The role of market-oriented institutions in the deployment of renewable energies: evidences from Europe

ABSTRACT

Focusing on European countries, this article investigates the link between market-oriented institutions, as measured by the Economic Freedom Index, and the production of energy from renewable sources. A dynamic panel approach shows that this correlation is positive and significant while the subcomponents of the Economic Freedom Index reveal that not all market-oriented institutions have a similar impact. Indeed, long-term price stability and freedom to trade boost the reliance on renewable energies whereas the importance given to markets rather than governments has no significant impact.

KEYWORDS:

• Renewable energy
• Economic Freedom Index
• market-oriented institutions
• dynamic panel data

JEL CLASSIFICATION:

• Q40
• Q48
• P48

Disclosure statement

No potential conflict of interest was reported by the author.

Notes

¹ See for example Agterbosch, Vermeulen, and Glasbergen (2004) for the Netherlands or Bergek, Mignon, and Sundberg (2013) for Sweden for country specific discussions on the emergence of these new entrepreneurs. See Zenewi and Stoltenberg (2010) and OECD (2011) for policy discussions on the importance of private investments in renewable energies.

² The missing-money problem (Joskow 2007) relates to the insufficient incentives to build new capacities, i.e. when the high up-front fixed cost owing to an investment is too large compared with the quantity and wholesale price of its output. The intermittence of renewable energies exacerbates this problem by increasing the reliance on the spot market and by creating additional uncertainties about the plant operation. In a competitive energy market, the electricity price might not be high enough to cover its marginal and fixed costs of production for the technology close to the marginal cost.

³ Using an extreme bounds analysis, Gassebner, Lamla, and Sturm (2010) observe a positive correlation between the Economic Freedom Index and the level of pollution. However, they do not look at the impact of each subcomponent.

⁴ Only 10 out of 263 observations are equal to zero. Due to the relatively small number of these corner outcomes, we have decided not to analyse them separately from our positive observation with a censored or a truncated approach.

⁵ Results available upon request.

⁶ Note that we also take into account that ratings are computed with two-year-old data at the time of publication.

⁷ According to the Hausman specification test, the coefficients using a fixed effect approach significantly differ from a random-effect approach. Mind the fact that we include time dummies in all our specifications but that it does not impact the quality of our results. Our results are driven by both intra and inter group variation. Standard errors are computed by assuming no correlation across countries in the idiosyncratic disturbances. This assumption is more likely to hold with time dummies. Note as well that these results do not vary if we do not include a lagged dependent variable as an independent variable.

⁸ Size of government is negatively correlated with the other 4 subcomponents of the index, while these same components are all positively correlated

⁹ Investor’s confidence level is an example of such time- and country-varying unobserved variable for which we do not have a good proxy. As this characteristic is potentially positively correlated with ecofree(t−1) and renewable, our original estimates would be positively biased.

¹⁰ As it analyses the determinants of renewable energy production, this spatial correlation has also been used as an instrument by Smith and Urpelainen (2014). However, in their case, the variable instrumented was the level of feed-in tariff, not the Economic Freedom Index.

Additional information

Funding

This work was supported by Labex Entreprendre from the University of Montpellier. I particularly want to thank Jean-Christophe Poudou and Benoit Mulkay for their constructive remarks. All remaining errors are mine.