36
This article was downloaded by: [University of Illinois at Urbana-Champaign] On: 30 September 2013, At: 06:26 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK The International Trade Journal Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/uitj20 FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India Ranjan Kumar Dash a & Chandan Sharma b a Research Consultant, RBI Unit, Institute of Economic Growth, Delhi b Assistant Professor, Department of Economics, National Institute of Financial Management (NIFM), Faridabad, India Published online: 30 Mar 2011. To cite this article: Ranjan Kumar Dash & Chandan Sharma (2011) FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India, The International Trade Journal, 25:2, 233-266, DOI: 10.1080/08853908.2011.554787 To link to this article: http://dx.doi.org/10.1080/08853908.2011.554787 PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or

FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

  • Upload
    chandan

  • View
    213

  • Download
    0

Embed Size (px)

Citation preview

Page 1: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

This article was downloaded by: [University of Illinois at Urbana-Champaign]On: 30 September 2013, At: 06:26Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,UK

The International TradeJournalPublication details, including instructions forauthors and subscription information:http://www.tandfonline.com/loi/uitj20

FDI, Trade, and GrowthDynamics: New Evidence fromthe Post-Reform IndiaRanjan Kumar Dash a & Chandan Sharma ba Research Consultant, RBI Unit, Institute ofEconomic Growth, Delhib Assistant Professor, Department of Economics,National Institute of Financial Management (NIFM),Faridabad, IndiaPublished online: 30 Mar 2011.

To cite this article: Ranjan Kumar Dash & Chandan Sharma (2011) FDI, Trade, andGrowth Dynamics: New Evidence from the Post-Reform India, The International TradeJournal, 25:2, 233-266, DOI: 10.1080/08853908.2011.554787

To link to this article: http://dx.doi.org/10.1080/08853908.2011.554787

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all theinformation (the “Content”) contained in the publications on our platform.However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness,or suitability for any purpose of the Content. Any opinions and viewsexpressed in this publication are the opinions and views of the authors, andare not the views of or endorsed by Taylor & Francis. The accuracy of theContent should not be relied upon and should be independently verified withprimary sources of information. Taylor and Francis shall not be liable for anylosses, actions, claims, proceedings, demands, costs, expenses, damages,and other liabilities whatsoever or howsoever caused arising directly or

Page 2: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

indirectly in connection with, in relation to or arising out of the use of theContent.

This article may be used for research, teaching, and private study purposes.Any substantial or systematic reproduction, redistribution, reselling, loan,sub-licensing, systematic supply, or distribution in any form to anyone isexpressly forbidden. Terms & Conditions of access and use can be found athttp://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 3: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

FDI, Trade, and Growth Dynamics:

New Evidence from the

Post-Reform India

Ranjan Kumar Dash

Research Consultant, RBI Unit, Institute ofEconomic Growth, Delhi

Chandan SharmaAssistant Professor, Department of Economics, National

Institute of Financial Management (NIFM), Faridabad, India

This article has two aims: first, review the related literature onthe relationship among Foreign Direct Investment (FDI), Trade,and Economic Growth, and second, empirically examine the link-age between these variables. Our review of the literature indicatesthat despite a large volume of literature on the relationship amongthese variables, the direction of causality among them is far fromover. Therefore, after observing a gap in the empirical literature,especially for developing economies, we investigate the linkage byusing a Vector Autoregression (VAR) model applying the Gangernon-causality test of Toda and Yamamoto (1995) for the period1991Q3 to 2006Q3. The evidence shows that there is bidirectionalcausality (two-way feedback) between FDI and economic growth.At the same time, there is also a unidirectional causality exists

Correspondence should be addressed to Chandan Sharma,Assistant Professor, Department of Economics, NationalInstitute of Financial Management (NIFM), Sector 48,Faridabad 121001, Haryana, India. E-mail: [email protected]

233THE INTERNATIONAL TRADE JOURNAL, Volume 25, No. 2, April–June 2011

ISSN: 0885-3908 print/1521-0545 online. DOI: 10.1080/08853908.2011.554787

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 4: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

234 THE INTERNATIONAL TRADE JOURNAL

between exports and FDI, which runs from the former to thelatter. Results of the test of causality between FDI and importsindicate the presence of a two-way feedback relationship betweenthe variables.

KEYWORDS FDI, trade, causality, Toda-Yamamoto test,linear decomposition

* * * * *

I. INTRODUCTION

Foreign Direct Investment (FDI)1 has figured prominentlyin the recent economic history of most developing countries,especially since the 1980s, when FDI flows to those countrieshave been directed increasingly at export-oriented projects. Sincethen, developing countries have been competing aggressively toattract foreign investors. Virtually all countries, through changesin their regulatory environments, have facilitated the expansionof foreign investment. The economic rationale for offering spe-cial incentives to attract FDI frequently derives from the beliefthat foreign investment produces externalities in the form oftechnology transfers and spillovers.

1According to the IMF, foreign direct investment (FDI) is the category ofinternational investment that reflects the objective of obtaining a lasting interest bya resident entity in one economy in an enterprise resident in another economy. Thelasting interest implies the existence of a long-term relationship between the directinvestor and the enterprise and a significant degree of influence by the investor onthe management of the enterprise. (See Balance of Payments Manual 5th Edition,IMF, http://www.imf.org/external/np/sta/bop/iss.htm)In India, FDI comprises of inflows through

(i) Government (the Secretariat for Industrial Assistance (SIA)/ForeignInvestment Promotion Board (FIPB)) route,

(ii) Reserve Bank of India (RBI) automatic route,

(iii) non-resident Indian (NRI),

(iv) acquisition of shares

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 5: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 235

Inflows of world FDI increased tremendously from US$200billion in 1990 to almost US$916 billion in 2005 (UNCTAD2006). There are ongoing discussions on the impact of FDI ona host country economy, as can be seen from recent surveys ofthe literature (Blomstrom et al. 1992; Borensztein, Gregorio andLee 1998; de Mello 1997, 1999; Fan 2002; Gregorio 1992; Lim2001). Most of the studies find positive effects of FDI on long-run economic growth through capital accumulation and technicalor knowledge transfers, especially under an open trade regime(e.g., Basu, Chakraborty and Reagle 2003). Romer (1994) arguesthat through technology transfer, foreign investment can ease thetransfer of technological and business know-how to poorer coun-tries. These transfers may have substantial spillover effects forthe entire economy. According to Summers (2000) “social ben-efits” can be made by the efficiency gains from the reallocationof capital from industrial to developing countries. The realloca-tion can improve living standards by mobilizing global savings tofinance investments in countries where the marginal productivityof investment is relatively high.

The positive relationship between FDI and growth, whichholds to varying degrees across various regions and over time,also depends crucially on a country’s absorptive capacity (Alfaroet al. 2003; Borzenstein et al. 1998; Durham 2004; Edison et al.2002). These initial conditions which capture the absorptivecapacity of host countries include the initial level of develop-ment (Blomstrom et al. 1992), existing human capital develop-ment (Borensztein et al. 1998), trade policy (Balasubramanyamet al. 1996), financial development (Alfaro et al. 2003; Durham2003), legal-based variables (Durham 2004; Edison et al. 2002),and general government policy (Edison et al. 2002). Recently,researchers recognized the fact that quality of FDI is more impor-tant than quantity alone. FDI is considered as higher-quality if itis export-oriented, transfers foreign technologies to the host coun-try, and induces economic spillover benefiting local enterprisesand workers (Enderwick 2005).

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 6: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

236 THE INTERNATIONAL TRADE JOURNAL

Developing countries around the world have been liberalizingtheir trade regimes and moving away from import-substitutioninvestment regimes to export-promotion development policies.Trade policy reform by countries often involves extensive invest-ment by governments in order to attract FDI, partly because ofa perceived link between FDI and the improved export competi-tiveness of the host country (see UNCTAD 2003). The potentialimportance of the export enhancing role of FDI for host countrieshas been recognized in a number of country-specific studies (forexample, the United Kingdom (Blake and Pain 1994), Portugal(Cabral 1995), Europe (Pain and Wakelin 1998), Ireland (Barryand Bradley 1997), a 52-cross-country study (UNCTAD 1999),Brazil (Oliveira 2002), China (Sun 2001; Zhang 2005). A cross-country analysis for 52 countries (UNCTAD 1999) suggested apositive relationship between FDI and manufactured exports;the relationship was stronger for developing than developedcountries. Similarly, a study by Vuksic (2006) for 13 transitioncountries of Central and Eastern Europe found that FDI playeda big role in export performance.

An empirical assessment of the role of FDI on host country’sexport performance is important, since exports have been for along time viewed as an engine of economic growth (Beckerman1965). There is a widely shared view that FDI promotes exportsof host countries by (a) augmenting domestic capital for exports,(b) helping transfer of technology and new products for exports,(c) facilitating access to new and large foreign markets, and(d) providing training for the local workforce and upgradingtechnical and management skills (Caves 1996; UNCTAD 2003).

Since India began economic reforms in early 1990’s, the econ-omy has gradually opened up to the rest of the world withincreasing FDI inflows and international trade. Its accession tothe World Trade Organization (WTO) in 1994–95 made it moreclosely integrated into the World economy. A high average annualGDP growth rate of 6.6 per cent from 1991 to 2006 and gradual

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 7: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 237

improvement of its market mechanisms attract world-wide atten-tion with emerging investment opportunities and a huge marketsize. At the same time, growth of foreign trade which was 17% ofGDP in 1991, increased to 44% by the end of 2005. Many multi-national enterprises have been shifting their production base toIndia due to the wage advantage along with availability of highlyskilled workers.

Given the importance of the impact of FDI on a developingcountry like India, however, there exists only a few studies, partic-ularly at macroeconomic level. For example, a number of studieshave investigated the export performance of foreign owned firmsin India for various periods of time (Kumar 1994; Lall andMohammad 1983). In general the export performance of the for-eign owned firms is found to be no better than that of locallyowned firms. Both groups appear to have targeted the shelteredprofitable domestic markets rather than the highly competitiveexport markets during the pre-1991 phase. Sharma (2000) exam-ined the contribution of FDI in India’s export performance for theperiod 1970–98. His empirical result indicates that FDI have notplayed any significant role in export growth. However, this studydoes not cover entire reform period. Aggarwal (2001) also reachedsimilar conclusion, using firm level data. Pailwar (2001) arguesthat India has not been able to attract FDI in export-orientedareas. Banga (2003) found that FDI not only led to export diver-sification but also indirectly improved export via export spilloversin India. According to the Asian Development Bank (ADB; 2004)FDI accounts for about 3% of India’s exports compared to 50%in various East Asian host countries.

Most of the studies on the growth impact of FDI in India arein striking contrast to the current facts. Pradhan (2002), whoestimates a Cobb-Douglas production function using aggregatedata for 1969–97 with FDI stocks as additional input variable,found that FDI has no significant impact on growth. However,a study by Sahoo and Mathiyazhagan (2003), using Johansen

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 8: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

238 THE INTERNATIONAL TRADE JOURNAL

co-integration technique for the period 1979–2001, finds a long-run relationship among Gross Domestic Product (GDP), FDI,and exports. Chakraborty and Basu (2002), using aggregate datafor 1974–1996, report a one-way causality runs from GDP to FDI.In the long-run, FDI is positively related to GDP and opennessto trade.

Dua and Rashid (1998) report similar results. Kumar andPradhan (2002) consider the FDI-growth relationship to beGranger neutral for India as the direction of causation is notpronounced. Bhat et al. (2004) report similar conclusions byapplying Granger causality. Agrawal (2005) found the growthimpact of FDI is negligible by applying panel estimates for fiveSouth Asian countries including India for the period 1965–1996.

It is clear from the previously mentioned fact that most ofthe studies confined to the mid-1990s may fail to capture theeffects of the changing policy environment in the post-reformperiod. Therefore, one can question the findings of these studiesin the present condition. Against this background, the presentstudy investigates the causal nexus between FDI, trade andindustrial production. Our study differs from existing study infive ways- First; we mainly focus on the reform period only;given the fact that before liberalization there was negligible levelFDI flows to India. Second, we use quarterly data for our pur-pose. Third, we have gone beyond the traditional two-variablerelationship by building four-variable VAR model to avoid pos-sible specification bias. Fourth, the methodology of Toda andYamamoto (1995) modified granger no-causality test is expectedto improve the standard F-statistics in the causality process.Fifth, Geweke decomposition test is used to find out the extentof linear feedback between FDI and trade.

The remainder of the article is organized as follows. Section 2deals with theoretical back ground of the article. Section 3presents a brief discussion on FDI policy and flows in India.Section 4 explains data source and empirical methodology. In

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 9: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 239

section 5, we discuss empirical results, followed by conclusionsand policy implications.

II. THEORETICAL BACKGROUND

FDI and Growth

In the Solow-type standard neoclassical growth models, FDIis traditionally conceived of as an addition to the capital stock ofthe host economy (Brems 1970). In this view, there are no sub-stantial differences between domestic and foreign capital. Moreimportantly, the impact of FDI on growth is similar to that ofdomestic capital. With diminishing returns to capital, FDI hasno permanent impact on the growth rate. FDI will have, however,a short-run impact on growth, which depends on the transitionaldynamics to the steady-state growth path.

However, new growth theories incorporate the role of knowl-edge or technology endogenously as a factor of production inits own right and provide for the possibility of non-diminishingreturns to capital (Grossman and Helpman 1991; Romer 1994.)There are a number of conceivable channels through which FDIpermanently affects the growth rate. A convenient way to thinkabout these effects is by separating out how FDI affects eachargument in the production function. FDI can affect output byincreasing the stock of capital. However, this impact is likelyto be small under the assumption of perfect substitutability.Although the empirical evidence on this matter is ambiguous(Hanson 2001), if foreign and domestic capital are complementsthe final impact of FDI on aggregate output will be larger as aresult of these externalities.

One can also think about the impact of FDI on labor. Onceagain, the expected impact is small and in this case it will be interms of job creation. Yet, the role of FDI as knowledge and tech-nology transfer becomes even more apparent as FDI has clearly a

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 10: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

240 THE INTERNATIONAL TRADE JOURNAL

more import role in the augmentation of human capital than onthe numbers employed. Consider the case in which foreign invest-ment is carried out in activities in which the host economy haslimited previous experience. In this case FDI will entail impor-tant knowledge transfers in terms of training of the labor force,skills acquisition, new management practices, and organizationalarrangements.

The last and arguably the most important venue throughwhich FDI could affect economic growth is transfer of technol-ogy. FDI inflows directly raise the levels of technology in thehost economy. That can be due to a variety of mechanisms.One plausible mechanism is that FDI inflows increase the vari-ety of intermediate products and types of capital equipment inthe host economy (Borensztein et al. 1998). In doing so, FDIinflows lead to an increase of the productivity in the host econ-omy. Another important mechanism through which FDI affectsgrowth is learning. FDI inflows diffuse knowledge about pro-duction methods, product design, and new organizational andmanagerial techniques. In this light, imitation becomes a crucialelement. Another important mechanism is that FDI raises theproductivity of domestic research and development activities.

FDI and Trade

There has been, traditionally, a divergence in terms of thedevelopment of the theories on FDI and international trade.Trade theories attempt to explain why countries trade with eachother and FDI theory tries to account for why firms produceabroad and invest in particular countries. In the neoclassicalapproach of the trade theory, the article of Mundell (1957) wasthe first to focus on the relationship between capital movementsand trade of commodities. In the Heckscher-Ohlin-Samuelson(HOS) framework, in taking account of the assumptions of per-fect competition and constant economies of scale, Mundell arguedthat a tariff protection would generate a perfect substitutionbetween capital movements and trade of commodities.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 11: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 241

Some earlier theoretical work has predicted either a sub-stitute or complementary relationship between FDI and trade.These models are based on the imperfect competition, theeconomies of scale, the difference in production technologies,etc. Some have focused mainly on either vertical or horizontalFDI.2 In the first case, firms separate geographically their dif-ferent stages of the value-added chain. In the second case, firmsduplicate the entire production process in several countries withan exception for headquarters activities.

The models of Helpman (1984) and Helpman and Krugman(1985) integrate vertical FDI into international trade theory.They show that FDI generates complementary trade flows offinished goods from foreign affiliates to parent companies or tothe home country and intra-firm transfers of intangible head-quarters services from parent companies to foreign affiliates. Onthe other hand, in the models based on horizontal FDI, suchas Markusen (1983), Brainard (1993), Horstmann and Markusen(1992), Markusen (1995), and Markusen and Venables (1998),foreign investments are alternatives modalities. The choice ofmultinational firms depends on the interaction among these keyelements: the firm specific advantages (activities of research anddevelopment, managerial know-how, etc.), the plant-level scaleeconomies, and transport, geographical, and cultural distancecosts. In these models, the substitutability between FDI andtrade prevails over complementarity. According to the mod-els of Brainard (1993) and Horstmann and Markusen (1992),when countries are identical in technologies, preferences, andfactor endowments, the higher the value of firm-level scaleeconomies and tariffs and transport costs relative to plant-level

2Horizontal FDI is defined as investment in the same industry abroad as a firmoperates in at home. In contrast Vertical FDI takes two forms:

(1) backward vertical FDI: where an industry abroad provides inputs for afirm’s domestic production processes,

(2) forward vertical FDI: in which an industry abroad sells the outputs of afirm’s domestic production processes.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 12: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

242 THE INTERNATIONAL TRADE JOURNAL

scale economies, the more likely is the presence of horizontalFDI. These models based on the trade-off between proximityand concentration postulate a substitution relationship betweenhorizontal FDI and trade.

Markusen and Venables (1998) further elaborated on thetheory to introduce asymmetries between countries in terms ofmarket size, factor endowments, and technologies. Countries’asymmetries make it possible for national and multinationalfirms and, therefore, trade and FDI to coexist. However, ascountries become more similar in market size, relative factorendowments, and technical efficiency, FDI will increase and inter-national economic activity will become increasingly dominated bymultinational enterprises (MNEs), which displace trade, providedthat transport costs are not very small.

The contributions within the theoretical literature show noambiguity in the relationship among FDI, international trade,and economic growth. However, empirical findings on this issueare inconsistent and often contrary to each other. Also, empiri-cal studies mostly presented evidence from the developed world.We fill this existing gap between theories and empirical evidenceand provide new evidence from a developing country, in our re-examination of the linkage for India. The findings of this study,hopefully, will fill the existing gap in the literature and provide anew insight to policy makers in India as well in other developingeconomies in formulating policies in this issue.

III. FDI POLICY AND FLOWS IN INDIA

FDI Policy in India

In the 1980s the Indian economy had witnessed a change ineconomic policy. However, a major shift in the attitude of gov-ernment is observed only after the 1991 economic reforms. Toaccelerate economic growth, many policies were undertaken; FDIregulation and liberalization were one of them. Since then, FDI

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 13: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 243

policy has been continuously revised and liberalized. As a result,most of the sectors of the economy are open to FDI throughthe automatic approval route. However, in a few sectors somerestrictions still exist.3 The restriction on the sectors is imposedmainly for a sector’s requirements and keeping in mind the coun-try’s defense concerns. Under the current policy, FDI can reachIndia via two routes, the automatic route and the prior govern-ment approval route. In some cases where the automatic routeis not permitted, investment can be approved by the ForeignInvestment Promotion Board (FIPB). Some other economic andindustrial policies such as repatriation of investment capital andprofits, change in labor laws, establishment of Special EconomicZones (SEZs),4 and taxation policy have helped India to becomean attractive destination of FDI.

FDI Inflows to India

With the liberalization of the Indian economy, inflows of FDIhave gone up substantially. The total cumulative inflow of FDIwas $54 billion for the period from August 1991 to March 2007(RBI Annual Report , 2007). The annual growth rate of FDI flowswas 51% during the same period.5 FDI has gained prominencein relative terms too. FDI flows accounted for 5.1% of gross fixedcapital formation in 2001–2006. Compared with all developingcountries (10.5% in 2004) and China (14.9% in 2004), this shareis still low (see UNCTAD online database, 2004). The reformsdid result in increased inflows of FDI during the 1990s. Even so,

3In a few sectors, there are restrictions on FDI investment in the form of equitycaps, divestment conditions, and lock in periods on investment. Moreover, FDI isnot allowed in gambling, betting retail, arms and ammunition, etc.

4Inspired by the success of China in attracting FDI (especially export-related)and in labor-intensive exports, India has started Special Economic Zones (SEZs) onan experimental basis.(for details, see Virmani 2004)

5Before the initiation of the economic reforms (1991), inflows of FDI into thecountry were negligible.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 14: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

244 THE INTERNATIONAL TRADE JOURNAL

Table 1Country-wise FDI Inflows and Corresponding Trade Share

Serial no. Country

Amount ofFDI Inflows

(In Million $)% With

Total Inflows

% ExportShare

(1991–2005)

% ImportShare

(1991–2005)

1. Mauritius 18, 481 44.25 0.34 N.A.2. United States 5, 283.29 15.19 18.73 7.353. Japan 2, 149.71 6.31 4.30 4.194. Netherlands 2, 034.26 6.05 2.18 0.945. U.K. 2, 054.24 6.01 5.31 4.606. Germany 1, 602.24 4.61 4.65 4.847. Singapore 1, 085.56 3.17 3.30 2.678. France 802.76 2.36 2.16 1.709. South Korea 763.89 2.07 0.01 4.55

10. Switzerland 649.80 1.87 0.80 4.55

Source: Department of Industrial Policy and Promotion, Ministry ofCommerce and Industry (http://dipp.nic.in/fdi statistics/india fdi index.htm),and WTO Online Database, 2006 (http://databank.worldbank.org).

the volume of FDI in India is relatively low compared to EastAsian countries and China.

Country-wise, FDI inflows indicate that Mauritius accountsfor the bulk of the total FDI inflows ($12.7 billion) and con-stituted 38.06% of the total FDI inflows from August 1991 toOctober 2007 (see Table 1). The United States stands a distantsecond with a 15.19% share. The remaining 47% is shared bythe rest of the countries. Increased FDI from Mauritius maybe attributed to the double taxation treaty that India signedwith Mauritius and also to the fact that most investment intoIndia from the United States is being routed through Mauritius.Although FDI inflows from Mauritius are the highest, Mauritiusis a minor trading partner of India, accounting for only 0.34%of India’s total export. However, the United States is the majortrading partner of India with 18.73% of the share in total exportand 7.35% of the share in total import.

The post-reform period is not only characterized by boomingFDI, but the sector- and industry-wise composition of FDI hasalso changed considerably. There is a sharp shift from FDI in

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 15: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 245

the primary and manufacturing sectors to FDI in services sincethe mid-1990s. Sector-wise FDI composition from August 1991 toJune 2006 reveals that electrical equipments (including computersoftware and electronics) accounts for the highest percentageshare (17%) in total FDI inflows, followed by telecommunica-tion (11%) and the service sector with 10% share (see Table 2).However, the only striking feature about the sectors receivingthe most FDI in India is that most FDI is coming into transport,electrical equipment, infrastructure, etc., but very little is flowinginto India’s leading export sectors. India should make an effortto attract FDI in its export sectors such as gems and jewelry,pharmaceuticals, textiles, and marine products and this wouldbe one way of stepping up FDI overall.

Table 2Sector-wise FDI Inflows in India (August 1991 to June 2006 )

Serial no. Sector

FDI Inflows(In Million

US$)

% Sharein FDIInflows

% Sharein TotalImports

% Sharein TotalExports

1. Electrical Equipments(including computersoftware &electronics)

5, 618.97 16.93 N.A. N.A.

2. Telecommunication 3, 691.57 11.01 4.6 0.893. Service Sector 3, 484.78 10.17 N.A. N.A.4. Transportation

Industry3, 360.41 9.86 10.71 3.06

5. Fuels (Power & OilRefinery)

2, 683.47 7.98 N.A. N.A.

6. Chemicals (Otherthan Fertilizers)

2, 161.04 6.04 6.42 6.41

7. Food ProcessingIndustries

1, 200.48 3.35 N.A. 0.70

8. Drugs andPharmaceuticals

1, 049.95 3.14 0.49 2.44

9. Cement and GypsumProducts

747.23 2.26 N.A. 0.99

10. MetallurgicalsIndustries

712.25 2.15 N.A. N.A.

Source: Department of Industrial Policy and Promotion, Ministry ofCommerce and Industry (http://dipp.nic.in/fdi statistics/india fdi index.htm), and WTO Online Database, 2006 (http://databank.worldbank.org).

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 16: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

246 THE INTERNATIONAL TRADE JOURNAL

Figure 1Exports and Imports in India (in $US millions), 1991:3–2006:3

0

10000

20000

30000

40000

50000

1991

Q1

1991

Q3

1992

Q1

1992

Q3

1993

Q1

1993

Q3

1994

Q1

1994

Q3

1995

Q1

1995

Q3

1996

Q1

1996

Q3

1997

Q1

1997

Q3

1998

Q1

1998

Q3

1999

Q1

1999

Q3

2000

Q1

2000

Q3

2001

Q1

2001

Q3

2002

Q1

2002

Q3

2003

Q1

2003

Q3

2004

Q1

2004

Q3

2005

Q1

2005

Q3

2006

Q1

2006

Q3

Exports Imports

According to the data on approvals for country-wise tech-nology transfers, the United States stands at the top with1704 approvals, followed by Germany (1099), U.K. (849), andJapan (849) (see Table 3). Approvals for sector-wise technologytransfers indicate that the electrical equipment sector (includ-ing computer software and electronics) has the highest numberof approvals (1250) followed by chemicals (other than fertilizer)(882) and industrial machinery (870). Quarterly FDI inflows areshown in Figure 1, which suggests a substantial growth in theinflow in the India in the reform period (since 1991). The inflowwas merely US $19 million in 1991:Q3, which increased to aroundUS $3600 million in 2006:Q3.

IV. DATA SOURCE AND EMPIRICALMETHODOLOGY

Empirical Hypothesis

To examine the FDI, export, import, and output (eco-nomic growth) relationship, we tested two important empiricalhypotheses as follows:

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 17: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 247

Table 3Country-wise Technology Transfer Approvals in India (August 1991

to June 2006)

Ranks Country

No. of TechnicalCollaborations

Approved Sector

No. of TechnicalCollaborations

Approved

1. UnitedStates

1, 704 ElectricalEquipments(includingcomputer software& electronics)

1, 250

2. Germany 1, 099 Chemicals (otherthan fertilizer)

882

3. U.K. 849 Industrial Machinery 8704. Japan 849 Transportation

Industry719

5. Italy 479 Misc. Mach.EngineeringIndustry

437

6. Othercountries

2, 812 Other sectors 3, 634

All countries 7, 792 Total of all Sectors 7, 792

Source: Department of Industrial Policy and Promotion, Ministry ofCommerce and Industry.

Hypothesis 1: FDI causes export, import, and economic growth.Hypothesis 2: export, import, and economic growth cause FDI.

Data and Data Sources

To test both of the set empirical hypotheses, we use quar-terly data of FDI, the Index of Industrial Production (IIP),and figures for total export of goods and services (export)and import of goods and services (import), which are culledfrom the International Monetary Fund’s International FinancialStatistics (IFS) CD-ROM, 2007. The sample range is from1991Q3 to 2006Q3, which comprises 61 observations. The year1991 was selected as the start of the sampling period as economic

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 18: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

248 THE INTERNATIONAL TRADE JOURNAL

Figure 2FDI in India (in $US millions), 1991:3–2006:3

0

1000

2000

3000

4000

1991

Q1

1991

Q3

1992

Q1

1992

Q3

1993

Q1

1993

Q3

1994

Q1

1994

Q3

1995

Q1

1995

Q3

1996

Q1

1996

Q3

1997

Q1

1997

Q3

1998

Q1

1998

Q3

1999

Q1

1999

Q3

2000

Q1

2000

Q3

2001

Q1

2001

Q3

2002

Q1

2002

Q3

2003

Q1

2003

Q3

2004

Q1

2004

Q3

2005

Q1

2005

Q3

2006

Q1

2006

Q3

FDI Inflows

reform was initiated in India in this year and subsequently FDIinflow rose significantly. We have used the Index of IndustrialProduction (IIP)6 as proxy for real GDP because quarterly dataon GDP were not available. All variables are seasonally adjusted(except FDI), and transformed into logarithms. Both export dataand import data are converted into real terms by deflating WPIseries because quarterly data on units of export and importprices are not available. The data of these variables are shown inFigures 1–3. Their movements suggest that there may be aco-movement among these variables.

Empirical Methodology

In this analysis we employ the Johansen and Juselius proce-dure of testing for the presence of multiple co-integrating vectors.The Johansen (1991) method of multivariate approach is a well-established model to trace out co-integration relationships amongthe time series variables. We utilize this approach in this arti-cle to find out the co-integrating relations among FDI, exports,

6IIP is a reliable leading indicator of business cycles in India and correlationbetween GDP and IIP is very strong (0.97 on 1% significant level) (Mazumdar2005).

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 19: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 249

Figure 3Index of Industrial Production (IIP) in India, 1991:3–2006:3

50

70

90

110

130

150

1991

Q1

1991

Q3

1992

Q1

1992

Q3

1993

Q1

1993

Q3

1994

Q1

1994

Q3

1995

Q1

1995

Q3

1996

Q1

1996

Q3

1997

Q1

1997

Q3

1998

Q1

1998

Q3

1999

Q1

1999

Q3

2000

Q1

2000

Q3

2001

Q1

2001

Q3

2002

Q1

2002

Q3

2003

Q1

2003

Q3

2004

Q1

2004

Q3

2005

Q1

2005

Q3

2006

Q1

2006

Q3

IIP

imports, and IIP. A causality test is done using the Toda andYamamoto (1995) modified causality test, and Geweke’s (1982)decomposition method is used to find the extent of linear depen-dence between FDI and trade. The first step of this processinvolves a test for stationarity to know the order of integrationof the variables. For this purpose, an Augmented Dickey-Fuller(ADF) test for unit roots has been carried out. Once the order ofeach variable is determined, we perform the co-integration analy-sis to determine whether the time series of these variables displaya stationary process in a linear combination.

Johansen’s Method of Co-Integration

The maximum likelihood approach of Johansen and Juselius(1990) is used to establish whether there is a long-run relationshipbetween the variables in the model. The model is based on theerror correction representation given by

�Xt = μ +p−1∑

i=1

�i�Xt−k + �Xt−k + εt(1)

Where Xt is an (nx1) column vector of p variables, μ is an (nx1)vector of constant terms, � and � represent coefficient matrices,

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 20: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

250 THE INTERNATIONAL TRADE JOURNAL

� is a difference operator, k denotes the lag length [Show k inequation (1)], and εt is a disturbance term independently andidentically distributed with zero mean and constant variance.The coefficient matrix � is known as the impact matrix andit contains information about the long-run relationships.

In order to apply Johansen’s technique one is required to esti-mate the VAR equation (1). The residuals from the estimatedequation are then used to compute two likelihood ratios (LR)test statistics that can be used in the determination of the uniqueco-integrating vectors of Xt. The first test, which considers thehypothesis that the rank of � is less than or equal to r (the num-ber of co-integrating vectors), is based on the trace test statistic(λtrace) given below:

λtrace (r) = −Tn∑

i=r+1

ln(1 − λi)(2)

where λi are the estimated values of characteristic roots or theeigen values, T is the number of observations, and n is the numberof variables. The second test statistic is known as the maximaleigen value test statistic(λmax) which tests the null hypothe-sis that there are exactly r co-integrating vectors in Xt and isgiven by:

λmax (r + 1) = −T ln(1 − λr)(3)

The asymptotic critical values for these likelihood ratio tests arecalculated via numerical simulations (see Johansen and Juselius1990).

Toda and Yamamoto Version of Granger Causality

Toda and Yamamoto (1995) proposed an alternative pro-cedure for testing Granger causality in an integrated and

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 21: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 251

co-integrated framework. This methodology is useful because itbypasses the need for potentially biased pre-tests for unit rootsand co-integration common to other formulations such as thevector error correction model.

The Toda and Yamamoto (1995) procedure uses a modifiedWald (MWALD) test to examine restriction on the parameters ofthe VAR (k) model, where k is the optimal lag length of the VARmodel. This test has an asymptotic chi-squared distribution withk degrees of freedom in the limit when a VAR [k + d(max)] isestimated (where d(max) is the maximal order of integration forthe series in the system). Two steps are involved for implementingthe procedure.

The first step includes determination of the lag length (k) andthe maximum order of integration (d) of the variables in the sys-tem. Measures such as the Akaike Information Criterion (AIC)or Schwartz Bayesian Criterion (SBC) can be used to determinethe appropriate lag structure of the VAR. Once the VAR (k) isselected, and the order of integration d(max) determined, levelsVAR can then be estimated with a total of p = [k + d (max)]lags. The second step is to apply standard Wald tests to thematrix of the first k VAR coefficient (but not the coefficientsof the lagged terms) to conduct inference on Granger causality.Rambaldi and Doran (1996) have shown that the MWALD testsfor testing Granger no-causality experience efficiency improve-ment when Seemingly Unrelated Regression (SUR) models areused in the estimation. Moreover, the MWALD test statistic isalso easily computed in the SUR system.

Decomposition of Causality

After establishing the causality between variables, it is impor-tant to know the extent of linear dependence of the variables.Geweke (1982) constructed a measure of linear dependencebetween two time series (Xt and Yt} based on the notion of

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 22: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

252 THE INTERNATIONAL TRADE JOURNAL

Granger causality.7 Geweke expanded the Granger’s notion ofcausality by including instantaneous feedback between Xt andYt. We briefly describe the methodology as follows:

Yt = α1 +p∑

s=1

β1sYt − s + u1t.(4)

Yt = α2 +p∑

s=1

β2sYt − s +p∑

s=1

λ2sXt − s+u2t.(5)

Yt = α3 +p∑

s=1

β3sYt − s +p∑

s=0

λ3sXt − s+u3t.(6)

Xt = α4 +p∑

s=1

λ4sXt − s+u4t.(7)

Xt = α5 +p∑

s=1

β5sYt − s +p∑

s=1

λ5sXt − s+u5t.(8)

Xt = α6 +p∑

s=1

β6sYt − s +p∑

s=0

λ6sXt − s+u6t(9)

Regression 4 and 5 measure the Granger causality runningfrom Xt to Yt and regression 7 and 8 measure the Granger causal-ity running from Yt to Xt. The instantaneous causality betweenYt and Xt} is measured either by regression 6 or 8. FollowingGeweke’s notation, the variance of residuals in each regressionis labeled as Var (u1t) = T1, Var (u2t) = T2, Var (u3t) = T3,Var (u4t) = T4, Var (u5t) = T5, Var (u6t) = T6. The measure of

7Zellner (1982) raised some serious questions on the use of the word “causality”in Geweke (1982) methodology and argued that it is identified with feedback ordependence with nearly correlation. Therefore, it is a loose usage of “causality.”However, in the standard literature, the term is frequently used by authors in thismethodology (for instance, see Chong and Calderon 2000; Calderon and Liu 2003;and Aizenman and Noy 2006).

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 23: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 253

Granger causality (Cx → y) from Xt to Yt is defined as Cx → y= ln (T1

T2 ). Similarly, the measure of Granger causality (Cy → x)from Yt to Xt is defined asCy → x = ln (T4

T5) and the instanta-neous causality (Cx ↔ y) between Xt to Yt is defined asCx ↔ y= ln (T2

T3) = ln (T5T6).

The overall measure of linear dependence Cx,y is calculated as

Cx ,y = Cx→y + Cy→x + Cx↔y(10)

Since the Geweke’s measures are simply the logarithm of like-lihood ratios, their statistical significance can be readily testedby applying conventional likelihood ratio tests. The null hypoth-esis (H0: C = 0) can be statistically examined using the χ2

distribution.

V. ESTIMATION RESULTS

To set the stage for the Toda-Yamamoto test, the order ofintegration of the variables is initially determined using the ADFtest. The testing procedures are based on the null hypothesisthat a unit root exists in the autoregressive representation of theseries. The test fails to reject the null hypothesis of a unit rootfor each of the series in level. First differences of log level of theseseries are stationary, suggesting that all series are integrated oforder one, 1(1) (see Table 4).

Since the order of the stationary I (1) is the same for all fourvariables, Johensen multivariate co-integration test is applied forfinding the order of co-integration d (max). Using the maxi-mum eigenvalue test as well as trace test, the null hypothesisof at most one co-integration can be rejected, indicating twoco-integrating vectors exist among the variables (see Table 5).Therefore, the result supports the hypothesis of co-integrationamong export, import, FDI, and IIP. We can conclude that there

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 24: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

254 THE INTERNATIONAL TRADE JOURNAL

Table 4Test for Unit Root Applying Augmented Dickey-Fuller (ADF) Test

VariablesTest statistics

(levels)Optimal Lags

(AIC)Test statistics(1st difference)

Optimal Lags(AIC)

FDI −1.40 1 −6.30∗ 1IIP 0.90 4 −5.64∗ 2Export 0.963 4 −2.99∗ 4Import 3.423 1 −3.48∗ 1

Notes: ∗denotes that the null hypothesis that the variable concerned is non-stationary and can be rejected at the 5% significance level. Asymptotic cut-offvalues for the 5% significance level are −3.41 when the trend term is includedand −2.86 when the trend tern is not included (see Davidson and Mackinnon1993).

Table 5Johansen Co-integration Test (Max-Eigenvalue and Trace Statistics)

Nullhypothesis

Test Statistics(Max-Eigenvalue) 5% CV

Test Statistics(Trace Statistics) 5% CV

r = 0 34.53∗ 28.2700 83.62∗ 53.48r≤1 30.83∗ 22.0400 49.08∗ 34.87r≤2 15.17 15.8700 18.25 20.18r≤3 3.08 9.1600 3.08 9.16

Notes: r is the number of co-integration vector under null hypothesis of noco-integration. We include a linear deterministic trend as suggested by theDoldado, Jenkinson, and Sosvilla-Rivero (1990) procedure. Both the trace testand the max-eigenvalue test indicate two co-integration vectors at the 5% level.The lag order in the VAR process is 1.

exists a long-run relationship among FDI, imports, exports, andthe industrial production in India.

Having established the long-run relationship, the next stepof the study is to estimate a Granger–causality test by usingthe Toda-Yamamoto (1995) procedure. The optimal lag length(k), chosen by the Schwarz Bayesian criterion (SBC), suggesteda VAR of order 1. Then we estimated a system of VAR atlevels, using total lag length 2 (dmax (1) + k (1)). The Toda

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 25: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 255

Yamamoto (1995) augmented the Granger causality test and hasbeen obtained in the present study by estimating the followingVAR model:

FDIt = a1 +3∑

j=1

b1jFDIt−j

3∑

j=1

c1jExportt−j

+3∑

j=1

d1j IIPt−j +3∑

j=1

e1j importt−j + u1t

(10)

Exportt = a2 +3∑

j=1

b2jFDIt−j

3∑

j=1

c2jExportt−j

+3∑

j=1

d2j IIPt−j +3∑

j=1

e2j importt−j + u2t

(11)

importt = a3 +3∑

j=1

b3jFDIt−j

3∑

j=1

cijExportt−j

+3∑

j=1

d3j IIPt−j +3∑

j=1

e3j importt−j + u3t

(12)

IIPt = a4 +3∑

j=1

b4jFDIt−j

3∑

j=1

c4jExportt−j

+3∑

j=1

d4j IIPt−j +3∑

j=1

e4j importt−j + u4t

(13)

The result of the causality MWALD test obtained from the SURestimation of the level VAR model and reported in Table 6,suggests bi-directional causality between exports and IIP. This

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 26: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

256 THE INTERNATIONAL TRADE JOURNAL

Table 6Test for Granger-causality Applying the Toda-Yamamoto MWALD

Test

Null Hypothesis Direction of Causality χ2 P-value

FDI does not Grangercause exports.

Exports ⇒ FDI 2.93 0.40

Exports do notGranger cause FDI.

FDI ⇒ Exports 8.29∗ 0.04

FDI does not Grangercause IIP.

FDI ⇒ IIP 17.48∗ 0.00

IIP does not Grangercause FDI.

IIP ⇒ FDI 17.21∗ 0.00

Exports do notGranger cause IIP.

Export ⇒ IIP 0.24 0.97

IIP do not Grangercause exports.

IIP ⇒ Exports 5.58 0.13

Imports do notGranger cause IIP.

Import ⇒ IIP 6.40∗ 0.09

IIP does not Grangercause imports.

IIP ⇒ Imports 6.09 0.10

Imports do notGranger cause FDI .

Import ⇒ FDI 9.68∗ 0.021

FDI does not Grangercause imports.

FDI ⇒ Imports 14.69∗ 0.00

Notes: The underlying model for the two equation system is a SUR model; thelag order is 3 (based on SBC criterion) for all the variables.∗denotes significant at 5% confidence level.

indicates the presence of a feedback between these two variables.Similarly, the assumption of non-causality from IIP to FDI andFDI to IIP is rejected at the 5% level, indicating bi-directionalcausality between industrial production and FDI. So the resultsuggests the presence of feedback between economic growthand FDI.

However, we could not find any significant impact of FDIon exports though the result suggests export stimulate inflowsof FDI to the country. This implies that contrary to the gen-eral euphoria, causality runs from exports to FDI, but not from

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 27: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 257

FDI to exports. On the one hand, FDI may replace exportswhen the motivation for the investment is market-seeking. Onthe other hand, FDI might promote exports when the motiva-tion for foreign investment is factor-seeking. This may supportthe hypothesis that FDI in India has remained domestic market-seeking.

Moreover, our result also suggests that there is a complemen-tarity between imports and FDI. We reject the null of Grangernon-causality from FDI to imports as well as from imports toFDI at a 5% level significance. So as FDI has increased, theimport content has intensified. Generally, imports are demandedmainly to provide inputs for domestic and export products, andto incorporate technology to foster economic development.

To validate the causality results, we also conduct Vector ErrorCorrection (VEC) analysis as all variables are I (1), and areco-integrated to each other. The results of this analysis are some-what in agreement with the results of causality and suggest FDIhas a small role in India in determining output or trade growth(see Table 7). To the contrary, output growth and trade per-formance have a significant role in determining the level of FDIflows in the country. The sign of the coefficient on exports isconstantly negative, which may indicate that FDI in India isdomestic market-seeking rather than export-seeking.

After having established the causality between FDI andtrade, we utilize Geweke’s (1982) decomposition method to findthe extent of linear dependence between FDI and trade. Thelag length of the estimation is selected on the basis of AkaikeInformation criterion. The result of decomposition of causalitybetween trade and FDI is presented in Table 8. It is clear thatmost of the linear feedback between export sand FDI (86%) canbe accounted for by Granger-causality from FDI flows to exports(29%) and from exports to FDI (57%). The remaining 14% isdue to instantaneous causality. As far as imports are concerned,

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 28: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Tab

le7

Tes

tsof

Lin

ear

VE

CA

naly

sis

Indep

enden

tV

ari

able

s�

FD

IIn

dep

enden

tV

ari

able

s�

IIP

Indep

enden

tV

ari

able

s�

Export

EC

M−0

.970

38E

CM

−0.3

6966

∗∗E

CM

0.02

121

(−0.

6815

)(−

4.52

89)

(0.0

731)

�E

xport

(−1)

−1.8

3048

∗∗�

Export

(−2)

0.11

596∗

∗�

FD

I(−

1)

0.04

402(1

.611

48)

(−1.

9731

)(2

.218

2)0.

0440

2(1

.611

48)

�E

xport

(−3)

−1.2

5139

∗�

Export

(−3)

0.11

063∗

∗�

Import

0.25

582∗

(−1.

7677

)(2

.417

61)

(−2)

(1.7

9210

)�

IIP

(−3)

4.25

9907

∗∗0.

0016

2∗∗

(2.3

7929

)�

FD

I(−

3)

(1.7

452)

�II

P(−

4)

3.23

7522

∗∗−0

.091

16∗∗

(2.1

0581

)�

Import

(−4)

(−2.

2667

)�

Import

(−2)

−1.1

9521

∗(−

1.70

127)

�Im

port

(−4)

−1.2

6549

∗(−

1.80

399)

R2

0.48

0641

R2

0.76

1855

R2

0.31

2262

AR

CH

(4)

0.52

58A

RC

H(4

)0.

7317

AR

CH

(4)

0.23

17

Note

s:T

he

dep

enden

tvari

able

isth

equart

erly

change

inth

evari

able

s.E

CM

isth

eeq

uilib

rium

corr

ecti

on

mec

hanis

mder

ived

from

Johanse

nco

-inte

gra

tion

analy

sis.

AR

CH

isth

eA

RC

H-L

Mte

stfo

rre

sidualhet

erosk

edast

icity,

ass

um

ing

one

and

four

lags

(pvalu

ein

pare

nth

eses

).A

range

ofdia

gnost

ics

test

ofth

em

odel

,i.e.

,st

andard

test

sfo

rnorm

ality

(Jarq

ue-

Ber

a),

het

erosk

edast

icity

(Whit

e),and

seri

alco

rrel

ati

on

(Q-st

ati

stic

,LM

)are

conduct

edand

they

show

that

the

resi

duals

are

wel

lbeh

aved

.t

valu

ein

pare

nth

eses

.∗∗

and

∗den

ote

signifi

cant

5and

10%

level

,re

spec

tivel

y.O

nly

stati

stic

ally

signifi

cant

(5and

10%

level

s)are

report

edin

the

table

.

258

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 29: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 259

Table 8Geweke’s Decomposition of Causality

Percent of Overall Linear Feedback

From FDI to Export 29From Export to FDI 57Simultaneous Feedback 14From FDI to Import 34From Import FDI 41Simultaneous Feedback 25

Note: For the decomposition of causality, we use the method-ology developed by Geweke (1982).

of the total dependence, 34% is due to causation flowing fromFDI to imports, 26% due to causation flowing from imports toFDI, and 40% is due to instantaneous causality.

VI. CONCLUSION AND POLICY SUGGESTIONS

This article has two objectives: first, to review the exist-ing literature of FDI, trade, and economic growth; second, totest causality among FDI, trade, and economic growth. Weobserve that the theoretical linkages among these variables arewell established; however, empirical literature on this issue iswidely inconsistent and scare, especially for developing coun-tries. As these three variables are, at least theoretically, related,instead of studying two variables separately at a time, we pre-fer to examine multivariate causalities among these variables. Aco-integration test substantiated the long-run relations amongthese variables, while our empirical findings based on the Toda-Yamamoto causality test seem to suggest that FDI has favorablegrowth effects. However, we find that causality is running fromexports to FDI and not vice versa. So the role of FDI in exportpromotion could not be substantiated. However, there exists acomplementary relation between FDI and imports. Our results

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 30: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

260 THE INTERNATIONAL TRADE JOURNAL

also indicate that FDI in India is domestic market-seeking ratherthan export-seeking.

A mutual dependence estimate shows that most of the lin-ear feedback between exports and FDI (86%) can be accountedfor by Granger-causality from FDI flows to exports (29%) andfrom exports to FDI (57%). The remaining 14% is due to instan-taneous causality. As far as imports are concerned, of the totaldependence 34% is due to causation flowing from FDI to imports,26% due to causation flowing from imports to FDI, and 40% isdue to instantaneous causality.

The above findings have important policy implications fordeveloping countries like India. Understanding the direction ofcausality between growth and FDI is crucial for formulatingpolicies that encourage private investors in developing coun-tries. Given the capital shortage coupled with cheap labor,FDI could be encouraged to invest in export-oriented indus-tries. This might help India to achieve export competitiveness atleast in some sectors where foreign investors have technologicaladvantages.

ACKNOWLEDGMENTS

We thank an anonymous referee for his or her useful com-ments and helpful suggestions on the previous version of thisarticle. Our thanks also to N. R. Bhanumurthy and ProfessorArup Mitra for their useful comments on an earlier draft of thisarticle.

REFERENCES

Aggarwal, A. 2001. Liberalization, Multinational Enterprises andExport Performance: Evidence from Indian Manufacturing.Working Paper no. 69. Indian Council for Research onInternational Economic Relations (ICRIER), New Delhi,India.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 31: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 261

Agrawal, P. 2005. Foreign Direct Investment in South Asia:Impact on Economic Growth and Local Investment. InE. M. Graham, ed., Multinationals and Foreign Investmentin Economic Development , pp. 94–118. Basingstoke, UK:Palgrave Macmillan.

Aizenman, J., and Noy, I. 2006. FDI and Trade—Two-wayLinkages? The Quarterly Review of Economics and Finance46: 317–337.

Alfaro, L., Chanda, A., Kalemli-Ozcan, S., and Sayek, S. 2003.FDI and Economic Growth: The Role of Local FinancialMarkets. Journal of International Economics 64: 89–112.

Asian Development Bank (ADB). 2004. Asian DevelopmentOutlook 2004. Part 3: Foreign Direct Investment inDeveloping Asia. Manila: ADB.

Balasubramanyam, V. N., Salisu, M., and Sapsford, D. 1996.Foreign Direct Investment and Growth in EP and ISCountries. Economic Journal 106: 92–105.

Banga, R. 2003. The Export-diversifying Impact of Japaneseand U.S. Foreign Direct Investments in Indian ManufacturingSector, Working Paper no. 110. New Delhi, India: ICRIER.

Barry, F., and Bradley, J. 1997. FDI and Trade: The Irish Host-country Experience. Economic Journal 107: 1789–1811.

Basu, P., Chakraborty, C., and Reagle, D. 2003. Liberalization,FDI, and Growth in Developing Countries: A Panel Co-integration Approach. Economic Inquiry 41: 510–516.

Beckerman, W. 1965. Demand, Exports and Growth. In W.Beckerman and Associates, eds., The Britain Economy in1975 , pp. 44–72. Cambridge, UK: The National Instituteof Economics and Social Research, Series 23, CambridgeUniversity Press.

Bhat, K. S., Tripura Sundari, C. U., and Raj, K. D. 2004. CausalNexus between Foreign Direct Investment and EconomicGrowth in India. Indian Journal of Economics 85(337):171–185.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 32: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

262 THE INTERNATIONAL TRADE JOURNAL

Blake, A., and Pain, N. 1994. Investigating Structural Changein UK Export Performance: The Role of Innovation andDirect Investment . National Institute of Economic and SocialResearch (NIESR) Discussion Paper No. 71. London: NIESR.

Blomstrom, M., Lipsey, R. E., and Zejan, M. 1992. WhatExplains Developing Country Growth. National Bureau ofEconomic Research (NBER) Working Paper Series no. 4132.Cambridge, MA: NBER.

Borensztein, E., De Gregorio, J., and Lee, J.-W. 1998. How DoesForeign Direct Investment Affect Economic Growth? Journalof International Economics 45(1): 115–135.

Brainard, S. L. 1993. A Simple Theory of MultinationalCorporations and Trade with a Trade-off between Proximityand Concentration. National Bureau of Economic Research(NBER) Working Paper no. 4269. Cambridge, MA: NBER.

Brems, H. 1970. Growth Model of International DirectInvestment. American Economic Review 60(3): 320–331.

Cabral, S. 1995. Comparative Export Behaviour of Foreign andDomestic Firms in Portugal. Banco de Portugal EconomicBulletin March: 69–78.

Calderon, C., and Liu, L. 2003. The Direction of Causalitybetween Financial Development and Economic Growth.Journal of Development Economics 72: 321–334.

Caves, R. 1996. Multinational Enterprises and EconomicAnalysis, 2nd ed. Cambridge, MA: Cambridge UniversityPress.

Chakraborty, C., and Basu, P. 2002. Foreign Direct Investmentand Growth in India: A Co-integration Approach. AppliedEconomics 34: 1061–1073.

Chong, A., and Calderon, C. 2000. Causality and Feedbackbetween Institutional Measures and Economic Growth.Economics and Politics 12(1): 69–81.

De Mello, L. R., Jr. 1997. Foreign Direct Investment inDeveloping Countries and Growth: A Selective Survey. TheJournal of Development Studies 34(1): 1–34.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 33: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 263

De Mello, L. R., Jr. 1999. Foreign Direct Investment-led Growth:Evidence from Time Series and Panel Data. Oxford EconomicPapers 51(1): 133–151.

Dua, P., and Rashid, A. I. 1998. FDI and Economic Activity inIndia. Indian Economic Review 33(2): 153–168.

Durham, J. B. 2003. Foreign Portfolio Investment, ForeignBank Lending, and Economic Growth. International FinanceDiscussion Papers 757. Washington, DC: Board of Governorsof the Federal Reserve System.

Durham, J. B. 2004. Absorptive Capacity and the Effects ofForeign Direct Investment and Equity Foreign PortfolioInvestment on Economic Growth. European EconomicReview 48: 285–306.

Edison, H. J., Levine, R., Ricci, L., and Slok, T. 2002. Inter-national Financial Integration and Economic Growth.Journal of International Money and Finance 21:749–776.

Enderwick, P. 2005. Attracting “desirable” FDI: Theory andEvidence. Transnational Corporations 14(2): 93–119.

Fan, E. X. 2002. Technological Spillovers from Foreign DirectInvestment—A Survey . Working Paper Series No. 33,Economics and Research Department (ERD). Manila,Philippines: Asian Development Bank.

Geweke, J. 1982. Measurement of Linear Dependence andFeedback between Time Series. Journal of the AmericanStatistical Association 77(378): 304–324.

Gregorio, J. 1992. Economic Growth in Latin America. Journalof Development Economics 39(3): 59–84.

Grossman, G., and Helpman, E. 1991. Innovation and Growth inthe Global Economy . Cambridge, MA: MIT Press.

Hanson, G. 2001. Should Countries Promote Foreign DirectInvestment? G-24 Discussion Paper Series No. 9. Geneva:United Nations Conference on Trade and Development.

Helpman, E. 1984. A Simple Theory of International Tradewith Multinational Corporations. The Journal of PoliticalEconomy 92(3): 451–471.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 34: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

264 THE INTERNATIONAL TRADE JOURNAL

Helpman, E., and Krugman, P. R. 1985. Market Structure andForeign Trade, Increasing Returns, Imperfect Competitionand the International Economy . Cambridge, MA: MIT Press.

Horstmann, I. J., and Markusen, J. R. 1992. Endogenous MarketStructures in International Trade. Journal of InternationalEconomic 32: 109–129.

Johansen, S. 1991. Estimation and Hypothesis Testing of Co-integrated Vectors in Gaussian Vector Autoregressive Models.Econometrics 59: 1551–1580.

Johansen, S., and Juselius, K. 1990. Maximum Likelihoodand Inference on Co-integration: With Applications to theDemand for Money. Oxford Bulletin of Economics andStatistic 52: 169–210.

Kumar, N. 1994. Multinational Enterprises and IndustrialOrganization: The Case of India. New Delhi: Sage.

Lall, S., and Mohammad, S. 1983. Foreign Ownership and ExportPerformance in the Large Corporate Sector in India. Journalof Development Studies 20: 56–67.

Lim, E.-G. 2001. Determinants of, and the Relation between,Foreign Direct Investment and Growth: A Summary of theRecent Literature. IMF Working Paper No. WP/01/175.Washington, DC: IMF.

Markusen, J. R. 1983. Factor Movements and Commodity Tradeas Complements. Journal of International Economics 14:341–356.

Markusen, J. R. 1995. The Boundaries of MultinationalEnterprises and the Theory of International Trade. Journalof Economic Perspectives 9: 169–189.

Markusen, J. R., and Venable, A. J. 1998. Multinational Firmsand New Trade Theory. Journal of International Economics46: 183–203.

Mazumdar, T. 2005. Capital Flows into India: Implications forits Economic Growth. Economic Political Weekly May 21:2183–2189.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 35: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

Dash and Sharma: FDI, Trade, . . . 265

Mundell, R. A. 1957. International Trade and Factor Mobility.The American Economic Review 47(3): 321–335.

Oliveira, L. H. 2001. Foreign Direct Investment and Export-Led Growth: A Challenge to Brazil . Working paper. Schoolof Business and Public Management. Washington, DC: TheGeorge Washington University.

Pailwar, V. 2001. Foreign Direct Investment Flows to India andExport Competitiveness. Productivity 42: 115–122.

Pain, N., and Wakelin, K. 1998. Foreign Direct Investment andExport Performance in Europe. In C. Milner, R. Read, andS. Thompson, eds., New Horizons in International Trade andIndustry , pp. 54–76. London: MacMillan Press.

Pradhan, J. P. 2002. Foreign Direct Investment and EconomicGrowth in India: A Production Function Analysis. IndianJournal of Economics 82(327): 582–586.

Rambaldi, A. N., and Doran, H. E. 1996. Testing forGranger Non-Causality in Co-integrated Systems MadeEasy . Working Papers in Econometrics and AppliedStatistics 88. Biddeford, ME: Department of Econometrics,The University of New England.

Romer, P. M. 1994. Idea Gaps and Objects Gaps in EconomicDevelopment. Journal of Monetary Economy 99: 543–573.

Sahoo, D., and Mathiyazhagan, M. K. 2003. Economic Growthin India: Does Foreign Direct Investment Inflow Matter?Singapore Economic Review 48(2): 151–171.

Sharma, K. 2000. Export Growth in India: Has FDI Played aRole? Center Discussion Paper no. 816. New Haven, CT:Economic Growth Center, Yale University.

Summers, L. H. 2000. International Financial Crises: Causes,Prevention and Cures. American Economic Review Papersand Proceedings 90(2): 1–16.

Sun, H. 2001. Foreign Direct Investment and Regional ExportPerformance in China. Journal of Regional Science 41(2):317–336.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13

Page 36: FDI, Trade, and Growth Dynamics: New Evidence from the Post-Reform India

266 THE INTERNATIONAL TRADE JOURNAL

Toda, H. Y., and Yamamoto, T. 1995. Statistical Inference inVector Autoregressions with Possibly Integrated Processes.Journal of Econometrics 66: 225–250.

United Nations Conference on Trade and Development(UNCTAD). 2000. World Investment Report 1999 . Geneva:UNCTAD.

United Nations Conference on Trade and Development. 2003.World Investment Report 2002 . Geneva: UNCTAD.

United Nations Conference on Trade and Development. 2006.World Investment Report 2005 . Geneva: UNCTAD.

Virmani, A. 2004. Accelerating Growth and Poverty Reduction—A Policy Framework for India’s Development . New Delhi:Academic Foundation.

Vuksic, G. 2006. Foreign Direct Investment and ExportPerformance of the Transition Countries in Central andEastern Europe. Paper presented at the 12th DubrovnikEconomic Conference, organized by the Croatian NationalBank, June 28–July 1, Dubrovnik, Croatia.

Zellner, A. 1982. Measurement of Linear Dependence andFeedback between Multiple Time Series: Comment. Journalof the American Statistical Association 79: 304–324.

Zhang, K. H. 2005. How Does FDI Affect a Host Country’sExport Performance? The Case of China. Paper presentedto International Conference of WTO, China, and the AsianEconomies, III, Xi’an, China, June 25–26.

Dow

nloa

ded

by [

Uni

vers

ity o

f Il

linoi

s at

Urb

ana-

Cha

mpa

ign]

at 0

6:26

30

Sept

embe

r 20

13