Asian Research Consortium

Asian Research Consortium Asian Journal of Research in Business Economics and Management Vol. 4, No. 11, November 2014, pp. 4662. ISSN 22497307 Asian Journal of Research in Business Economics and Management www.aijsh.org Determinants of Economic Growth in GCC Economies Mohamed Attaitalla Abdalla*; Hisham H. Abdelbaki** *Assistant Professor, University of Bahrain, Bahrain. **Professor Mansoura University, Egypt. Assistant Professor, University of Bahrain, Bahrain. DOI NUMBER10.5958/22497307.2014.00972.4 Abstract The paper investigates determinants of economic growth in GCC countries using VECM approach. The results report that in a 10year horizon, FDI accounts 15.87%, 10.29%, 16.58%, 0.89% and 6.88% for Bahrain, Kuwait, Qatar, Saudi Arabia and United Arab Emirates respectively in shocks in their economic growth compared to the contributions of exports (9.55%, 1.33%, 6.85%, 10.38% and 20.9%) and gross capital formation (0.83%, 0.07%, 9.88%, 1.98 and 17.98%) for the countries respectively. The results also show that the main determinants of economic growth are foreign direct investment and gross capital formation for Bahrain. For Kuwait, Qatar and Saudi Arabia, exports and gross capital formation are the main determinants of economic growth. Exports and foreign direct investment are the main determinants of economic growth in United Arab Emirates. Finally, there is no any evidence for shortrun or longrun unidirectional or bidirectional causality relationship for Oman. Keywords: Economic Growth, FDI, Exports, Investment, VAR, VECM, GCC Economies. 1. Introduction In order to accelerate their development and integrate more closely their economies, the governments of the GCC countries formed in 1981 the Gulf Cooperation Council (GCC) for Arab Gulf States. The present membership of this council consists (in alphabetical order) of Bahrain, 46

Vol. 4, No. 11, pp. 4662. Kuwait, Oman, Qatar, Saudi Arabia (KSA) and United Arab Emirates (UAE). The economies of these states witnessed major transformations since the early of 1970s. Table (1) shows some basic information about GCC countries. It illustrates the relative position of Saudi Arabia among the GCC member states in terms of land area and population size. Oman held a distant second position in this respect. In terms of GDP, UAE, Qatar and Kuwait held second positions. The six GCC countries hold about 39.5% of world oil reserves, with Saudi Arabia, Kuwait and UAE holding more than 90% of the region s oil reserves. Qatar, however, holds the largest natural gas reserves in the region, accounting to about 13.1% of world natural gas reserves. Jointly, Qatar, Saudi Arabia and UAE hold more than 90% of the region s natural gas reserves while together the six states account for about 22% of world natural gas reserves. The existence of oil and gas brought many similarities in the economies of these countries since oil constitutes the main source of exports and the main government revenue, with nearly 70% of the region s exports being oil exports while about 83.4% of government revenue being from oil. This large energy reserves provided the member countries with a most important source of comparative advantage, while oil revenues have made it possible for them to reach their present stage of economic development, including an underlying massive infrastructure. Table 1: Selected Indicators of the GCC Region and Member States, 2012 Indicator Bahrain Kuwait Oman Qatar KSA UAE GCC Area 000 0.7 17.8 309.5 11.5 2,250 83.6 2,673.1 km 2 Population 1,248.3 3,188.8 3,623 1,836.7 29,195.9 8,458 47,550.7 000 Nominal GDP 30,362.1 183,236 78,110.8 192,403.9 711,998.9 383,799 1,579,910.8 Mil $ Per capita 24,322.7 57,462.3 21,449.7 104,755.2 24,386.9 45,377 33,225.8 GDP Oil reserves 120 101,500 5,500 25,244 265,850 97,800 496,014 Mill Bar. Oil Res % of 0.0 8.1 0.4 2.0 21.2 7.8 39.5 world Gas reserves 92 1,784 950 25,202 8,151 6,091 42,270 B Cub M gas % of 0.0 0.9 0.5 13.1 4.2 3.2 22.0 world Oil % of 77.7 93.7 77.5 88.4 87.1 34.5 69.9 exports Oil % of total revenue 87.3 91.5 86.8 69.6 85.2 81.9 83.4 Source: Gulf Statistical Profile Gulf Organization for Industrial Consulting (GOIC), 2013. The large oil price increases of early 1970s marked a decisive turning point in the economies of the GCC countries. The large oil revenues resulting from these increases provided the resources needed to lay the foundations of a modern economic structure, and led to large amounts of spending which was directed to socioeconomic development and infrastructural project. This was aimed to lead over the years to the establishment of a fairly advanced nonoil sector which can 47

Vol. 4, No. 11, pp. 4662. constitute a second major source of income beside oil revenues, and accelerate economic growth in these countries. The main objectives of the paper are to investigate the main determinants of economic growth in GCC countries, to examine the relative strength of each determinant in explaining the change(s) in economic growth and to assess the causality relationship between these determinants. Apart from this introduction, this paper contains 6 more sections. Section 2 provides an overview of the GCC economies. Literature review is discussed in section 3. Section 4 presents the data and methodology used in this paper. The main empirical findings are analyzed in section 5, while concluding remarks are summarized in section 6. 2. GCC Economies: An Overview The importance of oil to GCC economies is seen in its contribution to Gross Domestic Product (GDP). However, to discuss the structure of these economies and their growth performance we divide the economy into two main sectors, namely the oil and nonoil sectors. By construction, the percentage shares of the two sectors in GDP add up to 100. To provide more in depth analysis of the nonoil sector, this sector is disaggregated into the public and private services sectors. The private services sector includes trade, restaurants, hotels, transportation, storage, housing, finance, insurance and banking. The public services sector includes the services which are traditionally provided by the government, beside the supply of water and electricity. The relative importance of these sectors is assessed in terms of their value added or contributions to the GDP. The statistical analysis is based on annual time series data covering the 19732012 periods. This entire time span provides a longterm view of the sectoral contributions to GDP and its growth rates. A shorterterm view is provided by four decades: 19731982, 19831992, 19932002 and 20032012. The sectors shares in GDP are shown in Figure 1 below. 48

Vol. 4, No. 11, pp. 4662. Figure 1: GCC Countries Sectors Shares in GDP Figure 1 shows that GDP was more than 50% in all member countries, except for Bahrain and UAE, during the first decade of 19731982. However, since then this share witnessed steady declines over all the reference time periods, pointing to a weakening in this sector's dominant position. The last period 200212, however, witnessed an increase in the contribution of this sector to GDP, but this is far lower than its share in the first decade 197382. The two decades following 1982 have seen the large decrease in oil prices of 198286 as well as the large increase in these prices following 1987. The two decades also witnessed considerable political events following the invasion of Kuwait by Iraq in 1990, and the subsequent two Gulf wars that followed. Such big events would certainly affect the stability of economies of these countries and might affect level of economic activity. This however is expected to affect all sectors of the economy and not only the oil sector. Moreover, the increase and decrease in oil prices within this period may cancel each 49

Vol. 4, No. 11, pp. 4662. other which indicate that the above estimates of oil shares in GDP are fairly reasonable. On the other hand, the share of nonoil sectors in GDP increased considerably during this period. Starting from the decade of the 1980s, the share of this sector was higher than share of oil for all states, except for Kuwait and Qatar during the last decade of 20022012. In Bahrain the private services sector contribution to GDP was higher than oil contribution for all time periods. The private services sector contribution in UAE was also higher than oil contribution during the last three decades. The contribution to GDP of manufacturing industry stayed at one digit percent except in Bahrain where it recorded two digit levels. Agriculture sector had the lowest share in GDP as expected, since it is constrained by unfavorable factor endowment. The two sectors are dropped in this analysis, but they would show in the nonoil sector. In general, noting the increasing value added to GDP over these four decades, the analysis suggests that the nonoil sector is gaining a leadership position in all GCC states. However, also noting the high contribution of oil to both exports and government revenue, as noted in section 1, the nonoil sector may need to take longer time to become a leading sector in this regard. To focus more on the performance of the economy, we analyze the growth rates of Real GDP and the main sectors of the economy as presented in Figure 2 below, where the yeargrowth rates of each sector are averaged over each decade and for each member state. Figure 2: GCC Countries Sectors average growth rates 50

Vol. 4, No. 11, pp. 4662. Over the first decades, 197382, in all countries except for Bahrain, figure 2 shows broadly similar growth rates for GDP and Oil sector. This reflects the dominant position of oil in the economies of the GCC countries. The impact of oil price fluctuations during the second decade 19831992 was clear on the growth rates of both GDP and oil sector where they both witnessed single digit growth rates and even negative rates in some countries., except in Kuwait where oil production was drastically increased in 1992 after the second Gulf war. However during this decade growth rates of GDP, although remained at single digit level, were higher than the oil growth rates, with negative GDP growth rate in Qatar only. This suggests that the impact on GDP of this decline in oil prices might have been offset by the stabilizing effect of the nonoil sector which recorded higher growth rates than the oil sector. However, the dominance of oil continued in the third and fourth decades with oil growth rates higher than that of GDP. This pattern continued over the entire 19732012 time period, the growth rates in oil was moderately higher than that of GDP growth rate. The first decade can be viewed as the period during which the nonoil sector was being established and the second decade as the period during which it became operative. Within this perspective, the changes in GDP in the second decade can be attributed to the growing contributions of the nonoil sector and the declining contributions of the oil sector. This raises the question what are the factors that determine economic growth in GCC countries. 3. Literature Review The issue of determinants of economic growth is long debated in literature at both theoretical and empirical levels. The discussion starts with the classical theory of Adam Smith and goes to neoclassical Solow model which emphasize the role of labor, capital, and technology in economic growth. A recent development in economic growth theories is the endogenous growth theory, developed and tested by Barro (1991), where growth is taken as an endogenous product rather than being determined by external factors outside the system. Empirical studies, on the other hand, have investigated a wide range of factors that determine economic growth. Using different methodology and conceptual framework, these studies identified a considerable number of factors affecting economic growth. Among variables investigated are saving and investment which were investigated in many studies. Foreign Direct Investment (FDI) has recently played a crucial role as a primary source of technology transfer and economic growth. FDI can influence economic growth in two ways; direct and indirect ways. The direct way in which foreign investments affect growth is by increasing production, employment, value added and exports. The indirect way is through the transition of technology and knowledge via license, imitation and job training which will lead to increase in GDP. Moreover, FDI and GDP can be positively correlated through strong and worldwide extended financial markets. The role of FDI as an important determinant of economic growth was also extensively examined. Openness to trade was also used as determinant of economic performance. Other variables investigated include human capital, research and development (R & D), as well as economic policies and institutional aspects in the country. Blomstoem, et al (1994) studied the reasons of economic growth in developing countries and found a unidirectional causal relationship between FDI inflows as a percentage of GDP and the growth of per capita GDP for all developing countries over the period 19601985. Using Granger causality analysis for ten Asian countries, Zhang (1999) examined the 51

Vol. 4, No. 11, pp. 4662. relationship between FDI and economic growth. The study found a unidirectional causality between FDI and economic growth with direction from FDI to GDP in Hong Kong, Japan, Singapore, Taiwan, a unidirectional causality between exports and economic growth with direction from economic growth to exports for Malaysia and Thailand. A bilateral causal relationship is found between FDI and GDP for Hong Kong and Indonesia. Moudatsou (2003) stated that FDI inflows have a positive effect on economic growth in European Union countries both directly and indirectly through trade reinforcement over the period 19801996. Dritsakis (2004) analyzed the relationship between exports, investments and economic development for Bulgaria and Romania. The results reported a strong Granger causal relation between economic growth and exports as well as investments and exports for the two countries. Ndambiri, H. K. at, al (2012) used a panel data of 19 subsaharan Countries to explore the determinants of economic growth. The results state that the physical capital formation, a vibrant export sector, and human capital formation contribute significantly to the economic growth. However, government expenditure, nominal discount rate and foreign aid have a negative effect on economic growth. Stylianou, T., (2014) investigated the relationship between economic growth, FDI and exports for the US economy. The results showed the presence of one cointegrating vector. Also the results stated the causality relationship between exports and FDI running from exports to GDP and FDI, and also from GDP to FDI. 4. Data and Methodology The paper seeks to study the determinants of economic growth in GCC countries and the relative effect of each of them in changes of economic growth. The data about exports, GDP and gross capital formation are from Gulf Organization for Industrial Consulting. Data about FDI are from World Development Indicators. Dependent Variable In this paper, economic growth is used as a dependent variable for the six GCC countries. Independent Variables Based on literature, the main factors determine economic growth are exports, foreign direct investment and gross domestic investment which are used in this paper. To investigate the impact of the independent variables on the dependent one, the following model is used: GR ic 0 1EXP 2FDI 3GFC t (1) Where GR is growth rate of GDP, EXP = exports as a ratio to GDP, FDI = foreign direct investment as a ratio to GDP, and GFC = gross capital formation as a ratio to GDP. All variables are adjusted by GDP deflator and all are annual covering the period 19802007. Empirical approach in this paper is based on methods of cointegration and Vector Auto Regression Model (VAR). In matrix notation, the VAR model for m variables can be expressed by the following: (2) 52

Vol. 4, No. 11, pp. 4662. Where, and A 1, A 2,. A s are matrix, = a m dimensional vector of errors with E ( ) = 0. In reduced form, the VAR model expressed as follows: Where, = the lag operator, = a vector consisting of appropriately transformed variables, and = a vector of innovations of these variables. 5. The Empirical Results To achieve the objectives of the paper, we employ the recently developed techniques of time series data cointegration. The empirical investigations start by examining the basic series properties to the data. Thomas R. L. (1997) states that classical regression techniques become invalid, if applied to variables that do not meet the definition of stationarity. Consequently, using data through the period 19802007, we, first, test to determine if the series are nonstationary, by employing Augmented DickerFuller () and PhillipsPerron () Tests, where the null hypothesis is that a series has a unitroot. If all of the series are stationary then model them in level form. However, if more than one series are nonstationary then the author will test to determine if the nonstationary series are integrated by applying the Trace and JohansenJuselius maximum likelihood method of cointegratrion (Granger, 1987 & Johansen, 1990). If they are not found to be so, then model them in differenced form. However, if the nonstationary series are cointegrated then the author will model them in the ECM form. 5.1. Unit Roots To clarify the relationship between included variables, the VAR model is used. This requires that the timeseries analyzed be stationary. Table (2) reports the unit roots tests based on the commonly Augmented Dickey Fuller and PhillipsPerron () procedures. The lag for each variable is presented for both indexes. The tests are conducted with and without timetrend. The tests results show that we cannot reject the null hypothesis of presence of a unit root for all variables, implying nonstationarity in the level for some variables. However, we strongly reject the null hypothesis for the existence of a unit root for the first difference series indicating stationarity. The results state that all firstdifference series are stationary. (3) 53

Vol. 4, No. 11, pp. 4662. Table (2): Augmented DickeyFuller () and PhillipsPerron () Tests Results Variables Level BAHRAIN BGR 0 4.116934** 4.047145** 1 First Difference 6.856293*** 15.37980*** EXP_GDP 0 2.491167 2.673468 0 5.694503*** 5.678253*** FDI_GDP 0 1.635918 1.635918 0 4.694381*** 4.678640*** GFC_GDP 4 0.146219 0.869469 0 3.736737** 3.736308** KUWAIT Level First Difference Variables BGR 1 5.386541*** 11.10319*** 3 4.714923 19.86037*** EXP_GDP 5 4.163171** 3.045868 5 2.153127 7.391440*** FDI_GDP 6 2.656316 0.688299 2 3.119939 GFC_GDP 1 4.413041*** 3.750816** 1 Variables Level OMAN 5.095882*** First Difference BGR 0 6.460473*** 9.825837*** 6 2.796584 EXP_GDP 2 2.342139 2.008542 6 4.008057** FDI_GDP 0 1.784024 1.674325 0 5.964747*** GFC_GDP 0 2.046151 2.046151 6 1.670960 Variables Level Qatar First Difference BGR 0 5.684477*** 11.59793*** 3 4.185647** 7.783112*** 8.937797*** 20.50302*** 13.35749*** 6.436198*** 4.928238*** 14.14236*** 4.646361*** EXP_GDP 1 1.578892 1.536469 0 4.631612*** FDI_GDP 1 2.206238 2.024418 0 4.058035** 3.952930** GFC_GDP 5 3.007655 2.474455 0 54

Vol. 4, No. 11, pp. 4662. Variables Level KSA 4.987692*** 5.242446*** First Difference BGR 0 5.284721*** 5.468439*** 3 3.547653** EXP_GDP 1 4.849687*** 3.592209** 1 4.039093** FDI_GDP 0 3.895844** 3.711374** 0 GFC_GDP 1 3.672564** 3.174276 0 Variables Level UAE BGR 0 6.003016*** 6.003016*** 0 EXP_GDP 4 1.723749 1.849395 0 FDI_GDP 1 4.421993*** 2.862315 2 GFC_GDP 6 3.908916** 2.764521 0 Note: *** and ** denote significance at the 1% and 5% levels respectively. 5.2. Test of Cointegration 4.812993*** 4.908768*** First Difference 9.494831*** 7.059988*** 5.325616*** 8.158240*** 10.92208*** 4.572990*** 3.954691** 5.567719*** 15.00007*** 9.380207*** 5.595667*** 8.158240*** To examine whether a longrun equilibrium exists amongst the independent variables and economic growth for each GCC country, the Johansen cointegration tests at the chosen lag level is conducted. If the series are cointegrated, this would imply that the series share a common stochastic trend and any deviations from longrun equilibrium is likely to lead to shortrun adjustment or realignment of the series to restore equilibrium. Table (3) states that only Oman in favor of zero cointegrating vector. Other countries experience mix evidence in number of cointegrating vectors. Table (3) Cointegration Test Summary Number of cointegration relation in level series: EXP, FDI and GFC Trace test Maxeigenvalue test No intercept No trend Intercept No trend Intercept Linear Trend No intercept No trend Intercept No trend Intercept Linear Trend Bahrain 2 1 0 2 0 0 Kuwait 1 2 0 1 0 0 Oman 0 0 0 0 0 0 Qatar 0 0 1 0 0 1 KSA 1 2 2 1 2 2 UAE 0 0 1 1 1 1 55

Vol. 4, No. 11, pp. 4662. 5.3. The VAR and VECM Estimates The presence of cointegration amongst the variables rejects the noncausality amongst them. This means that at least one of the variables reacts to deviations from the longrun relationship. Consequently, an investigation as to whether the comovements amongst the variables correct for disequilibrium is needed. The Vector Error Correction Model (VECM) allows distinguishing between the shortrun and longrun forms of causality. Thus, a VEC model is implemented as follows (Granger, 1987; Johansen, 1988). Based on results of unit roots and cointegration, tables 2 and 3,VECM is an appropriate model for all countries except for Oman. ECM provides a useful and meaningful link between the longrun and the shortrun approach to economic modeling. The ECM equation should be negatively signed, indicating movement towards equilibrium; positive sign indicates movement away from equilibrium. The coefficient value should lie between 0 and 1; where 0 suggests no adjustment one time period later and 1 reflects full adjustment. The estimations of VECM for the five countries are shown in table (4). The table reflects that the errorcorrection carries a negative sign for the two models and one is insignificant for the five countries. The estimation explains about 76%, 78%, 85, 64% and 89% of the variation in economic growth for Bahrain Kuwait, Qatar, Saudi Arabia and United Arab of Emirates respectively. The negative sign of the coefficient of the error correction term states that the models are stable. The coefficients of (0.57), (0.96), (0.53), (0.493) and (0.0296) suggest that 0.57%, 0.96%, 0.53, 0.49 and 0.03 are movements back towards equilibrium following a shock to the five countries, in the next period (Table 4). The relatively high numeric of the Fstatistic indicates that there is a relatively strong feedback effect or the presence of Granger bidirectional causality between the variables (see figure 3). Stability of the VEC model is an important issue because if it is unstable, certain results, such as impulse response standard errors, will be invalid making the model results and conclusions suspect. To test the VEC models stability, the test of inverse roots of the AR characteristic polynomial is conducted (Abdelbaki, 2013). The results reported that inverse roots of the AR polynomial have roots with modulus less than one and lie inside the unit circle. This indicates that the estimated VEC models are stable (stationary). Table (4) VECM Estimation BAHRAIN Variable Coefficient Std. Error tstatistic dexp_gdp (1) 0.22 0.181 1.266 dfdi_gdp (1) 0.334 0.415 0.81 dgfc_gdp (1) 3.421 60.78 0.56 ECM1(1) 0.57 0.74 2.112 C 0.71 2.534 0.28 Rsquared 0.76 Adjusted Rsquared 0.61 Fstatistic 5.01 KUWAIT Variable Coefficient Std. Error tstatistic dexp_gdp (1) 0.531 0.684 0.78 dfdi_gdp (1) 6.21 33.4 1.80 dgfc_gdp (1) 4.84 412.057 0.1 56

Vol. 4, No. 11, pp. 4662. ECM1 (1) 0.96 2.35 1.262 C 4.76 8.305 0.573 Rsquared 0.87 Adjusted Rsquared 0.75 Fstatistic 6.955 QATAR Variable Coefficient Std. Error tstatistic dexp_gdp (1) 0.27 0.66 0.41 dfdi_gdp (1) 13.33 4.391 3.034 dgfc_gdp (1) 15.587 56.172 2.592 ECM1 (1) 0.53 0.553 4.58 C 1.21 3.42 0.354 Rsquared 0.852 Adjusted Rsquared 0.757 Fstatistic 8.95 KSA Variable Coefficient Std. Error tstatistic dexp_gdp (1) 0.686 0.811 0.845 dfdi_gdp (1) 0.136 0.218 0.626 dgfc_gdp (1) 9.245 436.78 0.21 ECM1 (1) 0.493 0.911 0.541 C 0.576 3.316 0.174 Rsquared 0.64 Adjusted Rsquared 0.41 Fstatistic 2.748 UAE Variable Coefficient Std. Error tstatistic dexp_gdp (1) 0.351 0.223 1.572 dfdi_gdp (1) 12.1 5.932 2.04 dgfc_gdp (1) 25.423 92.65 3.19 ECM1 (1) 0.0296 0.072 0.41 C 0.532 3.911 0.136 Rsquared 0.887 Adjusted Rsquared 0.815 Fstatistic 12.27 For Oman, table (5) reflects the results of VAR model. The table shows that tstatistic figures are insignificant for all variables and independent variables explain only 62% of all variables in growth rate. Table (5) VAR estimates for Oman Variable Coefficient Std. Error tstatistic oexp_gdp 0.74 1.20 1.62 ofdi_gdp 2.42 5.3 1.46 ogfc_gdp 6.39 16.97 1.78 C 72.68 71.07 1.02 Rsquared 0.62 Adjusted Rsquared 0.55 Fstatistic 3.58 57

Vol. 4, No. 11, pp. 4662. 5.4 Variance Decomposition To examine the relative strength of each variable in explaining the changes in the dependent variable, the variance decompositions (VD) is applied. Table (6) shows that in a 10year horizon, FDI accounts for 15.87%, 10.29%, 16.58%, 0.89% and 6.88% for countries Bahrain, Kuwait, Qatar, Saudi Arabia and United Arab Emirates respectively in shocks in their economic growth compared to the contributions of exports (9.55%, 1.33%, 6.85%, 10.38% and 20.9%) and gross capital formation (0.83%, 0.07%, 9.88%, 1.98 and 17.98%) for Bahrain, Kuwait, Qatar, Saudi Arabia and United Arab Emirates respectively. Table (6) VEC Variance Decomposition of Economic Growth for GCC Countries BAHRAIN Period S.E. DBGR DBEXP_GDP DBFDI_GDP DBGFC_GDP 1 11.87896 100.0000 0.000000 0.000000 0.000000 2 12.62072 96.05265 0.076694 3.832122 0.038532 3 13.59569 83.81968 1.413862 14.49060 0.275854 4 15.65389 85.07159 3.312318 10.95411 0.661983 5 16.09960 81.42910 6.579860 10.97586 1.015178 6 16.47190 77.81340 7.319481 13.88944 0.977684 7 17.30236 76.12720 8.912480 14.07415 0.886169 8 17.46925 75.15064 9.455057 14.51949 0.874817 9 17.77785 74.23096 9.320902 15.58642 0.861713 10 18.18508 73.74690 9.550390 15.87414 0.828579 KUWAIT Period S.E. DKGR DKEXP_GDP DKFDI_GDP DKGFC_GDP 1 33.60877 100.0000 0.000000 0.000000 0.000000 2 50.51670 99.44772 0.164612 0.374609 0.013059 3 51.34996 96.37130 0.302805 3.307032 0.018860 4 53.05011 94.17603 0.306037 5.495771 0.022166 5 54.35318 90.11536 0.337308 9.513229 0.034100 6 57.01598 89.87524 0.805704 9.279888 0.039165 7 57.34979 88.83549 0.875525 10.24165 0.047334 8 57.59294 88.60386 0.872571 10.47139 0.052172 9 57.74433 88.38704 0.886961 10.66459 0.061406 10 58.82923 88.32057 1.326192 10.28823 0.065008 QATAR Period S.E. DQGR DQEXP_GDP DQFDI_GDP DQGFC_GDP 1 16.41843 100.0000 0.000000 0.000000 0.000000 2 22.24372 74.64145 6.503233 8.037229 10.81809 3 23.11979 74.97196 6.022159 8.949439 10.05644 4 27.11999 72.98542 4.818608 14.87853 7.317432 5 28.48419 66.27586 5.967453 18.77216 8.984527 6 28.72871 66.52331 5.867850 18.64970 8.959138 7 29.27446 66.12225 6.213656 17.96245 9.701644 8 29.83359 65.98481 6.667981 17.65704 9.690174 9 30.32067 66.44145 6.743264 17.13511 9.680175 58

Vol. 4, No. 11, pp. 4662. 10 30.84066 66.68086 6.851896 16.58287 9.884383 SAUDI ARABIA Period S.E. DSGR DSEXP_GDP DSFDI_GDP DSGFC_GDP 1 15.84030 100.0000 0.000000 0.000000 0.000000 2 16.35781 93.82458 2.237076 1.069754 2.868590 3 17.56959 86.11961 9.810559 0.952985 3.116846 4 19.38137 88.36674 8.081288 0.990544 2.561428 5 20.07213 88.36300 7.975249 1.159782 2.501967 6 20.81204 86.49108 9.936615 1.099321 2.472988 7 21.63309 86.79414 9.861749 1.043879 2.300232 8 22.37292 86.83708 9.986577 1.000901 2.175445 9 23.09266 86.79666 10.19703 0.945115 2.061193 10 23.77933 86.74613 10.38282 0.893526 1.977528 UNITED ARAB EMIRATES Period S.E. DUGR DUEXP_GDP DUFDI_GDP DUGFC_GDP 1 19.02605 100.0000 0.000000 0.000000 0.000000 2 28.48411 49.74690 24.34693 9.171896 16.73427 3 32.29329 56.60536 22.34537 7.990764 13.05850 4 33.55352 53.96156 24.01095 7.735044 14.29245 5 35.42571 55.09308 21.55262 7.079908 16.27439 6 36.56228 53.40160 23.96216 7.343304 15.29294 7 37.74325 52.52541 23.21227 7.166861 17.09547 8 38.62601 53.12399 22.51898 7.112348 17.24468 9 39.58370 53.55007 21.93436 7.160216 17.35536 10 40.67728 54.25337 20.89561 6.875181 17.97584 5.5 Granger Causality Test The Granger causality test is implemented within an errorcorrection framework to examine the existence of a longterm relationship between each two variables. The Granger causality test is used to indicate the direction of the effect between each two variables. Figure (3) represents the causality results with lag 2. The results show that unidirectional causality exists from foreign direct investment to economic growth and to exports. The figure indicates also that unidirectional causality exists from economic growth to gross capital formation and from exports to gross capital formation for Bahrain. For Kuwait, the unidirectional causality has been found from exports to economic growth and from gross capital formation to economic growth. However the bidirectional causality has been found between exports and gross capital formation. Economic growth has effects as unidirectional causality on both exports and gross capital formation and as bidirectional causality on foreign direct investment in case of Qatar. In case of Saudi Arabia, the unidirectional causality exists from exports to economic growth, foreign direct investment and to gross capital formation. The bidirectional causality has been found between economic growth and gross capital formation and finally gross capital formation contributes as unidirectional causality to foreign direct investment. The unidirectional causality has been found from exports to economic growth and the bidirectional causality exists between economic growth and foreign direct investment for United Arab Emirates. Surprisingly, there is no any evidence for shortrun or longrun unidirectional or bidirectional causality relationship for Oman. 59

Vol. 4, No. 11, pp. 4662. Figure 3: Pairwise Granger Causality Tests Summary BAHRAIN KUWAIT GR GR ** * EXP FDI ** EXP *** * ** *** * GFC GFC QATAR KSA GR EXP GR ** ** *** * FDI GFC *** ** *** *** FDI GFC *** *** UAE GR EXP * ** ** EXP FDI Source: designed by the author based on the Granger Causality Test results. *, ** and *** denote significance at 10%, 5% and 1% respectively. 60

Vol. 4, No. 11, pp. 4662. 6. Concluding Remarks The main objectives of the paper are to study the determinants of economic growth in GCC countries and the relative effect of each of them in changes of economic growth. Cointegration test states that only Oman is in favor of zero cointegrating vector. Accordingly, VECM is an appropriate model for all countries except for Oman. The variance decomposition reflects that in a 10year horizon, FDI accounts for 15.87%, 10.29%, 16.58%, 0.89% and 6.88% for Bahrain, Kuwait, Qatar, Saudi Arabia and United Arab Emirates respectively in shocks in their economic growth compared to the contributions of exports (9.55%, 1.33%, 6.85%, 10.38% and 20.9%) and gross capital formation (0.83%, 0.07%, 9.88%, 1.98 and 17.98%) for Bahrain, Kuwait, Qatar, Saudi Arabia and United Arab Emirates respectively. The results also show that the main determinants of economic growth are foreign direct investment and gross capital formation for Bahrain. For Kuwait, Qatar and Saudi Arabia, exports and gross capital formation are the main determinants of economic growth. Exports and foreign direct investment are the main determinants of economic growth in United Arab Emirates. Finally, there is no any evidence for shortrun or longrun unidirectional or bidirectional causality relationship for Oman. References Abdelbaki, H., (2013). The impact of Arab spring on stock market performance. British Journal of Economics, Management and Trade (BJEMT), Vol. 3, No. 3, JulySep. Aljebrain, M. A. and Ibrahim, M. A. (2012). Determinants of demand for imports in GCC countries International Journal of Economics and Finance, Vol. 4. Alomer, I. (2013). Economic growth and savings in GCC: a cointegration and causal relationship analysis International Journal of Humanities and Social Sciences: Vol. 3, No. 9. Barro R. J. (1999). Determinants of economic growth: implications of the global evidence for Chile Cuadernos de Economia, No. 36, No. 107, April 1999. Barro R. J. (2003). Determinants of economic growth in a panel of countries Annals of Economics and Finance 4. Barro R. J. (1991). Economic growth in a cross section of countries Quarterly Journal of Economics, 106, 2. Blomstoerm, M., Lipsey, R., & Zejan, M. (1994). What explains the growth of developing countries? Convergence of Productivity. Oxford University Press, Oxford. Dritsakis N. (2004). Exports, investments and economic development of preaccession countries of European Union: an empirical investigation of Bulgaria and Romania. Applied Economics, Vol. 36(16). Pp 18311838. Granger, E. (1987). Cointegration and error correction regression: estimation and testing. Econometrica. 1987;55(2):25176. 61

Vol. 4, No. 11, pp. 4662. Gulf Organization for Industrial Consulting, GOIC, Doha, Qatar, Gulf Statistical Profile, different years. Johansen S, Juselius. (1990). Maximum likelihood estimation and inference on cointegration: with application to the demand for money. Oxford Bulletin of Economics and statistics. 52(2):169210. Johansen S. (1988). Statistical analysis of cointegrating vectors. Journal of Economic Dynamic and Control. 12:231254. Metwally M. M. (1987). Determinants of external surplus of the member states of the Gulf Cooperation Council Applied Economics, 19, 305316 Moudatsou, A. (2003). Foreign direct investment and economic growth in the European Union. Journal of Economic Integration. 18(4). 689707. Ndambiri et al. (2012). Determinants of economic growth in SubSaharan Africa: a panel data approach International Journal of Economics and Management Sciences. Vol. 2, No. 2. Stylianon, T. (2014). Dynamic relationship between growth, foreign direct investment and exports in the US: an approach with structure breaks. The IUP Journal of Applied Economics, Vol. XIII, No. 2. Terkova, Marua an Tashevska, B. (2011). Determinants of economic growth in SouthEast Europe Perspectives of Innovations, Economics & Business, Vol. 7, Issue 1. Thomas R. L. (1997). Modern econometrics: an introduction. Essex: AddisonWesley. Longman Limited. World Bank World Development Indicators. Database. Zhang, H. (1999). FDI and economic growth: evidence from ten East Asian Economies. Economia Internatinale. 7(4). 517535. 62