University research is a catalyst for business sector innovations, according to several empirical studies (see for example, Agrawal 2001). Universities have a potential impact on firm innovation in a variety of ways including publication of fundamental research, university patenting and licensing, spin-offs and university incubators and science parks, joint research with firms, consultancy projects, the training of students and employees etc. As emphasised in a recent report by the Research Councils UK (2010), the capacity of university research to positively impact on private firm performance and innovative activity is an important argument in favour of public funding of university research in the UK.
From a policy perspective, a critical question is how to allocate money for university research. This involves choices over both how many universities should be funded and also the distribution of funding across universities. Should research be focused on the “best” universities and, of course, how should one determine the “best”? Is it important to ensure university research is spread across a wide range of regions? In other words, does the impact of knowledge transfer depend on the “quality” of research and is its effect geographically local?
Answers to these questions have direct implications for the allocation of public funding of university research. If “high-quality” research-intensive universities have a larger effect on innovation by private companies, regardless of the university’s geographical location, scarce public funding might be best allocated to a few world-class, research-oriented universities. Such a conclusion would be misplaced, however, if knowledge transfer were localised and the link between research “quality” and successful knowledge transfer was less pronounced.
In a recent paper (Helmers and Rogers 2010), we explore these questions empirically by analysing the impact of university research on the patenting activity of manufacturing firms in the UK.
Methodology: Finding a link between patents and universities
Our analysis considers the association between the number of patenting firms (as a measure of innovative firms) and the characteristics of relevant research in nearby universities. We break the UK down into 117 different (postcode) areas. For university research, we concentrate on engineering and biological science departments, looking at both quantity and quality. Quality is assessed according to the grade received in the 2001 Research Assessment Exercise. The data relates to research activity over the period 1996-2001, which allows for a considerable time lag in any effect of university research on patenting, which is measured for 2001. We distinguish between small firm versus large firm patenting. “Small firms” are defined as micro firms or small and medium enterprises. Our hypothesis is that geographical proximity to universities should matter less for large, sophisticated firms with access to wide networks, both nationally and internationally. In contrast, the impact of universities on smaller firms within a postcode region is likely to be important.
The analysis accounts for a range of other possible determinants of patenting in the postcode area, such as research and development, the numbers of firms and the proportion of skilled workers. We also investigate the possible channels of channels of technology transfer by using survey data direct from universities.
Our main finding is that the quality and quantity of engineering and biological university research is positively associated with small-firm patenting. We do not find any statistically significant correlation between university research and large-firm patenting. In contrast, the positive association of university research on small-firm patenting obtains for a range of different measures of university research, including the number of departments and the number of researchers. Moreover, we find evidence that research quality does matter. The departments with the top two research grades (grade 5 or 5*) drive the positive association with the number of small-firm patentees.
Our analysis of specific channels for knowledge transfer suggests that both formal and informal channels serve as a way of knowledge transfer, although the significance of specific channels differs between engineering and biological research. It also suggests that transfer channels that involve physical facilities to host companies in university proximity, i.e., science parks and incubators, may also play an important role in knowledge transfer.
To date, much of the evidence concerning the impact of university research on local innovation has been case study based (e.g. NESTA 2009). Our results provide evidence from aggregate, UK-wide data as a complement to such studies. The main policy conclusions are:
- Local university research has a positive association with the number local small firms that patent.
- The higher the research quality, the stronger this research-patentee association.
- Large firm patenting does not appear linked to local university research; instead large firms are likely to gain from national and international research networks.
Policy decisions concerning the concentration of research in selected universities, or regions, may thus have important effects on innovative small firms.
Agrawal A (2001), “University-to-industry knowledge transfer: literature review and unanswered questions”, International Journal of Management Reviews, 3(4):285-302.
Helmers C and M Rogers (2010), “The Impact of University Research on Corporate Patenting”, SERC Discussion Paper No. 54.
Kitson M, J Howells, R Braham and S Westlake (2009), “The Connected University: Driving Recovery and Growth in the UK Economy”, NESTA Research Report, April.
Vaitilingam R (2010), “Research for our Future: UK business success through public investment in research”, Research Councils UK.