Printer Friendly

Do overseas R&D laboratories in emerging markets contribute to home knowledge creation? An extension of the double diamond model.

Abstract:

* The double diamond model contends that both home and host locations affect MNE's international competitiveness. Drawing on the view that multinationals act as a link between home and host, we extend this framework and investigate the indirect impact of host on home location with reference to R&D internationalisation in emerging economies.

* By resorting to a sample of 221 large OECD regions from which R&D investments departed to the top six host emerging economies, we evaluate the contribution of different OECD R&D laboratories to the home knowledge creation of the OECD investing region.

* We test the complementarity between domestic R&D and different value-added R&D activities carried out by different technology-intensive R&D laboratories in terms of home knowledge creation of OECD investing regions.

* Our findings suggest that the activity of R&D laboratories focusing on adaptation complements domestic R&D in terms of knowledge creation regardless of the technological intensity of their operations, while the activity of medium technology-intensive R&D laboratories focusing on development is complementary to domestic R&D.

Keywords: Overseas R&D laboratories * Home region knowledge creation * Emerging markets

Introduction

To enhance their competitiveness, multinational enterprises (MNEs) rely on multiple geographical sources of knowledge by tapping into host location-specific advantages (e.g., Dunning 1977; Cantwell 1989; Rugman and Verbeke 1993), and benefit from them through reverse knowledge transfer from foreign subsidiaries to the parent (Ambos et al. 2006). An extensive body of literature has dealt with the contributions of different types of overseas R&D units to the MNE's knowledge base (Kuemmerle 1999; Pearce and Papanastassiou 1999; von Zedtwitz and Gassmann 2002; Cantwell and Mudambi 2005). However, this stream of research has primarily analysed R&D internationalisation from developed markets (DM) to other DM. A first reason is that the growth of overseas R&D laboratories in fast-growing emerging markets (EM) initiated only in the 1990s (von Zedtwitz 2006) and boomed during the early 2000s (UNCTAD 2005). A second reason is that R&D FDI in fast-growing EM was largely identified as a means to support the growing local market demand and, in this case, little contribution to the knowledge base of the whole MNE was expected (von Zedtwitz 2006). Meanwhile, the internationalisation of R&D activities by MNEs in EM has accelerated and changed facet (UNCTAD 2005). Traditionally, foreign R&D investments in those countries have been a means to adapt products or processes to local markets with little or no contribution to the knowledge home-base. Recent trends suggest a more strategic role of some EM in attracting R&D investments (e.g., D'Agostino et al. 2010) as a result of the emergence of talent pools and adequate expertise (Buckley and Ghauri 2004; Lewin et al. 2009) as well as of the technological upgrading in mature technologies of these countries (Athreye and Cantwell 2007; Ramamurti 2009).

This new trend of R&D internationalisation from DM to EM poses a number of challenges to international business (IB) research and calls for an analysis of the opportunities EM may offer to or the threat they may represent for DM location advantages (LAs). IB scholars converge on the idea that both home and host environments contribute to the firm's international competitiveness as this is strictly connected to the institutional set-up and spatially-bounded knowledge-flows of the external environment (Dunning 1977; Rugman and Verbeke 2001a). In particular, Rugman and Verbeke's (1993) double diamond model explains MNE's international competitiveness as a direct result of both home and host LAs. In line with this view and with reference to the internationalization of R&D in EM, we extend the double diamond model to account also for the indirect effect of host on home LAs. To this end, we draw on the view of the MNE as a differentiated network and argue that the MNE acts as a link between home and host locations, being able to transfer location-bound assets from one location to another (Bartlett and Ghoshal 1990). We focus on the impact of host EM LAs (as perceived by DM MNEs and reflected in their decision to locate specific types of R&D activities in EM) on DM LAs in terms of new knowledge creation. We answer a call for more IB research at sub- national level (Nachum 2000) and adopt the home region of the investing firm as unit of analysis, which the innovation literature recognises as a valuable unit of observation to describe the spatially-bounded factors that influence the innovation of local firms (Cooke 2005) and the systemic development of knowledge creation (Freeman 1987). Specifically, we investigate whether and how different types of R&D laboratories of OECD firms in the top six EM recipients (i.e., Brazil, Russia, India, China, Singapore and Taiwan- -BRICST) complement domestic R&D in terms of home knowledge creation of the investing region. To this end, we focus on different technology-intensive R&D laboratories carrying out in BRICST research, development and adaptation.

Drawing on a sample of 221 large OECD regions from which R&D investments departed to BRICST countries, our findings suggest that R&D activities of overseas laboratories in BRICST complement domestic R&D if they are low value-added (such as R&D adaptation) regardless of the technological intensity of the OECD laboratories operations. In addition, we find that the activity in BRICST of overseas medium technology-intensive R&D laboratories focusing on higher value-added R&D activities (such as development) is complementary to domestic R&D.

Theoretical Framework

LAs are the benefits associated with the localisation of certain activities in particular countries or regions. Extant IB research converges on the idea that home LAs are a major source of international competitiveness (Dunning 1977; Porter 1990; Rugman and Verbeke 2001a; Kuemmerle 2005). Drawing on the OLI paradigm (Dunning 1977), Rugman and Verbeke's (1993) double diamond model proposes that foreign input markets for valuable resources and/or foreign output markets for delivery of end products do contribute to international competitiveness. The double diamond model moves from the critique to Porter's (1990) diamond model and traditional theories of MNEs which contend that international competitiveness is solely grounded in the home country (Vernon 1966; Hymer 1990). This view was primarily motivated by the dominating FDI pattern whereby non-location-bound advantages were created in the home country and subsequently transferred to host country subsidiaries serving as a base for MNE's international competitiveness. However, in the last decades the significance of localised accumulated specialised resource pools and positive externalities for firm competitiveness has drawn attention to location-bound assets. These are assets that benefit a company only in a particular location (or set of locations) as they "cannot easily be transferred as an intermediate good and require significance adaptation in order to be used in other locations" (Rugman and Verbeke 2001b, p. 240). To account for this, the double diamond model explains international competitiveness in terms of both home and host location-bound assets. The implicit view of the MNE underlying this argument is that of an inter-organization network including both headquarters and different subsidiaries which are embedded in an external network of customers, suppliers, regulators and competitors with which they must interact (Bartlett and Ghoshal 1990). In this multi-hub 'integrated network' each unit contributes to the creation of new knowledge by relying on the knowledge available in its external environment (Gupta and Govindarajan 1994; Ghoshal and Nohria 1997). In particular, headquarters embeddedness at home enables to build the home base of the MNE international competitiveness by drawing on the home country location-bound assets. Foreign subsidiaries embeddedness in host locations enables the entire network to benefit from host country location-bound assets since the subsidiary assimilates and applies new external knowledge, and then transfers it within the MNE network (Andersson et al. 2001). Thus, the subsidiary plays a key role in the transformation of location-bound knowledge assets into ownership advantage for the whole MNE (Rugman and Verbeke 2001b).

[FIGURE 1 OMITTED]

With respect to R&D internationalisation, it has been argued that foreign R&D investments are motivated by the need to build upon and extend the extant core competences of the MNE or to access complementary assets that are lacking at home (Cantwell 1989). The implications of this reasoning are that both host and home locations of R&D activity can potentially enhance international competitiveness as in the double diamond framework. A number of studies have indeed documented that the MNE home location is an important source of valuable intangible assets and capabilities that can be exploited to prosper globally (Johansson and Vahlne 1977). Empirical evidence focusing on sub- national locations shows that there is a link between the technological capacities of the home region and the innovativeness of the MNEs that have their headquarters in the region (Cantwell and Iammarino 2003; Santangelo 2000). Similarly, evidence is available on the contribution of foreign R&D units to the creation of new technological competences at home through the leveraging of host-location specific knowledge assets (Birkinshaw and Hood 1998; Kuemmerle 1999; von Zedtwitz and Gassmann 2002). By means of reverse knowledge transfer foreign subsidiaries are able to transfer host country valuable knowledge to the parent company and, as a result, transform location-bound knowledge assets into ownership advantages for the whole MNE (Rugman and Verbeke 2001b).

Within the present trend of growing R&D internationalisation which has increasingly involved EM as new players, we extend the double diamond model to account also for the indirect effects of host on home LAs. The differentiated MNE network acts indeed as a link between home and host location being able to absorb and transfer location-bound knowledge assets from one location to another. In particular, host location- bound knowledge assets are increasingly transferred from the overseas units back to the headquarters embedded in the home environment where these assets then spill over affecting home location-bound knowledge assets. Our argument can be summarized in Fig. 1 where the direct home and host effects on the MNE are illustrated by the dotted line and the indirect effects on one location over the other are illustrated by the straight lines. In our analysis, we focus on the indirect effects of host on home LAs. (1)

Host LAs and Overseas R&D Laboratories

Many factors in the host location may affect the home-base knowledge of investing MNEs and, as a consequence, the MNE home LAs. In particular, the level of development of the host location, the availability of host specialised knowledge, the host supply of technical and research-base labour, the possibility to establish linkages with public research centres are factors that contribute to the host LAs and attract specific activities by foreign investors. These factors have a direct effect on MNE international competitiveness by means of reverse knowledge transfer from subsidiary to parent (Mansfield 1984) and an indirect effect on the home LAs as a result of the MNE embeddedness in its home market (Forsgren et al. 2005).

As far as R&D internationalisation is concerned, extant research has documented that the type of LAs that the host location can offer determine the types of incoming R&D laboratories (Pearce 1999). That is, the different types of foreign R&D laboratories established in the host location revealed the perception of the host LAs by the foreign investor. Highly innovative locations attract mainly foreign-R&D laboratories that augment home-base knowledge, and less technological dynamic locations attract R&D laboratories exploiting home-base knowledge (Dunning and Narula 1995; Florida 1997; Kuemmerle 1999; Papanastassiou 1999; Pearce and Papanastassiou 1999; Cantwell and Mudambi 2005). In host markets offering LAs related to the acquisition of new complementary local knowledge and/or to the monitoring of local scientific knowledge, foreign R&D laboratories have oriented their activity towards research aiming at augmenting the exiting home-base knowledge and eventually to create a possible commercial application of the research outcome, which can be exploited by the whole multinational network (Pearce and Papanastassiou 1999). Due to its strategic role, this type of R&D laboratories is likely to be located close to the parent company. However, the MNE may decide to undertake research activity in foreign locations where it can benefit from knowledge spillovers from local firms, universities or research institutes. Traditionally, these laboratories have been located in innovative clusters and technological dynamic regions of advanced countries (Cantwell and Janne 1999). In host markets primarily offering location advantages related to local skills and expertises (e.g., applied scientists, technicians, engineers), and opportunities to exploit economies of scale in R&D and market demand (Dunning 1993; Enright 2009), foreign R&D laboratories have traditionally focused on development of entirely new commercial products, and/or of specific product and/or process characteristics. This type of R&D activity has been mainly located in overseas R&D laboratories in advanced countries and has been traditionally undertaken by both laboratories with local mandate and laboratories that are more oriented toward the global R&D activities of the MNE (Pearce and Singh 1992; Pearce and Papanastassiou 1999). In host markets primarily offering LAs related to the domestic demand potentials, foreign R&D activity has traditionally offered technical support to local production for minor product or process adaptation. Being primarily market-driven, laboratories carrying out adaptation R&D type of activities exploit the home-base knowledge (Ronstadt 1977; Hood and Young 1982; Dunning and Narula 1995). With reference to the Triad countries, this type of activity is performed on a regular basis by a large share of foreign R&D laboratories typically holding a strong local mandate (Pearce and Singh 1992; Pearce and Papanastassiou 1999).

Extant literature has primarily examines the geographical configuration of different types of R&D laboratories by MNEs almost exclusively with reference to DM, where technological advanced regions are located (an exception is the work by D'Agostino et al. 2010). However, more recently EM and, especially, BRICST have received the majority of foreign R&D investments (UNCTAD 2005). The growing number of R&D inward FDI in EM during the early 2000s (UNCTAD 2005) makes wondering whether foreign R&D laboratories have evolved towards a more strategic role as a result of the evolution of host LAs, as also suggested by the rise of EM MNEs (Sauvant 2008).

Hypothesis Development

The indirect effects from host to home location illustrated in Fig. 1 by the straight lines take place depending on the specific characteristics of the foreign operations. Specifically, the contribution of different types of R&D laboratories of DM firms in BRICST to home regional knowledge creation is likely to vary depending on the type of R&D activity the overseas laboratories carry out and the technological intensity of the laboratories operations.

The wide diffusion of information and communication technology (ICT) has enabled firms to geographically diversify their knowledge production and eased technological modularization (Santangelo 2001). As a result, firms are fine slicing their global value chain by focusing at home on narrower activities associated with the highest value and dispersing lower value-added activities more cheaply and efficiently in locations willing to exploit them (Mudambi 2008). These receiving locations regard the opportunities to exploit the lower value-added activities as stepping-stones in the course to climb up the value chain.

With reference to R&D activities, EM such as BRICST benefit from high skills at relative low costs, strong educational institutions and ICT infrastructures together with effective government industrial policies (Arora et al. 2001; Ernst 2006; Manning et al. 2008). These conditions ensure that lower value-added activities along the R&D value chain such as adaptation can be more efficiently and cheaply carried out in these countries (Fifarek and Veloso 2010). Traditionally, the ultimate aim of R&D laboratories focusing on adaptation has been to offer technical support to local production with little or no contribution to the knowledge home-base. Foreign R&D laboratories established in the 1990s in South-East Asia and China were initially driven by new market- opportunities (yon Zedtwitz 2006). More recently, the dramatic growth of R&D FDI in these countries has been related to the improved local conditions, the rise of local innovative clusters, investments in infrastructures and availability of local talents (Buckley and Ghauri 2004). These countries have indeed reached a critical threshold of technological competences as shown by the local availability of talent pools and adequate expertise (Buckley and Ghauri 2004), which enable them to produce original knowledge (Athreye and Cantwell 2007). Accordingly, R&D activities in those countries are not exclusively motivated by the technical support to the local production, but progressively by opportunities to source local knowledge and technology (Dossani and Kenney 2003; Doh 2005; Manning et al. 2008). As a result, foreign R&D laboratories focusing on R&D adaptation can exploit host LAs related to knowledge creation and now contribute to the knowledge home-base.

This pattern is common across different technology-intensive sectors, despite the uneven technological upgrading of BRICST across sectors (Ramamurti 2009). Knowledge- intensive services (such as IT services and software) provide an emblematic example. In these technologies, BRICST and especially India (and more recently China) have been mainly engaged in customisation, which constitutes the lower value-added activity of the industry (Arora and Gambardella 2006). Similarly, Chinese software industry is still embryonic and Chinese top software companies exhibit weak capabilities, both in the organizational attainments and in the international quality standards of their products (Arora and Gambardella 2006). However, IT engineers and technicians in the overseas laboratory are able to find out improvements or new ideas that can be easily communicated at home while adapting extant service products to local market. Similarly, in medium and high technology-intensive sectors, overseas R&D laboratories focusing on adaptation in these countries are able to contribute to home knowledge creation as a result of the critical mass of technological competences now locally available (Athreye and Cantwell 2007).

Hypothesis 1: Regardless of the technological intensity of their operations, R&D laboratories of OECD firms focusing on adaptation in BRICST complement domestic R&D in terms of home region knowledge creation.

EM such as BRICST have undertaken a process of technological upgrading and have now entered into a phase of imitation and replication of foreign technologies, being ultimately able to originally produce new knowledge (Athreye and Cantwell 2007). Nonetheless, these countries are not at the technological frontier and are still catching up in more complex technologies (Altenburg et al. 2008; Ramamurti 2009). As a result of the global rationalization of knowledge creation across technologies, firms tend to geographically disperse abroad R&D activities in technologies that lie outside their own core competences while focusing at home on their technological core (Cantwell and Santangelo 2000). DM firms indeed have a traditional advantage in more tacit, less codified and more complex technologies whose development they retain at home, while EM such as BRICST have developed capabilities in technologies that are easy to codify and, as result, display a lower degree of complexity and geographical concentration (Cantwell and Santangelo 1999). Specifically, BRICST have now attained technological competences in mature technologies as documented by their dominance in medium technology-intensive sectors, where they benefit from (skilled) labour and resource-based advantages (Nolan 2004; von Zedtwitz 2006).

BRICST technological upgrading has enabled these countries to climb up the R&D value chain in medium technologies and develop expertise in less lower value- added R&D activities such as R&D development, where higher up knowledge creation takes place (Pearce and Singh 1992). Therefore, the technological upgrading of these countries secures LAs that transcend the more efficient and cheap conditions these locations offers (Fifarek and Veloso 2010) and relate to BRICST capabilities in higher up knowledge creation activities in less complex technologies.

Hypothesis 2: Medium technology-intensive R&D laboratories of OECD firms focusing on development in BRICST complement domestic R&D in terms of home region knowledge creation.

Methodology

Our sample consists of 221 large regions of 21 OECD countries (2), where the headquarters of the firms investing in R&D laboratory in BRICST is located. The BRICST countries are the top six host emerging economies over the period 2003-2005 of all R&D investments by OECD firms (fDi Markets database). The choice of the region as unit of analysis draws on the innovation literature, which has largely reported on the local nature of learning and on the systemic development of knowledge creation (Jaffe et al. 1993), which is the outcome of trust-based economic interactions and region-specific legal setting (Lundvall 1992; Cooke et al. 1997). Cantwell and Iammarino (2003) and Santangelo (2000) acknowledge indeed the link between the technological capacity of the home region and the competitiveness of the MNEs that have their headquarters in the region. Moreover, greater innovation differentials are documented across regions in DM (Usai 2010; Santangelo 2002).

Our dataset draws on three main sources: The OECD REGPAT database (January 2010), fDi Market database, and the OECD Regional Database (RDB).

REGPAT collects patent applications filed according to the Patent Cooperation Treaty (PCT) procedure and designating the European Patent Office (EPO) for the final grant. PCT procedure allows to request patent rights in many countries by a international application (Maraut et al. 2008). The main advantages of the REGPAT database are twofold. Firstly, it reduces possible domestic bias. EPO is generally considered non-bias toward a particular nation (Le Bas and Sierra 2002) and the international-based PCT procedure is less likely to suffer from domestic bias as confirmed by the equal contribution of European and US applicants to the growth of PCT applications (Khan and Dernis 2006). Secondly, in REGPAT the regional code of the inventor's residence is assigned to each patent according to the OECD (OECD 2008) regional divisions. In our analysis, we adopt Territorial Level 2. (3)

R&D FDI data are drawn from the fDi Market database, whose information are available since 2003. For each investment, fDi Market reports the sector and location of both the investing firm and foreign facility, as well as a description of overseas activity. For the sake of this study, we classified each OECD R&D laboratory in BRICST depending on the technological-intensity of the sector of operation and the type of R&D activity undertaken. To this end, we converted the SIC sector of foreign operation provided by fDi Market into the OECD technological-intensive sectors classification (Hatzichronoglou 1997): Low, medium-low, medium-high, and high technology-intensive sectors. Due to a very low number of observations in the former two categories, we aggregated the R&D laboratories in low, medium-low, and medium-high technology intensive sectors in one single sector, which we named medium technology-intensive. In addition, we relied on the EUROSTAT (Eurostat 2005) classification to classify R&D investments in services. We also used the description of the investment in fDi Markets to classify each R&D unit according to three categories: Research, development and adaptation. (4) Specifically, we surveyed previous studies (Ronstadt 1977; Hood and Young 1982; Pearce and Singh 1992; Dunning 1993; Pearce and Papanastassiou 1999) to identify specific keywords, and run a manual keyword search on the description of the investment. If incomplete, we integrate this description with online information. Thus, first we indentified R&D laboratories carrying out locally research activity defined as "basic", "scientific", "fundamental", "frontier technology" research, and application of such research that are potentially relevant for the activity of the MNE. In this category, we also included the activities of "hub", "centre of excellence", or part of a "global" network of R&D centres, as these laboratories perform mainly basic and applied research as part of integrated cross-border R&D activity (Hood and Young 1982; Pearce and Papanastassiou 1999). Second, we identified R&D laboratories locally carrying out development activity that refers to "development" and "solutions" of identifiable products or processes for commercialisation or engineering improvements (Ronstadt 1977; Hood and Young 1982; Dunning and Narula 1995). Finally, we identified R&D laboratories locally carrying out product and process adaptation which include the R&D laboratories that apply current products or technologies to the local "customer needs". We also included in this category the R&D laboratories that "support" local sales and marketing, and provides "technical services" (Dunning and Narula 1995).

The third source we rely on is the OECD RDB which provides data on regional R&D expenditures, as well as other socio-economic indicators. In addition, we also resort to the UNCTAD FDI database.

Variables

Dependent Variable

Our dependent variable is a proxy for home region knowledge creation (Griliches 1984, 1990; Jaffe et al. 1993). Patents are fairly good, although not perfect, measure of innovation activity at the sub-national level (Acs et al. 2002). They serve the purpose of our research as they provide information on the location of the invention (i.e., the inventor's address). Thus, we measured knowledge creation as the fractional count of PCT patent applications aggregated by the region i of residence of the inventor in the years 2006-2007 (2-year average) by thousand inhabitants and transformed in logarithm (Homekc) to ensure the normality of the distribution. (5)

Independent Variables

The independent variables are the 3-year average R&D expenditures as percentage of GDP in region i ([RDhome.sub.i]), and the 3-year total number of R&D investments made by MNEs, whose parent is located in the OECD region i ([RDhost.sub.ikz]), in BRICST in the technological-intensive sector k (with k equal to high and medium technology- intensive sectors, and knowledge-intensive services) and in the innovative activity z (with z equal to research, development and adaptation).

We want to test whether the R&D performed in the home OECD region and the R&D activities of laboratories of OECD firms in BRICST are complements in terms of home knowledge creation of the investing region. The complementarity between any two elements (e.g., activities of regions) implies that doing more of one element increases the payoff of doing the other element. We applied the productivity approach to complementarity (Mohnen and Roller 2005; Cassiman and Veugelers 2006) and transformed RDhome and [RDhost.sub.kz] into dummies. The choice of dummies rather than continuous variables primarily relies on the skewness of the variable [RDhost.sub.kz], making meaningless any continuous measure. We, therefore, created a dummy variable (Home) which takes value 1 if RDhome of region i is greater than the own country R&D intensity. R&D intensity is highly heterogeneous across OECD regions (OECD 2010) and these between-region differences reflect the disparity of countries in terms of economic development (Usai 2010). In addition, we created nine dummy variables [Host.sub.kz], which result from the combinations of the three k technology-intensive sectors and the three z types of R&D activities. Specifically, for each OECD region, we computed the average number of investments in each of the nine kz combinations and assigned value 1 to [Host.sub.kz] if the number of the kz-investments is higher than the average number of kz-investments departing from all OECD regions. Then, we created all possible interactions between Home and the nine [Host.sub.kz] variables. Specifically, [Homehost.sub.kz] equals 1 if Home equals I and the relevant [Hos.sub.k2] variable equals 1, and [Onlyhost.sub.kz] equals 1 if Home equals 0 and the relevant [Host.sub.kz] variable equals 1. In addition, [Onlyhome.sub.kz] equals 1 if Home equals 1 and [Host.sub.kz] equals 0, and [Nohomehost.sub.kz] equals 1 if Home and [Host.sub.kz] equal 0. We end up with nine groups of four categories to be used in nine different specifications of our model.

Controls

Home regional knowledge creation is affected by several home region-specific factors (Cooke et al. 1997; Acs et al. 2002). To proxy for the strength of the local inter-firm relationships, we assume that in agglomerations firms have more possibilities and occasions to interact, as for examples in cities (Feldman and Audretsch 1999) or clusters (Porter 1990), and we use the population density to proxy for the level of local agglomeration (Interfirm) (Sterlacchini 2008). To control for the role of public sector, we use a dummy (Public Sector) accounting for the fact that the region hosts the capital city, where considerable knowledge spillovers might arise between government research centres with large intramural R&D and local firms (Feldman 2003). We account for the closeness of the financial institutions to the local innovative firms with the share of employment in financial intermediation (Financial Sector) (Cooke et al. 1997). Since firms move away from relying exclusively on internal R&D towards a more 'open' R&D activity (Chesbrough 2003) and international technological collaborations increase (Hagedoorn and Duysters 2002), we control for openness of the home region R&D organisation by the share of patents with multiple inventors where at least one inventor is located in another country (Openness). In line with studies that highlight the role of university in regional innovativeness, we use the number of student enrohnents at the tertiary level (University) that is a proxy for the size of university (Anselin et al. 1997). We also control for the quality of human capital (Freeman 1987) by using the share of labour force with tertiary level of education (Human Capital). Business R&D controls for the private R&D expenditures, as customary in regional knowledge production function approaches (Anselin et al. 1997).

In addition, we control for different regional propensities to patent (Patent [Propensity.sub.j]) across technologies (Scherer 1983) by calculating a normalized revealed technological advantage (RTA) index for each of 5 broad technologies (Electrical Engineering (EE), Instruments (I), Chemistry (C), Mechanical Engineering (ME), and Other fields (O)). As regions vary in dimension, we control for regional size with population in logarithm (Population) (6). Finally, since the impact on the regional knowledge creation could be driven by R&D outward investments carried out in other countries (i.e., especially advanced economies), we use the number of R&D FDI in non-BRICST countries in the technological sector k taken in logarithm (R&D FD[I.sub.k]).

To account for the participation of the home country in global FDI flows, we introduce a binary variable taking value 1 for countries with positive balance between inward and outward FDI stocks, 0 otherwise (Attractiveness). We also control for home country effects with two dummies for North America (i.e., US and Canada) and Western European countries (EU), with Others serving as base category.

We introduce dummies variables to account for host country sector-specific characteristics which may affect FDI location choice (e.g., intellectual property right system) and MNEs' technological strategies (Lall 2003; Zhao 2006).

All controls refer to the period 2003-2005. Table 1 shows the correlation matrix and descriptive statistics.

The Model

As we are dealing with geographically-close cross-sectional data, spatial autocorrelation may be an issue (Anselin 1988). Thus, we firstly tested for the presence of spatial dependence in our data by means of Moran's I test with a binary contiguity matrix (W). (7) As the Moran's I test detected spatial autocorrelation, we performed the relevant tests (i.e., LMLAG and LM-ERROR) on the OLS models--which would be inefficient in the presence of spatial autocorrelation--to choose the more appropriate spatial model. Test results indicate the spatial lag model as superior over the spatial error model. Thus, our maximum likelihood spatial lag model estimates knowledge creation in the investing home OECD region i at time t ([Homekc.sub.it]) as a function of a vector of the combinations of R&D activities [C.sub.cit-1], a set of controls [X.sub.it-1], at time t-1 and a parameter [rho] which accounts for spatial autocorrelation. A positive and significant coefficient of [rho] means that the patenting of region i depends on the patenting of its neighbours (Acs et al. 2002; Moreno et al. 2005). Specifically, we estimated

[Homekc.sub.it] = [C.sub.cit-1][theta] + [X.sub.it-1][beta] + [rho]W[Homekc.sub.it] + [[epsilon].sub.i] (1)

where the subscript c refers to the four combinations of [Homehost.sub.kz] and [Onlyhost.sub.kz], [Onlyhome.sub.kz] and [Nohomehost.sub.kz].

Our test of complementarity is a Wald (one-sided) test of the following constraint:

[[theta].sub.11] - [[theta].sub.10] [greater than or equal to] [[theta].sub.01] - [[theta].sub.00] (2)

where the first subscript refers to Home and the second to [Host.sub.kz]. The rejection of the null hypothesis of equality and substitution would confirm our argument of complementarity between Home and [Host.sub.kz].

Results and Discussion

Table 2 reports the econometric results.

In all nine models, the variable [Nohomehost.sub.kz] has been dropped due to collinearity with the other three combinative categories. (8) The variable [Onlyhome.sub.kz] is always positive and significant across all specifications. The results for regions setting up high technology-intensive R&D laboratories (Models 1-3) show that [Homehost.sub.kz] is positive and significant. These results suggest that regions both showing great domestic R&D intensity and carrying out any type of R&D activities in BRICST are very innovative. For regions setting up medium technology-intensive R&D laboratories in BRICST (Models 4-6), [Homehost.sub.kz] is positive and significant only when the laboratories carry out R&D development in BRICST (Model 5). For regions setting up R&D laboratories in knowledge-intensive services (Models 7-9), [Homehost.sub.kz] is positive and significant when R&D laboratories focus on adaptation (Model 9). As far as the variable [Onlyhost.sub.kz] is concerned, this is negative and significant when the foreign R&D laboratories carry out adaptation in high and medium technology-intensive sectors (Models 3 and 6). This latter result illustrates that regions that both offshore R&D adaptation in the high and/or medium technology-intensive sectors and show a poor R&D intensity at home display a low patenting activity, ceteris paribus. [Onlyhost.sub.kz] is also positive and significant when overseas R&D laboratories conduct research activity in medium technology-intensive. Since individual significance and sign of the coefficients do not detect complementarity (Mohnen and Roller 2005), we rely on the direct test of complementarity illustrated in Eq. (2).

We found complementarity between domestic R&D and the activity of the R&D laboratories carrying out adaptation in all three technology-intensive sectors in BRICST. This confirms that there are synergic effects between domestic R&D and R&D adaptation in BRICST in terms of home region knowledge creation (Hypothesis 1). Thus, BRICST are optimal locations for lower value-added R&D activities along the global R&D value chain as overseas R&D laboratories focusing on these activities in these countries are now capable to contribute to the knowledge home-base. This result can be framed within the emergence of talent pools and adequate expertise in EM (Buckley and Ghauri 2004; Lewin et al. 2009). We also found complementarity between domestic R&D and the activity of medium technology-intensive R&D laboratories carrying out development in BRICST. This finding is consistent with the technological upgrading of EM in more mature technologies where these countries now have moved up along the R&D value chain (Athreye and Cantwell 2007; Ramamurti 2009).

Conclusion and Implications

By challenging the view that the location of R&D laboratories in EM is mainly motivated by the support to the growing local market demand with little or no contribution to the home-base, this study suggests that overseas R&D laboratories in BRICST do contribute to home knowledge creation depending on the type of R&D activity they carried out and the technological intensity of their operations as a result of the emergence locally of talent pools and adequate expertise, and the technological upgrading of these countries. The study offers three contributions to IB research. First, it extends the double diamond model by suggesting an indirect effect of host on home LAs drawing on the view that the MNE acts as a link between its home and host network of embedded relationships. Second, it enriches the literature on the effects of FDI, which has largely explored the impact of inward FDI in the host markets. As far as the impact at home is concerned, the focus has primarily been on the direct effects of overseas subsidiaries on the parent company. By contrast, our knowledge on the impact on the home region where the parent is embedded is scant despite the relevance of this level of analysis to account for the spatially-bounded factors that influence the innovation of local firms (Cooke 2005) and the systemic development of knowledge creation (Freeman 1987). A final contribution is closely related to the IB literature on EM, which has so far investigated the determinants of inward FDI from DM to EM (Ramamurti 2004; Lewin et al. 2009) and the impact of outward FDI from EM on host DM (e.g., Sauvant 2008), but, as far as our knowledge is concerned, an analysis of the impact of inward FDI from DM to EM on the home DM is lacking (an exception is the work of D'Agostino et al. 2010), despite its policy and managerial implications.

The study provides important policy implications as to whether DM regional governments should favour or discourage FDI into EM. National and sub-national government are increasingly engaging in location "tournament" to attract and promote FDI (Oxelheim and Ghauri 2003; Santangelo 2004). Given the increasing relevance of sub- national regions (Buckley and Ghauri 2004), there is a need to provide regional governments from developed countries with clear guidelines on whether and how to design policies for outward FDI to EM and on their expected impact at home. The study bears also significant managerial implications. Debate is ongoing on the threats or opportunities that the new players may offer to established technological leaders in DM. The fine slicing of firms activities across optimal locations calls indeed for an increasingly sophisticated decisionmaking process by MNE managers (Buckley and Ghauri 2004). Thus, managers need to be aware of the gains and losses that investments in BRICST may bring about. Our findings may be useful in both respects as they neatly call for sector-specific and activity-specific policies and corporate strategies.

The study suffers from a number of limitations. First, we are not able to distinguish greenfield from brownfield investments and fail to account for other types of entry modes into EM such as merger and acquisitions, and joint ventures. However, greenfields tend to be the most common entry mode to establish R&D facilities in EM (UNCTAD 2005). Second, our data do not allow us to have more precise information on the roles of overseas subsidiaries within the internal MNE's network, for example in terms of regional or world mandate. More detailed survey data would be necessary to this end. Nevertheless, we are confident that, despite these drawbacks, the study advances our knowledge on the role of EM in DM knowledge creation.

References

Acs, Z. J., Anselin, L., & Varga, A. (2002). Patents and innovation counts as measures of regional production of new knowledge. Research Policy, 31(7), 1069-1085.

Altenburg, T., Schmitz, H., & Stamm, A. (2008). Breakthrough China's and India's transition from production to innovation. Worm Development, 36(2), 325-344.

Ambos, T. C., Ambos, B., & Schlegelmilch, B. B. (2006). Learning from foreign subsidiaries: An empirical investigation of headquarters' benefits from reverse knowledge transfers. International Business Review, 15(3), 294-312.

Andersson, U., Forsgren, M., & Pedersen, T. (2001). Subsidiary performance in multinational corporations: The importance of technology embeddedness. International Business Review, 10(1), 3-23.

Anselin, L. (1988). Spatial econometrics: Methods and models. Dordrecht: Kluwer Academic Publishers.

Anselin, L., Varga, A., & Acs, Z. (1997). Local geographic spillovers between university research and high technology innovations. Journal of Urban Economics, 42(3), 422-448.

Arora, A., & Gambardella, A. (2006). From underdogs to tigers: The rise and growth of the software industry in Brazil, China, India, Ireland, and Israel. Oxford: Oxford University Press.

Arora, A., Arunachalam, V. S., Asundi, J., & Fernandes, R. (2001). The Indian software services industry. Research Policy, 30(8), 1267-1287.

Athreye, S., & Cantwell, J. (2007). Creating competition? Globalisation and the emergence of new technology producers. Research Policy, 36(2), 209-226.

Bartlett, C. A., & Ghoshal, S. (1990). The multinational corporation as an interorganizational network. Academy of Management Review, 15(4), 603-625.

Birkinshaw, J., & Hood, N. (1998). Multinational subsidiary evolution: Capability and charter change in foreign-owned subsidiary companies. The Academy of Management Review, 23(4), 773-795.

Buckley, P. J., & Ghauri, P. N. (2004). Globalization, economic geography and the strategy of multinational enterprises. Journal of International Business Studies, 35(2), 81-98.

Cantwell, J. (1989). Technological innovation and multinational corporations. Oxford: Blackwell.

Cantwell, J., & Iammarino, S. (2003). Multinational corporations and European regional systems of innovation. London: Routledge.

Cantwell, J., & Janne, O. (1999). Technological globalisation and innovative centres: The role of corporate technological leadership and locational hierarchy. Research Policy, 28(2/3), 119-144.

Cantwell, J., & Mudambi, R. (2005). MNE competence-creating subsidiary mandates. Strategic Management Journal, 26(12), 1109-1128.

Cantwell, J., & Santangelo, G. D. (1999). The frontier of international technology networks: Sourcing abroad the most highly tacit capabilities. Information Economics and Policy, 11(1), 101-123.

Cantwell, J., & Santangelo, G. D. (2000). Capitalism, profits and innovation in the new techno-economic paradigm. Journal of Evohttionao, Economics, 10(1/2), 131-157.

Cassiman, B., & Veugelers, R. (2006). In search for complementarity in the innovation strategy: Internal R&D and external knowledge acquisition. Management Science, 52(1), 68- 82.

Chesbrough, H. (2003) Open innovation: The new Imperative for creating and profiting from technology. Boston: Harvard Business School Press.

Cooke, P. (2005). Regionally asymmetric knowledge capabilities and open innovation: Exploring 'Globalisation 2': A new model of industry organisation. Research Policy, 34(8), 1128-1149.

Cooke, P., Uranga, G. M., & Etxebarria, G. (1997). Regional innovation systems: Institutional and organisational dimensions. Research Policy, 26(4/5), 475-491.

D'Agostino, L., Laursen, K., & Santangelo, G. D. (2010). The impact of R&D offshoring on the home knowledge production of OECD investing regions. DRUID Working paper.

Doh, J. P. (2005). Offshore outsourcing: Implications for international business and strategic management theory and practice. Journal of Management Studies, 42(3), 695-704.

Dossani, R., & Kenney, M. (2003). Went for cost, stayed for quality? Moving the back office to India. Berkeley Roundtable on the International Economy (BRIE) Research Paper. University of California, Berkley.

Dunning, J. H. (1977). Trade, location of economic activity and the MNE: A search for an eclectic approach. In: B. Ohlin et al. (Eds.), The international allocation of economic activity (pp. 395 431). London: Macmillan.

Dunning, J. H. (1993).Multinational enterprises and the global economy. Reading: Addison-Wesley.

Dunning, J. H., & Narula, R. (1995). The R&D activities of foreign firms in the United States. International Studies of Management and Organisation, 25(1/2), 39-73.

Enright, M. J. (2009). The location of activities of manufacturing multinationals in the Asia-Pacific. Journal of International Business Studies, 40(5), 818-839.

Ernst, D. (2006). Innovation offshoring: Asia's emerging role in global innovation networks. Honolulu: East-West Center Special Report.

Eurostat (2005). Eurostat, statistics in focus, science and technology, 4/2005, R&D Statistics. Luxembourg: EUROSTAT.

Feldman, M. P. (2003). The locational dynamics of the US biotech industry: Knowledge externalities and the anchor hypothesis. Industry & Innovation, 10(3), 3111-3329.

Feldman, M. P., & Audretsch, D. B. (1999). Innovation in cities: Science-based diversity, specialization and localized competition. European Economic Review, 43(2), 409-429.

Fifarek, B. J., & Veloso, F. M. (2010). Offshoring and the global geography of innovation. Journal of Economic Geography, 10(4), 559-578.

Florida, R. (1997). The globalization of R&D: Results of a survey of foreign- affiliated R&D laboratories in the USA. Research Policy, 26(1), 85-103.

Forsgren, M., Holm, U., & Johanson, J. (2005). Managing the embedded multinational: A business network view. Cheltenham: Edward Elgar.

Freeman, C. (1987). Technology policy and economic performance: Lessons from Japan. London: Pinter.

Ghoshal, S., & Nohria, N. (1997). The differentiated MNC: Organizing multinational corporation for value creation. San Francisco: Jossey-Bass.

Griliches, Z. (1984) R&D, patents and productivities. Chicago: University of Chicago Press.

Griliches, Z. (1990). Patent statistics as economic indicators. Journal of Economic Literature, 28(4), 1661-1701.

Gupta, A. K., & Govindarajan, V. (1994). Organizing for knowledge flows within MNCs. International Business Review, 3(4), 443-457.

Hagedoorn, J., & Duysters, G. (2002). External sources of innovative capabilities: The preference for strategic alliances or mergers and acquisitions. Journal of Management Studies, 39(2), 167-188.

Hatzichronoglou, T. (1997). Revision of the high-technology sector and product classification. Paris: OECD Publishing.

Hood, N., & Young, S. (1982). US multinational R&D: Corporate strategies and policy implications for the UK. Multinational Business, 2(2), 10-23.

Hymer, S. H. (1990). The large multinational 'corporation': An analysis of some motives for the international integration of business In: M. Casson (Ed.), Multinational Corporations (pp. 6-31). Cheltenham: Edward Elgar.

Jaffe, A., Trajtenberg, M., & Henderson, R. (1993). Geographic localization of knowledge spillover as evidenced by patent citations. The Quarterly Journal of Economics, 108(3), 577-598.

Johansson, J. K., & Vahlne, J.-E. (1977). The international process of the firm: A model of knowledge development and increasing foreign market commitments. Journal of International Business Studies, 8(1), 23-32.

Khan, M., & Demis, H. (2006). Global overview of innovative activities from the patent indicators perspective. OECD Science, Technology and Industry Working Papers, 2006/3. OECD Publishing.

Kuemmerle, W. (1999). Foreign direct investment in industrial research in the pharmaceutical and electronics industries: Results from a survey of multinational firms. Research Policy, 28(2/3), 179-193.

Kuemmerle, W. (2005). The entrepreneur's path to global expansion. Sloan Management Review, 46(2), 42-49.

Lall, S. (2003). Indicators of the relative importance of IPRs in developing countries. Research Policy, 32(9), 1657-1680.

Le Bas, C., & Sierra, C. (2002). Location versus home country advantages' in R&D activities: Some further results on multinationals' locational strategies. Research Policy, 31(4), 589-609.

Lewin, A., Massini, S., & Peeters, C. (2009). Why are companies offshoring innovation? The emerging global race of talent. Journal of International Business Strategy, 40(6), 901-925.

Lundvall, B.-A. (1992). National systems of innovation: Towards a theory of innovation and interactive learning. London: Pinter.

Manning, S., Massini, S., & Lewin, A. Y. (2008). A dynamic perspective on next- generation offshoring: The global sourcing of science and engineering talent. Academy of Management Perspectives, 22(3), 35-54.

Mansfield, E. (1984). R&D and innovation: Some empirical findings. In: Z. Griliches (Ed.), R&D, patents and productivity (pp. 127-154). Chicago: The University of Chicago Press.

Maraut, S., Dernis, H., Webb, C., Spiezia, V., & Guellec, D. (2008). The OECD REGPAT database: A presentation. OECD science, technology and industry working papers. OECD Publishing.

Mohnen, P., & Roller, L.-H. (2005). Complementarities in innovation policy. European Economic Review, 49(6), 1431-1450.

Moreno, R., Paci, R., & Usai, S. (2005). Spatial spillovers and innovation activity in European regions. Environment and Planning, 37(10), 1793-1812.

Mudambi, R. (2008). Location, control and innovation in knowledge-intensive industries. Journal of Economic Geography, 8(5), 699-725.

Nachum, L. (2000). Economic geography and the location of MNEs: Financial and professional service FDI to the US. Journal of International Business Studies, 31(3), 367- 386.

Nolan, P. (2004). China at the crossroads. Cambridge: Polity Press.

OECD (2008). Territorial grid of OECD member countries. Paris: OECD Public Governance and Territorial Development Directorat. http://www.oecd.org/dataoecd/35/60/42392313.pdf. Accessed 10 Feb 2012.

OECD (2010). OECD Science, Technology and Industry Outlook 2010. Paris: OECD Publishing. http://dx.doi.org/10.1787/sti_outlook-2010-en. Accessed 10 Feb 2012.

Oxelheim, L., & Ghauri, E N. (2003). European Union and the race for inward foreign direct investment in Europe. Oxford: Elsevier.

Papanastassiou, M. (1999). Technology and production strategies of multinational enterprise (MNE) subsidiaries in Europe. International Business Review, 8(2), 213-232.

Pearce, R. D. (1999). Decentralised R&D and strategic competitiveness: Globalised approaches to generation and use of technology in multinational enterprises (MNEs). Research Policy, 28(2/3), 157-178.

Pearce, R. D., & Papanastassiou, M. (1999). Overseas R&D and the strategic evolution of MNEs: Evidence from laboratories in the UK. Research Policy, 28(1), 23-41.

Pearce, R., & Singh, S. (1992). Globalizing research and development. London: Macmillan.

Porter, M. E. (1990). The competitive advantage of the nations. New York: Free Press.

Ramamurti, R. (2004). Developing countries and MNEs: Extending and enriching the research agenda. Journal of International Business Studies, 35(4), 277-283.

Ramamurti, R. (2009). What have we learned about emerging market MNEs? In R. Ramamurti et al. (Eds.), Emerging multinationals from emerging markets (pp. 3-23). Cambridge: Cambridge University Press.

Ronstadt, R. (1977). Research and development abroad by U.S. multinationals. New York: Praeger.

Rugman, A. M., & Verbeke, A. (1993). Foreign subsidiaries and multinational strategic management: An extension and correction of Porter's single diamond framework. Management International Review, 33(2), 71-84.

Rugman, A. M., & Verbeke, A. (2001a). Location, competitiveness, and the multinational enterprise. In: A. M. Rugman, et al. (Eds.), The Oxford Handbook of International Business (pp. 150-177). Oxford: Oxford University Press.

Rugman, A. M., & Verbeke, A. (2001b). Subsidiary-specific advantages in multinational enterprises. Strategic Management Journal, 22(3), 237-250.

Santangelo, G. D. (2000). Inter-European regional dispersion of corporate ICT research activity: The case of German, Italian and UK regions. International Journal of Economics of Business, 7(3), 275-295.

Santangelo, G. D. (2001). The impact of the information technology and communications technology revolution on the internationalisation of corporate technology. International Business Review, 10(6), 701-726.

Santangelo, G. D. (2002). The regional geography of corporate patenting in information and communications technology (ICT): Domestic and foreign dimension. Regional Studies, 36(5): 495-514.

Santangelo, G. D. (2004). FDI and local capabilities in peripheral regions: The Etna Valley case. Transnational Corporations, 13(1), 73-106.

Sauvant, P. (2008). The rise of transnational corporations from emerging markets. Threat or opportunity? Cheltenham: Edward Elgar.

Scherer, F. M. (1983). The propensity to patent. International Journal of Industrial Organization, 1(1), 107-128.

Smarzynka Javorcik, B. (2004). Does foreign direct investment increase the productivity of domestic firms? In search of spillovers through backward linkages. The American Economic Review, 94(3), 605-627.

Sterlacchini, A. (2008). R&D, higher education and regional growth: Uneven linkages among European regions. Research Policy, 37(6/7), 1096-1107.

UNCTAD (2005). World Investment Report: TNCs and the internationalization of R&D. Geneva: United Nations.

Usai, S. (2010). The geography of inventive activity in OECD regions. Regional Studies, 45(6), 711-731.

Vernon, R. (1966). International investment and international trade in the product cycle. Quarterly Journal of Economics, 80(2), 190-207.

Von Zedtwitz, M. (2006). International R&D strategies of TNCs from developing countries: The case of China. In UNCTAD fed.), Globalization of R&D and developing countries Proceedings of an Expert Meeting (pp. 117-140). New York: United Nations.

Von Zedtwitz, M., & Gassmann, O. (2002). Market versus technology drive in R&D internationalization: Four different patterns of managing research and development. Research Policy, 31(4), 569-588.

Zhao, M. (2006). Conducting R&D in countries with weak intellectual property rights protection. Management Science, 52(8), 1185-1199.

Endnotes

(1) Figure I also reports the indirect effects of home on host LAs, which we will not discuss as the literature on inward FDI and spillovers has investigated them extensively (Smarzynska Javorick 2004).

(2) The 21 OECD countries are: Australia, Austria, Belgium, Canada, Czech Republic, Finland, France, Germany, Greece, Hungary, Ireland, Italy, South Korea, Luxemburg, the Netherlands, Norway, Slovak Republic, Spain, Sweden, the United Kingdom, and the United States. Due to missing data, we excluded 9 regions (2 Canadian regions, 2 Spanish autonomous regions and the Canary Islands, 2 Italian autonomous provinces, and Alaska and Hawaii in the US).

(3) For the majority of the European Union countries, the Territorial Levels are equivalent to the Nomenclature of Territorial Units for Statistics developed by Eurostat (see Maraut et al. 2008).

(4) Information availability prevented us from adopting a finer classification of R&D laboratories (e.g., Pearce and Papanastassiou 1999).

(5) When multiple inventors participate to the patent, the patent is equally shared among them. None of the sample regions records zero patents.

(6) An alternative measure of economic size could be GDE However, GDP is highly correlated with population size in logarithm.

(7) Such N x N matrix (where N is the number of regions) takes value 1 when the pair of regions share a border or are separated by few kilometres of sea- or lake-water (e.g., the US and the Canadian states along the Great Lakes area), 0 otherwise.

(8) An alternative solution to collinearity would be to drop the constant. However, the spatreg STATA command used does not allow this option. Consequently, our complementarity test is performed on three categories ([Homehost.sub.kz], [Onlyhost.sub.kz] and [Onlyhome.sub.kz]) according to the following rule:

[[theta].sub.11] - [[theta].sub.10] [greater than or equal to] [[theta].sub.01] (3)

When [Nohomehost.sub.kz] is used as the benchmark against the three other dummies, [[theta].sub.00] = 0. Accordingly, the inequality tests involving four (Eq. (2)) or three categories (Eq. (3)), respectively, are equivalent.
Table 1: Correlation matrix and descriptive statistics

                                 1             2

 1. Interfirm                      1
 2. Financial sector               0.506 ***    1
 3. Openness                       0.069       -0.078
 4. University                     0.213 ***    0.426 ***
 5. Human capital                  0.245 ***    0.400 ***
 6. Business R&D                   0.030        0.412 ***
 7. Patent [propensity.sub.EE]     0.121 *      0.370 ***
 8. Patent [propensity.sub.I]      0.033        0.270 ***
 9. Patent [propensity.sub.C]      0 116        0.206 ***
10. Patent [propensity.sub.ME]    -0.182 ***   -0.428
11. Patent [propensity.sub.0]     -0.072       -0.271 ***
12. Population                     0.067        0.459 ***
13. R&D [FDI.sub.H]                0.195 ***    0.548 ***
14. R&D [FDI.sub.M]                0.173 ***    0.425 ***
15. R&D [FDI.sub.KS]               0.120 *      0.397 ***
Mean                             269.18        13.51
S. D.                            727.71         4.7

                                 3             4            5

 1. Interfirm
 2. Financial sector
 3. Openness                      1
 4. University                   -0.24 ***      1
 5. Human capital                -0.021         0.386 ***    1
 6. Business R&D                 -0.127 *       0.225 ***    0.145 **
 7. Patent [propensity.sub.EE]   -0.065         0.305 ***    0.271 ***
 8. Patent [propensity.sub.I]    -0.156 **      0.146 **     0.091
 9. Patent [propensity.sub.C]    -0.081         0.174 ***    0.083
10. Patent [propensity.sub.ME]    0.155 **     -0.465 ***   -0.262 ***
11. Patent [propensity.sub.0]    -0.195 ***     - 0.131     -0.174
12. Population                   -0.26,0 ***    0.220 ***    0.090
13. R&D [FDI.sub.H]              -0.095         0.235 ***    0.278 ***
14. R&D [FDI.sub.M]              -0.008         0.047        0.189 ***
15. R&D [FDI.sub.KS]             -0.092         0.165 **     0.159 **
Mean                              7.18          4.24        24.77
S. D.                             7.16          1.83         8.41

                                 6              7

 1. Interfirm
 2. Financial sector
 3. Openness
 4. University
 5. Human capital
 6. Business R&D                    1
 7. Patent [propensity.sub.EE]      0.297 ***    1
 8. Patent [propensity.sub.I]       0.128 *      0.058
 9. Patent [propensity.sub.C]       0.018        - 0.311
10. Patent [propensity.sub.ME]     -0.298 ***   -0.523 ***
11. Patent [propensity.sub.0]      -0.224 ***   -0.352 ***
12. Population                      0.530 ***    0.222 ***
13. R&D [FDI.sub.H]                 0.689 ***    0.382 ***
14. R&D [FDI.sub.M]                 0.414 ***    0.171 **
15. R&D [FDI.sub.KS]                0.695 ***    0.267 ***
Mean                             1709.461       -0.24
S. D.                            4146.14         0.27

                                 8            9            10

 1. Interfirm
 2. Financial sector
 3. Openness
 4. University
 5. Human capital
 6. Business R&D
 7. Patent [propensity.sub.EE]
 8. Patent [propensity.sub.I]     1
 9. Patent [propensity.sub.C]     0.045        1
10. Patent [propensity.sub.ME]   -0.265 ***   -0.327 ***    1
11. Patent [propensity.sub.0]    -0.219 ***   -0.037        0.198 ***
12. Population                    0.186 ***    0.304 ***   -0.335 ***
13. R&D [FDI.sub.H]               0.139 **     0.069       -0.397 ***
14. R&D [FDI.sub.M]               0.049        0.034       -0.055
15. R&D [FDI.sub.KS]              0.129 *     -0.026       -0.259 ***
Mean                             -0.1         -0.03         0.08
S. D.                             0.23         0.21         0.23

                                 11           12          13

 1. Interfirm
 2. Financial sector
 3. Openness
 4. University
 5. Human capital
 6. Business R&D
 7. Patent [propensity.sub.EE]
 8. Patent [propensity.sub.I]
 9. Patent [propensity.sub.C]
10. Patent [propensity.sub.ME]
11. Patent [propensity.sub.0]     1
12. Population                   -0.123 *     1
13. R&D [FDI.sub.H]              -0.250 ***   0.567 ***   1
14. R&D [FDI.sub.M]              -0.106       0.456 ***   0.569 ***
15. R&D [FDI.sub.KS]             -0.113 *     0.455 ***   0.660 ***
Mean                              0.1         7.65        0.29
S. D.                             0.31        1.06        0.63

                                 14          15

 1. Interfirm
 2. Financial sector
 3. Openness
 4. University
 5. Human capital
 6. Business R&D
 7. Patent [propensity.sub.EE]
 8. Patent [propensity.sub.I]
 9. Patent [propensity.sub.C]
10. Patent [propensity.sub.ME]
11. Patent [propensity.sub.0]
12. Population
13. R&D [FDI.sub.H]
14. R&D [FDI.sub.M]              1
15. R&D [FDI.sub.KS]             0.505 ***   1
Mean                             0.18        0.13
S. D.                            0.45        0.45

*** p [less than or equal to] 0.01; ** p [less than or equal to] 0.05;
* p [less than or equal to]0.10

Table 2: Econometric results

DV: Homekc                            High
                             technology-intensive
                                    sectors

                                     Model 1

                                    Research

                             Coef.    S. E.

Explanatory variables

[Homehost.sub.kz]             0.053   (0.026) **
[Onlyhomce.sub.kz]            0.038   (0.009) ***
[Onlyhost.sub.kz]             0.011   (0.027)

Home region controls

Interfirm                     0.000   (0.000)
Public sector                -0.040   (0.019) **
Financial sector              0.006   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.003   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.019   (0.020)
Patent [propensity.sub.I]     0.000   (0.016)
Patent [propensity.sub.C]    -0.057   (0.024) **
Patent [propensity.sub.ME]   -0.064   (0.024) ***
Population                   -0.016   (0.004) ***
R&D [FDI.sub.k]               0.000   (0.011)

Home country controls

Attractiveness               -0.005   (0.010)
North America                -0.028   (0.018)
EU                            0.029   (0.018)

Host country controls

[Brazil.sub.k]               -0.035   (0.038)
[Russian.sub.k]              -0.081   (0.040) **
[Indian.sub.k]                0.041   (0.021) **
[China.sub.k]                 0.033   (0.019) *
[Singapore.sub.k]            -0.056   (0.025) **
[Taiwan.sub.k]                0.000   (0.025)
Constant                      0.052   (0.049)
Rho constant                  0.090   (0.013) ***
Sigma constant                0.050   (0.002) ***
N                                     221

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                0.01
(one-side)

DV: Homekc                            High
                             technology-intensive
                                    sectors

                                    Model 2

                                  Development

                             Coef.    S. E.

Explanatory variables

[Homehost.sub.kz]             0.077   (0.032) **
[Onlyhomce.sub.kz]            0.036   (0.009) ***
[Onlyhost.sub.kz]             0.018   (0.035)

Home region controls

Interfirm                     0.000   (0.000)
Public sector                -0.047   (0.019) **
Financial sector              0.006   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.003   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.018   (0.020)
Patent [propensity.sub.I]     0.001   (0.016)
Patent [propensity.sub.C]    -0.058   (0.024) **
Patent [propensity.sub.ME]   -0.063   (0.024) **
Population                   -0.014   (0.004) ***
R&D [FDI.sub.k]              -0.004   (0.011)

Home country controls

Attractiveness               -0.004   (0.010)
North America                -0.027   (0.018)
EU                            0.031   (0.017) *

Host country controls

[Brazil.sub.k]               -0.043   (0.038)
[Russian.sub.k]              -0.089   (0.039) **
[Indian.sub.k]                0.044   (0.023) *
[China.sub.k]                 0.026   (0.019)
[Singapore.sub.k]            -0.059   (0.024) **
[Taiwan.sub.k]               -0.003   (0.025)
Constant                      0.035   (0.048)
Rho constant                  0.094   (0.013) ***
Sigma constant                0.049   (0.002) ***
N                                     221

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                0.61
(one-side)

DV: Homekc                            High
                             technology-intensive
                                    sectors

                                    Model 3

                                   Adaptation

                             Coef.    S. E.

Explanatory variables

[Homehost.sub.kz]             0.049   (0.029 *
[Onlyhomce.sub.kz]            0.035   (0.009) ***
[Onlyhost.sub.kz]            -0.057   (0.034) *

Home region controls

Interfirm                     0.000   (0.000)
Public sector                -0.044   (0.019) **
Financial sector              0.006   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.003   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.019   (0.020)
Patent [propensity.sub.I]     0.002   (0.016)
Patent [propensity.sub.C]    -0.059   (0.024) **
Patent [propensity.sub.ME]   -0.063   (0.024) ***
Population                   -0.015   (0.004) ***
R&D [FDI.sub.k]               0.000   (0.011)

Home country controls

Attractiveness               -0.004   (0.010)
North America                -0.028   (0.018)
EU                            0.028   (0.017)

Host country controls

[Brazil.sub.k]               -0.053   (0.039)
[Russian.sub.k]              -0.084   (0.039) **
[Indian.sub.k]                0.049   (0.020) **
[China.sub.k]                 0.044   (0.019) **
[Singapore.sub.k]            -0.042   (0.023) *
[Taiwan.sub.k]               -0.002   (0.026)
Constant                      0.048   (0.049)
Rho constant                  0.092   (0.012) ***
Sigma constant                0.049   (0.002) ***
N                                     221

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                3.85 **
(one-side)

DV: Homckc                          Medium
                             technology-intensive
                                    sectors

                                     Model 4

                                   Research

                             Coef.    S.E.
Explanatory variables

[Homehost.sub.kz]             0.077   (0.04)
[Onlyhomce.sub.kz]            0.041   (0.00) ***
[Onlyhost.sub.kz]             0.095   (0.037) **

Home region controls

Interfirm                     0.000   (0.000) *
Public sector                -0.038   (0.018) **
Financial sector              0.007   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.002   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.024   (0.(120)
Patent [propensity.sub.I]     0.004   (0.016)
Patent [propensity.sub.C]    -0.068   (0.024) ***
Patent [propensity.sub.ME]   -0.060   (0.025) **
Population                   -0.012   (0.004) **
R&D [FDI.sub.k]              -0.015   (0.011)

Home country controls        -0.006   (0.010)
                             -0.025   (0.018)
Attractiveness                0.029   (0.017)
North America
EU

Host country controls         0.053   (0.047)
                              0.005   (0.034)
[Brazil.sub.k]               -0.016   (0.020)
[Russian.sub.k]               0.035   (0.016) **
[Indian.sub.k]               -0.029   (0.049)
[China.sub.k]
[Singapore.sub.k]             0.024   (0.050)
[Taiwan.sub.k]                0.082   (0.013) ***
Constant                      0.049   (0.002) ***
Rho constant                          221
Sigma constant
N

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                0.94
(one-side)

DV: Homckc                          Medium
                             technology-intensive
                                    sectors

                                    Model 5

                                  Development

                             Coef.    S.E.
Explanatory variables

[Homehost.sub.kz]             0.101   (0.031) ***
[Onlyhomce.sub.kz]            0.036   (0.009) ***
[Onlyhost.sub.kz]             0.013   (0.031)

Home region controls

Interfirm                     0.000   (0.000)
Public sector                -0.028   (0.018)
Financial sector              0.006   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.003   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.021   (0.020)
Patent [propensity.sub.I]     0.005   (0.016)
Patent [propensity.sub.C]    -0.063   (0_024) ***
Patent [propensity.sub.ME]   -0.059   (0.025) **
Population                   -0.011   (0.004) **
R&D [FDI.sub.k]              -0.020   (0.010) *

Home country controls        -0.009   (0.010)
                             -0.021   (0.018)
Attractiveness                0.027   (0.017)
North America
EU

Host country controls         0.056   (0.043)
                             -0.001   (0.027)
[Brazil.sub.k]               -0.017   (0.024)
[Russian.sub.k]               0.012   (0.024)
[Indian.sub.k]               -0.025   (0.035)
[China.sub.k]
[Singapore.sub.k]             0.020   (0.050)
[Taiwan.sub.k]                0.088   (0.013) ***
Constant                      0.050   (0.002) ***
Rho constant                          221
Sigma constant
N

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                2.66 **
(one-side)

DV: Homckc                          Medium
                             technology-intensive
                                    sectors

                                    Model 6

                                  Adaptation

                             Coef.    S.E.
Explanatory variables

[Homehost.sub.kz]             0.020   (0.034)
[Onlyhomce.sub.kz]            0.037   (0.009) ***
[Onlyhost.sub.kz]            -0.083   (0.034) **

Home region controls

Interfirm                     0.000   (0.000)
Public sector                -0.037   (0.018) **
Financial sector              0.006   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.003   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.025   (0.020)
Patent [propensity.sub.I]     0.004   (0.016)
Patent [propensity.sub.C]    -0.062   (0.024) ***
Patent [propensity.sub.ME]   -0.050   (0.024) **
Population                   -0.012   (0.004) **
R&D [FDI.sub.k]              -0.020   (0.011) *

Home country controls        -0.006   (0.010)
                             -0.023   (0.017)
Attractiveness                0.030   (0.017) *
North America
EU

Host country controls         0.063   (0.047)
                              0.004   (0.026)
[Brazil.sub.k]                0.007   (0.020)
[Russian.sub.k]               0.072   (0.025) ***
[Indian.sub.k]               -0.028   (0.040)
[China.sub.k]
[Singapore.sub.k]             0.019   (0.050)
[Taiwan.sub.k]                0.089   (0.013) ***
Constant                      0.049   (0.002) ***
Rho constant                          221
Sigma constant
N

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                3.67 **
(one-side)

                              Knowledge-intensive
                                    services

                                    Model 7

                                    Research

                             Coef.    S. E.

Explanatory variables

[Homehost.sub.kz]            -0.004   (0.043)
[Onlyhomce.sub.kz]            0.034   (0.009) ***
[Onlyhost.sub.kz]            -0.032   (0.029)

Home region controls

Interfirm                     0.000   (0.000)
Public sector                -0.026   (0.019)
Financial sector              0.006   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.002   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.010   (0.020)
Patent [propensity.sub.I]     0.000   (0.016)
Patent [propensity.sub.C]    -0.062   (0.024) **
Patent [propensity.sub.ME]   -0.070   (0.024) ***
Population                   -0.017   (0.005) ***
R&D [FDI.sub.k]              -0.021   (0.015)

Home country controls

Attractiveness               -0.008   (0.010)
North America                -0.027   (0.018)
EU                            0.029   (0.017)

Host country controls

[Brazil.sub.k]               -0.208   (0.119) *
[Russian.sub.k]              -0.032   (0.061)
[Indian.sub.k]                0.035   (0.019) *
[China.sub.k]                 0.039   (0.030)
[Singapore.sub.k]            -0.051   (0.037)
[Taiwan.sub.k]               -0.051   (0.043)
Constant                      0.064   (0.050)
Rho constant                  0.086   (0.012) ***
Sigma constant                0.049   (0.002) ***
N                                     221

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                0.02
(one-side)

                              Knowledge-intensive
                                    services

                                     Model 8

                                  Development

                             Coef.    S. E.

Explanatory variables

[Homehost.sub.kz]             0.028   (0.027)
[Onlyhomce.sub.kz]            0.035   (0.009) ***
[Onlyhost.sub.kz]            -0.015   (0.028)

Home region controls

Interfirm                     0.000   (0.000)
Public sector                -0.027   (0.019)
Financial sector              0.006   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.002   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.012   (0.020)
Patent [propensity.sub.I]     0.000   (0.016)
Patent [propensity.sub.C]    -0.063   (0.024) ***
Patent [propensity.sub.ME]   -0.066   (0.024) ***
Population                   -0.016   (0.005) ***
R&D [FDI.sub.k]              -0.018   (0.015)

Home country controls

Attractiveness               -0.008   (0.010)
North America                -0.027   (0.018)
EU                            0.028   (0.017)

Host country controls

[Brazil.sub.k]               -0.212   (0.115) *
[Russian.sub.k]              -0.028   (0.058)
[Indian.sub.k]                0.026   (0.020)
[China.sub.k]                 0.030   (0.029)
[Singapore.sub.k]            -0.051   (0.036)
[Taiwan.sub.k]               -0.040   (0.046)
Constant                      0.055   (0.051)
Rho constant                  0.087   (0.012) ***
Sigma constant                0.049   (0.002) ***
N                                     221

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                0.08
(one-side)

                              Knowledge-intensive
                                    services

                                     Model 9

                                  Adaptation

                             Coef.    S. E.

Explanatory variables

[Homehost.sub.kz]             0.118   (0.028) ***
[Onlyhomce.sub.kz]            0.032   (0.009) ***
[Onlyhost.sub.kz]             0.001   (0.041)

Home region controls

Interfirm                     0.000   (0.000)
Public sector                -0.023   (0.019)
Financial sector              0.006   (0.001) ***
Openness                     -0.001   (0.000) ***
University                    0.002   (0.002)
Human capital                 0.000   (0.000)
Business R&D                  0.000   (0.000) ***
Patent [propensity.sub.EE]    0.007   (0.020)
Patent [propensity.sub.I]     0.000   (0.016)
Patent [propensity.sub.C]    -0.064   (0.023) ***
Patent [propensity.sub.ME]   -0.071   (0.024) ***
Population                   -0.015   (0.004) ***
R&D [FDI.sub.k]              -0.015   (0.015)

Home country controls

Attractiveness               -0.009   (0.009)
North America                -0.026   (0.017)
EU                            0.027   (0.017)

Host country controls

[Brazil.sub.k]               -0.289   (0.120) **
[Russian.sub.k]               0.011   (0.066)
[Indian.sub.k]                0.002   (0.017)
[China.sub.k]                -0.005   (0.027)
[Singapore.sub.k]            -0.064   (0.036) *
[Taiwan.sub.k]               -0.019   (0.044)
Constant                      0.046   (0.050)
Rho constant                  0.088   (0.012) ***
Sigma constant                0.048   (0.002) ***
N                                     221

Complementarity test: [Homehost.sub.kz] > [Onlyhost.sub.kz] +
[Onlyhome.sub.kz]

[chi square] Wald test                3.39 **
(one-side)

*** p [less than or equal to] 0.01; ** p [less than or equal to]
0.05; * p [less than or equal to] 0.10


DOI 10.1007/s11575-012-0135-2

Received: 30.03.2010 / Revised: 23.05.2011 / Accepted: 07.06.2011 / Published online: 16.03.2012 [c] Gabler-Verlag 2012

Dr. L. M. D'Agostino ([mail]) * Prof. G. D. Santangelo

Facolta di Scienze Politiche, University of Catania, Catania, Italy

e-mail: ldagosti@unict.it

Prof. G. D. Santangelo

e-mail: grsanta@unict.it
COPYRIGHT 2012 Gabler Verlag
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:RESEARCH ARTICLE
Author:D'Agostino, Lorena M.; Santangelo, Grazia D.
Publication:Management International Review
Geographic Code:4EUIT
Date:Mar 1, 2012
Words:11539
Previous Article:Firm specific advantage in developed markets dynamic capability perspective.
Next Article:Knowledge mobility in cross-border buyer-supplier relationships.
Topics:

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters |