Current Research Journal of Economic Theory 2(1): 1-7, 2010
ISSN: 2042-485X
© M axwell Scientific Organization, 2009
Submitted Date: October 27, 2009
Accepted Date: November 21, 2009
Published Date: January 30, 2010
The Role of the Economic Size of a Country in Attracting Foreign Direct
Investments: A Panel Data Analysis for OECD Countries (1980-2006)
U. Do—an, A. Volkan and G. Burcu
University of Selcuk Iktisadi ve Idari Bilimler Fakulesi Kampus, Konya, Turkey
Abstract: The long-run relationship between foreign direct investment and economic growth was examined
for 21 OECD member countries using panel data estimations (models) based on annual data for the period
1980–2006. During the analysis process, which started with panel unit root tests, it was observed that the
variables of foreign direct investment and economic growth had un it roots. In the next stage, cointegration was
found between the variables. The existence of a significant relationship was determined between foreign direct
investment and economic grow th based on the estim ated regression m odel.
Key w ords: Foreign direct investment, gross domestic product, panel unit root and panel cointegration
INTRODUCTION
The correlation between foreign direct investment
inflows and economic growth in a country both
theore tically and empirically constitute the focal point of
econ omic discussions. The gene ral opinion reg arding this
issue is that foreign direct investment has a positive effect
on economic growth. This positive effect shows itself as
an increase in productivity in the host country via certain
variab les which are provided by foreign direct
investments, such as techn ological adap tation (Alfora
et al., 2001 ).
The possible effects of foreign d irect inve stments on
econ omic grow th are examined under two main headings.
The first of these evaluates foreign direct investments as
a mechanism which helps countries to eliminate the
disadvantage which arises from the insufficiency of
dom estic savings to finance economic expansion.
Secondly, the most imp ortant parameter that foreign
capital enterprises look for when en tering a country is
econ omic growth (Mencinger, 20 03). This approach is
essen tially related to the fact that foreign capital
enterprises that are looking for more profitable areas of
inves tment desire to enter a developing country in order
to get a bigger share of the profit cake which occurs as the
result of the rapid economic growth in that country. The
studies conducted on the relationship between foreign
direct investment and economic growth do not always
yield similar results. As it will be pointed out in the
literature review section, while the mentioned relationship
is found to be positive in some studies, it could be
negative in some others. Carkovic and Levin point out
that if there is not a positive relationsh ip between foreign
direct investment inflows and economic growth, the
policies implemented by developing countries in order to
attract foreign direct investment might not have a
significant effect (Carko vic and L evin, 2002 ).
In the present study, the significance of economic
grow th in attracting foreign direct investment, an issue
which has been discussed in the literature for a long time,
was tested for 21 OECD member countries. First, the data
obtained for 21 OECD member countries were converted
into pane l data. In the nex t phase of the study, the
existence of panel unit root in the mentioned variables
was investigated and the process was reported.
Afterwards, the cointegration between the two variables
was tested b y building a mod el wh ich included both
variables and as the result of the tests carried out, the
existence of cointegration between the variables was
determined and reported. The autocorrelation problem
was corrected within the final model which was obtained
by performing the related tests. It was reported that there
was not a problem resulting from the presence of
hetero scedasticity in the model. As a result, it was
concluded that the economic size of the country was
effective in attracting foreign direct investment inflows
for 21 OE CD coun tries.
Literature Review: The theoretical basis of the
correlation between foreign direct investment inflows and
econ omic grow th is founded on the endogenous and neoclassical growth models. In neoclassical models of
growth, it is emphasized that foreign direct investments
increase the volume of investment and its efficiency
resulting in long-term growth. From this point of view,
foreign direct inv estment inflows contribute to the
economic growth of the country.
The endogenous growth theories consider long run
grow th as a function of technological progress, and
provide a framework in w hich fo reign d irect inve stments
can permanently increase the rate of growth in the host
economy throug h technolog y transfer, diffusion, and
spillover effects (Nair and W einbold, 200 1).
Corresponding Author: U. Do— an, University of Selcuk Iktisadi ve Idari Bilimler Fakulesi Kampus, Konya, Turkey
1
Curr. Res. J. Econ. Theory, 2(1): 1-7, 2010
Blomström et al. (1992) tested how the share of
foreign direct investments within the gross dome stic
product affected the income and growth rates in the
following five-year period using the 5-year data from
developing and developed countries. In their study, they
concluded that the share of foreign direct investments
within the gross domestic product had a positive effect on
econ omic growth. At the same time, they emphasized that
foreign direct investment had a more remarkable effect on
economic grow th in high income countries.
Balasubramanayam et al. (1996) examined the role
of foreign direct investments in the growth process in a
study conducted on eighteen developing countries and
found out that foreign direct investment inflows had a
positive effect on economic growth.
In a study on the relationship between foreign direct
investment and economic growth in five developing
countries including Turkey, Ba shir (1999) found out that
foreign direct investment had a positive effect on
economic growth.
DeMello (1997) stated that foreign direct investment
inflows did not have a significant effect on growth, and as
the difference in the level of technological development
became larger, the relationship between the two variables
would be negative. In the study, it was reported that the
foreign trade policy and the political atmosphere of the
host country highly affected the attractiveness of that
country to foreign d irect inve stment.
DeMello (1999) conducted a study on OECD
member and non-mem ber countries comprising the period
from 1970 to 1999 and reported that foreign direct
investment and economic growth had a negative
relationship depending on the technological gap between
foreign and d ome stic investment.
In their study regarding the effects of foreign direct
investment in Venezuela for the period between 1979 and
1989, Aitken an d Harrison (1999) found out that
technological spillover effect would be prevalent in
developed countries, but there would not be such a
transfer in developing countries.
Barthelemy and Demurger (2000) conducted a study
on a sam ple of 24 Chinese provinces for the period from
1985 to 1996 and found out that foreign direct investment
inflow s had a positive effec t on economic growth.
Zhang (2001) studied A sian and Latin American
countries based on the data for the period from 1 960 to
1997 and fo und out that improved human capital
conditions, liberalization process and export-oriented
foreign direct inv estments would have a positive effect on
economic growth.
In the study conducted by Lensink and Morrissey
(2001) on developing countries for the period between
1975 and 1 998, while it was determined that foreign
direct investmen t had a po sitive effect on economic
growth, it was also reported that the volatility of foreign
direct investment inflows had a negative impact on
economic growth.
In the study conducted by Obwona (2001) on Uganda
for the period between 1981 and 1995, it was found that
there was a positive correlation between foreign direct
investment inflows and economic growth.
Sadik and Bolbol (2001) investigated six Arab
countries in their study. They found that technology
transfer through foreign direct investments did not have
any man ifest positive spillover on total factor productivity
over and above those of other types of capital formation.
In their study, Ram and Zhang (2002) could not find
a significant relationship between foreign direct
investment inflows and economic growth.
Kumar and Prad han (200 2) investigated 107
developing countries using panel data for the period from
1980 to 199 9 and found that economic growth was
effective in attracting foreign direct investment inflows.
In a study of Asian countries for the period from
1960 to 2000, Chowdhury and Mavrotas (2005) used the
Tod a-Yama moto Causality Test and suggested that
econ omic growth caused foreign direct investments in the
case of Chile and not vice versa, while for both M alaysia
and Thailand, there w as a strong ev idenc e of a bidirectional causality between the two variables.
Roy and Von den Berg (2006) used a simultaneousequation model (SEM) to study the relationship between
foreign direct investments and U.S. economic growth.
FDI was found to have a significant, positive, and
econ omically imp ortant im pact on U.S. grow th.
De—er and E msen (2006) studied the relation ship
between foreign direct investments and growth in the
countries of Ce ntral Ea st Euro pe an d Central W est Asia
for the period between 1990 and 2002 using panel data
regression method and found that while the relationship
was positive in the countries of C entral East Europe, there
was not a significant relation ship between the variables in
the countries of Ce ntral W est Asia.
In this study the long-run relationship between
foreign direct investment and economic growth was
examined for 21 OECD member countries (Australia,
Austria, Belgium, Canada, Finland, France, Germany,
Iceland, Ireland, Italy, Japon, Mexico, Netherland, New
Zeland, Norway, Portugal, Spain, Sweeden, Turkey, UK,
USA) using panel data estimations (models) based on
annual data for the period 1980-2006.
MATERIALS AND METHODS
Cross-sectional data, time series and panel data consisting
of time series or the com bination of cro ss-sectional data
and time series are u sed for analyzing the relation ship
be tw e e n e c onomic va ria bles statistically and
econometrically. The functional form of pan el data
econometrics is as follows;
(1)
2
Curr. Res. J. Econ. Theory, 2(1): 1-7, 2010
As it can be seen in the function al form of pan el data
econometrics, t shows the time and i shows the sections.
In this equation, an individual effect exists. This effect
cannot be observed by independent variables, does not
change depend ing on time, bu t includes characteristics
peculiar to sections (B altagi, 2005).
In panel data econometrics, the next step after
converting cross-sectional data a nd time series data to the
panel system is to determine if the cross-section and
period effects can be explained by the fixed effects model
or the random -effects model. The fixed effects model
creates a different fixed value for each cross-sectional
unit. In the fixed effects model, it is assumed that the
slope coefficients which are shown with “$” do not
change, but fixed coefficients show d ifferences among
only cross-sectional data or time data or among both types
of data. If the differentiation occurs only depending on
time, these types of models are named as one-way time
dependent fixed effect mode ls. If a differentiation occurs
in panel data depending on both time and section, these
mod els are named as two-way fixed effects m odel.
How ever, because the cross-sectional effe ct is gen erally
investigated more in panel data studies, panel data mod els
are generally considered as one-way models (Hsiao,
2002). One-way an d two-way fixed effects mod els can be
seen in the Eq. (2) and (3) giv en be low:
(5)
Here, Eq. 5 shows the one-way random-effects model and
Eq. (6) shows the two-way random-effects model. The
error terms in random effects have two components. The
first of these components is the : i value of the crosssection i = 1,2,…… .,N, which does not vary over time,
and the v it value which signifies the rest of the crosssection where the values are correlated over time. In this
mod el, the : i value, which signifies the cross-section
effect, and the v it value, which includes the remaining
error terms, are independent from each other. In addition,
these two compon ents of the error term are independent
from an observed value of each independent variable. For
this reason, the ordinary least squares estimators are
consistent and unbiased estimators o f the error term
com ponents (: i and v it ) shown in Eq. (4) and (5) which
explain the random effects model (Özer and Çiftçi, 2008).
RESULTS AND DISCUSSION
The data used in the study was taken from the official
OECD database on an annual basis for 21 OECD m ember
countries including Australia, Austria, Belgium, Canada,
Finland, France, Germany, Iceland, Ireland, Italy, Japan,
Mexico, Holland, New Zealand, Norway, Portugal, Spain,
Sweden, Turkey, England and the United States. In the
study, the significance of the econo mic size of a co untry
in attracting foreign direct investment inflows was taken
as a criterion to determine the relationship between
foreign direct investment and economic growth.
In the study, the simple interaction between foreign
direct investment and economic growth in 21 OECD
member countries for the years between 1980 and 2006
can be seen in the graph (Fig. 1)
According to Granger and Newbold (19 74), a
regression analysis between the variables does not provide
reliable results in case non-stationary data is used. For this
reason, stationarity should be checke d before performing
the regression analysis. The studies cond ucted by Levin
and Lin (1992 ; 1993), Breitung and M eyer (1994), Quah
(1994), Maddala and W u (1999), H adri (2000) an d Im
et al. (2003) suggest the use of unit root tests in p anel data
models. Recently, the most com mon ly used unit root tests
in the studies performing panel data unit root tests on a
sectoral basis are Levin-Lin and Im Pe saran Shin Tests.
Unit root tests of Levin, Lin and Chu (LLC), Breitung,
Im, Pesaran and Shin (IPS), Augmented Dickey-Fuller
(AD F), PP (Phillips Peron ) and Hadri were used in our
study. The panel data unit root test results for foreign
direct inv estment are given in Tab le 1.
According to the unit root test results regarding FD I,
LLC and B reitung unit root tests pointed out that the FDI
variab le had a unit root. How ever, the results of IPS,
(2)
(3)
In this equation, it is considered that the error terms are
distributed independently and identically in a manner that
their variances equal to zero. In the fixed effects model,
the fixed effects estimator allows the fixed constant to
differ across cross-section units by estimating different
constants for each cross-sec tion (Baltagi, 2005 ).
The changes that occur depending on cross-sections
or both cross-sections and time are observed when they
are integrated into the model as a component of the error
terms. The advantage of random effects model over the
fixed effects model is that, without loss of degree of
freedom, the random effects model allows the inclusion of
the effects that are out of the sample to the model. The
functional relation for the mentioned models can be
dem onstrated as fo llows:
(4)
3
Curr. Res. J. Econ. Theory, 2(1): 1-7, 2010
Fig. 1: The course of FDI-GNI in 21 OECD member countries Graph (1980- 2006)
ADF, PP and Hadri unit root tests showed that panel unit
root did not exist in the dataset.
As a next step, it is possible to see if the G NI variable
has a unit roo t with the help of the results shown in
Table 2. It can be shown in Table 2, the GNI variable had
a unit root according to the results of the LLC, Breitung,
IPS, ADF , PP and Hadri statistics.
The existence of a unit root in both series was
detected as the result of the finding s obtained fro m un it
root tests and it was concluded that the series was nonstationary. For this reason, Pedroni, Kao and Johansen
Fisher Cointegration Tests were used in the remainder of
the study. In the next step after detecting the existence of
a panel unit root in the series, the presence of
cointegration was investigated with the help of Table 3.
As it can be shown in Table 3, the null hypothesis was
rejected in all the tests in the regression equation formed
by the mentioned variables and the alternative hypothesis
was accepted, that is, the existence of cointegration was
confirmed.
After confirming the existence of cointegration
through Pedroni Test, it would be possible to test the
existence of cointegration also by performing the Kao
Test (Table 4).
The null hypothe sis was accepted based on the resu lts
of the Kao Cointegration Test. That is, the existence of
cointegration was rejected. In the same way, the Johansen
Fisher Panel Co integration Test, which is another
technique to check the existence of cointegration, can be
analy zed w ith the he lp of Table 5.
4
Curr. Res. J. Econ. Theory, 2(1): 1-7, 2010
T ab le 1: Un it ro o t t es t r es ul ts fo r F D I
Method
Test Statistics
Lev in, Lin and Ch u t*
3.93850
Breitung t-stat
5.26857
Im, Pesaran and Shin W-stat
-2.86909
AD F - Fish er Ch i-square
110.772
PP - Fisher Chi-square1
71.566
Hadri Z-stat
0.81534
Heteroscedastic Consistent Z-stat
4.22439
Pro bab ility
1.0000
1.0000
0.0021
0.0000
0.0000
0.2074
0.0000
T ab le 2: Un it ro o t t es t r es ul ts fo r G N I
Method
Test Statistics
Lev in, Lin and Ch u t*
7.29815
Breitung t-stat
7.95299
Im, Pesaran and Shin W-stat
9.73160
AD F - Fish er Ch i-square
21.3043
PP - F isher C hi-squa re
9.06980
Hadri Z-stat
14.8044
Heteroscedastic Consistent Z-stat
10.7447
Pro bab ility
1.0000
1.0000
1.0000
0.9967
1.0000
0.0000
0.0000
According to the results of the Johansen Fisher Panel
Cointegration Test, the null hypothesis that there is no
correlation between the two variables was rejected and the
alternative hypothesis supporting the existence of
cointegration was accepted.
Of the three tests which w ere conducted in order to
determine the existence of cointegration in the model,
Pedroni and Johansen Fisher Panel Co integration Tests
indicated the existence of cointegration. However, the
Kao Cointegration Test revealed that there was no
correlation between the two variables. Since the majority
of the tests we performed revealed the existence of
cointegration in the model, it was accepted that for the
OECD countries, there was a correlation between foreign
direct investment and economic growth in the long run.
In the next stage, panel data regression estimations
were performed. The Hausman Test w as use d in ord er to
determine whether the fixed effects model or the randomeffects model was valid for the 21 OECD m ember
countries. According to the Hausm an Test, the fixed
effects model yielded more effective results for all the
countries subject to the study. Based on this report, the
fixed effects method was used in panel data regression
estima tion.
According to Tab le 6 in the mod el wh ich was
developed based on the fixed effects model, the
approxim ate value of the Durbin Watson (0.63) statistics
was found to be below the (2) value which was accepted
as significant. Therefore, an AR root was added to the
model in order to render the mod el statistically more
significa nt, but a statistically significant result could not
be obtained. For this reason, we tried to obtain a more
significant model by including the lags of the dependent
variab le in the m odel can be shown in Tab le 7.
W hen we included the lags of the dependent variable
in the model, the results of the Durbin W atson statistics in
Table 7 were found to be more significant comp ared to
the previous model (1,858391 > 0.934268). We continued
to include the lags of the dependent variable in the model
in orde r to increase the significance level.
It can also be seen in Table 8 that the inclusion of the
second lag of the dependent variable in the model yielded
statistically more significant results when compared to the
previous situation.
However, according to the assumption of a constant
variance, which is one of the significant assumptions of
the linear regression model, the error term variance stays
constant and does not vary depending on the changes that
occur in the dependent variable. However, the assumption
of a constant variance cannot always be maintained
especially in the studies based on cross-sectional and
panel data and heteroscedasticity might occur in some
cases (Ertek, 1996). The regression estimations carried
out in the present study were checke d for
heteroscedasticity using White Test and the problem of
hetero scedasticity was eliminated. The new model which
Table 3: Pedroni cointegration test results for FDI and GN I series
Method
Test Statistics
Pro bab ility
Panel v-¤statisti—i
3.794102
0.0001
Pane l rho- ¤statisti—i
-6.427852
0.0000
Pane l PP- ¤statisti—i
-6.504310
0.0000
Pane l AD F- ¤statisti—i
-10.54857
0.0000
Gro up rh o- ¤statisti—i
-7.468717
0.0000
Gro up P P- ¤statisti—i
-9.294297
0.0000
Gro up A DF - ¤statisti—i
-8.329224
0.0000
Table 4: Kao cointegration test results for LEC and LG DP series
Method
Test Statistics
Pro bab ility
ADF
0.953893
0.1701
Residual Variance
4.27E+20
HA C Variance
2.57E+20
Tab le 5: Johansen Fisher cointegration test results for LEC and L G D P
series
Hypothesized Fisher
Pro bab ility Fisher Statistics Pro bab ility
No. of C E(s) Statistics
(ma x-eig en te st)
(trace te st)
None
19 0.0
0.0000
14 2.2
0.0000
At mo st
11 33 .9
0.0000
13 3.9
0.0000
Tab le 6: F ixed effec ts pan el da ta reg ressio n es timatio n res ults
Coefficient
Standart Error
T-Statistics
Pro bab ility
C
-6.32E+08
1.49E+09
-0.423825
0.6719
G N I 0.016198
0.001367
11.84820
0.0000
R2: 0.628575
D-W Stat.: 0.934268
Tab le 7: Fixed effects panel data regression estimation results (Fdi (-1)
added)
Coefficient Standart Error T-Statistics
Pro bab ility
C
68807831
135E+09
0.050866
0.9595
GNI
0.007823
0.001374
5.693264
0.0000
FDI (-1) 0.544369
0.038901
13.99357
0.0000
R 2 : 0.734049
D.W. Stat.: 1.858391
Tab le 8: Fixed effec ts pan el da ta reg ressio n es timatio n res ults (fdi (-2)
adde d))
Coefficient Standart Error T-Statistics
Pro bab ility
C
-4.49E+08
1.43E+09
-0.313224
0.7542
GNI
0.009729
0.001497
6.498698
0.0000
FDI (-1) 0.631752
0.046040
13.72191
0.0000
FDI (-2) -0.189168
0.049750
-3.802394
0.0002
R 2 : 0.741374
D.W . ¤st.: 2.060753
5
Curr. Res. J. Econ. Theory, 2(1): 1-7, 2010
Tab le 9: T he n ew mo del fre e from the p rob lem o f hete rosc eda sticity
Coefficient
Standart Error T-Statistics
Pro bab ility
C
-4.49E+08
2.40E+09
-0.187148
0.8516
GNI
0.009729
0.003230
3.012279
0.0027
FDI (-1) 0.631752
0.239385
2.639067
0.0086
R 2 : 0.741374
D.W . ¤st.: 2.060753
REFERENCES
Alfaro, L., C. A reend am, Ô. Kalemli and S. Sayek, 2001.
FDI and Economic Growth: The Role of Financial
Markets. Harvard B usiness Schoo l, Working Paper,
pp: 01-083.
Aitken, B.J. and A. Harrison, 1999. Do domestic firms
benefit from direct foreign investment? Evidenve
From Venezuela. Am. Econ. Rev., 89: 605-618.
Baltagi, B.H ., 2005 . Econom etric Analysis of Panel Data.
3rd Edn., John W iley and Sons Gmbh, W est Sussex,
Eng land.
Balasubramanayam, V.N., A. Salisu Mohamm ed and
V.D. Sapsford, 1996. Foreign direct investment and
grow th in EP and IS cou ntries. Econ. J., 106(1):
92-105.
Berthélemy, J.C. and D. Sylvie, 2000. Foreign direct
investment and economic growth: theory and
application to China. Rev. Develop. Econ., 4(2):
140-155.
Bash ir, A., 1999. Foreign direct inv estment and eco nom ic
grow th in some MENA countries: Theory and
Evidence. From: http://www.luc.edu/orgs/meea/
volum e1/ba shir.pdf. (Acc essed : 05.05 .2009).
Blomström, M., R.E. Lipsey and V.Z. Mario 1992. What
Explains Developing Countries Growth?. National
Bureau of Economic Research (NBER), pp: 4132.
Breitung, J. and W . Meyer, 1994. Testing for un it roots in
panel data: Are wages on different barg aining levels
cointegrated?. Appl. Econ., 26: 353-361.
Carkovic, M. and L. Ross, 2002. Does foreign direct
investment accelerate economic growth? Financial
Globalization: A Blessing or a Curse, W orld Bank
Conference.
Chowdhury, A. and G. ve Mavrotas, 2005. FDI and
growth: A causal relationship. United Nation
Unive rsity, W IDE R (W orld Ins titute for
Development Econo mics R esearch), Research Paper
No. 2005/25.
DeM ello, Jr. and R. Luiz, 1997. Foreign direct investment
in developing countries and growth: A selective
surve y. J. Develo p. Stud ., 34(1): 1-34.
DeM ello, Jr. and R. Luiz, 1999. Foreign direct
investment-led growth: evidence from time series and
panel data. Oxford Economic Papers, 51(1): 133-151.
De—er, M.K. and E. Emsen, 2006. GeçiÕ Ekonomilerinde
D o—rudan Yabanc2 Sermaye Yat2r2mlar2 ve
Eko nom ik Büyüme ¤liÕkileri: Panel Veri A nalizleri
(1990-20 02). Çukurova Ü niversitesi ¤.¤.B.F. Dergisi,
7(2): 121-137.
Ertek, T., 1996. Ekonometriye GiriÕ, GeniÕletilmiÕ 2.
Bask2, Beta Yay2nlar2, ¤stanbul.
Grang er, C.W.J. and P. Newbold, 1974. Spurious
regressions in econometrics. J. Econ., 2: 111-120.
Hadri, K., 2000. Testing for stationarity in heterogeneous
panel data. Econ. J., 3: 148-161.
Hsiao, C., 2002. Analysis of Panel Data, Cambridge
University P ress, 2nd Edn., Ne w Y ork.
was rendered free from the problem of heteroscedasticity
is presented in Table 9.
As it can be show n in Table 9, based on the resu lts
obtained in the present study conducted for 21 OECD
member countries, we conclude that the economic size of
a country is effective in attracting foreign direct
investment inflows. Since the probability value is smaller
than the table value (0.0027 < 0.005).
CONCLUSION
The term ‘foreign direct investment’ has long been
the subject of various discussions particularly in the
context of econom ic grow th. W hile one group of
econ omists express their opinions regarding what effect
foreign direct investment will have on economic growth,
the other group has reached a consensus that the most
important economic factor that international capital looks
for when entering a country is economic growth.
In the present study, the second suggestion was
studied by investigating the economies of 21 OECD
member countries. Accord ingly, first, the pane l data
system consisting of time series and cross-sectional series
was used for analyzing the relationship between the
variables. In the next step, using panel data, the existence
of panel unit root in the mentioned variables was
investigated. The test results indicated the presence of a
unit root in the variables and the finding was reported.
After this step, a panel cointegration test was conducted
and it was reported that a pa nel cointegration relationship
existed among the variables consisting of the data
obtained for 21 OECD member countries. Afterw ards, it
was investigated whether the fixed effects model or the
random-effects model would be used in the model by
using the Hausman Test statistics. Based on the results of
the test statistics, it was reported that the fixed effects
model yielded more effective results for all the countries
subject to the study.
In the last step of the study, the Durbin Watson
coefficient was calculated and the lags of the dependent
variab le were included in the model in order to render the
model statistically more significant. In this way , a more
significant mod el wa s obtained. A fter improving the
Durbin Watson Test statistics, the problem of
hetero scedasticity was checked using the White Test and
the problem of heteroscedasticity was eliminated. Based
on the final model on w hich structural and diagno stic tests
were performed, it was concluded in this study of 21
OECD member countries that the economic size of a
country was an effective parameter in attracting foreign
direct inv estment.
6
Curr. Res. J. Econ. Theory, 2(1): 1-7, 2010
Im, K.S., M.H. Paseran and Y. Shin , 2003. Testing for
unit roots in heterogeneous panels. J. Econ., 115:
53-74.
Kum ar, N. and ve P. Pradhan, 2002. Foreign direct
investment externalities and economic growth in
developing countries. RIS Discussion Papers, 1: 0137.
Lensink, R. and M. Ve Oliver, 2001. Foreign direct
investment: flows, volatility and growth in
developing countrie. From: http://www.ub.rug.
nl/e ldoc/o m/e / 1 E 1 6 / 0 1 E 1 6 . p d f ( A c c e s s e d:
10.01.20 09).
Levin, A. and C. Lin, 1992. Unit Root Tests in panel data:
Asymp totic and finite sam ple properties. University
of California, San Diego Working Paper, pp: 23- 92.
Levin, A. an d C. Lin, 19 93. U nit Root tests in panel data:
New results. University of California, San Diego
W orking Paper, pp: 56- 93.
Maddala, G.S. and S. Wu, 1999. A comparative study of
unit root tests with panel data and a new simple test.
Oxford Bull. Econ. Stat., Special Issue, pp: 631- 652.
Mencinger, J., 2003. Does foreign direct investment
always enhance economic growth?. Kyklos, 56(4):
491-508.
Nair, R. and D. W einhold, 2001. Causality tests for
cross- country panels: A new look at FDI and
economic growth in developing coun tries. Oxford
Bull. Econ. Stat., 63(2): 153-171.
Obwona, M.B., 2001. D eterminants of FD I and their
impact on economic growth in Uganda. African
Development Bank Eco nomic Policy Research
Centre.
Özer, M. and N . Çiftçi, 2008. AR-GE Tabanl2 ¤çsel
Büyüme Modelleri ve AR-GE H arcamalar2n2n
Eko nom ik Büyüme Ü zerine Etkisi: OE CD Ülkeleri
Panel Veri Analizi. Selçuk Ü niversitesi ¤ktisadi ve
¤dari Bilimler F akültesi Sosyal ve Ekonom ik
AraÕt2rmalar Dergisi, Kony a. 9(16).
Quah, D., 1994. Exploting cross-section variations for
unit root inference in dynamic data. Econ. Lett., 44:
9-19.
Ram, R. and K.H. Zhang, 2002. Foreign direct investment
and economic growth: Evidence from cross-coun try
data for the 1990s. Econ. Develop. Cult. Change,
51(1): 205-215.
Roy, A.G. and H.F. Von den Berg, 2006. Foreign direct
investment and economic growth: A time series
approach . Glob al Eco n. J., 6: 1-19.
Sadik, A. an d V. ve Bolbol, 2001. Capital flows FDI and
technology spillow ers: Ev idence From Arab
Countries. World Develop., 29(12): 2111-2125.
Zhang, K. H ., 2001. Does foreign inv estment pro mote
econ omic growth? Evidence from East Asia and
Latin Am erica. C ontem. Econ. Policy, 19(2):
175-185.
7