Academic Entrepreneurs: Social Learning and Participation in University Technology Transfer

1
Academic EntrepreneursSocial Learning and
Participation inUniversity Technology Transfer

• Janet Bercovitz
• University of Illinois
• Maryann Feldman
• University of Georgia

2
Changing Environment for University-Industry
Relationships

• Universities Have Long Served as a Source of
  Scientific and Technical Knowledge
• Recent Environmental Changes. . .
• Emergence of New Technology Platforms
• Greater Knowledge-Based Competition
• Legislative Mandate -- Bayh-Dole Act of 1980
• Greater Budgetary Uncertainty
• Have Catalyzed a Shift in Emphasis
• Open Dissemination of Knowledge
• Commercialization of Academic Discoveries

3
(No Transcript)
4
University Technology-Transfer Process

• Inventor is a Faculty Member
• Eureka Moment!
• Faculty Files Invention Disclosure
• Federal requirement
• Low cost procedure, 0n-line forms
• Technology Transfer Office Evaluates
• Is it new? Useful? Non-obvious?
• If yes, then patent
• If Patent, then the Desired Outcomes
• Licenses
• Licensing revenues
• Start-up companies
• We care about outcomes, but they are predicated
  on faculty disclosing inventions

5
Results are Not Uniform

• Overall, a significant increase in the level and
  formalization of knowledge transfer activities at
  the university-industry interface
• However, there remains great variation in
  technology transfer activity across and within
  universities
• Why do some entities perform better than others?
• Not resources
• Not organizational initiatives
• Not incentives

6
Fundamental Question

• How Do Organizations (Places) Change?
• Change as an emergent rather than calculated
  phenomenon
• Collective rather than individual process
• Individual in context
• Localized learning?
• non-pecuniary sharing of information
• Groups of individual agents as conduits for
  organizational change
• Social Actors in the Geography of Innovation

7
Getting Faculty Invention Disclosures

• Seemingly Straightforward
• Its the law
• Articulated university goal
• Just about anything can be disclosed
• But, In Practice, Has Proven Difficult
• Only a subset of research with commercial
  potential is disclosed
• Perceived Barriers
• Basic research is not amenable (wrong)
• Risk of publication delays (wrong)
• Just not appropriate older norms of science
• Invention Disclosure Measures Adoption of Change
  to Entrepreneurial Behavior

8
Disclosures are Differentially Concentrated
within Medical School Department
9
Within Department Variation in Disclosure
10
Central Research Question

• What factors influence an individual faculty
  members disclosure decision?
• Technical Opportunity?
• Financial Incentives?
• Social Imprinting?
• Social Learning?
• What happens when individuals face dissonant
  situations?
• Lack of alignment
• Symbolic behavior
• Academic Entrepreneurship to study organizational
  change
• Understand Individual decision making in context

11
Imprinting Entrepreneurial Activity

• Training Effects
• An Individual is Shaped by the Norms and Values
  Prevalent
• In Key Social Institutions (Schein, 1985
  DiMaggio and Powell, 1983)
• During Formative Stages of Career Development
  (Ryder, 1965)

Training Institution Active in Tech-Transfer
H1 ()
Likelihood of Disclosure
Completed Training Recently
H2 ()
12
Social Learning Entrepreneurial Activity

• Individuals Learn How to Behave in Organizations
  by Observing the Behavior of Referent Others
  (Bandura, 1986)
• Leaders
• Build/Define Culture
• Act as Role-Model
• Peers
• Information Source
• Influence Decisions

Leader is Active in Tech-Transfer
H3 ()
Likelihood of Disclosure
Peers are Active In Tech-Transfer
H4 ()
13
Data

• Observation Individual Faculty Member
• Duke University and Johns Hopkins University
• Both late entrants in technology transfer
• Strong Medical Schools
• Same financial incentives at time under
  consideration
• Fifteen Matched Medical School Departments
• Basic, Nexus, and Clinical Departments
• Departmental fixed effects
• Research is expected from all faculty members
• 1779 Individuals
• Administrative Records
• Technology Transfer Office Database
• ISI Publications

14
PROBIT Model

• Two Period Model
• Dependent Variable
• Three-Year Window Academic Years 1996-1998
• Disclosure Activity Dummy Variable
• Independent Variables
• Independent variable individual characteristics
  and local context
• Activity in Previous Five-Year Window Academic
  Years 1991 1995
• Controls

15
Control Variables

• Quality
• Individual NIH Awards
• Departmental NIH Awards
• Number of Prior Disclosures
• Inventive Capacity
• Boundary Spanning
• Dual Degree
• Number of ISI publications
• Non-US Degree
• Type of Department (clinical omitted)
• Nexus Service Department
• Basic Science Department
• Academic Rank (Associate Professor omitted)
• Full Professor
• Assistant Professor
• University dummy variable

16
The Likelihood of Disclosing Increases

• Each additional publication 0.1.
• Strong Local Peer Effects
• 1 increase in the percentage of faculty
  disclosing within the relevant cohort increases
  the probability of an individual disclosing by
  12.
• Training Matters
• Pro Tech Transfer Institution 4 for every 10
  patents
• Stanford 27
• Dual Training (MD/PhD) 4
• Chairman influence weakest
• Chair active 4 (weakly significant)

17
Selection or Socialization?

• Department Chairs with a History of Disclosing
  were No More Likely to Hire Individuals
  Predisposed to Disclosing than Non-Active
  Chairs
• Robustness Checks
• Departmental Fixed Effects
• Number of Disclosures

18
Dissonant Situations

• What happens when training and current work
  environment provide mixed signals?
• H5 When individuals are faced with a situation
  where their individual training norms are not
  congruent with the localized social norms in
  their work environment, they conform to local
  norms.

19
Figure 1 Alignment between training norms and
localized social norms
20
Localized Learning Trumps Training

• Individuals are most responsive to local cohort
  pressures
• If not trained with entrepreneurial expectations,
  local cohort can catalyze
• If trained with entrepreneurial expectations,
  local cohort can suppress
• If neither training nor local pressure then
  entrepreneurship is a rare event
• Localized learning is a knowledge source for
  entrepreneurship

21
Symbolic versus Substantive Adoption

• Just enough to seem to be in compliance but not
  as much as might be done, ceteris paribus
• N 169 Symbolic Individuals
• N 136 Individuals
• H6 Symbolic compliers will respond to different
  influences than substantive adopters.

22
Symbolic vs. Substantive Adoption Participants

• Probit Model
• Dependent Variable Disclosure Filed (0, 1)
• Same Basic Specification
• Substantive Adoption Disclosures
• Local Peer Effect is Stronger
• Symbolic Disclosures
• Stronger Chair Effect
• NIH is positive and statistically significant

23
How to Change an Organization

• Creating Entrepreneurial Organizations
  Promoting Organizational Change
• Requires Understanding and Management of both
  Individual Motivations and Departmental
  Composition
• Individual decisions influenced by relevant
  others
• Sub-unit composition and dynamics are key
• Not just about leaders
• Not about hiring individuals with
• Appropriate training
• Prior experience
• Critical mass of symbolic participants
• Enforcement of rules and incentives
• Traction for creating local cohort
• Keep these individuals together then culture
  changes

24
Organizing for Entrepreneurial Success

• Academic Entrepreneurship is a team sport
• 40 Individual Efforts 60 Team Efforts
• Compared to linked academic publications the
  number of inventors on a disclosure is less than
  half the number of authors on a paper.
• Ave. publication team size is 5.33 (sd 1.81)
• Ave. disclosure team is size 2.11 (sd 1.31)
• Solo efforts
• Publications 3 of all inventors papers
• Disclosures 40 of disclosures

25
Broader Use of Disclosure Data

• Studying Disclosure Teams
• Same 2 Prominent East Coast Universities with
  Medical Schools
• From 1988 to 1998 July 1, 1988 to June 30, 1999
• Data from Tech Transfer Offices
• 2340 Disclosures Filed
• 4942 Unique Individual participated, all academic
  departments plus outsiders
• Configurations change
• Augmented with
• Web of Science/ISI Publication data
• Patent data
• Probit Model
• Dependent variables relevant outcomes patent,
  license, Royalty

26
All in One Hypotheses Results

• Technical Diversity Two Competing Influences
• Diversity in Knowledge is Key for Innovation ()
• But Diversity Raises Coordination Costs (-)
• We find Diverse Teams are Less Productive
• But Team Experience Matters The Negative effect
  is Reduced as the Team Gains Experience Together
• Organizational Diversity
• Diverse Networks Gives Access to Resources ()
• Having an Industry Team Member Matters
• Leadership Effect
• The Experience of the Leader Matters Directly ()
• Learning Effects beyond specific team
  configuration

27
What we are doing now

• Power Relationship
• Stars (Scientist) and their Constellations
• The Great Person or the Great Team
• Apprenticeship System
• Reconfigurations of teams
• Over trials, do teams become
• Larger or smaller
• More homogenous or more diverse
• More successful
• Stay tuned

28
Questions?