Upload
briana-haynes
View
212
Download
0
Embed Size (px)
Citation preview
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
The effect of innovation vouchers on science-industry interaction
Marc Van der SteegMaarten CornetBjörn Vroomen
CPB Netherlands Bureau for Economic Policy AnalysisThe Hague, The Netherlands
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Outline
The evaluation problem in innovation policy The innovation voucher Research question and data Analysis Current research Conclusions Discussion
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
The evaluation problem
A two-way relation► causal: from policy to innovation► correlation: from innovation to policy
How to disentangle these two relations?► add covariates to the regression equation► do highbrow econometrics► or...
Controlled experiment► experimental group and control group► random allocation► difference is causal impact
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
The innovation voucher
Goal: ► introduce SMEs to public research institutes► [market-oriented incentives for research institutes]
Characteristics► credit note, value max EUR 7,500, non-transferable► application-oriented research question► placed with a defined group of institutes► SMEs only► valid for 7 months► no restrictions on e.g. level of question or technology► 100 vouchers available; lottery if demand > supply
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Research question
What is the effect of the innovation voucher on the commissioning of projects to public knowledge institutes?► number ► size/value► account for timing effect
Beyond the scope of the paper► “John Henry effect”: effect on losers ► persistence effect (current research)► effect on innovation output (current research)
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Data (1)
1,044 applications on September 17th, 2004 Lottery: 100 winners, 944 losers Telephone interviews during May, 2005
► 100 winners► 500 randomly selected losers► questions about actual behaviour► questions about counterfactual behaviour
Response rate► 71 winners (71%)► 242 losers (48%)
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Data (2)
No significant differences between winners and losers in background characteristics (size, region, sector)
Before voucher scheme…..► 85% ever had contact with a public knowledge institute ► 55% ever commissioned a project to a public knowledge
institute
Reasons for never having commissioned an assignment:
► no research question (15%)► a research question, but...
– too expensive (42%)
– research conducted in-house (16%)
– other priorities (14%)
– research institution or contact person unknown (7%)
– usually commissioned to private organisations (2%)
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Data (3)
Type of research questions voucher winners:► 60% product-related vs. 40% proces-related► 80% technological vs. 20% non-technological► 90% applied vs. 10% fundamental
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Analysis (1): effect on number
Data► 62 out of 71 (= 87%) winners commissioned a
project► 20 out of 242 (= 8%) losers commissioned a project
Effect► 13% of the vouchers not used (= (71-62)/71)► 8% crowding out (= 20/242)► 79% impact (= 62/71 - 20/242)► standard errors are small
Counterfactual behaviour► 76% winners say: without voucher, fewer projects► 86% losers say: with voucher, more projects
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Analysis (2): effect on size
Actual behaviour: ► for most winners: size project = voucher value► almost no data for losers
Counterfactual behaviour:► 81% of winners and 60% of losers say: voucher does
not affect size project► difficult to interpret, but no indications for a large size
effect
Voucher value seems focal point► follow-up project instead of larger project?
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Analysis (3): timing effect
Few projects outside voucher period 11% of winners say: without voucher same
number of projects, but later This indicates a limited timing effect
► maybe one out of eight additional projects
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Current research (1)
Same set of winners and losers► new questionnaire in September 2006
Effect on innovation (output additionality)► 2 years after lottery: reasonable?► Community Innovation Survey “yes/no questions”
– ongoing and realised innovations– new or significantly improved products/processes
Persistence (behavioural additionality)► number of follow-up projects► size of follow-up projects
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Current research (2)
Two lotteries in 2005 ► March: 1900 applications for 300 vouchers ► September: 1400 applications for 450 vouchers
Effect on number of projects (input additionality)
Exactly the same questions as for 2004 lottery
Inn
ov
ati
on
vo
uc
he
r
EFPLSept. 30, 2006
Conclusions
Random allocation of innovation policy feasible► political and legal objections can be overcome► lottery if demand > supply and no further selection
information available
Convincing evidence, easy to communicate Input additionality: eight out of ten vouchers
► limited timing effect
Crowding out: one out of ten vouchers Current research
► into output and persistence effect for 2004 voucher► into input effect for two voucher lotteries in 2005