European Economics and Finance Society (EEFS) Conference 14th annual EEFS Conference, Brussels, Belgium, June, 2015

Transcription

1 European Economics and Finance Sociey (EEFS) Conference 14h annual EEFS Conference, Brussels, Belgium, June, 2015 Granger Causaliy Relaionship beween Foreign Direc Invesmens, Expors, Unemploymen and Economic Growh. A Panel Daa Approach for he New EU Members. Pavlos Samaiou and Nikolaos Drisakis Deparmen of Applied Informaics, Universiy of Macedonia, Economics and Social Sciences, 156 Egnaia Sree, Thessaloniki, Greece dris@uom.gr Absrac. Foreign direc invesmens (FDI) and expors play an imporan role in economic developmen worldwide. However, here is a widespread belief among economiss ha economic growh canno be enhanced if unemploymen raes are very high. This paper applies panel uni roo ess, panel coinegraion and panel causaliy es o invesigae he relaionship beween expors, FDI, unemploymen and economic growh in hireen new European Union (EU) members (Bulgaria, Czech Republic, Esonia, Croaia, Cyprus, Lavia, Lihuania, Hungary, Mala, Poland, Romania, Slovenia, and Slovakia) using annual daa for he period The FMOLS and DOLS are used o esimae he long run relaionship beween he examined variables. The resuls of he sudy indicae ha, in he long run, here is bidirecional causaliy beween expors and economic developmen and a unidirecional causaliy beween economic developmen and unemploymen wih direcion from economic developmen o unemploymen. Also, he resuls show ha here is a shor run unidirecional causaliy beween expors and FDI running from expors o FDI. Findings sugges ha expors are a vial force for increasing economic developmen and aracing foreign direc invesmens. Moreover, he resuls reveal ha economic growh is an imporan deerminan for increasing expors and reducing unemploymen. Keywords: Economic developmen, Foreign direc invesmens, Expors, Unemploymen, Panel coinegraion, FMOLS and DOLS mehods, Causaliy. JEL Classificaion: C22, E31, E50 1. Inroducion Since 2004, hireen new counries (Bulgaria, Czech Republic, Esonia, Croaia, Cyprus, Lavia, Lihuania, Hungary, Mala, Poland, Romania, Slovenia, and Slovakia), mainly of Cenral and Easern Europe, have acceded o he European Union (EU) and oher six (Albania, Iceland, Monenegro, Serbia, The former Yugoslav Republic of Macedonia and Turkey) are on he road o EU membership. Wih he accession of hese counries, in 2013 he real Gross Domesic Produc (real GDP) in he EU reached approximaely billion euro (AMECO 2015). The oal GDP of he hireen new members does no exceed 7% of European Union s GDP.

2 EU membership is he sronges guaranee for hese counries o achieve higher growh raes, a sable business environmen and generally a sable growh pah. I consiues he rigger for direc invesmens and new capial inflows since invesors gain more confidence for hese economies. The accession in EU symbolizes heir reurn o he European family and esablishes a balanced economic environmen. Among he hireen new members, here are counries such as Poland, Czech Republic, Slovakia, Hungary wih significan indusrial base, which hope o arac foreign direc invesmens from he res EU members (EU-15). Many companies locaed in he developed EU counries decide o creae new producion unis in he new neighboring EU members due o he low-cos labor supply. The increase in invesmens and in domesic producs in he new European Union counries will increase he flows of foreign rade and will promoe heir expors. The economic crisis which sared in 2008 has creaed serious concerns abou high unemploymen raes and negaive growh raes in many EU counries. In 2013, unemploymen raes in Bulgaria, Croaia, Cyprus, Lavia, Lihuania and Slovakia reached 13.0%, 17.3%, 15.9%, 12.0%, 11.8% and 14.2 respecively (AMECO 2015). Many economiss believe ha foreign direc invesmens (FDI) and inernaional rade are he key facors for enhancing economic developmen and reducing unemploymen. FDI can play a vial role due o he fac ha enhances privae invesmens, encourages he creaion of new jobs and ransfers knowledge and echnological skills in he workforce. Today, we canno say ha here is a universal agreemen concerning he impac of FDI in hos counries economies. The impac of FDI on economic performance remains a popular subjec among policy makers especially for some developing counries. The purpose of his paper is o examine he links beween FDI, expors, unemploymen and economic growh in he hireen new EU members (counries ha joined he EU afer May 1s, 2004 such as Bulgaria, Czech Republic, Esonia, Croaia, Cyprus, Lavia, Lihuania, Hungary, Mala, Poland, Romania, Slovenia, and Slovakia) over he period The conribuion of his paper is shown in he following paragraphs: I is used a VAR model o examine he relaions among foreign direc invesmens, expors of goods and services, unemploymen and economic growh for he hireen new EU members, viewed as group. Mos of hese counries were under communis regimes unil 1989 and he las 20 years are rying o conver heir economies o marke economies wihin he insiuional framework of EU. For hese counries, EU membership is he sronges guaranee for he esablishmen of marke economies wih high growh raes similar o EU-15. To our knowledge very few sudies have aken ino consideraion hese four variables ogeher and have used he causaliy analysis. Also, here are even fewer sudies ha carried ou heir analysis in he panel framework. The srucure of he paper is as follows: Secion 2 briefly reviews he lieraure. Secion 3 presens daa and mehodology and secion 4 presens he empirical resuls. Concluding remarks are given in he final secion. 2. Lieraure Review The grea recession of 2008 proved he inerdependence ha exiss beween he economies of European counries. Under adverse condiions, capial flows and inernaional financial inerconnecions beween hem can cause desabilizaion of heir real economies (Thalassinos 2008). However, economic crisis did no have he

3 same impac in all counries. This depended on he infrasrucure and he level of he domesic economy of each counry. The effec of FDI on domesic economy, rade, and unemploymen has been a key subjec of inense research during he las years. There is a widespread belief among policy makers ha foreign direc invesmens inflows in combinaion wih expors are he mos imporan facors in he process of economic developmen of a counry. Denisia (2010) believed ha FDI affecs economic growh boh direcly and indirecly hrough expors. The posiive effec of FDI on economic performance appears o be greaer in developed economies, since hey can absorb more easily new echnologies (Blomsrom, Lipsey and Zejan 1992). De Mello (1997) suppored ha FDI enhances he producion process, increases capial adequacy and ransfers knowledge and echnology boh in erms of labor raining and skill acquisiion. Also, in an open economy, echnology and knowledge can be ransferred hrough expors and impors and consequenly he economic growh can be promoed (Grossman and Helpman 1997). However, he fac ha he posiive effec of FDI appears greaer in developed economies was criicized by Rodrik. Rodrik (1999) suppored ha developed economies arac more FDI because hey have higher growh raes, are more producive and more profiable economies. On he oher hand, he benefis of FDI have been dispued, especially in he pas, (Reuber, Crookell, Emerson and Gallais-Hamonno 1973, Lall, 1974) because of he shor-erm unemploymen i ofen causes and he high adjusmen cos of domesic companies in order o cope he new compeiive condiions. Lee and Vivarelli (2005) suppored ha even FDI and rade are expeced o have posiive effec on employmen level, he creaion of new lobs canno be auomaically assured, since he employmen effec can be very diverse in differen areas of he world. In previous year, sudies on he relaionship beween foreign direc invesmens, expors, unemploymen and economic growh, using eiher ime series or panel daa, reveal differen resuls which depend on economeric mehods and he period he analysis were carried ou. Chang (2005) examined he dynamic ineracions beween FDI, economic growh, expors and unemploymen for Taiwan using ime series daa over he period This sudy found ha boh economic growh and expors have a posiive impac on FDI inflows. On he oher hand expors have a negaive impac on FDI ouflows. Also he resuls show ha FDI inflows have a posiive impac on expors and economic developmen. In addiion, here is a posiive relaionship beween economic growh and expors and a negaive relaionship beween economic growh and unemploymen. In conclusion, expors no only arac FDI inflows and reduce FDI ouflows bu also enhance economic growh. Akar and Ozurk (2009) invesigaed he relaionship beween FDI, expors, unemploymen and GDP for Turkey using ime series daa over he period They found ha FDI does no creae new job for his period. However, expors arac more FDI o he counry. Also, he resuls show ha economic growh does no reduce unemploymen rae eiher. Findings sugges ha Turkey should arac more and beer FDI, in order o boos economic performance hrough echnology ransfer. Ozughalu and Ogwumike (2013) sudied he relaionship among unemploymen, economic growh, expors and foreign direc invesmens in Nigeria over he period The resuls show ha foreign direc invesmens, expors and economic growh canno solve he problem of unemploymen in he counry. The Nigerian governmen should immediaely implemen policies which will help he counry o break he yoke of underdevelopmen.

4 Regarding sudies ha examine he ineracions beween hese variables in a group of counries, Hsiao and Hsiao (2006) invesigaed he relaionship among FDI, expors and GDP for eigh rapidly Eas and Souheas Asian economies using panel daa over he period The panel daa causaliy resuls revealed ha FDI influences GDP boh direcly and indirecly hrough expors. Also, hey found ha here is bidirecional causaliy beween expors and GDP for he group. The resuls sugges ha growh in domesic producs and he large amoun of inward FDI are he wo vial forces in promoing expors for hese eigh economies as a group. Also, he findings suppor he expor-led growh and FDI-led growh model. Drisakis and Samaiou (2014) invesigaed he relaionship beween FDI, expors and GDP in five Eurozone counries using panel daa for he period They found ha here is bidirecional causaliy beween expors and economic developmen, while here is no causaliy beween economic growh and FDI nor beween FDI and expors. The resuls of his sudy show ha an increase in domesic producs of hese counries will cause a dynamic impulse in expors and economic growh. A similar sudy (see also Mahmoodi and Mahmoodi 2014) invesigaed he casual links among GDP, expors and FDI in wo panels of Asian counries (hree developed and eigh developing counries) during he period The resuls from developed panel show unidirecional causaliy from GDP o expors, FDI o expors and from GDP o FDI. Findings sugges ha GDP and FDI are vial forces for increasing expors and ha economic growh is an imporan deerminan for aracing FDI. Resuls of developing counries indicae a unidirecional causaliy from GDP o FDI and bidirecional causaliy beween GDP and expors which means ha economic growh is an imporan deerminan for promoing expors and aracing FDI in his group of counries. 3. Daa and Mehodology 3.1 Daa The variables ha are used in his sudy are gross domesic produc (GDP), expors of goods and services (EXP) and foreign direc invesmens inflows (FDI inflows) measured in consan 2005 US dollars, expressed in million and unemploymen (UN) expressed as a percenage of civilian labor force. The daa are annual covering he period Daa are gahered from economic daabases Annual Macro-Economic (AMECO 2014), World Developmen Indicaor (WDI 2014) and Unied Naions Conference on Trade and Developmen (UNCTAD 2014). 3.2 Mehodology In order o examine he relaionships beween foreign direc invesmens, expors of goods and services, unemploymen rae and economic growh we specify a vecor auoregressive model (VAR), as implemened by Sims (1980). The general form of he VAR model is expressed in equaion (1): U = f (GDP, EXP, FDI, UN) (1) where: GDP is he economic developmen, EXP are he expors of goods and services, FDI are he foreign direc invesmens inflows and UN are he unemploymen raes.

5 Afer he specificaion of he VAR model, we coninue analyzing panel uni roo ess, panel coinegraion and panel causaliy es. The FMOLS (Fully Modified Ordinary Leas Square) and DOLS (Dynamic Ordinary Leas Square) mehods are used o esimae he long run relaionship beween he examined variables. The panel VECM (Vecor Error Correcion Model) is used in order o find he causaliy beween foreign direc invesmens, expors of goods and services, unemploymen rae and economic growh Panel Uni Roo Tess In he firs sep we should examine if he four variables (GDP, EXP, FDI, UN) are saionary. In economeric lieraure, here are several approaches for uni roo ess in panel daa. Considering ha hese mehods may give differen resuls, we apply he ess suggesed by Breiung (2000), Levin, Lin and Chu (2002) (LLC), Im, Pesaran and Shin (2003) W-es (IPS), ADF-Fisher Chi-square es (ADF-Fisher), PP Fisher Chi-Square es (PP-Fisher), Maddala and Wu (1999), and Hadri (2000). In all hese cases excep Hadri, he null hypohesis is ha he variable conains a uni roo (i.e., i is no saionary) (For deails on he descripion of he above mehods see Drisakis and Samaiou 2014) Johansen-Fisher Panel Coinegraion Tes Since panel uni roo ess have been applied, our nex sep is o use he Johansen- Fisher panel coinegraion es, developed by Maddala and Wu (1999), in order o examine he long run relaionship beween he variables. This es is based on he coinegraion race and maximum eigenvalue ess by Johansen (1988). Johansen s mehodology akes as saring poin a vecor auoregression (VAR) οf order of p: p Y AY e (2) i i i 1 This model can also be wrien as shown below: p 1 Y Y Y u (3) 1 i i i 1 where: p p ( Aj ) and i j 1 j i 1 A j, wih i 1... p Boh race and maximum eigenvalue ess are relaed o he rank r of marix. For he race es, he null hypohesis is ha here are a mos r coinegraion vecors agains he alernaive of full rank r k coinegraion vecors (where k is he number of variables included in he model). The null hypohesis of maximum eigenvalue es remains he same as before, however he alernaive is ha here are exacly r 1 coinegraion vecors. Maddala and Wu (1999), using Fisher s es (1932), proposed an alernaive mehod for esing coinegraion in panel daa. This es (Johansen-Fisher) combines he individual cross-secion ess, giving a saisic concerning he whole panel. Le

6 p i be he p-value obained from he individual coinegraion es of cross secion i. Then, under he null hypohesis for he panel: N 2 pi 2N (4) i 1 2 log( ) where N is he number of cross-secions (Hamori and Hamori 2010) Panel FMOLS and DOLS esimaes Since our variables are coinegraed he nex sep is he esimaion of he long-run equilibrium relaionship. The ordinary leas square (OLS) mehod leads o a biased and inconsisen esimaor when applied o coinegraed panels (Kao and Chiang 2000). Therefore, we esimae he long-run relaionship using he fully modified OLS (FMOLS) esimaor proposed by Pedroni (1999, 2001) and he dynamic OLS (DOLS) esimaor suggesed by (Kao and Chiang 2000) and Mark and Sul (2002). These esimaors allow for greaer flexibiliy in he exisence of heerogeneiy of he coinegraing vecors (Pedroni 1999, 2000, 2001, 2004). Consider he following fixed effecs panel regression: y x u for i = 1 N members and = 1 T (5) i i i i where y i is a marix (1,1), is a vecor of slopes ( k,1) dimension, i is he individual fixed effec, u i are he saionary disurbance erms. x i ( k,1) vecor. assumed o be an inegraed process of order one for all i, where xi xi 1 ei FMOLS and DOLS esimaors are consruced making correcions for endogeneiy and serial correlaion o he OLS esimaor (Phillips 1995). The equaion given below presens he esimaed coefficien of FMOLS: N T 1 N T * * FM ( xi xi )( xi xi ) ( xi xi ) yi T (6) i 1 1 i 1 1 where * y is he ransformed variable of i y i in order o achieve he endogeneiy * correcion and is he serial correlaion error correcion erm. The DOLS is an exension of Sock and Wason (1993) esimaor. Following equaion is used o obain he DOLS esimaor: j q y x c x v (7) i i i ij i, j i j q Where c ij represens he lead or lag coefficien of explanaory variables a firs differences. The DOLS esimaor is obained as following:

7 N T 1 N T * DOLS zz i i zi yi i 1 1 i 1 1 (8) where zi [ xi xi, xi, q,..., xi, q ] is 2( q 1) 1 vecor of regressors. Kao and Chiang (2000) suppored ha he DOLS esimaor is less biased and has superior small sample properies compared wih FMOLS esimaor Panel Causaliy Analysis The exisence of coinegraion beween he examined variables implies ha here is causaliy relaion among hem in a leas one direcion (Granger 1969, Engle and Granger 1987, Granger 1988). In order o deermine he direcion of causaliy he Vecor Error Correcion Model (VECM) is esimaed. The equaions ha are used o es Granger causaliy are he following: i, 1, i 1,1, i, k i, k 1,2, i, k i, k 1,3, i, k i, k 1,4, i, k i, k k 1 k 1 k 1 k 1 1, i i, 1 1, i, p p p p GDP a GDP FDI EXP UN ECT u (9) 2, i i, 1 2, i, p p p p FDIi, a2, i 2,1, i, k GDPi, k 2,2, i, k FDIi, k 2,3, i, k EXPi, k 2,4, i, k UNi, k k 1 k 1 k 1 k 1 (10) ECT u 3, i i, 1 3, i, p p p p EXPi, a3, i 3,1, i, k GDPi, k 3,2, i, k FDIi, k 3,3, i, k EXPi, k 3,4, i, k UNi, k k 1 k 1 k 1 k 1 (11) ECT u i, 4, i 4,1, i, k i, k 4,2, i, k i, k 4,3, i, k i, k 4,4, i, k i, k k 1 k 1 k 1 k 1 4, i i, 1 4, i, p p p p UN a GDP FDI EXP UN ECT u (12) where Δ is he firs difference operaor, k=1,,p is he opimal lag seleced by he Schwarz, ECTi, 1 is he esimaed lagged error correcion erm derived from he long run coinegraion equaion, ji, is he adjusmen coefficien (j = 1, 2, 3, 4) and u j,, i is he disurbance erm assumed o be uncorrelaed wih zero means. 4. Empirical Resuls The mulivariae panel framework comprises annual daa concerning expors of goods and services, foreign direc invesmens inflows, unemploymen rae and gross domesic produc for hireen new European Union members. We begin by esing he saionariy of hese variables.

8 4.1 Panel Uni Roo Tess Resuls Applying he uni roo ess of LLC, Breiung, IPS, ADF-Fisher, PP-Fisher and Hadri we presen he resuls in able 1. We have performed each es for he level and firs difference of he variables. All ess were performed for boh cases of including individual effecs and individual effecs and rend. Table 1: Panel Daa Uni Roo Tess Resuls Panel Level Series LLC Breiung IPS ADF PP Hadri LLC Level Uni Roo Tes Resuls GDP FDI EXP UN (0.152) (0.752) (0.965) (0.014)** (0.002)*** (0.031)** and rend and rend and rend and rend and rend and rend (0.995) (0.990) (0.149) (0.996) (0.190) (0.996) (0.999) (0.571) (0.001)*** (0.004)*** (0.001)*** (0.005)*** (0.001)*** (0.914) (1.000) (0.115) (0.999) (0.083) (0.999) (0.351) (0.001)*** (0.542) (0.192) (0.095) (0.448) (0.990) (0.372) Firs Difference Uni Roo Tes Resuls GDP FDI EXP UN and rend

9 Panel 1 s Dif Series Breiung IPS ADF PP Hadri and rend and rend and rend and rend and rend (0.003)*** (0.025)** (0.380) (0.091) (0.156) (0.666) Noes: 1. Panel daa include all counries 2. The numbers in parenheses denoe p-values 3. ***, **, denoes rejecion of null hypohesis a he 1% and 5% level of significance, respecively. 4. The null hypohesis of hese ess is ha he panel series has a uni roo (nonsaionary series) excep wih he Hadri es which has no uni roo in panel series. 5. Lag lengh selecion auomaic based on Schwarz crierion. As can be seen for able 1 he resuls showed ha GDP and EXP conain a uni roo (non-saionary) in levels, while he resuls for he oher wo variables (FDI and UN) are inconclusive. Evidenly, he resuls indicaed ha all variables are saionary in heir firs differences (i.e. I(1)). 4.2 Panel Coinegraion Tes Resuls I is commonly acceped, according o he exising lieraure, ha wo series wih differen inegraion order canno form a coinegraed series. However, i is no as well known ha more han wo series, wih differen inegraion order, i is possible o form a coinegraed series of lower order of inegraion. For example, suppose he series x and y where x I (0) and y I (1). In his case we say ha x and y are no coinegraed. Bu in he case we have more han wo series, suppose x, y, z where x I (1), y I (2) and z I (2), hen y and z can form a coinegraed series of order one I (1), which hen can coinegrae wih x o give a I (0) series (Pagan and Wickens 1989) (see also Ramirez and Sharma 2008, Alexiou, Tsaliki, Tsoulfidis 2014). Following Harris (1995), wih n variables here can be a mos n-1

10 linearly independen coinegraing relaions. Considering all he above, we coninue applying Johansen s Fisher panel coinegraion es in order o examine he long run relaionship beween FDI, EXP, GDP and UN. Table 2 presens he resuls of Johansen s Fisher es. Table 2: Johansen-Fisher Panel Coinegraion Tes Resuls Johansen-Fisher Panel Coinegraion Tes and rend Fisher Saisic Probabiliy Fisher Saisic Probabiliy (race es) (max eigen es) None *** *** A mos *** *** A mos *** *** A mos Noes: The null hypohesis is ha he variables are no coinegraed. Under he null ess, all variables are disribued normal, N(0, 1). *** and ** significan a he 1%, 5% levels, respecively. Fisher s es (1932) applied regardless of he dependen variable. Lag inervals for es: 1 1. Asympoic p-values are compued using X 2 disribuion. The resuls of able 2 (race es saisics and maximum eigenvalue saisics) suppor he presence of hree coinegraion vecors a 1% level of significance. We conclude ha here is a srong evidence of coinegraion among he examined variables. 4.3 Panel Coinegraion Esimaion Resuls Since ou variables are coinegraed, we proceed by esimaing he parameers of he long-run equilibrium relaionship. The resuls of Fully Modified OLS (FMOLS) and Dynamic OLS (DOLS) esimaions are provided in able 3. Table 3: Panel FMOLS and DOLS Esimaions Resuls for he Group of Counries (GDP as dependen variable) FMOLS DOLS Independen Variables Independen Variables EXP FDI UN EXP FDI UN Coeff (43.44)*** (4.63)*** (-8.73)*** (28.62)*** (4.33)*** (-8.21)*** R Adj. R Noes: The numbers in parenheses denoes -saisic. Asympoic disribuion of -saisic is sandard normal as T and N go o infiniy. *** significan a 1% level. Lag and lead mehod seleced by Akaike in DOLS. Table 3 presens he esimaed parameers from he long-run equaion where GDP is he dependen variable. In he oher cases he signs do no agree wih he economic heory and he coefficiens are no saisically significan.

11 The esimaion resuls shows ha expors as well as foreign direc invesmens have a posiive effec on economic growh in 1% level of significance. Also, unemploymen has a negaive impac on economic growh in 1% level of significance. 4.4 Panel VECM Causaliy Tess Resuls The resuls on panel daa for shor and long run causaliy relaionships are repored in Table 4. Table 4: Panel Causaliy Tess Resuls Dependen Source of Causaion (independen variables) -es Variable F-saisic Shor-run Long-run DGDP DEXP DFDI DUN ECT DGDP 2.915* *** DEXP 5.514** *** DFDI ** DUN 2.845* ** Noes: Δ denoed he firs difference operaor, ***, ** and * significan a he 1%, 5% and 10% levels, respecively. Shor-run causaliy is deermined by he saisical significance of he parial F-saisics associaed wih he righ hand side variables. Long-run causaliy is revealed by he saisical significance of he respecive error correcion erms using a -es. From he resuls of able 4 we see ha he esimaed coefficiens of ECT in equaions of DGDP, DEXP and DUN are saisically significan a 1% and 5% level, implying ha gross domesic produc, expors of goods and services and unemploymen could play an imporan adjusmen role in he long run equilibrium. In paricular, from he resuls of able 4 for he hireen counries viewed as a group we see ha here is a long run bidirecional causaliy relaion beween expors and economic developmen and a long run unidirecional causaliy beween economic developmen and unemploymen running from economic developmen o unemploymen. Also he resuls show ha, in he shor run, here is a unidirecional causaliy beween expors and FDI wih direcion from expors o FDI (see Figure 1). The knowledge abou he direcion of causaliy helps policy maker o develop a proper economic policy. Figure 1: Panel Daa Granger Causaliy Relaions for Thireen Counries

12 5. Conclusion and Policy Implicaions The European Union (EU) will coninue growing. The las few years we see ha more and more counries are ineresed in joining he EU. The accession in European Union symbolizes heir reurn o European family and esablishes a sable growh pah. EU membership is he sronges guaranee for hese counries o achieve recovery of heir economies, higher growh raes and a sable business environmen. This sudy invesigaes he causal relaionships among foreign direc invesmens, expors of goods and services, unemploymen and economic growh in hireen new European Union (EU) members (Bulgaria, Czech Republic, Esonia, Croaia, Cyprus, Lavia, Lihuania, Hungary, Mala, Poland, Romania, Slovenia, and Slovakia) using annual daa for he period We apply panel uni roo ess, panel coinegraion es and dynamic panel causaliy es wih error correcion model. Findings sugges ha here is a srong evidence of coinegraion among he examined variables, which indicaes ha he here is long run equilibrium relaionship. The FMOLS and DOLS mehods are used o esimae he long run relaionship beween he examined variables. The esimaion resuls, for he group of counries, show ha expors as well as foreign direc invesmens have a posiive effec on economic growh in 1% level of significance. Also, unemploymen has a negaive impac on economic growh in 1% level of significance. Then, an error correcion model was used o capure he shor run and long run dynamic relaionships. The obained resuls indicae ha in he long run, here is bidirecional causaliy beween expors and economic developmen and a unidirecional causaliy beween economic developmen and unemploymen wih direcion from economic developmen o unemploymen. Also, he resuls showed ha here is a shor run unidirecional causaliy beween expors and FDI running from expors o FDI. Findings sugges ha expors are a vial force for increasing economic developmen and aracing foreign direc invesmens. Moreover, he resuls reveal ha economic growh is an imporan deerminan for increasing expors and reducing unemploymen. An increase in domesic producs of hese counries will promoe expors and will help in reducing unemploymen. Moreover, long run dynamics show ha expors have an indirec effec on unemploymen hrough economic growh. Therefore, he governmens of hese counries should implemens policies o simulae heir expors. Expors no only no only arac FDI and enhance economic developmen bu also help in reducing unemploymen (hrough GDP). References Akar I. and Ozurk L. (2009): Can Unemploymen be Cured by Economic Growh and Foreign Direc Invesmen in TURKEY?, Inernaional Research Journal of Finance and Economics, 27, Alexiou C., Tsaliki P. and Tsoulfidis L. (2014): Classical Theory of Invesmen. Panel Coinegraion Evidence from Thireen EU Counries, Munich Personal RePEc Archive (MPRA) Paper, No Annual Macro-Economic Daabase - AMECO (2014), European Commission. Annual Macro-Economic Daabase - AMECO (2015), European Commission. Blomsrom M., Lipsey R.E. and Zejan M. (1992): Wha Explains Developing Counry Growh?, Naional Bureau of Economic Research (NBER) Working Paper, No

13 Breiung J. (2000): The Local Power of Some Uni Roo Tess for Panel Daa, in Balagi, B.(ed), Advances in Economerics, Non-saionary panels, panel coinegraion, and dynamic panels, Amserdam, JAI Press, Chang S.C. (2005) The Dynamic Ineracions among Foreign Direc Invesmen, Economic Growh, Expors and Unemploymen: Evidence from Taiwan, Economic Change and Resrucuring 38, De Mello L.R. (1997): Foreign Direc Invesmen in Developing Counries and Growh: A Selecive Survey, Journal of Developmen Sudies 34, 1, Denisia V. (2010): Foreign Direc Invesmen Theory: An Overview of he Main FDI Theories, European Journal of Inerdisciplinary Sudies 2, 2, Dickey D.A. and Fuller W.A (1979): Disribuions of he Esimaors for Auoregressive Time Series wih a Uni Roo, Journal of American Saisical Associaion, 74, 366, Dickey D.A. and Fuller W.A (1981): Likelihood Raio Saisics for Auoregressive Time Series wih a UniRroo, Economerica, 49, 4, Driakis N. and Samaiou P. (2014): Expors, Foreign Direc Invesmen and Economic Growh for Five European Counries: Granger Causaliy Tess in Panel Daa, Applied Economics Quarerly, 60, 4, Engle R.F., Granger C.W.J. (1987): Coinegraion and Error Correcion: Represenaion, Esimaion, and Tesing, Economerica, 55, 2, Fisher R.A. (1932): Saisical Mehods for Research Workers, Olliver and Boyd, Edinburgh, 4 h Ediion. Granger C.W.J. (1969): Invesigaing Causal Relaionships by Economeric Models and Cross-Specral Mehods, Economerica, 37, 3, Granger C.W.J. (1988): Some Recen Developmen in a Concep of Causaliy, Journal of Economerics, 39, 1-2, Grossman G.M. and Helpman E. (1997): Innovaion and Growh in he Global Economy, Cambridge, Massachuses: MIT Press. Hadri K. (2000): Tesing for Saionariy in Heerogeneous Panel Daa, Economerics Journal, 3, 2, Hamori S. and Hamori N. (2010): Inroducion of The Euro and he Moneary Policy of he European Cenral Bank, World Scienific, Singapore, Hackensack. Harris R. (1995): Using Coinegraion Analysis in Econoeric Modelling, Prenice Hall Publishing, Essex, England. Hsiao F.S.T. and Hsiao M.W. (2006): FDI, Expors, and Growh in Eas and Souheas Asia: Evidence from Time-Series and Panel Daa Causaliy Analyses, presened a 2006 Inernaional Conference on Korea and he World Economy V, Seoul, Korea. Im K.S., Pesaran M.H. and Shin Y. (2003): Tesing for Uni Roos in Heerogeneous Panels, Journal of Economerics, 115, 1, Johansen S. (1988): Saisical Analysis of Coinegraion Vecors, Journal of Economic and Conrol, 12, 2-3, Kao C. and M.H. Chiang (2000): On he Esimaion and Inference of a Coinegraed Regression in Panel Daa, Advances in Economerics, 15, Lall S. (1974): Less Developed Counries and Privae Foreign Direc Invesmen. A Review Aricle, World Developmen, 2, 4-5, Lee E. and Vivarelli M. (2005): Undersanding Globalizaion, Employmen and Povery Reducion, Indusrial and Labor Relaions Review, 59, 1. Levin A., Lin C.F. and Chu J. (2002): Uni Roo Tes in Panel Daa: Asympoic and Finie-Sample Properies, Journal of economerics, 108, 1-24.

14 Maddala G.S. and Wu S. (1999): A Comparaive Sudy of Uni Roo Tess wih Panel Daa and a New Simple Tes, Oxford Bullein of Economics and Saisics, 61, S1, Mahmoodi M. and Mahmoodi E. (2014): FDI, Expors and Growh: Evidence from Asian Counries, Journal of Educaional and Managemen Sudies, 4, 2, Mark N.C. and Sul D. (2002): Coinegraion Vecor Esimaion by Panel DOLS and Long-Run Money Demand, Naional Bureau of Economic Research (NBER) Working Paper, No Ozughalu U.M. and Ogwumike F.O. (2013): Can Economic Growh, Foreign Direc Invesmen and Expors Provide he Desired Panacea o he Problem of Unemploymen in Nigeria?, Journal of Economics and Susainable Developmen, 4, 1. Pagan A.R. and Wickens M.R. (1989): A Survey of Some Recen Economeric Mehods, Economic Journal, 99, 398, Pedroni P. (1999): Criical Values for Coinegraion Tess in Heerogeneous Panels wih Muliple Regressors, Oxford Bullein of Economics and Saisics, 61, Pedroni P. (2000): Fully Modified OLS for Heerogeneous Coinegraed Panels, Advances in Economerics, 15, Pedroni P. (2001): Purchasing Power Pariy Tess in Coinegraed Panels, Review of Economics and Saisics, 83, 4, Pedroni P. (2004): Panel Coinegraion: Asympoic and Finie Sample Properies of Pooled Time Series Tess wih an Applicaion o he PPP Hypohesis, Economeric Theory, 20, 3, Phillips P.C.B. (1987): Time Series Regression wih a Uni Roo, Economerica, 55, 2, Phillips P.C.B. and Perron, P. (1988): Tesing for a Uni Roo in Time Series Regression, Biomerika, 75, 2, Phillips P.C.B. (1995): Fully Modified Leas Squares and Vecor Auoregression, Cowles Foundaion for Research in Economics Paper No. 905, Yale Universiy. Ramirez M.D. and Sharma H. (2008): Remiances and Growh in Lain America: A Panel Uni Roo and Panel Coinegraion Analysis, Economics Deparmen Working Paper, No. 51, Yale Universiy. Reuber G.L., Crookell H., Emerson M. and Gallais-Hamonno G. (1973): Privae Foreign Invesmen in Developmen, Oxford: Clarendon Press. Rodrik D. (1999): The New Global Economy and Developing Counries: Making Openness Work, Policy Essay No. 24, Overseas Developmen Council, Washingon, D.C. Sims C. (1980): Macroeconomic and Realiy, Economerica, 48, 1, Sock J.H. and Wason M.W. (1993): A Simple Esimaor of Coinegraing Vecors in Higher Order Inegraed Sysems, Economerica, 61, Thalassinos E.I. (2008): Trends and Developmens in he European Financial Secor, European Financial and Accouning Journal, 3, 3, Unied Naions Conference on Trade and Developmen UNCTAD (2014), Geneva. World Developmen Indicaors The World Bank (WDI) (2014), Washingon.