2004/03 Research & Review

University-Industry Collaboration and the Importance of R&D-focused Small and Medium-sized Enterprises - Their Implications on Japan's Innovation System

Senior Fellow, RIETI


The Second Science and Technology Basic Plan, which outlines the basic direction of the nation's science and technology policies for the period spanning 2001 to 2006, presents a strong argument for the need to revamp Japan's innovation system to encourage companies to become more innovative. The promotion of university-industry collaboration is a core part of such an overhaul, and systemic reforms to promote relationships between universities and public research institutions and private companies in the field of research and development are being rapidly implemented. For example, the enactment of the 1998 Technology Licensing Organization Law has prompted the establishment of TLOs in universities and public research institutions, while the Law to Strengthen Industrial Technology Capacity enacted in 2000 includes measures that ease restrictions on the work university researchers can do outside their institutions and expand ways in which national universities can receive funding from the private sector.

In response to these recent policy measures, both universities and private firms are actively pursuing university-industry collaborations. However, such moves still lag behind those in the United States, which embarked on creating the systems necessary for such collaboration in the 1980s. The main reason for this is the nearly 20-year time lag between the actions of the two governments, but I believe it is also largely the result of the differences in innovation systems between Japan and the United States. In other words, cross-organizational R&D, including university-industry collaboration and venture businesses in universities, is more easily done in the United States due to such factors as capital markets that have been developed with venture capital and other means, and a high degree of labor mobility. In Japan, on the other hand, large companies which have their own research capabilities have spearheaded R&D efforts, and this sort of "do-it-yourself" attitude has been pointed out as being one obstacle to enhancing university-industry collaboration (Motohashi, 2001).

In order to accurately assess the rapid developments in the area of university-industry collaboration in recent years, RIETI conducted a Survey of University-Industry Collaboration Activities in February 2003. The results of this survey can be viewed on RIETI's website (RIETI, 2003), but in this column I would like to present the results of a more detailed quantitative analysis using data obtained from the survey. In terms of the Japanese innovation system mentioned above, it is important to focus on the role played by R&D-focused small and medium-sized enterprises. The "do-it-yourself" approach often seen at large companies is not a phenomenon unique to Japan. Indeed, there are those who point to the problems brought about by "NIH (Not Invented Here) Syndrome" in the United States. The major difference between the innovation systems of Japan and the United States is that venture businesses and R&D-focused SMEs play major roles in the latter. In this column, I would like to compare the role played by R&D-focused SMEs in university-industry collaboration in Japan with that played by large companies.

University-Industry collaboration is spreading to SMEs

RIETI's survey on university-industry collaboration activities covered all firms polled in the Ministry of Economy, Trade and Industry's Basic Survey on Business Structure and Activities (all companies engaged in manufacturing or wholesale and retail that have at least 50 employees and capital of ¥30 million or more) that conduct in-house R&D. The three main pillars of the RIETI survey are: (1) the extent to which the firms surveyed collaborate with other parties such as universities, public research institutions and private firms involved in R&D activities; (2) details of their collaboration with universities, including the number of joint research projects and the budgets allocated; and (3) qualitative issues such as their evaluation of university-industry collaboration and the problems they face. Here, I will conduct an analysis using linked data on the corporate level from this survey and the Basic Survey on Business Structure and Activities, which provides useful information concerning business performance.

First, according to results of analysis on the determinants of university-industry collaboration (the defining characteristics of companies that engage in university-industry collaboration), while large companies had been the main players in such collaboration, in the course of the past five years these moves have spread to SMEs. Looking at the details of their collaboration efforts, we see that there are only a few cases in which the relationship takes the form of technology and know-how transfer or patent licensing, and in most cases it is joint research. In other words, the scientific knowledge of universities is not something that companies can receive and utilize right away, but requires additional R&D at the firms with the assistance of the universities. Firm's decision to take part in UIC activities is not based on "make" or "buy" of university's knowledge but UIC activities require co-development based on it. In order to effectively proceed with university-industry collaboration, a company itself must have sufficient research resources, and as a result, large companies that have an abundance of such resources can utilize university-industry collaborations more effectively.

Looking at the analysis results, large companies with an abundance of in-house R&D resources were especially active in university-industry collaboration some five years ago. However, more recently, while the effect of business scale remains statistically significant as a determinant for such collaboration, its significance has slightly lessened. In addition, estimates that also include corporate age as a determinant show that the effects of business scale disappear and the younger the firm, the more active it is in university-industry collaboration. As this shows, the effect of corporate scale as a determinant for university-industry collaboration is gradually declining. To clarify this point further, when we look at the characteristics of firms that have launched university-industry collaboration projects in the past five years, we find no significant coefficient for the effects of corporate size. On the other hand, firms that do not have their own research facilities have a strong influence. Having in-house research institutions not only indicates that there is a substantial amount of R&D taking place within the firm, but also that the company's strategy is to handle even comparatively basic research with internal resources. In doing so, universities, which conduct basic research, could become a competitor with the firm's own research institute, and in recent years the effect of this competition may have been becoming more influential as a determinant for university-industry collaboration. In other words, as a recent trend, even SMEs that do not have sufficient in-house research resources are actively undertaking basic R&D through university-industry collaboration.

University-industry collaboration has a major impact on SMEs

Here I will discuss the effects of university-industry collaboration on corporate performance. First, I looked at the effects of such collaboration on corporate innovation activities. Innovation output can take such forms as the development of new products (product innovation) and an improvement in efficiency through new production methods (process innovation), but here I used the number of patents as a standard of measurement. For input, I included R&D budgets, whether R&D was outsourced and whether there was any university-industry collaboration and evaluated the effects of such collaboration on innovation productivity. The results showed that university-industry collaboration has a positive effect on innovation productivity. This analysis was also conducted on each of three groups of companies that were divided according to age. As a result, the analysis showed that the younger the company, the greater the effects of university-industry collaboration.

I also conducted an analysis of the impact university-industry collaboration has on production productivity, estimating the production function by using a company's value-added as output and items such as labor, capital, R&D investment and the existence of university-industry collaboration as input. The results showed that university-industry collaboration also has a positive influence on productivity as measured by the production function. Furthermore, as in the case of the analysis on the impact on innovation productivity, the effects of university-industry collaboration were greater for younger firms.

As this shows, university-industry collaboration is spreading to smaller and younger companies that are focused on R&D, and its impact on innovation and productivity is greater at these firms. The rise of SMEs with high innovation performance can also be seen in a detailed study by Toshihiro Kodama (2003) of the product development-focused companies in the TAMA (see note). The role of such regional networks is probably important in university-industry collaboration by SMEs, but data analysis on this point is an issue I would like to study in the future.

Japan's innovation system - implications on reforms

Research-oriented SMEs that are young and small in size can secure high productivity in R&D activities and production activities through university-industry collaboration. These companies are probably actively forging external alliances and aiming to join collaboration projects that are closer to commercialization, such as the development of new products, because they do not have an abundance of funds or human resources compared to large corporations. However, it is important to note that this high level of productivity is only relative compared to the same types of small businesses that do not undertake university-industry collaboration. Therefore, not all SMEs have seen success through university-industry collaboration, and it is more natural to interpret that for companies that do not have an abundance of management resources, firms that took the high risk of undertaking university-industry collaboration were rewarded with high returns as a result.

R&D-focused SMEs that take such risks and proceed with external R&D alliances such as university-industry collaboration not only have great growth potential but can also serve as a catalyst to alter Japan's innovation system, which is currently focused around large companies. The diagram on the left shows the typical position of R&D-focused SMEs in reforms of Japan's innovation system.

Diagram: The UICs of R&D-focused SMEs and their role in innovation reform

Japan's innovation system has systemic obstacles such as a rigid labor market and an underdeveloped technology market. Because of this, there have been few efforts to pursue external partnerships for R&D and as a result, large companies have played the central role through innovation using their own R&D resources. However, Japan may fall behind the innovation competition in fields of rapid technological advances, such as the information technology revolution, if it sticks to this sort of do-it-yourself approach. In addition, the R&D process for pharmaceuticals has greatly changed as a result of advances in biotechnology, and it has become important to create effective alliances with institutions like universities, which have scientific knowledge such as gene function analysis. As this shows, in such high-tech areas as IT and biotechnology, network-type systems that place priority on outside alliances are coming to gain a comparative advantage in innovation.

Because R&D-focused SMEs are not blessed with R&D resources like large companies, they have a strong incentive to overcome systemic obstacles and move toward university-industry collaboration that can lead to specific results such as the development of a new product. Furthermore, from the universities' side, compared to large companies, which lean toward basic research seeds, it is more likely that tie-ups with SMEs will satisfy their incentive to commercialize the results of their research. Such vigorous university-industry collaboration by R&D-focused SMEs that overcome systemic obstacles has a great chance of changing the entire Japanese system to make it more fluid. In recent years, we have observed progress in the university-industry collaborations of small, young firms, but this sort of R&D-type collaboration between universities and industry is also very beneficial for society, and should be pursued further on the policy front.

For details, please refer to "Economic Analysis of University-Industry Collaborations: The Role of New Technology Based Firms in Japanese National Innovation Reform."

>> Original text in Japanese

  • The Technology Advanced Metropolitan Area spanning southwest Saitama Prefecture, the Tama area of western Tokyo and central Kanagawa Prefecture.

August 11, 2004