slides - Francesco Lissoni

University autonomy, IP legislation and
academic patenting: Italy, 1996-2006
Francesco Lissoni1,2, Michele Pezzoni2,3, Bianca Potì4, Sandra Romagnosi5
– Université Bordeaux IV - France
– Università "L. Bocconi", Milan - Italy
3 Dept of Economics, Università Milano-Bicocca - Italy
4 CERIS-CNR, Rome - Italy
5 Parco ScientificoUniversità "Tor Vergata", Rome - Italy
APE-INV Final Conference
Paris, 3-4 / 9 / 2013
Motivation & Research Questions
Contribute to recent literature on academic patenting in Italy
(Europe) by:
1. What/Any trend in academic patenting?
• Weight of academic patenting on total domestic patenting
• Ownership: Universities’ share of IP over academic inventions
(vs individuals’, PROs’, and business companies’ share)
2. Exploring links between (1) and two policy changes:
• The granting of autonomy to universities (incl. financial
autonomy), in 1989 (effective kick-off: 1995)
• The introduction of the professor privilege, in 2001
Reasons for focusing on universities’ autonomy
• Policy: widespread diffusion of autonomy-granting/enhancing
reforms in all Europe (e.g. “loi Pecresse” in France, 2007); large
universities’ quest for more autonomy (e.g. EUA’s report, 2009)
• Scholarly research - in sociology: “entrepreneurial university”
(Clark, 1993); in economics: autonomy&competition  perfomance
link (Aghion et al., 2009)
 Increasing emphasis on “third mission”: is it materializing? (weight
of academic patenting)
 Decrease of “block grant” funding  project funding &
technology transfer as additional sources of revenues: do
universities look at IPRs as a source of revenue?
 Changes in academic profession’s status (from civil servants to
university employees): are universities seizing professors’ IPR
Reasons for focusing on the professor privilege
• Policy:
1. wave of abolitions in German-speaking and Scandinavian
countries since 2000  inefficient legal institution, standing
in the way of commercialization of academic research
2. BUT Italy has introduced it in 2001  incentive-setting
justification BUT contradiction with autonomy granting to
• Scholarly research – some recent advocacy for the privilege
(Kenney, 2009)
Conclusions /1
A. The absolute number of academic patents has increased, but
(i) their weight on total patenting by domestic inventors has not
(ii) the share of university-owned acad. patents has increased
B. The probability to observe an academic patent depends on:
- the technology considered
- the science-intensity of research,
- and the characteristics of the local innovation system
After controlling for these determinants:
(iii) the conditional probability to observe an academic patent
has declined over time.
Conclusions /2
C. The rise of university ownership is explained by:
(iv) the increasing share of public vs. private R&D
(v) the increased autonomy of Italian universities
introduction of explicit IP regulations
D. The introduction of the professor privilege in 2001 had no
impact at all on either trends  opposed and defeated by
universities, thanks to their newly gained autonomy
Methodology for data collection
1. Name disambiguation of inventors (EPO patent applications) 
free inventor database:
2. Professor-inventor name matching: 3 professors’ cohorts 
inventors 1996-2006 [academic patent  patent with at least 1
academic inventors]
3. Filtering of false matches by: (i) automatic criteria (ii) past surveys
(iii) ongoing survey (iv) probability estimates of no-responses
University autonomy in Italy:
a quick look
The professor privilege in Italy:
an even quicker look
University autonomy
• L.168/1989: basic principles and creation of ad-hoc Ministry
• Several laws/decrees 1990-1996.
• Financial autonomy
1. Key block grant: FFO ("Fondo di Finanziamento Ordinario"):
starts at 90% of all revenues  automatic decline
2. Universities become free to collect other revenues  great
3. No systematic tie with university-industry technology
transfer policy
4. (for a while) GERD grows faster than BERD
 (Epidemic) diffusion of IP regulations (IP_STATUTE) and TTOs at
the university-level
 Little correlation between the two diffusion processes
Weight of block funds (FFO) and public funds for scientific reserach
on Italian Universities’ totale revenues (sources: AQUAMETH, CNSVU)
Diffusion of IPR statutes and TechTransfer Offices in Italian Universities
(sources: own elaboration on NETVAL survey; CNSVU survey)
The professor privilege
• Introduced in 2001
• Unsolicited, indeed resisted by universities (unsuccessfully at legal
level; possibly successfully at IP regulation level)
• Reformed in 2005 (abolished for research co-sponsored by
Econometric Analysis
• 2-step Heckman Probit
• STEP1: probability of an Italian patent to be academic, 19962006 as a function of:
- time (year dummies)
- patent characteristics (IPC class, NPL backward citations, nr
- regional innovation system: BERD/GDP; universities’ and PROs’
share of R&D
- regional university system: diffusion of university IP statutes
and TTO; weight of FFO over total revenues;
 Estimate of academic patenting trend, conditional on changing
• STEP2: probability of an academic patent to be owned by the
inventor’s university, 1996-2006 as a function of:
- time(year dummies)
- patent characteristics & regional innovation system
- university’s characteristics:
- fixed effect (dummies)
- time-variant:
- adoption of IP statute
- TTO opening
- weight of FFO over total revenues (FFO_RATIO);
 Estimate of ownership trend, as a function of increasing
autonomy & conditional on changing environment
 Similar estimates for individual & business ownership
STEP1 (probability of an Italian patent to be academic)
- negative trend after controlling for patent characteristics (lessthan expected composition effect)
- “classic” results for patent characteristics
- Positive effect of both BERD/GDP (demand side) and
universities’ share of R&D (supply side)
- No effect of FFO_RATIO
STEP2 (probability of university ownership)
- positive trend after all controls ( unexplained trend)
- “classic” results for patent characteristics
- Positive effect of universities’ share of R&D (supply side)
- No effect of FFO_RATIO
- Positive effect of IP statute adoption vs no effect of TTO
Further research
1) The value of academic patents, by type of ownership
Lower? Lissoni and Montobbio (2013) + role of universities in weaker
Higher? Learning effect & increased autonomy ( see Flemish case)
2) Changes of property and markets for patents
3) Lessons for evaluation exercise (e.g. ANVUR)  Which patents do
count? Which patents shall we count?
 University-owned patents are a (non-representative?) subset of all
academic patents
 Counting university-owned patents may generate perverse
incentives in favour of patent filing / aggressive stances towards
business sponsors & faculty
 Use of public data such as PatStat / APE-INV

similar documents