Equity-based entry modes of emerging country multinationals: Lessons from Turkey

Equity-based entry modes of emerging country multinationals:Lessons from Turkey

Mehmet Demirbag a,*, Ekrem Tatoglu b, Keith W. Glaister a

a Management School, The University of Sheffield, 9 Mappin Street, Sheffield S1 4DT, United Kingdomb Bahcesehir University, Faculty of Economics and Administrative Sciences, Besiktas, Istanbul 34349, Turkey

Journal of World Business 44 (2009) 445–462

A R T I C L E I N F O

Keywords:

Institutional theory

FDI

Emerging country MNEs

Joint ventures

Wholly owned subsidiaries

Turkey

A B S T R A C T

Based on a sample of 522 foreign affiliates of Turkish multinational enterprises (MNEs)

with varying levels of Turkish equity ownership, this study provides an empirical analysis

of the determinants of equity-based entry mode strategies in host country markets. A

number of hypotheses are developed to examine the impact of institutional, transaction

specific and firm level variables on Turkish MNEs’ choice of equity ownership mode in

their foreign affiliates. The results reveal that institutional variables are important in

explaining the equity composition of foreign affiliates of Turkish MNEs. Particularly

important in determining equity ownership mode were found to be political constraints,

linguistic distance, knowledge infrastructure and the extent of parent diversity. Results

concerning the influences of the size of the affiliate are contrary to expectations and

contradict the findings of previous research. No support was found for the impact of

cultural distance on the equity ownership mode of Turkish MNEs in their foreign affiliates.

Apart from political constraints, equity ownership choice and its underlying determinants

do not vary between emerging and developed host country markets.

� 2008 Elsevier Inc. All rights reserved.

Contents lists available at ScienceDirect

Journal of World Business

journal homepage: www.e lsev ier .com/ locate / jwb

1. Introduction

A particular aspect of global competition is the increasingparticipation of firms from emerging economies. Althoughdeveloped country multinational enterprises (MNEs)account for the majority of global foreign direct investment(FDI), MNEs from emerging countries have shown anunprecedented increase both in numbers and the value ofoutward FDI. According to World Investment Report 2006, theamount of outward FDI from emerging and transitioneconomies reached $304 billion in 2007, representing about15.2% of worldwide outward flows—the highest level everrecorded. The value of the total stock of outward FDI fromdeveloping and transitional economies was estimated to be$2.5 trillion as of 2007, accounting for nearly 16.4% of the

* Corresponding author. Tel.: +44 114 222 3441; fax: +44 114 222 3348.

E-mail address: [email protected] (M. Demirbag).

1090-9516/$ – see front matter � 2008 Elsevier Inc. All rights reserved.

doi:10.1016/j.jwb.2008.11.009

world total (UNCTAD, 2008). While a small number ofsource economies are responsible for a large share of theseFDI outflows, companies from an increasing number ofcountries are seeking investment opportunities abroad todefend or build a competitive position. This surge has comemainly from the rapid pace of economic development andliberal market policies governments have implemented inthese countries.

There is a dearth of empirical research on the increasinginvolvement of emerging country firms in the globaleconomy and their impact on global competition. Whilethere is a well established literature on internationaliza-tion and market entry mode choice of developed countryMNEs (see Brouthers & Hennart, 2007 for a detailedreview), we will argue that these theoretical perspectivesmay not explain the behavior of emerging country firms inthe global market place (Child & Rodrigues, 2005; Khanna& Palepu, 2006; Luo & Tung, 2007; Mathews, 2006; Yiu,Lau, & Bruton, 2007).

Fig. 1. Trends in outward Turkish FDI firms over time. Source: General

Directorate of Banking and Exchange (GDBE).

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462446

Some researchers further argue that the developmentand internationalization paths of emerging country MNEsare different than those of developed country MNEs(Bonaglia, Goldstein, & Mathews, 2007; Chittoor & Ray,2007; Cuervo-Cazurra, 2007; Khanna & Palepu, 2006; Luo& Tung, 2007; Mathews, 2006). In the past developingcountry MNEs, so-called third world MNEs, targeted otherdeveloping countries and had more resource orientedactivities (Heenan & Keegan, 1979; Kumar, 1982; Kumar &McLeod, 1981; Lall, 1983; Lecraw, 1977, 1993; Wells,1983). Recent trends indicate that emerging country MNEstarget their operations toward both developed anddeveloping markets in both resource intensive and highervalue adding activities (Luo & Tung, 2007; OECD, 2007).Research on MNEs from emerging countries has so farfocused on governance (Filatotchev, Strange, Piesse, & Lien,2007), determinants of Chinese outward FDI (Buckley et al.,2007; Yiu et al., 2007) and FDI strategies of emergingcountry MNEs in other emerging countries (Chen, 1998;Kumar, 1982; Lecraw, 1977; Lee & Beamish, 1995; Raju &Prahalad, 1980; Yeung, 1998, 1999).

Although some empirical studies have examinedemerging country MNEs’ strategies (Bonaglia et al.,2007; Buckley et al., 2007; Filatotchev et al., 2007; Yiuet al., 2007), entry mode choice of these companies relatedto developed and other emerging and developing countrieshas yet to be studied (Brouthers & Hennart, 2007). AsBusiness Week reports, ‘‘they are shaking up entireindustries, from farm equipment and refrigerators toaircraft and telecom services, and changing rules of globalcompetition’’ (Business Week, 2006, p. 42). Given theincreasing number of MNEs from emerging marketeconomies entering other emerging and developed mar-kets, it becomes imperative to investigate their entry andequity ownership strategies.

This study aims to investigate factors affecting entrymode choice of Turkish MNEs. Given the emerging natureof the market and the transitional characteristics of theinstitutional environment (Cavusgil, Ghauri, & Agarwall,2002), Turkish outward FDI activity provides a good case totest a number of new dimensions alongside previouslytested variables on the determinants of equity-basedforeign market entry strategies of emerging country MNEs.Drawing on official data from the Turkish Treasury (the keyagency for Turkish outward FDI operations), this studyseeks to further our understanding of the determinants offoreign equity ownership strategies of MNEs from emer-ging countries.

2. Country background

According to UNCTAD (2006) the leading emergingcountry sources of FDI are China and Singapore, but otheremerging countries, including India, Malaysia, Mexico,South Korea and Turkey, are also becoming importantplayers in outward FDI. Among this group, Turkeypossesses important features, such as its geographicallocation, and associated cultural and linguistic proximitiesto Central Asian, European and Middle Eastern markets.The Turkish context provides an interesting researchsetting, characterized by attempts to become a more

Western style market economy and ongoing membershipnegotiations with the EU. Turkey is the first Muslimcountry to bid for EU membership.

Turkey has attracted a significant number of FDI entriesin recent years. Since the start of the accession negotiationsbetween Turkey and the EU in December 2004, total FDIentries have increased dramatically. FDI inflows between2001 and 2004 were US$ 9.72 bn, with the total value of FDIinflows jumping to a record level of US$ 20 bn as of 2007. Asimilar trend has also been observed in Turkey’s outwardFDI. Fig. 1 shows the formation of all outward FDI firmsover the 1990–2005 period. There were nearly 150outward FDI operations up to 1990. Since then sharpincreases have occurred in both the number and the valueof outward FDI activity. Drawing on the official statistics ofthe Turkish Treasury, as of July 2006 the number ofoutward FDI operations reached 2397 with the value ofcumulative outward FDI totaling nearly $8.6 bn. This surgein outward FDI has been caused by both economic andpolitical factors (Erdilek, 2003). Opening up of newmarkets in the EU, the USA, the Balkans, the former SovietUnion (FSU) countries including the Russian Federation,the newly independent Turkic Republics of Central Asia,and the Middle Eastern and North African countries haveencouraged Turkish investors to engage in FDI operationsin these countries. Domestic political and economic crisescoupled with intensifying competition spurred by Turkey’saccession to a customs union with the EU in 1996 are someof the push factors behind Turkey’s increasing outwardFDI.

Fig. 2 shows the distribution of Turkish FDI by hostcountry and regions. From the total value of $8.6 bnoutward FDI, EU countries including both developed andemerging EU countries accounted for nearly 57% of alloutward FDI. Of the EU countries, five countries includingGermany, the UK, France, the Netherlands and Italy werethe main destinations for FDI. These countries are amongthe leading trading partners of Turkey and are also themain sources of inward FDI to Turkey (Demirbag, Tatoglu,& Glaister, 2007; Tatoglu & Glaister, 1996; Tatoglu,Glaister, & Erdal, 2003). They have well-establishedpolitical and economic links particularly through the EU.The FSU countries constitute the second most attractivehost market, accounting for nearly 34% of all Turkish FDIoutflows. With the fall of communism in the FSU, a huge

Fig. 2. Distribution of Turkish FDI by host country/region (As of July 2006).

Source: General Directorate of Banking and Exchange (GDBE).

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462 447

market characterized by a relatively large and consump-tion-prone population has been opened to Turkishinvestors. The newly independent Turkish speakingCentral Asian countries have become the major recipientsof Turkish FDI due to their geographic and culturalproximity to Turkey (Demirbag, Gunes, & Mirza, 1998).

The sectoral distribution of Turkish outward FDI is asfollows: manufacturing (55%), including energy (28%),primary manufacturing industries (21%), telecom (4%) andconstruction (2%); services (45%), including banking andfinancial services (32%), trade (12%), and tourism andtransport (1%).

3. Literature review and hypotheses

Theoretical perspectives that explain the level andpattern of FDI or MNE activity range from mainstreameconomic theories (Caves, 1971; Hymer, 1976; Kindle-berger, 1969), internalization models (Buckley & Casson,1976) to Dunning’s eclectic paradigm (Dunning, 1988,2006). The most prominent are internalization theory andthe eclectic paradigm. These perspectives, however, arelargely based on the experiences of developed countryMNEs. The international business literature providesexplanations for internationalization and entry modedecisions of MNEs from developed countries. There is nosingle theory that can be used to explain outward FDI fromemerging country economies (Buckley et al., 2007; Luo &Tung, 2007; Yiu et al., 2007). Earlier work on developingcountry MNEs or third world MNEs initially used anadapted transaction cost explanation of internationaliza-tion (Kumar, 1982; Lee & Beamish, 1995). More recentlyinformed observers have noted that the emerging trend ismuch more complex and therefore requires insights from anumber of perspectives (Bonaglia et al., 2007; Luo & Tung,2007; Mathews, 2006).

The ownership/location/internalization (OLI) perspec-tive developed by Dunning has been extended to examineinternationalization of Chinese family enterprises (Erdener& Shapiro, 2005), Taiwanese and Singaporean firms (Sim &Pandian, 2003), and Korean firms (Lee & Slater, 2007).Dunning’s eclectic paradigm, as its name suggests, sets out

a holistic approach to explain the level and pattern ofinternational production (Dunning, 1988, 2006). Dunningcombines several factors that offer a greater explanation ofMNE or FDI activity in open markets than any singleapproach does (Grosse & Behrman, 1992, p. 117).Recognizing the importance of both structural andtransaction cost imperfections for MNE activity, Dunningadds ‘ownership advantages’ to the location and inter-nalization advantages previously suggested by internali-zation theory. Dunning’s approach consists of an attemptto analyze the ‘who’, ‘where’ and ‘why’ of FDI activity interms of ownership, location and internalization advan-tages (OLI) with each group of factors acting interdepen-dently (Dunning, 2006). Ownership advantages are thosethat are specific to a particular firm and that enable it totake advantage of investment opportunities abroad.Locational advantages are those advantages specific to acountry that dictate the choice of production site.Internalization advantages determine whether foreignproduction will be organised through markets (licensing)or hierarchies (FDI). The eclectic paradigm also posits thatthe pattern of international economic involvement or FDIactivity should vary by three contextual variables: countryof origin, industry, and firm-specific circumstances (Dun-ning, 2006).

While Dunning’s paradigm provides an appropriateframework to examine the FDI activity of emergingcountry MNEs, it has also been challenged for itsassumption that emerging country MNEs should possessownership advantages in order to operate in a foreigncountry (Luo & Tung, 2007; Mathews, 2006; Yiu et al.,2007). These MNEs may not necessarily have ownershipadvantages that can be exploited in the developed countrymarkets. The asset exploitation motive (which is based onownership advantages) is related more to market entry forother emerging and developing country markets, whereasfor developed country markets, emerging country MNEsmay not have such motives. Instead MNEs may be drivenby asset seeking motives (to develop ownership advan-tages) (Luo & Tung, 2007; Makino, Lau, & Yeh, 2002;Mathews, 2006; Liu, Buck, & Shu, 2005; Yiu et al., 2007).Both asset seeking and asset exploitation motives aredistinct but complementary motives that can be observedin the internationalization process of emerging countryMNEs. In the asset exploitation perspective, FDI is viewedas the transfer of a firm’s proprietary assets across borders.In the asset seeking perspective, FDI is perceived as ameans to acquire strategic assets (e.g., technology,marketing and management expertise, R&D capability,etc.) (Makino et al., 2002).

The technical aspects of the international market entrydecision are important. Equally important, however, is theinstitutional environment of both the home country of theemerging country MNE (Witt & Lewin, 2007) and the hostcountry market where a new venture is planned (Buckleyet al., 2007; Meyer & Nguyen, 2005). There is a growingbody of literature concerning the institution-based view ofstrategy (Hoskisson, Eden, Lau, & Wright, 2000; Wright,Filatotchev, Hoskisson, & Peng, 2005). Instead of focusingon the technical environments of individual transactionsas suggested by transaction cost theory, institutional

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462448

theory requires an investigation of different broaderinstitutional contexts across countries and their impacton ownership strategies of MNEs (Meyer, 2001). Moststudies of emerging market firms blend institutional theorywith one of the other three perspectives (Wright et al., 2005;Martinez & Dacin, 1999). Similarly, Dunning (2006) arguesthat MNEs in strategic asset seeking FDI do not seek to gainaccess to foreign resources, capabilities or market oppor-tunities only, but also to country specific institutions.‘‘Particularly, this is likely to be the case where the businessand social culture in the home and host countries aremarkedly different’’ (Dunning, 2006, pp. 203–204).

In emerging market economies, institutions and insti-tutional factors are particularly important because institu-tional immaturity raises transaction costs and risk level.These have a significant impact on entry mode choice(Child, Chung, & Davies, 2003; Meyer, 2001, 2004; Peng,2001). While some studies have examined entry modechoice of developed country MNEs by integrating institu-tional and OLI perspectives (Brouthers, 2002; Brouthers &Brouthers, 2000; Delios & Beamish, 1999) this study is thefirst attempt to integrate both perspectives from anemerging country MNE’s perspective. These two viewsshould not be considered mutually exclusive; rather theyshould be viewed as complementary. Recently, Dunning(2006) has attempted to extend the OLI framework byconsidering arguments from institutional theory. Further-more, ‘‘asset exploitation’’ and asset ‘‘augmentationmotives’’ may co-exist as internationalization drivers ofemerging country MNEs as they compete in mature andother emerging and developing economies. As Yiu et al.(2007, p. 522) argue, the driving force behind emergingcountry MNEs’ internationalization is not an ‘‘either-or’’situation of asset exploitation or asset augmentation, butrather how the concepts can be integrated in order toprovide the greater insight.

The transaction costs theory stresses that assetspecificity, behavioral and environmental uncertaintiescreate market transaction costs and organizational controlcosts. Within transaction cost theory, there are two mainviews on managing entry modes (Brouthers & Hennart,2007). The first view is articulated by a number of authors(Anderson & Gatignon, 1986; Erramilli & Rao, 1993; Hill,Hwang, & Kim, 1990; Woodcock, Beamish, & Makino,1994) who arrange arm’s length contracts (i.e., licensingand franchising) and hierarchical modes (i.e., JVs andWOSs) along a continuum of increasing control, risk, andresource commitment. This view of transaction coststheory is based on organizational control costs with WOSproviding the highest control and risk, and requiring themaximum commitment. A second perspective developedby Hennart (1991) within transaction costs theory does notview JVs as a step on the continuum between marketcontracts and equity ownership modes, with both JVs andWOSs in the equity category. Brouthers and Hennart(2007) argue that from a transaction costs perspective, theonly difference between WOSs and JVs is that the latterconsist of joint internalization whereas WOSs representsole internalization. This argument places a greateremphasis on market transaction costs in contrast to thosewho emphasize organizational control costs (Anderson &

Gatignon, 1986; Erramilli & Rao, 1993; Hill et al., 1990;Woodcock et al., 1994). While there are studies indicatingthat MNEs may use a number of non-ownership controlmechanisms (Woodcock et al., 1994), in the case ofemerging country MNEs, the applicability of such mechan-isms may be limited. Therefore, organizational controlcosts are more important as these new players have not yetbeen able to develop non-ownership control mechanisms(Luo & Tung, 2007). Non-ownership control mechanismsrequire a greater knowledge of the international market,experience and resources. They are slower, more costly andless efficient than ownership control. Compared todeveloped country MNEs, ownership based organizationalcontrol becomes particularly important for emergingmarket MNEs. These firms are generally weaker in termsof international experience and organizational familiaritywith the complex and risky host country environments.Based on this logic and confirming evidence, we adhere tothe perspective of organizational control theory.

Entry mode studies based on organizational controltheory have been developing a number of taxonomies ofentry mode structures ranging from WOSs to a smallshareholding in a venture (Anderson & Gatignon, 1986).Drawing on the premises of organizational control theory, itis argued that the same variables affect the choice betweenarm’s length contracts and equity based entry modes. Thisline of explanation of entry mode within transaction costtheory places a great emphasis on organizational controlcosts. Transaction costs are associated with the entry modeselected. A partner’s equity ownership in a JV is treated as anindicator of organizational control level (Woodcock et al.,1994). Different levels of equity ownership are posited toprovide specific control capabilities and capacities(Brouthers & Hennart, 2007; Woodcock et al., 1994). Froman organizational control theory perspective, each entrymode creates a different level of control costs that may berequired for ongoing management of the relationship. Forinstance, in a multi-partner JV there will be costs associatedwith the initial control relationship between parents(Woodcock et al., 1994). Similarly, there will be costsemanating from ongoing management of operations. Whenentry modes are viewed from an organizational controlperspective, a WOS option appears to be intrinsicallydistinct from a JV option (Woodcock et al., 1994).

Prior studies on entry mode choice of developed countryMNEs have identified a variety of factors, including bothfirm-related and host country-related, that may influencethe benefits and costs associated with alternative establish-ment modes. Some of the firm-related factors include thesize of the parent firm (Caves & Mehra, 1986; Kogut & Singh,1988), multinational experience of the parent firm (Hennart& Reddy, 1997; Kogut & Singh, 1988; Mudambi & Mudambi,2002), research and development intensity of the parentfirm (Brouthers & Brouthers, 2000; Kogut & Singh, 1988;Padmanabhan & Cho, 1995, 1999), relative size of theinvestment (Brouthers & Brouthers, 2000; Caves & Mehra,1986; Hennart & Park, 1993; Padmanabhan & Cho, 1995),degree of product diversity (Barkema & Vermeulen, 1998;Caves & Mehra, 1986), and foreign experience (Barkema &Vermeulen, 1998; Padmanabhan & Cho, 1999; Wilson,1980). Typical examples of the host country-related factors

Fig. 3. Conceptual framework for an emerging country MNE’s subsidiary

equity ownership mode.

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462 449

are the level of development and the size of the host countrymarket (Brouthers & Brouthers, 2000; Mudambi &Mudambi, 2002; Padmanabhan & Cho, 1995).

In the case of emerging country MNEs, we argue thatthese firms engage in FDI in developed and emergingcountries for different motives (Luo & Tung, 2007). Theylargely seek strategic assets when entering developedcountries but pursue market opportunities when operatingin emerging countries. Previous literature on FDI alsosuggests that there are significant differences betweendeveloped and emerging countries in terms of theinvestment environment and institutional and competi-tive factors that are likely to affect firms’ strategic choicesof international entry modes (Makino, Isobe, & Chan, 2004;Makino, Beamish, & Zhao, 2004).

The following subsections detail the rationale for thehypothesized effects of the institution and transactionspecific influences and firm level variables on emergingcountry MNEs’ ownership strategies of their foreignaffiliates. The proposed relationships between predictorvariables of transaction costs, institutional, firm andindustry specific variables and entry mode decision ofemerging country MNEs, are delineated in the conceptualframework shown in Fig. 3. It is argued that emergingcountry MNEs generally lack international experience andmay face control difficulties in risky and complex hostcountry environments. To compensate for such deficien-cies, emerging country MNEs may have a tendency tochoose entry modes that enable them to control theiroperating exposure to host country uncertainties andrisks.1 Our conceptual framework also allows us toexamine how the determinants of ownership choice byemerging country MNEs vary between developed andemerging host country markets.

3.1. Institutional and transaction specific influences

3.1.1. Political constraints

Institutional voids constitute an important factor inentry mode decisions. Political hazards (e.g., political

1 We thank an anonymous reviewer for highlighting this point.

instability, unpredictable regulatory changes, governmentinterference, ambiguous laws and regulations, etc.) con-stitute an important dimension of institutional environ-ment in a host country. An institutionally more efficientenvironment provides better opportunities for emergingcountry MNEs to exploit and upgrade their competitiveadvantages (Luo & Tung, 2007). As asset-augmentinginvestments are expected to go to developed marketeconomies, it is logical to expect lower political risk and anefficient institutional environment in such markets. In alow political risk and efficient institutional environment(e.g., Canada, Germany, UK, USA) it is also expected thatemerging country MNEs will service markets by directlyowned local production. Both the OLI paradigm andinstitutional theory posit that in countries experiencinghigh political risk, firms with asset exploitation motiveswill tend to opt for arm’s length servicing modes or lowerequity participation (Buckley et al., 2007; Henisz, 2000).

Empirical evidence supports the contention that thehigher the political risk the lower will be equity ownershipof MNEs. Miller (1992, 1993) and Kobrin (1976) empha-sized the importance of the dimension of political risk anduncertainties on MNEs’ operations and performance.Brouthers (2002) found a relationship between politicaluncertainties and MNEs’ entry mode decision and perfor-mance. Ahmed, Mohamad, Tan, and Johnson (2002) alsonoted that when perceived political risk was high, MNEstend to use low commitment entry modes. Yiu and Makino(2002) reported that in restrictive regulatory environ-ments, MNEs choose a joint venture over a wholly ownedsubsidiary.

Time is also another factor to consider. Prior researchsuggests that foreign investors view political stability froma longer-term perspective, rather than recently establishedstability (Root & Ahmed, 1979, p. 758; Murtha & Lenway,1994; Henisz, 2000; Bevan, Estrin, & Meyer, 2004). Thepolitical constraints index developed by Henisz (2000,2001, 2004) takes the time dimension into account andrecognizes that the regulatory restrictions and politicalconstraints change over time. Thus, we hypothesize that:

H1a. In an environment with high political constraints,emerging country MNEs will select a JV compared to awholly owned subsidiary.

H1b. In an environment with high political constraints,emerging country MNEs are more likely to choose a min-ority equity share rather than a 50% or a majority share JV.

3.1.2. Cultural distance

Following the work of Hofstede (1980), the concept ofcultural distance is well known and widely used forpredicting the entry mode of MNEs (Kogut & Singh, 1988).It is argued that cultural differences will increase the costof information flow and should be interpreted as atransaction cost factor. Bhardwaj, Dietz, and Beamish(2007) suggest that firm preferences for a particularnational culture, in aggregate, are manifested in the valueof total FDI inflows that a host country receives. Bothinstitutional theory and the OLI paradigm treat culturalproximity/cultural distance as an important determinant

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462450

of market entry strategy. Traditionally, emerging countryMNEs have entered markets where there is a large diasporacommunity, which helps minimize transaction costs(Buckley et al., 2007). Recent evidence, however, indicatesthat emerging country MNEs also enter markets wherethere are no immigrant/diaspora communities (Luo &Tung, 2007). For instance, firms from India, China, Mexicoand Turkey operate in markets where there appears to be alow presence of their diaspora communities and therefore,less dependence on ethnic ties (Luo & Tung, 2007). Luo andTung (2007) also argue that these latecomers to the globalmarket use acquisitions as an entry mode in order to beable to secure tacit knowledge from culturally distantmarkets. This line of argument appears to contradict theentry mode literature, which is largely based on theexperiences of developed country MNEs (Brouthers &Hennart, 2007). Considering asset augmentation motivesof emerging country MNEs, the learning argument maystill apply. Tacit knowledge may be secured by acquisitionsor wholly owned green field investments (Luo & Tung,2007) thus alleviating the impact of cultural distance onequity ownership of affiliates.

Studies on entry mode strategies of developed countryMNEs have found some support for the view that a highcultural distance results in a preference for a full ormajority ownership of foreign affiliates (Padmanabhan &Cho, 1999; Pan, 1996), while other empirical studies havesuggested that it leads to the adoption of JVs (Agarwal,1994; Erramilli, 1991; Erramilli & Rao, 1993; Hennart &Larimo, 1998; Kogut & Singh, 1988). This national culturalparadox has been well articulated in a recent study byBrouthers and Brouthers (2001). We take the view thatwhen emerging country MNEs enter host countries whichare culturally distant from their home countries, theywould have a particular need to learn how to do business inthe host country market, thus encouraging them to sharethe equity of the venture with a local partner. Based on thisreasoning and relying on the arguments of conventionalMNE theories we expect that:

H2a. An emerging country MNE is more likely to choose aJV over a WOS when facing greater cultural distancebetween home and host countries.

H2b. An emerging MNE is more likely to choose a minorityJV over a majority JV when facing greater cultural distancebetween home and host countries.

3.1.3. Linguistic distance

Cultural differences or distance measures have beenidentified as a key variable in entry mode decisions ofMNEs. However, with the exception of relatively fewstudies (Demirbag, Glaister, & Tatoglu, 2007; Dow &Karunaratna, 2006; West & Graham, 2004) a linguisticdistance measure has not been operationalized in entrymode research. Particularly for emerging country MNEs,due to the regional orientations of their operations, alinguistic distance measure may complement that ofcultural distance (Demirbag, Glaister, et al., 2007). It isargued that linguistic proximities or distance can be moredominant in business transactions (Dow & Karunaratna,

2006). Observed differences in managerial values couldalso be attributed to linguistic differences betweencountries (West & Graham, 2004). Chen, Sokal, and Ruhlen(1995) clustered genetic and linguistic relationshipsamongst 130 populations. It may be argued that familiaritywith the host country market is associated with suchclustering, in that firms tend to remain within theirlanguage group during the initial phase of internationalexpansion (Welch, Welch, & Marschan-Piekkari, 2001).Similarly, Nordstrom and Vahlne (1994) point out thatlinguistic and psychic proximity is instrumental forlearning and understanding the foreign environment.Bilateral affinity is known to be a determinant of bilateraltrade, and linguistic proximities tend to promote interna-tional trade (Disdier & Mayer, 2006). This logic can beextended to operations of MNEs.

Gomes-Casseres (1990) used this argument and found astrong relationship between environmental familiarityand joint venture likelihood. The logic behind thisargument is that familiarity with the host countryenvironment is likely to result in a better workingrelationship with local partners. Sohn (1994) examinedsocial knowledge of MNEs. Sohn argued that socialknowledge might supplement the MNEs control abilitythrough social means of control, which would reduce theneed to obtain an equity position to gain control (p. 314). InChina, Beamish (1993) found that most joint venturesoriginated from linguistically/ethnically related countries.Firms from Taiwan or other ethnically related countriespreferred substantially lower equity positions than theirWestern counterparts. Welch et al. (2001) argue that thereis a tendency for firms to engage in equity ventureformations in those countries with the same languagegroup at the initial stage of international expansion. Weextend this argument to the ownership structure choice ofMNEs. The greater the linguistic distance between homeand host countries will influence the entry mode decisionby influencing the risk perceptions of managers. Thereforewe hypothesize that:

H3a. The greater the linguistic distance between homeand host country, the more likely that an emerging countryMNE chooses a JV over a WOS.

H3b. The greater the linguistic distance between homeand host country, the more likely that an emerging countryMNE chooses a minority equity share rather than a 50% or amajority share JV.

3.1.4. Knowledge infrastructure

The national innovation system (NIS) of a host countrymay also influence equity ownership strategies of foreigninvestors from emerging market economies. An NISinvolves a set of institutions and organizations that areresponsible for the generation, diffusion and adoption ofnew technologies (Frederiksson & Liang, 2006). Differentactors, such as foreign firms, local firms, local and nationalgovernments, research institutes and universities are allcomponents of an NIS. In the face of high competition,retaining all technical activities at home may not besufficient for a firm to have a sustainable competitive

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462 451

position. As competitive pressure increases, firms will seekto gain access and integrate their global innovativecompetences through subsidiary-specific advantages indispersed countries. The existence of relative weaknessesand bottlenecks in the home country may force local firmsto invest in other locations that may strengthen emergingfirms’ innovative capacity. Resources required for technicaland development activities, mainly well-trained researchand development personnel, research institutes, anddesign experts, might be more available in a differentNIS (Frederiksson & Liang, 2006; Kumar, 2001). Newcommunication technologies and speed of data transmis-sions enable firms to locate parts of their activities beyondtheir immediate region. Further, competitive pressure mayforce firms to invest in locations that can offer comple-mentary or better conditions for certain activities (Freder-iksson & Liang, 2006).

Firms from emerging market economies will try tostrengthen ownership advantages particularly in countrieswith a high knowledge infrastructure. This infrastructureis essential for the continued existence of the firm (Tsang &Yip, 2007). Emerging country MNEs are latecomers in theglobal market place, and these firms may overcome thesedisadvantages by acquiring critical assets from matureMNEs to compensate for their competitive weaknesses(Bonaglia et al., 2007; Cuervo-Cazurra, 2007; Luo & Tung,2007). Such operations exist in mature markets, which alsohave a high knowledge infrastructure (Tsang & Yip, 2007).Availability of skilled engineers or scientists in acquiredoperations can be treated as elements of an assetaugmenting strategy (access to specific incorporatedknowledge), which is also associated with the OLIparadigm. Luo and Tung (2007) also argue that emergingcountry MNEs ‘‘are often not path dependent in selectingtheir entry modes and project location’’ (Luo & Tung, 2007,p. 482). Drawing largely on the arguments of the spring-board view of international expansion of emerging countryMNEs we developed the following two-part hypothesis:

H4a. The higher the knowledge infrastructure of a hostcountry, the more likely that an emerging country MNEwill choose a WOS over a JV.

H4b. The higher the knowledge infrastructure of a hostcountry, the more likely that an emerging country MNEwill choose a majority or a 50% equity share over a minorityshare JV.

3.2. Firm specific influences

3.2.1. Parent firm diversity

An investor entering a foreign market to manufacture aproduct that is outside of its main line of business is likelyto find that the necessary product-specific knowledge(technology, production know-how) is held or possessedby local firms. Such knowledge is difficult to replicate oracquire by contract. This knowledge is, therefore, mostefficiently obtained through a JV (Hennart, 1991). Anumber of studies have shown that when firms diversifyinto product areas outside of their core businesses, they aremore likely to share the ownership of their foreign

affiliates (Cleeve, 1997; Hennart, 1991; Stopford & Wells,1972). In contrast, other studies (Gomes-Casseres, 1989;Hennart & Larimo, 1998; Padmanabhan & Cho, 1999)found that the relatedness of an affiliate’s products to thoseof the foreign parent had no effect on the probability ofsetting up a JV.

Emerging country MNEs are mostly business groups.Some of these groups are widely diversified (Cuervo-Cazurra, 2007; Khanna & Palepu, 2006). ‘‘The challengethese widely diversified firms face in the internationaliza-tion process is that the firm lacks complementaryresources needed to operate at the larger scale requiredby expansion in the new country’’ (Cuervo-Cazurra,Maloney, & Manrakhan, 2007, p. 713). A new expansionoutside the main operations of emerging country MNEswould require an alliance with a partner to overcome theliability of expansion, or the lack of complementaryresources needed for a diversification (Cuervo-Cazurraet al., 2007). On the basis of this discussion and someconfirming evidence, we expect that emerging countryMNEs who diversify are more likely to choose a JV over aWOS, and also to favor a lower level of equity ownershipover a 50% or majority ownership of their JV. This leads tothe following hypothesis:

H5a. An emerging country MNE is more likely to choose aJV over a WOS when diversifying into areas outside of itsmain line of business.

H5b. An emerging country MNE is more likely to choose aminority share rather than a 50% or a majority share in a JVwhen diversifying into areas outside of its main line ofbusiness.

3.2.2. Capital size of subsidiary

The stages model of internationalization, based onbehavioral theory, assumes low resource commitmentsand lower control modes at the initial stage of internationalmarket entry (Johanson & Vahlne, 1977). The establishmentof a high resource/high control operation entails substantialinternal organization and bureaucratic costs, includinginvestment in administrative, legal, and operating infra-structures (Davidson & McFetridge, 1985). Transaction costtheorists argue that high switching costs stemming from thehigh overhead of such large-scale investments may alsodiminish the firm’s ability to shift to another operation(Erramilli & Rao, 1993). Prior to embarking on a large-scaleresource commitment, foreign investors must carefullyconsider the potential costs and returns of such investment,as the level of risk exposure is directly associated with thetotal investment in non-redeployable assets. Prior researchon developed country MNEs suggests that foreign investorsare more likely to choose lower equity ownership than amajority or full ownership of their subsidiaries when thecapital size of the operation is high (Gatignon & Anderson,1988; Pan, 1996; Shan, 1991). Empirical evidence, however,is equivocal. The findings of Hu and Chen (1993), Zhao andZhu (1998) and Mutinelli and Piscitello (1998) indicate thatforeign investors are likely to choose a full ownership or ahigher equity share in their affiliates as the size ofinvestment increases.

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462452

Transaction cost/internalization arguments are partlybased on developed country MNEs’ experience, whereasemerging country MNEs, as latecomers to global marketsappear to display a different pattern with respect to theirequity ownership and size of investment. These latecomersneed to accelerate their pace of internationalization tocatch up with that of their competitors. Luo and Tung(2007) argue that emerging country MNEs rapidly expandinternationally through high risk and high control equitymodes, such as acquisitions and greenfield investments.Particularly for asset seeking investments, acquisitions areused as vehicles to facilitate securing brands and gainingdirect and quick access to technology. While this argumentpartly contradicts the Uppsala model of internationaliza-tion (Johanson & Vahlne, 1977), the springboard approachassumes that the progressive nature of learning may stillhold true for emerging country MNEs, since their resourcecommitment, especially investment size, is not necessarilya function of time, experience or learning (Luo & Tung,2007). Luo and Tung (2007, p. 491) argue that ‘‘instead,their initial commitment tends to be large (owing to theuse of acquisitions or greenfield investment) and does notnecessarily involve many small steps (owing to strategicleapfrog from their latecomer position)’’. Based on mixedevidence from developed country MNEs and argumentsdeveloped through more recent theoretical frameworks(i.e., the springboard view of international expansion) weexpect that:

H6a. The greater the capital size of the affiliate, the morelikely that an emerging country MNE chooses a WOS over aJV.

H6b. The greater the capital size of the affiliate, the morelikely that an emerging country MNE chooses a majority ora 50% JV rather than a minority JV.

4. Research methods

4.1. Data

All Turkish foreign equity ventures are recorded by agovernment agency, the General Directorate of Banking andExchange (GDBE) within the Undersecretariat of Treasury.The database of GDBE consists of all Turkish outward FDIfirms established since 1990 and as of July 2006 includes2397 outward FDI firms. The GDBE database also providesinformation about the name of the host country, the sectorof operation, the proportion of Turkish equity shareholding,the amount of investment and the entry date.

The study first focuses on the equity-based entry modechoice that Turkish MNEs make between full ownership(setting up a wholly owned greenfield subsidiary orengaging in a full acquisition), and sharing ownershipwith a local firm (establishing a greenfield joint venture ormaking a partial acquisition). Next, we examine the choiceof equity level if the ownership structure is a JV. Assuggested by previous research (Contractor, 1990; Franko,1989; Kaynak, Demirbag, & Tatoglu, 2007) the equity sharedistribution is not exactly on a percentage continuum. Forinstance, there are very different strategic implications

between holding 23% and 24% equity ownership, andholding 49% and 50%; or between holding 50% and 51%equity ownership, and holding 49% and 50% (Pan, 1996).Hence, three separate comparisons for JVs are made: (1)comparing a minority foreign share with an equal foreignshare; (2) comparing a minority foreign share with amajority foreign share; (3) comparing an equal foreignshare with a majority foreign share. This enhances therobustness of the analysis by providing a better under-standing of the variables that impact the ownershipstrategies of Turkish MNEs in their foreign affiliates.

From the original list of 2397 outward FDI firms in thedatabase, a new dataset was compiled based on the capitalvalue of the subsidiary and the proportion of Turkishequity shareholding. Those ventures with capital value ofless than $500,000 were excluded. Most of these firms areowned by a single person or established by means ofordinary partnerships. For the purposes of this survey, itwas not considered feasible to include these firms in thesampling frame. This study also uses the 10% and 90% cut-off points to capture the alternative ownership structures.The investments with foreign ownership of less than 10%are considered to be portfolio investments and areexcluded from the database. A venture is defined as a JVwhen foreign equity ownership ranges from 10% to 90%,while a venture with foreign equity shareholding of over90% is considered to be a WOS. This range is consistentwith the definition of a JV used by the U.S. Department ofCommerce. Park and Ungson (1997) and Hladik (1985) alsofollowed the same definitions. In this study JVs are furtherclassified into three categories: minority foreign-owned(10–49%), co-ownership (50–50%), and majority foreign-owned (51–90%). Based on these criteria, the new datasetincluded a list of 522 Turkish FDI firms that constitutes thesample of this study.

The average foreign equity ownership is 69.3%. Thedistribution of ownership is as follows: minority foreign-owned, 28.2%; co-ownership, 10.2%; majority foreign-owned, 14.6%; and full ownership 47.1%. The average ageof affiliate is 8.33 years. In terms of the host country/regionorigin, the data set includes 202 subsidiaries in EUcountries, 48 in the USA, 119 in FSU countries, 67 inMiddle Eastern and African countries, and the remainder inother developed or emerging countries. The distribution ofthe data set in terms of the sector of operation is as follows:auto, transport and related equipment, 4.0%; electrical,electronics and durables, 6.5%; food and beverages, 9.6%;chemicals, 4.4%; textiles, apparel and leather, 8.2%; metal,iron and steel, 2.9%; mining, petroleum and gas, 4.2%; othermanufacturing, 5.2%; export-import trading, 14.8%; tour-ism, 3.1%; banking and financial services, 21.6%, construc-tion and logistics, 12.3% and other services, 5.0%.

The characteristics of the sample firms on the basis ofthe key dimensions of the data source are summarized inTable 1.

4.2. Operationalization of variables

4.2.1. Dependent variable

The equity structure of foreign subsidiaries (minorityJV, equal JV, majority JV and WOS) was treated as the

Table 1

Characteristics of the sample.

No. %

Foreign equity shareholding

Minority JV 147 28.2

Co-ownership JV 53 10.2

Majority JV 76 14.6

WOS 246 47.0

Broad sector of operation

Auto, transport and related equipment 12 2.3

Electrical, electronics and durables 34 6.5

Food and beverage 50 9.6

Chemicals 23 4.4

Textiles, apparel and leather 43 8.2

Metal, iron and steel 15 2.9

Mining, petroleum and gas 22 4.2

Other manufacturing 27 5.2

Export-import trading 77 14.8

Tourism 16 3.1

Banking and financial services 113 21.6

Construction and logistics 64 12.2

Other services 26 5.0

Time period of entry

1980–1995 143 27.4

1996 to present 379 72.6

Host country/region

EU countries 202 38.7

Non-EU European countries 17 3.3

USA 48 9.2

Other developed countries 3 0.6

Emerging EU countries 16 3.1

Emerging non-EU countries 44 8.4

FSU countries 119 22.8

Middle Eastern and African countries 67 12.8

Far Eastern countries 6 1.1

Size

C1: $500,000–$2,000,000 245 46.9

C2: $2,000,001–$10,000,000 174 33.3

C3: Greater than $10,000,000 103 19.8

Total 522 100.0

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462 453

dependent variable. We used the 10% and 90% cut-offpoints to capture the alternative ownership structures. The5% and 95% cut-off points were also tested, but nosignificant differences in the results were detectedbetween these and the results of the 10% and 90% cut-off points.

4.2.2. Independent variables

The political constraints index (POLCON) developed byHenisz (2000, 2001) measures the feasibility of a change inpolicy given the structure of a nation’s political institutions(the number of veto points) and the preference of actorsthat inhabit them (the partisan alignments of various vetopoints and the heterogeneity or homogeneity of thepreferences within each branch). The POLCON databasecovers almost all major nations and is calculated forvirtually all countries for the post-war period (1960–2004), which fits well with the database used for thisstudy. Scores range from 0, indicating a dictatorship, to 1,indicating democracy. The POLCON index represents thedegree to which checks and balances are present in acountry’s political system (Neumayer & Spees, 2005). As

the value of the index approaches 0, then an increase isexpected in the level of political constraints for a given hostcountry market.

Cultural distance (CULTDIST) is measured by using themethodology developed by Kogut and Singh (1988) basedon Hofstede’s (1980) measures of four dimensions ofnational culture: power distance, uncertainty avoidance,masculinity/femininity and individualism. Countries withsmall values of cultural distance are culturally similar toTurkey, with larger values signifying increasing dissim-ilarity.

There seems to be no developed measure of linguisticdistance (LINGDIST) between world languages in theinternational business or management literature.Researchers have mainly used dummy variables for acommon language (Arora & Fosfuri, 2000; Davidson &McFetridge, 1985; Srivastava & Green, 1986). Recentlythere has been an attempt to operationalize the concept toexplore the relationship between cultural distance dimen-sions and distance between languages (Dow & Karunar-atna, 2006; West & Graham, 2004). Chen et al. (1995)developed a measure of linguistic distances between 130localities around the world. This study follows the samepath in operationalizing the linguistic distance by adoptingthe dendrogram developed by Chen et al. (1995). Theyconstructed this dendrogram by using the UPMGA methodbased on unequal branch length linguistic distances. Thedendrogram’s tree includes 130 localities around theworld, which covers all languages used by investors inthe database used for this study. The raw data and thematrix developed by using the raw data were kindlyprovided by these authors for this study. For multilingualcountries, as suggested by West and Graham (2004),weighted averages based on percentages and mixes oflanguages were calculated and used. The World Wide Webversion of Grimes (1992) catalog of human languages wasused to identify the language or languages of a particularcountry. Turkish was used as the focal language incalculating linguistic distance between countries.

An appropriate measure for host country knowledgeinfrastructure (KNOWLEDGE) was compiled from datapublished in multiple editions of World Economic Forum’sGlobal Competitiveness Report (WEF, 2003; 2004; 2005).Global Competitiveness Report is an annual multidimen-sional analysis of how far national environments areconducive or detrimental to the global competitiveness ofcompanies operating in the countries involved. Based onthree year average figures, we have created a compositeindex of five knowledge indicators for each host countrymarket in our data set. Knowledge indicators used for thecomposite index are ‘availability of scientists and engi-neers’, ‘technological readiness’, ‘quality of scientificresearch institutions’, ‘firm level technology absorption’and ‘quality of mathematics and science education’.

Parent diversity (DIVER) was measured using Rumelt’s(1974) categories, i.e. single business, dominant business,related business, and unrelated business. An ordinalvariable was created that takes the value from 1 to 4 torepresent each category respectively.

Capital size of affiliate (LNSIZE) is measured using thelogarithm of the amount of total investment in US dollars.

Table 2

Correlation matrix.

Variable name Definition 1 2 3 4 5 6 7 8 9 10 11 12

1. POLCON Political constraints index 1.00

2. CULTDIST Cultural distance 0.34* 1.00

3. LINGDIST Linguistic distance 0.30* 0.32* 1.00

4. KNOWLEDGE Knowledge infrastructure 0.22 0.33* 0.42** 1.00

5. DIVER Extent of parent diversity 0.01 0.11 0.07 0.12 1.00

6. LNSIZE Logarithm of the capital size 0.06 0.12 0.13 0.06 0.23 1.00

7. DATE Period of entry 0.01 0.12 0.03 �0.06 �0.05 �0.01 1.00

8. GROUP1_IND Group 1 manufacturing industries �0.12 �0.08 �0.09 �0.22 �0.12 0.01 0.07 1.00

9. TERTIARY_IND Tertiary industries 0.16 0.04 0.06 0.21 �0.02 �0.03 �0.10 �0.32* 1.00

10. HOST_COUNTRY Broad host country groups 0.25* 0.33* 0.32* 0.48** 0.12 0.15 0.01 �0.17 0.15 1.00

11. R&D_INT R&D intensity of industry 0.01 0.12 0.08 0.07 0.10 �0.10 �0.01 �0.30* �0.29* 0.11 1.00

12. SUBS_DENS Subsidiary density 0.16 �0.14 �0.29* �0.05 0.02 0.03 �0.11 �0.01 0.04 �0.08 �0.01 1.00

N = 522.* p < 0.01.** p < 0.001 (two-tailed test).

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462454

The logarithmic transformation is generally used tonormalize the size variable, which might otherwise bebadly skewed.

4.2.3. Control variables

This study controls for date of entry, industry, hostcountry, industry R&D, and subsidiary diversity. A dummyvariable for the timing of entry (DATE) was assigned tocontrol for variation that might result from the character-istics of Turkey’s economic and business developmentprogram first initiated in the early 1980s and acceleratedlater in the mid 1990s with the start of the customs unionwith the EU in 1996. A value of 0 was given for outwardTurkish FDI entries during the 1980–1995 period and 1 forpost-1995 entries.

To control for industry variations, three broad industrialgroupings are introduced as dummy variables. Group 1manufacturing industries included food, beverages, textile,apparel, leather, metal, iron, steel, mining, petroleum andgas, Group 2 manufacturing industries comprised of auto,transport and related equipment, electrical, electronics,durables and chemicals. Group 3 industries includedexport-import trading, tourism, banking and financialservices, construction, logistics and other services.

A dummy variable (HOST_COUNTRY) was assignedwith 1 = Developed host country markets and 0 = Emer-ging host country markets. The former group of hostcountries involved developed EU and non-EU countries,the USA and other developed countries, while the lattergroup included emerging EU and non-EU countries, FSUcountries, Middle Eastern and African countries.

In this study, the measure of industry R&D intensity wasadapted from the UK Department of Trade and Industry(DTI) index of R&D intensity of industries (DTI, 2006). TheDTI scoreboard is based on the ratio of R&D expenditures tosales at the four digit Standard Industrial Classification(SIC). R&D intensity of industries varies from 15%(pharmaceuticals) to 1.5% (food production) to 0.5%(forestry and paper).

Subsidiary density in a country was operationalized asthe number of Turkish subsidiaries in the host country forthe year before the entry, i.e. the number of Turkishsubsidiaries at the end of the calendar year before the entry

date. Similar measures have also been used in previousstudies (Miller & Eden, 2006) in order to indicateisomorphic tendencies of firms (Li, Yang, & Yue, 2007;Yiu & Makino, 2002).

5. Results and discussion

The correlation matrix of the independent variables inthe study is shown Table 2. The pairwise correlations donot seem to present serious collinearity problems for themultivariate analysis, as none of the variables havecorrelation coefficients above 0.50. Wetherill (1986)recommends an analysis of VIF when three or morevariables are involved. In the near dependency thecorrelations between relevant pairs of variables need notbe large (Wetherill, 1986). This is where VIF may play animportant role and should not be larger than ten. Since thehighest VIF value for independent variables was signifi-cantly lower than this cut-off point, collinearity in theexplanatory variables for the data set does not seem to be aproblem.

The hypotheses were tested by conducting binomiallogistic regressions on the functional relationshipsbetween the hypothesized effect variables and varioussubsidiary ownership strategies of foreign investors. Theanalysis is conducted in three stages. The first stageincludes an investigation of the Turkish investors’ choicebetween full ownership and shared-equity ownership. Indoing this, four separate comparisons are made: (1)comparing a minority ownership JV to a WOS, (2)comparing an equal ownership JV to a WOS, (3) comparinga majority ownership JV to a WOS, and (4) comparing a fullsample of JVs to a WOS. Four sets of coefficients areestimated and reported in Table 3, where the choicebetween each category of JV ownership and a WOS ispresented in Models 1, 2, 3, and 4. A value of 0 is assignedfor the JV as the base mode including each category of JVand the full sample of JVs.

The estimated coefficients represent the utility ofchoosing a WOS over a JV. A positive coefficient for anindependent variable means that it increases the prob-ability of a WOS compared to a JV. A negative coefficientmeans that a JV is more likely than a WOS. All four sets of

Table 3

Logistic regressions for the ownership pattern of JVs and WOSs.

Variable name Definition Model 1 Model 2 Model 3 Model 4

Minority JV vs.

WOS (WOS = 1)

50% Ownership JV

vs. WOS (WOS = 1)

Majority JV vs.

WOS (WOS = 1)

All JVs vs.

WOS (WOS = 1)

Coefficient Wald

statistics

Coefficient Wald

statistics

Coefficient Wald

statistics

Coefficient Wald

statistics

POLCON Political constraints index 0.56* 1.84 2.43*** 5.09 1.00 1.36 0.83** 4.19

CULTDIST Cultural distance �0.04 0.05 �0.12 0.25 0.05 0.05 �0.01 0.01

LINGDIST Linguistic distance �0.36** 4.14 �0.17 0.29 �0.26* 2.61 �0.32** 4.56

KNOWLEDGE Knowledge infrastructure 0.02 0.18 0.43*** 5.48 0.33** 3.77 0.16** 4.51

DIVER Extent of parent diversity �0.42*** 19.48 �0.21 2.11 �0.10 0.75 �0.29*** 13.01

LNSIZE Logarithm of the capital

size

0.27*** 10.77 0.17 1.75 0.02 0.03 0.19*** 6.99

DATE Period of entry 0.48** 3.57 �0.44 1.15 0.57* 3.31 0.37* 3.03

GROUP1_IND Group 1 manufacturing

industries

�0.18 0.12 �0.17 0.04 �0.55 0.86 �0.44 1.01

TERTIARY_IND Tertiary industries �0.01 0.01 �1.14* 2.76 0.12 0.05 �0.33 0.68

HOST_COUNTRY Broad host country groups �0.07 0.03 �0.75 1.38 �1.32** 3.58 �0.68* 2.94

R&D_INT R&D intensity of industry �0.20 1.72 �0.35* 2.85 �0.33** 3.69 �0.31** 6.16

SUBS_DENS Subsidiary density 0.01*** 8.59 0.01*** 7.54 0.00 1.59 0.01*** 11.64

Intercept �2.63 2.29 �1.92 0.60 0.44 0.04

Model chi-square 52.69*** 32.88*** 20.24** 48.64***

Sensitivity 0.85 0.98 0.99 0.57

Specificity 0.44 0.11 0.07 0.66

Correct ratio 0.70 0.86 0.80 0.63

Prop’al chance criterion 0.52 0.69 0.63 0.50

Cox & Snell R2 0.13 0.10 0.06 0.09

Nagelkerke R2 0.17 0.17 0.09 0.12

N = 522; Minority JV = 147; 50% Ownership JV = 53; Majority JV = 76; JV (all) = 276; WOS = 246.* p < 0.1.** p < 0.05.*** p < 0.01 (two-tailed test).

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462 455

models have high overall explanatory power with sig-nificant chi-square values (p < 0.01). All the models have agood fit with classification rates ranging from 86% to 63% ofthe observations. Rates that are higher than that would beexpected by chance. The specificity (i.e., capacity tocorrectly predict JV categories) of the four models rangesfrom 66% to 7%, while their sensitivity (i.e., capacity tocorrectly predict WOSs) ranges from 99% to 57%. Pseudo R-square measures confirm that the models have adequateexplanatory power.

The second stage includes an analysis of the choicebetween the three levels of JV ownership and involvescomparing (1) minority ownership and equal ownership,(2) minority ownership and majority ownership, and (3)equal ownership and majority ownership. Three sets ofcoefficients are presented in Models 5, 6 and 7 in Table 4.Model 5 shows the results of the binomial logistic modelwhere the minority ownership JV is taken as the base andgiven a zero value. In Model 6, a minority ownership JV isset as the base and assigned a zero value, while in Model 7an equal ownership JV is used as the base and given a zerovalue. All three models have high overall explanatorypower with significant chi-square values (p < 0.01). Thethree models also have satisfactory classification rateswith levels that are well above the proportional chancecriterion. While the sensitivity of the three models leavesroom for improvement (78–23%), their specificity is good(95–51%). Pseudo R-square measures indicate satisfactoryexplanatory power.

The third stage includes a further assessment of howthe underlying determinants of ownership choice ofTurkish MNEs in their affiliates vary between developedand emerging host country markets. In terms of FDImotivation, emerging country MNEs may be largelyseeking strategic assets when entering developed hostcountry markets, while they may be pursuing marketopportunities when serving emerging host country mar-kets. It is commonly accepted that institutional constraintsin host country markets exist in both developed andemerging countries, although the degree or threat of suchconstraints may differ between the two groups. It istherefore reasonable to argue that the distinctions in theinstitutional context of host country markets stemminglargely from the nature of host country market (developedvs. emerging country) is likely to exert some influence onthe determinants of ownership strategies of emergingcountry MNEs. The overall sample is therefore split intotwo sub-samples one composed of developed hostcountries and the other of emerging host countries. Thesample size did not permit an examination of theownership choice between various categories of JVs. Abinomial logistic analysis was undertaken to explain onlythe ownership choice of Turkish investors between the fullsample of JVs and WOSs where the full sample of JVs istaken as the base and given a zero value. Table 5 presentstwo sets of coefficients in Models 8 and 9, which report thefindings for both sub-samples of emerging and developedhost country markets, respectively. Models 8 and 9 have a

Table 4

Logistic regressions for the ownership pattern of JVs.

Variable name Definition Model 5 Model 6 Model 7

Minority JV vs. 50% Owner-

ship JV (50% Ownership = 1)

Minority JV vs. Majority

JV (Majority JV = 1)

50% Ownership JV vs.

Majority JV (Majority JV = 1)

Coefficient Wald statistics Coefficient Wald statistics Coefficient Wald statistics

POLCON Political constraints index 1.93* 3.53 0.79 0.61 3.62*** 5.26

CULTDIST Cultural distance 0.03 0.01 �0.08 0.09 �0.08 0.05

LINGDIST Linguistic distance �0.33 0.67 �0.10 0.11 �0.07 0.02

KNOWLEDGE Knowledge infrastructure 0.55** 3.78 0.52** 3.69 0.10 0.06

DIVER Extent of parent diversity �0.33** 4.88 �0.36*** 7.85 �0.11 0.37

LNSIZE Logarithm of the capital size 0.22 1.99 0.36*** 8.00 0.24* 2.64

DATE Period of entry 0.51 1.42 0.06 0.03 �0.90* 3.23

GROUP1_IND Group 1 manufacturing industries 0.74 0.69 0.54 0.66 0.46 0.23

TERTIARY_IND Tertiary industries 1.71** 4.32 0.23 0.12 �0.82 0.89

HOST_COUNTRY Broad host country groups 0.73 1.15 �0.17 0.08 �1.22 2.04

R&D_INT R&D intensity of industry 0.44 2.68 0.19 0.78 0.03 0.02

SUBS_DENS Subsidiary density �0.01 1.87 0.00 0.34 0.01** 5.45

Intercept �2.21 0.64 �4.27* 3.01 �3.64 1.15

Model chi-square 24.21*** 26.18*** 29.34***

Sensitivity 0.23 0.26 0.78

Specificity 0.95 0.86 0.51

Correct ratio 0.76 0.66 0.67

Prop’al chance criterion 0.59 0.53 0.52

Cox & Snell R square 0.11 0.11 0.20

Nagelkerke R square 0.17 0.15 0.27

N = 276; Minority JV = 147; 50% Ownership JV = 53; Majority JV = 76.* p < 0.1.** p < 0.05.*** p < 0.01 (two-tailed test).

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462456

relatively good fit with classification rates being 62% and63% of the observations, respectively. The specificity ofboth models is 67%, while their sensitivity is 56% and 57%.Pseudo R-square measures confirm that the models haveadequate explanatory power.

As hypothesized by H1a, the coefficient of POLCON ispositive and significant in three models (Model 2 atp < 0.01; Model 4 at p < 0.05; and Model 1 p < 0.1)providing moderate support for the hypothesis. Thissuggests that in a business environment with fewerpolitical constraints, foreign investors tend to chooseWOSs over JVs. However, when the influence of the type ofhost country market is taken into account, there is avariation with regard to the sign of the coefficient onPOLCON. The coefficient is positive and significant for thesub-sample of developed host country markets (Model 9 atp < 0.05), while it is negative and significant for the sub-sample of emerging host country markets (Model 8 atp < 0.1). This finding suggests that in developed hostcountry markets with fewer political constraints Turkishinvestors would be more willing to choose WOSs over JVs,whereas they would be more in favor of adopting JVs thanWOSs in emerging host country markets characterized byfewer political constraints.

There is also some support for H1b. The coefficient ofPOLCON is positive in all three models, but significant inonly two (Model 7 at p < 0.01; Model 5 at p < 0.1),suggesting that the foreign partner is more likely to choosea majority JV over an equal share JV or an equal share JVover a minority JV, when the host country environment ischaracterized by having fewer political constraints.

There is no support for H2a and H2b. The coefficient ofCULTDIST is not significant in any of the nine models,indicating that the Turkish investors’ choice of a particularlevel of equity stake in their foreign affiliates is not relatedto the cultural distance between home and host country.

The coefficient of LINGDIST is negative in Models 1–4and Models 8 and 9, and significant in four (Models 1, 4 and9 at p < 0.05; Model 3 at p < 0.1). This provides moderatesupport for H3a and suggests that the extent of linguisticdistance between home and host country is negativelyrelated to the probability of the venture being whollyowned. Further, the sign of the coefficient on linguisticdistance is the same for both emerging markets anddeveloped host country markets (Models 8 and 9,respectively). The coefficient of LINGDIST is negative inboth models but significant only in Model 9 (p < 0.05) forthe sub-sample of developed host country markets.

No support is found for H3b. The coefficient onLINGDIST, is negative, but is not significant in all threemodels (Models 5–7). These data indicate that the TurkishJV partner’s choice of equity level is not related to whetherthe JV is established in a host country that exhibits greaterlinguistic distance to Turkey.

There is relatively strong support for H4a. The positiveand significant coefficient for KNOWLEDGE (Model 2 atp < 0.01; Models 3, 4 and 8 at p < 0.05), confirms the viewthat Turkish investors would prefer a WOS over a sharedequity investment in host country markets characterizedby a high level of knowledge infrastructure. This choice isparticularly emphasized in emerging host country marketswith a high level of knowledge infrastructure, where

Table 5

Logistic regressions for the ownership pattern of JVs: emerging host country markets and developed host country markets.

Variable name Definition Model 8a Model 9b

All JVs vs. WOS (WOS = 1)

Emerging countries

All JVs vs. WOS (WOS = 1)

Developed countries

Coefficient Wald statistics Coefficient Wald statistics

POLCON Political constraints index �0.32* 2.98 1.99** 5.26

CULTDIST Cultural distance 0.38 1.30 0.01 0.01

LINGDIST Linguistic distance �0.07 1.12 �0.33** 4.87

KNOWLEDGE Knowledge infrastructure 0.38** 4.94 0.05 0.04

DIVER Extent of parent diversity �0.18 2.11 �0.36*** 11.22

LNSIZE Logarithm of the capital size 0.08 0.50 0.24** 6.26

DATE Period of entry 0.76** 5.55 0.11 0.13

GROUP1_IND Group 1 manufacturing industries 0.14 0.05 �1.05 2.51

TERTIARY_IND Tertiary industries �0.05 0.01 �0.70 1.54

R&D_INT R&D intensity of industry �0.26 1.73 �0.37** 4.90

SUBS_DENS Subsidiary density 0.01*** 11.27 0.01* 3.49

Intercept �3.39 2.27 �2.38 0.73

Model chi-square 29.58*** 30.26***

Sensitivity 0.56 0.57

Specificity 0.67 0.67

Correct ratio 0.62 0.63

Prop’al chance criterion 0.49 0.50

Cox & Snell R sq 0.11 0.11

Nagelkerke R sq 0.15 0.14

* p < 0.1.** p < 0.05.*** p < 0.01 (two-tailed test).a Emerging host country markets (N = 252; JV (all) = 131; WOS = 121).b Developed host country markets (N = 270; JV (all) = 145; WOS = 125).

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462 457

Turkish investors are more willing to adopt full ownershipof their subsidiaries (Model 8 at p < 0.05). Similarly, somesupport is also found for H4b. The coefficient of KNOWL-EDGE is positive and significant in two models (Models 5and 6 at p < 0.05).

There is some support for H5a. The coefficient of DIVERis negative in all four models and significant in two models(Models 1 and 4 at p < 0.01). In a similar vein, there is alsosome support for H5b with negative and significantcoefficients in two of the three models (Model 5 atp < 0.05; Model 6 at p < 0.01). No host country effect isfound for DIVER, in that the coefficient of DIVER is negativein both Models 8 and 9 and significant in Model 9(p < 0.01) for the sub-sample of developed host countrymarkets.

The positive and significant coefficients on LNSIZE intwo of the four models (Models 1 and 4 at p < 0.01) showthat Turkish investors are more likely to choose a WOSover a JV as the capital size of the affiliate increases. Thesefindings support H6a. Similar support has also been foundfor H6b with the coefficient on LNSIZE is significant andpositive in two models (Model 6 at p < 0.01; Model 7 atp < 0.1). Turkish firms would prefer a majority JV over aminority JV or an equal ownership JV. For both sub-samples of emerging and developed host country markets,the coefficient of LNSIZE is positive, but significant only forthe developed host country sub-sample (Model 9 atp < 0.05).

Examination of the control variables shows that fourvariables including period of entry (DATE), R&D intensityof industry (R&D_INT), subsidiary density (SUBS_INT) and

broad host country groups (HOST_COUNTRY) have somesignificant coefficients. The coefficient of DATE is positiveand significant in four of the five models (Models 1 and 8 atp < 0.05; Models 3 and 4 at p < 0.1). Turkish investorsincreasingly have been more in favor of establishing WOSsthan JVs in host country markets since the mid-1990s.Though positive, the coefficient is significant only for theemerging host country sub-sample. However, with respectto the choice between an equal ownership JV and amajority JV, Turkish investors have had more interest inhaving an equal ownership rather than a majority own-ership in their foreign affiliates over the post-1995 period(see Model 7 at p < 0.05).

The coefficient of R&D_INT is negative and significantin four of the nine models (Model 1 at p < 0.1; Models 3,4 and 9 at p < 0.05) suggesting that that Turkishinvestors tend to prefer establishing JVs rather thanestablishing WOSs as the R&D intensity of the industryincreases. Considering the two sub-samples (i.e., emer-ging and developed host country markets), the coeffi-cient of R&D_INT is negative in both models (Models 8and 9), though it is significant only in Model 9, indicatingthat Turkish investors have a tendency to choose lowerequity modes over full ownership particularly indeveloped host country markets as the industry R&Dintensity increases.

Regarding the impact of subsidiary density, thecoefficient of SUBS_INT is positive and significant in sixof the nine models (Models 1, 2, 4 and 8 at p < 0.01; Model7 at p < 0.05; and Model 9 at p < 0.1). The impact ofsubsidiary density is much stronger in emerging host

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462458

country markets (p < 0.01) than in developed host countrymarkets (p < 0.1).

In terms of broad host country groups, the coefficient ofHOST_COUNTRY is negative in four models and significantin three models (Model 2 at p < 0.01; Model 3 at p < 0.05;and Model 4 at p < 0.1). These findings suggest thatTurkish investing firms have a greater preference forestablishing JVs in developed host country markets ratherthan emerging host country markets. In contrast, they aremore in favor of having full ownership of their foreignaffiliates in emerging host country markets.

6. Conclusions

This study draws upon institutional and transactioncost theories, as well as more recent arguments positedby the springboard approach to internationalization ofemerging country MNEs, to estimate equity-based own-ership modes of Turkish MNEs, in several host countrymarkets. A database of Turkish MNEs’ subsidiaries invarious host country markets allowed us to test rigor-ously a set of hypotheses based on these theoreticalperspectives.

With respect to the choice between a JV and a WOS, ingeneral, a relatively good level of support has been foundfor the study’s hypotheses. The results show that the maininstitutional, transaction specific and subsidiary leveldeterminants of foreign equity ownership influence theownership mode of Turkish MNEs’ foreign affiliates. Thelevel of political constraints, linguistic distance, and thelevel of knowledge infrastructure in the host countrymarket and parent diversity have the expected impact onthe Turkish MNE’s choice between a JV and a WOS. Nosupport has been found for the impact of cultural distanceon the level of Turkish MNEs’ equity shareholding of theirforeign affiliates. The results concerning the impact ofsubsidiary level influence of parent diversity has theexpected effect on ownership mode strategies of TurkishMNEs. However, the effect of capital size of subsidiary onthe level of Turkish MNEs’ equity ownership of theirforeign affiliates contradicts the findings of previousresearch.

With regard to the choice of equity level of JV, theinstitutional variables of political constraints and knowl-edge infrastructure were found to be relatively strongdeterminants of foreign ownership level. Similarly, somesupport has been found with regard to the impact ofsubsidiary level variables of parent diversity and capitalsize of subsidiary on the level of Turkish MNEs’ owner-ship of their foreign affiliates. No support, however, wasfound for the impact of cultural distance on the TurkishMNEs’ equity ownership level of JVs in host countrymarkets.

Further investigation was undertaken in order toidentify how the determinants of ownership choice ofTurkish MNEs in their affiliates vary between developedand emerging host country markets. Apart from politicalconstraints, no significant variation was noted between thetwo sub-samples of host country markets with regard tothe determinants of equity choice. In general, our resultstend to differentiate better between JVs and WOSs than

between different categories of JVs. This might be partlyexplained by the existence of a hierarchical process inentry mode decision-making (Pan & Tse, 2000). Foreigninvestors have a tendency first to decide whether to chooseeither full ownership or shared ownership in theiraffiliates. Then, they decide the extent of ownership intheir shared equity investments.

Drawing on a relatively large data set involving bothdeveloped and emerging host country markets, thisstudy develops the seemingly scant literature on theforeign equity ownership strategies of MNEs from anemerging country. Secondly, we adopted measures thathave not been employed in most of the prior studiesexamining entry modes or ownership patterns ofMNEs. Amongst these, linguistic distance and politicalconstraints were important dimensions which weoperationalized and tested to explain equity ownershipmodes of MNEs from an emerging country. The level ofknowledge infrastructure is another novel dimensionoperationalized and tested in this paper. Also, analyzingthe host country effect on the underlying determinantsof emerging country MNEs provides a valuablecontribution to the extant literature in a number ofways.

In terms of explaining behaviors of emerging countryMNEs, our empirical results provide support for bothconventional explanations and more recent theoreticalviews such as the springboard perspective. Overall,however, our findings indicate that the recent theoreticallenses emerge as more powerful explanations of entry todeveloped country markets than do the conventional MNEand entry mode theories. Our findings (Models 2–6)provide strong support to the central argument of thespringboard perspective which posits that ‘‘emergingcountry MNEs use outward investment as a springboardto acquire strategic assets needed to compete moreeffectively against global rivals’’ (Luo & Tung, 2007, p.482). The springboard perspective also argues thatemerging country MNEs are neither path dependent norevolutionary in selecting their entry modes. Our findingsindicate that in seeking knowledge (knowledge infra-structure). Turkish MNEs are not path dependent and tendto opt for higher equity ownership (WOS rather than JV, orhigher equity commitment in the case of JVs). Findingsalso indicate that Turkish MNEs, particularly whenentering developed countries where the asset-seekingmotive is the main driver, behave in a radical way in termsof investment size and commitment, which providesfurther support to the springboard explanation of emer-ging country MNEs. Turkish MNEs’ investments indeveloped markets require greater size and level ofcommitment (Model 9), whereas the same level ofsignificance with regard to size of operation is notobserved for investment in emerging countries (Model8). Thus, there appears to be a difference in the size ofoperations depending on whether the investment islocated in emerging or developed countries, which tendsto support the central argument of the springboardperspective.

We do not find support for arguments related to theimpact of cultural distance on equity ownership. In

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462 459

contrast, linguistic distance emerges as an importantfactor in our analyses. In this sense, our study clearlyjustifies the need to revise conventional MNE theories tointegrate not only institutional factors such as culture,but also psychic distance factors such as linguisticproximity/distance between home and host countries ofMNEs.

Our findings (Models 8 and 9) provide partial supportfor Dunning’s investment development path (IDP)approach (Dunning, 2006; Galan, Gonzalez-Benito, &Zuniga-Vincente, 2007). Findings suggest that lack ofpolitical constraints in developed countries have apositive impact (making a WOS more likely) while inemerging countries they have a negative impact (a JV ismore likely) on emerging country MNEs’ ownershipstrategies, as the IDP approach predicts. Models 8 and 9also provide some support for new perspectives such asthe springboard approach. Particularly when emergingcountry MNEs enter other emerging countries, motivatedby knowledge acquisition or asset augmentation, thespringboard perspective predicts a higher level ofcommitment. From the knowledge infrastructure pointof view, findings presented in Model 8 supports thisargument for emerging country MNEs operations in otheremerging countries. Our findings therefore indicate thatboth conventional and new theories are relevant toexplaining ownership modes of MNEs. This should notbe viewed as being contradictory, but rather points to thecomplementary nature of both conventional and newtheories. Conventional theories appear to be morerelevant when explaining entry to emerging countries,while new perspectives provide a better explanation ofentry to developed country markets by emerging countryMNEs.

6.1. Managerial relevance

It is widely recognized that the choice of a particularownership mode for outward FDI is one of the mostimportant decisions that managers of many emergingcountry MNEs have to make. Within the context of bothconventional and new theories explaining the strategicbehavior of emerging country MNEs on equity ownershipmode, it is assumed that several institutional, transactionspecific and firm level variables may have a crucialinfluence on global competitiveness and, hence, long-term survival of numerous emerging country firms.Emerging country MNEs, as late comers to internationalexpansion, attempt to compensate for their competitivedisadvantages especially when investing in developedcountries. They seek superior technology or know-howby having full or majority ownership of their outwardFDI. Unlike developed country MNEs, emerging countryMNEs have benefited immensely from inward FDI athome by using different linkages such as joint venturingor original equipment manufacturing with global playersthat have transmitted technological and managerialskills, leading them to undertake their internationaliza-tion later in some unconventional ways. For instance,Turkish MNEs (such as Sabanci and Koc Groups) havelong used joint ventures and OEM manufacturing as

common initial strategies to mitigate risk and gain accessto superior technological and managerial know-how oftheir foreign partners. Following experiential learningand developing their own R&D in their joint ventures,Turkish MNEs prefer eventual full ownership by buyingout their foreign partners. This has been witnessed notonly in the outward joint ventures but also in the inwardjoint ventures of large Turkish conglomerates such as theSabanci Group, which used its inward joint ventures withDuPont as a springboard for its outward joint ventures,which recently became its wholly owned ventures(Erdilek, 2008). Other examples include full acquisitionof the German company Grundig, and also the acquisi-tions of white goods manufacturers Blomberg, ElektraBregenz, Leisure, Flavel and Arctic by Turkey’s largestconglomerate, Koc Group; Eczacibasi Group’s acquisitionof the highly prestigious Villeroy & Boch tile division; andthe acquisition of the Godiva Chocolatier business for$850 million by Ulker Group, which is a diversifiedTurkish food giant.

Emerging country MNEs use international expansionthrough higher equity modes as a springboard to securepreferential treatment offered by emerging countrygovernments and also to bypass trade barriers intodeveloped country markets. Some Turkish textile andclothing companies invested in Egypt, as a springboard togain preferential financial and non-financial treatment,and Jordan as a gateway to increase exports to the USA. Itshould, however, be borne in mind that it is of the utmostimportance for emerging country MNEs to invest time andmoney in investigating as many factors as possible prior tothe choice of a particular equity ownership in theiroutward investments. In some circumstances, the wrongchoice may threaten the survival of the investing firm.Emerging country investors should ensure that theyconcentrate on the factors that are most relevant to theirown situation.

6.2. Limitations

It is necessary to note the limitations of the paper. First,the focus on outward direct investment from a singleemerging country to numerous host countries has bothstrengths and weaknesses. One strength is that the studyis able to test whether political constraints, linguistic andcultural distances influence the choice of foreign owner-ship structure. However, a weakness is the lack of data onpotentially important home and host country specificmarket and industry conditions. We were unable to obtaindata for variables that have been shown to be significantdeterminants of the entry mode choice, such as entrymethod (acquisition vs. greenfield), industry concentra-tion, and multinational experience, for a sample ofsubsidiaries based in 65 different host countries. Sinceour data is cross-sectional in nature, we were unable toanalyze the impact of institutional changes over time onownership patterns. Also, there has been criticism of thenational cultural distance concept for methodologicalweaknesses and cultural biases (Brouthers & Brouthers,2001). Prior studies relating cultural distance to modechoice have generally used aggregate national cultural

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462460

distance measures, such as Hofstede’s measures.2 How-ever, as suggested by Stottinger and Schlegelmilch (1998)individual perceptions of country distance might bedifferent from aggregate national cultural distances basedon measures such as Hofstede’s scale. It would be usefulfor future studies to investigate whether aggregatenational cultural distance measures and individuallyperceived cultural distance measures have a similar ordifferent impact on entry mode choice (Shenkar, 2001).Shenkar, Luo, and Yeheskel (2008) criticize ‘‘culturaldistance’’ measure by arguing that the distance metaphorby its very nature is a static one and therefore unable tocapture changes which may emerge as a result of culturalencounter between cultural carriers in a FDI context. Theyfurther argue that the ‘‘cultural distance’’ metaphorshould be replaced with another, namely ‘‘culturalfriction’’ which may facilitate a shift of focus from abstractdifferences towards contact between specific entities.Also, in future studies country-level data could beaccompanied by cognitive cultural distance measures(e.g., Sullivan & Bauerschmidt, 1990) as was undertaken ina survey by Boyacigiller (1990), where executives wereasked to rank adjustment difficulties in countries wherethey had served in the past. It may also be noted that ourclassification of geographic location into developed andemerging countries may be too crude. While we examinesome variation among this broad categorization ofcountry groups, it would be useful to further investigatea more fine grained distinction between various countrieswith respect to emergent country MNEs’ entry modechoice.

6.3. Future research

Further research would extend the investigation inseveral directions. First, research analyzing changesover time in ownership patterns would enhance ourunderstanding of the impact of institutional changes onequity compositions of subsidiaries. Second, the linguis-tic distance measure used in this study could bereplicated in other contexts involving MNEs fromvarious emerging countries, which would serve tofurther validate the findings of this study. Third, withineach category of determinants (institutional and firm-level), there may be other determinants (e.g., regulatorydistance, political stability, and diplomatic relationsunder the category of institutional determinants) thatcan be considered in future work. Finally, comparisonstudies of countries with a similar development level tothat of Turkey and diverse culture dimensions wouldprovide greater insight into the role of the institutionalenvironment and organizational factors in ownership

2 It is important to acknowledge the limitation of Hofstede’s national

level cultural scores for the Turkish context. As the original four

dimensions are not updated for Turkey, we were unable to use the fifth

dimension in calculating cultural distance between Turkey and host

countries. Further, despite recent improvements in Hofstede’s old

dataset, his cultural scores may still have some bias in interpreting

corporate level cultural differences between investing parties. We thank

an anonymous reviewer for highlighting this point.

structures of foreign subsidiaries of emerging countryMNEs.

References

Agarwal, S. (1994). Socio-cultural distance and the choice of joint ven-tures: A contingency perspective. Journal of International Marketing,2(2): 63–80.

Ahmed, Z. U., Mohamad, O., Tan, B., & Johnson, J. P. (2002). International riskperceptions and mode of entry: A case study of Malaysian multinationalfirms. Journal of Business Research, 55: 805–813.

Anderson, E., & Gatignon, H. (1986). Modes of foreign entry: A transactioncost analysis and propositions. Journal of International Business Studies,17(Fall): 1–26.

Arora, A., & Fosfuri, A. (2000). Wholly owned subsidiary versus technologylicensing in the worldwide chemical industry. Journal of InternationalBusiness Studies, 31(4): 555–572.

Barkema, H. G., & Vermeulen, F. (1998). International expansion throughstart-up or acquisition: A learning perspective. Academy of ManagementJournal, 41(1): 7–26.

Beamish, P. W. (1993). The characteristics of joint ventures in the People’sRepublic of China. Journal of International Marketing, 1(2): 29–48.

Bevan, A., Estrin, S., & Meyer, K. E. (2004). Foreign investment location andinstitutional development in transition economies. International Busi-ness Review, 13: 43–64.

Bhardwaj, A., Dietz, J., & Beamish, P. W. (2007). Host country culturalinfluences on foreign direct investment. Management InternationalReview, 47(1): 29–50.

Bonaglia, F., Goldstein, A., & Mathews, J. A. (2007). Accelerated internatio-nalization by emerging markets’ multinationals: The case of white goodssector. Journal of World Business, 42(4): 369–383.

Boyacigiller, N. (1990). The role of expatriates in the management of inter-dependence. Journal of International Business Studies, 21(3): 357–381.

Brouthers, K. D. (2002). Institutional, cultural and transaction cost influenceson entry mode choice and performance. Journal of International BusinessStudies, 33(2): 203–221.

Brouthers, K. D., & Brouthers, L. E. (2000). Acquisition or greenfield start-up?Institutional, cultural and transaction cost influences. Strategic Manage-ment Journal, 21: 89–97.

Brouthers, K. D., & Brouthers, L. E. (2001). Explaining the national culturalparadox. Journal of International Business Studies, 32(1): 177–189.

Brouthers, K. D., & Hennart, J. F. (2007). Boundaries of the firm: Insights frominternational entry mode research. Journal of Management, 33(3): 395–425.

Buckley, P. J., & Casson, M. C. (1976). The future of the multinational enterprise.London: Macmillan.

Buckley, P. J., Clegg, L. J., Cross, A. R., Liu, X., Voss, H., & Zheng, P. (2007). Thedeterminants of Chinese outward foreign direct investment. Journal ofInternational Business Studies, 38(4): 499–518.

Business Week (2006). Emerging Giants. Cover Story, July 31.Caves, R. E. (1971). International corporations: The industrial economics of

foreign investment. Economica, 38: 1–27.Caves, R. E., & Mehra, S. K. (1986). In M. E. Porter (Ed.), Entry of foreign

multinationals into U.S. manufacturing industries (pp. 449–481). Compe-tition in Global Industries.

Cavusgil, S. T., Ghauri, P. N., & Agarwall, M. R. (2002). Doing business inemerging markets. Sage.

In T. J. Chen (Ed.), Taiwanese firms in Southeast Asia: Networking acrossbordersEdward Elgar: Cheltenham.

Chen, J., Sokal, R. R., & Ruhlen, M. (1995). Worldwide analysis of genetic andlinguistic relationships of human populations. Human Biology, 67(4):595–612.

Child, J., Chung, L., & Davies, H. (2003). The performance of cross-border unitsin China: A test of natural selection, strategic choice and contingencytheories. Journal of International Business Studies, 34: 242–254.

Child, J., & Rodrigues, S. B. (2005). The internationalization of Chinese firms:A case for theoretical extension? Management and Organization Review,1: 381–418.

Chittoor, R., & Ray, S. (2007). Internationalization paths of Indian pharma-ceutical firms: A strategic group analysis. Journal of International Man-agement, 13(3): 338–355.

Cleeve, E. (1997). The motives for joint ventures: A transaction costs analysisof Japanese MNEs in the UK. Scottish Journal of Political Economy, 44(1):31–43.

Contractor, F. J. (1990). Ownership patterns for U.S. joint ventures abroadand the liberalization of foreign government regulations in the 1980s:Evidence from the benchmark surveys. Journal of International BusinessStudies, 21(1): 55–73.

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462 461

Cuervo-Cazurra, A. (2007). Sequence of value-added activities in the multi-nationalization of developing country firms. Journal of InternationalManagement, 13(3): 258–277.

Cuervo-Cazurra, A., Maloney, M. M., & Manrakhan, S. (2007). Causes of thedifficulties in internationalization. Journal of International Business Stu-dies, 38: 709–725.

Davidson, W. H., & McFetridge, D. G. (1985). Key characteristics in the choiceof international technology transfer mode. Journal of International Busi-ness Studies, 16(Summer): 5–21.

Delios, A., & Beamish, P. (1999). Ownership strategy of Japanese firms:Transactional, institutional and experience influences. Strategic Manage-ment Journal, 20: 711–727.

Demirbag, M., Tatoglu, E., & Glaister, K. W. (2007a). Dimensions of Europeandirect investment activity in Turkey: Patterns and prospects. Interna-tional Journal of Emerging Markets, 2(3): 274–297.

Demirbag, M., Glaister, K. W., & Tatoglu, E. (2007b). Institutional andtransaction cost influences on MNE’s ownership strategies of theiraffiliates: Evidence from an emerging market. Journal of World Business,42(4): 418–434.

Demirbag, M., Gunes, R., & Mirza, H. (1998). Political risk management: Acase study of Turkish companies in Central Asia and Russia. In Mirza(Ed.), Global competitive strategies in the new world economy: Multilater-alizm, regionalization and the transnational firm (pp. 269–283). Cheltnam:Edward Elgar.

Disdier, A. C., & Mayer, T. (2006). Je t’aime, moi non plus: Bilateral opinions andinternational trade. CEPII working paper no 2006-01, Paris.

Dow, D., & Karunaratna, A. (2006). Developing a multidimensional instru-ment to measure psychic distance stimuli. Journal of International Busi-ness Studies, 37(5): 578–602.

DTI. (2006). The R&D Scoreboard: The top 800 UK & 1250 Global Companies byR&D Investment, vol. 1. London: Department of Trade and Industry.

Dunning, J. H. (1988). The eclectic paradigm of international production: Arestatement and some possible extensions. Journal of International Busi-ness Studies, 19(1): 1–31.

Dunning, J. H. (2006). Towards a new paradigm of development: Implica-tions for the determinants of international business. Transnational Cor-porations, 15(1): 173–227.

Erdener, C., & Shapiro, D. M. (2005). The internationalization of Chinesefamily enterprises and Dunning’s eclectic MNE paradigm. Managementand Organization Review, 1(3): 411–436.

Erdilek, A. (2003). A comparative analysis of inward and outward FDI inTurkey. Transnational Corporations, 12(3): 79–105.

Erdilek, A. (2008). Internationalization of Turkish MNEs. Journal of Manage-ment Development, 27(7): 744–760.

Erramilli, M. K. (1991). The experience factor in foreign market entrybehavior of service firms. Journal of International Business Studies,22(3): 479–501.

Erramilli, M. K., & Rao, C. P. (1993). Service firms’ international entry modechoice: A modified transaction-cost analysis approach. Journal of Mar-keting, 57: 19–38.

Filatotchev, I., Strange, R., Piesse, J., & Lien, Y. C. (2007). FDI by firms fromnewly industrialized economies in emerging markets: Corporate gov-ernance, entry mode and location. Journal of International BusinessStudies, 38(4): 556–572.

Franko, L. G. (1989). Use of minority and 50–50 joint ventures by UnitedStates multinationals during the 1970s: The interaction of host countrypolicies and corporate strategies. Journal of International Business Studies,20(1): 19–40.

Frederiksson, T., & Liang, G. (2006). Features of national innovation systemsand determinants of international R&D locations: Conceptual frame-work and empirical exploration. Paper presented at the AIB AnnualConference, 23–26 June, Beijing.

Galan, J. I., Gonzalez-Benito, J., & Zuniga-Vincente, J. A. (2007). Factorsdetermining the location decisions of Spanish MNEs: An analysis basedon the investment development path. Journal of International BusinessStudies, 38: 975–997.

Gatignon, H., & Anderson, E. (1988). The multinational corporation’s degreeof control over foreign subsidiaries: An empirical test of a transactioncost explanation. Journal of Law, Economics, and Organization, 4(2): 305–336.

Global Competitiveness Report (2003). World Economic Forum.Global Competitiveness Report (2004). World Economic Forum.Global Competitiveness Report (2005). World Economic Forum.Gomes-Casseres, B. (1989). Ownership structures of foreign subsidiaries.

Journal of Economic Behavior and Organization, 11: 1–25.Gomes-Casseres, B. (1990). Firm ownership preferences and host govern-

ment restrictions: An integrated approach. Journal of International Busi-ness Studies, 21: 1–22.

Grimes, B. (1992). Ethnologue: Languages of the world (12th ed.). Dallas, TX:Summer Institute of Linguistics.http://www.sil.org/ethnologue.

Grosse, R., & Behrman, J. N. (1992). Theory in international business. Trans-national Corporations, 1(1): 93–126.

Heenan, D. A., & Keegan, W. J. (1979). The rise of the third world multi-nationals. Harvard Business Review, 57(Jan–Feb): 101–109.

Henisz, W. J. (2000). The institutional environment for multinational invest-ment. Journal of Law, Economics and Organization, 16(2): 334–364.

Henisz, W. J. (2001). Uncertainty, imitation and plant location: Japanesemultinational corporations 1960–1996. Administrative Science Quarterly,46: 443–475.

Henisz, W. J. (2004). Political institutions and policy volatility. Economics andPolitics, 16(1): 1–27.

Hennart, J. F. (1991). The transaction costs theory of joint ventures: Anempirical study of Japanese subsidiaries in the United States. Manage-ment Science, 37(4): 483–497.

Hennart, J. F., & Park, Y. R. (1993). Greenfield vs. acquisition: the strategy ofJapanese investors in the United States. Management Science, 39(9):1054–1070.

Hennart, J. F., & Reddy, S. (1997). The choice between mergers/acquisitionsand joint ventures: The case of Japanese investors in the United States.Strategic Management Journal, 18: 1–12.

Hennart, J. F., & Larimo, J. (1998). The impact of culture on the strategy ofmultinational enterprises: Does national origin affect ownership deci-sions. Journal of International Business Studies, 29(3): 515–538.

Hill, C. W., Hwang, P., & Kim, W. C. (1990). An eclectic theory of the choice ofinternational entry mode. Strategic Management Journal, 11: 117–128.

Hladik, K. J. (1985). International joint ventures: An economic analysis of U.S.foreign business partnerships. Lexington MA: Lexington Books.

Hofstede, G. (1980). Culture’s consequences. Beverly Hills, CA: Sage Publica-tions.

Hoskisson, R. E., Eden, L., Lau, C. M., & Wright, M. (2000). Strategy inemerging economies. Academy of Management Journal, 43(3): 249–267.

Hu, M. Y., & Chen, H. Y. (1993). Foreign ownership in Chinese joint ventures:A transaction costs analysis. Journal of Business Research, 26(2): 149–160.

Hymer, S. H. (1976). The international operations of national firms: A study ofdirect foreign investment. Cambridge, MA: MIT Press.

Johanson, J., & Vahlne, J.-E. (1977). The internationalization process of thefirm - a model of knowledge development and increasing foreign marketcommitments. Journal of International Business, 8(1): 23–32.

Kaynak, E., Demirbag, M., & Tatoglu, E. (2007). Determinants of ownershipbased entry mode choice of MNEs: Evidence from Mongolia. Manage-ment International Review, 47(4): 505–530.

Khanna, T., & Palepu, K. G. (2006). Emerging giants: Building world-classcompanies in developing countries. Harvard Business Review, 84: 2–10.

Kindleberger, C. P. (1969). American business abroad. New Haven, CT: YaleUniversity Press.

Kobrin, S. J. (1976). Environmental determinants of foreign manufacturinginvestment: An ex post empirical analysis. Journal of International Busi-ness Studies, Fall/Winter: 29–42.

Kogut, B., & Singh, H. (1988). The effect of national culture on the choice ofentry mode. Journal of International Business Studies, 19(3): 411–432.

Kumar, K., & McLeod, M. (Eds.). (1981). Multinationals from developingcountries. Lexington: D.C. Heath.

Kumar, K. (1982). Third world multinationals: A growing force in interna-tional relations. International Studies Quarterly, 26(3): 397–424.

Kumar, N. (2001). Determinants of location of overseas R&D activity ofmultinational enterprises: The case of US and Japanese corporations.Research Policy, 30: 159–174.

Lall, S. (1983). The new multinationals: The spread of third world enterprises.New York: Wiley.

Lecraw, D. (1977). Direct investment by firms from less developed countries.Oxford Economic Papers, 29: 445–457.

Lecraw, D. (1993). Outward direct investment by Indonesian firms:Motivation and effects. Journal of International Business Studies,24(3): 589–600.

Lee, C., & Beamish, P. W. (1995). The characteristics and performance ofKorean joint ventures in LDCs. Journal of International Business Studies,26(3): 637–654.

Lee, J., & Slater, J. (2007). Dynamic capabilities, entrepreneurial rent-seekingand the investment development path: The case of Samsung. Journal ofInternational Management, 13(3): 241–257.

Li, J., Yang, J. Y., & Yue, D. R. (2007). Identity, community and audience: Howwholly owned foreign owned subsidiaries gain legitimacy in China.Academy of Management Journal, 50(1): 175–190.

Liu, X., Buck, T., & Shu, C. (2005). Chinese economic development, the nextstage: Outward FDI? International Business Review, 14: 97–115.

Luo, Y., & Tung, R. L. (2007). International expansion of emerging marketenterprises: A springboard perspective. Journal of International BusinessStudies, 38(4): 481–498.

Makino, S., Lau, C. M., & Yeh, R. S. (2002). Asset-exploitation versus asset-seeking: Implications for location choice of foreign direct investment

M. Demirbag et al. / Journal of World Business 44 (2009) 445–462462

from newly industrialized economies. Journal of International BusinessStudies, 33(3): 403–421.

Makino, S., Isobe, T., & Chan, C. M. (2004a). Does country matter? StrategicManagement Journal, 25: 1027–1043.

Makino, S., Beamish, P. W., & Zhao, N. B. (2004b). The characteristics andperformance of Japanese FDI in less developed countries. Journal ofWorld Business, 39: 377–392.

Martinez, R. J., & Dacin, T. M. (1999). Efficiency motives and normativeforces: Combining transaction costs and institutional logic. Journal ofManagement, 25(1): 75–96.

Mathews, J. A. (2006). Dragon multinationals: New players in 21st centuryglobalization. Asia Pacific Journal of Management, 23: 5–27.

Meyer, K. E. (2004). Perspectives on multinational enterprises in emergingeconomies. Journal of International Business Studies, 35(4): 259–276.

Meyer, K. E. (2001). Institutions, transaction costs and entry mode choice inEastern Europe. Journal of International Business Studies, 32(2): 357–367.

Meyer, K. E., & Nguyen, H. V. (2005). Foreign investment strategies and sub-national institutions in emerging markets: Evidence from Vietnam.Journal of Management Studies, 42(1): 63–93.

Miller, K. D. (1992). A framework for integrated risk management in inter-national business. Journal of International Business Studies, 23(2): 311–331.

Miller, K. D. (1993). Industry and country effects on managers’ perceptions ofenvironmental uncertainties. Journal of International Business Studies,24(4): 693–714.

Miller, S. R., & Eden, L. (2006). Local density and foreign subsidiary perfor-mance. Academy of Management Journal, 49(2): 341–355.

Mudambi, R., & Mudambi, S. M. (2002). Diversification and market entrychoices in the context of foreign direct investment. International BusinessReview, 11: 35–55.

Murtha, T. P., & Lenway, S. A. (1994). Country capabilities and the strategicstate: How national political institutions affect multinational corpora-tions’ strategies. Strategic Management Journal, 15: 113–129.

Mutinelli, M., & Piscitello, L. (1998). The entry mode choice of MNEs: Anevolutionary approach. Research Policy, 27: 491–506.

Neumayer, E., & Spees, L. (2005). Do bilateral investment treaties increaseforeign direct investment to developing countries? World Development,33: 1567–1585.

Nordstrom, K. A., & Vahlne, J. E. (1994). Is the globe shrinking? Psychicdistance and the establishment of Swedish sales subsidiaries during thelast 100 years. In M. Landeck (Ed.), International Trade: Regional andGlobal Issues (pp. 41–56). New York: St. Martin’s Press.

OECD. (2007). International investment perspectives: Freedom of investment ina changing world. Paris: OECD.

Padmanabhan, P., & Cho, K. R. (1995). Methodological issues in internationalbusiness studies: The case of foreign establishment mode decisions bymultinational firms. International Business Review, 4(1): 55–73.

Padmanabhan, P., & Cho, K. R. (1999). Decision specific experience in foreignownership and establishment strategies: Evidence from Japanese firms.Journal of International Business Studies, 30(1): 25–44.

Pan, Y. (1996). Influences on foreign equity ownership level in joint venturesin China. Journal of International Business Studies, 27(1): 1–26.

Pan, Y., & Tse, D. (2000). The hierarchical model of market entry methods.Journal of International Business Studies, 31(4): 535–554.

Park, S. H., & Ungson, G. R. (1997). The effect of national culture, organiza-tional complementarity and economic motivation on joint venturedissolution. Academy of Management Journal, 40(2): 279–307.

Peng, M. W. (2001). The resource based view and international business.Journal of Management, 27(6): 803–829.

Raju, M. K., & Prahalad, C. K. (1980). The emerging multinationals: Indianenterprise in the ASEAN region. Madras, India: MK Raju ConsultantsPrivate Ltd.

Root, F., & Ahmed, A. (1979). Empirical determinants of manufacturing directforeign investment in developing countries. Economic Development andCultural Change, 27(4): 751–767.

Rumelt, R. P. (1974). Strategy structure and economic performance. Boston:Division of Research, Harvard Business School.

Shan, W. (1991). Environmental risks and joint venture sharing arrange-ments. Journal of International Business Studies, 22(4): 555–577.

Shenkar, O. (2001). Cultural distance revisited: Towards a more rigorousconceptualization and measurement of cultural differences. Journal ofInternational Business Studies, 32(3): 519–535.

Shenkar, O., Luo, Y., & Yeheskel, O. (2008). From ‘‘distance’’ to ‘‘friction’’:Substituting metaphors and redirecting intercultural research. Academyof Management Review, 33(4): 905–923.

Sim, A. B., & Pandian, J. R. (2003). Emerging Asian MNEs and their inter-nationalization strategies: Case study evidence on Taiwanese and Sin-gaporean firms. Asia Pacific Journal of Management, 20: 27–50.

Sohn, J. H. D. (1994). Social knowledge as a control system: A proposition andevidence from the Japanese FDI behavior. Journal of International BusinessStudies, 25(2): 295–324.

Stopford, J. M., & Wells, L. T. (1972). Managing the multinational enterprise.New York: Basic Books.

Srivastava, R. K., & Green, R. T. (1986). Determinants of bilateral trade flows.The Journal of Business, 59(4): 623–640.

Stottinger, B., & Schlegelmilch, B. B. (1998). Explaining export developmentthrough psychic distance: Enlightening or elusive? International Market-ing Review, 15(5): 357–372.

Sullivan, D., & Bauerschmidt, A. (1990). Incremental internationalization: Atest of Johanson and Vahlne’s thesis. Management International Review,30: 19–30.

Tatoglu, E., Glaister, K. W., & Erdal, F. (2003). Determinants of foreignownership in Turkish manufacturing. Eastern European Economics,41(2): 5–41.

Tatoglu, E., & Glaister, K. W. (1996). Trends and patterns of European foreigndirect investment in Turkey. European Business Review, 96(6): 11–21.

Tsang, E. W. K., & Yip, P. S. L. (2007). Economic distance and survival offoreign direct investments. Academy of Management Journal, 50(5):1156–1168.

UNCTAD (2006). World Investment Report 2006. New York and Geneva:United Nations.

UNCTAD (2008). World Investment Report 2008. New York and Geneva:United Nations.

Welch, D. E., Welch, L. S., & Marschan-Piekkari, R. (2001). The persistentimpact of language on global operations. Prometheus, 19(3): 193–208.

Wells, L. T. (1983). Third world multinationals: The rise of foreign investmentfrom developing countries. Cambridge, MA: MIT Press.

West, J., & Graham, J. L. (2004). A linguistic based measure of culturaldistance and its relationship to managerial values. Management Inter-national Review, 44(3): 239–260.

Wetherill, G. B. (1986). Regression analysis with applications. London: Chap-man and Hall.

Wilson, B. (1980). The propensity of multinational companies toexpand through acquisitions. Journal of International Business Studies,12: 59–65.

Witt, M. A., & Lewin, A. Y. (2007). Outward foreign direct investment asescape response to home country institutional constraints. Journal ofInternational Business Studies, 38(4): 579–594.

Woodcock, P. C., Beamish, P. W., & Makino, S. (1994). Ownership based entrymode strategies and international performance. Journal of InternationalBusiness Studies, 25(2): 253–273.

Wright, M., Filatotchev, I., Hoskisson, R. E., & Peng, M. (2005). Guest editor’sintroduction—Strategy research in emerging economies: Challengingconventional wisdom. Journal of Management Studies, 42(1): 1–33.

Yeung, H. W. C. (1998). Transnational corporations and business networks:Hong Kong firms in the ASEAN region. London: Routledge.

Yeung, H. W. C. (1999). The globalization of business firms from emergingeconomies, 2 vols. Cheltenham: Edward Elgar.

Yiu, D., & Makino, S. (2002). The choice between joint venture and whollyowned subsidiary: An institutional perspective. Organization Science,23(6): 667–683.

Yiu, D., Lau, C. M., & Bruton, G. (2007). International venturing by emergingeconomy firms: The effects of firm capabilities, home country networksand corporate entrepreneurship. Journal of International Business Studies,38(4): 519–540.

Zhao, H., & Zhu, G. (1998). Determinants of ownership preference of inter-national joint ventures: New evidence from Chinese manufacturingindustries. International Business Review, 7: 569–589.