Policies for innovation networks: do requirements on network composition work?

Federica Rossi, Annalisa Caloffi and Margherita Russo

Innovation policies increasingly support innovation networks involving firms and other organisations, in a number of ways – alongside traditional interventions targeted at individual firms, like Research and Development (R&D) tax credits and innovation vouchers, more contemporary policies are emerging, like R&D consortia, networks of excellence and university-industry partnerships. For the last three decades, the EU Framework Programmes supported networks of cooperating organisations with funding, but policies supporting innovation networks are also increasingly popular at regional and national levels.

The objectives of such policies are primarily to support joint R&D, technology transfer activities or even networking, with a view to create a critical mass of experts or users in a certain field or technology area. Many of these policy interventions require the participants to comply with a number of relational features that should lead to successful collaborative innovation. For instance, innovation networks are often required to include a minimum number of SMEs and universities. However, the effects of such interventionist approach to network design have rarely been tested in practice.

Case Study – Tuscany, Italy

By utilising micro data on the 1127 organisations participating in a set of nine consecutive regional policy interventions implemented in the Italian region of Tuscany between 2002 and 2008, we investigated whether the imposition of requirements on network composition had any effect on the subsequent networking behaviour of participating organisations – that is, whether asking participants to comply with specific network designs improved their ability to network with other organisations after the policy had ended.

Our analysis suggests that participating in networks that, on average, had higher minimum size requirements led organisations to form larger and more heterogeneous networks; while participating in networks that, on average, had higher minimum heterogeneity requirements led organisations to form less heterogeneous and smaller networks. Further results suggest that this requirement was interpreted as a guideline to follow closely, rather than a minimum threshold, reducing experimentation and leading organisations to adhere strictly to the recommended requirement. Once requirements were removed in the second period, organisations experimented with greater variability in network composition, and there was no effect on the average size and heterogeneity of an organisation’s networks in the second period of the average size and heterogeneity requirement imposed in the first.

Going forward

A number of lessons can be learned from this analysis. While there is a general consensus on the benefits of heterogeneous networks, the nature of the organisations that may best contribute to the partnership relies heavily on the content of the project that the network intends to realise. The requirement to involve specific types of organisations may force participants to include types of partners that are not needed for the purposes of their project, creating unnecessary complications. Our findings also suggest that they may discourage them from experimenting with different, possibly more appropriate, network compositions. Looser requirements like the imposition of a minimum number of partners may have fewer drawbacks, since they still leave participants with some freedom to design their partnership according to their needs, but they still impose additional transaction costs and require the project resources to be spread to partners who may not be necessary for the project’s success.

Therefore, requirements should be used sparingly, and particularly very specific requirements should be avoided, as they can be counterproductive to fostering network effectiveness, as well as learning and experimentation.

To support organisations’ ability to interact with a broader and more heterogeneous set of collaborators, other strategies may be more productive. For example, outreach actions to inform organisations of the benefits of networking with other partners, and to help them exploit the opportunities offered by policy interventions; or additional measures, alongside the funding of innovation networks, to foster participants’ learning from the experience of successful projects and to mediate communications between organisations of different types that find it difficult to work together.

The detailed empirical analysis on which these results build is presented in:

Rossi, F., Caloffi, A. Russo, M. (2016) Networked by design: can policy requirements influence organisations’ networking behaviour?, Technological Forecasting and Social Change, 105:203-214.