Empirical determinants of eco-innovation in European countries using the community innovation survey

G Model EIST-191; No. of Pages 14

ARTICLE IN PRESS Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

Environmental Innovation and Societal Transitions journal homepage: www.elsevier.com/locate/eist

Empirical determinants of eco-innovation in European countries using the community innovation survey Jens Horbach The Augsburg University of Applied Sciences, P.O. Box 110605, D-86031 Augsburg, Germany

a r t i c l e

i n f o

Article history: Received 8 January 2015 Received in revised form 28 July 2015 Accepted 26 September 2015 Available online xxx JEL classification: Q55 O33 C25 Keywords: Eco-innovation Probit models Country analysis

a b s t r a c t The determinants of eco-innovation activities have been widely explored for single countries. However, there is a lack of analyses comparing different countries. The Community Innovation Survey (CIS 2008) allows for an analysis of the determinants of eco-innovation in 19 different European countries. This paper focuses on Eastern Europe as the determinants of eco-innovation have not yet been systematically analyzed for these countries. Concerning the introduction of eco-innovations, this analysis shows that regulation activities and environmentally related subsidies seem to be more important for the Eastern countries than they are for the “richer” Western European countries. Furthermore, the Eastern European countries rely more on competitors and external R&D as information sources, indicating a technology transfer from West to East. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Eco-innovations alleviate environmental impacts or lead to a reduction in energy use and are, therefore, crucial for climate protection. They help to remedy negative external environmental effects of economic activities. In many cases, these negative external effects have to be internalized using regulatory measures. Young and dynamic eco-innovation fields such as the development of renewable energies can also be economically benign because these eco-innovations may lead to cost-savings. Recently, the determinants of eco-innovation activities have been widely explored for single countries but there is still a lack of country comparisons (Johnstone 2007; Frondel et al., 2007 or Horbach et al., 2013 as exceptions). This is mainly due to data restrictions. Country comparisons can be used to analyze the importance of different development levels, sector structures, regulation measures or innovation systems for the realization of eco-innovations. For the first time, the Community Innovation Survey (CIS) 2008 allows for such an analysis for 19 European countries including Eastern European countries because of the inclusion of a separate module on eco-innovation. A major focus of the paper is the analysis of the determinants of eco-innovation of Eastern European countries. Compared to the “rich” Western European countries, these Eastern European countries are characterized by a lower development level associated with a lower level of R&D input, other environmental priorities, lower environmental awareness among the population or a higher energy intensity in the economy. As far as possible, these country-specific determinants of eco-innovation are analyzed for different eco-innovation fields using a group of richer countries as a reference. The analysis helps to improve the innovation system and to identify regulation deficits in these countries.

E-mail address: [email protected] http://dx.doi.org/10.1016/j.eist.2015.09.005 2210-4224/© 2015 Elsevier B.V. All rights reserved.

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

2 Table 1 Determinants of eco-innovation. Supply side

• Technological capabilities • Appropriation problem and market characteristics

Demand side

• (Expected) market demand (demand pull hypothesis) • Social awareness of the need for clean production; environmental consciousness and preference for environmentally friendly products

Institutional and political influences

• Environmental policy (incentive-based instruments or regulatory approaches) • Institutional structure: e.g. political opportunities of environmentally oriented groups, organization of information flow, existence of innovation networks

Source: Horbach (2008).

The paper is organized as follows: Section 2 analyzes the theoretical background of the determinants of eco-innovation, focusing in particular on country-specific differences. Section 3 deals with the data basis and descriptive results for the specialization of the 19 countries on different eco-innovation fields. The econometric results for the determinants of ecoinnovation for country groups and nine eco-innovation fields are examined in Section 4. Section 5 summarizes and concludes the findings. 2. Determinants of eco-innovation—differences between countries In an EU-funded research project called “Measuring Eco-Innovation” (MEI), eco-innovation was defined as follows (Kemp and Pearson 2008: 7): “Eco-innovation is the production, application or exploitation of a good, service, production process, organizational structure, or management or business method that is novel to the firm or user and which results, throughout its life cycle, in a reduction of environmental risk, pollution and the negative impacts of resource use (including energy use) compared to relevant alternatives”. This definition is based on the outcomes of innovation activities. If innovations lead to positive environmental effects, they are defined as eco-innovations. The MEI definition also categorizes innovations as “green” if the underlying innovation activities were not intended to improve the environment. Therefore, “. . .it does not matter if environmental improvements have been the primary goal of a new product or process, or came about as a by-product or simply by chance. Eco-innovations can thus be the result of other economic rationales such as increasing market share or reducing costs” (Horbach et al., 2012: 113). In the following section, the different elements of the existing eco-innovation theory (see e.g. Jaffe and Palmer, 1997; Hemmelskamp, 1999; Cleff and Rennings, 1999; Rennings, 2000; Jaffe et al., 2002; Brunnermeier and Cohen, 2003; Mazzanti and Zoboli, 2006; Del Rio Gonzalez, 2009; Horbach, 2008; Kesidou and Demirel, 2012; Horbach et al., 2012; De Marchi, 2012; Cecere et al., 2014; Ghisetti et al., 2015) are used to derive country-specific determinants of eco-innovation. The general innovation theory accentuates the relevance of technology push factors and market or demand pull factors for the explanation of innovation activities (see Table 1 for an overview). These factors are, of course, also relevant for eco-innovations. However, as most environmental problems represent negative external effects, there may be no clear market incentives to develop new environmentally benign products and processes. Therefore, (environmental) policy measures and institutional factors may play an important role in the realization of eco-innovations. In many empirical studies, regulation has been identified as an important determinant of eco-innovation (e.g. early studies from Green et al., 1994; Cleff and Rennings, 1999; Rennings and Zwick, 2002; Brunnermeier and Cohen, 2003; Popp, 2006; Rennings and Rexhäuser, 2011; Lanoie et al., 2011; Veugelers, 2012) and is known as the “regulatory push/pull effect” (Rennings, 2000)1 . In light of this positive role of regulation for ecoinnovation, Acemoglu et al. (2012) show that if “dirty” and “clean” inputs are substitutable, a temporary taxation of the dirty inputs would be sufficient to redirect the production process to a more sustainable path. The authors state that “. . .delay in intervention is costly: the sooner and the stronger the policy response, the shorter will the slow growth transition phase be” (Acemoglu et al., 2012: 159). Despite the growing harmonization of environmental policies within the EU, countries may set different priorities according to sector structures, energy intensities or environmental impacts. These different priorities may stem from the problem that some countries are predominantly locked-in in pollution-intensive technologies (Zeppini and van den Bergh, 2011; Cecere et al., 2014). The resulting path dependencies make it difficult to change the innovation system of a country towards a more environmentally friendly path. In such a country, environmental policy may be more important for the introduction and implementation of eco-innovations compared to countries that are already locked-in in environmental technologies. In light of the above-mentioned argument expressed by Acemoglu et al. (2012), temporary regulation measures may be sufficient to re-direct the innovation system to a more sustainable path.

1

See Barbieri et al. (2015) or Horbach (2015) for overviews, a summary of this vast literature is beyond the scope of this paper.

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

3

Table 2 Country performance in terms of core economic, environmental and innovation indicators. Countries

Relative GDP per capita (EU27 = 100) 2008

Industry share in% of all branches 2008

R&D in% of GDP 2008

Bulgaria Cyprus Czech Republic Germany Estonia Finland France Hungary Ireland Italy Lithuania Luxembourg Latvia Malta Netherlands Portugal Romania Sweden Slovakia Total (EU 27)

44 100 81 116 69 119 107 64 132 104 64 264 59 81 134 78 47 124 73 100

22.0 9.5 29.6 25.7 20.3 24.8 13.7 25.1 23.6 20.7 21.5 9.4 14.1 17.0 19.8 17.3 25.8 21.6 29.0 19.8

0.47 0.43 1.30 2.69 1.28 3.7 2.12 1.0 1.45 1.21 0.8 1.66 0.62 0.55 1.77 1.50 0.58 3.7 0.47 1.91

Share of renewables in electricity consumption in%

Energy intensitya

Environmental awarenessb

2008

2011

2008

2012

9.7 0.3 5.2 15.1 2.1 27.3 14.4 5.3 11.1 16.6 4.8 3.6 38.7 0.0 7.1 34.6 28.1 53.5 18.0 16.6

12.9 3.4 10.6 21.3 12.3 29.2 16.5 6.4 17.6 23.5 9.0 4.1 44.7 0.6 9.8 46.5 31.1 59.6 19.8 21.7

718.2 186.1 370.8 142.4 463.7 209.3 151.1 287.8 89.4 123.1 366.3 138.3 299.3 176.3 149.5 157.4 412.2 156.4 377.8 141.6

71 88 62 84 68 76 87 79 83 85 69 89 70 86 81 80 77 81 70 83

Source: EUROSTAT (2013), European Commission (2013). a Gross inland consumption of energy divided by GDP (kg of oil equivalent per 1000 EUR). b Importance of a product’s impact on the environment in purchasing decisions, share of “very or rather important” in%.

According to Cecere et al. (2014), the main lock-in factors are costs, technologies and stakeholders. “The costs related to the adoption of a cleaner technology result in the persistence of an old technological regime, in which the existing knowledge, competences, infrastructure, and capital lead to entrenchment in a pollution-intensive trajectory” (Cecere et al., 2014: 1045). Because of these high costs for the transition from dirty to clean technologies, the lock-out from an old technological regime is highly dependent on the economic state of a country. However, this is not the only crucial factor; the country also needs new environmentally oriented knowledge and infrastructure. From an empirical point of view, the Eastern European countries still seem to be characterized by pollution-intensive technologies. These countries show very high levels of energy intensity (especially Bulgaria) indicating a great need or even potential for renewable energies in the future (see Table 2). The image of the share of renewable energies from 2008 to 2011 shows an interesting pattern. Sweden, Portugal, Finland and even Romania (31%) and Latvia (45%) are already characterized by a high proportion of renewable electricity resources, not least because of geographical pre-conditions. Sweden and Romania specialize in hydroelectric power stations. In Sweden, 49% of the electricity supply is generated by water power plants (Deutsch-Schwedische Handelskammer, 2010; Brunklaus et al., 2013; Deutsch-Rumänische AHK, 2012). From 2008 to 2011, countries such as Germany, Italy, Estonia, Lithuania and the Czech Republic showed a dynamic development in their use of renewable energy whereas Hungary, France, Bulgaria, Luxembourg and Slovakia can be characterized as stagnating countries (see Table 2). Technological capabilities play an important role in the realization of (eco)-innovation. On the basis of German panel data, Horbach (2008) shows that the improvement of technological capabilities (“knowledge capital”) by R&D triggers ecoinnovations. Canon de Francia et al. (2007) find that “. . .availability of greater technical knowledge within a company moderates its vulnerability in the face of the demands of new environmental regulations.” Recent literature shows that the proximity to the best knowledge infrastructure and further regional and location conditions favor eco-innovations (Cainelli et al., 2011; Horbach, 2014). Many fields of eco-innovation are relatively new (e.g. renewable energies, electro mobility) and are, therefore, more dependent on external sources of information and basic research activities compared with more established innovation fields. Therefore, universities and other research institutions play a significant role when it comes to eco-innovation. These institutions contribute to the regional availability of highly skilled employees who have an upto-date education in new research fields such as new energy technologies. As information flows seem to be particularly important for new technologies, local cooperation networks may also focus in particular on promoting eco-innovation (De Marchi, 2012). Ghisetti et al. (2015) enforce this argument. The authors claim that eco-innovation requires a broad knowledge base because the introduction of eco-innovations often requires institutional changes, the adoption of specific management systems, the implementation of complex regulations and diverse sustainable suppliers. Therefore, differences in eco-innovation activities and performance in different countries may stem from different levels of R&D input, availability of knowledge infrastructures, cooperation networks and activities, existence of regional “sustainable” suppliers and different “cultures” relating to the realization of organizational changes in a country’s firms. Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model EIST-191; No. of Pages 14

4

ARTICLE IN PRESS J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

Unfortunately, there is no data for R&D in the field of eco-innovation but the figures for general R&D reveal considerable differences among countries. Apart from the Czech Republic (1.3%), the R&D share (in% of GDP) in the Eastern European countries is clearly below the EU-average (1.9%), indicating a high dependence on foreign knowledge sources (see Table 2). Eco-innovations are also dependent on the environmental consciousness of the consumers and firms, which may be interpreted as an environmentally oriented demand pull effect. Since eco-friendly products are still more expensive (Rehfeld et al., 2007) compared to other goods, a disproportionally low consumption of eco-products may stem from a low per capita income in a country. From a public choice perspective, the environmental awareness of the population becomes even more important because environmentally sensitive citizens may put more pressure on governments to pass stricter environmental regulations (Cecere et al., 2014; Horbach, 1992). Opinion polls conducted by the European Commission (2013) show a lower environmental awareness in Eastern European countries compared to the European average. In the EU, 83% (in 2012) of the persons questioned stated that a product’s impact on the environment is very or rather important for them when deciding on a purchase. These figures amounted to only 62% in the Czech Republic and 71% in Bulgaria (European Commission, 2013; see Table 2). Furthermore, cost-savings as a motivation, especially those caused by the reduction of material and energy use may also be more important for eco-innovations because, in many cases, they are connected with less severe environmental impacts. For example, less material consumption signifies a reduction of waste and energy savings are normally accompanied by reductions of CO2 -emissions (Horbach et al., 2013). This motivation to introduce eco-innovations may be especially important for countries that have large manufacturing sectors such as the Czech Republic, Germany, Hungary, Romania or Slovakia (see Table 2). Based on the preceding theoretical and empirical considerations, we can derive the following hypotheses: Hypothesis 1. Regulation measures triggering eco-innovation are especially important for Eastern European countries because they are still locked-in in energy intensive technologies. Hypothesis 2. As eco-innovation requires a broad knowledge base, the still under-developed technological capabilities of the Eastern European countries need specific technology transfers from other countries and competitors. Hypothesis 3. Due to the high costs for the transition to cleaner production, the Eastern European countries are disproportionally dependent on subsidies.

3. Data basis and descriptive results for 19 EU countries The analysis uses data from the Community Innovation Survey (CIS) of the European Commission conducted in 2009 (CIS 2008). A separate module on eco-innovations was introduced for the CIS 2008. This special module largely follows the definition developed in the MEI project mentioned in Section 2. The CIS-questionnaire defines an eco-innovation as follows: “An environmental innovation is a new or significantly improved product (good or service), process, organizational method or marketing method that creates environmental benefits compared to alternatives. The environmental benefits can be the primary objective of the innovation or the result of other innovation objectives. The environmental benefits of an innovation can occur during the production of a good or service, or during the after-sale use of a good or service by the end user.” The following table contains a list of environmental benefits that an eco-innovation could have if it is produced either within the firm or from the after-sale use of a product by the user. In this survey, firms were asked whether any of these eco-innovations were introduced in response to existing or expected environmental regulations. They were also asked about the availability of financial support by governments, demand from customers, or voluntary codes or industry agreements. The micro-level firm data for the voluntary special module on eco-innovation is only available in the Safe Centre on the premises of EUROSTAT in Luxembourg and covers 19 countries and 121.395 observations. A descriptive analysis of this data with respect to different environmental innovation fields shows that in nearly all countries, the reduction of energy use is an important innovation field. This is relevant in particular for Germany, Hungary and Sweden (see Table 3). Furthermore, the recycling sector seems to be important for the Czech Republic, Germany, Hungary, Ireland, Luxembourg, and Portugal. On average, the Eastern European countries, except Hungary, are less eco-innovative compared to the other countries. This was to be expected considering the low level of expenditure on R&D in these countries. Apart from in Bulgaria, the recycling sector plays an important role in all of these countries. That is also plausible because the recycling sector predominantly requires low-skilled workers who usually receive low wages. A breakdown according to different trigger factors (see Table 4) shows that regulation activities seem to be much more important for Eastern European countries (EEC’s) when introducing eco-innovation (Czech Republic 61%, Hungary 61%, Lithuania 70%, Romania 66%, Slovakia 69%). This is in line with the observation that the environmental awareness of the population is lower in these countries (see Section 2) and that, apart from Hungary, the Eastern European countries are also more dependent on subsidies. This confirms the significant influence of the State on the realization of eco-innovation in these countries. However, the market demand already plays an important role in triggering eco-innovation, especially in Hungary. Interestingly, Environmental Management Systems (EMS) are already very important in the Eastern European countries.

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

5

Table 3 Specialization of 19 countries on different innovation fields. Countries

Environmental benefits within the enterprise

Material

Energy

Environmental benefits from after sales use of a good or service by the end user

CO2

Dangerous substances

Air, noise, soil, water

Recycling

Energy

Air, noise, soil, water

Recycling

6.0 5.4 14.5 33.3 6.2 19.2 13.4 19.9 22.0 13.0 9.9 32.7 7.8 5.1 16.6 22.0 12.8 30.4 7.3 15.3

10.0 5.7 16.9 24.4 10.3 17.2 16.1 34.1 19.4 16.0 12.9 30.6 9.4 7.3 22.4 30.9 12.0 30.5 10.9 17.8

10.6 8.3 22.4 37.6 15.2 16.8 15.3 34.5 17.6 23.7 11.8 27.8 12.3 5.1 20.9 35.5 18.6 29.6 13.8 20.8

8.7 8.6 31.4 36.7 14.5 23.3 23.8 30.3 34.2 25.1 10.0 48.0 6.5 12.1 22.9 46.9 18.9 27.6 14.8 24.7

8.8 3.4 21.4 35.2 10.8 23.2 15.2 21.2 20.6 23.9 9.7 31.6 8.5 9.2 21.9 28.0 16.7 33.5 11.2 19.8

8.1 3.9 21.0 27.7 8.6 14.8 10.7 20.0 14.9 23.9 9.4 20.7 9.9 3.5 18.9 28.8 17.0 26.8 10.7 17.5

6.2 3.7 20.4 23.2 7.5 15.4 11.6 13.4 22.6 22.5 7.5 29.9 4.9 7.4 15.2 31.4 11.8 21.9 10.5 16.3

Shares of answers “yes” in% Bulgaria Cyprus Czech Rep. Germany Estonia Finland France Hungary Ireland Italy Lithuania Luxembourg Latvia Malta Netherlands Portugal Romania Sweden Slovakia Total

11.6 6.8 20.4 36.6 15.2 23.0 17.9 37.3 18.9 10.3 12.6 26.5 9.7 7.8 13.8 27.9 16.9 31.0 11.1 17.8

13.6 8.7 24.3 42.4 16.4 23.6 18.5 42.7 22.5 14.5 14.3 31.6 10.2 9.6 19.7 30.0 18.3 36.6 12.9 20.7

Source: CIS 2008, own calculations. Table 4 Determinants of eco-innovations and introduction of environmental management systems (EMS). Countries

Eco-innovations from 2006 to 2008 were introduced in response to: Existing regulations or taxes

Introduction of EMS

Expected regulations

Subsidies

Market demand

Voluntary codes

22.8 23.5 43.6 30.4 45.4 33.1 27.5 54.5 28.9 30.3 53.1 23.9 27.6 38.7 23.5 25.8 36.5 28.9 52.0 31.7

10.2 13.0 15.6 10.6 11.8 10.5 12.8 6.5 13.7 24.9 23.0 8.6 14.6 16.2 16.1 8.9 17.8 9.0 12.8 14.4

16.9 16.7 24.1 32.3 39.5 55.3 32.4 48.8 37.8 28.1 47.3 26.9 19.5 25.0 33.0 31.3 32.8 37.4 25.4 32.2

21.8 51.2 42.1 30.3 59.1 49.1 42.5 49.7 41.2 31.9 39.8 67.4 64.3 27.5 31.2 56.7 32.7 35.8 37.1 39.6

Shares of answers “yes” in% Bulgaria Cyprus Czech Rep. Germany Estonia Finland France Hungary Ireland Italy Lithuania Lux. Latvia Malta Netherlands Portugal Romania Sweden Slovakia Total

36.3 29.0 61.0 34.7 56.8 30.1 41.0 60.9 37.5 42.3 70.1 22.6 51.4 38.7 26.4 42.5 66.4 21.8 69.4 43.4

8.2 24.2 27.2 21.8 34.1 – 16.7 26.1 – – 20.0 – 24.2 9.3 22.2 29.5 – 70.2 21.5 –

Source: CIS 2008, own calculations.

4. Econometric analysis To analyze the determinants of eco-innovation for different environmental technology fields and countries, binary probit models are applied. For each environmental field, a firm has to decide whether to introduce an innovation related to this field (Y = 1) or not (Y = 0). According to our theoretical analysis, different factors such as regulation or cost savings summarized by a vector x influence this decision. Therefore, we need an estimation of the probability



Prob Y =

1 x



= F(x, ˇ)

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model EIST-191; No. of Pages 14

6

ARTICLE IN PRESS J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

Due to the binary nature of our dependent variable, we use the probit model assuming normal distribution. The ␤ parameters reflect the impact of changes in x on the probability (Greene, 2008: 772). We calculate marginal effects that allow comparisons of the different eco-innovation areas. As the outcomes of the different eco-innovation fields may be correlated leading to inconsistent estimates of the simple probit models, multivariate probit models have been estimated as robustness checks (Appendix B). The multivariate probit model is similar to the seemingly unrelated regression (SUR) with the exception that the dependent variables are binary (Cappellari and Jenkins, 2003). The simulated maximum likelihood estimator of Geweke-Hajivassiliou-Keane (GHK) is used to evaluate the multivariate normal distribution. Our dependent variables are derived as described in the following section (see Appendix A for an exact definition of all variables). For each environmental innovation field (see Table 5), the firms had to respond to the following question: “During the three years 2006–2008, did your enterprise introduce a product (good or service), process, organizational or marketing innovation with any of the following environmental benefits” (EUROSTAT, 2009: 10). The answers were coded as yes/no options meaning that the resulting variables are dummies. To capture the determinants of the different eco-innovation fields, the following groups of correlated variables2 are considered: • • • • • • •

Regulation measures, subsidies, Market factors: Market demand, cost savings, Innovation inputs, Innovation objectives, Information sources, Organizational innovations, Control variables (size, sector and country dummies).

Regulation activities, environmentally related subsidies (envsubsidies) and general subsidies capture the influence of state activities on different eco-innovation fields. The variable voluntary denotes the relevance of voluntary codes or agreements for environmentally friendly practice. The current or expected market demand from customers, export activities and cost savings as a motivation for introducing eco-innovations are market-oriented factors that indicate the competition pressure the firm has to face. Innovation inputs include internal and external R&D activities, the acquisition of advanced machinery, equipment and computer hardware or software (equip), the purchase or licensing of patents and non-patented inventions (extknowledge), internal or external training for the personnel, activities for the market introduction of new or significantly improved products and processes (marketintro) and other activities such as feasibility studies or industrial engineering (otherinput). The variable cooperation represents cooperation arrangements on innovation. The econometric analysis also considers the innovation objectives denoting an increase in the range of goods and services (productrange), a replacement of outdated products or processes (replace), an entry into new markets (newmarkets), an increase of the market share (marketshare), an improvement of the quality of goods and services (quality), an improvement of the flexibility, an increase in the capacity for producing of goods or services, an improvement in health or safety, or a reduction in labor costs per unit of output. The importance of the following different information sources for the realization of the firm’s innovation activities are considered: Sources within the firm or firm group (internal), suppliers of equipment, materials, components or software, customers, competitors or other firms in the same sector, consultants or commercial laboratories or private R&D institutes, universities or other higher education institutions, government or public research (pubresearch), conferences, trade fairs or exhibitions, scientific journals and other publications, and professional and industry associations. Furthermore, the influence of organizational measures such as new business practices (e.g. supply chain management or lean production) (business org.), new methods of organizing work responsibilities and decision-making (e.g. team work, education or training systems) (labor org.) and new methods of organizing external relations (e.g. outsourcing or subcontracting) (orgexr) are tested. Country and sector dummies were also included. In a first step, probit models including all 18 countries3 were estimated (see Table 5). The analysis confirms a significant outcome of recent literature (e.g. Horbach et al., 2012) for single countries: There are highly positive marginal effects for regulations concerning “traditional end-of-pipe” fields such as air, noise, soil, water (22.8%) and dangerous substances (17.1%), whereas the respective values for material (2.4%) and energy savings (5.1%) are much lower. For the last-mentioned fields, cost savings are more important as motivation (material (7.4%) and energy (6.4%)). Environmental subsidies (envsubsidies) are especially important for innovations reducing CO2 . Not surprisingly, market demand plays an important role in eco-product-innovations, but also in the reduction of dangerous substances

2 We use the term “correlated variables” because the results of the models have to be interpreted with caution. There may be endogeneity problems especially concerning the choice of different innovation inputs. For these variables we explicitly test for endogeneity. In all cases where the Smith–Blundell test indicated the existence of endogeneity, we estimated the respective models without the variables under suspicion as a robustness check. In all analyzed cases the results did not change significantly. 3 Ireland could not be included because of incomplete data.

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

7

Table 5 Determinants of eco-innovation by different environmental areas (all countries). Determinants

Regulation Envsubsidies Demand Voluntary Cost savings Size Subsidies Export

Environmental benefits within the enterprise

Environmental benefits from after sales use of a good or service by the end user

Material

Energy

CO2

Dangerous substances

Air, noise, soil, water

Recycling

Energy

Air, noise, soil, water

Recycling

2.4** 2.6** 8.5** 5.8** 7.4** 1.1** – –

5.1** 8.6** 6.5** 9.5** 6.4** 2.0** – 1.3+

11.3** 12.5** 10.9** 11.0** 3.7** 2.5** – –

17.1** 3.6** 14.7** 11.4** 1.8* 0.9** – −1.4+

22.8** 8.9** 8.8** 12.7** 2.9** 1.8** 2.2** 1.4+

13.3** 4.0** 5.9** 15.5** – 1.2** −2.0* –

3.2** 12.2** 17.2** 5.7** 3.7** 0.6** – −3.9**

15.9** 10.8** 17.5** 9.5** 3.0** 0.6** – −2.7**

10.8** 5.9** 14.2** 13.3** – 0.3** −3.1** −3.4**

4.6** 1.4+ 1.8* 1.8* 1.7* 3.5** 1.7* −1.6*

3.3** 2.9** 2.7** – – 1.3+ 1.8** –

– – – 2.5** – 1.9** 2.0** –

1.8* – – – – 2.5** 1.8** –

1.7* – 1.8* – 1.5* – – –

– 2.0** 1.8* 1.9** – 1.4* – –

1.8* – – 1.9* – 4.8** 1.4+ –

2.0** – – 2.7** – 4.5** – –

– – – – 2.8** 4.4** – −2.4**

– 3.1** −1.6* – – – 1.5+ 2.1** 8.3**

−2.5** – – – – – 3.9** 2.4** 6.5**

−2.5** – −1.4+ – – 1.8*– 1.7* 9.1** 1.6+

– 3.1** – – −2.1** – −1.9* 10.9** –

−1.8* – – −1.9* – – 3.7** 12.6** −1.7+

– – – – – – 1.8* 7.0** –

– – 1.3+ 1.7* – 1.3+ – 2.6** 3.2**

−2.2** – 2.8** – – – – 11.8** –

– – 1.6* – – – – 7.7** –

1.3+ −1.4+ – – −2.9* – – −1.9+ – –

– – – – −4.2** – – −3.1** 2.7* –

– – – – −3.3** – 3.5+ −2.5* – –

−2.1** – 1.7* – – – – – 4.0** –

−1.5* −1.5+ – −3.0** – – – – – –

−1.4* – – – – – – – 2.6+ –

−2.3** – – – −4.1** – – −2.4* – –

−2.2** −1.8* – −3.2** – – – – – 3.2*

−2.6** – – −1.9+ – – −4.6* – – 3.5*

3.4** 2.9** 2.1 ** 26268 Chi2 (2) = 4.8+

2.2** 2.6** 2.5** 26281 Chi2 (2) = 3.05

– 3.6** 2.6** 26256 Chi2 (2) = 2.2

2.2** 2.7** 2.3** 26269 Chi2 (2) = 3.1

– 2.2** 1.6+ 26285 Chi2 (2) = 4.8+

4.4** 4.6** 2.3** 26281 Chi2 (2) = 3.1

– 2.8** 2.4** 26259 Chi2 (2) = 26.0**

– 2.2** – 26242 Chi2 (2) = 14.4**

– 3.7** 4.4** 26251 Chi2 (2) = 31.0**

Innovation inputs Internal R&D External R&D Equip Extknowledge Training Marketintro Otherinput Cooperation Objectives Productrange Replace NewMarkets MarketShare Quality Flexibility Capacity Health Laborcosts Info sources Internal Suppliers Customers Competitors Consultants Universities Pubresearch Conferences Journals Associations Org. innovation Business org. Labor org. Orgexr Number of obs. Smith–Blundell test of exogeneity (internal and external R&D)

Source: CIS 2008, own estimations. Notes: Marginal effects are reported (in%). The marginal effects for the continuous independent variables were calculated at their means. Concerning dummy variables the values report the change in probability for a discrete change of the dummy variable from 0 to 1. Only significant marginal effects (at least at 10% level) are considered. + ,* , ** denote significance at the 10%, 5% and 1% level, respectively. Sector and country ¨ that the marginal effect is not significant. Only eco-innovators are considered. dummies are included but not reported. ¨–means

(14.7%). This may be due to a growing environmental awareness among customers (see also Section 2, Table 2) meaning that green characteristics of products play a more important role. Interestingly, the export orientation is not significantly correlated to eco-innovations. The marginal effects for eco-product-innovations are all negative, which is in line with results from recent literature (De Marchi and Grandinetti, 2012). Concerning innovation inputs, internal R&D seems to be especially important for material (4.6%) and energy (3.3%) savings. That is plausible because material and energy savings often stem from changes of the individual production process. External knowledge is positively correlated to CO2 related innovations and eco-product-innovations. In fact, CO2 reduction technologies are relatively young meaning that knowledge acquired from basic research activities is necessary. Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

8

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

The objectives of eco-innovations were also tested: An increase in the range of goods and services does not seem to be a relevant objective for eco-innovations (negative or insignificant marginal effects). Interestingly, the introduction of material and energy saving innovations is also accompanied by a reduction in labor costs per unit of output indicating the fact that material, energy and savings in labor costs are a result of the change of the production process. Not surprisingly, health and safety objectives also trigger eco-innovations especially in the fields of dangerous substances and air, soil, noise and water-related eco-innovations. An analysis of the information sources shows no clear picture for the sample of all countries, whereas organizational innovations seem to be very relevant for nearly all types of eco-innovations. In a further step, specific aspects relating to Eastern European countries are analyzed. To this end, additional probit models were estimated for a group of “rich” countries (Germany, France, Netherlands, Luxembourg) (Table 6) and for the available Eastern European countries (Table 7). The econometric results show that for “traditional fields” such as air, noise, soil, water (25.2% versus 23.2%), recycling (17.2% versus 11.5%), and dangerous substances (19.2% versus 18.0%), the marginal effects for the importance of regulation measures are slightly higher for the Eastern European countries supporting Hypothesis 1 (see Section 2). This may be explained by a higher importance of pollution-intensive sectors connected with lower environmental technology standards in these areas in the Eastern countries. The “rich” countries, by contrast, have already adapted their abatement technologies in past years. For new emerging fields such as material or energy savings within firms, regulation measures do not yet play a significant role in Eastern Europe compared to the “rich” countries characterized by important regulation measures in these areas such as the renewable energy law in Germany. This may indicate regulation deficits in these countries. For nearly all environmental fields (except recycling), voluntary codes or agreements for environmentallyfriendly practice are less important in Eastern Europe, indicating a lower level of environmental awareness. This confirms the results of opinion polls (see Section 2). Apart from energy saving measures, environmentally related subsidies (envsubsidies) seem to be quantitatively more important for the Eastern European countries highlighting the lower financial performance of Eastern European firms (Hypothesis 3). In the “young” environmental technology field of energy saving measures, the Eastern European countries rely more on competitors as information sources (5.9% versus -5.1%). This argument is also supported by the fact that Eastern European countries are slightly more dependent on external R&D measures (external R&D 4.5%, equip 5%) indicating a technology transfer from West to East (Hypothesis 2). In the “rich” countries, internal information sources are significant whereas this is not the case in the Eastern European countries. The dependence of Eastern European firms on alliances and partnerships is also documented by the significant marginal effects of orgexr, denoting new methods of organizing external relations. This variable is significant for material (4.6%) and energy (3.1%) savings, CO2 -reductions (4.8%), dangerous substances (5.5%), energy saving products (4.4%) and recycling of products (4.5%). The results for innovation objectives in the two country groups are very similar. 5. Summary and conclusions The paper analyzes the determinants of eco-innovation activities for 19 different countries in nine technology fields. In 2009, a special module on eco-innovation was included in the Community Innovation Survey (CIS) allowing for such an analysis. A descriptive analysis of this data with respect to different environmental innovation fields shows that in nearly all countries, the reduction of energy use is important. This is especially relevant for Germany, Hungary and Sweden. Furthermore, the recycling sector seems to be important for the Czech Republic, Germany, Hungary, Ireland, Luxembourg and Portugal. On average, the Eastern European countries, apart from Hungary, are less eco-innovative compared to the other countries. This was to be expected considering the low level of expenditure on R&D in these countries. A breakdown according to different trigger factors shows that regulation activities seem to be much more important for Eastern European countries (EEC’s) for the introduction of eco-innovation. This is in line with the observation that the environmental awareness of the population is lower in these countries and that, apart from Hungary, the Eastern European countries are also more dependent on subsidies. This confirms the more important role of the State in realizing eco-innovation. From a policy perspective, it would be useful to redirect already existing transfer payments from Western to Eastern European countries to a more eco-friendly innovation system, thus supporting the technological capabilities of these countries. The econometric analysis for all involved countries confirms an important outcome of recent literature for single countries: Regulations are very important for “traditional end-of-pipe” fields such as air, noise, soil, water and also dangerous substances whereas their influence on material and energy savings is much lower. For the last-mentioned fields, cost savings are more important as motivation. Environmentally related subsidies are especially important for innovations reducing CO2 . Concerning innovation inputs, internal R&D seems to be particularly important for material and energy savings. That is plausible because material and energy savings often stem from changes of the individual production process. The objectives of eco-innovations were also tested: Interestingly, the introduction of material and energy saving innovations is also accompanied by the reduction in labor costs per unit of output. This highlights the fact that material, energy and savings to labor costs are an overall result of the change in the production process. In a further step, the specificities of the Eastern European countries are analyzed. To this end, additional probit models were estimated for a group of “rich” countries (Germany, France, Netherlands, Luxembourg) and for the available Eastern European countries. The econometric results show that for “traditional fields” such as air, noise, soil, water, recycling and dangerous substances, the marginal effects for the importance of regulation measures are slightly higher for the Eastern Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

9

Table 6 Determinants of eco-innovation by different environmental areas—results for Germany, France, Netherlands, Luxembourg. Determinants

Regulation Envsubsidies Demand Voluntary Cost savings Size Subsidies Export

Environmental benefits within the enterprise

Environmental benefits from after sales use of a good or service by the end user

Material

Energy

CO2

Dangerous substances

Air, noise, soil, water

Recycling

Energy

Air, noise, soil, water

Recycling

2.5* – 7.7** 9.4** 5.3** 1.2** – –

7.8** 9.8** 5.1** 12.4** 5.6** 1.8** – –

9.7** 12.9** 10.6** 13.5** 3.1* 2.3** – –

18.0** – 11.5** 12.9** – 1.1** – –

23.2** 6.9** 7.8** 15.4** – 1.3** – 2.7*

11.5** 2.7+ 7.2** 14.8** – 0.9** −2.3+ –

5.7** 12.5** 17.5** 8.1** – 0.6** – −4.8**

16.5** 8.7** 16.4** 11.7** – 0.5* – –

10.5** – 14.0** 14.0** – 0.3** – −3.0**

4.5** – – 2.3+ 3.9** 3.5** – –

3.2** 3.7** – 2.2+ – – – –

3.0* – – 2.2+ – – – –

– – – – – – 3.1* –

– 2.3+ – – – – – –

−2.4+ – 3.0** – – – – –

2.8* – – – 3.0** 2.6* – –

– – – – 2.2+ 2.3+ – −2.8**

– – – 3.2** – – – −2.5*

– 3.5** – 2.2+ – – – 4.6** 8.6**

– – – – – – 2.9* 3.1* 4.9**

– – −2.4* 2.3+ – – – 9.0** 2.8*

– 3.4** – – −2.7* – – 13.2** –

– – – – – – 4.8** 13.7** –

– – – – 3.2** – 2.3+ 7.3** –

– – – 3.0* – – – – 2.8*

−2.5* – 3.7** 2.1+ – – – 11.4** –

−2.0+ – 2.0+ 2.2* – – – 7.5** –

3.3** – – – – – 6.1+ – – −4.2+

2.5* −2.3+ – −5.1** – – – −4.4* – –

– – – – – – – – – –

– – 2.3+ – – – – – 4.4+ –

– −2.3+ – – – – – – – –

– – – – – – – 3.2+ – –

– – – – – – – – – –

– – – −3.6* – – – – – –

– – – −3.5* – – – – – –

4.2** 3.2** – 10539 Chi2 (2) = 1.7

– – 2.3+ 10539 Chi2 (2) = 1.1

– 2.5* 3.7** 10539 Chi2 (2) = 0.5

2.4+ – – 10539 Chi2 (2) = 3.3

2.5* – 2.2+ 10539 Chi2 (2) = 4.7+

5.3** 5.6** – 10539 Chi2 (2) = 0.7

– 2.9* – 10523 Chi2 (2) = 15.8**

– – – 10523 Chi2 (2) = 7.5*

2.9** 2.1+ 3.0** 10523 Chi2 (2) = 12.2**

Innovation inputs Internal R&D External R&D Equip Extknowledge Training Marketintro Otherinput Cooperation Objectives Productrange Replace NewMarkets MarketShare Quality Flexibility Capacity Health Laborcosts Info sources Internal Suppliers Customers Competitors Consultants Universities Pubresearch Conferences Journals Associations Org. innovation Business org. Labor org. Orgexr Number of obs. Smith–Blundell test of exogeneity (internal and external R&D)

Source: CIS 2008, own estimations. Notes: Marginal effects are reported (in%). The marginal effects for the continuous independent variables were calculated at their means. Concerning dummy variables the values report the change in probability for a discrete change of the dummy variable from 0 to 1. Only significant marginal effects (at least at 10% level) are considered. + ,* , ** denote significance at the 10%, 5% and 1% level, respectively. Sector and country ¨ that the marginal effect is not significant. Only eco-innovators are considered. dummies included but not reported. ¨–means

European countries. This may be explained by lower environmental technology standards connected with a higher concentration of pollution-intensive technologies in the Eastern countries. The “rich” countries, on the other hand, have already adapted their abatement technologies and reduced their energy intensity during the past years. For new emerging fields such as material or energy saving, regulation measures within firms do not yet play a significant role in Eastern Europe compared to the “rich” countries characterized by important regulation measures in these areas such as the renewable energy law in Germany. This indicates a lack of regulation. An implementation of similar regulation measures in the Eastern European countries would be useful but very cost-intensive. In light of the hypothesis of Acemoglu et al. (2012), temporary measures would perhaps be sufficient to redirect the economies of the Eastern European countries to a more eco-friendly innovation Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

10

Table 7 Determinants of eco-innovation by different environmental areas—Results for Bulgaria, Czech Republic, Estonia, Hungary, Lithuania, Latvia, Romania, Slovakia. Determinants

Regulation Envsubsidies Demand Voluntary Cost savings Size Subsidies Export

Environmental benefits within the enterprise

Environmental benefits from after sales use of a good or service by the end user

Material

Energy

CO2

Dangerous substances

Air, noise, soil, water

Recycling

Energy

Air, noise, soil, water

Recycling

– 7.4** 6.4** 3.4* 11.2** – −3.8* –

– 8.2** 3.5* 6.9** 9.9** 1.0+ – –

15.8** 10.4** 7.2** 8.5** 3.8** 1.6** – –

19.2** 7.2** 13.5** 10.9** 3.8* 1.3* – –

25.2** 11.8** 9.4** 9.8** 6.1** 4.5** 3.4* –

17.2** 8.1** 4.8** 15.4** – 2.1** −3.0+ –

– 9.0** 19.7** 3.7** 5.6** – – −5.0**

14.5** 10.7** 21.5** 6.7** 4.8** – – −4.9**

11.4** 11.3** 15.9** 10.0** – 0.8+ −5.8** −3.1*

5.1** – 3.9* – – 3.4* – –

– 4.5** 5.0** – – 2.5+ – 2.5+

– – – – – – – –

2.9* – – – – 3.9** – –

– – – – – – – –

– 4.1** – 2.6+ – – 3.2* –

– – – – – 5.3** 2.3+ –

– – – 3.0* – 7.1** 2.7+ –

– – – 3.1* – 6.1** – −2.3**

– 3.6** – – – – – – 9.1**

– – – – – – 5.1** – 6.9**

– – – – – 9.4** –

– 3.9** – – – – −4.1* 7.9** –

– – – – – – – 10.6** −4.7**

– 2.3+ – – – – 2.8+ 6.0** –

– – – – – – – – 3.8*

−2.5+ – 4.5** – – – – 9.4** –

– – – – – 4.1** – 3.8** –

– – – – – – – – – –

−2.2+ – – 5.9** −5.1* – – −3.3+ – –

– – – – – – 7.8* – −3.3+ –

– – – – – – – – – –

– – – −4.9** – – – – – –

– – – – – – – – – –

−4.3** – – 3.8* −5.7** – – – – 5.2+

– −3.8** – −3.8* – – – – – 8.1**

– – – – – – – – – 7.1**

– 4.9** 4.6** 7258 Chi2 (2) = 1.4

– 3.9** 3.1* 7258 Chi2 (2) = 3.7

– 4.0** 4.8** 7258 Chi2 (2) = 4.5

– – 5.5** 7258 Chi2 (2) = 1.2

– – – 7258 Chi2 (2) = 0.5

4.7** 4.2** – 7258 Chi2 (2) = 1.1

– 2.6+ 4.4** 7258 Chi2 (2) = 10.9**

– – – 7258 Chi2 (2) = 11.6**

– 3.2* 4.5** 7258 Chi2 (2) = 8.0*

Innovation inputs Internal R&D External R&D Equip Extknowledge Training Marketintro Otherinput Cooperation Objectives Productrange Replace NewMarkets MarketShare Quality Flexibility Capacity Health Laborcosts Info sources Internal Suppliers Customers Competitors Consultants Universities Pubresearch Conferences Journals Associations Org. innovation Business org. Labor org. Orgexr Number of obs. Smith–Blundell test of exogeneity (internal and external R&D)

Source: CIS 2008, own estimations. Notes: Marginal effects are reported (in%). The marginal effects for the continuous independent variables were calculated at their means. Concerning dummy variables the values report the change in probability for a discrete change of the dummy variable from 0 to 1. Only significant marginal effects (at least at 10% level) are considered. + ,* , ** denote significance at the 10%, 5% and 1% level, respectively. Sector and country ¨ that the marginal effect is not significant. Only eco-innovators are considered. dummies included but not reported. ¨–means

system. For nearly all environmental fields (except recycling), voluntary codes or agreements for environmentally friendly practice are less important in Eastern Europe, indicating a lower level of environmental awareness in these countries. Acknowledgements The author is grateful to an anonymous referee and the editor for valuable comments. The data of the Community Innovation Survey 2008 (CIS 2008) was provided by EUROSTAT. In this regard, I would like to thank Geneviève Villette and Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

11

Corina Mihaela Niculet for their valuable support. All results and conclusions are mine and not those of EUROSTAT, the European Commission or any of the national authorities whose data has been used. I dedicate this work to the memory of Klaus Rennings who unexpectedly passed away on September 16th 2015. Klaus was a very good friend of mine. The field of environmental innovation research has lost one of its most important representatives.

Appendix A. Table A1

Table A1 Definition of variables and descriptive statistics. Variables

Description

Mean

St. Dev.

Material Energy CO2 DangSubstances AirNoiseSoilWat Recycling

Environmental benefits within the enterprise (1 yes 0 no) Reduced material per unit of output Reduced energy use per unit of output Reduced CO2 footprint (total CO2 production) Replaced materials with less polluting substitutes Reduced soil, water, noise, or air pollution Recycled waste, water or materials

0.43 0.51 0.37 0.43 0.51 0.60

0.50 0.50 0.48 0.50 0.50 0.49

Energyprod EmissionsProd RecyclingProd

Environmental benefits from after sales use (1 yes 0 no) Reduced energy use Reduced air, water, soil or noise pollution Improved recycling of product after use

0.48 0.43 0.40

0.50 0.49 0.49

Determinants Regulation Envsubsidies Demand Voluntary Cost savings Size Subsidies Export

Existing environmental regulations (1 yes 0 no) Government grants for eco-innovation (1 yes 0 no) Current or expected market demand (1 yes 0 no) Voluntary Codes or agreements (1 yes 0 no) Reduce costs per unit of output as motivation (1 high 0 other) Number of employees in 2008, in 1000 Financial aid from regional, national or EU (1 yes 0 no) Selling goods or services in other (EU)-countries (1 yes 0 no)

0.41 0.14 0.29 0.37 0.23 0.32 0.16 0.52

0.49 0.35 0.46 0.48 0.42 2.38 0.37 0.50

Innovation inputs Internal R&D External R&D Equip Extknowledge Training Marketintro Otherinput Cooperation

Intramural R&D (1 yes 0 no) Extramural R&D (1 yes 0 no) Acquisition of machinery (1 yes 0 no) Acquisition of external knowledge (1 yes 0 no) Training for innovative activities (1 yes 0 no) Market introduction of innovation (1 yes 0 no) Other activities (e.g. feasibility studies) (1 yes 0 no) Cooperation arrangements on innovation (1 yes 0 no)

0.48 0.26 0.63 0.22 0.50 0.34 0.38 0.41

0.50 0.44 0.48 0.42 0.50 0.47 0.48 0.49

Objectives Productrange Replace NewMarkets MarketShare Quality Flexibility Capacity Health Laborcosts

Increased range of goods or services (1 high 0 other) Replace outdated products or processes (1 high 0 other) Enter new markets (1 high 0 other) Increase market share (1 high 0 other) Improve quality of goods or services (1 high 0 other) Improve flexibility for prod. Goods/services (1 high 0 other) Increase capacity for prod. goods/services (1 high 0 other) Improve health and safety (1 high 0 other) Reduce labor costs per unit output (1 high 0 other)

0.33 0.25 0.25 0.29 0.37 0.21 0.21 0.19 0.21

0.47 0.43 0.43 0.45 0.48 0.41 0.40 0.39 0.41

Info sources Internal Suppliers Customers Competitors Consultants Universities Pubresearch Conferences Journals Associations

Sources within the firm (1 high 0 other) Suppliers of equipment, materials (1 high 0 other) Clients or customers (1 high 0 other) Competitors of other firms of same industry (1 high 0 other) Consultants, commercial labs (1 high 0 other) Universities, other higher education institutes (1 high 0 other) Government or public research institutes (1 high 0 other) Conferences, trade fairs, meetings (1 high 0 other) Scientific journals (1 high 0 other) Professional and industry associations (1 high 0 other)

0.31 0.14 0.19 0.08 0.05 0.03 0.02 0.08 0.06 0.04

0.46 0.35 0.39 0.27 0.22 0.18 0.14 0.27 0.23 0.19

Endogenous var.

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model EIST-191; No. of Pages 14

ARTICLE IN PRESS J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

12 Table A1 (Continued) Variables

Description

Mean

St. Dev.

Org. innovation Business org. Labor org. Orgexr

New business practices for organizing work (1 yes 0 no) New methods of workplace organization (1 yes 0 no) New methods of organizing external relations (1 yes 0 no)

0.41 0.44 0.25

0.49 0.50 0.43

Sector dummies Sec1 Sec2 Sec3 Sec4 Sec5 Sec6 Sec7 Sec8 Sec9 Sec10 Sec11 Sec12 Sec13 Sec14 Sec15 Sec16 Sec17 Sec18 Sec19

1 yes, 0 no (for all sector dummies) Agriculture, forestry and fishery Mining Food products, beverages and tobacco Textiles, leather Wood, paper, printing Chemical industry, rubber and plastics, glass Basic metals and fabricated metals (Electrical) machinery and apparatus, motor vehicles Furniture and other products Electricity, gas, steam Water collection and treatment, sewerage, waste Construction sector Wholesale and retail trade Transport and logistics Accommodation, restaurants Information and communication Services: banking sector, assurances, real estate etc. Architectural and engineering offices Public sector and other services

0.003 0.01 0.06 0.04 0.05 0.10 0.07 0.12 0.05 0.01 0.03 0.08 0.13 0.07 0.02 0.04 0.04 0.05 0.02

0.06 0.10 0.24 0.19 0.23 0.30 0.26 0.32 0.22 0.12 0.18 0.27 0.34 0.25 0.15 0.19 0.18 0.21 0.14

Countries Bulgaria Cyprus Czech Republic Germany Estonia Finland France Hungary Ireland Italy Lithuania Luxembourg Latvia Malta Netherlands Portugal Romania Sweden Slovakia

1 yes, 0 no (for all country dummies) Bulgaria Cyprus Czech Republic Germany Estonia Finland France Hungary Ireland Italy Lithuania Luxembourg Latvia Malta Netherlands Portugal Romania Sweden Slovakia

0.03 0.004 0.08 0.08 0.01 0.03 0.18 0.03 0.02 0.17 0.01 0.007 0.006 0.007 0.12 0.09 0.08 0.04 0.01

0.17 0.06 0.27 0.27 0.12 0.16 0.38 0.16 0.15 0.38 0.11 0.09 0.07 0.08 0.33 0.29 0.27 0.19 0.12

Note: The descriptive statistics are related to the 41211 eco-innovative firms.

Appendix B. Table A2 Table A2 Determinants of eco-innovation (results for all countries): Comparison of results between multivariate and simple probit models. Determinants

Environmental benefits from after sales use of a good or service by the end user Energy

Regulation Envsubsidies Demand Voluntary Cost savings Size Subsidies Export

Air, noise, soil, water

Recycling

Probit

Multivariate Probit

Probit

Multivariate Probit

Probit

Multivariate Probit

0.08** 0.31** 0.43** 0.14** 0.09**

0.08** 0.31** 0.43** 0.14** 0.09** 0.02** – −0.10**

0.41** 0.28** 0.45** 0.24** 0.07** 0.01** – −0.07**

0.41** 0.28** 0.45** 0.24** 0.07** 0.01** – −0.07**

0.28** 0.15** 0.37** 0.35** – 0.00** −0.08** −0.09**

0.28** 0.16** 0.37** 0.35** – 0.00** −0.08** −0.09**

0.02** −0.10**

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model

ARTICLE IN PRESS

EIST-191; No. of Pages 14

J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

13

Table A2 (Continued) Determinants

Environmental benefits from after sales use of a good or service by the end user Energy

Air, noise, soil, water

Recycling

Probit

Multivariate Probit

Probit

Multivariate Probit

Probit

Multivariate Probit

Innovation inputs Internal R&D External R&D Equip Extknowledge Training Marketintro Otherinput Cooperation

0.04** – – 0.05* – 0.12** 0.03+ –

0.05* – – 0.05* – 0.12** 0.03+ –

0.05** – – 0.07** – 0.11** – –

0.05** – – 0.07** – 0.11** – –

– – – 0.07** – 0.12** – −0.07**

– – – 0.08** – 0.12** – −0.07**

Objectives Productrange Replace NewMarkets MarketShare Quality Flexibility Capacity Health Laborcosts

– 0.03+ 0.04* – 0.03+ – – 0.06** 0.08**

– 0.03+ 0.04* – 0.03+ – – 0.07** 0.08**

−0.06** – 0.07** – – – – 0.30** –

−0.06** – 0.07** – – – – 0.30** –

– – 0.04* – – – – 0.20** –

– – 0.04* – – – – 0.20** –

Info sources Internal Suppliers Customers Competitors Consultants Universities Pubresearch Conferences Journals Associations

−0.06** – – – −0.10** – – −0.06* – –

−0.06** – – – −0.11** – – −0.06* – –

−0.06** −0.05* – −0.08** – – – – – 0.08*

−0.05** −0.04* – −0.08** – – – – – 0.09*

−0.07** – – −0.05+ – – −0.12* – – 0.09*

−0.07** – – −0.05+ – – −0.12* – – 0.09*

Org. innovation Business org. Labour org. Orgexr Number of obs.

– 0.07** 0.06** 26259

– 0.07** 0.06** 26199

– 0.06** – 26242

– 0.05** – 26199

– 0.10** 0.12** 26251

– 0.10** 0.12** 26199

Notes: Only significant coefficients are considered. +,*, ** denote significance at the 10%, 5% and 1% level, re-spectively. Sector and country dummies are ¨ that the coefficient is not significant. Multivariate Probit Model: Likelihood ratio test of rho21 = rho31 = rho32 = 0: Chi2 included but not reported. ¨–means (3) = 5927.1**.

References Acemoglu, D., Aghion, P., Bursztyn, L., Hemous, D., 2012. The environment and directed technical change. Am. Econ. Rev. 102 (1), 131–166. Barbieri, N., Ghisetti, C., Gilli, M., Marin, G., Nicolli, F., 2015. A Survey of the Literature on Environmental Innovation Based on Main Path Analysis. SEEDS Working Paper Series 7/2015. University of Ferrara, Ferrara. Brunklaus, B., Hildenbrand, J., Saransini, S., 2013. ECO-innovative Measures in Large Swedish Companies—An Inventory Based on Company Reports. VINNOVA 2013: 03. Chalmers University of Technology, Gothenburg. Brunnermeier, S.B., Cohen, M.A., 2003. Determinants of environmental innovation in US manufacturing industries. J. Environ. Econ. Manag. 45 (2), 278–293. Cainelli, G., Mazzanti, M., Zoboli, R., 2011. Environmental innovations, complementarity and local/global cooperation: evidence from North-East Italian industry. Int. J. Technol. Policy Manag. 11 (3/4), 268–328. Canon de Francia, J., Garcés-Ayerbe, C., Ramírez-Alesón, M., 2007. Are more innovative firms less vulnerable to new environmental regulation? Environ. Resour. Econ. 36, 295–311. Cappellari, L., Jenkins, S.P., 2003. Multivariate probit regression using simulated maximum likelihood. Stata J. 3 (3), 278–294. Cecere, G., Corrocher, N., Gossart, C., Ozman, M., 2014. Lock-in and path dependence: an evolutionary approach to eco-innovations. J. Evol. Econ. 24 (5), 1037–1065. Cleff, T., Rennings, K., 1999. Determinants of environmental product and process innovation—evidence from the Mannheim innovation panel and a follow-up telephone survey. Eur. Environ. 9 (5), 191–201. De Marchi, V., Grandinetti, R., 2012. Who are the green innovators? An empirical analysis of firm’s level factors. In: DRUID 2012 Conference Paper, Copenhagen. De Marchi, V., 2012. Environmental innovation and R&D cooperation: empirical evidence from Spanish manufacturing firms. Res. Policy 41, 614–623. Del Rio Gonzalez, P., 2009. The empirical analysis of the determinants for environmental technological change: a research agenda. Ecol. Econ. 68, 861–878. Deutsch-Rumänische, AHK, 2012. Erneuerbare Energien in Rumänien: Elektrische Energie, Bukarest. www.ahkrumaenien.ro. Deutsch-Schwedische Handelskammer, 2010. Die Klimaschutzziele Schwedens und Deutschlands sind ambitioniert, Stockholm. http://www.handelskammer.se/de/news/energiepolitische-ziele-deutschland-und-schweden. European Commission, 2013. EU skills panorama. http://euskillspanorama.ec.europa.eu/. EUROSTAT, 2009. The Community Innovation Survey 2008 (CIS 2008)—The Harmonised Survey Questionnaire. EUROSTAT, Luxembourg.

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005

G Model EIST-191; No. of Pages 14

14

ARTICLE IN PRESS J. Horbach / Environmental Innovation and Societal Transitions xxx (2015) xxx–xxx

EUROSTAT, 2013. Statistical Data. EUROSTAT, Luxembourg http://epp.eurostat.ec.europa.eu. Frondel, M., Horbach, J., Rennings, K., 2007. End-of-pipe or cleaner production? An empirical comparison of environmental innovation decisions across OECD countries. Bus. Strat. Environ. 16 (8), 571–584. Ghisetti, C., Marzucchi, A., Montresor, S., 2015. The open eco-innovation mode. An empirical investigation of eleven European countries. Res. Policy 44, 1080–1093. Green, K., McMeekin, A., Irwin, A., 1994. Technological trajectories and R&D for environmental innovation in UK firms. Futures 26, 1047–1059. Greene, W.H., 2008. Econometric Analysis, Sixth ed. Pearson International, New Jersey. Hemmelskamp, J., 1999. Umweltpolitik und technischer Fortschritt, Schriftenreihe des Zentrums für Europäische Wirtschaftsforschung, Physica, Heidelberg, New York. Horbach, J., 1992. Neue Politische Ökonomie und Umweltpolitik, Campus, Frankfurt, New York. Horbach, J., 2008. Determinants of environmental innovation—new evidence from German panel data sources. Res. Policy 37, 163–173. Horbach, J., Rammer, C., Rennings, K., 2012. Determinants of eco-innovations by type of environmental impact—the role of regulatory push/pull, technology push and market pull. Ecol. Econ. 78 (2012), 112–122. Horbach, J., Oltra, V., Belin, J., 2013. Determinants and specificities of eco-innovations—an econometric analysis for France and Germany based on the community innovation survey. Ind. Innov. 20 (6), 523–543. Horbach, J., 2014. Do eco-innovations need specific regional characteristics? An econometric analysis for Germany. Revi. Regional Res. 34 (1), 23–38. Horbach, J., 2015. The role of environmental policy for eco-innovation: theoretical background and empirical results for different countries. In: Crespi, F., Quatraro, F. (Eds.), The Economics of Knowledge. Innovation and Systemic Technology Policy, Routledge. Jaffe, A., Palmer, K., 1997. Environmental regulation and innovation: a panel study. Rev. Econ. Stat. X, 610–619. Jaffe, A.B., Newell, R.G., Stavins, R.N., 2002. Environmental policy and technological change. Environ. Resour. Econ. 22, 41–69. Johnstone, N. (Ed.), 2007. Environmental Policy and Corporate Behaviour. Edwar Elgar, Northhampton. Kemp, R., Pearson, P., 2008. Final report MEI project about measuring eco-innovation, Maastricht, www.merit.unu.edu\MEI. Kesidou, E., Demirel, P., 2012. On the drivers of eco-innovations: empirical evidence from the UK. Res. Policy 41, 862–870. Lanoie, P., Laurent-Lucchetti, J., Johnstone, N., Ambec, S., 2011. Environmental policy, innovation and performance: new insights on the porter hypothesis. J. Econ. Manag. Strat. 20 (3), 803–842. Mazzanti, M., Zoboli, R., 2006. Examining the factors influencing environmental innovations, FEEM working paper series 20. 2006, Milano. Popp, D., 2006. International innovation and diffusion of air pollution control technologies: the effects of NOx and SO2 regulation in the US, Japan, and Germany. J. Environ. Econ. Manag. 51 (1), 46–71. Rehfeld, K.M., Rennings, K., Ziegler, A., 2007. Integrated product policy and environmental product innovations: an empirical analysis. Ecol. Econ. 61 (1), 91–100. Rennings, K., 2000. Redefining innovation—eco-innovation research and the contribution from ecological economics. Ecol. Econ. 32, 319–332. Rennings, K., Zwick, T., 2002. The employment impact of cleaner production on the firm level—empirical evidence from a survey in five european countries. Int. J. Innov. Manag. (IJIM) 6 (3), 319–342. Rennings, K., Rexhäuser, S., 2011. Long-term impacts of environmental policy and eco-innovative activities of firms. Int. J. Technol. Policy Manag. 11 (3/4), 274–290. Veugelers, R., 2012. Which policy instruments to induce clean innovating? Res. Policy 41, 1770–1778. Zeppini, P., van den Bergh, J.C.J.M., 2011. Competing recombinant technologies for environmental innovation: extending Arthur’s model of lock-in. Ind. Innov. 18 (3), 317–334.

Please cite this article in press as: Horbach, J., Empirical determinants of eco-innovation in European countries using the community innovation survey. Environ. Innovation Soc. Transitions (2015), http://dx.doi.org/10.1016/j.eist.2015.09.005