The Impact of Energy Prices on Green Innovation.

AuthorLey, Marius
PositionReport - Statistical data
  1. INTRODUCTION

    Despite the fact that climate change should ideally increase the demand for green technologies, firms still have little incentive to invest in green technologies as there is a 'double externality problem' (see, e.g., Beise and Rennings 2005, Faber and Frenken 2009, Hall and Helmers 2011). Firstly, financial market imperfections that are normally associated with innovation activities (see Arrow 1962, p. 172) are even more pronounced for green innovation. Green innovations carry a large technical risk as they often imply investing in technologies that lie beyond the firm's traditional technological scope. Additional commercial uncertainty arises from unclear market developments (Aghion et al. 2009). Hence, potential external investors are hardly willing to invest in such projects and financial markets are usually not ready to finance such risky technological investments. As a consequence, access to external capital to finance green innovation is likely to be constrained. Secondly, because the greatest benefits from green inventions are likely to be public rather than private, the customers' willingness to pay for these innovations is low. In line with this literature, recent studies have shown at the firm and industry levels that green innovations currently have lower returns than non-green innovations (see Marin 2014, Soltmann et al. 2014). These results indicate that--given the current level of green promotion--free market incentives alone are not sufficient to allow the green innovation activities of industries to rise considerably. However, technological innovations are needed to solve environmental problems. "Without significant technological development of both existing low-carbon technologies and new ones, climate change is unlikely to be limited to anything like 2[degrees]C" (see Helm 2012, p. 213). Accordingly, policy intervention is required to stimulate green innovation activities.

    This paper focuses on energy prices and investigates whether energy prices increase the probability of producing green inventions. More concretely, we investigate whether the effects of energy prices are different for 'green' inventions than for 'non-green' inventions. The fluctuation of energy prices has a price component and a policy component. As the price component is primarily driven by international market prices and we control for between-country variation, the remaining variation in the energy price in our model is mainly due to energy taxes. Consequently, energy prices can be interpreted as an environmental policy instrument (see Aghion et al. 2012 for a similar argumentation). Nevertheless, as we cannot fully capture the price component, the econometric results of this paper cannot be interpreted as a 'pure' policy effect. Our results refer to fluctuation of the end-use price including taxes. (1)

    Empirical research linking environmental policy and innovation is related to a small but growing literature. A first group of studies uses pollution abatement control expenditures (PACE) as a proxy for environmental regulation stringency. Brunnermeier and Cohen (2003) found for the US that PACE is positively related to environmental innovation. Based on a data set that includes 17 countries Lanjouw and Mody (1996) also found a positive correlation between PACE and environmental innovation. However, the use of PACE as a measure for policy stringency in a crosscountry study is questionable due to the heterogeneity in the definitions and sampling strategies (see Johnstone et al. 2012, p. 2161). To overcome this problem Johnstone et al. (2012) used survey data. Based on this data they again found that environmental innovation is positively affected by environmental policy stringency.

    Most other studies overcome the problem of comparability by using energy prices as proxy for environmental regulation. Most of them focus on a single industry. Aghion et al. (2012) investigated the significance of energy prices for technological change by looking at the car industry using patent data between 1978 and 2007. They found that higher energy prices increase the propensity of 'clean' innovation in the car industry. Moreover they stated that the price effect is stronger for firms with a large stock of 'dirty' patents. Newell et al. (1999) looked at the level of product characteristics in the air-conditioning industry and found that energy prices had an observable effect on energetic features of the products offered for sale. Lanzi and Sue Wing (2011) found a positive relationship between energy prices and innovations in renewable technologies in the energy sector of 23 countries.

    Rather than focusing on a single industry, Popp (2002) focused on a single country. He looked at 11 different technologies including supply (e.g. solar energy, fuel cells) and demand technologies (e.g. recovery of waste heat for energy, heat pumps) for the USA and found that energy prices and the existing knowledge stock have a strong and significant positive effect on innovation.

    It is unclear in all these studies whether the results also hold for other industries and/or countries. Only a few studies are based on data for more than one country and more than one industry. Johnstone et al. (2010) analyzed how different policies (among others energy prices) affect innovation for five different renewable energy technologies. Verdolini and Galeotti (2011) investigated the impact of energy prices on technological innovation (12 technologies like in Popp 2002) for a panel of 17 countries and found a positive sign. However, as both studies are based on data that is either aggregated to the country-level or technology-level, there is a concern that there may be other country-specific shocks correlated with both innovation and the energy price (see Aghion et al. 2012, p. 5).

    This study contributes to the existing literature primarily in two respects. First, the breadth of our data set allows us to draw much more general conclusions than was possible in previous studies, which have focused on single industries or countries. This enables us to generate an industry-level data set that covers the whole manufacturing sector (grouped into 10 industries), the most important countries for green innovation (18 OECD countries that are responsible for more than 95% of all green patents and total patents worldwide) and this over a period of 30 years. Secondly, in contrast to previous studies, we greatly reduce the probability of omitted variable bias, which makes our results more reliable.

    In line with previous studies, we use patent data to identify green and non-green inventions according to the OECD Indicator of Environmental Technologies (see OECD 2012), (2) however, we switch from the technology level to the industry level by using the Schmoch et al. (2003) concordance scheme. In contrast to studies that stay on the technology-level, this approach allows us to include industry-level control variables (e.g., capital and number of employees). Furthermore, we reduce the potential problem of omitted variable bias by controlling for industry and country fixed effects. Additionally, we calculate industry specific energy prices, which allow us to include country-specific time fixed effects. Compared with previous studies on a more aggregated level (i.e. country level), country-specific shocks that are correlated with both innovation and the energy prices (see Aghion et al. 2012, p.5) do not bias the results in this particular study. An important concern here is that national governments may have introduced policies directly supporting green innovation (such as research subsidies) simultaneously with higher energy taxes. In such a scenario, not controlling for country-specific attributes would tend to bias estimates for the price effect on innovation. (3)

    With respect to our main variable, green inventions, we find that energy prices stimulate both the level of green invention as well as the share of green invention. In our model, a 10% increase in the average energy prices over the previous five years results in a 3.4% and 4.8% increase of the number of green inventions and the ratio of green inventions to non-green inventions, respectively. Knowledge about potential political instruments to stimulate invention in this area is of great importance. This study shows that energy prices may serve as such an instrument. An increase in energy prices may stimulate the building of a green knowledge stock that: (a) would help to achieve a country's climate targets; and, (b) may help to establish a cleantech market for which long-term growth is predicted.

  2. CONCEPTUAL BACKGROUND AND HYPOTHESES

    The idea that an increase in the relative price of a production factor will direct innovation efforts towards technologies that are less intensive in the production factor becoming more expensive can be attributed to Hicks (1932, as quoted e.g. in Binswanger et al. 1978): "A change in the relative prices of the factors of production is itself a spur to innovation, and to innovation of a particular kind - directed to economizing the use of a factor which has become relatively expensive."

    This intuitively appealing assertion has been known as the induced innovation hypothesis. Subsequent research attempted to provide microeconomic foundations for this claim and to assess its relevance for traditional welfare economics (Binswanger et al. 1978, ch. 4). Induced innovation is generally thought to exacerbate the effects of externalities not properly taken into account. In particular, the exploitation of fossil fuels has undesirable side effects as C[O.sub.2] emissions negatively affect global climate. Two harmful mechanisms are at work as a result of not having adequately priced these energy resources (by failing to take into consideration their negative externalities, e.g. by charging a C[O.sub.2] tax): price signals not only affect entrepreneurs' choice of input combinations...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT