R&D Internationalization and Knowledge
Development: Evidence from China
Master’s Thesis Final Version
Anna Hrisztov 10993126
Date of Submission: 29 January 2016
Program: MSc. in Business Administration
Track: International Management
First supervisor: Dr Lori DiVito
Second supervisor: Dr Ilir Haxhi
S
TATEMENT OF ORIGINALITY
This document is written by Student Anna Hrisztov who declares to
take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is
original and that no sources other than those mentioned in the text and
its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the
supervision of completion of the work, not for the contents.
A
BSTRACTMultinational enterprises (MNEs) traditionally engaged in internationalization of their low-value activities, however since the 1970s high-low-value activities, such as research and development (R&D), are also increasingly offshored. Changes in MNEs’ motives and market trends have led to the emergence of new destinations too i.e. MNEs shift their focus from developed economies to emerging ones. Consequently, academic literature also pays more attention on operating R&D overseas. Yet still little is known about its host country implications. The current study aims to add to the academic literature in this regard by examining how foreign R&D, measured by patenting activities, affects knowledge development in China. To test the hypotheses a qualitative regression analysis was conducted based on secondary data provided by the World Intellectual Property Organization (WIPO) database and the China Statistical Yearbook in the period between 2005 and 2012. The results reveal that foreign R&D does not provide an incentive for local population to get involved in scientific education, but is has a positive effect on the awareness of locally created scientific knowledge and increases domestic innovative performance. The study calls attention to the need to develop IPR protection and the Chinese education system, to which MNEs may also contribute. Decision makers also have to deal with the increasing domestic competition in China that emerges due to foreign R&D.
T
ABLE OFC
ONTENTS1.INTRODUCTION ... 4
2.CONCEPTUAL BACKGROUND AND HYPOTHESIS DEVELOPMENT ... 6
2.1 Motives and performance implications of R&D internationalization ... 6
2.3 Entry mode decisions ... 11
2.4 Location choice for R&D activities ... 15
2.4.1 Foreign R&D in China ... 17
2.4.2 Attractiveness of China ... 18
2.5 Effects of R&D internationalization ... 20
2.5.1 Home country effects ... 21
2.5.2 Host country effects ... 23
2.5.3 Theoretical framework and hypotheses ... 27
3.DATA AND METHODOLOGY ... 29
3.1 Research design ... 29 3.2 Variables ... 30 3.3 Data analysis ... 35 4.RESULTS ... 35 5.DISCUSSION ... 39 5.1 Implications ... 40
5.1.1 Implications for academics ... 40
5.1.2 Implications for practitioners ... 41
5.1.3 Implications for policy makers ... 42
6.CONCLUSION AND LIMITATIONS ... 43
6.1 Conclusion ... 43
6.2 Limitations and future research ... 44
1.INTRODUCTION
International business refers to the practice when companies engage in cross-border economic activities (Rugman et al., 2011). Initially, multinational companies internationalize the comparatively low-value activities of their value chain. Evidence shows however, that globalization of high-value activities such as research and development, is gaining popularity since the 1970s as well (Dunning and Narula, 1995).
“Research and experimental development (R&D) comprise creative work undertaken
on a systematic basis in order to increase the stock of knowledge, including knowledge of man, culture and society, and the use of this stock of knowledge to devise new applications”
(OECD, 2002, p. 30). R&D is not the only component of innovation, but it is the most developed statistical indicator of industrial innovation activities (UNCTAD, 2005). It covers (1) basic research, which aims to acquire new knowledge without particular application plan, (2) applied research, which is also about gaining new knowledge, but directed towards a specific goal and (3) experimental development that includes activities in order to produce new materials, products and systems or improve already existing ones (OECD, 2002).
While R&D internationalization started in the 1970s, it became a popular phenomenon only in the 1980s (Cantwell, 1995). Later on since the 1990s R&D decentralization has been evolving to a strategic decision in order to enhance innovative capability of the firm (Chiesa, 1995). While Narula (2001) argues that firms still prefer to locate their R&D activities in their home countries, Meyer and Mizushima (1989) conclude that R&D globalization became indispensable as a result of the localization of competition, the shortened product life and the need for proximity to technological know-how.
Originally, multinational enterprises (MNEs) were more likely to offshore R&D to developed countries, but emerging economies, especially China started to attract foreign R&D too. The main incentives are the rapidly growing market, availability of low-cost but
high-quality labour and the favourable policies (Sun et al., 2006). The type of offshored R&D has also changed recently. Sun (2003) argues that foreign R&D units in emerging economies are not only tactical anymore, but MNEs establish strategic R&D facilities as well. The main differences between tactical and strategic R&D is that while the former one serves domestic markets, involves only the development and conduct short term projects, strategic R&D units serve the global markets, conduct original research and run crucial R&D projects that contribute to the long-term innovative performance of the whole MNE (Sun, 2003).
As R&D globalization gained attention among managers it also arouse the interest of researchers. The topic of R&D internationalization has been touched upon from several perspectives, including the most important trends, motives, performance implications as well as home and host country effects. Still, implications of foreign R&D in host locations need to be investigated more deeply to better understand its effects on specific areas. The current study therefore analyses the effects of R&D internationalization, specifically focusing on knowledge development in an emerging economy, China. Knowledge development is measured by the changes in scientific talent pool, the international recognition of locally created knowledge and innovation performance. The aim of this study is to answer the following questions related to knowledge development based on a quantitative correlational analysis in an eight-year period between 2005 and 2012; Is there a causal relationship between foreign R&D and the development of local scientific talent pool? Does foreign R&D increases the international recognition of domestically generated scientific knowledge? Does foreign R&D increases local innovative performance? Data to test the hypotheses was collected from the WIPO database and the China Statistical Yearbook. The results indicate that foreign R&D increases international awareness of locally created knowledge and contributes to domestic innovative performance. However, no causal relationship was found between foreign R&D and local scientific talent pool, indicating that although scientific
graduates attract international R&D, the attractiveness of scientific education does not increase as a result of foreign R&D in the country. The study contributes to academic literature by extending previous research on host country effects of R&D internationalization. Implications for practitioners include the need to take part in improving education more actively and to pay attention on the increasing domestic competition caused by foreign R&D. The results also provide important insights for the government. If China wishes to benefit from international R&D policies need to strengthen IPR protection and improve the educational system in the country.
The remaining of this paper is structured as follows. In the second section I discuss R&D internationalization focusing on its drivers, performance implications, location choice and home and host country effects. Building on the existing academic literature I address a research gap and present my theoretical framework and hypotheses. The third section covers the data and methodology I used to test my hypotheses, paying special attention on explaining the variables. Based on the SPSS analysis the fourth part discusses the results and provides implications for academics and policy makers. Finally, in section five, I draw the conclusions, address the limitations of my study and provide agenda for future research.
2.CONCEPTUAL BACKGROUND AND HYPOTHESIS DEVELOPMENT
2.1 Motives and performance implications of R&D internationalization
Dunning (1998) argues that the critical choice for a firm is to internalize intermediate product markets within the home country or in a foreign location. Investment motives of MNEs have traditionally been explained by market seeking, efficiency seeking or resource seeking. The modern approach also adds strategic asset seeking as an incentive, which originate from MNEs’ desire to catch-up, to diversify or provides an R&D springboard
(Dunning, 1998). In the case of R&D globalization drivers have changed significantly over the years and several motives played an important role in the offshoring decision.
While traditionally R&D is not among the most internationalized activities, trends regarding R&D globalization suggest a change (UNCTAD, 2005). Such trends are the overall trend of internationalization of services, which leads to the globalization of R&D activities as well. The degree of manufacturing outsourcing is an important determinant too. As production became more dependent on external sources, the need for innovation coming outside from the company may increase alike. The increasing pace of R&D internationalization, especially towards Asian host countries, changes the previous practices as well. Additionally, types of R&D activities in emerging economies do not only include local adaptation anymore, but complex stages of R&D are located overseas. Finally, companies from emerging economies also engage in R&D internationalization (UNCTAD, 2005).
Further, the emergence of new communication technologies and management practices has reduced transaction costs and made R&D globalization an attractive choice. New research methodologies, which enable the greater codification and standardization of R&D practices also enforce internationalization. Besides the need for local adaptation and accessibility of knowledge centres, new drivers, such as proximity to local talent, reducing R&D costs and speeding up the technology process incite MNEs to locate R&D centres abroad. Furthermore, centralization of R&D may generate additional costs as well, because R&D facilities may lose flexibility and isolate themselves from other business units (UNCTAD, 2005).
Dörrenbächer and Wortmann (1991) distinguish two levels of MNE’s motives for creating new R&D facilities abroad or enlarging their existing units. The authors refer to R&D related motives if the chosen location is the most efficient within the framework of the
corporate R&D system. Within R&D related motives two further types can be separated (1) R&D abroad that supports local production and (2) R&D that generates new technology, which will be used throughout the company. In cases, when the internationalization of R&D activities do not relate to the development of the overall R&D system Dörrenbächer and Wortmann (1991) talk about R&D unrelated motives. These include government requirements to preserve the R&D potential of their country, precondition for market approval or improving the image of the company by creating job opportunities abroad.
Chiesa (1995) presents another classification of incentives behind R&D globalization. The author distinguishes demand factors and technology supply factors. The former one refers to the need for accessing markets, responding to local needs and increasing proximity to customers. Technology supply related factors on the other hand include the access to technology and qualified personnel in the foreign location.
Lewin et al. (2009) approach R&D offshoring from human resources perspective. They conclude that the need to access science and technology personnel, coupled with the emerging shortage of qualified talent in the U.S. explain R&D globalization. The European Union faces similar problems in human resources, where 700,000 scientists and engineers would have been needed to meet the target of devoting 3% of GDP to R&D in 2005 (UNCTAD, 2005).
Gassmann et al. (2008) summarize the main motives of the R&D internationalization. The authors argue that the first set of incentives are science- and technology-related, including factors such as the proximity of local scientific community, lower R&D costs, and better R&D environment. On the other hand, sales and output efficiency incentives refer to for example the better connection to local markets, improvement of image, better compliance with regulatory requirements and tax optimization.
Regardless of the changes in the trends and in MNEs’ attitudes, R&D was one of the least globalized functions in 2005. One of the reasons of it is that MNEs choose to conduct R&D activities in their home country when the costs of transferring knowledge are high. Gassmann et al. (2008) point out that the most cited arguments against the trend include factors that (1) support central R&D, such as economies of scale, synergy effects, better communication and control over the process and legal protection, and (2) obstacles to international R&D: redundant development, differences in language and culture, loss of specific know-how, increasing costs due to political risk, establishment, coordination and access to information.
Other incentives against internationalization include loss of control over innovation processes, and the fear of knowledge spillovers to competitors (UNCTAD, 2005). Based on a research in Canadian MNEs Rugman (1981) found that in order to protect their knowledge advantage MNEs prefer home country locations for their R&D units. Another argument against internationalization of R&D concerns economies of scale and emphasizes problems in reaching the critical mass at the decentralized units. Furthermore, the increased costs of coordination and control make R&D internationalization less attractive to many companies (Chiesa, 1995).
Demirbag and Glaister (2010) argue that the new opportunities that emerged due to the increased options for R&D locations and the access to local talent pool result in higher uncertainty for MNEs. Consequently, MNEs need to make trade-offs between those opportunities and their inherent risks. Gassmann et al. (2008) also highlight the increased complexity resulting from the management of cross-border R&D activities. The authors suggest that synergy effects, like decreased time-to-market, improved effectiveness, and enhanced learning capabilities can compensate for the extra costs of coordination.
Empirical evidence suggests that MNEs are compensated for the emerging costs of global R&D. Surveying 253 CIOs and CFOs of the largest European firms, Oshri and Kotlarsky (2011) conclude that outsourcing innovation is critical to achieving high performance for the majority of the firms. Analyzing the relationship between offshoring R&D and performance, Nieto and Rodrigez (2011) found that R&D internationalization is positively associated with innovation performance, and this relationship is greater in product innovation than in process innovation. Similarly, Bertrand and Mol (2013) show that offshore outsourcing of R&D activities leads to more positive innovation outcomes compared to domestic R&D and again such effect is the strongest in product innovation. Focusing on Korean manufacturing firms Han and Bae (2014) also argue that R&D intensity is positively associated with firm performance in high technology intensive industries. Furthermore, the authors highlight that the ratio of researchers with Ph.D. further strengthen this positive effect.
However, there seems to be a right extent of R&D internationalization. Examining German firms Grimpe and Kaiser (2010) found evidence for an inverse U-shape relationship between outsourcing R&D and innovation performance. Reviewing existing literature on R&D outsourcing Hsuan and Mahnke (2011) also propose the same results, further dividing the curve in three phases, in which modularity effect (phase 1), learning effect, distance effect (phase 2) and intermediation effect (phase 3) guide the performance implication. Similarly, Adalikwu (2011) suggests an inverse U-shape relationship between R&D internationalization and innovation performance measured by the introduction of new products and sales growth in innovative products. Chen et al. (2012) however, propose an S-curve relationship. The authors argue that innovative performance increases in the decentralization stage, decrease in the transition stage and increase again in the recentralization stage as a result of R&D
internationalization. Further, their results suggest that organizational slack negatively moderates the relationship.
Conversely, based on patent data from more than 1100 firms Singh (2008) states that geographically dispersed R&D is negatively associated with average value of innovation due to difficulties associated with knowledge integration. The author however points out that MNEs can overcome the problems by using cross-unit integrative mechanisms among innovative teams.
2.3 Entry mode decisions
When companies engage in internationalization of their R&D activities they may locate R&D projects overseas by foreign direct investment, or by outsourcing. The former one includes the establishment of a new R&D facility or the acquisition of an existing R&D unit in order to take over its innovation potential (UNCTAD, 2005). Consequently, FDI through acquisition is either a main motive or a side effect of a strategic decision (Dörrenbächer and Wortmann, 1991). On the other hand, R&D outsourcing refers to the practice of conducting innovation processes outside the boundaries of the firm (Hsuan and Mahnke, 2011). Depending on the location of outsourcing, onshore outsourcing concerns contracting within the home location, while offshore outsourcing entails conducting the activity abroad. Contracts between two firms from different countries, and alliances by enterprises for joint R&D can also lead to the globalization of R&D (UNCTAD, 2005).
Kumar and Subramanian (1997) suggest that managers use some form of hierarchy when making entry mode decisions. Pan and Tse (2000) go further and establish the hierarchical model of market entry modes that distinguishes the two main types of entry modes. First, non-equity modes include export (direct export, indirect export and others) and contractual agreements (licensing/franchising, turnkey projects, R&D contracts and co-marketing). Equity modes i.e. foreign direct investment (FDI), cover joint ventures (minority,
50/50 and majority joint ventures) and wholly owned subsidiaries (Greenfield, acquisition and others). The model suggests that multinational companies first start at highest hierarchical level and choose either a non-equity or an equity mode of entry. The advantages of non-equity entry modes are fewer costs, less risk and less sensitivity to location-specificity. On the other hand equity modes provide the advantages of internalization and signals strategic commitment to the market (Pan and Tse, 2000).
Several authors have examined the factors that influence the entry decision of companies. One of the most influential streams of thoughts, which integrates those analysed factors (Agarwal and Ramaswami, 1992), is the eclectic paradigm introduced by Dunning in the late 70s (Buckley and Hashai, 2009; Dunning, 1988). The phenomenon states that the interaction of three interdependent variables determines the extent, geography and industrial composition of multinational companies’ foreign production. First, ownership specific advantages build on the competitive advantages of the firm (Dunning, 2000). Similarly to the resource-based view O-specific advantages look at the factors inside the organization to answer the question of why some firms perform better than others (Peng, 2001). The second variable is the location attraction, which highlights the importance of the home country’s location-specific assets (Dunning, 2000). L-advantages relate to the institution-based view of international business. Peng (2002) defines institutions as ‘the rules of the game in a society’ that differs across countries, therefore cause liability of foreignness for new entrants. Finally, the internalization advantages are built on the transaction cost theory. Firms evaluate the alternative ways in which they create and exploit their core competencies and vote for arms’ length (exporting and importing) or intra-firm trade. Dunning (2000) concludes that when all the three ‘OLI’ advantages exist, multinational companies engage in foreign direct investment. Pan and Tse (2000) follow a more specific approach by listing eight differentiate factors that grounds the choice between non-equity and equity entry mode decisions. These
are the prioritized location, host country risk, risk orientation, power distance, the extent of interaction between host and home country, and industry factors.
Additional factors also influence the entry mode decision regarding R&D. For example, Narula (1999) highlights that choosing the entry mode for conducting R&D abroad is highly dependent on the kind of innovative activity and the state of the technological paradigm. The UNCTAD (2005) report lists further determinants of how firms enter a foreign location. The higher the extent of the tacit nature of knowledge and the coordination required to manage R&D activities the less likely the firm will outsource them. Similarly, companies prefer to keep strategic and core functions, such as the majority of innovations, in-house.
Nonetheless, offshore outsourcing of R&D activities became a broadly used concept among MNEs. The most cited reasons for innovation outsourcing concerns lower corporate R&D costs and access to new talented researchers (Chou and Chou, 2011). Oshri and Kotlarsky (2011) argue that decision-makers are increasingly focusing on the latter one, driving back cost considerations to the second place. Hsuan and Mahnke (2011) summarize the most important changes in corporate R&D practices, which have led to R&D contracting. First, companies have realized that knowledge and innovation are also created outside the company. Second, MNEs became aware that centres of excellence in R&D are established all over the world, with the increasing importance of the BRIC (Brazil, Russia, India and China) countries. Moreover, developed information and communication technologies (ICT) play an important role in coordinating and contribute to the spread of outsourcing R&D through lower transaction costs. Other external factors also affect the spread of contracting in R&D. Such factors include the rising risks of R&D activities, the increasing complexity of innovation, the need for rapid innovation and cutting costs due to competitive pressures, and the need for expensive routine engineering and testing. The growing number of R&D providers and their increasingly specialized supply also play an important role in outsourcing
(UNCTAD, 2005). The opportunities in contracting R&D activities are limited though – too much outsourcing may result in loss of knowledge, or create competition (UNCTAD, 2005).
The second option for R&D internationalization concerns equity mode entry decisions i.e. foreign direct investment (FDI). Kuemmerle (1999) argues that there are two main reasons for engaging in FDI in R&D, augmentation or exploitation of existing knowledge base. The former one is driven by the need to access new resources and capabilities, and to capture externalities abroad. On the other hand, exploitation motives drive the process when MNEs aim to benefit from their existing capabilities Kuemmerle (1999). Whereas augmentation motives leads to the establishment of R&D units close to universities, knowledge exploitation is best achieved by locating the R&D unit close to existing manufacturing facilities and markets.
FDI can take the form of Greenfield investment or acquisition. The choice between the two types depends on the purpose of the R&D, the availability of suitable target firms and the competitiveness of the market. For example, in case of adaptive R&D Greenfield investment is more common, whereas in asset-seeking FDI in R&D acquisition dominates. Due to the higher innovative capabilities of firms in developed economies acquisition is a more attractive choice in such countries. Similarly, MNEs tend to choose acquisitions if target firms with strong and similar capabilities are available. Finally, market structures influence the decision. A more concentrated market structure as well as oligopolistic competition incites MNEs to enter the market by acquisition (UNCTAD, 2005). It seems however, that all in all the preferred way of R&D globalization within the internalized options is Greenfield investment (Kuemmerle, 1999; UNCTAD, 2005).
Finally, R&D globalization may occur by alliances. Studying the agreements of the world’s biggest information technology firms based on their patenting activities in the U.S, Cantwell and Colombo (2000) conclude the complementary technological competence
profiles of firms are directly proportional to their willingness to form alliances. R&D alliances may also be established based on competitive forces, as to explore new technological developments more rapidly (UNCTAD, 2005). Examining 276 international R&D alliances over the period 1995-2000 in China, Li and Zhong (2003) found that the location, the number of partners, the nature of local partner and the origin of foreign partner are significant determinants of alliance formation for R&D activities.
2.4 Location choice for R&D activities
Location choice for R&D activities can be analysed from the macroeconomic perspective, including home and host country determinants, or from the microeconomic perspective, i.e. the firms’ viewpoint. An important external factor that determines the offshore location is the strength of the destination country’s national innovation system (NIS). The NIS includes knowledge institutions, such as R&D laboratories, universities and institutions for standards and quality assurance, local and foreign R&D enterprises and the institutional framework for innovation (UNCTAD, 2005). Part of the institutional framework is the intellectual property right (IPR) regime. IPR protection plays an important role in R&D location choice, however Maskus (1998) points out that the importance of IPRs varies significantly by industry and market structure. Zhao (2006) indicates that weak IPR in a host location results in low returns to innovation and underutilization of talent.
The policy environment of the host country also has a great impact on R&D internationalization. As an UNCTAD secretariat (2005, p.16) summarizes: “A stable and
good general policy environment, including macro-economic and political stability, as well as consistent and transparent investment, trade and industrial policies, are important. Good communication systems and other infrastructural facilities are equally important for the dispersed R&D activities of TNCs.” Besides host country policies, home country initiatives
concessions, support for FDI, training, matching services, alliances and provision of resource access (UNCTAD secretariat, 2005).
Neither strong IPRs nor government incentives alone are strong enough to attract foreign investment; other determinants have to be in place as well (Maskus, 1998; UNCTAD, 2005). On the macroeconomic level, the likelihood of a location to be chosen as an offshore destination is positively related to its market size and income per capita. Factors like lower R&D wage costs compared to the home country, large pool of experts, and the level of technology know-how and knowledge infrastructure all attract foreign R&D (Kumar, 2001; Demirbag and Glaister, 2010 and Gersbach and Schmutzler, 2011). Examining MNEs from the U.S. and Japan Kumar (2001) goes even further, suggesting that the lack of adequate IPR or restrictive trade regime does not reduce attractiveness of a location if other advantages exist. Indeed, Zhao (2006) argue that MNEs establishing R&D units in countries with weak IPR protection use alternative mechanisms to protect their knowledge. The empirical analysis based on more than 1500 U.S. companies reveals that MNEs use internal organizations as a substitute for external institutions.
Although acknowledging the importance of country and industry factors, Feinberg and Gupta (2004) analyse offshore location choice from a different perspective, based on firm specific factors. The authors argue that MNEs anticipate potential knowledge spillovers from competitors and they base their decision on the opportunities to assess such spillovers across countries and across categories of competitors within the same country.
Chiesa (1995) explains that when choosing a location for R&D internationalization, companies consider both R&D related and non-R&D related factors. R&D related factors include (1) input factor costs: cost of personnel, equipment and facilities, (2) internal and external transportation costs, (3) organizational costs, which are associated with the establishment of a new R&D centre at a certain location and (4) input resource quality, which
refers to skills and quality of technical people, and the need to achieve critical mass and economies of scale. Surveying foreign R&D affiliates in the U.S. Florida (1997) also found that internationalization of innovation is largely dependent on technology factors, especially on gaining access to the scientific talent pool. Non-R&D related factors, such as existing business locations, local infrastructure and the company’s managerial culture are also taken into consideration (Chiesa, 1995). Demirbag and Glaister (2010) point out that the company’s prior R&D experience in the host country has an effect on offshore location choice. Not only R&D experience, but prior manufacturing or sales activities also influence the decision, driven by the need of proximity to production facilities (Gersbach and Schmutzler, 2011; UNCTAD, 2005). Chiesa (1995) conclude that balancing these factors results in the preferred location.
2.4.1 Foreign R&D in China
While traditionally mostly developed economies were taken into consideration, Qu et al. (2013) argue that emerging economies became the most likely destinations for multinational companies for offshoring their R&D activities and specifically China is the most popular choice. Similarly, Huggins et al. (2007) point out that Asia Pacific, especially China and India, host more than half of the FDI investments in R&D and almost three-fourth of job-creation through FDI took place in these countries (UNCTAD, 2005).
However, not only the quantity, but the pace of foreign R&D dissemination followed an impressive trend. Based on surveys at more than 200 multinational companies conducted in 2005 Thursby and Thursby (2006) found that almost 70 percentage of respondents anticipate an increase in R&D employment in China. By the end of 2010, more than 1300 foreign owned R&D centres had been set up in the country (Business Times, 2011). This meant a more than six and half fold increase within ten years (China 360, 2013). A year later Chou and Chou (2011, p. 352) report that “[t]here are about 300 MNCs that have set up
R&D centers in China, including big names such as Microsoft, Nokia, General Electric, Unilever and Alcatel-Lucent, etc. [...] [C]omputer MNCs in the IT industry such as Google, IBM, Motorola and Intel conduct their advanced research and fund projects at China's universities.” In 2013, 400 out of Fortune 500 firms have already established R&D centres in
the country (Tiak, 2013). The trend seems to continue, since the number of R&D centres in the country exceeded 1500 by today and the estimations predict a 20-percentage growth by 2018 (Jolly et al., 2015).
Most of the investments within the country are centred in Beijing and Shanghai, with over 550 and 350 R&D centres respectively (Business Times, 2011; China 360, 2013; Sun et al., 2006). Based on interviews in 18 foreign R&D facilities Sun et al. (2006) highlight that although Shanghai is an attractive location, and its industrial structure is more diversified, the latter city hosts the majority of foreign R&D. The main reason behind it is that the largest IT enterprises are based in Beijing and the majority of foreign R&D is in the IT industry. Furthermore, the best universities are located in the Beijing, providing a better access to local talent for the international companies. As a result of local incentive systems, other tier-two and tier-three cities, such as Xi’an, Chengdu and Chongqing, are getting more attractive locations for multinational companies next to the two main destinations (China 360, 2013).
2.4.2 Attractiveness of China
The main reason behind the wide spread of foreign R&D in China is that the country offers all the incentives that make a host location attractive. As China’s innovation system developed, motivation of multinational companies to locate their R&D there has changed accordingly. Initially, MNEs were attracted by the low costs, but later on the incentives changed to the desire to cater the local market as well (Jolly et al., 2015). Thursby and Thursby (2006) rank the factors that influence R&D location choice and find that those companies, which chose an emerging economy as a host country for R&D activities were
attracted mostly by the growth potential of the market. The second most cited reason is the quality of R&D personnel, since China is the world leader in terms of the number of scientific graduates (Yang, 2012). It is expected that by 2020 the number of Chinese citizens who received university and college education will exceed 120 million, of which 25-30% is in the field of engineering and science (Sigurdson, 2004). However, Lewin et al. (2009) draw attention to the fact that the commonly held phenomenon about the unlimited talent supply is wrong. The authors argue that China faces a growing shortage of high-quality science and engineering graduates. Furthermore, MNEs have to cope with the low level of English understanding, which in turn can decrease the attractiveness of the country.
Other incentives, listed in the order of the importance, include costs, expertise of university faculty and the ease of collaboration with universities (Thursby and Thursby, 2006). Surveying 50 R&D centres established in China Jolly et al. (2015) come to a similar ranking. The authors conclude that 54 percentages of respondents had “market-driven” incentives, 28 percentages are engaged in “knowledge-driven” R&D, while only 18 percentages have chosen China for its cost advantages (“cost-driven R&D”). Walsh (2005) further argues that the proximity to already offshored manufacturing and production in China attracts foreign R&D too.
Chinese legal environment also plays an important role in the attractiveness of the country for R&D investment, which signifies the gradual shift away from manufacturing toward innovation and intellectual know-how development (Yang, 2012). The ‘National Medium- and Long-Term Program for Science and Technology Development’ program released in 2006 by the national government aims to increase R&D expenditure to 2,5 percentage of GDP, derive more than 60 percentage of economic growth from technical progress and position China as fifth worldwide for patents and citations of publications by 2020 among other goals (China 360, 2013).
IPR protection in the country needs to be improved though. Due to the lower assurance in terms of intellectual property (IP) in emerging economies respondents in the study of Thursby and Thursby (2006) stated that only 22 percentage of the cites are established for new science – compared to the over 45 percentage in developed economies. Other research also recognized that the quality of IP protection in China is behind compared to the developed economies (China 360, 2013), but the increasing number of patent and invention applications underpins the efforts of the Chinese government. For example, intellectual property protection was enforced by the 12th Five-Year-Plan on National Intellectual Property Development starting in 2011, or the 12th Five-Year-Plan on Patents (Yang, 2012). The government also supports foreign direct investment for R&D activities by the Chinese tax system that offers tax benefits and provides financial support for companies that establish R&D facilities in the country (Yang, 2012). Such tax incentives include the reduced corporate income tax rate of 15%, super deduction, tax concession, customs duty and VAT exemption or refund, and tax concessions on technology transfers (KPMG, 2012). In specific cities or regions, for example in Shanghai or in Chongqing and other western areas additional local subsidies may also be available (China 360, 2013).
2.5 Effects of R&D internationalization
On the macroeconomic level R&D globalization may affect both the home and the host country’s economy. The net outcome of R&D offshoring on both locations is difficult to predict. The implications depend primarily on the extent R&D offshoring impacts the countries national innovative capabilities (UNCTAD, 2005). However, Alazzawi (2012) indicates that the effects of R&D internationalization highly depend on the nature of the home and host countries. The author concludes that inward and outward FDI positively affect domestic innovation for technological followers, whereas for technology leaders the positive
implications arise from outward FDI. Inward FDI in the latter case increases competition, but contributes to domestic productivity.
2.5.1 Home country effects
By the advantages gained from internationalization of their R&D activities, such as improved technological performance, MNEs indirectly increase the competitiveness and growth potential of their home countries. By offshoring the less sophisticated R&D activities, thus allowing the home country to focus on the higher value added ones, efficiency is further improved by R&D globalization (UNCTAD, 2005). Examining 262 EU regions, Castellani and Pieri (2013) suggest that R&D offshoring results in high regional productivity growth, especially if the offshoring takes place in South East Asia.
Another positive consequence of R&D globalization is the so called inter-firm reverse technology transfer, which refers to the transfer of new foreign technology to other companies in the offshoring company’s home country (Criscuolo, 2009). However, such effect appears to be significant only in case of technologically advanced host countries and depends on the purpose of the R&D too (UNCTAD, 2005). Offshoring R&D does not only have a positive influence through technology transfer, but also by increasing demand for other inputs and services originating from the home location. Therefore, as a result of localization strategies by MNEs, they promote market expansion of their home country (UNCTAD, 2005).
The influence of R&D globalization on the home country has been examined from the perspective of knowledge creation as well. Piscitello and Santangelo (2009) observe a positive impact when offshoring takes place from home OECD countries towards emerging economies. Similarly, D’Agostino et al. (2013) investigate whether offshoring R&D from high-income countries to emerging economies has an effect on regional knowledge production in the home country. Assuming that high-income countries have advantage in
high-tech R&D and destination countries lead in medium/low R&D, the authors confirm that when these advantages are utilized complementary relationship exists between home location and foreign investment in R&D. More specifically, D’Agostino and Santangelo (2012) suggest that regardless of their intensity R&D units focusing on adaptation, and medium technology-intensive R&D laboratories focusing on development complement knowledge creation in the home country.
Another stream of research has been focusing on analysing the effects of offshoring on domestic employment. The academic literature includes two main effects in this regard, namely the productivity effect and the downsizing effect. Agnese (2009) for example found that offshoring materials and services to Japan negatively affects the labour market through the lay off of workers. However, the author highlights that the overall effects of offshoring are not significant, rather a reassignment of tasks appear. Furthermore, by offshoring activities to Japan, more productive activities, such as skill upgrading present on the local market. Cai and Zhang (2011) also found positive association between offshoring and labour productivity with a more significant effect of service offshoring compared to material offshoring. Egger et al. (2015) point out the importance of costs of offshoring and conclude that domestic welfare may fall when the associated costs are high, but the effects are reversed when offshoring costs are low. Similarly, Moser et al. (2015) report mixed results. They encompass German establishments in their research and demonstrate the positive productivity effect of offshoring on local employment, while they also find evidence on the negative downsizing effect.
The increasing supply of low-cost, but skilled human resource from the host countries may result in the loss of research jobs and can have a downward pressure on researchers’ wages in the domestic market. Another concern is the risk of technology leakage, referring to that R&D abroad may reduce the demand for domestic products on the short term, while may
affect the strategic position of the home country on the long term, because it loses control over key technologies (UNCTAD, 2005).
Other negative home country effects have been noticed as well. A possible consequence of R&D internationalization is that it replaces domestic R&D, which in turn leads to the deterioration of domestic national innovative system and loss of skills (UNCTAD, 2005). Fifarek et al. (2008) also suggest that internationalization of R&D activities has a negative effect on the home country’s development. They conclude that the level of R&D and productivity of innovation processes decrease in the domestic market as a result of offshoring innovation.
In spite of the possible negative effects, there is still a long way to go until knowledge-intensive R&D in emerging economies become serious competitors of developed home countries (UNCTAD, 2005). Castellani and Pieri (2013) also conclude that the net effect of carrying R&D abroad is rather positive from the home location’s perspective.
2.5.2 Host country effects
Offshoring R&D also has important benefits from the host country’s perspective. Foreign R&D causes structural changes in the national innovation system (NIS), by increasing its complexity (UNCTAD, 2005). For example Chen (2006) confirms that FDI in R&D enriches and restructures the innovation system in Shanghai. MNEs also connect developing economies into global technology development activities by linking the local innovation system to the global R&D network the MNE operates in (Reddy, 1997).
Foreign R&D also creates the innovative enterprise sector, which is primarily relevant in developing host location, where R&D is mostly performed in universities and government research institutions. The NIS is further enforced by new methodologies and skills brought in by foreign investment and by the demand-creating effect of foreign R&D for related services (UNCTAD, 2005).
The host location’s human resources also benefit from international R&D. Since one of the main incentives of MNEs is to tap on scientific talent pool, they increase employment and provide in-house training for the their employees (Yuan, 2005). Higher education also benefits from incoming R&D activities, through the internship and fellowship programs as well as specific courses by foreign MNEs. When the trained employees move to a local company or start their own business spillover effects may also arise. Analysing FDI in Shanghai Chen (2006) argues that as a result of the localization of foreign R&D units, trained engineers move to local firms and new research fields are created through collaboration with universities. One of the most important benefits of international R&D, especially in China and India, is the “reverse brain drain” effect. The term refers to the practice of those individuals, who tapped on the new career opportunities, worked abroad in R&D and returned back, bringing new knowledge and skills to their home country (UNCTAD, 2005).
Foreign R&D may also help the host country to move up the value chain (UNCTAD, 2005). Kaplinsky and Morris (2001) distinguish four types of upgrading, which may all be a result of R&D internationalization. First, process upgrading refers to the increased efficiency of internal processes. Second, product upgrading refers to improvements in product development processes. The effect of foreign R&D on process and product upgrading highly depends on whether the outcomes of R&D activities are applied in the host country. The third type, functional upgrading involves changes in activity mix that in turn increases the value added. MNEs contribute to functional upgrading by transferring resources, providing demand for R&D outcomes and stimulating the innovative system in the host country. Finally, chain upgrading i.e. moving to another value chain may occur through good reputation of a host location by MNEs, or by empowering host locations to build up more knowledge-intensive industries.
Another positive consequence is that R&D globalization contributes to industrial cluster formation in the host country. Manning (2008) describes the process of cluster development, starting by a pioneer MNE, which customizes the science and engineering clusters. The clusters therefore develop specialized resources and service capabilities in accordance with the pioneer MNE’s requirements, and attract follower companies who have the same sourcing needs. However, Reddy (2005) warns that in developing economies barricaded high-technology territories may emerge and inhibit knowledge dissemination.
Extensive literature focuses on knowledge spillovers from MNEs to the host country, however evidence is mixed in this regard. Cassiman (2004) argue that it is not the international character of MNEs per se that causes knowledge spillovers to local economy, but their access to the international technology market. In the U.S semiconductor industry Almeida (1996) for example found that foreign firms learn from, as well as contribute to the host county. The developed management systems and methods brought by MNEs may also have positive spillover effects on local management (Yuan, 2006). However, unintended spillovers may also result in the emergence of new competitors (Kozhikode, 2009). Similarly, results are mixed regarding the implications of R&D fragmentation by MNEs. While on the one hand it can lure away learning opportunities for the host country and reduce spillover benefits, it can also lead to economies of scale in R&D specialization and attract more foreign investors (UNCTAD, 2005).
It is important to address the negative implications of international R&D in the host country too. When foreign firms enter the market through acquisition a possible consequence is that they close up existing R&D units, thus reduce R&D activity (UNCTAD, 2005). Since China is a member of the World Trade Organization crowding out local laboratories is an even more important threat for the country (Yuan, 2005). However, Cassiman et al. (2005) explain that the effects depend on whether the merging firms have complementary
technologies. It seems that negative effects only appear in cases when the acquired entity had substituting technologies.
Foreign MNEs may also harm the local innovation system by unfair compensation for the jointly developed intellectual property, which is a result of unbalanced bargaining power, information asymmetry, market failures or institutional deficits. The low costs of human resources and government incentives may impel MNEs to unethical behaviour and race to the bottom (UNCTAD, 2005). Several examples are listed about unethical behaviour in the pharmaceutical industry regarding clinical trials. The most immoral cases pertain to experimental drug testing in humans without establishing the critical safety requirements. Examples for such behaviour include the practice of Sun Pharmaceuticals, Shanta Biotechnics and Biocon in India, or Viral Genetics in China. However, the case of the Miami test centre in the 21st century, where Latin American immigrants were treated unethically highlights that examples are not limited to developing countries (SOMO, 2008). MNEs can also crowd out in the local labour market by attracting researchers from local R&D institutions (UNCTAD, 2005).
Castellani et al. (2015) provides an extensive summary about host country effects by distinguishing direct and indirect effects of R&D internationalization. The former one derives from the fact that foreign MNEs are more productive compared to domestic firms, and that they usually concentrate in higher average productivity sectors. Direct effects influence aggregate productivity and innovation positively in the host country. On the other hand, indirect effects result from the technological externalities, crowding-out and business stealing effects that offshoring MNEs cause. Despite the extensive literature regarding the influence of indirect effects, no clear consensus about its influence exists. Reddy (2005) also points out that direct and indirect effects as well as the type of the investment are important to consider to understand host country implications.
2.5.3 Theoretical framework and hypotheses
R&D internationalization has been broadly investigated by researchers, and its possible implications on the macroeconomic level have been analysed too. Still little is known about the relationship between R&D internationalization and the host location’s knowledge development. The aim of this study is to shed light on this relationship by examining three possible effects of foreign R&D in China. The importance of such analysis is expressed by Thomson (2013), who addresses the need for investigating the consequences of R&D offshoring in the destination country. While previous studies showed how employment is affected, they all focus on implications for those who already graduated. As discussed above the attractiveness of a location for foreign R&D is highly dependent on the available scientific talent pool, specifically tapping on the local knowledge is one of the most cited reasons for internationalization of R&D activities. Thomson (2013) for example argues that national scientific capacity is an important factor attracting offshored R&D for host countries. Meanwhile, foreign R&D measured by the number of non-resident patent grants may also has a reverse effect on scientific talent pool. The increasing foreign R&D activity in China gives rise to the number of R&D institutions in the country by FDI, as well as by stimulating the domestic innovation system. Those institutions generate supply and provide an attractive career opportunity for students graduating in the field of science. As argued among experts foreign R&D inflow impacts the employment prospects of trained personnel by providing job opportunities for skilled workforce (UNCTAD secretariat, 2005). A case study on Texas Instruments in India also reveals that new job opportunities emerge in the interface of science and business for local researchers (UNCTAD, 2005). It is therefore expected that R&D institutions provide an incentive to students to get involved and graduate in scientific education.
H1. The number of non-resident patent grants is positively
associated with the number of scientific graduates in China. The
number of R&D institutions in the country mediates this positive
association.
Existing literature proves that international R&D in the host location connects the domestic national innovation system to the global innovation network and improves local NIS (Reddy, 1997; UNCTAD, 2005). On the other hand the question whether the awareness of the developed innovation capacity and knowledge increases is still unanswered. This paper therefore investigates this issue by measuring the increased awareness of scientific knowledge creation by the number of scientific paper citations. The underlying assumption of hypothesis 2 is that on the base of more developed local knowledge more scientifically relevant research is published. As foreign R&D enters the market and improves the host country’s national innovation system, more attention will be paid on these local scientific publications. Relatedly, to make these researches available internationally, they will be taken in major foreign reference systems.
H2. The number of non-resident patent grants is positively
associated with the number of scientific paper citations in China.
The current study also aims to broaden the literature on local innovative performance as a result of R&D globalization by analysing the development of innovation system. By entering the Chinese market foreign companies are likely to cause knowledge spillovers, which increase innovative performance in the host country. For example based on the increasing domestic R&D spending in India Reddy (2005) demonstrates the positive effect of foreign R&D. Previous research mostly applied qualitative analysis on how national innovative system develops, here however innovative performance is measured by national patenting activity and a quantitative approach is used to address the host country effect.
H3. The number of non-resident patent grants is positively
associated with the number of national patent grants in China.
Figure 1 summarizes the above-discussed hypotheses and shows the predicted relationships between the variables.
3.DATA AND METHODOLOGY
3.1 Research design
The effects of R&D internationalization is addressed by a quantitative correlational research in the period between 2005 and 2012 on a macro level in China. Data was collected yearly in the 8-year time period, which results in a total 568 observations. The next section elaborates on how these observations were aggregated and analysed. R&D internationalization is measured by the number of non-resident patent grants, which refers to the foreign owned patents granted in Chinese filing offices. The dependent variables – scientific graduates, scientific paper citations and national patent grants – on the other hand proxies the level of knowledge base in the host country. The theoretical framework includes a mediator variable as well, namely R&D institutions, which mediates the relationship between non-resident patent grants and scientific graduates.
Knowledge development in the host country
Scientific graduates
Scientific paper citations
National patent grants
R&D
internationalization
Non-resident patent grants
R&D institutions
H1 + H2 + H3 +
The empirical analysis is based on secondary data provided by the World Intellectual Property Organization’s (WIPO) Statistics Database and the China Statistical Yearbook (CSYB). The WIPO Statistics Database collects data from national and regional IP offices through yearly questionnaires, generates data from international filing and registration systems, and complies data from the Patstat database (WIPO Statistics). The CSYB is an annual publication of the National Bureau of Statistics about economic and social development of China.
The highly reliable sources ensure the quality of the collected data, which is publicly available for the whole period (2005-2012). In CSYB from 2011 the ‘large- and medium sized industrial enterprises’ include enterprises with the sales revenue above 20 million RMB, compared to the previous 5 million RMB limit, which affects the mediator variable. Although it raises the question of comparability among the years, we can assume that the sales revenue of enterprises with R&D institutions exceed 20 million RMB1. Besides the above-mentioned change in regulation there has been no other reported change that may affect the comparability of the variables through the analysed years.
3.2 Variables
Non-resident granted patents. On a macroeconomic level patent indicators serve as
good proxies to measure inventive performance, and are the most commonly used variables to address the output of R&D activities. The reason for such broad usage of them originates from such advantages as their close link to invention, the broad coverage of technologies and the relatively easy accessibility from patent offices (OECD, 2009). The WIPO Statistic Database provides resident and non-resident count by filing office, which enables to measure the extent of foreign presence within R&D activities. The term ‘resident’ refers to the filings made by applicants in their national or regional office. For example, a Chinese applicant that
files a patent in a Chinese office counts as a resident filing in China. In contrast, ‘non-resident’ filings are received from a foreign applicant in a Chinese filing office. The intellectual property grant data uses the same concept as the data on applications (WIPO). Since the patents are validated by their registration, the number of granted patents is a more sophisticated measure for successful R&D activities compared to patent applications. For this reason, the current study uses the number of granted patents as the independent variable in the model. The measure includes both the direct and the Patent Cooperation Treaty (PCT) national phase entries, where the latter refers to “the decision by a PCT applicant to enter the
national phase before a national or regional patent office is referred to as national phase entry” (WIPO Glossary).
It is important however, to address the limitations of using patents as measure of innovative performance. Given the definition of ‘non-resident’ filings, the variable does not take into account those foreign inventions that were created in China, but were patented in a foreign IP office. Therefore, the activities of institutions that carry on research and development, but the innovations filed by another IP office are excluded from the analysis. Although such practice occurs, for example in the pharmaceutical industry, the effects of foreign R&D activities, which would not result in patent applications in the host country are smaller, thus negligible. Another limitation is that the value distribution of patents is highly uneven, as some of them provide higher value for the society as others. Furthermore, some inventions may be dismissed due to their unpatentable nature or of being protected by other methods. Finally, the differing propensity to patents and the different patent regulations across countries and years suggest using different indicators (OECD statistics). Even though, the above-mentioned drawbacks provide some limitations, they do not significantly affect the steadily increasing pattern of the number of ‘non-resident’ patents in the analysed period.
Furthermore, since the host country is limited to China, country level differences can be also omitted.
Scientific graduates. The variable on the number of scientific graduates is an
aggregated variable that consists of four components: (1) Number of postgradual students graduating both in regular colleges and research institutions, (2) Number of graduate students in adult institutions of higher education run by adult institutions of higher education or run by regular institutions of higher education, (3) Number of students graduated in graduate in undergraduate and junior colleges and (4) Number of graduate students enrolled in internet-based courses. Regular institutions of higher learning include full-time universities, colleges, high professional schools and short-term professional universities (CSYB definitions). Graduates in institutions of higher learning for adults provide similar qualifications as graduates in the regular institutions of higher education. Such institutions include radio and TV universities, schools of high education for staff and workers and peasants, colleges for management cadres, pedagogical colleges, and independent correspondence colleges (CSYB definitions). Each component takes into account the graduates in the field of science as published by the CSYB each year.
Paper citations. The variable on paper citations refers to the Chinese scientific papers
taken by major foreign referencing systems, such as the Science Citation Index (SCI), the Engineering Index (EI) and the Conference Proceedings Citation Index – Science (CPCI-S, called ISTP before 2010). Data on the scientific paper citations provided by the CSYB was compared to the top fifteen industrial sectors by overall R&D investment (Hernández et al., 2014) and to patent applications by top field of technology (WIPO database). Table 1.1 lists all the disciplines within Chinese scientific papers taken by foreign referencing systems. Table 1.2 lists the industries and their R&D intensity taken from the EU R&D Scoreboard 2014. Overall R&D intensity in the different sectors is based on the 2500 scoreboard
companies, from which almost three-fourth of all the companies located in Europe, the U.S. and Japan (Hernández et al., 2014). Since the current study focuses on the effects of R&D offshoring, the number of scientific paper citations is only relevant in the disciplines in which R&D intensity is high. Consequently, scientific paper citations are taken into account only in those disciplines that are listed in the top fifteen sectors by R&D intensity. Therefore, the following seventeen disciplines are selected for further analysis: Mechanics; Information, Systems Science; Chemistry; Protective Medicine; Basic Medicine; Pharmacy; Clinic Medicine; Special Medicine; Livestock, Veterinary Medicine; Engineering & Basic Technology Science; Machinery, Instrument; Power & Electrical Engineering; Nuclear Technology; Electronics, Communication & Automation; Computer; Chemical Engineering; and Aviation and Aerospace.
Table 1.3 lists patent applications by top fields of technology in the period 1993-2013 based on the WIPO statistical database. Considering these fields, three additional disciplines were added to the analysis: Energy; Metallurgy, Metallography and Food. Therefore, in total twenty out of forty disciplines from Table 1.1 were taken into account for further analysis. To make them easily distinguishable, they are highlighted with light green background.
National patent grants. The third dependent variable measuring knowledge
development in China is the number of national patent applications. Similarly to the independent variable, data is collected from the WIPO Statistics Database, which provides patent data by country of origin. Country of origin refers to the resident of the applicant or investor (WIPO Statistics). The variable includes the number of total patent grants, including direct and PCT national phase entry, originating from China. This means that the variable contains patents that were granted in China, as well patents from a Chinese applicant granted in a foreign filing office. On average, however 95 percentages of patents were granted in the home location (WIPO Statistics). As discussed above, patent indicators are broadly used to
measure innovative performance, and thus provide a great proxy for the innovative development in China.
Table 1.2: Ranking of the top 15 industries by overall R&D intensity
Source: Hernández et al. (2014)
Table 1.3: Patent applications by top fields of technology in China (1999-2013)
Source: WIPO database
Table 1.1: Disciplines as a base for data on scientific paper citations
R&D institutions. The number of R&D institutions represents the mediator variable in
the model. It consists of three sub-variables: (1) Number of R&D institutions subordinated to central and local level in units, (2) Number of R&D institutions in higher education for science and technology activities, and (3) Number of R&D institutions of large- and medium-sized industrial enterprises. Due to inconsistency in the yearly editions in CSYB about the latter sub variable, data was collected from the statistical yearbook that provided data for almost all the years. Therefore CSYB 2011 provides the input for years 2006-2010, while for the last two years the latest editions were used, CSYB 2013 and CSYB 2014. Due to lack of data for the year 2005, the number of science and technology institutions was used from CSYB 2006.
3.3 Data analysis
To test the first hypothesis (H1) simple mediation was performed using the PROCESS macro for SPSS and SAS by Andrew F. Hayes. The single mediator model in the PROCESS add-on is based on an ordinary least squares or logistic regression-based path analytic framework (processmacro.org). The second and third hypotheses (H2 and H3) were tested using simple regression in SPSS. Both simple mediation and regression enables to predict a variable from another, therefore are good methods to test the predicted casualties (Field, 2009).
4.RESULTS
Table 2.1 reports the descriptive statistics i.e. means, standard deviations and correlations of the variables. The numbers indicate a high positive relation between the number of non-resident patent grants and the dependent variables. High positive relation is measured between scientific paper citations and scientific graduates (Pearson’s correlation coefficient = .802), which suggests that scientific papers are mainly written by local
