Since the Bayh-Dole Act of 1980, universities that receive federal research funding have been encourage to patent their research and license it to private industry. The benefits of the patenting rule are still very much in dispute, but surely everyone can agree that the spread of academic research into real firms has been a net positive? At many universities, faculty are even explicitly encouraged to commercialize their research, through start-ups and other means.
Astebro et al point out that something is missing here, though. Google is a university spinoff, but spun off by two PhD students at Stanford. Facebook and Microsoft are university spinoffs, but spun off by undergraduate students. Indeed, faculty don’t seem to be very entrepreneurial at all: the median top-100 US university in any given year has zero such spinoffs.
Using data from a (somewhat) longitudinal survey of university undergraduates, along with existing faculty spinoff data, the authors point out that students, within a few years of graduating, are twice as likely as faculty during that period to form their own business. Since there are so many more students, this means that 24 times more startups come from recent university grads than from faculty. And these are not low-quality startups or disguised unemployment: 36 percent of the businesses are still around in a follow-up survey two years later. Earnings from these startups is particularly strong for those students who claim their business is in an area related to their studies, and for students at research universities in the NRC top 10 for doctoral research.
There is very little in the way of identification in this paper, so read this as only a first go through the data. Nonetheless, the basic point is clear: if you are a region who wants to leverage universities for new business growth, developing better and more entrepreneurial students seems to trump encouraging faculty to run businesses. Indeed, to the extent that faculty create the human capital these students will use in their businesses, such faculty-biased policies may be counterproductive. The authors discuss a couple case studies, including the interesting E-school program at Chalmers in Gothenberg, Sweden, which perhaps provide a way forward here.
A broader takeaway, which hopefully is well-known already: the link between urban policy and invention/entrepreneurship policy is ridiculously important.
http://www.andrew.cmu.edu/user/sbrag/ABB.pdf (Draft – final version in Research Policy 41.4 (2012). No IDEAS page available.)