Hang the low-hanging fruit even lower - Evidence that energy efficiency matters for corporate financial performance

Journal of Cleaner Production 147 (2017) 66e74

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Hang the low-hanging fruit even lower - Evidence that energy efficiency matters for corporate financial performance Anne Bergmann, Jan Niklas Rotzek, Martina Wetzel, Edeltraud Guenther* Technische Universitaet Dresden, Faculty of Business and Economics, Chair of Environmental Management and Accounting, Muenchner Platz 1/3, 01062, Dresden, Germany

a r t i c l e i n f o

a b s t r a c t

Article history: Received 6 June 2016 Received in revised form 13 January 2017 Accepted 13 January 2017 Available online 17 January 2017

Energy efficiency measures are often called low-hanging fruit. First, they significantly lower the energy consumption and therefore contribute to combat global warming. Second, they are considered shortterm and cost-effective investments. However, there still exists a gap between existing profitable investments to increase energy efficiency and the corporate reality where organizations do not implement these measures, the so-called energy efficiency gap. This analysis aims to prove that an increasing level of corporate energy efficiency is directly related to an improved corporate financial performance. The study bases on a multiple regression analysis and considers the manufacturing industry worldwide. Findings indicate a significant positive link. Hence, the study reveals that managers should pay more attention to the implementation of energy efficiency measures, even though they incorporate investment costs. The analysis further contributes to recent research as it takes into account the impacts on corporate financial performance from activities along the corporate value chain. The regression model specifically includes the nine activities of the corporate value chain as variables to control for effects on corporate financial performance. Since gained results provide a higher predictive power, the study calls future research to explicitly consider specific value chain activities when analyzing impacts on corporate financial performance. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Energy efficiency Corporate financial performance (CFP) Corporate value chain Manufacturing industry Energy efficiency gap Corporate environmental performance (CEP)

1. Introduction “There's a lot of low-hanging fruit – this is the area where we can have the greatest environmental impact while making sure that we're creating good jobs and saving businesses and consumers money. So it's a win-win.” Barack Obama before signing the US Energy Efficiency Improvement Act of 2015 (White House, 2015, no pages) While the IEA (2015, p. 36) states that “energy production and use accounts for around two-thirds of global greenhouse gas (GHG) emissions today, of which carbon dioxide (CO2) is the great majority” and calls for a reduction across the board, the world wide

* Corresponding author. E-mail addresses: [email protected] (A. Bergmann), jan_niklas. [email protected] (J.N. Rotzek), [email protected] (M. Wetzel), [email protected] (E. Guenther). http://dx.doi.org/10.1016/j.jclepro.2017.01.074 0959-6526/© 2017 Elsevier Ltd. All rights reserved.

energy consumption and related GHG emissions continue to increase. These emissions have already heavily affected global warming and continue to do so today (IPCC, 2014). Energy efficiency and its further increase represent one of the pathways in order to mitigate climate change (European Commission, 2012). A direct reduction of 25% could be reached by using currently available best technologies. An additional 20% increase of energy efficiency could be reached through innovation (IPCC, 2014). Prior research comes to the conclusion that, very often, corporations neither recognize the direct impacts of climate change nor take appropriate action to hinder those impacts (Schmid, 2004). In addition to that lack of information (Sardianou, 2008), energy efficiency measures are often related to investment costs (Brunke et al., 2014). Since, on the other hand, the costs for emitting GHGs are still too low to have an effect on corporate decisions, these costs still represent a type of external effect which corporations are not accountable for (TEEB, 2010). Providing incentives for corporations to increase their current level of energy efficiency represents a solution to reduce current

A. Bergmann et al. / Journal of Cleaner Production 147 (2017) 66e74

GHG emissions (Worrell et al., 2009). According to Schmid (2004), there are two options: either the internalization of external effects, e.g. by improving the current GHG trading certificate system; or corporations discover other relevant motives which are related to lower costs and, as a result, to an improved competitiveness (Olson, 2014). Thollander and Ottosson (2010) further conclude that a higher level of energy efficiency leads to improved corporate financial performance (CFP). Considering that CFP is an essential indicator for firm performance and corporate survival in the long term (Hamann et al., 2013), corporations should be quite willing to implement energy efficiency measures. However, there still exists a gap between existing profitable investments to increase energy efficiency and the corporate reality where organizations do not implement these measures, the socalled energy efficiency gap (first analyzed by Hirst and Brown, 1990). These insights have even found their way into the encyclical letter, ‘Laudato si’, of Pope Francis (2015). Corporations often focus on other, more economically relevant issues (Sardianou, 2008), such as reducing personnel expenses. They further lack appropriate decision support tools regarding measures for energy efficiency (Trianni et al., 2016) and appropriate cost-benefit analyses (Bunse et al., 2011). This study tackles exactly the above-described gap and aims to prove that an increasing level of corporate energy efficiency is directly related to an improved CFP and, thus, helps corporations to survive in the long run. To the best of our knowledge, there exists no prior empirical study regarding this relationship. The paper's novelty stems from the narrow perspective on the link between energy efficiency and financial performance of corporations. The study applies a multiple regression analysis and considers the manufacturing industry worldwide. Findings will emphasize that benefits in terms of improved CFP outweigh related efforts to increase energy efficiency. This will foster corporate decision-makers to redirect more attention on energy efficiency and related improvement measures (Pye and McKane, 1999). An increasing level of energy efficiency will also contribute to cleaner production and climate change mitigation (Virtanen et al., 2013). In addition to prior studies that assess a relationship to CFP, this analysis further presents a first approach to systematically consider impacts of corporate value chain activities (Porter, 1985) on CFP. The study adds the nine value chain activities as control variables in the estimated regression model. Therewith, current practices of controlling for financial impacts are extended and the study contributes to recent research approaches. The present paper is organized as follows: The following Section 2 provides a literature review on the topics corporate energy efficiency, its barriers, and analyzes studies which focus on the link between increasing corporate energy efficiency and CFP. It further determines impacts of corporate value chain activities on CFP. As a result, two hypotheses are deduced. Section 3 then presents the chosen method and material. The results of the multiple regression analysis as well as the discussion will be presented in Section 4. Concluding remarks are provided in Section 5. 2. Literature review and hypotheses development 2.1. Corporate energy efficiency and its barriers “Energy efficiency is simply the ratio of energy services out to energy input.” (Herring, 2006, p. 11) Phylipsen et al. (1997, p. 717) specify energy services as “the amount of human activity (e.g. heating a room to a certain temperature, transporting goods over a certain distance, producing a certain amount of steel)”. Prior research provides a bouquet of indicators to express the term of corporate energy efficiency and highlights that the application of

67

existing indicators depends on the situation and decision to be made (Bunse et al., 2011). Scientists and policy-makers encourage corporations to enhance their current level of energy efficiency, e.g. via Energy Management Systems along the standard ISO 50001:2011 (Karcher and Jochem, 2015). In general, management attention needs to be directed towards the diverse benefits related to an increase in the current energy efficiency level (Pye and McKane, 1999). Many prior studies conclude that energy efficiency fully addresses all three aspects of the triple bottom line (Bunse et al., 2011). Energy efficiency measures also contribute to energy related as well as nonenergy related benefits (Trianni et al., 2014). Well-known examples for non-energy related benefits are increased profitability and improved product quality and output (Henriques and Catarino, 2016). Yet, there still exists a gap between the potential that energy efficiency measures entail and its implementation by organizations. This phenomenon is defined as the energy efficiency gap and has been researched by several scholars (Chai and Baudelaire, 2015). To date, barriers which hamper the implementation and cause the energy efficiency gap have been the main research focus. For instance, Trianni et al. (2016) provide an overview of empirical studies on industrial energy efficiency barriers. Recent research identifies economic barriers as almost always the primary issue (Cagno and Trianni, 2014). Among the several reasons for the existence of economic barriers, “technical risks, limited access to capital, and other priorities for financial investments” (Brunke et al., 2014, p. 514) tend to dominate. In addition, hidden costs of energy efficiency investments play a major role (Schmid, 2004). Since prior analyses also identified economic issues such as cost reductions (Apeaning and Thollander, 2013) and potential access to funding (Meath et al., 2016) as the most important driving forces for energy efficiency (Lee, 2015), this study will focus on the economic aspect. However, existing research that considers barriers and drivers often elaborates on the underlying motives and mechanisms (Hrovatin et al., 2016). For instance, Cagno and Trianni (2014) consider the evaluation's perspective and find that barriers may differ when evaluating them by technology area or at the company level. In contrast, this study takes up the idea of the research stream which investigates the relationship between environmental and financial performance within corporations. This research stream bases on the idea that a redirection of management attention towards the analyzed issue is possible when an overall positive effect on CFP can be proven because CFP is an essential indicator for firm performance and corporate survival in the long term (Hamann et al., 2013). We challenge the hypothesis that energy efficiency represents a low-hanging fruit as it reduces energy costs as well as GHG emissions at the same time. Instead, we aim to hang the lowhanging fruit even lower by contributing with the first empirical proof that an increasing level of corporate energy efficiency is also directly related to an improved CFP. To do so, we first summarize prior empirical studies from the field. Afterwards, we present our particular research goal and its contribution to research in more detail. 2.2. Prior research on the relationship between energy efficiency and CFP For more than 40 years, scholars have analyzed the relationship between corporate environmental performance (CEP) and CFP (Guenther et al., 2011). Meta-studies provide evidence that corporations managing their CEP provide a higher CFP than competitors which do not place as much value on corporate environmental issues (Endrikat et al., 2014). One of the most recently published

68

A. Bergmann et al. / Journal of Cleaner Production 147 (2017) 66e74

meta-analyses includes about 2200 studies and findings indicate an overall positive link from environmental, social, and governance criteria to CFP (Friede et al., 2015). This leads to an overall win-winsituation, for corporations as well as for the natural environment. Despite these overall positive tendencies, the level of analysis for the CEP construct matters since CEP itself represents a multidimensional construct and consists of two interrelated dimensions, namely environmental management performance and environmental operational performance (Trumpp et al., 2015). Energy efficiency can and should be controlled on both dimensions and is of great importance regarding the worldwide goal to mitigate climate change (Lee, 2015). We expect to find many prior empirical studies which analyze in how far an increase of energy efficiency contributes to CFP in a corporate environment. A conducted systematic literature review did not detect one single study with our chosen focus. However, several articles analyze the link from energy efficiency to productivity (Boyd and Pang, 2000) and find a substantial impact from energy efficient technologies on productivity (Kounetas et al., 2012). A first group of identified empirical studies considers the Chinese iron and steel industry. Here, scholars detect a positive relationship between energy efficiency and gross industrial output value as a measure for productivity (Zhang and Wang, 2008). The findings of He et al. (2013) support these results, but the authors also find a low level of average energy efficiency, namely 61.1% between the years 2001e2008. An improved energy efficiency level could be attained by environmental regulations, which would further support productivity growth (Zhang et al., 2011). A second group focuses on the US iron and steel industry and identifies several energy efficiency measures where the productivity benefits are quantifiable. A review of more than 70 case studies comes to the conclusion that investments in energy efficiency measures can lead to a significant productivity boost (Worrell et al., 2003). The study further highlights that the identified overall increase of productivity can be associated with nonenergy benefits, such as reduced input material and waste, improved quality of products, or improved worker morale - all leading to direct as well as indirect economic benefits (Worrell et al., 2003). A direct relationship from energy efficiency to CFP can be assumed (Dangelico and Pontrandolfo, 2015). To date, a quantification of those co-benefits represents a research gap (Worrell et al., 2009). Support for this statement stems from a recent study investigating the relationship between the number of energy saving technologies implemented to return on sales, finding no consistent significant impact (Pons et al., 2013). The case for energy efficiency could be greatly strengthened if efficiency advocates understood how managers decide, especially regarding financial aspects (Pye and McKane, 1999). The present study follows this consideration that energy efficiency measures will gain more attention by managers when a direct link between an increasing level of corporate energy efficiency and an increased CFP can be demonstrated. The study's focus on CFP stems from the fact that many corporations are confronted with shareholder pressure on shorter-term financial performance (Porter and Kramer, 2011). Considering that energy efficiency leads to direct and non-direct benefits, we formulate the following central hypothesis: H1: Increasing corporate energy efficiency positively affects corporate financial performance. 2.3. Impacts of value chain activities Although this study analyzes the link from an increasing level of corporate energy efficiency to CFP, it is of great importance to

control for other issues which also might significantly impact CFP. Above-presented prior studies which assess the CEP-CFP relationship often control for industry, firm size, or country effects. In addition, further financial indicators such as financial risk, R&D, or advertising intensity are used to control for an effect on CFP (Endrikat et al., 2014). We agree with the approach of controlling for impacts on CFP by using control variables. In the following we describe why we aim to extend the currently used bundle of control variables by adding the full range of corporate value chain activities. The value chain structures the activities of a company into strategically relevant fields in order to identify strengths and weaknesses of the company to deal with existing and potential opportunities and threats (Porter, 1985). Whereas primary activities focus on the core process concerning the production of a good or service and its sale, secondary activities support these. In order to integrate the idea of a circular economy, disposal can be added as a primary activity. The traditional linear concept of the value chain by Porter (1985) has received critique as it assumes closed-loop resources instead of closed-loop processes (Hartman and Stafford, 1998). A circular presentation of the value chain (DIN, 2014) brings this character more to the forefront and better signifies the close interconnection of all processes. As distinguished from supply chain thinking, value chain thinking “provides the enabling (internal) business environment for the development of sustainable competitive advantage” (Fearne et al., 2012, p. 576). Taking all activities of the value chain together, they contribute to the value of a corporation, which are reflected in the performance and competitiveness of an organization (Guenther and Scheibe, 2005). Hence, when studying the relationship to CFP, the corporate value chain and its impact should be considered as well. Furthermore, we follow Kung et al. (2012), who state that research also needs to take the individual impacts of the value chain elements into consideration. We add proxies for all nine activities of the value chain as control variables and predict: H2: Each corporate value chain activity individually affects corporate financial performance. 3. Material and method The current analysis aims to study the relationship between increasing corporate energy efficiency as an independent variable and CFP as the dependent variable by also taking into account the impacts of specific corporate value chain activities as variables to control for effects on CFP. This section describes the sample and all variables in detail. 3.1. Sample The chosen sample focuses on the manufacturing industry following Klassen and Whybark (1999, p. 599), who state that “Customers, suppliers, and the public are increasingly demanding that businesses in general, and manufacturing firms in particular, minimize any negative impact of their products and operations on the natural environment.” This might also be one of the reasons why so many prior studies analyzing corporate energy efficiency focus on that industry (Trianni et al., 2016). The focus, moreover, allows to better control for several striking features of that particular sector (Klassen and Whybark, 1999). We also enhance comparability of gained results and do not have to control for industry effects (Porter and Kramer, 2011). Since the majority of existing studies which aim to link CEP to disclosure or CFP focus on one distinct country (predominantly the United States) (Guenther et al., 2015), the current analysis overcomes this limitation as we do not limit our sample to a specific

A. Bergmann et al. / Journal of Cleaner Production 147 (2017) 66e74

country (Pons et al., 2013). Data stems from the Thomson Reuter databases Datastream Worldscope and Asset4, which cover financial as well as environmental data from worldwide stock listed corporations. The study utilizes multiple regression analysis and runs calculations with SPSS. The first screening of data leads to a total data set of 2648 corporations from the manufacturing industry. Incomplete data for the independent variable led to a reduced final sample of 650 corporations. The resulting sample had been winsorized to 5% (Dixon and Yuen, 1974). Fig. 1 depicts the sample, which is dominated by Japanese companies, followed by the United States and European countries. In Table 1, the six sub-sectors of the manufacturing industry are reported.

3.2. Variables 3.2.1. Independent variable: Increasing corporate energy efficiency (EE) Among the large range of indicators for measuring corporate energy efficiency, the ratios from energy consumption either to a monetary value (‘energy intensity’) or to a specific unit (‘specific energy consumption’) represent the most typical ones (Bunse et al., 2011). Both indicators concentrate on the operational performance level of CEP (Trumpp et al., 2015). A lot of studies focus on that level “to measure the actual results of environmental management and to avoid the subjectivity that is often associated with EMP [environmental management performance] evaluation” (Trumpp and Guenther, 2017, p. 51). This study also focuses on that level and uses the measure of total direct and indirect energy consumption as provided by the database Thomson Reuters Asset4. As the variable should also control for the energy demand of a particular corporation, it seems to be appropriate to relate it to the size of a corporation. There are basically three possibilities to measure size: via employees, assets, or sales (Weinzimmer et al., 1998). We refer to sales as this measure represents a more up-todate measure in contrast to employees. Moreover, sales represents an economic measure which should be used for depicting value added (He et al., 2013). Following Modi and Mishra (2011), we

69

Table 1 Industry affiliation of sample firms. Sub-sector of the manufacturing industry

Observations

Relative proportion

Industrial Basic Materials Cyclical Consumer Goods/Services Non-Cyclical Consumer Goods/Services Technology Healthcare Total

175 156 125 76 67 51 650

26.9% 24.0% 19.2% 11.7% 10.3% 7.8%

rely on net sales and follow Anderson et al. (2003) that it depicts a proxy for volume. We place net sales in the numerator as the comparison of the development of this formula between two periods of time results in a positive value when the corporation increases its level of energy efficiency. We measure the corporate increase in energy efficiency by relying on the formula which relates net sales in US$ to total energy consumption in GJ and compare the development of this formula from one year prior to the year of the CFP.

3.2.2. Dependent variable: Return on assets (ROA) In general, CFP can be depicted either by accounting- or marketbased measures. Following Endrikat et al. (2014, p. 740), “Accounting-based measures capture a firm's efficiency at using their assets to generate value (Peloza, 2009) and reflect internal capabilities and performance”. As we specifically focus on efficiency, we rely on an accounting-based measure. Regarding prior research on the CEP-CFP relationship, return on assets (ROA) represents not only the most commonly used measure (Guenther et al., 2011), but it also makes sense to rely on this indicator as energy efficiency improvements and investments are often related to changing assets. According to Kaplinsky and Morris (2001), ROA is also more indicative of corporate financial characteristics than return on equity. We measure ROA by dividing net income by total assets at the beginning of the year.

Fig. 1. Country affiliation of sample firms. ‘Other’ includes: Chile (2), Luxembourg (2), Portugal (2), Singapore (2), Bermuda (1), Indonesia (1), Israel (1), Malaysia (1), Thailand (1), Turkey (1).

70

A. Bergmann et al. / Journal of Cleaner Production 147 (2017) 66e74

3.2.3. Control variables As outlined in the literature review, we analyze the relevance of the value chain activities' effects on CFP. The value chain comprises procurement, production, sales, and disposal as primary activities; R&D/innovation, logistics, HR/organization/employees, marketing, and accounting are considered secondary activities (DIN, 2014). Marketing will be matched with sales while logistics will be included in the variable for measuring procurement. A description is provided in Table 2. Drawing from prior empirical studies in our field, we further control firm size, financial measures that have an impact on CFP, and the countries' performances regarding climate change (Table 3). The resulting model reads as follows:

ROA2012 ¼ b0 þ b1 EE20112012 þ b2 PROC LOG2011 þ b3 PROD2011 þ b4 SAL MARK2011 þ b5 DISP2011 þ b6 R&D2011 þ b7 HR2011 þ b8 ORG2011 þ b9 ACC LEV2011 þ b10 ACC LIQU2011 þ b11 SIZE2011 þ b12 CCPI2011 þ ε2011

4. Results and discussion 4.1. Findings Table 4 presents descriptive statistics. Results of the estimated regression model are depicted in Table 5. Since variables provide different units of measurement, the results section refers to values of standardized beta. Table 5 also contains values for

unstandardized beta. We tested for multicollinearity (Petter et al., 2007) and found none, because each variance inflation factor (VIF) is below the threshold of 10 (Table 5). The F-statistics indicate a statistically significant regression model (Adj. R2 ¼ 0.463, F ¼ 41.547, p ¼ 0.000). There is a significant positive relationship between EE and ROA in 2012 (stand. b ¼ 0.133, p < 0.01). In addition, six from nine value chain variables, which we consider as control variables, provide significant b-coefficients. Only PROD, R&D, and ACC_LEV do not significantly influence the model. Regarding the two further control variables, CCPI has a significant influence, SIZE does not. As hypothesized, there is a strongly significant positive relationship between EE and ROA. Focusing on our deduced formula for measuring EE, there are three possible cases for interpretation as to why EE has a positive influence on ROA. First, if net sales (numerator) stay at the same level while consuming less total energy (denominator), this might lead to lower energy costs and, subsequently, to a resulting higher return and ROA. This can also be achieved through growing net sales while keeping the total energy consumption on the same level. Considering the third case, nominator and denominator can change simultaneously and might provide a combined effect. To sum up, hypothesis 1 can be proven. Hypothesis 2 states that each step of the corporate value chain individually affects CFP. The following analysis considers each value chain variable in detail. The first corporate value chain variable PROC_LOG has a negative and significant influence on CFP, as expected (stand. b ¼ 0.166, p < 0.01). Specifically, a more intensive use of material and energy raises the capital intensity while it negatively affects CFP, supporting results of King and Lenox (2001). As our investigation measures a short-term horizon, corporations might further consider investments as financial loads, while those investments need more than two years to amortize.

Table 2 Variable definition for corporate value chain activities. Activity

Variable and measurement

Expected relationship to CFP

 Companies with a higher capital intensity have to take more efforts to integrate efficiency measures as more assets and material uses have to be managed (Hambrick, 1983)  Negative relationship of capital intensity to CFP is expected (King and Lenox, 2001)  Since Zeng et al. (2010) identify a positive link between cleaner proProduction Asset newness: net properties divided by gross properties; proxy for duction and business performance; we also expect a positive relation(PROD) innovators applying the newest technology and yielding a cleaner ship of asset newness and CFP production Sales activities connect the company to the customer; chosen variable of  We follow Luo and Bhattacharya (2006) that a higher customer satisSales and faction and loyalty has a positive influence on CFP customer satisfaction from Thomson Reuters (2013) allows to include the Marketing supporting marketing activity (SAL_MARK) Disposal Natural logarithm of total amount of environmental expenditures  Regarding a wide range of studies with a negative link between the (DISP) impact of environmental expenditures and CFP (Guenther et al., 2011), we also expect a negative relationship Natural logarithm of R&D expenses divided by net sales  R&D represents the innovativeness of a company and is identified as a Research & significant driver for CFP (McWilliams and Siegel, 2001) Development/  Companies with higher efforts for innovation might strive for an Innovation increasing CFP in the future; we expect a negative relationship to (R&D) short-term CFP (Tidd, 2001) Human Resources Employee satisfaction from Thomson Reuters (2013)  Employees directly influence CFP negatively through their wages and (HR) salaries (Hansson, 2004) and positively through their motivation and the resulting productivity (Huselid, 1995)  We expect an overall positive relationship of employee satisfaction to CFP because positive effects dominate the negative effects Organization Monitoring of key performance indicators: variable for monitoring resource  We expect that companies with a better monitoring and penetration of the organizational processes also have a better CFP (Perego and (ORG) efficiency, i.e. management performance, multiplied as an interaction term Hartmann, 2009) with the variable for energy efficiency (EE) Liquidity: net cash flow divided by total assets (beginning-of-the-year)  Liquidity represents the flexibility of a company in terms of slack Accounting: resources (Artiach et al., 2010) and we expect a positive relationship Liquidity and to CFP Leverage Leverage: total debt to total assets  For leverage, a negative relationship is expected (Trumpp and Guenther, (ACC_LIQU) 2017) (ACC_LEV) Procurement and Logistics (PROC_LOG)

Capital intensity: capital expenditures divided by beginning-of-the-year total assets

A. Bergmann et al. / Journal of Cleaner Production 147 (2017) 66e74

71

Table 3 Variable definition for further control variables. Variables

Measurement and description

 Larger firms are more profitable than smaller corporations (Trumpp and Guenther, 2017) as they are based on more resources and are more able to focus on generating positive CFP in terms of accounting-based measures (Modi and Mishra, 2011)  Measurement: natural logarithm of number of employees in the year prior to CFP (Barnett and Salomon, 2012) Liquidity and Leverage (ACC_LIQU)  ACC_LIQU and ACC_LEV serve as proxies for accounting and as further variables to control for effects on CFP (ACC_LEV)  Measurement: see Table 2 Climate Change Performance Index  CCPI by Germanwatch assesses different countries considering their climate change policy, emission levels, and performance (CCPI) regarding efficiency and renewables (Burck et al., 2015)  CCPI also allows to control for origin and country effects (Guenther et al., 2015) Firm size (SIZE)

Table 4 Descriptive statistics. Variable

Mean

SD

Median

Pearson correlation to ROA

ROA EE PROC_LOG PROD SAL_MARK DISP R&D HR ORG ACC_LIQU ACC_LEV SIZE CCPI

0.049 0.034 0.048 0.464 70.865 18.728 0.040 69.992 10.55 0.104 0.242 10.071 56.183

0.064 0.151 0.037 0.144 23.311 2.438 0.045 25.658 29.56 0.068 0.138 1.153 7.923

0.042 0.032 0.038 0.438 77.490 18.063 0.023 81.590 0.43 0.091 0.234 10.110 53.100

0.077** 0.050 0.062* 0.061* 0.185*** 0.089** 0.054* 0.089** 0.618*** 0.276*** 0.082** 0.042

*p < 0.1 **p < 0.05 ***p < 0.01

Table 5 Regression results. ROA Variable

b

SE b

Stand. b

Sig.

VIF

EE PROC_LOG PROD SAL_MARK DISP R&D HR ORG ACC_LIQU ACC_LEV SIZE CCPI Constant Adj. R2 F-Test Observations

0.056*** 0.283*** 0.016 0.000** 0.003*** 0.073 0.001* 0.000*** 0.629*** 0.026 0.002 0.001** 0.005 0.463 41.547*** 565

0.013 0.060 0.015 0.000 0.001 0.047 0.000 0.000 0.033 0.016 0.002 0.000 0.031

0.133*** 0.166*** 0.036 0.073** 0.100*** 0.052 0.054* 0.107*** 0.677*** 0.056 0.028 0.069** e

0.000 0.000 0.306 0.024 0.004 0.124 0.088 0.001 0.000 0.103 0.405 0.030 0.880

1.029 1.296 1.267 1.108 1.262 1.182 1.068 1.131 1.364 1.238 1.158 1.065 e

*p < 0.1 **p < 0.05 ***p < 0.01 Note: The particular analysis for ROA 2012 could only be conducted for 565 corporations due to missing values. The sample size of 650 refers to all conducted analyses together, including robustness checks.

Although a positive and significant relationship from asset newness to CFP was predicted, the variable PROD does not indicate a significant effect on ROA. Among the several possible reasons, the negative impacts of affording new assets might overlap the predicted positive impacts of cleaner production (Zeng et al., 2010). The variable SAL_MARK, measured by customer loyalty, has a

significant positive relationship to ROA with a small b-coefficient (stand. b ¼ 0.073, p < 0.05). We can follow Luo and Bhattacharya (2006) that customer liability and satisfaction causes higher sales and, as a consequence, CFP increases as well. The variable for environmental expenditures DISP has a strong negative influence on ROA (stand. b ¼ 0.100, p < 0.01). It follows that companies with fewer end-of-pipe solutions avoid increasing environmental expenditures and reach improved CFP. This result is in line with recent meta-analyses, which find an overall positive CEP-CFP link (Endrikat et al., 2014). Considering the variable R&D, the analysis can not statistically confirm that R&D significantly influences ROA because the p-value is above the threshold of 0.1. In contrast, HR provides a small, but positive impact on ROA (stand. b ¼ 0.054, p < 0.1). The productivity and efficiency of employees can increase through more satisfied employees, which in turn leads to improved CFP (Huselid, 1995). ORG has a positive and significant influence in the estimated regression model (stand. b ¼ 0.107, p < 0.01). This means that companies with an implemented monitoring of their resources and processes possess a stronger organization and see improved CFP. The results are in line with Perego and Hartmann (2009, p. 405), who state that “information related to benchmarking of a firm's environmental practices against its competitors and industry standards allows performance to be monitored and improved, whether the goal is to achieve industry standards or exceed them.” Considering the corporate value chain activity of accounting, the variable ACC_LIQU has the highest b-coefficient and is strongly significant (stand. b ¼ 0.677, p < 0.01), whereas ACC_LEV does not have a significant effect on ROA. The analysis only partially confirms that our chosen proxies for accounting impact CFP. Summarizing the discussion of the nine analyzed corporate value chain activities, we find hints that most value chain aspects (six out of nine) are actually determining factors for CFP as considered in hypothesis 2. We additionally ran a multiple regression without including the value chain activities as control variables and the results still provide a highly significant, but lower predictive power (Adj. R2 ¼ 0.138, F ¼ 19.727, p ¼ 0.000). To sum up, the consideration of value chain activities leads to a more comprehensive investigation. Regarding CCPI, which was included as a further control variable, the force of expression is not as strong as for other coefficients, but it still represents a significant variable (stand. b ¼ 0.069, p < 0.05). Corporations which are headquartered in countries with improved features for policy of climate change, emission levels, and performance regarding efficiency and renewables also seem to have an improved CFP in comparison to corporations from lower ranked countries. The model also included SIZE and reveals no significant impact, which is in line with a recent investigation by Dangelico and Pontrandolfo (2015).

72

A. Bergmann et al. / Journal of Cleaner Production 147 (2017) 66e74

4.2. Robustness analyses As further analyses can increase the robustness of gained findings (Widener, 2007), we conduct two additional investigations. First, we used the same model as indicated above, but switched the models' year to 2011 and 2013. For the year 2011, the F-statistics indicate that the regression model is statistically significant (Adj. R2 ¼ 0.513, F ¼ 51.024, p ¼ 0.000). The variable EE has a significant positive influence on ROA (stand. b ¼ 0.087, p < 0.01). Significance for a positive relationship between EE and ROA also holds true for the year 2013 (stand. b ¼ 0.101, p < 0.01) with a statistically significant model (Adj. R2 ¼ 0.561, F ¼ 56.730, p ¼ 0.000). The first additional investigation confirms the significant influence from an increasing energy efficiency level to CFP in terms of ROA for the years 2011 and 2013. Within the second additional analysis, we changed the variable for depicting CFP (Artiach et al., 2010) and used return on investment (ROI) and return on equity (ROE). Considering the impact of EE on ROI in 2012, regression results reveal a significant influence (stand. b ¼ 0.123, p < 0.01) and a statistically significant model (Adj. R2 ¼ 0.382, F ¼ 31.075, p ¼ 0.000). The analysis further points out that EE has a significant impact on ROE in 2012 (stand. b ¼ 0.110, p < 0.01). The F-statistics for the investigation of ROE 2012 as a dependent variable also indicate a statistically significant regression model (Adj. R2 ¼ 0.280, F ¼ 19.883, p ¼ 0.000). To sum up, both additional investigations confirm hypothesis 1. Detailed results tables for the additional regression analyses are provided as supplementary material. 4.3. Limitations and directions for future research Even though the estimated regression model and additional robustness analyses provide evidence for a significant link from an increasing level of corporate energy efficiency to CFP, the analysis is not complete without presenting limitations and pathways for future research. First of all, the chosen variables provide some limitations. Being aware that there is no general approach to measuring energy efficiency (Virtanen et al., 2013), our chosen measure includes total direct as well as indirect energy consumption, but does not differentiate the consumed energy into heat or electricity. By including this information, more findings regarding the impact on climate change by reducing GHG emissions could be gained (Morfeldt et al., 2015). Moreover, the energy efficiency variable contains net sales as a proxy for volume (Anderson et al., 2003). Future studies could rely on a more direct measure for value added. Regarding chosen accounting-based measures for CFP, future studies could also rely on market-based measures that also include reputational effects and intangible assets (Endrikat et al., 2014) and, furthermore, are appropriate for analyzing long-term effects (Zeng et al., 2010). Moreover, the chosen variables for corporate value chain activities are not without limitations. Since the analysis includes some nonsignificant relationships between value chain variables and CFP, future studies should rely on other proxies for depicting the specific activities. The non-significance might also lie in the short-term perspective or our chosen operational level. Although the analysis represents the first empirical approach which systematically considers impacts of corporate value chain activities on CFP, a second limitation stems from the fact that the study does not include that activities of the value chain might be also affected by energy efficiency improvements. This assumption should be captured by future research, e.g. by controlling for it in the statistical analysis via interaction terms. Including such interaction terms would simultaneously tackle another weakness of the study: namely that it remains unclear whether the positive effects

of energy efficiency on CFP are rather due to an increase in net sales or due to a decrease in energy consumption. Even though this is not the study's focus, it could be advantageous for future studies to convince firms to invest in energy efficiency improvements by providing insights about the channels through which energy efficiency leads to higher CFP (Morfeldt et al., 2015). The third recommendation for future research refers to the analysis of reverse causality due to resource slack (Modi and Mishra, 2011). For example, it could be that a company which correctly anticipates high CFP in one year is more likely to invest in better energy efficiency in the following year. This phenomenon is better known as the slack resources hypothesis and future research could capture the combination of both directions (Henriques and Catarino, 2016). A fourth reason that could limit the information value represents the fact that we do not differentiate results to different subsectors of the manufacturing industry. In order to reply to this possible limitation, we ran additional regression analysis for every sub-sector and found high significances for the sectors ‘Industrial’ and ‘Technology’. Future research could investigate possible reasons and deduce appropriate decision support-tools considering every sub-sector in detail. Scholars can also investigate which kind of energy efficiency improvement measures provide the most farreaching implications, for instance by conducting in-depth case studies (Kubule et al., 2016). Here, scholars can build on prior publications which analyze suitable energy efficiency improvement measures (Apeaning and Thollander, 2013), either on the technology or management level (Bunse et al., 2011). Even though we acknowledge that our findings cannot be generalized to other industries, they provide a good basis for further studies. It is necessary to replicate the study within other industries and to focus more on country-specific effects, especially when regarding the impacts of and to public policies (Bunse et al., 2011). Future research should also include the effects from an increasing energy efficiency on the level of corporate GHG emissions (Worrell et al., 2001), besides the effects on CFP. Herewith, an even stronger proof to corporate decision-makers can be made. 5. Conclusion The study empirically proves that an increasing level of corporate energy efficiency is directly related to an improved CFP. It applies a multiple regression analysis and considers the manufacturing industry worldwide. Findings indicate a significant positive link and provide the first empirical proof for the investigated relationship. The empirical analysis reveals that corporate decision-makers should consider the implementation of energy efficiency measures, even though they incorporate investment costs. Besides a relationship to CFP, increasing corporate energy efficiency also helps to reduce GHG emissions, which are responsible for climate change and related worldwide impacts. To conclude, increasing energy efficiency supports corporate wants of an improved CFP and further contributes to mitigate climate change. This win-win situation also entails even more energy and non-energy related benefits and is further known as a form of low-hanging fruit. We hope that our findings further encourage research as well as practice to decrease total energy consumption and to combat global warming. Finally, the analysis contributes to recent research as it takes into account the impacts on CFP from activities along the corporate value chain. The regression model specifically includes the nine activities of the corporate value chain as variables to control for effects on CFP. Since gained results provide a higher predictive power, the study calls future research to explicitly consider specific value chain activities when analyzing impacts on financial

A. Bergmann et al. / Journal of Cleaner Production 147 (2017) 66e74

performance. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.jclepro.2017.01.074. References Anderson, M.C., Banker, R.D., Janakiraman, S.N., 2003. Are selling, general, and administrative costs ‘sticky’? J. Acc. Res. 41 (1), 47e63. http://dx.doi.org/10.1111/ 1475-679X.00095. Apeaning, R.W., Thollander, P., 2013. Barriers to and driving forces for industrial energy efficiency improvements in African industries e a case study of Ghana's largest industrial area. J. Clean. Prod. 53, 204e213. http://dx.doi.org/10.1016/ j.jclepro.2013.04.003. Artiach, T., Lee, D., Nelson, D., Walker, J., 2010. The determinants of corporate sustainability performance. Acc. Financ. 50 (1), 31e51. http://dx.doi.org/10.1111/ j.1467-629X.2009.00315.x. Barnett, M.L., Salomon, R.M., 2012. Does it pay to be really good? Addressing the shape of the relationship between social and financial performance. Strateg. Manag. J. 33 (11), 1304e1320. http://dx.doi.org/10.1002/smj.1980. Boyd, G.A., Pang, J.X., 2000. Estimating the linkage between energy efficiency and productivity. Energy Policy 28 (5), 289e296. http://dx.doi.org/10.1016/S03014215(00)00016-1. Brunke, J.-C., Johansson, M., Thollander, P., 2014. Empirical investigation of barriers and drivers to the adoption of energy conservation measures, energy management practices and energy services in the Swedish iron and steel industry. J. Clean. Prod. 84, 509e525. http://dx.doi.org/10.1016/j.jclepro.2014.04.078. € nsleben, P., Brülhart, M., Ernst, F.O., 2011. Integrating Bunse, K., Vodicka, M., Scho energy efficiency performance in production management e gap analysis between industrial needs and scientific literature. J. Clean. Product. 19 (6e7), 667e679. http://dx.doi.org/10.1016/j.jclepro.2010.11.011. Burck, J., Hermwille, L., Bals, C., 2015. The Climate Change Performance Index Background and Methodology. Germanwatch, Bonn, Germany. Cagno, E., Trianni, A., 2014. Evaluating the barriers to specific industrial energy efficiency measures: an exploratory study in small and medium-sized enterprises. J. Clean. Product. 82, 70e83. http://dx.doi.org/10.1016/ j.jclepro.2014.06.057. Chai, K.-H., Baudelaire, C., 2015. Understanding the energy efficiency gap in Singapore: a motivation, opportunity, and ability perspective. J. Clean. Product. 100, 224e234. http://dx.doi.org/10.1016/j.jclepro.2015.03.064. Dangelico, R.M., Pontrandolfo, P., 2015. Being “green and competitive”: the impact of environmental actions and collaborations on firm performance. Bus. Strategy Environ. 24 (6), 413e430. http://dx.doi.org/10.1002/bse.1828. DIN, 2014. Stakeholder Engagement e Guidelines for Decision Making Processes Dealing with Climate Change (DIN SPEC 35810:2014). Beuth-Verlag, Berlin, Germany. Dixon, W.J., Yuen, K.K., 1974. Trimming and winsorization: a review. Stat. Hefte 15 (2e3), 157e170. http://dx.doi.org/10.1007/BF02922904. Endrikat, J., Guenther, E., Hoppe, H., 2014. Making sense of conflicting empirical findings: a meta-analytic review of the relationship between corporate environmental and financial performance. Eur. Manag. J. 32 (5), 735e751. http:// dx.doi.org/10.1016/j.emj.2013.12.004. European Commission, 2012. Directive 2012/27/EU of the European Parliament and of the Council of 25 October 2012 on Energy Efficiency. eur-lex.europa.eu/ LexUriServ/LexUriServ.do?uri¼OJ:L:2012:315:0001:0056:en:PDF (Accessed 02 September 2016). Fearne, A., Garcia Martinez, M., Dent, B., 2012. Dimensions of sustainable value chains: implications for value chain analysis. Supply Chain Manag. Int. J. 17 (6), 575e581. http://dx.doi.org/10.1108/13598541211269193. Friede, G., Busch, T., Bassen, A., 2015. ESG and financial performance: aggregated evidence from more than 2000 empirical studies. J. Sustain. Financ. Investig. 5 (4), 210e233. http://dx.doi.org/10.1080/20430795.2015.1118917. Guenther, E., Guenther, T., Schiemann, F., Weber, G., 2015. Stakeholder relevance for reporting: explanatory factors of carbon disclosure. Bus. Soc. 55 (3), 361e397. http://dx.doi.org/10.1177/0007650315575119. Guenther, E., Hoppe, H., Endrikat, J., 2011. Corporate financial performance and corporate environmental performance: a perfect match? Z. Umweltpolit. Umweltr. 34 (3), 279e296. Guenther, E., Scheibe, L., 2005. The hurdles analysis as an instrument for improving environmental value chain management. Prog. Ind. Ecol. Int. J. 2 (1), 107e131. http://dx.doi.org/10.1504/PIE.2005.006780. Hamann, P.M., Schiemann, F., Bellora, L., Guenther, T.W., 2013. Exploring the dimensions of organizational performance: a construct validity study. Organ. Res. Methods 16 (1), 67e87. http://dx.doi.org/10.1177/1094428112470007. Hambrick, D.C., 1983. High profit strategies in mature capital goods industries: a contingency approach. Acad. Manag. J. 26 (4), 687e707. http://dx.doi.org/ 10.2307/255916. Hansson, B., 2004. Human capital and stock returns: is the value premium an approximation for return on human capital? J. Bus. Financ. Acc. 31 (3e4), 333e358. http://dx.doi.org/10.1111/j.0306-686X.2004.00542.x.

73

Hartman, C.L., Stafford, E.R., 1998. Crafting ‘enviropreneural’ value chain strategies through green alliances. Bus. Horiz. 41 (2), 62e72. http://dx.doi.org/10.1016/ S0007-6813(98)90036-2. He, F., Zhang, Q., Lei, J., Fu, W., Xu, X., 2013. Energy efficiency and productivity change of China's iron and steel industry: accounting for undesirable outputs. Energy Policy 54, 204e213. http://dx.doi.org/10.1016/j.enpol.2012.11.020. Henriques, J., Catarino, J., 2016. Motivating towards energy efficiency in small and medium enterprises. J. Clean. Prod. 139, 42e50. http://dx.doi.org/10.1016/ j.jclepro.2016.08.026. Herring, H., 2006. Energy efficiencyda critical view. Energy Second Bienn. Int. Workshop Adv. Energy Stud. 31 (1), 10e20. http://dx.doi.org/10.1016/ j.energy.2004.04.055. Hirst, E., Brown, M., 1990. Closing the efficiency gap: barriers to the efficient use of energy. Resour. Conserv. Recycl. 3 (4), 267e281. http://dx.doi.org/10.1016/09213449(90)90023-W. Hrovatin, N., Dolsak, N., Zori c, J., 2016. Factors impacting investments in energy efficiency and clean technologies: empirical evidence from Slovenian manufacturing firms. J. Clean. Prod. 127, 475e486. http://dx.doi.org/10.1016/ j.jclepro.2016.04.039. Huselid, M.A., 1995. The impact of human resource management practices on turnover, productivity, and corporate financial performance. Acad. Manag. J. 38 (3), 635e672. http://dx.doi.org/10.2307/256741. IPCC, 2014. Summary for Policymakers, in: IPCC, Climate Change 2014: Mitigation of Climate Change. In: Contribution of Working Group III to the Fifth Assessment Report of the Intergovernmental Panel on Climate. Cambridge University Press, Cambridge, New York, USA. IEA, 2015. Energy and Climate Change. World Energy Outlook Special Report. www. iea.org/publications/freepublications/publication/ WEO2015SpecialReportonEnergyandClimateChange.pdf (Accessed 02 September 2016). Kaplinsky, R., Morris, M., 2001. A Handbook for Value Chain Research. IDRC, Ottawa, Canada. Karcher, P., Jochem, R., 2015. Success factors and organizational approaches for the implementation of energy management systems according to ISO 50001. TQM J. 27 (4), 361e381. http://dx.doi.org/10.1108/TQM-01-2015-0016. King, A.A., Lenox, M.J., 2001. Does it really pay to be green? An empirical study of firm environmental and financial performance. J. Ind. Ecol. 5 (1), 105e116. http://dx.doi.org/10.1162/108819801753358526. Klassen, R.D., Whybark, D.C., 1999. The impact of environmental technologies on manufacturing performance. Acad. Manag. J. 42 (6), 599e615. http://dx.doi.org/ 10.2307/256982. Kounetas, K., Mourtos, I., Tsekouras, K., 2012. Is energy intensity important for the productivity growth of EET adopters? Energy Econ. 34 (4), 930e941. http:// dx.doi.org/10.1016/j.eneco.2011.05.006. Kubule, A., Zogla, L., Ikaunieks, J., Rosa, M., 2016. Highlights on energy efficiency improvements: a case of a small brewery. J. Clean. Prod. 138 (Part 2), 275e286. http://dx.doi.org/10.1016/j.jclepro.2016.02.131. Kung, F., Huang, C., Cheng, C., 2012. Assessing the green value chain to improve environmental performance: evidence from Taiwan's manufacturing industry. Int. J. Dev. Issues 11 (2), 111e128. http://dx.doi.org/10.1108/14468951211241119. Lee, K.-H., 2015. Drivers and barriers to energy efficiency management for sustainable development. Sustain. Dev. 23 (1), 16e25. http://dx.doi.org/10.1002/ sd.1567. Luo, X., Bhattacharya, C., 2006. Corporate social responsibility, customer satisfaction, and market value. J. Mark. 70 (4), 1e18. http://dx.doi.org/10.1509/ jmkg.70.4.1. McWilliams, A., Siegel, D., 2001. Corporate social responsibility: a theory of the firm perspective. Acad. Manag. Rev. 26 (1), 117e127. http://dx.doi.org/10.5465/ AMR.2001.4011987. Meath, C., Linnenluecke, M., Griffiths, A., 2016. Barriers and motivators to the adoption of energy savings measures for small-and medium-sized enterprises (SMEs): the case of the ClimateSmart Business Cluster program. J. Clean. Prod. 112, 3597e3604. http://dx.doi.org/10.1016/j.jclepro.2015.08.085. Modi, S.B., Mishra, S., 2011. What drives financial performanceeresource efficiency or resource slack? J. Oper. Manag. 29 (3), 254e273. http://dx.doi.org/10.1016/ j.jom.2011.01.002. Morfeldt, J., Silveira, S., Hirsch, T., Lindqvist, S., Nordqvist, A., Pettersson, J., Pettersson, M., 2015. Improving energy and climate indicators for the steel industry e the case of Sweden. J. Clean. Prod. 107, 581e592. http://dx.doi.org/ 10.1016/j.jclepro.2015.05.031. Olson, E.L., 2014. Green innovation value chain analysis of PV solar power. J. Clean. Prod. 64, 73e80. http://dx.doi.org/10.1016/j.jclepro.2013.07.050. Peloza, J., 2009. The challenge of measuring financial impacts from investments in corporate social performance. J. Manag. 35 (6), 1518e1541. http://dx.doi.org/ 10.1177/0149206309335188. Perego, P., Hartmann, F., 2009. Aligning performance measurement systems with strategy: the case of environmental strategy. Abacus 45 (4), 397e428. http:// dx.doi.org/10.1111/j.1467-6281.2009.00297.x. Petter, S., Straub, D., Rai, A., 2007. Specifying formative constructs in information systems research. MIS Q. 31 (4), 623e656. Phylipsen, G.J.M., Blok, K., Worrell, E., 1997. International comparisons of energy efficiency - methodologies for the manufacturing industry. Energy Policy 25 (7e9), 715e725. http://dx.doi.org/10.1016/S0301-4215(97)00063-3. Pons, M., Bikfalvi, A., Llach, J., Palcic, I., 2013. Exploring the impact of energy efficiency technologies on manufacturing firm performance. J. Clean. Prod. 52,

74

A. Bergmann et al. / Journal of Cleaner Production 147 (2017) 66e74

134e144. http://dx.doi.org/10.1016/j.jclepro.2013.03.011. Pope Francis, 2015. Encyclical Letter Laudato Si’ of the Holy Father Francis on Care for Our Common Home. Vatican Press, Vatican, Vatican State. Porter, M.E., 1985. The Competitive Advantage: Creating and Sustaining Superior Performance. Free Press, New York, USA. Porter, M.E., Kramer, M., 2011. Creating shared value. Harv. Bus. Rev. 89 (1/2), 62e77. Pye, M., McKane, A., 1999. Enhancing Shareholder Value: Making a More Compelling Energy Efficiency Case to Industry by Quantifying Non-energy Benefits. American Council for an Energy-Efficient Economy, Washington, D.C., USA. Sardianou, E., 2008. Barriers to industrial energy efficiency investments in Greece. J. Clean. Prod. 16 (13), 1416e1423. http://dx.doi.org/10.1016/ j.jclepro.2007.08.002. Schmid, C., 2004. Energieeffizienz in Unternehmen: eine wissensbasierte Analyse von Einflussfaktoren und Instrumenten. vdf Hochschulverlag AG, Zürich, Switzerland. TEEB, 2010. The Economics of Ecosystems and Biodiversity. Mainstreaming the Economics of Nature: a Synthesis of the Approach, Conclusions and Recommendations of TEEB. UNEP, Geneva, Switzerland. Thollander, P., Ottosson, M., 2010. Energy management practices in Swedish energy-intensive industries. J. Clean. Prod. 18 (12), 1125e1133. http://dx.doi.org/ 10.1016/j.jclepro.2010.04.011. Thomson Reuters (Ed.), 2013. Asset4 ESG Data Glossary (Accessed via the database 06 June 2016). Tidd, J., 2001. Innovation management in context: environment, organization and performance. Int. J. Manag. Rev. 3 (3), 169e183. http://dx.doi.org/10.1111/14682370.00062. , S., 2016. Barriers, drivers and decision-making process Trianni, A., Cagno, E., Farne for industrial energy efficiency: a broad study among manufacturing small and medium-sized enterprises. Appl. Energy 162, 1537e1551. http://dx.doi.org/ 10.1016/j.apenergy.2015.02.078. Trianni, A., Cagno, E., De Donatis, A., 2014. A framework to characterize energy efficiency measures. Appl. Energy 118, 207e220. http://dx.doi.org/10.1016/ j.apenergy.2013.12.042. Trumpp, C., Endrikat, J., Zopf, C., Guenther, E., 2015. Definition, conceptualization, and measurement of corporate environmental performance: a critical examination of a multidimensional construct. J. Bus. Ethics 126 (2), 185e204. http://

dx.doi.org/10.1007/s10551-013-1931-8. Trumpp, C., Guenther, T., 2017. Too Little or too much? Exploring U-shaped relationships between corporate environmental performance and corporate financial performance. Bus. Strategy Environ. 26, 49e68. http://dx.doi.org/ 10.1002/bse.1900. Virtanen, T., Tuomaala, M., Pentti, E., 2013. Energy efficiency complexities: a technical and managerial investigation. Manag. Acc. Res. 24 (4), 401e416. http:// dx.doi.org/10.1016/j.mar.2013.06.002. Weinzimmer, L.G., Nystrom, P.C., Freeman, S.J., 1998. Measuring organizational growth: issues, consequences and guidelines. J. Manag. 24 (2), 235e262. http:// dx.doi.org/10.1177/014920639802400205. White House, 2015. Remarks by the President before Signing the Energy Efficiency Improvement Act of 2015. www.whitehouse.gov/the-press-office/2015/04/30/ remarks-president-signing-energy-efficiency-improvement-act-2015 (Accessed 02 September 2016). Widener, S.K., 2007. An empirical analysis of the levers of control framework. Acc. Organ. Soc. 32 (7e8), 757e788. http://dx.doi.org/10.1016/j.aos.2007.01.001. Worrell, E., Bernstein, L., Roy, J., Price, L., Harnisch, J., 2009. Industrial energy efficiency and climate change mitigation. Energy Effic. 2 (2), 109e123. http:// dx.doi.org/10.1007/s12053-008-9032-8. Worrell, E., Laitner, J.A., Ruth, M., Finman, H., 2003. Productivity benefits of industrial energy efficiency measures. Energy 28 (11), 1081e1098. http:// dx.doi.org/10.1016/S0360-5442(03)00091-4. Worrell, E., Price, L., Martin, N., 2001. Energy efficiency and carbon dioxide emissions reduction opportunities in the US iron and steel sector. Energy 26 (5), 513e536. http://dx.doi.org/10.1016/S0360-5442(01)00017-2. Zeng, S.X., Meng, X.H., Yin, H.T., Tam, C.M., Sun, L., 2010. Impact of cleaner production on business performance. J. Clean. Prod. 18 (10e11), 975e983. http:// dx.doi.org/10.1016/j.jclepro.2010.02.019. Zhang, C., Liu, H., Bressers, H.T.A., Buchanan, K.S., 2011. Productivity growth and environmental regulations - accounting for undesirable outputs: analysis of China's thirty provincial regions using the MalmquisteLuenberger index. Ecol. Econ. 70 (12), 2369e2379. http://dx.doi.org/10.1016/j.ecolecon.2011.07.019. Zhang, J., Wang, G., 2008. Energy saving technologies and productive efficiency in the Chinese iron and steel sector. Energy 33 (4), 525e537. http://dx.doi.org/ 10.1016/j.energy.2007.11.002.