Inside or Outside the IP System? Business Creation in Academia Scott Shane (CWRU)
Academic Entrepreneurship, Innovation, and Policy Academic research is a key engine of economic growth and competitive advantage But university research often distant from real economic needs Since academics respond to economic incentives Adopt policies to facilitate knowledge transfer and commercialization the Bayh-Dole story
Trends in Academic Entrepreneurship New U.S. Patent Applications Filed Licenses/Options Executed 12,000 6,000 10,000 5,000 8,000 4,000 6,000 3,000 4,000 2,000 2,000 1,000 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year Start-ups established Gross License Income Received 600 500 Milions 1,400 1,200 400 1,000 300 800 600 200 400 100 200 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year 0 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 Year Source: AUTM (1996-2006)
The Situation in 2006 Activity USA (AUTM N=161) Licenses and options executed 4,038 Startups 484 Invention disclosures 6,384 New US patent applications 10,183 License income ($ 000) 1,249,082,798 Source: AUTM (2006)
Academic Entrepreneurship is a hot topic 6000 5000 "University technology transfer" "Minimum wage" "Behavioral finance" "Agency theory" 4000 3000 2000 1000 0 1997 1999 2001 2003 2005 2007 Year Source: Google scholar
The venture creation process in academia Research Invention Disclosure to TTO Patent License or Start up based on Patent So polices are based on presence of formal IPRs, particularly patents But formal IPRs neither effective nor used in many sectors, leading lots of commercial, knowledge transfer activities by academics not to be patented Nevertheless, we study what we can measure Academic research and public policy evaluation focuses on IP-based measures of entrepreneurship (patents, licensing, spinoffs)
What Previous Research Tells Us Increase Academic Entrepreneurship Individual Age Being male Higher academic rank Being in biological sciences and engineering Emphasize research; high publication rate University Institution quality Amount invested in R&D Ties to industry Located in a major city with venture capital activity
Questions Just how big is academic entrepreneurship outside the IP system? How does inside the IP-system entrepreneurship differ from outside the IP system entrepreneurship? Are IP-based policies appropriate for non-ip-based forms of academic entrepreneurship? Is the extent of academic entrepreneurship systematically underestimated, and current analyses distorted?
E-survey to 58,321 tenured or tenure track faculty members and post docs at all Carnegie I and II during second half of 2007 Voluntary participation Four follow-up electronic messages sent to non-respondents 11,572 responses (20%) Data and Research Design
Data and Research Design Questions Demographics: Gender; Age; Academic rank; Experience + Uni., Dept. Academic activities in 2006-2007: Percentage of time in research, teaching; interaction with industry; research funding; cumulated publ. record Commercial activities: Invention disclosures, patents issued, licenses Equity New businesses started Overall On patent And not based on patent
Data and Research Design Additional info on universities: Carnegie Foundation + AUTM + US News Ownership, Age, Size, Research Expenditures, Localization, Tech-transfer support mechanisms, TT outputs, Ranks/Tiers Use this info to analyze peculiarities of IP and No-IP based business creation in academia Wealth of info at multiple levels, big sample size But......Selection?...Cross-sectional data?
Data and Research Design Selection: Limited on school, tech area, gender, age Sample representativeness cdf 0.2.4.6.8 1 cdf 0.2.4.6.8 1 0 50 100 150 200 School codes 1 2 3 4 Tech codes Respondents Non Respondents Respondents Non Respondents density 0 5 10 15.5.6.7.8.9 1 Prob(resp)_probit Respondents Non Respondents Resp. vs. 1,000 random non-resp. Prob(resp) = f(school, tech. area, sex, age). Use inverse of est. prob. as IPW in regressions
Data and Research Design Cross sectional data Descriptive study But value in new info unveiled, for policy and research
Magnitude A lot of academics start businesses - 16 percent of the sample have done so Most do not start businesses based on a patented invention two outside the IP-based system start-ups for every one inside the IPbased system start-up Researchers and policy makers are missing a lot of the academic entrepreneurship
Commercialization and the underground entrepreneurs In conjunction with Variable Tot Only activity Disclosure Patent License Equity New business on patent Disclosure 2,950 793 US Patent issued 2,166 99 1,822 License (on US patent) 1,146 6 988 1,071 Equity 957 45 691 611 409 New business on patent 682 12 560 560 358 493 New business not on patent 1,266 623 432 314 172 343 138 Correlation table Variable Disclosure Patent License Equity New business on patent Disclosure 1 US Patent issued 0.67 1 License (on US patent) 0.46 0.64 1 Equity 0.31 0.34 0.33 1 New business on patent 0.37 0.41 0.37 0.59 1 New business not on patent 0.07 0.05 0.04 0.24 0.06
Look at Differences Predict start-ups, start-ups based on patents, start-ups not based on patents, and start-ups based on patents versus start-ups not based on patents Linear probability models with robust standard errors Inverse probability weighting Include variables about individuals age and gender, source of funding, publications, time allocation, academic field, university rank, location, university R&D, age and size of technology transfer office, and whether university is public or private.
So What s Different? Inside the IP-System Entrepreneurs are: More likely to be female Younger Publish more More likely to raise money from industry Spend more time on research Spend less time on teaching More likely to be located in the Northwest Less likely to be in a large town More likely to be from a higher ranked university More likely to be in a biomedical field
New businesses and Gender Fraction of academics who have started a business on patent, by gender 0.05.1.15.2 Fraction of academics who have started a business, by gender Female Male 0.05.1.15.2 0.05.1.15.2 Female Male Fraction of academics who have started a business not on patent, by gender Female Male
New businesses and Age Fraction of academics who have started a business on patent, by age group 0.05.1.15.2 Fraction of academics who have started a business, by age group Age20-29 Age30-39 Age40-49 Age50-59 Age60-69 Age70-79 0.05.1.15.2 0.05.1.15.2 Age20-29 Age30-39 Age40-49 Age50-59 Age60-69 Age70-79 Fraction of academics who have started a business not on patent, by age group Age20-29 Age30-39 Age40-49 Age50-59 Age60-69 Age70-79
New businesses and University rank Fraction of academics who have started a business on patent,, by US news tier 0.05.1.15.2 Fraction of academics who have started a business, by US news tier 1-20 21-50 51-100 >100 0.05.1.15.2 0.05.1.15.2 1-20 21-50 51-100 >100 Fraction of academics who have started a business not on patent,, by US news tie 1-20 21-50 51-100 >100
New businesses and Technological areas Fraction of academics who have started a business on patent, by technological area Engineering Bio, Med 0.05.1.15.2.25.3 Fraction of academics who have started a business, by technological area Engineering Bio, Med Social Sc., Human. Math, Phys. Stats. Social Sc., Human. Math, Phys. Stats. Fraction of academics who have started a business not on patent, by technological ar Engineering Bio, Med Social Sc., Human. Math, Phys. Stats. 0.05.1.15.2.25.3 0.05.1.15.2.25.3
New businesses and Location Fraction of academics who have started a business on patent, by location Fraction of academics who have started a business, by location Large city Large town Fraction of academics who have started a business not on patent, by location Large city Large town Mid-size city Not assigned Rural Small town Large city Large town Mid-size city Not assigned Rural Small town 0.05.1.15.2 0.05.1.15.2 0.05.1.15.2 Mid-size city Not assigned Rural Small town
New businesses and Academic activities: research Average % of time spent in research for respondents starting a business on patent 0 10 20 30 40 50 Average % of time spent in research for respondents starting a business 0 1 0 10 20 30 40 50 Average % of time spent in research for respondents starting a business not on patent 0 10 20 30 40 50 0 0 1 1
New businesses and Academic activities: teaching Average % of time spent teaching for respondents starting a business on patent 0 10 20 30 40 50 Average % of time spent teaching for respondents starting a business 0 1 0 10 20 30 40 50 Average % of time spent teaching for respondents starting a business not on patent 0 10 20 30 40 50 0 0 1 1
Implications We are missing more than we are measuring: two thirds of academic entrepreneurship is off the radar of researchers and policy makers Outside the IP system academic entrepreneurship is different from within the IP system academic entrepreneurship across a number of dimensions Individual characteristics Technology/discipline School rankings School location Approach to academic job We have a problem in both research and public policy that needs to be fixed Understanding impact of academic entrepreneurs? Revisiting what we think matters? Adjusting Federal legislation? Changing university technology transfer policies?