Effects of university characteristics on scientists interactions with the private sector

1
Effects of university characteristics on
scientists interactions with the private sector

• Paper presented to the T2S annual conference in
  Atlanta, GA, September 2006

2
Research question

• Academic quality, total and industrial RD,
  patenting are shown to be associated with higher
  level of institutional involvement in
  commercially relevant outcomes (e.g. startups,
  licensing revenue, patents)
• Are these institutional characteristics
  indicative of environment that increases the
  propensity of individuals employees in these
  institutions to engage in interactions with the
  private sector?
• Do such institutional characteristics moderate
  the effect of individual level determinants of
  interactions with the private sector?

3
Motivation

• Econometric relationships are well established
• But what is the underlying explanation?
• Are differences in commercial involvement due to
  unmeasurable differences among institutions?
• Or do certain institutional characteristics in
  fact stimulate individual behaviors that can
  produce the commercially relevant outcomes (at
  the aggregate)

4
Hypotheses

• These institutional characteristics positively
  influence the propensity of scientists to
  interact with the private sector
• Academic quality
• Total RD expenditures
• Total patents
• Industrial RD expenditures
• ?Rationale research capacity facilitates
  undertaking commercially relevant behaviors

5
Data

• Survey data from 1653 scientists in 13
  disciplines employed in the nations 150
  research extensive universities
• Extant university level data (WebCASPAR, USPTO)

6
Dependent variable the industrial involvement
scale
7
Independent variables individual

• Tenure status
• Productivity
• Affiliation with a university research center
• Discipline
• Number of students supported through grants
• Number of active grants
• Number of collaborators
• Gender

8
Independent variables contextual

• Total RD expenditures, FY 2002
• Total industry RD expenditures, FY 2002
• Total number of patents produced by the
  university
• Average number of citations per paper at
  university

9
Methodology

• Hierarchical linear model (HLM)
• Efficient way to estimate the effect of
  institutional characteristics on the dependent
  variable but also on the magnitude of effects of
  the individual level independent variables
• Random slopes and intercepts model

10
(No Transcript)
11
MS Students
PhD Students
Tenured
Engineering
Physical
12
Results

• Academic quality has direct negative effect on
  scientists propensity to interact with the
  private sector
• Industry RD has direct positive effect on
  scientists propensity to collaborate with the
  private sector
• Industrial RD moderates the effect of rank (-)
  and MS students supported through grants ()
• Academic quality moderates the disciplinary
  effects and the effect of supporting graduate
  students
• Patenting moderates the effect of Physical
  sciences and number of PhD students supported
  through grants (-)

13
Conclusions

• Institutional characteristics such as patenting
  and total RD have no direct effect on private
  sector-oriented behavior
• Consistent with the institutional differences
  hypothesis
• Academic quality and patenting have positive
  effect on private sector-oriented behavior
• Consistent with institutional environment
  /support /incentives hypothesis
• The effect of lower level variables is moderated
  academic quality (ENG PHYS ?) and patenting
  (PHYS? - )
• Indirect evidence that the macro-level analysis
  institutional analysis should be supplemented
  with micro-level analysis

14
Conclusions contd.

• Applied friendly contexts more intensive
  industrial presence diminishes the effect of
  institutional barriers (e.g. the tenure system)
  for faculty to undertake more applied research
• Academic quality important, but at the
  discipline level. Interacts with unknown
  disciplinary characteristics that impact
  propensity to interact with the private sector.
• Future research needs to identify what particular
  discipline level characteristics affect the
  propensity of scientists to interact with the
  private sector