FISCAL POLICY AND ECONOMIC GROWTH IN ROMANIA

Transcription

1 FISCAL POLICY AND ECONOMIC GROWTH IN ROMANIA Cosmin Enache ABSTRACT: The connecion beween fiscal policy and economic growh is no a srong one, aking ino accoun ha fiscal policy is no a fundamenal source of growh. Even so, governmen auhoriies could use fiscal policy o affec in an indirec manner he economic growh. We will ry o highligh his indirec connecion and is srengh on Romania s case using a specific economeric mehodology which akes ino accoun he resricions imposed by he governmen s budge consrain and we will idenify he specific fiscal policy measures which could enhance economic growh in Romania. Key words: fiscal policy, economic growh, governmen budge consrain JEL codes: E62, O40, C20 Inroducion Theoreical and empirical sudies from he lieraure envisage he fac ha beween fiscal policy promoed by he governmens and economic growh process could be a connecion, bu, his is an ambiguous one, from he perspecive of is srengh and is lengh. This ambiguiy of he connecion beween fiscal policy and economic growh is due, in he main par, o he fac ha fiscal policy is no a fundamenal source of economic growh. The fundamenal sources of economic growh are represened by economic facors like accumulaion of physical capial, labor force, human capial and echnological knowledge, and by non-economic, social, culural, poliical geographical facors, like he qualiy of insiuions, he availabiliy of naural resources or he dominan culural paradigm in he sociey. From his poin of view, fiscal policy represens jus a ool for he governmen auhoriies, which could be used o influence hese fundamenal sources of economic growh. As Vio Tanzi (997) pus i: while a mulipliciy of facors, some of which are of a noneconomic naure, could plausibly affec he performance of an economy from period o period, a counry s growh over a reasonably long period of ime is ulimaely deermined by hree facors: () given he sae of echnical know-how in ha counry, he efficiency wih which any exising sock of resources is uilized (which would depend, among oher hings, on culural, insiuional, and poliical, as well as economic, parameers); (2) he accumulaion over ime of producive resources (which would include human and oher forms of inangible capial); and (3) echnological progress (which for mos counries would depend, among oher hings, on heir abiliy o absorb new echnology from abroad). Even if regarding he effecs of fiscal policy on long-erm economic growh (in he saionary sae of he economy) he diversiy of he models from he lieraure do no converge on he same resuls, all hese models prove ha fiscal policy could influence a leas he level (if no he growing pace) of some imporan economic macro-variables, like income per capia, he capial sock per capial or he consumpion per capia. We consider ha he imporance of his finding of hese economic growh models was a lile bi negleced in he empirical and heoreical sudies realized so far, because he vas majoriy of hese sudies were focused on he developed economies, which have high values for hese variables ha could be affeced during he ransiion phase o saionary sae of he economy. Of course, in developing counries, he level of income per capial Wes Universiy of Timisoara, Faculy of Economics and Business Adminisraion, cosmin.enache@feaa.uv.ro 502

2 or he level of consumpion per capia are exremely imporan, because boh direcly influence he qualiy of life of he individuals. So, he imporance of fiscal policy for he economic growh process should no be underesimaed, a leas for he developing economies. More, i worh consider he fac ha in economic growh models wihou scale effecs, he process of he asympoic convergence of economic growh rae o he level corresponding o he saionary sae of he economy is exremely slow, which implies a considerable lengh for he ransiion period o he saionary sae of he economy, and by consequence, a sronger and more persisen impac of fiscal policy on he level of income per capia (Eicher and Turnovski, 999). Lieraure review Having in mind ha fiscal policy is no a fundamenal source of economic growh, i only could influence in an indirec manner he process of economic growh, hrough is main insrumens (public expendiure, public revenues, budgeary balance and public deb). Public expendiures could have a direc impac on long-erm economic growh if hey maerialize in goods ha ener as argumen in he producion funcion of economic agens (infrasrucure public expendiures) or in he uiliy funcion of he individuals. Bu, more imporan are he indirec effecs of he public expendiures on economic growh, which are due o heir influence on privae capial formaion, and o heir influence on he produciviy of he privae inpus in he producion funcion. Regarding he public expendiure effecs on privae capial formaion one could idenify several well documened ransmission mechanisms, like: he well-known crowding-ou effec, he complemenariy effec (as shown in Aschauer (989), Agénor (2004) or Sala-i-Marin, Doppelhofer and Miller (2004) he infrasrucure public expendiures could posiively - or negaively, when heir level is no he opimal one - affec he produciviy of privae capial usage in he economy), he effec on adjusmen coss (Turnovsky (996) and Agénor (2006)) and he effec on he durabiliy of privae capial (Agénor and Moreno- Dodson (2006), Agénor (2008)). The public expendiures could influence also he produciviy of he privae inpus in he producion funcions, and hence, he economic growh (Afonso and Alegre (2008)). On he one hand, public infrasrucure could posiively influence he produciviy of he privae physical capial as shown in heoreical sudies like Caning and Pedroni (999) or in empirical sudies like Demeriades and Mamuneas (2004), Fedderke and Bogeic (2006) or Zou e al (2008). On he oher hand, he produciviy of he human capial is srongly influenced by public educaion (Bils and Klenow (2000), de la Fuene (2003), Blankenau and Simpson (2004), Creedy and Gemmel (2005)) and by publicly provided healh services (Arora (200) and Agénor (2008)). More, public expendiures wih general services, naional defense, public order and naional securiy, housing and communiy ameniies represens core expendiures, absoluely necessary for limiing he inefficiencies induced by diverse marke failures and for a good funcioning of he economy (Tanzi and Schuknech, 2003). Having in mind all hese issues, Barro (990) group he public expendiures based on heir impac on economic growh in producive public expendiures (which have a posiive impac on economic growh), unproducive public expendiures (which are neural or have an insignifican impac on economic growh), and oher public expendiures (which have an insignifican impac on economic growh). 503

3 Table no. Public expendiures classificaion based on he impac on economic growh Producive public Unproducive public Oher public expendiures expendiures expendiures Educaion Healh Social securiy conribuions Jusice Recreaion, culure and religion Oher General public services Oher economic acions Housing Transpor and communicaions The possible effecs of public revenues (especially fiscal revenues) on long-erm economic growh, and he ransmission mechanisms of hese effecs Regarding public revenues, from a general perspecive, any ax has a poenial or real disorsionary impac on economic growh, because i affecs he choices of individual economic subjecs (firms and individuals) regarding he aciviies which hey carry on (producion, invesmen, consumpion, or savings). The corporae ax negaively affecs he incenives and he invesmen resources of he firms, he ax on wage income negaively influences boh individual consumpion and saving, and individual invesmen in human capial. The axes on consumpion affec individual choices beween work ime and leisure ime (Mendoza e al. (997), Milesi-Ferrei and Roubini (998)). Having in mind all hese feaures of he axaion, Barro and Sala-i-Marin (2004) group he public revenues based on heir impac on economic growh in: disorsionary public revenues (which have negaive effecs on economic growh), non-disorsionary public revenues (which are neural or have an insignifican impac on economic growh), and oher public revenues expendiures (which have an insignifican impac on economic growh). Table no. 2 Public revenues classificaion based on he impac on economic growh Disorsionary public revenues Corporae ax Income ax Social securiy axes Wealh axes Non-disorsionary public revenues Value-added ax (General axes on sales) Excises Oher public revenues Cusom duies Oher fiscal revenues Non-fiscal curren revenues Capial revenues Oher revenues Finally, he budge defici could influence economic growh hrough a ransmission channel represened by he ineres rae. More, he coninuous accumulaion of public deb could undermine long-erm susainabiliy of he fiscal policy promoed by he governmen auhoriies, wih a negaive impac on economic growh also. Research mehodology In order o envisage he connecion beween fiscal policy and economic growh in he Romania s case, we will follow he nex mehodological seps: firs, we will use he neoclassical growh model o idenify he main deerminans of economic growh for he Romania s case, and hen we build an economeric model for he evoluion of real GDP growh rae; second, we will idenify he specific public expendiures and revenues which have an impac on economic growh, and we will include hem in he original model; hird, we will es he effecs of some changes in public expendiures and revenues on he 504

4 real GDP growh rae dynamic, and, in he same ime, keeping in order he budgeary consrain. For a specific counry, he economic growh deerminans could be esed using he following general regression: where: VD α β VI ε VD n = α+ βivi i= i + ε - dependen variable; - free erm (consan); - vecor of lengh n of independen variables coefficiens; - vecor of lengh n of independen variables; - vecor of lengh n of sochasic perurbaions. Applying his very simple economeric mehodology on he Romania s case is a difficul hing o realize, due o unavailabiliy of long enough ime series for he relevan variables. In order o surpass his difficuly we used he exising ime series for seleced variables exended wih some forecased values, so he ime span covers he period. Having such a shor period of analysis we consider as primary deerminans for economic growh in Romania (quanified by he annual real GDP growh rae - RPIBR) only he physical capial accumulaion (quanified by he fixed capial formaion in % of GDP - FBCF) and labor force accumulaion (quanified by he annual rae of employed populaion - RPO). The sources for he saisical daa used are: - for real GDP growh rae IMF Counry Repors for Romania; - for fixed capial formaion in % of GDP, and for growh rae of employed populaion Naional Saisical Insiue and Naional Prognosis Commission. In order o highligh he long erm induced effecs of hese economic growh deerminans, he original daa were cleaned by uni-periodic shocks, aking ino accoun only heir rend. The esimaion mehodology for he rend is based on weighed moving average (MMP), because such approach offers he possibiliy o ake ino accoun he possible srucural breaks in he daa ses: ( X) = i= j= So, he relaion which will be esed is he following: i MMP i Xi (2) j RPIBR= α+β FBCF+β RPO+ ε (3) The obained resuls obained using E-Views 5.. are repored in he following able: Table no. 3 Regression resuls for he deerminans of economic growh in Romania Mehod: Leas Squares Sample: FBCF RPO C () 505

5 R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Akaike info crierion Sum squared resid Schwarz crierion Log likelihood F-saisic Durbin-Wason sa Prob(F-saisic) Resuls generaed using E-Views 5.. As i could be observed, he general level of significance of he model is no exremely high (R2 = 0,466772). The coefficiens for he wo independen variables were correcly esimaed and have saisical significance. The sign for he fixed capial formaion coefficien is he expeced one (+), confirming he direc relaion wih real GDP growh rae. The sign for he growh rae of employed populaion coefficien is (-), which indicaes an inverse relaion wih real GDP growh rae. This resul is a odds wih heoreical predicions, bu i could be explained if one have in mind he srucural adjusmens realized in Romania during he ransiion period, which imposed a rescaling of labor force o he real economic performance. The impac of he public secor size (quanified by he public expendiures in % of GDP - CHP_T), and of he ways of financing hem (given by he public revenues VEN_T, and budgeary balance SB, boh in % of GDP) on economic growh in Romania could be envisaged by esing he following relaions: RPIBR= α+β FBCF+β RPO+β CHP _ T+ ε (4) RPIBR= α+β FBCF+β RPO+β VEN _ T+β SB+ ε (5) 4 The resuls obained for he wo regressions are he following: Table no. 4 Esimaed resuls: he impac of public secor size on economic growh in Romania Mehod: Leas Squares Sample: FBCF RPO CHP_T C R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression.0050 Akaike info crierion Sum squared resid Schwarz crierion Log likelihood F-saisic Durbin-Wason sa Prob(F-saisic) Resuls generaed using E-Views

6 The obained resuls show a negaive and significan impac of he public secor size on economic growh. Moreover, inroducing public expendiures in he original model raises he overall significance level of he model (R 2 = 0,899895). Table no. 5 Esimaed resuls: he impac of public revenues and budgeary balance on economic growh in Romania Mehod: Leas Squares Sample: FBCF RPO VEN_T SB C R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Akaike info crierion Sum squared resid Schwarz crierion Log likelihood F-saisic Durbin-Wason sa Prob(F-saisic) Resuls generaed using E-Views 5.. The obained resuls show a negaive and significan impac on economic growh in Romania of public revenues. In he same ime, he size of budgeary balance in posiively correlaed wih real GDP growh rae, fac which indicaes ha a reducion of he budge defici has a posiive impac on economic growh. More, inroducing public revenues and budgeary balance in he iniial model leads o an increased level of significance (R2 = 0, 97707). Nex, in order o envisage he specific effecs of diverse budgeary variables on economic growh, we grouped he public expendiures and he public revenues following Barro and Sala-i- Marin (2004). So, he public revenues were divided in disorsionary (VEN_DIST), nondisorsionary (VEN_NOND) and oher public revenues (VEN_ALTE) (see Table no. 2). Due o he unavailabiliy of he appropriae ime series for he public expendiures, i was impossible o group hem using he scheme presened in Table no.. As an alernaive soluion, we use economic classificaion of he public expendiures, considering as producive he capial expendiures (CHP_PROD), as unproducive he expendiures wih wages, wih he acquisiion of goods and services, subvenions and ransfers (CHP_NEPR) and as oher he ineres paid for he public deb and oher public expendiures, (CHP_ALTE). In order o decide which of hese aggregaed budgeary variables could be included in he iniial economic growh model wihou any reducion in he level of is saisical significance, we realized for every one of hem an omied variable es. This es indicaes in which measure he iniially omied variable adds o he model explanaory power. The resuls of he omied variable es are synhesized in he following able: 507

7 Table no. 6 Esimaed resuls: Omied variable es Null hypohesis: The variable is no significan for he model Omied variable F saisic Log likelihood Log likelihood F saisic probabiliy rae probabiliy CHP_PROD 9, , , ,00984 CHP_NEPR 33,9793 0, ,3273 0,00000 CHP_ALTE 0, , , , VEN_DIST 27, , , , VEN_NOND 34, , , ,00000 VEN_ALTE 86,4800 0, , , SB 5, , ,6526 0, Synhesis of he resuls generaed using E-Views 5.. The resuls show ha we could no o include he oher public expendiures in he iniial model (he values for F saisic and Log likelihood are exremely low, and probabiliies ha he null hypohesis o be rue are 62,74%, and, respecively, 58,63%). Now, we can consruc and esimae a model for he real GDP growh rae which includes along he iniial variables he seleced aggregaed budgeary variables (all exceping oher public expendiures). The general relaion of he model is he following: where: g VE VB g n m i = α+ βivei + i= j= δ VB j j + ε - real GDP growh rae; - he vecor of economic independen variables; - he vecor of budgeary independen variables. i (6) The specific relaion of he model is he following: RPIBR= α+β FBCF+β2RPO+β3CHP _ PROD+β4CHP _ NEPR+ + β VEN _ DIST+β VEN _ NOND+β VEN _ ALTE+β SB+ ε (7) Esimaion resuls for he presened model are repored in he following able: Esimaion resuls: Regression wih relevan budgeary variables Mehod: Leas Squares Sample: Table no. 7 FBCF RPO CHP_PROD CHP_NEPR VEN_DIST VEN_NOND VEN_ALTE SB C

8 R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Akaike info crierion Sum squared resid Schwarz crierion Log likelihood F-saisic Durbin-Wason sa Prob(F-saisic) Resuls generaed using E-Views 5.. Analyzing he obained resuls one could noice ha he general saisical significance of he model is high (R2 = 0, 98590). Moreover, only disorsionary public revenues and budgeary balance have a significan impac on real GDP growh rae. The disorsionary public revenues have he expeced negaive impac on economic growh a % raise of hese public revenues deermines a 2.6% reducion in he real GDP growh rae. The budgeary balance has a posiive impac on economic growh a % reducion in he budge defici deermines a 3.33% raise in he real GDP growh rae. Anoher fac worh noiced is ha he producive public expendiures coefficien, alhough is no saisically significan, has a negaive value, which is agains heoreical predicions. On he one hand, his siuaion could be due o he inclusion of capial expendiure in his caegory. Because hese expendiures conribue o he fixed capial formaion in he economy, his could induce some redundancies ino he model. This siuaion could no be avoided because here were no available daa on privae fixed capial formaion. On he oher hand, such a resul could be appreciaed as an expression of he inefficien way of spending public money in Romania. The same siuaion repeas for he oher public revenues variable, is coefficien being a posiive one. Nex, we could highligh he impac on economic growh of some changes in he srucure of public expendiures and in he srucure of public revenues. Bu, if we have in mind ha in he m relaion 6, he budgeary consrain implies VB 0, i follows ha a leas one of he budgeary j= j = variables has o be excluded from he model in order o avoid he perfec colineariy (Kneller e al. (999)). This exclusion also offers a proper way o inerpre any changes in a budgeary variable included in he model: he change is realized based on a corresponding change of he omied variable from he model, such as he budgeary consrain o hold. Examining he resuls of he real GDP growh rae model which includes he budgeary variables, one could noiced ha of all public expendiures and revenues, he unproducive public expendiures and non-disorsionary public revenues have he lowes saisical significance, as i is prediced by he heoreical predicions. So, nex, we will eliminae boh hese variables from he model, one a a ime, and we will analyze he resuls. The resuls obained from he esimaion of he model given in relaion (7), changed by eliminaion of unproducive public expendiures are repored in he following able: Table no. 8 Esimaion resuls: Regression wih budgeary variables omied variable: Unproducive public expendiures Mehod: Leas Squares Sample:

9 FBCF RPO CHP_PROD VEN_DIST VEN_NOND VEN_ALTE SB C R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Akaike info crierion Sum squared resid Schwarz crierion Log likelihood F-saisic Durbin-Wason sa Prob(F-saisic) Resuls generaed using E-Views 5.. Analyzing he obained resuls one could noice ha, of all budgeary variables coefficiens, only hose for disorsionary public revenues and budgeary balance are saisically significan. Having in mind he budgeary consrain, which is supposed o hold: - a % reducion (increase) of disorsionary public revenues compensaed by a similar reducion in unproducive public expendiures will deermine a 0.96% raise (decline) in he real GDP growh rae; - a % reducion (increase) in he budgeary balance (equivalen o a % increase (decrease) of he budge defici) used o finance some unproducive public expendiures will deermine a 2.59% reducion (increase) of he real GDP growh rae. The resuls obained from he esimaion of he model given in relaion (7), changed by eliminaion of non-disorsionary public revenues are repored in he following able: Table no. 9 Esimaion resuls: Regression wih budgeary variables omied variable: Nondisorsionary public revenues Mehod: Leas Squares Sample: FBCF RPO CHP_PROD CHP_NEPR VEN_DIST VEN_ALTE SB C

10 R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Akaike info crierion Sum squared resid Schwarz crierion Log likelihood F-saisic Durbin-Wason sa Prob(F-saisic) Resuls generaed using E-Views 5.. If we omi from he model only he non-disorsionary public revenues, he only budgeary variable for ha he esimaed coefficien is saisically significan is represened by he budgeary balance. Hence, a % increase (decrease) in budgeary balance (equivalen o a % reducion (increase) of he public defici) based on a corresponding raise in non-disorsionary public revenues will rigger a 2.40% increase (reducion) of he real GDP growh rae. Conclusions In conclusion, in Romania, here are several fiscal policy measures ha could lead o an increase in he real GDP growh rae: a reducion of disorsionary public revenues compensaed by a reducion of unproducive public expendiures; a reducion of he budge defici compensaed by a reducion of he unproducive public expendiures; a reducion of he budge defici compensaed by a corresponding increase in he nondisorsionary public revenues. The resuls of his sudy should be inerpreed wih cauion. On he one hand, he reduced lengh of daa ses used in he esimaions could lead o a drop in he overall saisical significance of he esimaions. On he oher hand, he base model used in our economeric esimaions, which was augmened o ake ino accoun he budgeary variables, is a reduced form of he neoclassical growh model and does no include a variable for echnological progress, a very imporan source of economic growh. This omission could inroduce some disorions in our esimaions. Even so, he obained resuls give us an indicaion on he main possible posiive effecs of fiscal policy on economic growh in Romania. References.Afonso A., Alegre J. G., 2008, Economic Growh and Budgeary Componens. A Panel Assessmen for he EU, European Cenral Bank Working Papers, 848, January. 2.Agénor P.-R., 2004, The Economics of Adjusmen and Growh, second ediion, Harvard Universiy Press, Boson, Massachuses. 3.Agénor P. R., 2008, Fiscal policy and endogenous growh wih public infrasrucure, Oxford Economic Papers, Oxford Universiy Press, vol. 60(), p Agénor P.-R., Moreno-Dodson B., 2006, Public Infrasrucure and Growh: New Channels and Policy Implicaions, Policy Research Working Paper Series, 4064, The World Bank. 5.Arora S., 200, Healh, Human Produciviy, and Long-Term Economic Growh, Journal of Economic Hisory, LXI, p Aschauer D. A., 989, Is Public Expendiure Producive?, Journal of Moneary Economics, 23, p Barro R. J., 990, Governmen Spending in a Simple Model of Endogenous Growh, Journal of Poliical Economy, 98, p. s03-s25. 5

11 8.Barro R. J., Sala-i-Marin X., 2004, Economic Growh, 2nd ed., MIT Press, Cambridge, Massachuses. 9.Bils M., Klenow P., 2000, Does Schooling Cause Growh?, American Economic Review 90, p , December. 0. Blankenau W. F., Simpson N. B., 2004, Public Educaion Expendiures and Growh, Journal of Developmen Economics, Elsevier, vol. 73(2), p , April.. Canning, D., Pedroni, P., 999, Infrasrucure and Long Run Economic Growh, Mimeo, World Bank, Washingon DC. 2. Creedy J., Gemmell N., 2005, Publicly Financed Educaion in an Endogenous Growh Model, Journal of Economic Sudies, 32, 2, p De La Fuene A., 2003, Human Capial in a Global and Knowledge-Based Economy, Par II: Assessmen a he EU Counry Level, European Commission, Direcorae-General for Employmen and Social Affairs. 4. Demeriades P. O., Mamuneas T. P., 2000, Ineremporal Oupu and Employmen Effecs of Public Infrasrucure Capial: Evidence from 2 OECD Counries, Economic Journal, 0, p Eicher T. S., Turnovsky S. J., 999, Convergence Speeds and Transiion Dynamics in Non-Scale Growh Models, Journal of Economic Growh, 4, Fedderke J. W., Bogeic Z., 2006, Infrasrucure and Growh in Souh Africa: Direc and Indirec Produciviy Impacs of 9 Infrasrucure Measures, World Bank Policy Research Working Paper, 3989, Augus. 7. Howi P., 2005, Healh, Human Capial and Economic Growh: A Schumpeerian Perspecive, în Lopez-Casasnovas G., Rivera B., Currais L. (ed.) Healh and Economic Growh, Findings and Policy Implicaions, MIT Press. 8. Kneller R., Bleaney M., Gemmell N., 999, Fiscal Policy and Growh: Evidence from OECD Counries, Journal of Public Economics, 74, Mendoza E., Milesi-Ferrei G.M., Asea P., 997, On he Ineffeciveness of Tax Policy in Alering Long-Run Growh: Harberger s Superneuraliy Conjecure, Journal of Public Economics, 66, p Milesi-Ferrei G. M., Roubini N., 998, Growh Effecs of Income and Consumpion Taxes, Journal of Money Credi and Banking, 30, p Sala-i-Marin X., Doppelhofer G., Miller R., 2004, Deerminans of Long-Term Growh: A Bayesian Averaging of Classical Esimaes (BACE) Approach, American Economic Review, 94, p Tanzi V., Schuknech L., 2003, Public Finances and Economic Growh in European Counries, Fosering Economic Growh in Europe Conference, Vienna, June 2-3, Tanzi V., Zee H. H., 997, Fiscal Policy and Long-Run Growh, Inernaional Moneary Fund Saff Papers, 44, 2, June, 997, p Turnovsky S. J., 996, Fiscal Policy, Adjusmen Coss, and Endogenous Growh, Oxford Economic Papers, 48, p Zou W., Zhang F., Zhuang Z., Song H., 2008, Transpor Infrasrucure, Growh and Povery Alleviaion: Empirical Analysis of China, Annals of Economics and Finance, 9-2, p