increase in real ineres raes consisen wih evidence in Ford and Laxon (1999) who find ha a 12.5 percen increase in deb in he OECD increased real ineres raes by 100 basis poins (on new deb) during he 1980s. Assumes aging coss of 4.5 percen of GDP by 2050 in he euro area and 6.0 percen in he Unied Saes, wih deb increasing by 17.5 percen by Achieving fiscal susainabiliy requires addiional effors beyond hose in he coaliion s ax reform package. As se ou in is Sabiliy Program, he governmen aims o move owards srucural balance over he medium-erm. In pracical erms, his would require abou a ½ percenage poin of GDP reducion in he defici per year during (some 2 percen of GDP in oal). However, addiional adjusmen he size of which depends on he ype of adjusmen measure is required o mainain deb below he Maasrich limi for he enire period unil We characerize he size of he adjusmen needed, if adjusmen is fron loaded, beyond achieving srucural balance by 2011 for each measure in Table 3. Srucural fiscal adjusmen could be aained hrough various combinaions of expendiure and revenue measures (Table 3). GFM is used o compare he effecs of differen consolidaion mehods: (i) lower governmen consumpion; (ii) lower governmen ransfers; (iii) higher worker social securiy conribuions; (iv) higher employer social securiy

20 18 conribuions; (v) higher personal income ax raes; (vi) higher VAT; (vii) raising social securiy conribuions of workers in combinaion wih income ax base broadening measures; and (viii) higher corporae income ax raes. In addiion, he effecs of a combinaion of some of hese measures labeled a package are examined. Table 3. Germany: Alernaive Fiscal Adjusmen Sraegies: / (In percen unless oherwise indicaed) VAT Sauory rae Effecive ax rae Revenue (in percen of GDP) Labor income ax workers Effecive ax rae Revenue (in percen of GDP) Base broadening and higher labor income ax workers Base broadening 2/ Reducing effecive exempion rae (in percen of GDP) Labor income ax Effecive ax rae Overall revenue effec (in percen of GDP) Labor income ax employers Effecive ax rae Revenue (in percen of GDP) Personal income ax Effecive ax rae Revenue (in percen of GDP) Corporae income ax Effecive ax rae Revenue (in percen of GDP) Social securiy ransfers In he absence of fiscal adjusmen (aging effec) 3/ Spending (in percen of GDP) Wih fiscal adjusmen Spending (in percen of GDP) Governmen consumpion Spending (in percen of GDP) Percenage reducion (relaive o 2007) Package of measures (in percen of GDP) Spending Social securiy ransfers Revenue from reducing effecive exempion rae VAT (in percen) Effecive rae Sauory rae Source: IMF saff esimaes. 1/ Adjusmen of 1/2 percen of GDP during in addiion o he coaliion agreemen. All adjusmen scenarios achieve srucural balance by 2011 wih subsequen adjusmen o mainain governmen deb wihin he Maasrich limi unil / Reducion in he fracion of average labor income ha is ax exemp. 3/ Aging coss are projeced o decline in Germany during as a resul of recen reforms (see Figure 1).

21 19 Figure 5. Macroeconomic Effecs of Alernaive Fiscal Adjusmen Sraegies 1/ (Deviaion from iniial seady sae) Revenue measures Expendiure measures 4 3 Adjusmen package Indirec axaion Direc axaion Real GDP (percenage poins) 4 3 Adjusmen package Spending Transfers Real GDP (percenage poins) Consumpion (percen) 14 Consumpion (percen) Consumpion poor (rule-of-humb) minus wealhy (opimizing) (percenage poins) 5 Consumpion poor (rule-of-humb) minus wealhy (opimizing) (percenage poins) Source: GFM simulaions. 1/ All adjusmen sraegies achieve srucural balance by 2011 wih subsequen adjusmen o mainain governmen deb below he Maasrich deb limi. See Table 3 for deails on each adjusmen sraegy. Indirec axaion relies on higher VAT raes; direc axaion relies on higher corporae income axaion.

22 20 Figure 6. Deb Dynamics in Germany Wih Fiscal Adjusmen (in percen of GDP) Deb dynamics wih adjusmen Maasrich deb limi Deb dynamics wihou adjusmen Source: GFM simulaions. Calibraion resuls from GFM sugges ha he shor-run growh slowdown of achieving srucural balance varies wih he ype of consolidaion measure (Figure 5). The impac on shor run growh varies beween -0.2 percen and -0.3 percen each year beween , depending on he specific measures, heir disorionary effec, and he impac on domesic demand. All individual consolidaion measures, as well as he package, lead o he same deb profile unil Revenue increases. The negaive growh impac of he differen ax measures ranges from -0.1 percen o -0.3 percen. The VAT is less disorionary han payroll axes, because i also axes accumulaed savings (i.e., reaches a broader base, including reirees) in addiion o affecing he labor-leisure choice. Increasing payroll axes on workers is more disorionary han reducing ax exempions (base broadening) owing o marginal ax raes on workers exceeding he average rae. Raising corporae or personal income axes is roughly equally disorionary in erms of oupu loss. Tha payroll axes are less disorionary han axaion of capial is a resul found in oher

23 21 sudies as well (Baylor, 2005), and reflecs generally inelasic labor supply in paricular among males in mos indusrial counries, including in Germany. Figure 7: Macroeconomic Effecs on Germany of Adjusmen Package 1/ (Deviaion from iniial seady sae in percenage poins unless oherwise noed) 4.0 Real GDP 1 Consumpion (percen) Consumpion Opimizing consumers Rule-of-humb consumers Facors of producion Relaive prices Real wage (percen) Real exchange rae (percen) Real ineres rae (righ axis) Invesmen (percen) Capial sock Labor effor (righ axis; percen) Overall defici and curren accoun balance Governmen deb and ne foreign asses Governmen deb (percen of GDP) Ne foreign asses Overall governmen defici Curren accoun balance Source: GFM simulaions 1/ See Table 3 for deails on he adjusmen package.

24 22 Expendiure cus. Lowering social securiy ransfers has he smalles growh impac per year. This relaively modes growh effec occurs when he benefis ha are reduced are disribued in a lump sum manner reducing ransfers ha cause economic disorions (such as unemploymen benefis) would imply growh losses similar o hose observed for higher payroll axes in so far ransfers can be inerpreed as negaive axes. Par of he decline in consumpion demand is absorbed by rading parners via reduced impor demand. In conras, reducions in oher governmen consumpion would lead o a larger slowing in growh, which reflecs he fac ha curren governmen spending in mos counries is heavily biased owards domesic goods, or nonradables ( home bias ). If no specific sraegy is proposed, rising aging-relaed expendiure pressures would likely resul in higher direc axaion. German law sipulaes ha he social securiy accouns have o mainain balance and as a resul, under curren rules, growing expendiures mus be me wih equivalen social securiy conribuions. While his prevens runaway fiscal deficis and he buildup of deb shown in Figures 2 o 4, i implies higher payroll axes as he defaul policy response. This defaul sraegy is no desirable from an efficiency perspecive as shown in Figure 5. Achieving he 2 percen of GDP adjusmen beween by relying exclusively on jus one of he eigh adjusmen measures appears difficul, and he governmen likely will need o choose a combinaion of measures. For insance, reducing governmen spending wheher on goods and services or social securiy ransfers by 2 percenage poins of GDP by 2011 implies unrealisically large cus in discreionary spending. Similarly, furher increases in he VAT revenue are also limied (including hrough EU regulaions), alhough furher base broadening would be possible by placing fewer iems under he lower (7 percen) VAT rae. Raising direc axaion is disorionary and runs couner o he governmen s inenions o increase incenives for labor paricipaion and invesmen. The macroeconomic effecs of a mixed policy package are also oulined in Figure 5, comprising lower governmen spending, lower social securiy ransfers hrough enilemen reforms, reducions in income ax exempions (base broadening), and a small furher increase in he effecive VAT, o a sauory rae of 20 percen. Such a package compares favorably o raising direc axes. Eliminaing he srucural defici by 2011 hrough such a package lowers medium-run growh by abou 0.2 percen per year (Figure 5 repors he corresponding decline in he level of oupu relaive o he level ha would perain in he absence of adjusmen, and of aging). The effecs on consumpion are more sizable, paricularly for hose consumers ha rely relaively more on labor income as lower social securiy ransfers in paricular lower heir consumpion opporuniies. Neverheless, such a package is no necessarily worse for hese rule-of-humb consumers compared o he long-erm effecs of raising corporae income axaion. The laer ax affecs consumpion by he opimizing consumers more in he shor erm as hese consumers hold he equiy in he firms, bu in he long erm he ax

25 23 incidence shifs o workers as well hrough lower real wages. Also, he shor-erm decline in oupu from fiscal adjusmen could be larger if wages and prices are sicky. These adjusmen sraegies achieve a considerable improvemen in deb dynamics (Figure 6). Eliminaing he srucural defici by 2011, wih furher adjusmen hrough 2017, implies a subsanial improvemen in he overall fiscal balance. In a sense, he adjusmen package prefunds (o a subsanial degree) fuure aging coss: a significan reducion in governmen deb over he medium erm reduces he ineres burden of he governmen o he exen ha rising aging coss can be accommodaed wih limied increases in deb up unil Alhough he adjusmen proposed here prevens deb from rising above he Maasrich deb limi, Figure 6 also suggess, however, ha owards he end of he period, furher measures will be required. Prefunding fuure aging coss resuls in a reducion in real ineres raes, by abou 40 basis poin in he medium erm when governmen deb reaches a rough he magniude of his effec depends criically on how non-ricardian consumers are. Alhough his effec is larger in Germany han would be observed in oher euro area counries, i is relaively modes as Germany s naional savings rae has only a limied effec on he global savings and invesmen balance (Figure 7). A higher VAT and base broadening of income axaion imply reduced incenives for labor paricipaion, as indicaed by he decline in hours worked and he increase in he real wage. Invesmen responds posiively hrough crowding-in afer deb sars o decline. Oupu gradually increases as a resul, alhough i akes a long ime before GDP reurns o he level observed before he fiscal adjusmen was iniiaed. As consumpion declines, he real exchange rae depreciaes, and he rade balance records sronger surpluses. Higher naional saving resuls in increasing claims on he res of he world hrough accumulaion of ne foreign asses. Higher ineres earnings from he ne foreign asse posiion imply ha he curren accoun remains above he iniial seady sae level for a considerably longer period han consumpion and he rade balance. V. GLOBAL AGING AND FISCAL ADJUSTMENT: INTERNATIONAL COOPERATION? Since deb is seen o be unsusainable in Germany, he res of he euro area, and he Unied Saes under unchanged policies, we analyze he implicaions of each counry/region implemening an adjusmen package similar o he package implemened in Germany, alhough he size and composiion will depend on he saring fiscal posiion and fuure aging coss. The focus is on he impac of such an adjusmen package on oher counries and regions via spillover effecs hrough financial and rade channels. The sequencing of he fiscal consolidaion in differen counries is paricularly ineresing. To analyze he poenial coss and benefis of fiscal cooperaion, we underake simulaions based on wo alernaive consellaions: (i) ha he adjusmen package is implemened immediaely, albei gradually; or (ii) i will be implemened afer 10 years. These scenarios

26 24 capure he relaive coss of fiscal cooperaion, including he poenial risks of a simulaneous decline in global demand, and he benefis, versus he incenives o poenially free ride on fiscal adjusmen in oher pars of he world. In addiion, we consider he effecs of a sequenial, sop-and-go policy ha feaures a more gradual adjusmen. The deails of he package in he euro area and he Unied Saes could be differen from he one in Germany, given he differences in he srucural characerisics and saring poins (see Table 4 for deails on he package in he euro area and he U.S.). We assume, as was done for Germany, ha he package of adjusmen measures in he euro area mainains governmen deb below he Maasrich limi unil 2050 and avoids increases in direc axaion if implemened wihou delay. Alhough he iniial fiscal balance and he magniude of aging coss are differen in he res of he euro area compared o Germany, he adjusmen will reduce he real ineres raes more in he euro area as a whole since i is a larger region (compared o Germany). This essenially allows he size and composiion of he package (o sabilize deb a abou 55 percen) o be similar o ha in Germany. For he U.S., o be consisen wih Germany and he euro area, he package is designed such ha deb remains somewha below is curren level unil However, despie he U.S. being a larger economy compared o Germany, given he subsanially higher aging coss including Medicare and Medicaid coss in he U.S. he package needs o include furher measures, assumed here o consis of furher increases in indirec axaion. To assess he benefis of cooperaive acion, we analyzed a number of varians for he early adjusmen and he lae adjusmen scenarios: we firs consider all hree counries/regions underaking early join acion, and hen each one delaying by 10 years, while he oher wo implemen early adjusmen. We also explore pair-wise delays (e.g., U.S. and res of euro area delay by 10 years, while Germany coninues wih adjusmen), and hen all hree delaying. A key resul is ha Germany benefis subsanially from early fiscal adjusmen in he res of he euro area, and in he U.S. (Figure 8). Alhough expors decline relaive o he iniial seady sae level following lower consumpion in he wo regions and herefore lower demand for impors from Germany, his is more han offse by he decline in real ineres raes hrough financial linkages beween Germany and he euro area and he U.S. The resul is an invesmen rebound and higher domesic demand in Germany The differenial effec on real ineres raes beween large and small economies implies an ineresing hypohesis: large economies should have a greaer incenive for pruden fiscal polices han small economies. Small economies wih inegraed capial markes have a smaller incenive o implemen a fiscal conracion, as he real ineres rae will no decline much relaive o larger open economies, or small closed or financially less inegraed economies. On he oher hand, and absracing from he role of moneary policy, large open or relaively closed economies have a smaller incenive o use fiscal policy as a demand sabilizing insrumen because of he sronger crowding ou effecs and smaller mulipliers.