ASSESSING THE IMPACT OF CORRUPTION UPON THE ROMANIAN ECONOMY

Transcription

1 ASSESSING THE IMPACT OF CORRUPTION UPON THE ROMANIAN ECONOMY Sorin-Daniel, Manole 1, Raluca, Erdniç 2 Absrac: This paper is dedicaed o esimaing he influence of corrupion upon Romania s economic growh by means of an economeric model ARMA componen. In order o quanify he impac, firsly some indicaors have been idenified o properly assess he economic condiion and corrupion. The mos imporan economic growh indicaor is real GDP growh rae (or chain index of real GDP). In order o esimae he level of corrupion, he auhors have used he Corrupion Percepions Index, annually launched and calculaed by Transparency Inernaional. The model chosen for his paper has an ARMA componen and expresses he dependence of he economic variable on he corrupion variable by a sraigh-line relaionship. The model shows ha one of he facors having a significan negaive impac upon he economic growh is corrupion. Key words: corrupion; Corrupion Percepions Index; real GDP; Romania. JEL Classificaion: D73, K42, E6. 1. Inroducion Corrupion is a social, economic and poliical issue spread all over he world o various exens. As well as oher conceps, when defining corrupion here is no inernaional consensus. In broad erms, corrupion is public power abuse for privae reasons and concerns (Lambsdorff, 2007, 16). According o he Explanaory Dicionary of he Romanian Language, corrupion is he sae of violaing moraliy, honesy and duy. Corrupion basically occurs in four main forms (Rohwer, 2009): bribery, embezzlemen, fraud and exorion. Bribery is undersood as he paymen ha is made or received in a corrup relaionship. These are all noions of corrupion in erms of he money paid or favours done o employees in privae enerprises, public officials and poliicians. Embezzlemen is he hef of resources by people who are responsible for managing hem. Fraud involves he manipulaion or disorion of informaion, facs and experise by public officials for heir own benefis. Exorion is money and oher resources exraced by he use of coercion, violence or hreas o use force. In order o esimae a counry s corrupion level, wha is mos frequenly used is: he Corrupion Percepions Index of he Transparency Inernaional non-governmenal organizaion and he Worldwide Governance Indicaors of he World Bank ha comprise six headings of which one is he Conrol of Corrupion. More recenly, new insrumens o measure corrupion have emerged such as: he Global Corrupion Baromeer of he Transparency Inernaional organizaion and he Global Inegriy Index of he Global Inegriy organizaion. Worldwide Governance Indicaors o assess he corrupion degree measure he level up o which public power is exered for privae gains, including low and high forms of corrupion, as well as "capuring" of he sae by elies and privae ineress. The Corrupion Percepions Index (CPI) laid down by he Transparency Inernaional every year assess he degree o which corrupion is perceived in he public 1 Associae Professor, Ph.D., Faculy of Managemen Markeing in Economic Affairs, Rm. Vâlcea, "Consanin Brâncoveanu" Universiy of Pieşi, danielsorinmanole@yahoo.com. 2 Juris, Ph.D., Managemen Auhoriy of General Programme on Solidariy and Managemen of Migraion Flows, ralucagpopa@yahoo.com. 4

2 secor of every counry. The index was launched in 1995, wih Romania having been assessed ever since I is a composie index ha is calculaed according o he informaion supplied by many sources, made up of daa regarding corrupion aken from specialized surveys underaken by famous independen insiuions and from scores prepared by expers (hazard assessmen expers, counry analyss). The Corrupion Percepions Index reflecs opinions from all over he world including hose of expers living in he counries already assessed. The aggregaion mechanism of informaion abou a counry s or erriory s corrupion level and urning he informaion ino a score included on a scoring scale from 0 o 10 or from 0 o 100, as in 2012, are complicaed. The 0 minimum value means a very high corrupion level and he 10 or 100 maximum value means he oal lack of corrupion. Once he Corrupion Percepions Index of he assessed counries has been announced, he Transparency Inernaional also inroduces a ranking prepared according o he descending measures of he index. The implicaions of his social issue are manifold as suggesed by so many empirical sudies in he specialized lieraure. Mo (2001) analyzes he relaionship beween corrupion and economic growh for 54 counries and he finds ha a 1% increase in he corrupion level reduces he growh rae by abou 0.72% and he mos imporan channel hrough which corrupion affecs economic growh is poliical insabiliy which accouns for abou 53% of he oal effec. Ali and Hodan (2003) indicae ha corrupion has a subsanial explanaory power for economic growh and higher economic growh does no guaranee lower corrupion in he fuure. Pellegrini and Reyer (2004) use sraigh-line regression models in order o quanify he effec of corrupion on economic growh, boh in a conex wih and wihou oher independen variables (invesmen, schooling, rade openness, poliical insabiliy). Using a sample of African counries in heir survey, d Agosino, Dunne and Pieroni (2012) assess how corrupion affecs he relaionship beween governmen spending and economic growh. 2. Corrupion in Romania The main causes riggering corrupion in Romania are: he lack of poliical will, he weakness of public insiuions, and radiion. Moreover, here has been fauly implemenaion of he seps aken over he years o decrease he scale of corrupion, which has led o such seps low efficiency (Radu and Gulyas, 2010). Corrupion leads o he inefficien allocaion of resources a naional level, lowers he adminisraive sysem s efficiency, holds back he economic growh and conribues in he poliical sysem s decay (Iamandi and Voicu-Dorobanţu, 2007). According o he Naional Anicorrupion Direcorae (2011), here is a high level of perceiving he connecions beween organized crime and corrupion. Organized crime and corrupion suppor each oher and when organized crime rules, he public secor corrupion is very likely o increase. According o he same survey, he mos imporan premises of he connecions beween organized crime and corrupion are ha: public adminisraion appoinmens do no occur due o meris, here are oo many links beween poliicians and he business environmen, and poliicians do no ac enough o sop he issue. In he year 2010, Romania go a 3.7 score, i was he 69 h in a general ranking (of 178 moniored counries) and he second o las in he ranking of European Union counries before Bulgaria (3.6) and Greece (3.5). In 2011, our counry s score was 3.6 and ranked he 75 h a world level, he same as China (of 182 moniored counries). I was he second o las in he ranking of European Union counries as i had he previous year, higher han Greece (3.4) and Bulgaria (3.3). 5

3 In he year 2012, 176 counries were ranked according o he Corrupion Percepions Index, and Romania had a 44-poin score as well as Saudi Arabia and Kuwai, ranking he 66 h. I was he firs year Romania had repored a significan increase of is score and ranking. I also advanced one more level in he European Union counries op, ranking before Ialy (42), Bulgaria (41) and Greece (36). Alhough survey ses and mehodology have changed over he years, he progress of he Corrupion Percepions Index during is ineresing. Figure 1 shows he progress of he Corrupion Percepions Index in Romania along wih he average progress in he European Union (for he year 2012 he index and he average have been divided by 10 in order o acquire uniform daa) Romania EU avarage Figure 1. Romania vs. European Union average Source: Prepared by he auhors based on he daa from Transparency Inernaional (hp:// By comparing Romania s Corrupion Percepions Index wih he EU average, i can be seen ha our counry s index during was below he average wih few excepions, ranging from 3.6 o 3.8 poins (3.23 he lowes in 1999, and 3.91 he highes in 2002). Alhough during Romania s difference from he EU average can be said o have remained quie consan, he same difference sared o significanly decrease in 2007 and ended up o 1.97 poins in Anoher concerning issue is he inerdependence beween corrupion expressed by he Corrupion Percepions Index and economic crime expressed by he number of economic and financial offences. The correlaion coefficien beween he ime series of he Corrupion Percepions Index and he number of economic offences in he case of Romania is r (Erdniç and Manole, 2013, pp ), hence beween corrupion and economic crimes here is a direc significan correlaion in erms of is inensiy. 3. Empirical Research 3.1. Seing Significan Variables for he Presen Survey In order o esimae he impac of corrupion upon he economic growh, one should firs find indicaors o express economic growh and corrupion as well as hey can. In erms of an economic growh index, he auhors have chosen he chain index of real GDP. The real value of a macroeconomic index is calculaed by comparing he index 6

4 nominal value o he corresponding price index. Addiionally, in order o assess he corrupion level, he auhors have chosen he Corrupion Percepions Index as i is he mos complee insrumen o measure corrupion. This paper includes he ime series of he chain index of real GDP in Romania for (Figure 2.) and he ime series of Romania s Corrupion Percepions Index for (Figure 1.) Figure 2. The Chain Index of Real GDP in Romania during Source: Prepared by he auhors based on processing he informaion aken from he Romanian Saisical Yearbook of and from web sie hp://epp.eurosa.ec.europa.eu/gm/able.do?ab= able&ini=1&plugin=1&language=en&pcode=ec00115 The wo series wih all heir erms being sricly posiive have been logarihmized. Therefore, inerpreing he coefficiens in he regressions compleed based on hese series is easier as hey are elasiciies. The necessary economeric assessmens have been performed by means of he EViews 7.0 programme package Saionarizing he Time Series An imporan aspec when analyzing he ime progress of a chronological series is saionariy. If a given series is no saionary, successive differencing is applied unil a saionary series is achieved (Dickey and Fuller, 1979). The bes-known surveys for he sudy of saionariy are Augmened Dickey-Fuller (ADF), Phillips-Perron (PP), Leybourne-McCabe (LM) and Kwiakowski-Phillips- Schmid-Shin (KPSS) (Balagi, 2008, pp ). Applying he ADF es means comparing he es values o he criical values corresponding o 1%, 5% and 10% significance hresholds. If he saisic value obained is higher han he criical value corresponding o a significance level, hen a null hypohesis is acceped which means he series is non-saionary. Oherwise, he chronological series is saionary. By he use of Augmened Dickey-Fuller, one can es he logarihmized series saionariy of he chain index of real GDP (LI_GDP) and i can be noiced ha he series is saionary (Table 1.). 7

5 Table 1. Resuls of ADF Uni Roo Tes for he Firs Difference of he Logarihm of GDP Variable -Saisic Prob.* Augmened Dickey-Fuller es saisic Tes criical values: 1% level % level % level *MacKinnon (1996) one-sided p-values. The same es helps show ha he naural logarihms series of he Corrupion Percepions Index (L_CPI) is no saionary (Table 2.), whereas is former difference (DL_CPI) proves he saionariy requiremen (Table 3.). Table 2. Resuls of ADF Uni Roo Tes for he Logarihm of CPI Variable -Saisic Prob.* Augmened Dickey-Fuller es saisic Tes criical values: 1% level % level % level *MacKinnon (1996) one-sided p-values. Table 3. Resuls of ADF Uni Roo Tes for he Firs Difference of CPI Variable Logarihm -Saisic Prob.* Augmened Dickey-Fuller es saisic Tes criical values: 1% level % level % level *MacKinnon (1996) one-sided p-values Model o Assess Corrupion Impac upon he Economy I is believed ha economic growh of a cerain ime is influenced by he economic growh of recen previous imes. Moreover, i is believed his here is a sraigh-line inerdependence so ha he model has an AR componen (auoregressive process). Furhermore, one can sar a he premise ha he process level during a cerain ime depends on he average deviaions in he pas, so ha here is also an MA componen (moving average process). Hence, he model has an ARMA componen. Therefore, here is only one explanaory variable signifying corrupion and he model form is y p q 0 i y i c j j i1 j 1 where = ime expressed in years; y = he firs differences of he logarihms of he real GDP a ime ; c = he firs differences of he logarihms of he CPI a ime ; (1) 8

6 i, j, = he unknown coefficiens, i 0,1,, p, j 1,2,, q ; 2 = error erms, variables ha are normally disribued wih mean 0 and variance. The ime series of he residual variable mus be saionary and mee he following requiremens: 2 2 E, E, ; 0 0 E s,,s, s. Afer several aemps, he order of AR pars of model p 6 and he order of MA pars q 2 have been chosen. The mos frequenly used mehods o esimae he parameers of economeric models are he leas squares mehod and he mehod of maximum likelihood. Solving his model has been achieved by he leas squares mehod Resuls and Analyses The informaion relaed o he esimaion of coefficiens and diagnosic saisics of he model is provided in he following able (Table 4.). Table 4. Esimaion Resuls and Diagnosic Saisics of he Model Dependen Variable: LI_GDP Mehod: Leas Squares Dae: 05/11/13 Time: 23:42 Sample (adjused): Included observaions: 15 afer adjusmens Convergence achieved afer 29 ieraions MA Backcas: Variable Coefficien Sd. Error -Saisic Prob. C LI_GDP(-1) LI_GDP(-3) LI_GDP(-6) DL_CPI MA(1) MA(2) 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 Hannan-Quinn crier F-saisic Durbin-Wason sa Prob(F-saisic) Invered MA Roos i i By replacing he esimaed values of he coefficiens, one can ge his equaion: y y y y c (2) The model coefficiens are significanly differen from zero (wih higher han 0.95 probabiliy), as he corresponding values of he significance level (he values in he Prob. column) are lower han 0.05, excep coefficien Since he corresponding ˆ3 9

7 probabiliy of his coefficien slighly exceeds he 0.05 level, being , i can be said ha 3 oo is significanly differen from zero. The raio of he explained variaion o he oal variaion, expressed by he coefficien of deerminaion (R-squared) is 80.39%. The adjused value of his coefficien almos having he same significance bu penalizing he occurrence of independen variables ha have lower relevance upon a dependen variable, is high enough (65.68%). The null hypohesis of he F-es (all he regression coefficiens are zero) has low probabiliy (Prob(F-saisic)), even lower han 5%, hence a leas some of he regression parameers are nonzero wih high probabiliy (more han 0.95). In addiion, he process mus be inverible, which means he inverse roos of he MA characerisic polynomial should lie inside he uni circle (Andrei e al, 2008, pp ). This requiremen is me as i can be seen in he las par of he able above, wih he inverse roos i and i The resuls obained afer applying such economeric ess lead o model validaion and admiance. The esimaed value of he coefficien of variable expressing corrupion ˆ shows ha a 1% decrease in he corrupion level resuls in % economic growh. The previous saemen can be inerpreed ha a 1% increase in he Corrupion Percepions Index leads o a % increase in he real GDP. I is esimaed he real GDP shall rise by 3.1% in 2013 and amoun o 649 billion lei (he Naional Prognosis Commission, 2012). Thus, as o he year 2013, a 1% decrease in he corrupion level namely an increase in he Corrupion Percepions Index from 44 o 45 migh resul in a GDP value equalling billion lei, ha is a GDP increase by million lei, which means million U.S. dollars (a he annual average exchange rae in 2013 of 1 US Dollar = lei). I proves ha he negaive influence of corrupion upon he economic growh is significan. 4. Conclusions The presen paper has researched he effec of corrupion upon Romania s economic growh. In order o assess he impac of corrupion upon he economic growh, he auhors have used an economeric model wih an ARMA componen, where he dependen variable is a logarihm of chain index of real GDP, and he only explanaory variable is he firs difference of he logarihms of he CPI. The es resuls reveal ha here is a saisically significan negaive relaionship beween corrupion and economic growh. Therefore, a 1% increase in he Corrupion Percepions Index leads o an increase in he real GDP by %. As o he year 2013, a 1% of he Corrupion Percepions Index (from 44 o 45) migh resul in a GDP increase of million U.S. dollars (a he annual average exchange rae in 2013 of 1 US Dollar = lei). As far as a more deailed sudy of he corrupion influence upon he economic growh is concerned, oher indices of he corrupion level should be equally considered (Worldwide Governance Indicaors, Global Corrupion Baromeer and Global Inegriy Index), as well as oher independen variables (invesmen, schooling, rade openness). 10

8 Besides, he effecs of such a complex social issue are manifold and is economic coss are immense, as shown by empirical surveys in he specialized lieraure (he majoriy based on muliple regression models). References 1. Ali, A. M., Hodan, S. I. (2003), Deerminans of Economic Corrupion: A Cross-Counry Comparison, Cao Journal, Vol. 22, No. 3, pp Andrei, T., Sancu, S., Iacob, A. I., Tuşa, E. (2008), EViews-Assised Basics of Economerics, Economic Publishing House, Buchares. 3. Balagi, B. (2008), Economerics, Springer-Verlag Berlin Heidelberg, 4h Ediion. 4. Naional Prognosis Commission (2012), Forecas of Main Macroeconomic Indices during , available a hp:// primavara_2012.pdf, (accessed on 22 November 2013). 5. D Agosino, G., Dunne, J. P., Pieroni, L. (2012), Governmen Spending, Corrupion and Economic Growh. A Souhern Africa Labour and Developmen Research Uni, Working Paper, No. 74, Cape Town: SALDRU, Universiy of Cape Town. 6. Dickey, D. A., Fuller, W. A. (1979), Disribuion of he Esimaors for Auoregressive Time Series wih a Uni Roo, Journal of he American Saisical Associaion, No. 74, pp Naional Anicorrupion Direcorae (2011), Survey on he Relaionship beween Organized Crime and Corrupion, available a (accessed on 25 Sepember 2013). 8. Erdniç, R., Manole, S. (2013), Organized Economic and Financial Crime Impac upon Growh. Case of Romania, Universiaria Publishing House, Craiova. 9. Iamandi, I., Voicu-Dorobanţu, R. (2007), Corrupion A Hazard o Romania in he European Union, The Romanian Economic Journal, Year 10, No. 24, pp Lambsdorff, J. G. (2007), The Insiuional Economics of Corrupion and Reform: Theory, Evidence, and Policy, Cambridge Universiy Press, Cambridge. 11. Mo, P. H. (2001), Corrupion and Economic Growh, Journal of Comparaive Economics, Vol. 29, No. 1, pp Pellegrini, L., Reyer, G. (2004), Corrupion s Effec on Growh and Is Transmission Channels, Kyklos, Vol. 57, pp Radu, L., Gulyas, G. (2010), Corrupion A Problem wihou Soluions in Romania?, Transylvanian Review of Adminisraive Sciences, Vol. 26, No. 2, pp Rohwer, A. (2009), Measuring Corrupion: A Comparison beween he Transparency Inernaional s Corrupion Percepions Index and he World Bank s Worldwide Governance Indicaors, CESifo DICE Repor 3/2009, available a hp:// pls/poral/docs/1/ pdf, (accessed on 9 November 2013). 15. hp://epp.eurosa.ec.europa.eu/gm/able.do?ab=able&ini=1&plugin=1&language=en&pcode= ec hp:// 11