Efficiency and evaluation analysis of a network of technology transfer brokers

Abstract

Technology transfer (TT) brokers are a type of innovation intermediary that is regarded as a facilitator of the interaction between donor and recipient within an innovation process. Yet, little is known about how to analyse their performance and efficiency. This paper aims at modelling a proper production function for the TT brokers, providing insights on how to compare their performance and investigating the influence of external factors on their efficiency, in order to draw useful managerial and policy implications. Data drawn from a survey conducted among the nodes belonging to the technology transfer network IRC (Innovation Relay Centre) have been utilised to investigate how endogenous and exogenous factors may affect brokers’ efficiency. Results showed how brokers can pursue more productive and targeted strategy in order to increase their impact and improve their performance (inter alia, by taking into account a ‘pull’ approach to technology and the need of a long-lasting relationship with the clients – small and medium sized enterprises (SMEs) in particular). At the policy level, interesting insights are drawn on how to tune existing innovation support initiatives and design new ones from scratch.

Keywords:

• technology transfer
• innovation intermediaries
• brokers
• efficiency
• data envelopment analysis (DEA)
• ordered probit

Notes

Brokers are that type of intermediary whose aim is to bring parties together to achieve a transaction (Chesbrough 2006).

The highest benefits should be found, at least, in achieving an easier first contact with potential clients.

The questionnaire is available from the authors upon request.

The DEA technique is affected to a certain extent by outliers. Therefore, its degree of precision may be affected in the presence of results differing greatly from others in the same sample.

It refers to people who have successfully completed education at the third level (ISCED ’97 version levels 5a, 5b or 6) and are employed in an S&T occupation (ISCO ’88 COM codes 2 or 3). Source: Eurostat (http://ec.europa.eu/eurostat).

For further information, see http://www.trendchart.org/

In particular, the RRSII index is calculated as an average of the following indexes: Human Resources in Science and Technology – Core (% of population), Participation in life-long learning (per 100 population aged 25–64), Public R&D expenditures (% of GDP), Business R&D expenditures (% of GDP), Employment in medium-high and high-tech manufacturing (% of total workforce), Employment in high-tech services (% of total workforce), EPO patents (per million population). Further information can be found in Hollanders (2006).

Our survey also contains data from countries that do not belong to the EU (Israel and Turkey) or that did not belong to EU in 2006 (Bulgaria and Romania). For the regions of those countries, we calculated missing data, in particular on ‘R&D’, ‘RRSII’ and ‘R&D_hrst’, through a proper proportion with EU-25 values.

For further information, see Wolfe, ‘OMODEL: Stata modules to perform tests on ordered probit and ordered logit’, 1997 (downloadable from http://econpapers.repec.org/software/bocbocode/s320901.htm).

The LR (likelihood-ratio) test compares the restricted model (in our case model A) with the unrestricted one; in our case, this test always rejects the null hypothesis that restrictions are valid. Therefore, model A is ruled out by this test.

For the computation of the marginal effects, variables are made standardised, thus with the mean value equal to zero and standard deviation equal to 1.

On this matter, consult the METTTES project, focusing on fostering the adoption of innovative solutions in the environmental field (http://metttes.innovationmalta.com/).