Renewable energy innovations in Europe: a dynamic panel data approach

Abstract

We investigate the determinants of renewable energy R&D intensity and the impact of renewable energy innovations on firm performance, using several dynamic panel data models. We estimate these models using a large data set of European firms from 19 different countries, with some patenting activity in areas related to renewable energies during the 1987 to 2007 period. Our results confirm our priors on the determinants of the rapid development of renewable energy R&D intensity during the past decades. Additionally, we find evidence that renewable patent intensity has a significant dynamic impact on the stock market value of firms.

Keywords:

• R&D intensity
• stock market value
• patents
• count data
• dynamic panel data

JEL Classification:

• C15
• C31
• C32
• C33
• CA1

Acknowledgements

We thank PIUNA and Deloitte for the financial support. We also thank Mirko Abbritti, Luis Albériko Gil-Alana, Germán López Espinosa, Antonio Moreno, and Jerome Vandenbussche for their comments on previous versions of this article. This article was presented at the XXXIV Simposio de la Asociación Española de Economía, Valencia, December 2009 and at the MKE Annual Conference, Budapest, December 2010. In this context, the authors thank Gábor Kőrösi for helpful discussion and comments.

Notes

¹ See Hausman et al. (1984), Pakes (1985), Lanjouw et al. (1998) or Trajtenberg (2002).

² We use these two variables alternatively as patent counts in our empirical application.

³ An extension of this work could be the application of a more complete firm specific data set, which would result a more complete panel after matching firm data with the PATSTAT database.

⁴ We have two alternative choices for the endogenous variable in the patent count data model. We shall estimate two alternative specifications for each count data model. (See Sections II and IV.)

⁵ We shall be more precise regarding the Z _it term in the empirical applications section.

⁶ We control for initial conditions because we are in a short-panel setup in this article.

⁷ We shall be more specific regarding the firm performance variable in our empirical application section.

⁸ For example in solar energy sector where the cost of developing performing photocells is particularly high.

⁹ We exchange oil prices to EUR as our sample includes companies from the EU and EFTA that account their costs in EUR.