3. Theoretical Approaches to Dynamic Efficiency in Policy Contexts: The Case of Renewable Electricity

However, some authors are doubtful about the relative importance of dynamic efficiency criteria compared with more traditional, static efficiency criteria. For example, Parry et al. (2003) stress that the welfare gain from innovation is sometimes not much greater than the welfare gain of efficiently abating pollutants by means of conventional technologies. Requate (2005) observes that resources to engage in R&D are scarce. Hence, environmental technological progress may crowd out other strands of welfare enhancing technological progress. Finally, Fischer and Newell (2008) argue that the underlying process of technological change turns out to be far less important than the incentives to use technology efficiently to reduce emissions.

Following Marechal (2007), we use the word traditional to avoid the problems arising from the somewhat ambiguous use of the term neoclassical. Traditional economics refers to the Walrasian model of welfare economics, which can be defined as the theoretical synthesis of the Marshallian approach with marginal production theory (Marechal 2007).

A stream of the economic literature on climate change mitigation has applied an evolutionary approach with the aim of emphasising the inertia in current technological systems (Kemp 1996; Unruh 2000, 2002; Marechal 2007; del Río and Unruh 2007; Rip and Kemp 1998; Foxon et al. 2005).

Assessment in terms of system functions is one of the main approaches of the systems of innovation literature. Other innovation system studies have placed more emphasis on structural analyses (Carlsson et al. 2002; Jacobsson and Johnson 2000). Currently, some authors are concentrating on the integration of both approaches (Markard and Truffer 2008).

For example, in his analysis of wind energy deployment and policy in Denmark, Spain and Sweden, Meyer (2007) provides empirical evidence of the role of the coalition of forces in encouraging wind energy in Spain.

The assessment of Astrand and Neij (2006) shows that early inflexible steering of technology and market development, together with a lack of comprehensive, long-term strategy, lack of continuity in policy interventions and weak combinations of policy programmes and measures, have contributed to very limited wind power development in Sweden.

For example, Junginger et al. (2006) show that for technologies developed on a local level (e.g. biogas plants), learning-by-using and learning-by-interacting are important learning mechanisms whereas for CHP plants utilising fluidised bed boilers, upscaling is probably one of the main mechanisms behind cost reductions. Nemet and Baker (2010) show that certain components of the costs of solar PV improved with R&D investment, whereas others responded to increased deployment of the technology.

Some authors have stressed the difficulties in building learning curves for some renewable energy technologies (Junginger et al. 2006) or criticised the learning curve model itself (Kahouli-Brahmi 2008).

For a recent analysis of (observed) learning rates for various electricity supply technologies, see IEA (2008) and Kahouli-Brahmi (2008), among others.

The literature seems to be too polarised in this respect, with theoretical and empirical studies following either one or the other approach. Exceptions are Rogge and Hoffmann (2010) and Walz and Schleich (2009).

Indeed, learning effects introduce nonlinearities and positive feedbacks into the models in which they are used (the more a technology is used, the greater the incentive for using it more) (McDonald and Schrattenholzer 2001).

Watanabe et al. (2000) convincingly showed that the political environment behind Japanese government support for PV innovation was critical in developing the interindustry partnerships basic public research and broad-based market promotion for this fledging industry which in turn led to and was a result of learning effects. The authors analysed the Japanese solar PV Sunshine Project which aimed to encourage the broad involvement of cross-sectoral industry, stimulate inter-technology stimulation and cross-sectoral technology spillover and induce vigorous industry investment in PV R&D, leading to an increase in industry’s PV technology knowledge stock. They showed that an increase in this technology knowledge stock contributed to a dramatic increase in solar cell production. These increases led to a dramatic decrease in solar cell production price, and this decrease induced a further increase in solar cell production. An increase in solar cell production induced further PV R&D, thus creating a “virtuous cycle” between R&D, market growth, learning effects and price reduction.

For example, in the analysis of the Dutch wind-offshore sector, Smit et al. (2007) argued that there was weak learning-by-interacting by the actors from the industrial part of the technology system, who should get better access to actors in academia and actors in the oil and gas industry. The authors showed that there were several barriers hindering this interaction process. In contrast, they also showed that in the Danish case, learning-by-interacting occurred between knowledge institutes, component suppliers, project operators and turbine manufacturers and Danish policies contributed to the formation of these interactions.