Rethinking Network and Location in economics of innovation: Theory and public policy evaluation. – RENEWAL

This project presents new methodologies which take into account network effects (motivating firms to collaborate in R&D) or spatial effects (while targeting some territories) for the evaluation of innovation policies. There are more and more such policies in the world, and evaluation studies of these policies do not take into account, in an appropriate way, indirect effects associated with the collaborative and geographic structuration of innovation activities. However, if there are some positive effects due to cooperation and the spatial proximity, there should be also some distortions between firms and territories that lead to an uncertain total effect. Two kinds of theoretical literature have been mobilized to identify situations that justify the use of these policies. The first, Game theory is applied to the R&D collaboration network. The second, the New Economic Geography and Growth theories are dedicated to the analysis of innovation policies in a spatialized context implying location choices and local dynamics. This project proposes new evaluation methods based on this analytical framework.
Several methodologies have been presented to measure direct and indirect effects of innovation policies. These indirect effects are due to the networks of collaborations between firms or the interactions between neighboring firms.
- At a microeconomic level, recent literature on the econometrics of network is used to explain the collaboration choices and measure the impact of the policies on firms’ performances while taking into account interactions between firms within these R&D collaborations networks.
- At an aggregate level, interactions model are used in order to identify the effect of R&D policies. More precisely we study the impact of policies on the performance of territories by taking into account the spatial and collaborative dependencies between them.

Beyond the theoretical results that have been published, our main results regard new methods for the evaluation of spatial and network effects associated with innovation policies. New databases have been set up at the firms and the territorial levels for the evaluation of diverse French and European policies implemented during the 2000's.
Applying our methods on these data, we obtain new results about the impact of a number of R&D and Innovation public policies.

First obtained results have rich potential implications and open ways to fine-tune the design of policies that so far have been focused on all-location/all-cooperation notably through the diverse clusters policies.
The recongnition that these first results have received in the international academic communauty show the high potential that has been opened for future researches in this field mixing competencies in games theory, economic geography of growth and spatial/network applied econometrics.

Numerous developments using the databases computed during this project are still possible.
In particular, we plan to implement some econometrics methodologies issued from the most recent developments of the econometrics of network field. These are methods that can allow us to measure indirect network effects of policies with endogenous et dynamics networks. Works on this topic are still in progress.

This project has also allowed us to build new national and international partnerships, in a more interdisciplinary perspective:
- a new proposal for a ANR project , on the one hand.
- the insertion within a European proposal issued from the GeoInnov network, on the other hand (to be concretized at the GeoInnov2018 conference in Barcelona)

This short project leads to several international academic publications.

1. Billand P., Bravard C., Kamphorst J., Sarangi S., Network Formation when Agents Seek Confirmation of Information, Mathematical Social Sciences, forthcoming.
2. Billand P., Bravard C., Chakrabarti, S., Sarangi S. (2016). Business intelligence and multi- market competition. Journal of Public Economic Theory, vol. 18, N. 2, pp. 248-267, numéro spécial « Networks and Externalities ».
3. Billand P., Bravard C., Durieu J., Sarangi S. (2015). Efficient Networks for a Class of Games with Global Spillovers. Journal of Mathematical Economics, vol. 61, pp. 203-210.
4. Autant-Bernard C., S. Hazir & J. LeSage The Role of R&D Collaboration Networks on Regional Knowledge Creation: Evidence from Information and Communication Technologies Papers in Regional Science, 2017, forthcoming
5. B. Montmartin, M. Herrera, N. Massard, «R&D policy regimes in France: New evidence from a spatio-temporal analysis« en révision pour Research Policy.

Other works are available as working papers. These working papers (7) are submitted or will be submitted to international journals. These works have been presented in international academic conferences (19) but also to large audience conferences (5). Also they have been central to the organization of special sessions and conferences (3).
Moreover these methods have already been valorized with national and local public decision makers for contractual evaluation studies. Due to contractual conditions, some studies cannot be diffused at this stage.