The Casual Relationship Between Information and Communication Technology (ICT) and Foreign Direct Investment (FDI)

Transcription

1 Association for Inforation Systes AIS Electronic Library (AISeL) ECIS 2003 Proceedings European Conference on Inforation Systes (ECIS) 2003 The Casual Relationship Between Inforation and Counication Technology (ICT) and Foreign Direct Investent (FDI) Roghieh Gholai National University of Singapore, Sang-Yong To Lee National University of Singapore, Alas Heshati The Unitied Nations University, Follow this and additional works at: Recoended Citation Gholai, Roghieh; Lee, Sang-Yong To; and Heshati, Alas, "The Casual Relationship Between Inforation and Counication Technology (ICT) and Foreign Direct Investent (FDI)" (2003). ECIS 2003 Proceedings This aterial is brought to you by the European Conference on Inforation Systes (ECIS) at AIS Electronic Library (AISeL). It has been accepted for inclusion in ECIS 2003 Proceedings by an authorized adinistrator of AIS Electronic Library (AISeL). For ore inforation, please contact

2 Gholai, R. To Lee,S. Heshati A. The Causal Relationship Between Inforation and Counication Technology (ICT) and Foreign Direct Investent (FDI) Roghieh Gholai National University of Singapore Departent of Inforation Systes Sang-Yong To Lee National University of Singapore Departent of Inforation Systes Alas Heshati The United Nations University World Institute for Developent Econoics Research (WIDER) Abstract This paper investigates the siultaneous causal relationship between investents in inforation and counication technology and foreign direct investent, with reference to its iplications on econoic growth. For the epirical analysis we use data fro 23 ajor countries with heterogeneous econoics developent for the period The results of unit roots and Johansen co-integration tests indicate variations in degrees of integration aong the saple countries. Our causality test results suggest that there is a causal relationship fro ICT to FDI interpreted as the higher level of ICT investent leads to increased inflow of FDI. ICT contributes to econoic growth indirectly by attracting ore foreign direct investent. In developed countries there already exist a build up ICT capacity which causes inflow of FDI, while in developing countries ICT capacity ust be build up to attract FDI. The inflow of FDI causes further increases in ICT investent and capacity. Keywords Foreign direct investent (FDI), inforation and counication technology (ICT), stationarity, co-integration, causality, LSDV, and 2SLS

3 . Introduction The growth of world foreign direct investent (FDI) in recent years has been exceptional. The US dollar value of world FDI inflows reached a record US$.3 trillion in 2000 fro just over US$200 billion in 993. In 980 the FDI stock represented the equivalent of only five percent of world GDP; while this percentage had alost tripled to 4 percent by the end of 990s, (UNCTAD 2000). The share of developing countries in FDI inflows has been raised fro 7. percent in to 2.4 percent in Since the 980s, attracting FDI has been one of the ost iportant policy goals of both developing and developed countries. To achieve the objective, a nuber of countries have not only liberalized restrictions on FDI but also provided incentives to attract FDI. The contribution of FDI to doestic productivity has been studied by previous research and there is a general agreeent about the positive ipacts of foreign direct investent on econoic developent (Baldwin et al 999; Eaton and Kortu 997; Haddad and Harrison 993; Aitken and Harrison 999). Though soe found negative results (Levine et al 2000) but ost epirical studies found a positive relation between FDI, productivity and growth (Markusen and Venables 999; Xu 2000; Borensztein et al. 998; OECD 998; Blostro et al 994; Soto 2000). A variety of factors are cited in the literature including infrastructure, huan capital, political instability as deterinants of FDI, however we ust also take account of deeper and broader changes in the global econoy, especially the spread of the New Econoy, and the new inforation and counication technology (ICT). This factor is reshaping the global syste. There is a large literature on FDI, soe of it dating to 40 years or ore. But the global econoy has undergone assive change over the last 20 years, and what was relevant to attracting FDI in the 970s ay no longer be the case today (Addison, and Heshati, 2002). Countries that successfully adopt ICT ay be able to overcoe barriers that have long held the back, particularly the constraint of a reote geography and an unfavorable cliate that ay otherwise adversely affect their ability to participate in global trade (Addison and Rahan 2002). In fact, the ajor global shift of the last twenty years is technological. The rapid spread of internet and use of the World Wide Web has opened up the possibility of accessing coercial and political inforation that was previously unavailable or severely restricted. In particular ICT has reduced any of the transactions costs of participating sub-contracting (through B2B interaction), and it is facilitating the operations of low-cost suppliers of IT services based in developing countries (Matabalya and Wolf, 200). Therefore ICT need to be considered into any explanation of FDI flows. In a recent study Addison and Heshati (2002) using a large saple of countries, explored the deterinants of FDI. Their findings suggest that ICT increases inflows of FDI to developing countries. ICT lowers the transaction and production costs of foreign investors, as well as their inforation about investent opportunities in poorer econoies. Findings also suggest that there is weak cause effect fro ICT on FDI in developing countries. Motivation of this study is to exaine the existence and nature of any causal relationship between ICT and FDI inflows. This issue can be analyzed using tie series and panel data

4 analysis tools. If non-stationary tie series variables are not co-integrated, then high degree of correlation between two variables does not ean a causal relationship between the variables. Tie series ethodology epowers us to recognize and avoid spurious results, which ight happen using siple OLS ethod. To our knowledge, no attepts have so far been ade to investigate the causal relationship between FDI and various deterinants of FDI inflows based on relatively large saple of countries, long tie series and various causality analysis including Tie Series Granger Causality, Least Squares Duy Variable (LSDV), and Instruental Variables Method for panel causality analysis. Here our focus is particularly on the relationship between ICT and FDI and its iplications on econoic growth. The key feature of this paper is its contribution to analysis of causality aong the key variables of interest in a siultaneous fraework. The hypothesis to be tested is whether the rich Inforation and counication technology infrastructure of the host country attracts ore FDI. ICT variable is fro ITU s, World Telecounication Indicators Database (2002) and other variables are fro the World Developent Indicators (World Bank 2002). We have chosen 23 developed and developing countries observed for the period based on data availability. The organization of the paper is as follows. Following the introduction and a brief review of the literature, we explain the data and ethodology for tie series analysis in Section 3 and 4. In Section 5 various approaches to Panel Causality Test are outlined and discussed. In Section 6 the estiation results fro the Granger-Causality tests is presented. Having done that, conclusion appears in the next section. 2. A Brief Review of the Literature 2.. FDI and Econoic Growth The recent trend of FDI has posed opportunities and challenges for developent and econoic growth, especially for developing countries. International investent can ipact econoic growth in any ways, but it is possibly for increasing productivity by iproveent in technology, including anagerial knowledge and skills, is perhaps the ost iportant one (Baldwin et al. 999; Saggi, 2000; UNCTAD, 999). The econoic rationale for offering special incentives to attract FDI frequently derives fro the belief that foreign investent produces externalities in the for of technology transfers and spillovers. Roer (993), for exaple, argues that there are iportant idea gaps between rich and poor countries. He notes that foreign investent can ease the transfer of technological and business know-how to poorer countries. This transfer ay have substantial spillover effects for the entire econoy. Thus, foreign investent ay boost the productivity of all firs - not just those receiving foreign capital (Rappaport, 2000). The growth theories have identified the factors that play role in prooting econoic growth as follows: Savings and investent (classical odels), technical progress (neoclassical odels), R&D, huan capital, accuulation and externalities (new growth theory). FDI The countries studied include Austria, Australia, Brazil, Canada, Colobia, Denark, Finland, France, Iceland, India, Indonesia, Ireland, Italy, Japan, Korea, Malaysia, Mexico, Norway, Singapore, Sweden, Turkey, United Kingdo, and United States.

5 has been integrated into theories of econoic growth and there is a gains-fro-fdi approach. To the extent that FDI adds to the existing capital stock, it ay have growth effects siilar to that of doestic investent. FDI ay iprove exports and help the access of doestic enterprises to international arkets. Moreover there will be spillover effect through the diffusion of the transferred technology. The entry of a ultinational corporation (MNC) represents soething ore than a siple iport of capital into a host country, which is generally how the atter is treated in odels rooted in traditional trade theory. 2.2 What Does FDI Offer to a Host Country? There is a general agreeent about the positive ipacts of foreign direct investent on the welfare of countries receiving FDI. The benefits of FDI concerning the capital arket, technology transfer, arket access, investent opportunities and export prootion are aong the factors attracting FDI inflows fro a host country perspective. Capital: Multinational Enterprises (MNEs) invest in long-ter projects, taking risks and repatriating profits only when the projects yield returns. Technology: Evidence provided by the vast ajority of econoic studies dealing with the relationship between FDI on the one hand and productivity and econoic growth on the other hand, as found that technology transfer via FDI has contributed positively to productivity and econoic growth in host countries (OECD, 99). Market access: MNEs can provide access to export arkets. The growth of exports itself offers benefits in ters of technological learning, copetitive stiulus etc. Increased doestic investent: Agrawal (2000) exained the data on five south Asian countries and found out that the increase in FDI inflows were associated with a any-fold increase in the investent by national investors, (Borensztein et al. 998; Agosin and Mayer 2000; McMillan 999; Alfaro et al. 200). Export prooting: It sees that FDI could be associated with export trade in goods, and the hosting country can benefit fro an FDI-led export growth (Goldberg and Klein 999, Urata et al 998, OECD, 998). A tie series study on China indicates a two-way Granger causality running between output growth and FDI inflows (Shan J. et al. 997). Blostro, Lipsey, and Zejan (994), Borensztein, De Gregorio, and Lee (998) show that there are any econoetric specifications in which FDI is positively linked with long-run growth.

6 2.3 Deterinants of Foreign Direct Investent Traditional Factors: Many factors have been considered in the literature as deterinants of FDI. However, the selection of deterinants is often done ad hoc. The selection process is deterined by the availability of data and the nature of the relations studied. The key deterinants frequently appearing in the literature and their expected ipacts, including natural resources, arket size, socio-political instability, business operating conditions, wage costs; exchange rate, trade barriers, export orientation, openness of developing host countries, deocratization and risk, in addition one ay control for several other observable and unobservable tie-specific and country-specific effects (see Dunning 980; Lunn 980; Root and Ahed 979; Chakrabarti 200; Dollar 992). A coprehensive study of deterinants of FDI is beyond the scope of this paper. Here, the focus is on the causal relationship between FDI, ICT and econoics growth. We are currently working those issues. New Factor: Inforation and Counication Technology is considered as the ain new deterinant of FDI (Addison and Heshati 2002). The world is rapidly oving toward an econoic syste based on the continuous and pervasive availability of inforation. Recent advances in inforation and telecounication technology has been an iportant vehicle in peritting inforation exchange to develop as a valuable coodity. Countries and sectors equipped with the requisite telecounications systes have been rapidly oving into postindustrial, inforation-based econoic growth. ICT offers a unique opportunity for countries to free theselves fro the doination of geography. Siilarly, goods and services fro such countries can be offered on the global arket as efficiently as those fro any other country through the use of ICT. The ever-evolving and increasingly powerful ICT had fundaentally changed the nature of global relationships, sources of copetitive advantage and opportunities for econoic and social developent. Technologies such as the Internet, personal coputers and wireless telephony had turned the globe into an increasingly interconnected network of individuals, firs, and governents counicating and interacting with each other through a variety of channels 2. For the developing world, a odern telecounications infrastructure is not only essential for doestic econoic growth, but also a prerequisite for participation in increasingly copetitive world arkets and for attracting new investents. In the advanced industrial countries of Europe and North Aerica, universal telecounications services have penetrated every sector of society. In any developing countries the liited availability of service is constraining econoic growth. Econoic developent policies in the industrial countries increasingly include telecounications as an essential coponent of the econoic infrastructure. This realization has been initiated by industry's deand for advanced telecounications equipent for copetitive reasons. The lesser-developed countries have begun to recognize that inadequate telecounications services will be a disincentive to new investent and place existing industry at a copetitive disadvantage. 2 Report of the Regional Round Table on Inforation Technology and Developent, New Delhi, 2-22 June (2000).

7 Few doestic businesses and no international activities could operate copetitively without odern telecounications. The priary benefits include reduced transport costs, reduced transaction costs, iproved arketing inforation and increased efficiency of industrial production. A wide range of studies indicates that expanded telecounications investent is essential, not only for growth, but also to reain copetitive within the increasingly inforation-oriented global econoy. 3. The Data The data used in this study consists of a saple of 23 countries observed for the period 976 to 999. ICT variable is fro International Telecounication Union s; World Telecounication Indicators Database Annual investent in Telecounications is a proxy for ICT. Following the tradition in the literature, we define FDI as net inflows of foreign direct investent expressed as a percentage of GDP (World Developent Indicators 2002, World Bank). Data is a balanced panel and is chosen based on availability for ICT variable. 4. Tie Series Granger Causality Analysis Most of the econoic variables are not stationary at their levels, therefore to study the long run and short-run relationship we need to know each series degree of integration. Therefore, first we have perfored the Augented Dickey-Fuller (ADF) test to establish the order of integration of the variables (Dickey and Fuller, 979; 98). Now, for the co-integration tests, we use Johansen (988) odel, which was extended by Johansen and Juselius (990) and Johansen (99). This ethod applies the axiu likelihood procedure that is appropriate in a ultivariate fraework analysis. Table presents the results of the unit root tests. For each of the series exained, the test statistics suggest that the levels of the series are not stationary. They are integrated of order, or 2, which, eans the first or second differences of the series are stationary. Now that we know the level of integration of the series we can proceed to test of co-integration. In testing the long-run relationship between ICT and FDI, the null hypothesis states there is no co-integration relation. If the null hypothesis cannot be accepted, we will test the hypothesis that there is at ost one cointegration vector. The results of Johansen trace and axial eigenvalue tests are provided in Table 2. Results suggest that there is not enough evidence of long run relationship between FDI and ICT in ost of the countries in our saple (8 out of 23 countries). Even for few countries like Denark, Japan, Malaysia, Singapore and Norway the significance level is weak. ICT and FDI don t see to have a long-run relationship. In the absence of the long-run relationship aong econoic variables, it still reains of interest to exaine the short-run linkages aong the (Manning and Adriacanos, 993). That is, even though a long-run relationship between the two variables cannot be established for this tie period, it ay still be possible that the variables are causally related in the short-run. Systeatic testing and deterination of causal directions becae possible, after Granger (969) and Sis (972) developed the operational fraework. In econoetrics, the ost widely used

8 operational definition of causality is the Granger definition of causality, which is defined as follows: x is a Granger cause of y (denoted as x y), if present y can be predicted with better accuracy by using past values of x rather than by not doing so. (Chareza and Deadan, 992:90). After getting the stationary series (stationary series obtained fro differencing), we use following vector autoregression (VAR) odels estiated for each country separately. Here we have a nuber of key deterinants of FDI, such as: ICT investent, openness, GDP growth. Openness of a country is the trade share of GDP (iports plus exports); there is a positive association between openness and FDI. GDP growth also has a positive ipact on FDI. We have chosen theses key variables, which are the ost coon variables considered in previous studies. n () FDI t = ai FDI t i + bj ICTt j + ckgdpt k + d lopent l + ut i= n j= (2) ICT t = ei ICTt i + f j FDI t j + g kgdpt k + hlopent l + vt i= j= k = k= where t indicate tie period. We selected the lag structure of the odel based on, Akaike Inforation Criteria (AIC), at 5% level reported by E-views Panel Causality Analysis A. Least Squares Duy Variable (LSDV) Approach The introduction of a panel data diension allows using both cross-sectional and tie series inforation to test the causality relationships between y and x. In particular, it leads to give the researcher a large nuber of observations, increasing the degree of freedo and reducing the collinearity aong explanatory variables. So, it noticeably iproves the efficiency of Granger causality tests. Pooling cross-sectional units does have certain advantages. Like, the assuption of tie stationarity can be relaxed. We consider the following VAR odel: M N (3) FDI = it α + afdi i + b t nicti + clgdpi +, d t l kopeni + u, t n,, t k it = M N n= (4) ICT = it α + e ICTi + f nfdi t i i t j + g lgdpi t l + 2, hk OPENi t l + ω,,, it = n= L l= L l= Where FDI is the FDI share of GDP of country i (i=,, N) in period t (t=,, T), u it is the error ter. The error ter follows a two-way error coponent structure (Baltagi 200) and can be broken down into an unobservable country specific (µ i ), a tie specific (λ t ), and a rando error ter (v it ) coponents as: (5) u it = µ i + λt + vit The error ter v it represents easureent errors in the dependent variable and oitted explanatory variables. The error ter is assued to be independently and identically distributed l= l= K k= K k =

9 with zero ean and constant variance, σ 2. The country and tie specific effects µ i (country duies) and λ t, are factors representing country heterogeneity and exogenous technological change respectively and assued to be independent of each other and regressors. In the literature, the tie effects λ t are often replaced with a tie trend reducing the two-way error coponent odel to one-way error coponent odel. In the panel literature the estiation of the odel (3) has been developed in 2 directions, the fixed effect (FE) odel where µ it is assued to be fixed and correlated with explanatory variables and rando effects (RE) odel which µ it is assued to be rando and not correlated with the explanatory variables. In this study we use fixed effects odel since we have a relatively sall saple of countries not chosen randoly. Furtherore, the country heterogeneity effects are iportant concerning the flow of FDI. B. A ethod of Instruental Variables (2SLS) To date, ost causality tests have used tie-series data. However, it is difficult to control for easureent errors and oitted variable probles. To overcoe these probles, we apply an instruental variable 2SLS technique to conduct the causality test. The idea is to account for the endogeneity of regressors using instruental variable ethods. This ethod can be used when standard regression estiates of the relation of interest are biased because of reverse causality, selection bias, easureent error, or the presence of uneasured confounding effects. The central idea is to use a third, instruental variable to extract variation in the variable of interest that is unrelated to these probles, and to use this variation to estiate its causal effect on an outcoe easure. 2SLS estiator increases coputational efficiency without significantly detracting fro its effectiveness. A typical exaple of traditional panel data causality testing is Holtz-Eakin et.al (988). The Holtz-Eakin et al odel is: (6) y it = α + αj yi + t j δj xit j + fi + u it j= j= where i = N. In order to eliinate the fixed effects, to the odel: f ithe authors difference the data leading it j ( it j it j j it j it j it it j= j= (7) y y = α y y ) + δ ( x x ) + ( u u ) it This specification introduces a proble of siultaneity because the error ter is correlated with the regressor y it j yit j. Therefore, a 2SLS instruental variables procedure with a tievarying set of instruents is used to estiate the odel.

10 Anderson-Hsiao (982), suggest IV on the differenced odel using y lagged twice, y (-2) and differenced x s as instruents (dx). The authors then equate the question of whether or not x causes y with a test of the joint hypothesis: (8) δ δ =... = δ 0 = 2 = In the estiation, soe attention is paid to the validity of the instruents. Here are two probles with instruental variables ethods: first, the instruents should be uncorrelated with the error ter, or the orthogonality conditions should be satisfied by the data (exogeneity requireent); second, the instruents should have a strong correlation with the regressors of the odel (relevance requireent). For estiation purposes, we have used the 2SLS estiation procedure available in EVIEWS. 6. Results of Causality Test Table and 3 present the results of Granger causality test using tie series data, LSDV and 2SLS ethods using panel data. H denotes the alternative hypothesis that ICT does not Granger Cause FDI, and H 2 denotes the alternative hypothesis that FDI does not Granger Cause ICT. Results significantly suggest that there is a causal relationship between ICT and FDI in the saple countries. Results of causality test fro the Least Squares Duy Variable approach (LSDV) based on pooled data suggest that in developed countries ICT causes FDI, which eans in developed countries there is a basic ICT infrastructure which the host country can invest in and attracts foreign investors to coe and invest, whereas in developing countries FDI causes ICT eans inflows of FDI generate new ICT investent to facilitate production potential. Developing country is poor of ICT infrastructure and has no possibilities to internally finance such by theselves. FDI could cause ICT, given that the flow FDI is to branches producing product requiring advanced (ICT) technologies. Causality results fro 2SLS approach only suggest that in developed countries ICT causes FDI. Increases in inforation and knowledge, result in ore efficient cooperation and coordination. Coerce is essentially an inforation processing activity. Effective buying, selling and brokerage rely on access to current inforation on the availability and price of goods and services. Telecounications increases the available inforation thereby, increases the efficiency of coercial activity. Considering the findings of this paper, which suggest a causal relationship fro ICT to FDI, it sees ICT contributes to econoic growth indirectly by attracting ore foreign direct investent. Telecounications can also reduce transactions costs, widening the scope of arkets and thereby increasing copetition and efficiency. Another possible interpretation is that the growth in ICT is siply a passive consequence of developent. The advanced nations have ore telephones because they are able to afford the. In all econoic sectors anufacturing and services-advanced telecounications systes are becoing an integral part of business operations. It sees the lesser-developed countries should accelerate their application of telecounications technology or fall further behind in econoic copetitiveness.

11 7. Conclusion The relationship between foreign direct investent and econoic growth has been thoroughly studied by the previous research and there is a general agreeent about the positive ipacts of foreign direct investent on econoic developent of the host countries through the capital, technology transfer, arket access, investent opportunities and export prootion. So governents especially in developing countries have not only liberalized restrictions on FDI but also provided incentives to attract FDI. A variety of factors are cited in the literature including infrastructure, huan capital, political instability as deterinants of FDI, however we ust also take account of deeper and broader changes in the global econoy, especially the spread of the New Econoy, and the new inforation and counication technology (ICT), Recent studies show ICT has a positive effect on FDI inflows. In this study, we exained this issue using the tie series analysis tools, panel causality including LSDV and instruental variables (2SLS) estiation ethods. Results fro Granger causality test indicate that there is a significant causal relationship in the saple countries. In developed countries existing ICT infrastructure attracts FDI; higher level of ICT investent leads to higher level of FDI inflows suggesting ICT contributes to productivity and econoic growth indirectly by attracting ore foreign direct investent. But in developing countries the direction of causality goes fro FDI to ICT. In developed countries there already exist a build up ICT capacity which causes inflow of FDI, while in developing countries ICT capacity ust be build up to attract FDI. The inflow of FDI causes further increases in ICT investent and capacity. The rapid expansion of world FDI resulted fro several factors including technical progress in telecounication services and ajor currency realignent. Technical progress in telecounication services facilitates international counications involving parent copanies and their overseas affiliates, while ajor currency realignent has provided copanies with the opportunities for aking profits by undertaking FDI.

12 References Addison T., and Heshati A. (2002), Deocratization and New Counication Technologies as deterinants of foreign direct investent in developing countries, World Institute for Developent Econoics Research of the United Nations University Addison T., and Rahan A. (2002), Capacities to Globalize: Why are soe countries ore globalized than others? paper presented at the 5 th annual conference on Globalization, Growth and (In) Equality, (Forthcoing as WIDER Discussion paper, UN university) Agosin M, and Mayer R. (2000), Foreign Investent in developing countries. Does it crowd in doestic investent? UNCTAD Papers No. 46, Geneva Agrawal, Pradeep. (2000), Econoic Ipact of Foreign Direct Investent in South Asia, World Bank Aitken B., and Harrisson A. (999), Do Doestic Firs Benefit fro Direct Foreign Investent? Evidence fro Venezuela, Aerican Econoic Review, Vol. 89: 605:68. Alfaro, L., Areenda C., Sebne K, and Selin S. (200), "FDI and Econoic Growth: The Role of Local Financial Markets." Harvard Business School. Working Paper Anderson, T.W. and Cheng Hsiao, (982), Forulation and Estiation of Dynaic Models Using Panel Data, Journal of Econoetrics, vol 8 Baldwin R., Barconier H., and Forslid R., (999), Multinationals, endogenoud growth and technological spillovers: theory and evidence, CEPR discussion paper no.255. Baltagi B. H. (200), Econoetric Analysis of Panel Data, John Wiley and Son s. LTD. Blostro, M., Lipsey, R. and M. Zejan (994), What Explains the Growth of Developing Countries? in Bauol, W., Nelson, R. and E. Wolff (eds) Convergence of Productivity: Cross-National Studies and Historical Evidence, Oxford, Oxford University Press Borensztein E., De Gregorio J., and Lee J. (998), How does foreign direct investent affect econoic growth, Journal of International Econoics, Vol. 45: 5-35 Chakrabarti, A. (200), The Deterinants of Foreign Direct Investent: Sensitivity Analysis of Cross-Country Regressions, Kyklos 54 (): 89 3 Chareza, W.W. and Deadan, D.F. (992) New directions in econoetric practice. Edward Elgar, England Dickey, D.A, and Fuller, W. A. (979) Distribution of the estiations for autoregressive tie series with a unit root Journal of the Aerican Statistical Association, 74, Dickey, D. A. and W. A. Fuller (98), "Likelihood Ratio Test for Autoregressive Tie Series with a Unit Root", Econoetrica, 49, Dollar D. (992), Outward Oriented Developing Econoies Really Do Grow More Rapidly: Evidence fro 95 LDCs, , Econoic Developent and Cultural Change Dunning, J.H. (980), Toward an elective theory of international production: soe epirical tests, Journal of International Business Studies, (Spring/Suer), pp 9-3 Eaton J., and Kortu S. (997), Trade in ideas: productivity and patenting in the OECD, Journal of International Econoics, Vol. 40: Goldberg L, and Klein M. (997), Foreign Direct Investent, Trade and Real Exchange Rate Linkages in Southeast Asia and Latin Aerica, National Bureau of Econoic Research, Working Paper No. 6344, Cabridge, Mass, NBER Granger, C.W. (969), Investigating causal relations by econoetric odels and cross-spectral ethods, Econoetrica, 37 (3), Haddad M., and Harrison A. (993), Are there positive spillovers fro direct foreign investent? Evidence for panel data for Morocco, Journal of Developent Econoics, Vol.42: 5-74 Holtz-Eakin, D.; W. Newey and H. Rosen, (990), Estiating Vector Autoregressions with Panel Data, Econoetrica 56,

13 Johansen S, (988), Statistical Analysis of Co-integrating Vectors, Journal of Econoic Dynaics and Control, 988, 2, Johansen S., and K. Juselius, (990), Maxiu Likelihood Estiation and Inference on Co-integration With Applications for the Deand for Money, Oxford Bulletin of Econoics and Statistics, 52,69 20 Lunn J. (980), Deterinant of U.S. Investent in the EEC: Further Evidence, European Econoic Review 3, pp McMillan S. (999), Foreign direct investent in three regions of the South at the end of the twentieth century, New York: St. Martin's Press Manning, L.M. and Andriacanos, D. (993), Dollar oveents and inflation: a cointegration analysis, Applied Econoics, 25, Matabalya F., and Susanna W. (200), Perforance of SMEs in East Africa: Case studies fro Kenya and Tanzania, Center for Developent Research (ZEF), Bonn Markusen J., and Venables A. (999), Foreign direct investent as a catalyst for industrial developent, European econoic review, Vol. 43: OECD (99), Technology and Productivity: The Challenge for Econoic Policy. Paris OECD (998), Open Markets Matter: The benefits of trade and investent liberalisation, Paris: OECD Rappaport, J. (2000). "How Does Openness to Capital Flows Affect Growth?" Mieo, Federal Reserve Bank of Kansas City, June Roer, P., (993), Idea gaps and object gaps in econoic developent, Journal of Monetary Econoics 32, No.3.Deceber Root F.R. and A.A. Ahed (979), Epirical Deterinants of Manufacturing Direct Foreign Investent in Developing Countries, Econoic Developent and Structural Change, 27, Saggi K. (2000), "Trade, Foreign Direct Investent, and International Technology Transfer: A Survey," Working Papers - International Econoics. Trade, capital flows, World Bank Shan J., Tian G., and Sun F, (997), The FDI-led growth hypothesis: further econoetric evidence fro China, National Center for Developent Studies, The Australian National University, Research school of Pacific and Asian studies Sis, C. (972), Money, incoe and causality, Aerican Econoic Review 62, Soto, M. (2000), Capital Flows and Growth in Developing Countries: Recent Epirical Evidence, OECD Developent Centre Technical Paper No. 60, Paris UNCTAD, (2000), World Investent Report: Cross-border Merges and Acquisitions and Developent, Geneva: United Nations developent, Oxford Bulletin of Econoics and Statistics4 (4) World Developent Indicators, World Bank 2002 Xu B. (2000), Multinational Enterprises, Technology Diffusion, and Host Country Productivity Growth, Journal of Developent Econoics, Vol. 62, p

14 Country Series Degree of integration Hypothesis F statistics Result Country Series Degree of integration F statistics Result Australia Austria Brazil Canada Colobia Denark Finland France Iceland India Indonesia Ireland FDI I() H FDI I(0) GDP I(2) No No GDP I() Italy ICT I(2) H 2 ICT I(2) FDI ICT ** FDI ICT Open ** I(2) FDI I(2) H FDI I(2) GDP I() ICT FDI ** GDP I(2) No Japan ICT I() H 2 ICT I(2) No Open I() FDI ICT ** FDI I(2) H FDI I(2) GDP I() ICT FDI ** No GDP I(0) Korea ICT I(2) H 2 ICT I() FDI ICT ** No FDI I() H FDI I(2) GDP I(2) No ICT FDI ** GDP I() Malaysia ICT I() H 2 ICT I(2) No No Open I() FDI I() H FDI I() GDP I() No No GDP I() Mexico ICT I() H 2 ICT I(2) Open I() FDI ICT ** No Open I() FDI I(2) H FDI I() No.7730 No GDP I(0) GDP I(2) Norway ICT I(2) H 2 ICT No I() No Open I() FDI I() H FDI I() GDP I(2) ICT FDI ** GDP I(2) ICT FDI ** Singapore ICT I(2) H 2 ICT I() FDI ICT **.2395 FDI ICT ** FDI I(2) H FDI I(2) GDP I(2) No GDP I(2) No Sweden ICT I() H 2 ICT I(2) No FDI ICT Open ** I(2) FDI I() H FDI I() GDP I(2) ICT FDI ** No GDP I(2) Turkey ICT I(2) H 2 ICT I() No No FDI I(0) H FDI I() GDP I() No GDP I(2) ICT FDI ** UK ICT I() H 2 ICT I(2) No No FDI I(2) H FDI I(2) GDP I() ICT FDI ** GDP I() ICT FDI ** US ICT I(2) H 2 ICT I(2) No No FDI I(2) H GDP I(2) No ICT I(2) H 2 Open I() FDI ICT ** Table. Results of ADF Unit Root and Granger-Causality Tests

15 Country Nuber of Cointegrating Equation(s) Eigenvalue Trace Statistic 0% CV 5% CV Max-Eigen Statistics 0% CV 5% CV Australia Brazil Canada Colobia Denark France Indonesia Ireland Japan Kenya Korea Malaysia Norway Singapore Sweden Turkey US None At ost None At ost None At ost None At ost None * At ost * None At ost None At ost None At ost None ** At ost * None At ost None At ost None ** At ost * None * At ost None * At ost * None At ost None At ost None At ost Table 2. Results of Johansen Co-integration Test Notes: **, * 5% and 0% significance levels respectively

16 Group Hypothesis F- statistics (LSDV) Result F- statistics (2SLS) Result Developed Countries Developing Countries H ICT FDI ** ICT FDI ** H No No H No No H FDI ICT ** No Table 3. Results of panel causality tests fro LSDV and Instruental Variable Estiation Notes: **, 5% significance level, H denotes the alternative hypothesis that ICT does not Granger Cause FDI, and H 2 denotes the alternative hypothesis that FDI does not Granger Cause ICT