1. Introduction
To stay ahead in an era of incredibly rapid change and stiff global completion, firms have had to pursue access to exclusive resources, such as new knowledge. As a consequence, industry-university collaboration has emerged as a sustainable enabling solution for technology transfer and commercialization, and as a fertile breeding ground for new inventions [
1,
2,
3,
4].
Universities have important roles to play and contributions to make to the creation of knowledge and technology at the economic and social levels of a nation. University-industry collaboration may begin in response to a firm’s need for new technology, or conversely, when new knowledge generated within a scientific field yields practical applications. By means of this collaboration, it is possible to make a bridge for the exchange of new knowledge and technologies generated within the university, and the real-world practice and marketing experience of particular organizations [
5,
6]. Numerous prior studies argue that university-industry collaborations, in particular where these are formed within science and technology-based industrial sectors, should be recognized as an important impetus for economic growth [
7]. Reference [
6] argued that the more experience accumulated by biotech with university partners, the greater the performance impact of these university-industry collaborations, while [
8] demonstrate that firms located in proximity to universities are likely to grow faster than other firms thanks to university-industry collaboration. In particular, these collaborations have the potential to act as a bridge in transforming academic discoveries into commercially successful innovations [
6].
Absorbing new knowledge and technology from universities as well as maximizing their capacity to use that knowledge for performance gains is critical for a firm’s survival through the utilization of open innovation [
9,
10,
11]. Not surprisingly, university activities such as technology transfer and commercialization have received much attention in the literature in recent years [
12,
13,
14,
15]. It is firms, rather than universities, which act as the main entities for commercializing innovations.
Patenting of new inventions via industry-university collaboration is the first step in a long process of commercialization. In fact, what is ‘commercialization’ from the standpoint of the firm, becomes ‘knowledge transfer’ from the point of view of the university. Technology commercialization is the act of reaching the market, using the outputs of research [
16]. In order to implement the process of commercialization, there are numerous difficulties and hurdles to be overcome. Patents covering the output of innovation can be used as a tool not only for protection in innovation management but also for incentive creation, collaboration, negotiation and licensing, etc., [
17,
18,
19].
Patent propensity can also be expected to vary according to industry, along with other factors. The object of this study is to discover whether there is any empirical evidence that characteristics of the industry, characteristics of the firm, InnoBiz (an innovative firm certified by the government), age of the firm, research division or the academic credentials of the CEO might affect a firm’s patent propensity. This paper will expound upon the ways in which innovation polices can be targeted to influence the firm’s growth. It also sheds new light on various methods of technology transfer and commercialization and on innovation polices aimed at influencing firms’ growth in a particular region. In addition, while there are a few studies which have analyzed technology transfer commercialization by using hard data, in this paper it is analyzed on the level of the firm. Also, the focus of this paper on a specific stage—patenting—sets it apart from most prior studies, which have approached industry-university collaboration from the standpoint of the overall process of technology transfer and commercialization, by means of an exploratory and qualitative approach.
In order to achieve the research goal, we considered a sample of 263 small and medium enterprises (SMEs) in manufacturing industries—including some high-tech enterprises—operating in Gwangju, South Korea, that have engaged in joint research with universities. Specifically, this paper used data on firms that had received a government designation of ‘innovative’ sometime between 2014 and 2016. The author collected the data from multiple institutions (i.e., SMEs center, the Association of Managerial Foundation in Gwangju and the InnoBiz Association). According to the findings, the CEO’s academic credentials constitute the primary factor that is positively correlated with patent propensity. Surprisingly, with regard to the characteristics of the industry, both automotive parts, and electronics and electricity are less likely to patent innovations, reflecting the fact that these industries are more interested in cost comparativeness. The lack of any significant differences in innovation activities and outcomes between InnoBiz firms and general SMEs is highly unexpected. Another factor of importance is the number of researchers, which has a positive effect on patent propensity.
The remainder of the study is structured as follows.
Section 2 presents the literature review and the hypotheses we investigated.
Section 3 presents our data sets, variables, and estimation methodology.
Section 4 presents the results and discusses them. Finally,
Section 5 presents the conclusions and implications of the research.
3. Data and Research Methodology
As described above, the object of this study was to establish the relationship between innovative firms and patent propensity among firms in Gwangju that had engaged in any form of industry-university collaboration. The assembled dataset was composed of 15 variables that served as ways of characterizing each firm, including financial performance, over the three years from 2012 to 2014. We considered the regional industry conditions and collected data from the automotive parts and machinery, electronic and electricity, and photonics industries, accounting for form representativeness.
The data set is unique. We collected the patent data from the Korea Institute of Patent Information (KIPI), an affiliated public institute of the Korean Intellectual Property Office. The KIPI has been designated as a specialized institute that provides information on industrial property rights to build an information infrastructure for Korea’s intellectual property. We obtained the data for the independent variables (e.g., R&D intensity, firm age, CEO’s academic degree and number of researchers) from the Korean Investment Service, a Moody’s affiliate, and the Korean Small and Medium Business Administration (SMBA). The SMBA has supported Korea’s SME-led innovation, as a dynamo of the national economy. Its primary mission has been to foster challenging and innovative SMEs in order to maximize Korea’s growth potential. In particular, SMBA’s regional branch office keeps data such as CEOs’ career histories, where firm employees come from, each firm’s financial state and main product lines. We utilized these data to analyze innovative outputs as a means of complementary data.
The data came from 263 firms that had participated in industry-university collaboration programs more than once within a decade, collected from official databases and covering the period from 2012 to 2015. This formed an appropriate sample for the exploration of our research hypotheses, given that these SMEs are well-known for their high entrepreneurial and manufacturing capabilities. Concerning firm size, we used SMBA’s definition that an SME has fewer than 350 employees; the firms for our analysis employed from 1 to 303 workers, with a mean of 36. One hundred and twelve firms (43%) held InnoBiz certification. The frequency distribution of the industry variables was 151 (58%), 51 (19%) and 61 (23%) firms in the automotive parts and machinery, electronics and electricity, and photonics industries, respectively; the distribution (
Figure 2) shows that Gwangju was home to the largest automotive firm, KIA, and thus, the first and second tiers of KIA are comprised of more than 50% of the regional firms.
Figure 2 shows that the ratio of industry is balanced in representing the regional firms. As a global car maker, KIA has been pivotal in its contribution to the regional economy and its growth; there are many automotive product SMEs in this particular value chain. Traditionally, the electronics and electricity industry has been the second largest contributor to Gwangju’s regional economy.
From the raw data, we found that the ratio of InnoBiz firms—43%—was relatively high, particularly in view of the fact that InnoBiz certification began only in 2001; we can assume that young firms with less than ten years of operation are mainly InnoBiz firms. The data also suggests that young firms could be more innovative than others: considering the number of patents for each firm in the three years of the study, younger firms had more patents than older firms. Thus, InnoBiz firms are likely to engage in more innovative activity and to generate more patents. The Korean government selected the photonics industry as one of its high-tech industries for the 21st century and has established plans for substantial investment and systematic promotion of this industry, and in its 2001 regional economic policy, the central government designated Gwangju as a photonics industry complex area in order to boost the regional SMEs. The Korea Association for Photonics Industry Development (KAPID) was established with the purpose of developing South Korea’s photonics industry and promoting and expanding relevant businesses. As part of the government’s strong support, the KAPID was set up as a major civil organization to promote the industry, a high value-added cutting-edge industry based on new technologies.
As mentioned above, in accordance with the government’s plan to promote the photonics industry, the nation’s largest photonics-oriented industrial complex was established in Gwangju.
Table 1 shows that relatively few firms (84 out of 263, 32%) had a research division, considering that 112 firms were InnoBiz certified; the data suggests that some of the InnoBiz firms do not have a research division, and certainly, just because a firm has a research division does not mean it is innovative. The average firm age was 12 years, which is relatively young. With regard to the CEOs’ academic credentials, only a few had PhDs. For the photonics industry, cutting-edge technology and pure science knowledge are required, and it can be expected that the higher the CEO’s credentials, such as a PhD, the greater the likelihood of their innovation success, but only 28 CEOs had a PhD. Interestingly, more than half of the firms in our data set had no patents; many firms have not been seeking to obtain patents as an important innovative output from the data description.
Table 2 reports the summary statistics of the data. The correlation matrix of the dependent and independent variables (
Appendix A) showed that the correlations among the independent variables were not strong, suggesting that multicollinearity and confounded effects were not a serious problem.
Based on the findings from the review of literature on SMEs’ innovation activities and innovative outcomes (patents) and the availability of representative data, we may postulate a model of patent propensity as follows:
where the subscript
i indicates firm,
ResFun,
Age,
ResDiv,
Researcher,
InnoBiz,
IndCode and
CEOdeg are research fund, age of firm, research division,
InnoBiz certified, industry code and
CEO degree, the
β’s are the effects of the explanatory variables on the number of patents and
ε is an error term that captures the effects of left-out variables, measurement error in patents and random events.
IndCode and
CEOdeg are categorical variables. For the purpose of conducting a sensitivity analysis of the relationships between patents and the explanatory variables, we estimated approximations to multiple models using multivariate regression analysis and ordinary least squares (OLS) with robust standard errors. The results are presented below.
In summary, the hypothesis that InnoBiz firms are more creative in patents than general SMEs (H1) was statistically rejected. SMEs’ research funds and having a research division are important in patent propensity in terms of SMEs’ performance, which was also statistically rejected with respect to the research institute (H2). The hypothesis that older firms show greater patent propensity than younger firms (H3) was statistically insignificant and thereby rejected. The industry sector’s characteristics are related to its patents (H4) was significantly negative for both the automotive parts and machinery, and electronics and electricity industries. Finally, Hypothesis 5 that CEOs’ academic careers are related to patents was statistically significantly positive for CEOs with PhDs but not below.
4. Results and Implications
Patents are frequently considered to indicate the outputs of research and development, and they can be considered useful in valuing the output of research activities. However, there are some shortcomings to patents as an indicator. For example, some companies or individuals may not want to disclose the results of their research in order to preserve exclusive or monopoly rights to their invention or innovation. In this paper, we tested five research hypotheses; in order to elaborate on the empirical findings, we also used interviews.
Table 3 presents the empirical results from our patent propensity model estimates. In investigating Hypothesis 1, we found that InnoBiz firms being more creative in patents than general SMEs, was statistically rejected. As we mentioned previously, InnoBiz firms were designated by the government as more innovative than general SMEs based on their innovative performance, which was assessed based on the Oslo manual’s standardized list of innovation and innovativeness indicators. Thus, once an SME obtains InnoBiz certification, it is considered an innovation-driven firm, and as such, this finding was unexpected. Of course, innovation output can be measured by various aspects including patents and share of sales attributed to new products and processes. In light of the interesting findings of Holgersson’s study, the result for Hypothesis 1 is somewhat understandable. Reference [
40] conducted a survey on patenting and entrepreneurship in the United States and analyzed the responses of 1332 early-stage technology companies that had been founded since 1998. In the findings from that study, the traditional theory that patents provide an incentive to invent was subtly adjusted; surprisingly, the start-up executives’ responses indicated that patents generally provide relatively weak incentives to conduct innovative activities.
Previous research findings show that in order to boost growth, a firm needs to have good internal research organization and capacity including a research division [
28]. However, Hypothesis 2 that having a research fund and a research division are important in patent propensity in terms of SMEs’ performance was also statistically rejected with respect to having a research division. In other words, it can be assumed that firms’ research divisions may not pursue patenting but instead focus on other performance activities such as designing new product lines and creating new marketing strategies.
As mentioned above, the functions of research divisions vary and include identifying suitable strategies for diversifying activities for firm survival, strategically cooperating with other firms with respect to R&D projects, absorbing more external knowledge [
41] and building capacity to implement complementary strategies with other firms or R&D institutions [
31]. Second, it can be assumed that research divisions may not facilitate research and development but may give priority to survival strategies as do other management divisions. According to the interviews with researchers, there are a number of reasons for not seeking patent protection: the technology may not be patentable; the costs may be too high to prosecute and enforce the patent; there may also be the perception that with reverse engineering, patents afford relatively weak protection; and there can be the fear of disclosure. However, a somewhat expected result was that having a research fund and a number of researchers as control variables was positively related with patenting significantly.
Small or young firms tend to have higher patent-to-R&D ratios than large ones; one explanation could be that small firms have higher R&D yields but lower patent propensities than large firms. Another explanation could be that innovation activities in large firms are underestimated when measured with patent statistics whereas those in small firms are underestimated when measured by R&D statistics.
Hypothesis 3, that older firms show greater patent propensity than younger firms, was found to be statistically insignificant and thereby rejected. Considering the data analysis, we expected this finding given that more than 80% (157) of the younger firms that were less than ten years old had no patents. Even though young firms are more likely to invest in R&D through collaboration, the rejection of Hypothesis 3 maybe explained by unmeasured variables such as external effects. Based on the discussions with entrepreneurs and researchers, the most common reasons start-ups chose not to patent their innovations included not wanting to disclose the information; the cost of obtaining the patent, including attorneys’ fees; that competitors could easily invent around the patent; that they believed a trade secret was adequate protection; the cost of enforcing the patent, including court actions; that they did not believe the technology was patentable; and that they had no need for legal protection. The literature review showed that [
25,
42] found patent propensity to vary by firm characteristics, and the patentable-invention-to-R&D ratio varied across industries.
Hypothesis 4, that a firm’s characteristics are related to its patents, was significantly negative for both the automotive parts and machinery, and electronics and electricity industries. Based on the discussion with entrepreneurs and researchers, we found that regardless of firms’ characteristics, the reasons they did not want to patent were the cost of filing and the belief that the technology was not patentable; this finding shows that compared with industry conditions in neighboring regions such as Changwon, Gwangju is lagging behind. Our analysis of the raw data for the automotive parts and machinery industry showed that these firms originated in automotive mold manufacturing, and the mold industry has been known as the representative industry in Gwangju; in general, in the mold industry, what is important is not patenting but manufacturing with cost competitiveness. We found that core automotive and machinery parts were brought from SMEs in the Changwon area rather than Gwangju. In order to improve the automotive industry tier systems, there have recently been a number of InnoBiz firms started up with support from the local government. Electrical and electricity SMEs in Gwangju show far fewer patents than those in Changwon over the three years of the study period. However, it is interesting that electrical and electricity SMEs are notably highly productive in patenting. The local government has focused on nurturing the household robot industry in the next generation of growth engines based on the existing infrastructure, including the production technology of the biggest home appliance cluster in the country. In particular, in order to boost the household robot industry, a specialized center was established in 2011.
Surprisingly, in this study, the automotive parts and electronics and electricity industries had fewer patents. According to the information gathered through personal interviews, this finding could be attributed to KIA, a Korean car maker, and Samsung. The automotive industry is characterized by a vertically integrated structure at the center of the assembler; if a small firm as the first and/or second tier of the assembler cooperates with a car maker, this firm should follow the car maker’s requirements. Therefore, small firms appear not to implement their own managerial strategies for innovation; rather, they are likely to be subordinates of the car maker. In other words, instead of using their own innovation strategies spontaneously, SMEs mainly focus on the car maker’s requirements. That is, small firms are primarily interested in enhancing their comparative cost advantage, and as such, many SMEs are less likely to patent. Another reason for the smaller number of patents in the automotive parts industry is that assemblers require parts to be continuously produced for more than ten years once production begins, regardless of the possibility of a hiatus or cessation in car production. Nevertheless, it is very difficult to catch up with the development of new car models so that they keep in line with old products. Importantly, car makers do not like SMEs with patent propensity strategies, as revealed by the interviews. However, the appliance factory pulled out in 2008, and SMEs collaborated with Samsung to change their product lines; thus, we found no clustering effect from cooperative innovation in the empirical results.
Gwangju is located south of Seoul and has relatively few firms compared with other industrial cities in Korea; except for KIA, a large automotive firm that is representative of Korean automotive firms along with Hyundai, there are only a few firms. Since the beginning of 2000, the central government has been providing support in order to bring about a revitalization of regional industry, but in Gwangju, the most rapidly expanding industry is the photonics industry, which, owing to public policy support, has been growing rapidly since 2001. As of 2014, 364 firms and more than 8000 new employment opportunities had been created over the previous ten years. In that sense, these firms should be more innovative than other industry sectors, but we found no evidence for higher patent propensity in the photonics industry.
Managers make numerous strategic choices when attempting to capture returns from innovation investments, including a determination that the appropriation strategy is used and whether or not to patent, strategic choices that depend on the firm size among other things. A CEO’s education potentially impacts his or her abilities in three non-mutually exclusive ways. First, education could potentially contribute to the CEO’s knowledge, perspective and ability to understand technical and abstract concepts. Second, higher education could be a sign of the CEO’s intellect and ability to persevere through challenging intellectual activities. Finally, the social networks acquired in college and graduate school can be helpful professionally in the future. As discussed above, however, education is just one of the determinants of a CEO’s ability. Hence, it is a priori unclear how much impact CEO education has on firm performance.
Hypothesis 5 that CEOs’ academic careers are related to patents displayed a statistically significant positive correlation for CEOs with PhDs, but not for those with lower level degrees, a finding with important implications. First, CEOs who have a doctoral degree may tend towards innovation. In reviewing the previous studies, patents can contribute to firms’ performance through improvements in innovation, productivity and market value. In contrast, a CEO’s academic background is not related to the performance of the firm in Korea [
35], but age is related in that younger CEOs are likely to perform better than older CEOs. Younger CEOs are also more likely to have higher academic degrees (e.g., more than 90% of CEOs of venture firms in Korea have PhDs) than their older counterparts. One reason for the higher academic credentials in Gwangju is likely to be the presence of the photonics industry. In addition, it is commonly known that the photonics industry is innovative, and as such, there are many InnoBiz (designated innovative) firms in Gwangju compared with other regions in Korea.
Gwangju city has nurtured the photonics industry by promoting it as a strategic regional industry since 2000, and has established an international photonics industry cluster where industrial, academic and research functions are integrated based on a high-tech scientific industrial zone structure; here, photonics-related InnoBiz firms established by CEOs with PhDs enhance the national photonics development technology infrastructure. This resulted in the establishment of the Advanced Photonics Research Institute, which includes the Korea Photonics Technology Institute. From the empirical results of this study, we note that the research fund to turnover ratio, the number of researchers, and high-caliber CEOs are positively related to patent production and are statistically significant.
This paper is distinct from prior studies in that this empirical analysis focused on the perspective from the point of view of firms, because it is firms which are the main agent in the commercialization of technology. In the majority of previous studies, technology transfer and commercialization were analyzed according to characteristics of the university (e.g., public vs. private, size, location, research driven) or the presence of a university technology transfer office [
8,
9,
42], as well as the effects on technology transfer and commercialization by organizational function.
5. Final Remarks and Policy Suggestions
Little is known in empirical terms about patenting resulting from university-industry collaboration. Patents have been used to study a broad range of measures reflecting the performance of a firm, such as growth and survival. The ultimate policy goal with respect to industry-university collaboration (IUC) is to enhance the firm’s innovation strength. According to the empirical results it is only the CEO’s academic credentials that are found to be positively related to propensity to patent regardless of industry sector and innovation types. Patent propensity is likely to relate to the underlying managerial decision of whether to apply for patent protection for an invention or not. Empirical findings also reveal that a patent is an effective way to technology transfer from universities into firms. One reason for this might be the fact that patenting is related to technology commercialization. In a dynamic setting, to promote collaboration activities, policy supports are needed. Since the industry-university collaboration law was enacted in 2003, the promotion of industry-university collaboration has been a major regional innovation policy in Korea. Just one year after this, the policy program, New University for Regional Innovation (NURI) was enacted, with the aim of regional development stimulated by central government. Even though universities are the new knowledge creators as both partners and facilitators of small medium enterprise (SME) innovation, the key players that can directly contribute to regional development are SMEs located within specific regions. This is because the primary role of universities is to teach and to conduct basic research. In order to nurture SMEs, taking Korea’s industry ecology system into consideration, large firms are important partners for technology commercialization.
Under NURI, the roles of the university should be enlarged according to the Triple Helix Model, in that a university becomes a catalyst linking small firms and large firms. This is because according to the in-depth interview, a small or new regional venture firm can be expected to face many difficulties meeting business partners for technology commercialization. Based on data from the in-depth interviews, what is desperately needed in launching the market for SMEs on a regional basis is not only technology transfer, but support for the new policy, such as sharing risk systems (e.g., information about new markets, market trends and market competition conditions).
It is a time of incredible technological change. SMEs look to the source of new knowledge for commercialization each day. Knowledge transfer in industry-university collaboration can be either formal or informal [
43,
44]. While patenting propensity is likely to depend on the formal channel of knowledge transfer (i.e., contract), informal channels (i.e., personal contacts) between industry and academic researchers also have the potential to influence commercialization [
45]. Universities have conspicuously expanded their missions from their original task of preserving knowledge to producing new technology and developing and disseminating new knowledge to the firms for their use. Knowledge sharing between universities and firms is critical to keep sustainable growth for both.
Technology commercialization often demands collaboration between two or more functional units that are separated by structural, cultural and organizational boundaries. The author suggests intellectual property rights alliances (IPR ALLIANCE) among universities as a bridge connecting university and industry. IPR alliance, offering patent strategic services for small firms without sufficient capabilities using the innovation outputs, can be an important hub for commercialization partners with firms. IPR Alliance can play a strategic win-win role for both universities and industries. It can be a new regional innovation ecosystem.
One of the limitations of this study was the narrow scope of patents used as a means of unveiling the knowledge from university-industry collaborations, despite the wide variety of methods associated with knowledge transfer and commercialization. Patenting propensity is also more likely to depend on managerial aspects. Even though a firm has new inventions instead of patenting, know-how or secrecy also can be associated with knowledge transfer. However, this may act as a stimulus for additional research directed specifically towards the commercialization processes relating to universities.