Paper to be presented at the ICSB 2011 World Conference on

“Back to the Future: Changes in Perspectives of Global Entrepreneurship and Innovation“ at Stockholm, Sweden, June 15 - 18, 2011

BRIDGING THE GLOBAL AND THE LOCAL? MULTINATIONAL ENTERPRISES, LABOR MARKET MOBILITY AND LOCALIZED LEARNING

Bernd Ebersberger Management Center Innsbruck, Innsbruck, Austria [email protected] Sverre J. Herstad NIFU STEP Oslo, Norway [email protected] Olavi Lehtoranta VTT Technical Research Centre of Finland Espoo, Finland [email protected]

Abstract The paper investigates how multinational enterprises link territorial innovation systems to external information, knowledge and networks. As the corporate networks of multinationals span different business contexts, their employees are exposed to richer knowledge flows and broader social networks than are the employees of uninational firms. Much of these resources follow the individual employee. Hence, the presence of multinational network nodes which are linked to the economy of location through the labor market may represent an important empirical manifestation of the “local buzz-global pipeline” metaphor. Using the Finnish Innovation Survey, and longitudinal employer-employee data (1995 to 2004) covering all employees in Finland, we find that labor mobility inflows from MNEs impact innovation performance, whereas mobility inflows from uninational firms do not. ____________ Prof. (FH) Dr. Bernd Ebersberger holds the professorship for Innovation Management and Economics at MCI, the Management Center Innsbruck, Internationale Hochschule GmbH, Innsbruck, Austria. Dr. Sverre J. Herstad is currently senior advisor at the Norwegian Ministry of Trade and Industry, Department of Research and Innovation policy. He holds a senior researcher position at NIFU Nordic Institute for Studies in Innovation, Research and Education, and a Dr. Polit (PhD) degree in Innovation Studies, Centre for technology, innovation and Culture, University of Oslo, and a Cand. Polit degree in Economic Geography. Mr. Olavi Lehtoranta is senior researcher at VTT Technical Research Centre of Finland, Innovation Studies. He holds a Licentiate degree in Economics, University of Helsinki.

1

Introduction There is a strong and growing interest in two fundamental aspects of localized learning and territorial system dynamics. The first aspect concerns the mechanisms by which knowledge diffuses within territorial systems, and it concerns the way the composition of the industrial structure determines the impact of such diffusion on innovation and growth (Frenken, Oort, &Verburg, 2007; Meyer &Sinani, 2009). The focus of the debate has distinctively shifted away from interactive learning, by means of collaboration, towards localized knowledge diffusion through the mobility of people in local labor markets (Malmberg& Power, 2005). In doing so, the debate draws inspiration from the classic dispute of Marshallian economies of specialization versus Jacobian economies of diversification (Beaudry&Schiffauerova, 2009; Frenken et al., 2007). The second aspect concerns the impact of different external corporate and value chain linkages on territorial systems (Sturgeon, 2003; Sturgeon, Biesebroeck, &Gereffi, 2008). The second aspect also concerns potential interdependencies between the internal dynamics of territorial systems and its openness to information and knowledge from outside (Bathelt, Malmberg, &Maskell, 2004; Boschma & Iammarino, 2009; Owen-Smith & Powell, 2004). The openness of regions to ideas and information from outside is more and more considered a key determinant of the regions’ ability to relate to an increasingly complex international technological landscape (Simmie, 2003, 2004). This is visible not least in the recent literature on gatekeepers, that is, actors which by means of their networking behavior (Giuliani, 2005) combine internal and external system linkages (Graf, 2010; Lazaric, Longhi, & Thomas, 2008; Morrison, 2008). The purpose of this paper is to investigate empirically the extent to which the affiliates of multinational enterprises serve as gatekeepers of external networks and knowledge. We 2

first discuss knowledge diffusion and learning in multinational corporate networks against the background of the localized learning concept from economic geography. We then discuss how these two spheres are linked. Based on this discussion, we develop five hypotheses about the relationship between labor market mobility, innovativeness, and the presence of multinational corporations in territorial systems. The subsequent empirical analysis uses a unique data set consisting of innovation and labor mobility data from a representative sample of 1,528 Finnish firms. Lastly, we discuss the findings against the background of conceptual debates and recent empirical studies (Bathelt, Malmberg, & Maskell, 2004; Giuliani & Bell, 2005; Henderson, 2007; Meyer & Sinani, 2009; Morrison, 2008; Meyer & Sinani, 2009; Balsvik, forthc.; Maliranta, Mohnen, &Rouvinen, 2009).

The multinational enterprise The MNE is commonly conceptualized as a differentiated network (Nohria & Ghoshal, 1997), that is, a complex set of relational ties by which linkages—direct and indirect, formal and informal—are created across diverse business contexts (Currah & Wrigley, 2004). Research has found that this multinationality affects firm performance (Bellak, 2004; Goerzen & Beamish, 2003) and is conducive to innovation at the enterprise level (Frenz, Girardone, & Ietto-Gillies, 2005; Frenz & Ietto-Gillies, 2007). This positive effect exists, because the parent network can internalize information and knowledge originating in one context or location and make it available to activities in others, more efficiently than uninational enterprises (Dachs, Ebersberger, & Loof, 2008; Ebersberger & Herstad, 2011). This transfer of knowledge and information is neither a trivial task, nor an inevitable outcome (Persaud, 2005). The distribution of activities across different territorial systems not only increases the overall exposure of MNEs to diverse networks, information and 3

knowledge. It also reinforces the challenges of building communicative capacity conducive to knowledge diffusion and application within the group (Forsgren, 1996; Ghoshal, Korine, & Szulanski, 1994; Hansen, 2002). The post-war MNE reduced this tensions by home-base exploiting internationalization strategies seeking market access or low-cost production platforms, or by developing formalized systems of control and monitoring (Bartlett & Ghoshal, 1998; Geppert, Williams, & Matten, 2003). These systems limited the need for communicative capacity by reducing the amount of information, which had to be processed by headquarters or the group network as a whole. Due to intensifying innovation-based competition and the shift towards knowledge seeking internationalization strategies, many MNEs now forced to build such capacity. They do so by focusing on the internal creation of place (Ebersberger & Herstad, 2011) through linking formal top down initiated integrative mechanisms and organizational structures to bottom-up initiatives (Coe & Wrigley, 2007). This entails the establishment of meeting places, of support functions and of cross-unit organizational labor markets. It also necessitates the development of a set of shared values, objectives and belief systems across MNE units (Björkman, Barner-Rasmussen, & Li, 2004). Through such initiatives, MNE managers attempt to nurture those inter-unit personal ties (Szulanski, 1996; 2003) that are critical to the flow of information (Gupta & Govindarajan, 2000; Ivarsson, 2002; Persaud, 2005) on a broader basis than what can result from administrative design and traded linkages within the group. Consequently, richer information is diffusing through broader networks within the modern MNE, than within uninational corporate groups (Ebersberger & Herstad, 2011; Frenz et al., 2005; Frenz & Ietto-Gillies, 2007).

4

Localized learning Economic geography has firmly established how regions serve as containing social structures, within which information sharing and collective knowledge development is nurtured by personal network formation, by labor market mobility (Agrawal, Cockburn, & McHale, 2006; Dahl & Pedersen, 2004) and by the formation of trustful collaborative ties (Helper, DacDuffie, & Sabel, 2000). Such localized linkages give rise to territorial specialization, and they tie innovativeness to the properties of places. The flip side of this coin is that actors might over-embedded in a specific region. It might also cause lock-ins to paths of diminishing returns, induced by a combination of high cost of establishing extra-regional linkages and of low marginal cost of continuing to use existing ones (Grabher, 1993; Narula, 2002). Additionally it might encourage actors to focus excessively on existing information sources (Katila&Ahuja, 2002) and known collaboration partners (Bathelt, Malmberg, & Maskell, 2004). Through their linkages to other regions the presence of MNE network nodes may expose territorial systems to richer inputs. Yet, while the literature on knowledge spillovers from MNE presence is vast, it remains inconclusive (Döring & Schnellenbach, 2006; Görg & Greenway, 2004; Henderson, 2007; Meyer & Sinani, 2009; Unctad, 2005). This is partly because much of the empirical literature attempts to measure the impact of spillovers, without relating this directly to the mechanism at play in diffusing and absorbing knowledge. This confuses the presence of (latent) spillovers, the mechanism by which knowledge may or may not diffuse, and the ability of local economies to induce and absorb these spillovers. With respect to the latter, the context itself, for instance the diverse information diffusion ecologies of different territorial systems, will mediate both the MNE behavior within it (Lowe & Wrigley, 2010) and the effects spillovers have on it (Lazaric et al., 2008; Meyer &Sinani, 5

2009). This context can be characterized in terms of overall levels of development (Meyer & Sinani, 2009) or in terms of industrial structure composition (Beaudry & Schiffauerova, 2009; Frenken et al., 2007). Knowledge diffusion mechanisms can be characterized in terms of the density of labor market mobility between firms (Boschma, Eriksson, & Lindgren, 2009; Eriksson & Lindgren, 2009; Maliranta, Mohnen, & Rouvinen, 2009), and in terms of the extent to which such indirect diffusion is reinforced by localized collaborative networks (Ebersberger & Herstad, 2011; Fritsch & Franke, 2004), institutionalized trust (Helper et al., 2000) and local conventional-relational assets (Storper, 1997). These are characteristics specific to different industrial contexts. Consider first localized innovation collaboration. Traditionally, theory has put a very strong emphasis on knowledge diffusion by means of networking (Fritsch, 2003; Fritsch & Franke, 2004; Henderson, 2007). The importance of collaborative relationships have been stressed by e.g. new growth theory (Krugman, 1991; Romer, 1990) and different innovation system approaches (Edquist, 1997; Lundvall, 1992). The work on regional innovation systems and industrial districts (Asheim, 1996; Asheim & Isaksen, 2002) in particular reflects the work of Piore & Sabel (1984) on networked production in Germany and Northern Italy. Against this background, the process of corporatization (Asheim & Herstad, 2005) of territorial systems through MNE activity is considered problematic because it will often force management and researcher attention (Ocasio, 1997) towards the corporate network, at the expense of attention towards external collaborative knowledge development (Blanc & Sierra, 1999; Phelps & Fuller, 2000). Initial post-acquisition tensions may be strong; and attention may subsequently be directed towards internal politics and bargaining between elite actors in different parts of the MNE (Currah & Wrigley, 2004). As this may contribute to ‘decoupling’ the MNE subsidiary from local collaboration networks, concern has been expressed

6

concerning the extent to which MNE activity contribute to fragmentation of local collaboration networks with subsequent loss of local industrial dynamics (Asheim & Herstad, 2005; Ebersberger & Herstad, 2011; Phelps & Fuller, 2000; Stewart, 1976).

Labor market mobility Yet, it is increasingly argued that we are witnessing a shift towards global production and innovation network linkages on a broader basis, by which territorial systems are deconstructing as sets of traded user-producer relationships because the advantages of proximity in collaborative work is outweighed by the availability of more diverse competences abroad (Coe, Dicken, & Hess, 2008; Cotic-Svetina, Jaklic, & Prodan, 2008). This suggests that the approaches to the understanding of MNE’s impacts which emphasize traded linkages become increasingly irrelevant regardless of whether they focus on effects in the home economy or in host economies. These approaches are less relevant for such traded linkages will—or will have to—internationalize irrespective of MNE activity to keep the territorial economy up to date and avoid systemic lock-in (Narula, 2002). Yet, while the benefits of traded linkages to firms can be harnessed with relative ease once partners have been identified (Adams, 2002; Sturgeon, 2003), the tendency of economic activity to cluster in space does not only remain unchanged but it appears to be reinforced. This suggests that there are informal knowledge diffusion mechanisms at play which not only resist the forces of globalization, but increase in importance with it (Maurseth & Verspagen, 2002; Verspagen & Schoenmakers, 2000). These are often referred to as the untraded interdependencies of regional economies (Storper, 1995). It is suggested that these interdependencies stem from knowledge diffusion and upgrading effects attributable to the flow of people and tacit knowledge between firms through localized personal network formation and labor market mobility, that is the local buzz 7

or local broadcasting capacity of territorial systems (Isaksen & Onsager, 2010; Maurseth & Verspagen, 2002; Storper & Venables, 2004; Trippl, Todtling, & Lengauer, 2009). This means that the question of linkages between localized learning processes and external environments provided by MNEs should be conceptualized first and foremost as a question of social network formation (Giuliani & Bell, 2005) and knowledge diffusion working through the labor markets, which exist around the local MNE network node. This approach to the impact of MNE presence on localized learning is supported by the growing stock of research which links labor market dynamics to firm level performance (Boschma et al., 2009; Eriksson & Lindgren, 2009; Eriksson, Lindgren, & Malmberg, 2008; Maliranta et al., 2009). Eriksson & Lindgren (2009) argue that labor market externalities derived via local job mobility produces more powerful effects for involved firms than do economies of co-location, diversity and scale. The mobility of researchers in Norway has been found to be an important mechanism for diffusion of technological knowledge (Møen, 2005). Similarly, the inflow of workers from R&D departments into non-R&D functions in other firms has been found to increase both productivity and profitability among the latter (Maliranta et al, 2009).

Productivity increases among non-MNE firms has also been

attributed to their inflow of employees from multinationals (Balsvik, forthc.). More generally, Almeida and Kogut (1999) argue that the mobility of researchers and engineers has driven knowledge diffusion in Silicon Valley. Onsager & Isaksen (2010) align with Malmberg & Power (2005) and argue that the core of localized learning processes is the informal and indirect knowledge diffusion mechanisms through personal networks and local buzz (Storper & Venables, 2004). Findings consistent with this are presented by Fallick, Fleischman, & Rebitzer (2006), while Boschma et al (2009) ad that local mobility enables firms to absorb

8

more diverse competences than does the mobility of personnel across larger geographical distances.

Hypotheses The mobility of individuals between different economic activities does not only lead to i) the diffusion of knowledge embodied in individuals through the mobility event in itself, but also to ii) the formation of social networks between the constituents of the labor market (Agrawal et al., 2006; Breschi&Lissoni, 2001; Stefano Breschi&Lissoni, 2003; Dahl & Pedersen, 2004). The impact of the latter may outlast the mobility event as such. Based on this recognition we develop five hypotheses. First, we assume that the inflow of new employees serve to trigger innovation activity in the focal firm (Eriksson & Lindgren, 2009). This holds because new employees enrich the network linkages the employing company has into the surrounding economy. This exposes the company to non-public information about technological opportunities, market threats and competitor moves. Such information should serve to trigger innovation activity: H1: The inflow of new employees increases the likelihood that the focal is innovation active. However, this effect can be assumed to be contingent on the direct exposure of the firm to diverse non-public information by other means, notably through own internationalization. Hiring firms who are themselves multinationals, are exposed to more diverse inputs and networks than firms which are uninational, independently of labor inflow (see e.g. Frenz, Girardone, & Ietto-Gillies, 2005; Frenz & Ietto-Gillies, 2007). This may moderate the impact of such inflow on the likelihood of innovation activity: H2: The inflow of new employees has a stronger impact on the likelihood of innovation activity if the focal firm is uninational, compared to focal firms who are multinationals. 9

The broadening of the search space of the hiring firm (Hansen, 1999; Katila, 2002; Katila & Ahuja, 2002) which follow from labor inflow in general can be assumed to be positively associated also with innovativeness (Laursen & Salter, 2006). Evolutionary theorists have argued that the more alternatives there are to select from, the better are the alternatives selected (Nelson & Winter, 1982). This highlights the importance of extending search beyond one’s own organizational boundaries (Rosenkopf & Nerkar, 2001), value chain linkages, and technological domains (Hargadon & Sutton, 1997). Because search is inherently uncertain (Fleming, 2001), much rest on the exposure of firms to information beyond existing supply chain and innovation collaboration linkages, e.g. through employee’s informal networks. Similarly, new employees come with competences which diversify the knowledge base of the hiring firm, and broaden its capacity to assimilate, transform and exploit new information and knowledge (Van den Bosch et al., 1999; Nooteboom, Van Haverbeke, Duysters, Gilsing, & van den Oord, 2007; Zahra & George, 2002). All this increases the likelihood that firms are successful in their innovation efforts once initiated: H3: The inflow of new employees increases the likelihood of innovation success. As described above, we assume that this works partly by means of creating network linkages which extend beyond the hiring firm, and continues to feed information into it well after the mobility event as such. The size of this effect will be contingent on the networks created by the new employee during her previous position of employment. As job mobility is most intense within regions, the impact of labor inflow can at the outset be considered limited by the technological specialization profile of the region as a whole. Employees who have been exposed to the internal corporate networks of the modern multinational enterprises are therefore likely to contribute broader networks and more specialized knowledge to the hiring firm than employees with a background in non-MNEs (see e.g. Balsvik, forthc. ). As

10

explained above, this holds because the modern MNE increasingly is forced to emphasize the internal creation of place which is conducive to knowledge diffusion between the different contexts it spans (Currah& Wrigley, 2004; Gupta &Govindarajan, 2000; Ivarsson, 2002; Lowe & Wrigley, 2010; Persaud, 2005). Consequently, MNE employees become embedded in social networks which cut across different territorial contexts, and accumulate unique knowledge which rarely can be fully privatized by the employer. As this unique knowledge and network linkages follow the individual employee through the labor market, we assume that a multinationality effect (Bellak, 2004) is at play in labor inflow: H4: The inflow of new employees from multinational enterprises has a stronger impact on the innovation success of the hiring firm, than the inflow from uninationals. The concept of absorptive capacity (Cohen & Levinthal, 1990) suggests that novel competences entering organizations are only mirrored in innovation success to the extent that the hiring firm is able to effectively assimilate, transform and exploit them. This is partly a question of the extent to which it already possesses knowledge which is complementary to that which enters from outside; i.e. different enough to yield novel insights but similar enough to enable mutual understanding (Nooteboom, 2000; Nooteboom, Van Haverbeke, Duysters, Gilsing, & van den Oord, 2007), and partly a question of how internal routines and organizational processes are able to exploit them (Lane & Lubatkin, 1998; Zahra & George, 2002). Similarly, the new informal network linkages created by new employees can only translate into innovation success to the extent that these internal routines and organizational processes are sensitive to the information they channel. These two challenges of competence absorption and search space broadening are of course reinforced when the content of knowledge sourced and network linkages established reflect contextual conditions in countries and regions which are not necessarily familiar to the sourcing firm. As this aspect

11

of absorptive capacity reflects prior experiences embedded in organizational routines, it develops cumulatively (Van den Bosch et al., 1999). The impact of inflow from multinationals may therefore be contingent on the internationalization status of the hiring firm, and thus its direct experience with relating to diverse inputs from diverse business contexts: H5: The positive impact of labor inflow from MNEs on innovation success is greater for multinational enterprises, than for uninational enterprises.

Data and methodology The empirical analysis is based on data which enable us to link information on the innovation success of Finnish companies to their previous inflow of personnel from different sources. Information on innovation success is gathered from the Finnish Innovation Survey. This together with R&D surveys and business register data serve as the source of most of the controls. The labor mobility inflow indicators variables are constructed from the Finnish Longitudinal Employer Employee Database.

Finnish Innovation Survey The Finnish Innovation Survey is a national implementation of the European Community Innovation Survey (CIS), which has been used for analysis in industrial organization (e.g. Cassiman & Veuglers, 2002, 2006; Almus & Czarnitzki, 2003; Cefis & Marsili, 2006; Czarnitzki et al, 2007), management studies (e.g. Laursen & Salter, 2004, 2006, Cassiman & Veugelers 2006) and economic geography (Asheim, Ebersberger, & Herstad, 2011; Ebersberger & Herstad, 2011; Simmie, 2003, 2004). The CIS2006 round was carried out in 2007, and covers innovation activities and innovation output in the three-year period from 2004 to 2006. It contains information about innovation activities and output in a 12

representative sample of 1,528 companies, which we use to construct the dependent variables and some of the control variables in the analysis.

Selection & Dependent variables The selection variable (INNO_ACT) captures whether or not the focal firm is innovation active. It takes on the value one if the firm has positive R&D expenditures, or if it during the reference period successfully introduced a new product or production process. It also takes on the value one if the firm carried out but abandoned innovation projects during the reference period. This definition of innovation activity follows the tradition of e.g. Tether (2002), Veugelers & Cassiman (2005), Cassiman & Veugelers (2006), and Rammer et al. (2009). The selection variable is used in the first stage of the regressions, in which we estimate the likelihood of innovation activity. The dependent variable INNO_MKT captures whether or not the company has brought a product innovation (a good or a service) to the market which is new not only to the company but also to this market. Market novelties have previously been used as a measure of innovation success in various studies, either as a dummy variable indicating the event of introducing the product, or as the share of sales subsequently generated by them (Belderbos, Carree, & Loshkin, 2004; Cassiman & Veugelers, 2006; Frenz & Ietto-Gillies, 2009; Laursen & Salter, 2006). We focus on novel product innovations because these provide the best indicator of successful linking between technological possibility and market opportunity (Danneels, 2002; Dougherty, 1992), i.e. the technological impact of labor inflow sought captured. The event of introducing a new product is chosen over the option of turnover generated. This is because the latter can be assumed more contingent on decisions made by the firm and its value chain partners after the innovation process is finished (e.g. concerning

13

marketing, logistics and production), and on conditions outside the control of the individual focal firm (e.g. competitor responses to new products introduced) (see Ebersberger & Herstad, 2010 for a discussion).

Selection instrument & Control variables In the selection stage, we draw on the information available on all firms to determine the likelihood of innovation activity. Incentives to engage in innovation activities are related to the size of the market on which the firm can commercialize the innovation and the capacity to engage in such activities are related to the size of the firm. International orientation and export activities positively influence the perceived market size and hence increase the firm’s incentive to innovate (Baldwin & Gu, 2004; Harris & Li, 2005). We capture international orientation of the firm by its export share (EXPSHR) and by a dichotomous variable which takes the value one if the most important market is stated to be outside Finland (INTMKT). Size is captured as the natural log of number of employees (LN_EMP). In addition we can identify if the companies are part of a corporate group. The multinationality of the group (MULTI) is included in the selection instrument, because it is known to influence innovation activity (Frenz & Ietto-Gillies, 2007). As different sector are characterized by different incentives to engage in innovation activities (Leiponen & Drejer, 2007; Malerba & Orsenigo, 1993; Marsili & Verspagen, 2002; Pavitt, 1984), and these contextually determined opportunities or constraints cannot be assumed captured by firm-level information available, we include nine sector dummy variables in both selection and outcome equations. In the subsequent outcome regressions, we use the richer information available to introduce a set of additional controls for behavioral characteristics of the focal firm. R&D intensity (RDINT) as the fraction of turnover spent on internal and external R&D activities

14

captures the effects of investments in systematic knowledge exploring efforts on product innovation (Cassiman &Veuglers, 2002; Ebersberger, Marsili, Reichstein, & Salter, 2010; Reichstein, Dahl, Ebersberger, & Jensen, 2010; Roper, Du, & Love, 2008), independent of size effects. This follows from the definition of R&D in the survey, which in accordance with the guidelines in the OECD ’Oslo Manual’ on the measurement of technological innovation (OECD, 2005).

As external innovation collaboration provides a mechanism by which

external competences and networks can be accessed without labor market mobility events, we control for whether or not the firm engage in such using a dummy variable (COLL). Firms which are affiliated with enterprise groups have direct access to information flows within the group network (Dachs, Ebersberger, & Lööf, 2008). They may form collaborative relationships towards other units which are part of a common administrative system, source R&D with less own management attention required and use other units as a platform for external sourcing or collaboration (Asheim, Ebersberger, & Herstad, 2011; Maskell, Pedersen, Petersen, & Dick-Nielsen, 2007). These factors may influence the impact of all external interface measures, and thus be associated with the dependent variable (see e.g. Ebersberger & Herstad, 2009; Frenz & Ietto-Gillies, 2007). Thus, the variable MULTI which captures MNE affiliation is included also in the base outcome regressions. Similarly, as the strength and breadth of firm’s internal competences is related to its size and will impact its innovation success, we also control for size in the outcome regressions by using the variable LN_EMP. By linking innovation survey data to data derived from the national business register we are able to construct a control variable which captures the average annual growth of the focal firm during the 10 years1 preceding the time covered by the innovation survey. This

1

For companies younger than 10 years we compute the average growth rate since their founding.

15

variable enables us to calculate labor inflow impacts at given levels of growth during the periods in which the inflow is occurring. Table 1 Sectoral break down of the sample Number of observations Innovation active comp. Sectors Mining & Quarrying Food, Beverages & Tob. Textiles, Leather, Wood Chemicals, Rubber & Pl Basic Metal Machinery & Equipment Electronics & Instrumen. Manufacturing nec. KI Services Total

Total 1,528 952 1,4 6,3 18,4 7,5 18,9 11,1 8,6 8,3 19,5 100%

Longitudinal Employer-Employee Data Labor mobility indicators are constructed on the basis of Finnish Longitudinal Employer–Employee Data. This data from this source has previously been used by Maliranta et al. (2009), while comparable data from other Scandinavian countries have been used by Eriksson & Lindgren (2009) (Swedish data) and Balsvik (forthc.) (Norwegian data). The database holds information about all employees in Finland, which are linked to their employing firm on a yearly basis. This means that mobility can be captured as year-toyear changes in employing firm. As each employing firms can be uniquely identified, additional firm level information can be derived from other sources, in our context notably the business register. This includes information about whether companies are part of a multinational group, or part of a uninational group or independent Finnish companies. In addition, this allows us to link the employer-employee data to the innovation survey data discussed above. Given the fact that we derive our measures on annual excerpt of register

16

data, mobility indicators are constructed based on all mobility events within the Finnish economy where the firm-employee relationship exceeds one year. Table summarizes the universe of firms, employees and mobility events.

17

Table 2 Employees and firms in the Longitudinal Employer-Employee Data (1995-2004)

Employees Mobile employees Firms … part of MNC … part of NAT … independent

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

1,040,436 17,639

1,064,236 41,195

1,154,589 16,197

1,320,914 21,414

1,364,240 37,458

1419,946 54,393

1,440,193 42,510

1,432,759 31,349

1,437,956 34,411

1,460,062 40,031

86,077 1,265 654 84,158

88,754 1,347 659 86,748

96,483 1,671 979 93,833

164,660 2,084 1,213 161,363

167,146 2,354 1,839 162,953

169,953 2,598 1,979 165,376

171,215 2,787 2,030 166,398

172,082 3,308 3,758 165,016

173,634 3,977 4,202 165,455

175,891 4,189 4,194 167,508

Note: MNC – multinational corporate group, NAT – national corporate group

18

We then trace the employees of all 1,528 firms in the innovation survey. As we are interested in knowledge diffusion in broad terms, we acknowledge that positive impacts are not necessarily limited to the turnover of defined R&D personnel or other key persons, as assumed in most recent research (Boschma et al., 2009; Power & Lundmark, 2004). Hence, we follow Balsvik (forthc.) in focusing on employee flows at all levels of education for all functions within the firm. For the years 1995 to 2004 we approximate the mobility (MOB) by calculating the number of newly hired employees who directly originate from Finnish firms, which are part of a multi-national corporate group (_MNC). We also calculate the number of employees coming directly from national corporate groups (_NAT). We aggregate the annual flows using a discount rate of 15%. Scaling these measures with the current size of the company eliminates size effects. These raw indicators are aggregated to the total mobility inflow (MOB_TOT), and to the total mobility inflow coming from MNCs and uninational corporate groups (MOB_MNC and MOB_NAT).

19

Table 3 Descriptive statistics of the sample able 1:

All companies Innov, active companies (INNO_ACT=1) N=1,528 N=952 Variable Mean Std. Dev. Min Max Mean Std. Dev. Min Max + INNO_MKT 0.298 0.458 0 1 0.479 0.500 0 1 INNO_ACT 0.623 0.485 0 1.000 MOB_TOT++ 0.071 0.217 0 2.000 0.102 0.251 0 2.000 MOB_MNC++ 0.061 0.201 0 2.000 0.087 0.233 0 2.000 ++ MOB_NAT 0.011 0.073 0 1.039 0.015 0.087 0 1.040 GRWTH 0.089 0.287 -3.848 4.335 0.105 0.306 -0.711 4.335 LN_EMP 4.014 1.365 2.303 10.081 4.344 1.450 2.303 10.081 INTMKT+ 0.557 0.423 0 1 0.549 0.416 0 1 EXPSHR 0.193 0.297 0.000 1.000 0.225 0.315 0.000 1.000 GROUP+ 0.508 0.500 0 1 0.574 0.495 0 1 MULTI+ 0.236 0.425 0 1 0.306 0.460 0 1 RDINT 0.031 0.112 0.000 1.000 COLL+ 0.664 0.473 0 1 Note: + indicates dummy variables. Descriptives are not reported for the 10 sector dummies, ++ measured 1995-2004

Methodology The decision to engage in innovation activity (INNO_ACT=1) determines whether or not information on the various innovation activities (COLL, RDINT) and outcomes are available from the questionnaire. This translates into a selection bias in the regressions (Greene, 2000; Heckman, 1979), which must be accounted for. Estimating the effect of inflow on innovation success based only on the sample of innovating companies would result in a biased estimate (Greene 2000). We therefore apply the Van de Ven and Van Praag (1981) version of the Heckman (1979) selection model, in which the dependent variable of the outcome equation is dichotomous and the outcome equation is modeled as a probit regression. The selection (INNO_ACT) and outcome (NEWMKT) stages are consequently estimated simultaneously (table 4). As the coefficient estimates of a probit regression are not informative about the size of the influence we report marginal effects of the impact of labor mobility inflows originating 20

from companies affiliated to a multinational or to a uninational group. To be able to test whether the effects depend on the multinationality (hypothesis 2 & 5) of the receiving firm we report the average marginal effects for the group of multinational companies in the sample and for the group of uninational companies in the sample.

Findings Table 4 reports the coefficient estimates of the base models. Notable control variable

estimates include the significant impacts of firm size, export share and innovation collaboration on the likelihood of innovation success. In Model I we observe that the overall inward mobility significantly influences the decision about innovation activities and the likelihood that a new product is introduced on the market. This is consistent with Hypothesis 1 and 3. In the subsequent models, we distinguish between uninational and multinational sources of labor inflow, and run the analysis for all firms (Model II) and SMEs only (Model III). Consistent findings are obtained in that both forms of mobility increase the likelihood of innovation activity, whereas it is only the inflow from multinationals that increases the likelihood of innovation success. This is consistent with Hypothesis 4.

21

Table 4 Regression results for all (N=1,528) and SMEs (N=1,292)

INNO_MKT=1 MOB_TOT

Model I.1

Model II.1

Model III.1

All Firms

All Firms

SMEs

b/se

b/se

b/se

0.361* 0.187

MOB_MNC

—

MOB_NAT

—

—

—

0.353*

0.438*

0.197

0.235

0.402

0.629

0.493

0.507

0.166*

0.184

MULTI

0.166* 0.099

0.099

0.115

GROWTH

0.044

0.043

0.241

0.145

0.146

0.186

0.139***

0.138***

0.040

0.046 0.398 0.318

0.047 0.399 0.319

0.063 0.624 0.392

EXPSHR

0.269**

0.270**

0.312**

COLL

0.133 0.338*** 0.108

0.134 0.339*** 0.110

0.156 0.290*** 0.101

_cons

-1.369***

-1.365***

-1.072***

LN_EMP RDINT

0.263

0.275

0.288

Model I.2

Model II.2

Model III.2

INNO_ACT=1

b/se

MOB_TOT

1.065***

b/se —

b/se —

0.214 MOB_MNC

—

1.032*** 1.646*** 0.239 0.307 MOB_NAT — 1.226** 1.209** 0.567 0.574 MULTI 0.272** 0.273*** 0.200* 0.107 0.107 0.118 GROWTH 0.350** 0.352** 0.358** 0.168 0.169 0.166 LN_EMP 0.312*** 0.312*** 0.250*** 0.034 0.034 0.044 EXPSHR 0.186 0.187 0.189 0.133 0.133 0.144 GROUP -0.146* -0.146* -0.125 0.080 0.080 0.082 LL -1515.000 -1514.912 -1293.179 Chi2 117.21*** 115.54*** 82.65*** Note: *** (**, *) indicate significance at the 1% (5%, 10%) level. Model x.1 and Model x.2 are estimated simultaneously. 10 sector dummies included in the regressions, not reported here.

22

In the next step we investigate whether the multinationality of the receiving firm affects the impact of the mobility flows. We calculate the marginal effects of the mobility inflow MOB_MNC and MOB_NAT for both multinational receiving firms and for uninational receiving firms. The effects on innovation activity are displayed in Table 5 and the effects on innovation success are displayed in Table 6. Let us first turn to the effects on innovation activity in Table 5. The Wald tests show that although both forms of inflow significantly affect the likelihood of a company taking up innovation activities, the effect is significantly stronger for uninational receiving firms. This holds for all firms, and for the subsample of SMEs only. It implies that direct exposure to diverse business contexts by means of own presence abroad reduces the impact of indirect exposure to such diversity by means of labor inflow, lending support to Hypothesis 2. Table 5 Marginal effects on innovation activity All firms MOB_MNC MOB_NAT

[1] Receiver: Multinational firm (N=362) me/se

[2] Receiver: Uninational firm (N=1,166) me/se

0.342***

0.249***

0.371***

0.079

0.059

0.085

0.406**

0.296**

0.441**

4.44**

0.138 [3] Receiver: Multinational SME (N=222) me/se

0.203 [4] Receiver: Uninational SME (N=1,070) me/se

[3]=[4]

0.594***

0.512***

0.611***

11.02***

0.109

0.096

0.112

0.436**

0.376**

0.449**

All Firms (N=1,528) me/se

0.187

SMEs

SME (N=1,292) me/se

MOB_MNC MOB_NAT

[1]=[2]

Chi2 15.22***

Chi2

3.68*

0.206 0.179 0.212 Note: Average marginal effects of the mobility inflow. Standard errors in italics. *** (**, *) indicate significance at the 1% (5%, 10%) level. +reports the Chi2 test statistic for testing the equality of the marginal effects.

23

The next question is whether or not the impact of the two forms of inflow on innovation success is contingent on the internationalization status of the receiving firm, as suggested by hypothesis 5. Table 6 shows that once the decision to engage in innovation activity has been taken, it is only the inflow from multinationals that affects the likelihood of innovation success. We find no significant differences in the effect between multinational and uninational receiving firms in the sample as a whole and in the subsample for SMEs. This suggests that both multinational receiving firms and uninational receiving firms are equally equipped to benefit from the positive effects labor inflow with a previous multinational experience have. Table 6 Marginal effects on innovation performance of innovation active firms All innovation active firms

All Firms (N=952) me/se

[1] Receiver: Multinational firm (N=291) me/se

[2] Receiver: Uninational firm (N=663) me/se

[1]=[2]

Chi2+

MOB_MNC

0.120* 0.064

0.133* 0.073

0.113* 0.060

1.74

MOB_NAT

0.136

0.152

0.129

0.53

0.165 Innovation active SMEs

SME (N=741) me/se

MOB_MNC MOB_NAT

0.185 [3] Receiver: Multinational SME (N=158) me/se

0.158 [4] Receiver: Uninational SME (N=583) me/se

0.140*

0.161*

0.134*

0.073

0.085

0.070

0.201

0.232

0.193

[3]=[4] Chi2+ 2.53 1.28

0.160 0.186 0.153 Note: Average marginal effects of the mobility inflow. Standard errors in italics. *** (**, *) indicate significance at the 1% (5%, 10%) level. +reports the Chi2 test statistic for testing the equality of the marginal effects.

Consistent with our basic hypothesis, we have find that the inflow of new employees increases the likelihood of innovation activity (Hypothesis 1), and the likelihood of innovation success (Hypothesis 3). Both reflect how mobility inflows broaden both the competence base of the hiring firm, and its external search space. While the impact of inflow on innovation 24

activity is present for both inflows from multinational and uninational sources; it is significantly stronger when the sourcing firm is itself a uninational. This is a strong support for Hypothesis 2, which suggests that exposure to diverse competences and networks from inflow matters less as a trigger for innovation activity when the firms are already present in foreign business contexts. Similarly, while the impact of inflow from multinational sources on innovation success is present for both uninational and multinational sourcing firms there is no significant difference between the receiving multinational firms and the receiving uninational firms. Hence we find no support for Hypothesis 5. We do not find that the ability to effectively assimilate, transform and exploit mobility inflow is contingent on the internal routines that develop cumulatively as a byproduct of the firms’ experience with diverse external environment (Van den Bosch et al., 1999). Yet, most importantly we have found unambiguous support for Hypothesis 4. The organizational context of the MNC exposes its employees to information, knowledge and networks which directly reflect its presence in numerous contexts – but its boundaries are porous, causing knowledge developed and networks formed within to leak out into its surroundings. This forms the basis for innovation success in other firms.

Conclusion To our knowledge, only one prior study has directly linked the outflow of labor from multinational firms to performance impacts at the inflow end. However, this (Balsvik, forthc.) and other recent studies of the link between labor market mobility and economic performance more broadly (Boschma et al., 2009; Eriksson & Lindgren, 2009) use accounting data to estimate productivity impacts, and focus on whether or not these impacts indicate spillover effects which are or are not internalized by the labor market. This entails that the empirical

25

literature has been less interested in the mechanism by which a knowledge inflow translates into productivity improvement, namely innovation. By contrast, we have linked labor market induced knowledge diffusion out of MNC to its impact on innovation activity and innovation success at the receiving end - irrespective of possible labor market internalization effects (Maliranta et al, 2009; Møen, 2005) and later productivity impacts. The inflow of new personnel with fresh experiences, insights and networks exert a significant impact on the likelihood that a firm decides to engage in innovation activities. This holds irrespectively of whether or not new employees originate in national or multinational firms. However, once we control for the selection bias created and consider the impact on various measures of innovation success, we find that it is only the inflow of employees with MNC background which has a positive impact on innovation success. Thus, we have contributed novel insights in the role of multinationals in localized learning: As it is well established that mobility is most intense within regional labor markets, the lack of impact from inflow out of uninational firms point to the real danger of lock-in at the territorial economy level if no multinationals are present. In addition, we have also contributed to the theory of organizational absorptive capacity by finding the impact of inflow into SMEs to be contingent

on

the

experiences

gained

and

routines

developed

through

own

internationalization. Our analysis has notable limitations. First, we know from previous studies that mobility patterns per se as well as their subsequent impact are sensitive to the composition of the industrial structure, and the nature of the competences which diffuse. As this is not accounted for by our analysis, future work should carefully disentangle the effects of multinationality found by us and others (Balsvik, forthc. ) from the effects of overall industry

26

(Beaudry & Schiffauerova, 2009; Frenken et al., 2007) composition on mobility patterns per se and the effects of inflow composition on subsequent performance (Boschma et al., 2009). Second, we have built on previous research, which suggests that it is not the nationality of the multinational group which is present, but multinationality as such, which matter for performance (Bellak, 2004). We therefore stress that our findings should not be interpreted as pointing to the importance of foreign firm presence. Rather on the contrary, we see obvious reasons to expect that there could be important headquarter location effects at play in determining impacts at the level of surrounding economies (Asheim, Ebersberger, & Herstad, 2011; Ebersberger & Herstad, 2011; van Pottelsberghe de la Potterie & Lichtenberg, 2001), which point to the importance of multinationality by means of territorial system FDIbased expansion abroad. While MNCs attempt to create internal network mechanisms which transcend geographical space, new ‘internal’ geographies of closeness and distance may also be created and result in asymmetric exposure to information and knowledge floating within the network as a whole.

Headquarters (Gupta & Govindarajan, 2000), ‘back region’

business process support and development functions (Currah & Wrigley, 2004) and corporate R&D laboratories (Benito, Larimo, Narula, & Pedersen, 2002), can be assumed to be exposed to richer, and more accurate, information flows than those units which are primarily exposed through formal design or several layers of personal ties (Hansen, 2002). Last, the transfer of managerial and technical talent across the MNC network (Coe & Wrigley, 2007) may be asymmetric, and anchored in HQ and other home-base functions. If so, this causes informal networks to converge on HQ and – by means of mobility - favor the immediate context around it. Future studies should therefore pay particular attention to the issue of how the mode of FDI-based territorial innovation system internationalization impact internal system dynamics.

27

References Adams, J. D. (2002). Comparative localization of academic and industrial spillovers. Journal of Economic Geography, 2, 253-278. Agrawal, A., Cockburn, I., & McHale, J. (2006). Gone but not forgotten: knowledge flows, labor mobility, and enduring social relationships. Journal of Economic Geography, 6(5), 571-591. Almeida, P., & Kogut, B. (1999). Localization of Knowledge and the Mobility of Engineers in Regional Networks. Management Science, 45(7), 905-917. Almus, M., & Czarnitzki, D. (2003). The effects of public R&D subsidies on firms’ innovation activities: The case of Eastern Germany. Journal of Business and Economic Statistics, 21(2), 226-236. Asheim, B. T. (1996). Industrial districts as 'learning regions': a condition for prosperity. European Planning Studies, 4, 379-400. Asheim, B. T., & Herstad, S. (2005). Regional innovation systems, varieties of capitalisms and non-local relations: Challenges from the globalising economy. In R. A. Boschma & R. C. Kloosterman (Eds.), Learning from Clusters: A critical Asessment for an Economic-Geographical Perspective. Dordrecth: Springer. Asheim, B.T.,& Isaksen, A. (2002). Regional Innovation Systems: The Integration of Local ‘Sticky’ and Global ‘Ubiquitous’ Knowledge. Journal of Technology Transfer, 27, 77-86. Asheim, B. T., Ebersberger, B., & Herstad, S. (2011). MNCs between the global and the local: Knowledge bases, proximity and globally distributed knowledge networks. In M. Heidenreich (Ed.), Innovation and Institutional Embeddedness of Multinational Companies. Cheltenham: Edward Elgar (in print). Baldwin, J. R., & Gu, W. L. (2004). Trade liberalization: Export-market participation, productivity growth, and innovation. Oxford Review of Economic Policy, 20(3), 372-392. Balsvik, R. (forthc. ). Is labor mobility a channel for spillovers from multinationals? Evidence from Norwegian manufacturing. Review of Economics and Statistics Bartlett, C., & Ghoshal, S. (1998). Managing across borders: The transnational solution. London: Random House. Bathelt, H., Malmberg, A., & Maskell, P. (2004). Clusters and Knowledge: Local Buzz, Global Pipelines and the Process of Knowledge Creation. Progress in Human Geography, 28(1), 31-56. Beaudry, C., & Schiffauerova, A. (2009). Who's right, Marshall or Jacobs? The localization versus urbanization debate. Research Policy, 38(2), 318-337. Belderbos, R., Carree, M., & Loshkin, B. (2004). Cooperative R&D and firm performance. Research Policy, 33, 1477-1492. Bellak, C. (2004). How domestic and foreign firms differ and why does it matter? Journal of Economic Surveys, 18(4), 483-514. Björkman, I., Barner-Rasmussen, W., & Li, L. (2004). Managing knowledge transfer in MNCs: The impact of headquarter control mechanisms. Journal of International Business Studies, 35, 443-455. Blanc, H., & Sierra, C. (1999). The internationalisation of R&D by multinationals: A trade-off between external and internal proximity. Cambridge Journal of Economics, 23, 187-206. Boschma, R., & Iammarino, S. (2009). Related Variety, Trade Linkages, and Regional Growth in Italy. Economic Geography, 85(3), 289-311.

28

Boschma, R., Eriksson, R., & Lindgren, U. (2009). How does labour mobility affect the performance of plants? The importance of relatedness and geographical proximity. Journal of Economic Geography, 9, pp. 169–190. Breschi, S., & Lissoni, F. (2001). Knowledge spillovers and local innovation systems: A critical survey. Industrial and Corporate Change, 10(4), 975-1005. Breschi, S., & Lissoni, F. (2003). Mobility and Social Networks: Localised Knowledge Spillovers Revisisted. CESPRI working paper, 142. Brown, J. S., & Duguid, P. (1996). Organizational learning and communities-of-practice: Towards a unified theory of working, learning and innovation. In M. D. Cohen & Sproul (Eds.), Organizational learning. New York: Sage Publications. Cassiman, B., & Veugelers, R. (2002). Cooperation and spillovers: Some empirical evidence from Belgium. American Economic Review, 9, 1169-1184. Cassiman, B., & Veugelers, R. (2006). In search of complementarity in innovation strategy: Internal R&D and external knowledge aquisition. Management Science, 52(1), 68-82. Cefis, E., & Marsili, O. (2006). Survivor: The role of innovation in firms' survival. Research Policy, 35(626641). Coe, N. M., & Wrigley, N. (2007). Host economy impacts of transnational retail: The research agenda. Journal of Economic Geography, 2007(7), 341-371. Coe, N. M., Dicken, P., & Hess, M. (2008). Global production networks: Realizing the potential. Journal of Economic Geography, 8, 271-295. Cohen, W. M., & Levinthal, D. A. (1990). Absorptive capacity - a new perspective on learning and innovation. Administrative Science Quarterly, 35(1), 128-152. Cotic-Svetina, A., Jaklic, M., & Prodan, I. (2008). Does collective learning in clusters contribute to innovation? Science and Public Policy 35(5), 335-345 Currah, A., & Wrigley, N. (2004). Networks of organizational learning and adaption in retail TNCs. Global Networks, 4(1), 1-23. Czarnitzki, D., Ebersberger, B., & Fier, A. (2007). The relationship between R&D collaboration, subsidies and R&D performance: Empirical evidence from Finland and Germany. Journal of Applied Econometrics, 22, 1347-1366. Czarnitzki, D., Hanel, P., & Rosa, J. M. (2011). Evaluating the impact of R&D tax credits on innovation: A microeconometric study on Canadian firms. Research Policy, 40(2), 217-229. Dachs, B., Ebersberger, B., & Loof, H. (2008). The innovative performance of foreign-owned enterprises in small open economies. Journal of Technology Transfer, 33(4), 393-406. Dahl, M. S., & Pedersen, C. Ø. R. (2004). Knowledge flows through informal contacts in industrial clusters: myth or reality? Research Policy, 33(10), 1673-1686. Danneels, E. (2002). The dynamic of product innovation and firm competencies. Strategic Management Journal, 23, 1095-1121. Dougherty, D. (1992). A practice-centered modell of organizational renewal through product innovation. Strategic Management Journal, 13, 77-92.

29

Döring, T., & Schnellenbach, J. (2006). What do we know about geographical knowledge spillovers and regional growth? A survey of the literature. Regional Studies, 40(3), 375-395. Ebersberger, B., & Herstad, S. (2010). Product innovation and the complementarities of external interfaces. Knowledge in Organizations Conference, May 30th - June 3rd, Monte Verita, Switzerland. Ebersberger, B., & Herstad, S. (2011). Go abroad or have strangers visist? On organizational search spaces and local linkages. Journal of Economic Geography, doi: 10.1093/jeg/lbq057 Ebersberger, B., Marsili, O., Reichstein, T., & Salter, A. (2010). Into thin air - using a quantile regression approach to explore the relationship between R&D and innovation. Journal of Applied Economics, 24(1), 95-102. Edquist, C. (1997). Systems of innovation : technologies, institutions and organizations. London: Pinter. Eriksson, R., & Lindgren, U. (2009). Localized mobility clusters: impacts of labour market externalities on firm performance. Journal of Economic Geography, 9(1), 33-53. Eriksson, R., Lindgren, U., & Malmberg, G. (2008). Agglomeration mobility: effects of localisation, urbanisation, and scale on job changes. Environment and Planning A, 40(10), 2419-2434. Fallick, B., Fleischman, C. A., & Rebitzer, J. B. (2006). Job-hopping in Silicon Valley: Some evidence concerning the microfoundations of a high-technology cluster. Review of Economics and Statistics, 88(3), 472-481. Fleming, L. (2001). Recombinant uncertainty in technological search. Management Science, 47(1), 117-132. Forsgren, M. (1996). The advantage paradox of the multinational corporation. In I. Björkman & M. Forsgren (Eds.), The Nature of the International Firm (pp. 69-85). Copenhagen: Copenhagen Business School Press. Frenken, K., Oort, F. V., & Verburg, T. (2007). Related Variety, Unrelated Variety and Regional Economic Growth. Regional Studies, 41(5), 685-697. Frenz, M., Girardone, C., & Ietto-Gillies, G. (2005). Multinationality matters in innovation: The case of the UK financial services. Industry & Innovation, 12(1), 65-92. Frenz, M., & Ietto-Gillies, G. (2007). Does multinationality affect the propensity to innovative? An analysis of the Third Community Innovation Survey International Review of Applied Economics, 21(1), 99-117. Frenz, M., & Ietto-Gillies, G. (2009). The impact on innovation performance of different sources of knowledge: Evidence from the UK Community Innovation Survey. Research Policy, 38(7), 1125-1135. Fritsch, M. (2003). Does R&D Cooperation Behavior Differ between Regions? Industry & Innovation, 10(1), 25 - 39. Fritsch, M., & Franke, G. (2004). Innovation, regional knowledge spillovers and R&D cooperation. Research Policy, 33(2), 245-255. Geppert, M., Williams, K., & Matten, D. (2003). The social construction of contextual rationalities in MNCs: An Anglo-German comparision of subsidiary choice. Journal of Management Studies, 40, 617-640. Ghoshal, S., Korine, H., & Szulanski, G. (1994). Interunit communication in multinational corporations. Management Science, 40(1), 96-110. Goerzen, A., & Beamish, P. W. (2003). Geographic scope and multinational enterprise performance. Strategic Management Journal, 24, 1289-1306.

30

Görg, H., & Greenway, D. (2004). Much ado about nothing? Do domestic firms really benefit from foreign direct investment? The World Bank Research Observer, 19(2), 171-197. Grabher, G. (1993). The weakness of strong ties. The lock-in of regional development in the Ruhr area. In G. Grabher (Ed.), The embedded firm: on the socio-economics of industrial networks. London: Routledge. Graf, H. (2010). Gatekeepers in regional networks of innovators. Camb. J. Econ. Greene, W. H. (2000). Econometric Analysis (4 ed.). Upper Saddle River: Prentice Hall Griliches, Z. (1979). Issues in assessing the contribution of research and development to productivity growth. Bell Journal of Economics 10, 92–116. Giuliani, E., & Bell, M. (2005). The micro-determinants of meso-level learning and innovation: evidence from a Chilean wine cluster. Research Policy, 34, 47-68. Gupta, A. K., & Govindarajan, V. (2000). Knowledge flows within multinational corporations. Strategic Management Journal, 21, 473-496. Harris, R. I. D., & Li, Q. C. (2005). Establishment Level Empirical Study Of Links Between Exporting, Innovation, and Productivity. Report to UKTI. Hansen, M. T. (1999). The search-transfer problem: The role of weak ties in sharing knowledge across organizational subunits. Administrative Science Quarterly, 44(1), 82-111. Hansen, M. T. (2002). Knowledge networks: Explaining effective knowledge sharing in multiunit companies. Organisation Science, 13, 232-248. Hargadon, A., & Sutton, R. I. (1997). Technology brokering and innovation in a product development firm. Administrative Science Quarterly, 42(4), pp.716-749. Heckman, J. (1979). Sample selection bias as a specification error. Econometrica, 47 (1): 153–61. Helper, S., DacDuffie, J. P., & Sabel, C. (2000). Pragmatic Collaborations: Advancing knowledge while controlling opportunism. Industrial and Corporate Change, 9(3), 443-487. Henderson, J. V. (2007). Understanding knowledge spillovers. Regional Science and Urban Economics, 37, 497508. Herstad, S. J., Bloch, C. W., Ebersberger, B., & Van De Velde, E. (2008). Open Innovation and Globalization; Theory, Evidence and Implications. Helsinki. Retrieved from Hipp, C., & Grupp, H. (2005). Innovation in the service sector: The demand for service-specific innovation measurement concepts and typologies. Research Policy, 34(4), 517–535. Isaksen, A., & Onsager, K. (2010). Regions, networks and innovation performance: The case of knowledgeintensive industries in Norway. European Urban and Regional Studies, 17(3), 227-243. Ivarsson, I. (2002). Transnational corporations and the geographical transfer of localised technology: a multiindustry study of foreign affiliates in Sweden. Journal of Economic Geography, 2(2), 221-247. Ivarsson, I. (2002). Transnational corporations and the geographical transfer of localised technology: A multiindustry study of foreign affiliates in Sweden. Journal of Economic Geography, 2, 221-247. Jaffe, A. (1986). Technological opportunity and spillovers of R&D: evidence from firms’ patents, profits and market value. American Economic Review 76, 984–1001.

31

Katila, R. (2002). New Product Search over Time: Past Ideas in Their Prime? The Academy of Management Journal, 45(5), 995-1010. Katila, R., & Ahuja, G. (2002). Something Old, Something New: A Longitudinal Study of Search Behavior and New Product Introduction. The Academy of Management Journal, 45(6), 1183-1194. Krugman, P. (1991). Geography and Trade. Cambridge, MA: MIT Press. Lane, P. J., & Lubatkin, M. (1998). Relative absorptive capacity and interorganizational learning. Strategic Management Journal, 19(5), 461-477. Laursen, K., & Salter, A. (2006). Open for innovation: The role of openness in explaining innovation performance among UK manufacturing firms. Strategic Management Journal, 24, 131-150. Lazaric, N., Longhi, C., & Thomas, C. (2008). Gatekeepers of knowledge versus platforms of knowledge: From potential to realized absorptive capacity. Regional Studies, 42(6), 837-852. Leiponen, A., & Drejer, I. (2007). What exactly are technological regimes? Intra-industry heterogeneity in the organization of innovation activities. Research Policy, 36, 1221-1238. Lowe, M., & Wrigley, N. (2010). The "Continously Morphing" Retail TNC During Market Entry: Interpreting Tesco's Expansion into the United States. Economic Geography, 86(4), 381-408. Lundvall, B.-Å. (1992). National Systems of Innovation: towards a theory of innovation and interactive learning. London: Pinter. Maliranta, M., Mohnen, P., & Rouvinen, P. (2009). Is inter-firm labor mobility a channel of knowledge spillovers ? Evidence from a linked employer-employee panel. Industrial and Corporate Change, 18(6), 1161-1191. Malmberg, A., & Power, D. (2005). (How) Do (Firms in) Clusters Create Knowledge? Industry and Innovation, 12(4), 409-431. Malerba, F., & Orsenigo, L. (1993). Technological regimes and firm behaviour. Industrial and Corporate Change, 2(1), 45-71. Marsili, O., & Verspagen, B. (2002). Technology and the dynamics of industrial structures: An empirical mapping of Dutch manufacturing. Industrial and Corporate Change, 11(4), 791-815. Maskell, P., Pedersen, T., Petersen, B., & Dick-Nielsen, J. (2007). Learning Paths to Offshore Outsourcing: From Cost Reduction to Knowledge Seeking. Industry and Innovation, 14(3), 239-257 Maurseth, P. B., & Verspagen, B. (2002). Knowledge spillovers in Europe: A patent citations analysis. Scandinavian Journal of Economics, 104(4), 531-545. Meyer, K., & Sinani, E. (2009). Where and When Does Foreign Direct Investment Generate Positive Spillovers? A Meta Analysis. Journal of international Business Studies, 40(7), 1075-1094. Miranda, A., & Rabe-Hesketh, S. (2006). Maximum likelihood estimation of endogenous switching and sample selection models for binary, count, and ordinal variables. Stata Journal, 6(3), 285-308. Møen, J. (2005). Is Mobility of Technical Personnel a Source of R&D Spillovers? Journal of Labor Economics, 23(1), 81-114. Morrison, A. (2008). Gatekeepers of knowledge within industrial districts: Who they are, how they interact. Regional Studies, 42(6), 817-835.

32

Narula, R. (2002). Innovation systems and 'intertia' in R&D location: Norwegian firms and the role of systemic lock-in. Research Policy, 31(5), 795-816. Nelson, R., & Winter, S. (1982). An evolutionary theory of economic change. Cambridge, MA: Harvard University Press. Nelson, R.R. (1987). Understanding Technical Change as an Evolutionary Process. North-Holland, Amsterdam. Nohria, N., & Ghoshal, S. (1997). The differentiated network: Organising multinational corporations for value creation. San Fransisco: Jossey-Bass. Nooteboom, B., Van Haverbeke, W., Duysters, G., Gilsing, V., & van den Oord, A. (2007). Optimal cognitive distance and absorptive capacity. Research Policy, 36(7), 1016-1034. Ocasio, W. (1997). Towards an attention-based view of the firm. Strategic Management Journal, 18, 187-206. OECD. (2005). Oslo Manual: Guidelines for collecting and interpreting innovation data. Paris: OECD & Eurostat. Owen-Smith, J., & Powell, W. W. (2004). Knowledge networks as channels and conduits: The effects of spillovers in the Boston biotechnology community. Organization Science, 15(1), 5-21. Pavitt, K. (1984). Sectoral patterns of technical change - towards a taxonomy and a theory. Research Policy, 13(6), 343-373. Persaud, A. (2005). Enhancing synergistic innovative capabilitie in multinational corporations: An empirical investigation. Journal of product innovation management, 22, 412-429. Phelps, N. A., & Fuller, C. (2000). Multinationals, Intracorporate Competition, and Regional Development*. Economic Geography, 76(3), 224-243. Ramani, S., Elaroui, M., & Carrere, M. (2008). On estimating a knowledge production function at the firm and sector level using patent statistics. Research Policy, 37(9), 1568-1578. Rammer, C., Czarnitzki, D., & Spielkamp, A. (2009). Innovation success of non-R&D-performers: substituting technology by management in SMEs. Small Business Economics, 33, 35-58. Reichstein, T., Dahl, M. S., Ebersberger, B., & Jensen, M. B. (2010). The devil dwells in the tails. Journal of Evolutionary Economics, 20(2), 219-231. Romer, P. M. (1990). Endogenous Technological Change. The Journal of Political Economy, 98(5, Part 2), 71102. Roper, S., Du, J., & Love, J. H. (2008). Modelling the innovation value chain. Research Policy, 37(6-7), 961977. Rosenkopf, L., & Nerkar, A. (2001). Beyond local search: Boundary-spanning, exploration, and impact in the optical disk industry. Strategic Management Journal, 22(4), 287-306. Simmie, J. (2003). Innovation and urban regions as national and international nodes for the transfer and sharing of knowledge. Regional Studies, 37(6-7), 607-620. Simmie, J. (2004). Innovation and clustering in the globalised international economy. Urban Studies, 41(5-6), 1095-1112. Stewart, J. C. (1976). Linkages and foreign direct investment. Regional Studies, 10(245-258).

33

Storper, M. (1995). The resurgence of regional economies, 10 years later: the region as a nexus of untraded interdependencies. European Urban and Regional Studies, 2, 191 - 221. Storper, M., & Venables, A. J. (2004). Buzz: face-to-face contact and the urban economy. Journal of Economic Geography, 4(4), 351-370. Sturgeon, T. J. (2003). What really goes on in Silicon Valley? Spatial Clustering and dispersal in modular production networks. Journal of Economic Geography, 3, 199-225. Sturgeon, T. J., Biesebroeck, J. V., & Gereffi, G. (2008). Value chains, networks and clusters: Reframing the global automotive industry. Journal of Economic Geography, 8, 297-321. Szulanski, G. (1996). Exploring internal stickiness: Impediments to the transfer of best practice within the firm. Strategic Management Journal, 17, 27-43. Szulanski, G. (2003). Sticky knowledge: Barriers to knowing in the firm: Sage Publications. Tether, B. S. (2002). Who co-operates for innovation, and why - an empirical analysis. Research Policy, 31(6), 947–967. Trippl, M., Todtling, F., & Lengauer, L. (2009). Knowledge Sourcing Beyond Buzz and Pipelines: Evidence from the Vienna Software Sector. Economic Geography, 85(4), 443-462. Unctad. (2005). World investment report 2005: Transnational corporations and the internationalization of R&D. New York: United Nations. Van den Bosch, F. A. J., Volberda, H. W., & De Boer, M. (1999). Coevolution of Firm Absorptive Capacity and and Forms Environment : Organizational Combinative Capabilities. Organization Science, 10(5), 551-568. Van De Ven, W. P., & Van Praag, B. M. (1981). The demand for deductibles in private health insurance - A probit model with sample selection. Journal of Econometrics, 17(1), 229-252. van Pottelsberghe de la Potterie, B., & Lichtenberg, F. (2001). Does foreign direct investment transfer technology across borders? The review of Economics and Statistics, 83(3), 490-497. Verspagen, B., & Schoenmakers, W. (2000). The spatial dimension of patenting by multinational firms in Europe. Journal of Economic Geography, 4(1), 23-42. Veugelers, R., & Cassiman, B. (2005). R&D cooperation between firms and universities. Some empirical evidence from Belgian manufacturing. International Journal of Industrial Organization, 23, 355 - 379. Zahra, S. A., & George, G. (2002). Absorptive capacity: A review, reconceptualization and extension. Academy of Management Review, 27(2), 185-203.

34