Firms started by university graduates as commercial spin-offs

advertisement
THE EFFECTIVENESS OF UNIVERSITY KNOWLEDGE
SPILL-OVERS:
PERFORMANCE DIFFERENCES BETWEEN UNIVERSITY
SPIN-OFFS AND CORPORATE SPIN-OFFS
Karl Wennberg
Stockholm School of Economics & Ratio
Johan Wiklund
Syracuse University
Mike Wright
CMBOR & EMLyon
Context
The Bayh-Dole Act of 1980
“A wealth of scientific talent at American colleges and
universities – talent responsible for the development of
numerous innovative scientific breakthroughs each year –
is going to waste as a result of bureaucratic red tape and
illogical government regulations…What sense does it
make to spend billions of dollars each year on
government-supported research and then prevent new
developments from benefiting the American people
because of dumb bureaucratic red tape?”
U.S. Senator Birch Bayh, 1980
Unversities own the Intellectual Property (IP) generated
through university research
Bayh-Dole implications
• Creation of Technology Transfer Offices (TTO)
• Explosion of university patent activity
University Patents as a Share of All Patents with Domestic Assignees
0.04
0.035
0.03
0.02
0.015
0.01
0.005
University Patent Issue Year
(Mowery 2005)
19
96
19
94
19
92
19
90
19
88
19
86
19
84
19
82
19
80
19
78
19
76
19
74
19
72
19
70
19
68
19
66
19
64
19
62
19
60
19
58
19
56
19
54
19
52
19
50
0
19
48
Share %
0.025
How does knowledge spill over from universities?
Knowledge spill-over through
entrepreneurship?
• AUTM reports annual mean of 426 startups
from U.S. Universities, 1998-2004
• MIT TTO reported 29 startups (2001)
• Stanford TTO reported 6 startups (2001)
• Based on AUTM data, one startup generated
per $368 million of R&D
Broader perspective needed
Our paper
How does knowledge spill over from universities through
entrepreneurship?
Steve graduates from Stanford, works at Stanford, spins off a company
Lisa graduates from Stanford, works at Fairchild, spins off a company
To date we have focused on the Steves of the world but not the Lisas
Unfortunate
Much entrepreneurship is probably carried out by the Lisas
Even if we are mainly interested in the Steves, the Lisas provide a
relevant baseline for comparison
We suggest both Steve and Lisa represent university
knowledge spill-over through entrepreneurship
They differ in terms of their experience
Both Steve and Lisa hold unique knowledge that can be
valuable for knowledge-intensive entrepreneurship
Ranked #1 worldwide
forentrepreneurship
publications in
academic
entrepreneurship
Which path to
is more
productive?
Ranked #1How
worldwide
for publications
in academic
entrepreneurship
does experience
inside/outside
university influence
subsequent spin-off performance?
Ranked #1 worldwide for
publications
academic entrepreneurship
University
spinoffs in
(USOs)
Corporate spinoffs (CSOs)
Individuals with University Degree
University
Company
Entrepreneur
Performance
Gap
• Academic entrepreneurship literature emphasizes venture
creation
– Not subsequent performance
• Little attention to indirect spillover from university to
entrepreneurship
– CSOs provide alternative mechanism to USOs
• Policies to stimulate USOs has not used suitable baseline
• Comparison of USOs to CSOs relevant
Performance differences
• Potential for successful entrepreneurship depends
on technological and market knowledge
• Prior experience of customer relationships essential
to generate market knowledge
– More likely in commercial environment than university
• CSOs have greater exposure to commercial
interaction and thus greater market knowledge
• H1: Firms started by university graduates as
commercial spin-offs perform better than firms
started by university graduates as university spinoffs.
Knowledge Sources
• Schooling: Increases general human capital
• Prior industry experience: provides relevant product related
knowledge
• Previous entrepreneurial experience: Specific knowledge of
running a business
• CSO founders complement university knowledge with market
knowledge unlike USO founders
• Prior industry experience, schooling and entrepreneurial experience
can compensate for lack of market knowledge
• This compensation more valuable to USOs than to CSOs
• H2: Firms started by university graduates as university spin-offs
benefit more from the knowledge sources [a] years of education,
[b] years of industry experience in the same sector, [c] years of
entrepreneurial experience, than firms started by university
graduates as commercial spin-offs.
Parent organizational context
• We build on the spawning literature
• Examines entrepreneurship depending on parent organization
characteristics
• spin-offs usually inherit both general technological and
market-related knowledge from their parents
• We suggest there are important differences between
university context and corporate context that influences spinoff performance
Parent organizational context
• University technological knowledge broad but far from market
• Corporate technological knowledge narrow and closer to market
 CSOs more affected by parent knowledge
• Universities generally bureaucratic
 USOs less affected by parent organization
• Large corporations especially those with many divisions have
spawnable activities that are peripheral and difficult to control and
incentivize yet which may have good underlying performance
prospects
• H3: Firms started by university graduates as commercial spin-offs
benefit more from the spawning environments of the parent
organization (size and breadth of technological knowledge) than
firms started by university graduates as university spin-offs
Research design
•
•
•
•
Symmetric data on USOs and CSOs
Follow firms from inception and onwards
Performance data that are robust across industries
Data from Sweden
– Complete data
– IP remains with inventor for USOs
Data I
• Official data maintained by Statistics Sweden
• Annual observations, great detail
• All private incorporated companies started in Sweden
1994-2001 in knowledge-intensive sectors
– Excluded other sectors and legal forms
– Ensures relevance
• Match with individual data on founders
• Individuals completing university degree lasting at least
3 years in any field
• Worked for university or a private company at least
some time in 1993-2000
• Leave employer at T and start firm at T+1
– USO if employer = university
– CSO if employer = private firm
Data II
• Must hold majority stake in first year and work there
full time.
• Team spinoffs require that 50% or more work at same
employer at T
• Identified 528 USOs (6%), 8,663 CSOs (94%)
Dependent variable
• Three different indicators of performance
– Employment growth
– Sales growth
– Survival [discontinued, excluding M&A]
Independent Variables
• Design allows detailed assessment of human capital
– Entrepreneurial experience (team mean)
– Same industry experience (team mean)
– Education length (team mean)
• Parent organization
–
–
–
–
–
Revenues
Employment
# establishments
Share of employees with engineering degree
Share of employees with PhD
Control Variables
• Team size
• Capital invested
• Social capital (prior years in location where business was
started;
• Industry affiliation (ISIC 3-digit level)
Industry of spinoffs
Academic Spinoffs
Corporate Spinoffs
Chemicals and fibre manufacturing
0.30%
0.19%
Electrical and optical equipment
2.37%
3.42%
Transport equipment
0.30%
0.12%
Networks, radio and TV
2.96%
2.27%
Finance
4.44%
4.65%
Real estate business
0.31%
12.07%
Computers/software
19.53%
16.25%
Research and Development
11.54%
1.18%
Accounting / auditing
3.25%
6.16%
Construction / engineering
5.03%
11.09%
Advertising
3.25%
7.78%
Management consulting
19.53%
14.30%
Law firms
1.18%
5.03%
Other consulting services
3.25%
6.82%
Education
13.30%
0.02%
institutions (elderly/children/care)
0.30%
1.64%
Private Health care
4.14%
5.14%
News and entertainment
5.03%
1.86%
Analyses
• Random effects generalized least squares regression for
growth DVs
• Cox regression for survival
H1
H2
H3
CSOs outperform USOs
Dependent variables
Emp
Survival
Sales growth
growth
Supported
Supported Supported
(a) Education matters more for USOs
not supported
not
supported
not
supported
weak
support
(b) Ind Exp matters more for USOs
Supported
Supported
(c) Ent Exp matters more for USOs
Supported
Supported
Supported
Employment size matters more for
CSOs
Supported
Supported
weak support
Revenue size matters more for CSOs
Supported
Supported
Supported
# Establishments matters more for
CSOs
Supported
Supported
Supported
# Engineers matters more for CSOs
weak support Supported
Supported
# PhDs matters more for CSOs
Supported
not
supported
not
supported
Conclusions & Implications
• CSOs outnumber USOs 14 to 1 and perform better
– Universities educate people who become entrepreneurs but mainly
indirectly
– CSOs seem to be good mechanism for knowledge-intensive
entrepreneurship
– Why assign policies that single out and support USOs?
• USOs benefit more from entrepreneurial and industry
experience
– Connecting USO entrepreneurs to experienced people could be
important task for TTO
– Do TTOs have expertise to build these teams?
• Parent size has positive effect on CSO performance
– Large firms seems like good seed bed for spinoffs
Policy Implications
• Implications for IP policy
– Broader view of university knowledge spillover
• Assist USOs in building viable teams with commercial
experience to succeed
– TTO competence needs augmentation
•
Wider policy debate on how knowledge spills over into new
organizations
– Policy support to facilitate spinning out from established
organizations?
Policy Implications
• Why exclusive policy focus on entrepreneurs who work at
universities?
– Realizing growth among knowledge-intensive firms involves general
problems that apply to the whole population of spinoffs
– Policy favoring the establishment of growth-oriented entrepreneurship
in general, not just USOs?
Download