FISCAL POLICY SUSTAINABILITY IN ROMANIA

Transcription

1 FISCAL POLICY SUSTAINABILITY IN ROMANIA Professor PhD. Ioan Talpoş, Wes Universiy of Timisoara, Lecurer PhD Candidae Cosmin Enache, Wes Universiy of Timisoara, ABSTRACT: The paper is focused on aspecs regarding fiscal policy susainabiliy in Romania. Analyzing he heoreical foundaions of fiscal policy susainabiliy i could be idenified a mehodology for economeric ess o be carried ou and o pu forward de faco public deb longrun siuaion a he counry level. In his direcion, we carried ou saionariy ess for he soc of he gross public deb and coinegraion ess for he public revenues end public expendiures flows for he case of Romania. Keywords: fiscal policy, susainabiliy, saionariy ess, coinegraion ess JEL codes: H20, H63, C22, C32. INTRODUCTION A lo of conemporary economies run public deficis and accumulae significan socs of public deb, so he problem of long-erm susainabiliy of fiscal policy is an imporan and acual one. The relevance of he fiscal policy susainabiliy was envisaged a he beginning of he pas cenury, afer de Firs World War by he English lord John Maynard Keynes in his sudy regarding he problem of he French public deb: he sae liabiliies ( ) have reached an excessive proporion of naional income (Keynes, 938). Following his firs approach, in a modern vision, he fiscal policy underaen by he governmen could be defined as susainable if i is a policy such he raio of deb o GNP evenually converges o is iniial level (Blachard, Chouraqui, Hagemann, Saror, 990). Theoreically, a cerain level of public deb could be considered accepable if i could be absorbed by he fuure primary surpluses of he public budge or, in oher words, if his level is equal o he presen value of he fuure public budge balances. In order o disinguish his complex opic of fiscal policy susainabiliy, we will mae an exposiion similar o hose in he relaed lieraure (see Blachard, Chouraqui, Hagemann, Saror, 990; Cuddingon, 996; Afonso, 2004). where: 2. THEORETICAL BACKGROUND The governmen is facing he following budgeary consrain: G + (+r )B - = R + B () G B R public expendiures of he period, no including he ineres expendiures aached o he public deb; public deb soc of he period; public revenues of he period; 233

2 r real ineres rae. Taing ino accoun he fac ha budgeary periods have (a leas virually) an infinie succession, we could rewrie he equaion () for he periods +, +2, +3,, and we could solve he resuling equaions sysem for B. Therefore, i could be derived he iner-emporary budgeary consrain for governmen auhoriies: R G B B = lim (2) = ( ) ( ) j= + r+ j + r j= + j Supposing ha he real ineres rae is saionary and wih a mean r (as he empirical sudies on his maer show very clearly, his is exremely probable), he oal public expendiures (E ) is given by he following relaion: E = G + (r r)b - (3) In hese condiions, he relaion (2) becomes: B (4) + = R E + = B lim + ( ) ( ) r ( r) The firs erm from he righ side of relaion (4) represens he presen value of he fuure primary budgeary surpluses. If he second erm of relaion is zero, han he value of he public deb curren soc is equal wih he presen value of fuure primary budgeary surpluses. In his case, he fiscal policy promoed by he governmen could be considered susainable. Therefore, i is necessary o impose he following ransversaliy condiion: lim B + ( + r) = 0 (5) The ransversaliy condiion in (5) forces he governmen auhoriies no o engage in Ponzigames regarding public finances ( a governmen engages in Ponzi finance if, afer some dae, i never runs a primary (non-ineres) budge surplus despie having a posiive soc of deb ousanding - (Buier, Klezer, 994)). In oher words, his condiion imposes ha he presen value of he public deb o converge o zero in an indefinie fuure. The fiscal policy susainabiliy condiion could be also wrien in raes o GDP. Therefore, we could rewrie he relaion () as follows: where: Y GDP; y real growh rae of GDP; B = Y b ( + r ) B G R + ( + y ) Y Y Y (6) or ( + r ) b + g q ( + y ) = (7) 234

3 b, g, q public deb, public expendiure and respecively public revenues relaive values o GDP. Considering he real ineres rae as saionary and wih mean r an he economic growh rae y, he iner-emporal budgeary consrain is given by he following relaion: y + y b = ( q + e+ ) + lim b + (8) = 0 + r + r Nex, o ensure ha he inheried public deb soc is equal o indefinie flow of fuure budgeary surpluses, we have o impose he following condiion: lim b + + y + r + = 0 (9) An alernaive indicaor of fiscal policy susainabiliy derived using he same concepual and mehodological framewor as he one already presened is Blachard s susainabiliy index (I B ) (Blachard, Chouraqui, Hagemann, Saror, 990). This index is given by he difference beween he susainable level of axaion ( * ) and he curren level of axaion (). I B = * - (0) In heir approach, he susainable level of axaion could be deermined if he expeced level of public expendiures and he iniial level of public deb are boh nown. This level of axaion will be given by he raio of acual values of public expendiures o GDP supplemened by he difference beween he ex ane ineres rae and he economic growh rae, muliplied wih he raio of public deb o GDP. The inerpreaion of his indicaor is exremely simple: if he susainable level of axaion exceeds he curren level of axaion i is obvious ha in he fuure is necessary an increase in global ax burden, unil he susainable level of axaion is aained. Moreover, he exac size of he necessary increase is nown by he governmen auhoriies because is given by he susainabiliy index. In conclusion, he fiscal policy measures regarding he budge balance and he public deb could be considered susainable if: he presen value of fuure budgeary surpluses is equal o he curren soc of public deb; he presen value of public deb approaches zero in an infinie ime horizon. 3. METHOD AND RESULTS The fiscal policy susainabiliy could be proven empirically hrough: ess for he saionariy of he firs differences of public deb soc; ess for a firs-order coinegraion relaion beween he firs differences of oal public expendiures (including deb s ineres) an oal public revenues. 3.. Saionariy ess In order o verify he susainabiliy condiion in (5) is sufficien o es if he ime series of firs differences of public deb soc is saionary (ha is if is mean and variance are consan over ime and he value of covariance beween wo periods depends only on he disance or lag beween 235

4 wo ime periods and no on he acual ime a which he covariance is compued (Gujarai, 995)). In order o do his, we used he ADF (Augmened Dicey-Fuller) saionariy es. The ime series used for he es was composed by 33 observaions, represening he quarerly gross consolidaed public deb soc of Romania for he period beween 2000 quarer I and 2008 quarer III. The saisical daa source was he EU saisical daabase Eurosa. The resuls of ADF es for he firs differences of public deb soc are he following: The resuls of ADF es for he firs differences of public deb soc Null Hypohesis: D(PDRO) has a uni roo Exogenous: Consan Lag Lengh: 0 (Auomaic based on SIC, MAXLAG=8) -Saisic Table no. Prob.* Augmened Dicey-Fuller es saisic Tes criical values: % level % level % level *MacKinnon (996) one-sided p-values. Augmened Dicey-Fuller Tes Equaion Dependen Variable: D(PDRO,2) Mehod: Leas Squares Sample (adjused): 2000Q3 2008Q2 Included observaions: 32 afer adjusmens Variable Coefficien Sd. Error -Saisic Prob. D(PDRO(-)) C R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Aaie info crierion Sum squared resid Schwarz crierion Log lielihood F-saisic Durbin-Wason sa Prob(F-saisic) The obained resuls confirm he saionariy hypohesis for he firs differences of public deb soc (τ saisic value obained hrough ADF es -6,90 is lower han he MacKinnon criical values (-3,65, - 2,96, -2,62) and he probabiliy ha he saionariy hypohesis no o be verified is 0%. Therefore, he fuure evoluion of Romania s public deb soc will converge o a saionary level and will no be permanenly affeced by evenual shocs Coinegraion es The exisence of a co-inegraion relaion beween wo variables proves ha beween hese variables exiss a long-erm (equilibrium) relaion. Therefore, even if he ime series of he respecive variables are non-saionary, a linear combinaion of hese wo ime series could be saionary. In order ha iner-emporal budgeary consrain (4) o be saisfied, i is necessary ha he variables R (public revenues) and E (public expendiures) o be coinegraed. Tha is, here is a relaion beween hese variable of he following ind: 236

5 R = a + be + u () where: a,b consans; u sochasic variable wih zero mean, consan variance σ 2 and non-self-correlaed. In hese condiions, we will es if he ime series of public revenues and public expendiures (including public deb s ineres) are coinegraed. The ess resuls could be inerpreed as follows: if here is no a coinegraion relaion beween he wo ime series, hen fiscal policy is no susainable; if here is a coinegraion relaion beween he wo ime series and he coefficien b=, hen he fiscal policy is susainable; if here is a coinegraion relaion beween he wo ime series and he coefficien b<, hen he fiscal policy is no susainable. For he coinegraion es we used monhly daa for public revenues and public expendiures from he period beween January 997 and augus he source of hese daa is NBR. For sar, we esed if he ime series of he wo variables, public revenues (TVRO) and public expendiures (TCRO) are saionary or non-saionary. The resuls of he ADF saionariy ess for he ime series of public revenues are he following: 237

6 The resuls of he ADF es for he ime series of public revenues Null Hypohesis: TVRO has a uni roo Exogenous: None Lag Lengh: 7 (Auomaic based on SIC, MAXLAG=3) -Saisic Table no. 2 Prob.* Augmened Dicey-Fuller es saisic Tes criical values: % level % level % level -.65 *MacKinnon (996) one-sided p-values. Augmened Dicey-Fuller Tes Equaion Dependen Variable: D(TVRO) Mehod: Leas Squares Sample (adjused): 997M M08 Included observaions: 32 afer adjusmens Variable Coefficien Sd. Error -Saisic Prob. TVRO(-) D(TVRO(-)) D(TVRO(-2)) D(TVRO(-3)) D(TVRO(-4)) D(TVRO(-5)) D(TVRO(-6)) D(TVRO(-7)) R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Aaie info crierion Sum squared resid Schwarz crierion Log lielihood Durbin-Wason sa I could be noiced ha he obained resuls confirm wih 99.85% probabiliy ha he public revenues ime series (TVRO) has a uni roo. So, he ime series is no saionary and i has a random wal. In oher words, TVRO is no I(0). We realized he same es for he firs differences of he public revenues ime series. The resuls are he following: 238

7 Table no. 3 The resuls of he ADF es for he firs differences of he public revenues ime series Null Hypohesis: D(TVRO) has a uni roo Exogenous: None Lag Lengh: 6 (Auomaic based on SIC, MAXLAG=3) -Saisic Prob.* Augmened Dicey-Fuller es saisic Tes criical values: % level % level % level -.65 *MacKinnon (996) one-sided p-values. Augmened Dicey-Fuller Tes Equaion Dependen Variable: D(TVRO,2) Mehod: Leas Squares Sample (adjused): 997M M08 Included observaions: 32 afer adjusmens Variable Coefficien Sd. Error -Saisic Prob. D(TVRO(-)) D(TVRO(-),2) D(TVRO(-2),2) D(TVRO(-3),2) D(TVRO(-4),2) D(TVRO(-5),2) D(TVRO(-6),2) R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Aaie info crierion Sum squared resid Schwarz crierion Log lielihood Durbin-Wason sa The obained resuls show ha he firs differences of public revenues ime series is saionary (or TVRO is I()). The saionariy es for public expendiures ime series are he following: 239

8 The resuls of he ADF es for he ime series of public expendiures Null Hypohesis: TCRO has a uni roo Exogenous: None Lag Lengh: (Auomaic based on SIC, MAXLAG=3) -Saisic Table no. 4 Prob.* Augmened Dicey-Fuller es saisic Tes criical values: % level % level % level *MacKinnon (996) one-sided p-values. Augmened Dicey-Fuller Tes Equaion Dependen Variable: D(TCRO) Mehod: Leas Squares Sample (adjused): 998M0 2008M08 Included observaions: 28 afer adjusmens Variable Coefficien Sd. Error -Saisic Prob. TCRO(-) D(TCRO(-)) D(TCRO(-2)) D(TCRO(-3)) D(TCRO(-4)) D(TCRO(-5)) D(TCRO(-6)) D(TCRO(-7)) D(TCRO(-8)) D(TCRO(-9)) D(TCRO(-0)) D(TCRO(-)) R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Aaie info crierion Sum squared resid Schwarz crierion Log lielihood Durbin-Wason sa As in he public revenues case, he obained resuls confirm wih a 00% probabiliy ha he public expendiure ime series (TCRO) has a uni roo, which means ha he ime series is no saionary and is a random wal. In oher words, TVCO is no I(0). The same es was applied for he firs differences of he public expendiures ime series and he obained resuls were he following: 240

9 Table no. 5 The resuls of he ADF es for he firs differences of he public expediure ime series Null Hypohesis: D(TCRO) has a uni roo Exogenous: None Lag Lengh: 4 (Fixed) -Saisic Prob.* Augmened Dicey-Fuller es saisic Tes criical values: % level % level % level *MacKinnon (996) one-sided p-values. Augmened Dicey-Fuller Tes Equaion Dependen Variable: D(TCRO,2) Mehod: Leas Squares Sample (adjused): 997M M08 Included observaions: 34 afer adjusmens Variable Coefficien Sd. Error -Saisic Prob. D(TCRO(-)) D(TCRO(-),2) D(TCRO(-2),2) D(TCRO(-3),2) D(TCRO(-4),2) R-squared Mean dependen var Adjused R-squared S.D. dependen var S.E. of regression Aaie info crierion Sum squared resid 2.0E+08 Schwarz crierion Log lielihood Durbin-Wason sa The obained resuls confirm ha firs differences of he public expendiures ime series is saionary (or TCRO is I()). Boh public revenues ime series and public expendiures ime series are I(), so we could apply he coinegraion es. The obained resuls were he following: 24

10 The resuls of he coinegraion es Sample (adjused): 997M M08 Included observaions: 35 afer adjusmens Trend assumpion: No deerminisic rend Series: TCRO TVRO Lags inerval (in firs differences): o 4 Table no. 6 Unresriced Coinegraion Ran Tes (Trace) Hypohesized Trace 0.05 No. of CE(s) Eigenvalue Saisic Criical Value Prob.** None * A mos * Trace es indicaes 2 coinegraing eqn(s) a he 0.05 level * denoes rejecion of he hypohesis a he 0.05 level **MacKinnon-Haug-Michelis (999) p-values Unresriced Coinegraion Ran Tes (Maximum Eigenvalue) Hypohesized Max-Eigen 0.05 No. of CE(s) Eigenvalue Saisic Criical Value Prob.** None * A mos * Max-eigenvalue es indicaes 2 coinegraing eqn(s) a he 0.05 level * denoes rejecion of he hypohesis a he 0.05 level **MacKinnon-Haug-Michelis (999) p-values Unresriced Coinegraing Coefficiens (normalized by b'*s*b=i): TCRO TVRO Unresriced Adjusmen Coefficiens (alpha): D(TCRO) D(TVRO) Coinegraing Equaion(s): Log lielihood Normalized coinegraing coefficiens (sandard error in parenheses) TCRO TVRO

11 ( ) Adjusmen coefficiens (sandard error in parenheses) D(TCRO) (0.5956) D(TVRO) ( ) The obained resul show wih a 00% probabiliy ha here is a leas wo coinegraion relaions beween he wo ime series, and he coinegraion equaion is he following: TCRO = 0,9877TCRO + u (2) Because he coefficien b is less han (b=0,9877), bu very close o, we could conclude ha in Romania, if he deerminans of he public revenues and public expendiures remain unchanged, he public deb will no have a divergen evoluion and he fiscal policy will be susainable in he long run. 4. CONCLUSIONS Since he beginning of he ransiion process, redefining he public auhoriies role in he economy and sociey necessiaed significan insiuional and srucural reforms. In hese condiions, our counry governmen auhoriies were and sill are facing a srong consrain: o saisfy increasing public needs wih limied fiscal resources, due o a small ax base. This paper was no focused on shor-erm managemen of a negaive budge balance. Insead, i was rying o apply an appropriae mehodology in order o reveal he cumulaive long-erm effecs of he governmen auhoriies decisions regarding he budge balance and he public deb. The fiscal policy susainabiliy was verified hrough wo mehods: a saionariy es for he public deb soc ime series and a Johanssen coinegraion es for he public revenues and public expendiures ime series. The ADF saionariy es for he quarerly public deb ime series revealed ha Romania s fiscal policy is susainable in he long-run. Applying he Johanssen coinegraion es for he monhly public revenues and public expendiures ime series we find, wih a probabiliy of almos 99%, ha here is a leas wo coinegraion relaions beween he wo ime series. Because he resuled coefficien b in he coinegraion equaion is less han, bu very close o, we conclude ha in Romania, if he deerminans of he public revenues and public expendiures remain unchanged, he public deb will no have a divergen evoluion and he fiscal policy will be susainable in he long run. References. Afonso A. (2005) Fiscal Susainabiliy: he Unpleasan European Case, FinanzArchiv: Public Finance Analysis, Mohr Siebec, Tübingen, vol. 6(), pages 9-, March. 2. Blanchard O., Chouraqui J-C., Hagemann R. P., Saror N. (990) The Susanabiliy of Fiscal Policy : New Answers o an Old Quesion, OECD Economic Sudies, no Buier W. H., Klezer K. M (994).- Ponzi Finance, Governmen Solvency and he Redundancy or Usefulness of Public Deb, Cowles Foundaion Discussion Paper, no Cuddingon J. T.(996) Analysing he Susainabiliy of Fiscal Deficis in Developing Counries, World Ban Policy Research Woring Paper No Gujarai D. N.(995) Basic Economerics, 3 rd ed., McGraw-Hill Boo Co. 6. Keynes J. M.(938) A Trac of Moneary Reform, cied in Anonio Afonso Fiscal Susainabiliy: he Unpleasan European Case,