The impact of green innovation on competitive strategy: an empirical
assessment of the moderating role of the board of directors
Master thesis MSc BA Strategic Innovation Management Faculty of Economics and Business
Supervisor: prof. dr. J. Surroca Co-assessor: dr. J.D. van der Bij
Lucas Hartman S2006995 Steenhouwerskade 2A20
9718 DA Groningen l.j.hartman@student.rug.nl
2
Abstract
Prior research has reported mixed findings on the impact of green innovation on competitive advantage. This paper investigates the effect of the board of directors, an important corporate governance mechanism, on the relation between green innovation and competitive strategies that can be employed to achieve competitive advantage. Specifically, I explore the influence of control-oriented boards and advice-oriented boards, as monitoring and giving advice are generally regarded as the two primary functions of the board. I address their roles across different types of green innovation and their impact on different strategic outcomes using an integrated perspective of the resource based view (RBV), agency theory, and resource dependence theory. Based on a sample of 262 innovating firms in the United States (U.S.), the results show that green process innovations enable firms to gain competitive advantage through cost leadership strategies, and that green product innovations enable firms to achieve competitive advantage through differentiation and focus strategies. In addition, I find that control-oriented boards have a negative moderating effect on the relationship between green product innovation and differentiation, suggesting that boards emphasizing efficiency may be counterproductive when firms strive to obtain differentiation advantages based on green products. Furthermore, the results show that advice-oriented boards have a positive moderating effect on the relationship between green product innovation and differentiation, suggesting that diversity within the board may enhance competitive advantage through differentiation for firms engaging in green product innovation.
Keywords: innovation; green innovation; green process innovation; green product innovation; board of
3
Table of contents
1. Introduction ... 4
2. Theory and hypotheses ... 6
2.1 Green innovation and competitive strategies... 7
2.2 The control and advisory roles of the board of directors ... 9
2.2.1 The control-oriented board ... 10
2.2.2 The advice-oriented board ... 12
3. Methodology ... 14
3.1 Research setting ... 14
3.2 Data collection and sample ... 15
3.3 Measurements ... 15 3.3.1 Dependent variables ... 16 3.3.2 Independent variables ... 16 3.3.3 Moderating variables ... 16 3.3.4 Control variables ... 20 3.4 Method of analysis ... 22 4. Results ... 23
4.1 Descriptive statistics and correlations ... 23
4.2 Regression results ... 23
4.3 Additional analyses ... 28
5. Discussion ... 29
5.1 Theoretical implications ... 29
5.2 Managerial implications ... 33
5.3 Limitations and future research ... 34
6. Conclusion ... 35
Acknowledgements ... 36
References ... 36
4
1. Introduction
Since the Industrial Revolution, adverse environmental effects caused by industrial and/or infrastructural projects are a growing global problem (Chen, 2008). Policies to attenuate and counteract the environmental damage have been implemented in the last decades, but the global environment’s current state presents an unprecedented challenge: either change the nature of economic activity or risk permanent damage to ecological systems on a global scale (Hart, 1995). Firms are ever more challenged to create new concepts of strategy, and the basis for achieving competitive advantage is increasingly rooted in capabilities such as green product design, waste minimization, and technology collaboration. As a result, the concept of ‘green innovation’ has recently received much attention in literature on environmental management and is increasingly being pursued in order to minimize/eliminate environmental problems (Chang, 2011; Chan, Yee, Dai, & Lim, 2015).
Green innovation is “the improvement of products or processes about energy-saving, pollution-prevention, waste recycling, green product designs, and corporate environmental management” (Chang, 2011, p. 363). Although it is widely accepted that innovation is a source of competitive advantage (Russo & Fouts, 1997; Sharma & Vredenburg, 1998), research on green innovation provides inconclusive findings. Numerous scholars argue that engaging in green innovation enables firms to pursue strategies of cost reduction (e.g., Chen, 2008) and differentiation (e.g., Christmann, 2000), which can reduce total costs and increase total revenues, ultimately resulting in higher financial performance (e.g., King & Lenox, 2002). However, some scholars find that green innovation does not always lead to improved firm performance; green innovation may lead to an increase in costs and customers may be unwilling to purchase the new products, consequently lowering financial performance (e.g., Mathur & Mathur, 2000; Aguilera-Caracuel & Ortiz-de-Mandojana, 2013).
These inconclusive results may suggest that a more appropriate approach could be to examine the conditions under which green innovation leads to competitive advantage. Prior research has proposed several factors to explain why green innovation will lead some firms to reap higher profits than others. For example, Russo and Fouts (1997) find industry growth to moderate the relationship between environmental performance and economic performance in such a way that the returns on environmental performance are higher in high-growth industries. In addition, capabilities for process innovation and implementation (Christmann, 2000), and bundles of manufacturing methods (King & Lenox, 2002) have been identified as complementary resources to positively moderate the relationship between green innovation practices and firm performance.
5
perspective of agency theory and resource dependence theory, the board of directors –as a corporate governance mechanism– may (1) reduce the agency costs associated with the implementation of investment in green process innovation (Howarth, Haddad, & Paton, 2000; Jaffe, Newell, & Stavins, 2004), and (2) obtain access to external resources required for the implementation of investment in green product innovation (Hillman & Dalziel, 2003). Second, the RBV argues that it is inappropriate to conclude that investment in green innovation automatically lead to competitive advantage (Aguilera-Caracuel & Ortiz-de-Mandojana, 2013) and has long noted the critical role of resource complementarity in improving firm performance (Barney, 1991). Rooted in this aspect of the RBV, recent research on the returns of green innovation emphasizes organizational resources complementary to green innovations, including the establishment of intra- and interorganizational structures (Petruzzelli, Maria Dangelico, Rotolo, & Albino, 2011). The board of directors represents such an organizational structure (Denis & McConnell, 2003), which therefore may stand out as a complementary resource for green innovation. Hence, lack of research on the moderating role of the board of directors may be the reason that empirical evidence on the link between green innovation and competitive advantage is inconclusive.
This paper aims to address this literature gap by empirically assessing the role of the board of directors as a complementary resource to green innovation, developing a model that links green innovation, competitive strategies, and corporate governance. Based on the literature gap and the aims of my research, I have formulated two research questions. First, inconclusive empirical evidence on the effect of green innovation on competitive advantage leads to my first research question:
RQ1: What is the relationship between green innovation and competitive advantage?
Second, the goal to empirically examine the role of the board of directors as a complementary resource to green innovation leads to my second research question:
RQ2: Does the board of directors moderate the relationship between green innovation and
competitive advantage?
6
These predictions were tested on a total sample of 262 innovating U.S. firms using data from the ASSET4, Orbis, MSCI, Compustat, and Execucomp databases. The findings show that green process innovation enables firms to achieve competitive advantage through cost leadership strategies and that green product innovation enables firms to achieve competitive advantage through differentiation and focus strategies. The results further suggest that control-oriented boards negatively moderate the relationship between green product innovation and differentiation strategies. Advice-oriented boards were found to positively moderate the relationship between green product innovation and differentiation strategies. Surprisingly, boards that have features of both control-oriented and advice-oriented boards (in this study referred to as intermediate boards) have a positive moderating effect on the relationship between both green process innovation and focus strategies, and between green product innovation and focus strategies.
The theoretical contribution of this research is fourfold. First, I substantiate the increasing body of literature on green innovation that argues green innovation can enhance firm performance by testing the direct impact of green process innovation and green product innovation on gaining competitive advantage through cost leadership, differentiation, and focus strategies. Second, I extend the literature on organizational resources that enhance green innovation’s profitability by adding the board of directors as one of these resources. Third, I link RBV, agency, and resource dependence theories by demonstrating how the board of directors may reduce agency costs and provide the firm with critical external resources; thus, it can serve as a complementary resource to green innovation. This understanding helps to better integrate green innovation and corporate governance literature, as prior research on green innovation examined the benefits of green innovation without directly relating the analyses to a firm’s board of directors. Fourth, I extend the literature on corporate governance, specifically related to the board of directors, by examining the direct impact and interaction effects of different board compositions on competitive advantage. Overall, the paper highlights the importance of carefully designing the board’s composition in order to reduce agency problems and provide a firm with valuable resources, and gives advice to management on how to improve decision making with respect to the implementation of investments in different types of green innovation.
The remainder of this study is structured as follows. The next section reviews relevant theories and literature used to theoretically ground the hypotheses. The succeeding section elaborates on the methodology. Next, the results are presented. This is followed by a discussion on the findings, after which the conclusion finalizes the paper.
2. Theory and hypotheses
7
directors and discusses the possible effects of different types of boards on the relationship between green innovations and competitive strategies, leading to the main hypotheses.
2.1 Green innovation and competitive strategies
Corporate social responsibility (CSR) refers to actions that appear to further some social good, beyond the interests of the firm and that which is required by law (McWilliams & Siegel, 2001). In line with McWilliams and Siegel’s (2001) perspective on the theory of the firm, green innovation can be viewed as a form of CSR investment for which the management of publicly held firms attempts to maximize profits. In this context, a firm can create a certain level of green innovation by greening its production process or by embodying its products with green attributes. The former refers to green process innovation, which is defined as the firm’s capacity to effectively reduce the emission of hazardous substances or waste, reduce the consumption of water, electricity, coal or oil, and reduce the use of raw materials in the production process (Chang, 2011). The latter refers to green product innovation and is defined as the firm’s capacity to exploit market opportunities by improving product design, quality, and reliability with respect to environmental concerns (Chang, 2011). Investment in these two types of green innovation can be assessed by looking at its potential to achieve competitive advantage through different competitive strategies (McWilliams & Siegel, 2001; Christmann, 2000). Porter's (1980) distinction between cost and differentiation advantages provides a useful framework for discussing these strategies and helps to explain how firms can obtain higher profits by means of engaging in green innovation through cost leadership, differentiation, or both.
8
ambiguous). It can lead to significant cost reductions resulting from increased productivity and efficiency. In a similar vein, Shrivastava (1995) states that “environmental technologies are a potential strategic resource because they affect the value chain at multiple points” (p. 190). They are capable of providing firms with unique and inimitable advantages at each stage of the value chain. In the input system, competitive advantage accrues from materials, labour, and energy conservation. In this stage, green process innovations systematically conserve inputs to minimize costs. In the throughput system, an important source of competitive advantage is manufacturing for the environment, which improves production efficiencies and minimizes waste and pollution. Green process innovations in this stage make production lean, thereby enabling cost reduction.
Differentiation captures a firm's efforts to differentiate itself from its competitors using a range of marketing and marketing-related activities (Porter, 1980). It refers to the degree to which a product and its improvements are perceived as unique (Berman et al., 1999). The key to making this strategy successful is an ability to charge above market prices, which is possible because of the customer's perception that the product is special in some way. Green product innovation can enable the production of environmentally friendly products with enhanced quality, design, and reliability (Hart, 1995; McWilliams & Siegel, 2000, 2001). In addition, it enables advertising the environmental benefits of those products (Christmann, 2000). As a result, a firm is better able to enhance its differentiation, which creates the potential to increase product prices and obtain better profit margins (Chen, 2008; Chen, Lai, & Wen, 2006; Petruzzelli et al., 2011). The RBV also helps to explain such differentiation advantages. Green product innovation can support firms in gaining a pro-environment reputation (Russo & Fouts, 1997), which is a widely known valuable, rare, and difficult to replicate (socially complex) intangible resource that allows a firm to differentiate its products from competitors (Rivera, 2002). Moreover, Hart (1995) argues that the strategic capability product stewardship –the ability to design products with the objective of minimizing the environmental impact of product systems– enables firms to create a base from which to build reputation and differentiate its products by establishing the firm as an early mover in new green product domains. Shrivastava’s (1995) conceptualization of environmental technologies as a tool for competitive advantage also applies to green product innovation; in the output system of the value chain, environmental technologies create competitive advantage through better product designs and business portfolios. In this stage, green product innovations allow firms to incorporate environmental considerations into portfolio analysis, resulting in more robust green portfolios, consequently enhancing differentiation.
9
H1a: Green process innovation leads to cost leadership. H1b: Green product innovation leads to differentiation.
2.2 The control and advisory roles of the board of directors
Though less numerous than the studies indicating that “it pays to be green”, some research has pointed out that green innovation does not always lead to improved financial performance. Customers may refuse to accept the green products (Casey, 1992), and green innovation can lead to an increase in the costs of training, quality control and safety (Gelb & Strawser, 2001), promotion (Mathur & Mathur, 2000), and research and risk prevention (López, García, & Rodríguez, 2007). In addition, environmental performance can have a negative effect on short-term financial performance (Sarkis & Cordeiro, 2001). Reconciling insights from agency theory and resource dependence theory helps to further the understanding of the potential missing link between green innovation and competitive advantage.
According to principal-agent theory, managers may act in their own interest, against the benefit of shareholders (e.g., Fama & Jensen, 1983a; Jensen & Meckling, 1976). From this notion, managers may intentionally make inappropriate decisions concerning the implementation of green process innovation’s investments. Howarth et al. (2000) found that managers who incur expenses of investing in green process innovation without reaping the benefits (i.e., no cost savings in manager’s department) may refuse to implement such innovations even though the firm as a whole would profit. In addition, managers may decide to reject or suboptimally implement energy-saving green process innovations, because such projects are often small and difficult to monitor (Jaffe et al., 2004). Furthermore, managers may deliberately imitate other managers’ investment decisions to enhance their professional reputations, with little or no regard to their firms’ green innovation needs (Graham, 1999; Scharfstein & Stein, 1990). These agency problems may occur because interests between shareholders and managers are misaligned and shareholders are at an information disadvantage, making it difficult to evaluate decisions made by managers (Jaffe et al., 2004). As a result, investment in green process innovations may not materialize into the attainment of an effective cost leadership strategy.
10
Integrating agency theory and resource dependence theory into a unitary framework enables the identification of a corporate governance mechanism that can both reduce agency costs and provide the firm with external resources: the board of directors (Hillman & Dalziel, 2003; Oehmichen, Schrapp, & Wolff, 2017; Ashwin, Krishnan, & George, 2016; Desender, Aguilera, Lópezpuertas-Lamy, & Crespi, 2016; Loukil & Yousfi, 2016). The board of directors is defined as a decision-making unit that is empowered to appoint, support, evaluate and dismiss the CEO and other senior managers, to call shareholders' meetings and implement its resolutions, and to determine internal management systems (Huang, 2010). The board of directors can reconcile the problems associated with the implementation of cost leadership and differentiation strategies through its two key functions: (1) monitoring managers to ensure that they do not underinvest or overinvest regarding the implementation of competitive strategies, and (2) providing managers with valuable advice necessary for the successful implementation of competitive strategies (Ashwin et al., 2016; Hillman & Dalziel, 2003; Desender et al., 2016; Zahra & Pearce, 1989; Johnson, Daily, & Ellstrand, 1996). Accordingly, Bordean, Borza, and Maier (2011) state that “the engagement of board members in strategy implementation seems to be beneficial since they are often industry experts and operate at the interface of the firm’s internal and external environment” (p. 989). As such, the board of directors may be actively involved in monitoring the implementation of green technology investments and play a significant role in bringing the needed information and skills in green innovation management. However, the role of the board of directors as a complementary resource to green innovation has received little attention in research on green innovation, as indicated by several reviews of that literature (Schiederig, Tietze, & Herstatt, 2012; Molina-Azorín, Claver-Cortés, López-Gamero, & Tarí, 2009).
2.2.1 The control-oriented board
11
(1993) concurs with this view, stating that boards become less effective monitors when they comprise more than seven or eight members. Additionally, the presence of women on the board may also improve monitoring efforts (Terjesen, Couto, & Francisco, 2016; Loukil & Yousfi, 2016). Furthermore, the ability of boards to monitor managers and contribute to improved efficiency can be improved by the presence of board members that operate in a similar industry as the focal firm (Balsmeier, Buchwald, & Stiebale, 2014). In this line, directors from similar industries may be better able to recognize valuable green process innovations and ensure that they are implemented. In addition, directors from banks often have a strong focus on numbers, which may make them better at monitoring expected efficiency improvements from the implementation of green process innovations (Aoki & Patrick, 1994; Kaplan & Minton, 1994). In both cases, it becomes clear that outside directors, which are those who are not employed by the firm, may play an important role in monitoring managers. In fact, outside, nonaffiliated directors are believed to be particularly effective in monitoring compared to inside directors, because they are more objective (Carpenter, Pollock, & Leary, 2003). Moreover, outside directors have the incentive to safeguard their reputation and avoid legal action (Klein, 2002). Boards characterized by a majority of such directors lack the disincentive to monitor (Hillman & Dalziel, 2003), which may make them better in ensuring managers maximize production efficiency. So, a board of directors that focuses on efficiency may alleviate the agency costs of implementing green process innovations, consequently facilitating the attainment of a cost leadership strategy.
The RBV has long noted the critical role of resource complementarity in improving firm performance (e.g., Barney, 1991). Complementary resources are defined as resources or capabilities that support firms to capture the benefits associated with a strategy, a technology, or an innovation. Control-oriented boards may stand out as a complementary resource for green process innovation by reducing green process innovation’s agency problems associated with the implementation of a cost leadership strategy. Specifically, a control-oriented board may provide the firm with the capability to thoroughly evaluate management decisions, which “can be deployed to serve the goal of creating competitive advantage around environmental innovation” (Orsato, 2006, p. 129). Deploying this capability to optimize implementation of investment in green process innovation can produce valuable, rare, and inimitable synergistic effects, which ultimately helps to achieve competitive advantage through cost leadership. In sum, both prior literature and theoretical perspectives from agency theory and the RBV converge to suggest that control-oriented boards can play a significant role in mitigating the agency problems associated with the implementation of green process innovations, thereby helping the firm achieve competitive advantage through cost leadership strategies. Thus, I hypothesize:
H2a: A control-oriented board positively moderates the relationship between green process
innovation and cost leadership.
12
contributions could be counterproductive when managers require strategic advice (Ashwin et al., 2016). Following this line of reasoning, a control-oriented board may cause friction when a firm requires valuable advice on improving green product innovations in order to obtain differentiation advantages. The lower frequency of meetings by a small group of directors in control-oriented boards may be insufficient to provide managers with high-quality advice (Dalton, Daily, Ellstrand, & Johnson, 1998). Moreover, outside, nonaffiliated directors that mainly focus on efficiency and low prices may lack the incentive to provide managers with valuable advice (Hillman & Dalziel, 2003). Such advice can be crucial if the firm needs to understand how to improve the customer’s perception that the firm’s green products are unique, which lies at the base of a differentiation strategy (Berman et al., 1999). In fact, a board focused on efficiency can by its very nature be detrimental to achieving differentiation advantages; the key to making a differentiation strategy for green product innovation successful is an ability to command premium prices (Chen, 2008; Chen et al., 2006; Petruzzelli et al., 2011), however, a control-oriented board focused on efficiency strives to reduce costs. Moreover, control-control-oriented boards may be incapable of providing strategic advice as they often lack a diversity of knowledge sources (Hillman & Dalziel, 2003). Such knowledge sources are particularly important to obtain knowledge about product improvements and new innovative technologies, which are critical for pursuing a differentiation strategy (Berman et al., 1999). In sum, control-oriented boards are expected to get in the way of implementing green product innovations, which attenuates the positive effect of green product innovation on differentiation. Thus, I hypothesize:
H2b: A control-oriented board negatively moderates the relationship between green product
innovation and differentiation. 2.2.2 The advice-oriented board
13
a variety of knowledge inputs and related connections with other, outside experts to draw on. When shared within the board, diversity of the knowledge at hand benefits the board’s overall knowledge base (Hillman, Cannella, & Paetzold, 2000). This allows the board to provide advice on product improvements and new innovative technologies, which are essential for the differentiation of green products (Chen, 2008; Chen et al., 2006; Petruzzelli et al., 2011). In a similar vein, diversity within the board can stimulate the firm to scan the broader environment for new trends and developments (Geletkanycz, Boyd, & Finkelstein, 2001), which the board can recombine to provide managers with valuable advice on product characteristics, market characteristics, and competitors, all of which are sources of a competitive advantage through differentiation (Orsato, 2006).
Furthermore, from the notion of RBV complementarity (Barney, 1991), an advice-oriented board may stand out as a complementary resource for green product innovation by reducing the firm’s dependency on external resources, which may interfere with the implementation of a differentiation strategy. Specifically, diversity in human capital serves as a source of competitive advantage because it creates value that is both difficult to imitate and rare (Richard, 2000). Novel knowledge and expertise enhance the board’s ability to provide managers with high-quality advice (Peng, 2004), which can produce synergistic effects when combined with the firm’s incumbent knowledge on green product innovation, thereby improving the firm’s ability to differentiate its products. In sum, both prior literature and theoretical perspectives from resource dependence theory and the RBV converge to suggest that advice-oriented boards can play a significant role in mitigating the resource dependence problems associated with the implementation of green product innovations, thereby helping the firm achieve competitive advantage through differentiation strategies. Thus, I hypothesize:
H3a: An advice-oriented board positively moderates the relationship between green product
innovation and differentiation.
14
fraught with issues such as social loafing and increased coordination costs, and hence are not good monitors (Pham, Suchard, & Zein, 2009). In sum, advice-oriented boards are expected to interfere with the implementation of green process innovations, which attenuates the positive effect of green process innovation on cost leadership. Thus, I hypothesize:
H3b: An advice-oriented board negatively moderates the relationship between green process
innovation and cost leadership.
The conceptual model depicted in Figure 1 provides an overview of the constructs and hypotheses tested in this study. All six hypothesized relationships between constructs are tested at the firm-level.
Figure 1. Conceptual model
3. Methodology
This section describes and justifies the methodological choices that were made for the purpose of this study. The first part clarifies the choice for the research setting. The second part describes the data collection process and how this led to the final sample. Next, the variables, their theoretical justification, and their measurements are presented. The final part describes the method used to test the hypotheses.
3.1 Research setting
15
(Raheja, 2005; Desender et al., 2013; Triantis & Daniels, 1995). So, the board of directors is more likely to have a key role in corporate affairs such as control and advice in U.S. firms, which is the reason for this research setting.
3.2 Data collection and sample
The data for this study was collected as follows. First, a list (LA4CTYUS) was obtained from the ASSET4 database through Datastream containing all U.S. firms for which data on environmental, social, and corporate governance was available for the year 2012. Using ISIN codes, this list was then uploaded to Orbis. Here, patent data was used to create a population of innovating U.S. firms from the ASSET4 list. Firms were included in the population of innovating U.S. firms if they had published at least one patent between 01/01/2012 and 31/12/2012. Through this process, a list containing 460 innovating U.S. firms was generated. Next, the ISIN codes of these 460 U.S. firms were converted into CUSIP codes, after which the list was uploaded to the Compustat database. In Compustat, firms that failed to meet the following requirements were removed from the list. First of all, information on cost leadership, differentiation, and focus strategies must be available for the year 2013 (1-year lag). Second, information on firm size, firm growth, firm age, R&D intensity, free cash flow, leverage, market share, and industry type must be available for the year 2012 to control for those variables. To control for firm growth, information on the total sales in 2011 also needed to be available. Applying these criteria left 328 firms. Next, the Ticker symbols of these 328 firms were obtained from Compustat in order to upload the list to the ExecuComp database. Here, firms were removed from the list that had no information available on executive compensation for the year 2012. As a result, 287 firms were left. Next, using Ticker symbols, this list was uploaded to the MSCI database. In this database, firms that did not have information available on the composition of the board of directors and the control variable ownership concentration for the year 2012 were removed. Applying these criteria left a sample of 268 firms. Next, using Bureau van Dijk (BvD) ID codes, the list containing 268 U.S. firms was uploaded to Orbis. Here, firms were removed from the list that had no information available on board education for the year 2012, leaving a final sample of 262 firms. Finally, in order to obtain information about green process and green product innovation, all ‘green patents’ published by these 262 firms in 2012 were identified, resulting in a total of 1,580 green innovations. To determine whether a patent should be classified as a process or product patent, two methods were used. First, the classification of the patent according to the Cooperative Patent Classification (CPC) system can distinguish between process and product patents (i.e., the Y02P class indicates only green process inventions). Second, the abstract of each patent was studied to obtain more detailed information on each patent. Using these two methods, a total of 853 green process innovations and 727 green product innovations were identified.
3.3 Measurements
16
selected and defined based on prior literature. For clarity, an overview of the used data sources and measures per variable is provided in Table A1 of the Appendix.
3.3.1 Dependent variables
Cost leadership. I used the approach of Hambrick (1983) and Berman et al. (1999) to measure the
strategy constructs, because it parsimoniously captures the competitive strategic dimensions put forward by Porter (1980). To assess a firm’s cost leadership position, I used the measure cost efficiency (Hambrick, 1983; Berman et al., 1999). This measure was one minus the ratio of the cost of goods sold to total sales.
Differentiation. The differentiation strategy was captured using the measure selling intensity,
defined as a firm’s willingness to spend on marketing- and selling-related activities in an effort to differentiate itself from its competitors (Hambrick, 1983; Berman et al., 1999). Selling intensity was calculated as the ratio of general, selling, and administrative expenses to total sales.
3.3.2 Independent variables
Green process innovation. To measure a firm’s engagement in green process innovation, patent data is
used, which has become the most common measure of innovation output (Giuri et al., 2007). Using patents to measure green process innovation is convenient, because patents indicate invention counts and provide detailed information on the content of new ideas and technological development (Aguilera-Caracuel & Ortiz-de-Mandojana, 2013). Whether a patent should be considered a green innovation depends on the invention’s environmental effects, meaning that the patent should provide information on environmental benefits or environmental gain must pre-exist within a specific patent class (Arundel & Kemp, 2009). The CPC system identifies the Y02 patent class, which contains patents that indicate technologies or applications for mitigation or adaptation against climate change. So, this study uses the CPC Y02 patent class to identify green innovations. Specifically, to identify green process innovations, each patent’s abstract was read to determine whether the invention resembles the definition of green process innovation provided earlier by Chang (2011).
Green product innovation. The measurement of a firm’s green product innovation follows a
similar approach to that of green process innovation. The only difference is that the CPC Y02 patent’s abstract had to resemble the definition of green product innovation provided earlier by Chang (2011).
3.3.3 Moderating variables
Control-oriented board/advice-oriented board/intermediate board. Based on prior research on board
17
Board independence. This measurement refers to the percent of outside, nonaffiliated directors
on the board (e.g., Dalton et al., 1998). I expect high levels of board independence to represent more control-oriented boards, in line with prior research that has generally linked high levels of board independence to improved monitoring abilities (e.g., Boyd, 1994; Fama, 1980)
Board gender diversity. To measure gender diversity, the ratio of women to total board size was
calculated (Walls et al., 2012). Since prior studies have shown that the presence of women directors on the board can improve the board’s monitoring function (Terjesen et al., 2016; Loukil & Yousfi, 2016), I expect boards with relatively high gender diversity to represent control-oriented boards.
Board size. This measurement indicates the number of directors sitting on the board (Walls et
al., 2012). Large board may be a sign of inefficiency (e.g., Jensen, 1993). Hence, I follow Lipton and Lorsch (1992) and expect relatively small boards with a maximum of nine directors to represent more control-oriented boards.
Board meetings. This measurement represents the annual frequency of board meetings (e.g.,
Dalton et al., 1998). I expect relatively few board meetings to indicate control-oriented boards and a higher number of meetings to indicate advice-oriented boards, in line with existing research that argues intense board activity may indicate monitoring inefficiency (Vafeas, 1999), but can provide managers with higher quality advice (Cramton & Hinds, 2005).
Board mean education level. To measure the board’s mean level of education, first, a coding
scheme was used in which each director was assigned a value 1–8, depending on the level of education attained by the individual director (Pegels et al., 2000). Next, the mean level of education for each board was calculated (Pegels et al., 2000; Wiersema & Bantel, 1992). Boards with high mean levels of education may be better able to actively influence a firm’s decision-making process in terms of receptivity to change and willingness to assume risks (Pegels et al., 2000; Wiersema & Bantel, 1992; Boeker, 1997). Hence, I expect boards with high mean educational levels to represent more advice-oriented boards. Coded values are listed in Table 1.
Table 1. Education levels
Code Highest level of education completed by director
1 Secondary school only
2 Some college
3 Associate’s degree
4 Bachelor’s degree
5 More than one bachelor’s degree
6 Master’s degree or professional certification (MFA, CPA, CPCM, etc.)
7 More than one master’s degree or master’s plus a professional certification
8 Doctorate degree or professional certification (PhD, DBA, JJD, JD, MD, etc.)
Board education level heterogeneity. The assessment of the board’s diversity in terms of level
18
et al., 2000). The larger the standard deviation, the greater the diversity of the board on the education level. Hence, I expect higher levels of education level heterogeneity to represent more advice-oriented boards.
Board educational background heterogeneity. The measurement of board’s educational
background diversity was a two-step process. First, educational background was measured as the percent of board members whose primary education had been in any of the five areas shown in Table 2 (Wiersema & Bantel, 1992; Pegels et al., 2000).
Table 2. Educational classifications Code Type of education
1 Arts
2 Sciences
3 Engineering
4 Business and economics
5 Law
Next, board educational background heterogeneity was calculated by the Herfindahl index (Michel & Hambrick, 1992; Blau, 1977). The formula is:
𝐻 = 1 − ∑ 𝑝𝑖2 5
𝑖=1
In this formula, H is the heterogeneity measure and 𝑝 represents the percentage of board members in each of the five educational background categories. H can take on values from 0 to 1, with high values indicating that a board is heterogeneous. Given that diverse boards have access to more diverse knowledge sources (Van Knippenberg & Schippers, 2007), I expect boards with high diversity in educational background to represent more advice-oriented boards. Table A2 in the Appendix provides an overview of all educational backgrounds in this study and their classifications.
Board educational specialization. To determine each board’s dominant educational discipline,
I included the variable board educational specialization (Wiersema & Bantel, 1992). This was measured by categorizing each board on the basis of its directors’ mode specialization, after which a dummy variable was developed to capture boards dominated by directors with educational backgrounds in science and engineering (Surroca, Prior, & Tribó Giné, 2014; Wiersema & Bantel, 1992). Wiersema and Bantel (1992) argue that those disciplines are more concerned with progress, innovation, and change in competitive strategy than backgrounds in arts, business, and law. Thus, I expect boards dominated by scientists and engineers to indicate advice-oriented boards.
19
boards, advice-oriented boards, or intermediate boards in the case that boards display characteristics of both control-oriented and advice-oriented boards. The results of the application of this theory-based methodology are shown in Table 3. In order to test if the identified board types truly exist, I examined the between-group variability for each variable (ANOVA) and for all the variables simultaneously (MANOVA) (Fiegenbaum & Thomas, 1990). The ANOVA statistics are significant for six out of eight variables, suggesting that the dimensions that firms emphasize differ across board types. The MANOVA results complement these findings, demonstrating the ability of this theory-based board classification to separate the observations among clusters.
Table 3. Final cluster means of eight board-related variables
Clusters ANOVA 1 2 3 Total F test Board independence 0.641 (0.192) 0.614 (0.203) 0.555 (0.207) 0.614 (0.201) 3.8**
Board gender diversity 0.140
(0.103) 0.138 (0.096) 0.116 (0.111) 0.134 (0.103) 1.25 Board size 7.450 (1.246) 11.566 (1.561) 10.305 (2.144) 9.310 (2.404) 173.29*** Board meetings 6.488 (1.522) 6.658 (1.621) 12.424 (3.318) 7.874 (3.219) 181.24***
Board mean education level 7.062
(0.536) 7.218 (0.429) 7.206 (0.477) 7.140 (0.498) 3.05**
Board education level heterogeneity 0.904 (0.439) 0.944 (0.337) 0.929 (0.402) 0.921 (0.402) 0.25
Board educational background heterogeneity 0.233 (0.122) 0.358 (0.172) 0.376 (0.153) 0.303 (0.158) 29.09*** Board educational specialization 0.055 (0.229) 0.605 (0.492) 0.610 (0.492) 0.340 (0.475) 66.61*** Number of boards 127 76 59 262 Control-oriented board Intermediate board Advice-oriented board
Standard deviations in parentheses; * p < 0.10, ** p < 0.05, *** p < 0.01; MANOVA tests between-group
variability for all variables simultaneously. Its F-statistic (38.97) was significant at the 1 percent level.
20
Cluster 2 comprises boards that are the largest in size and have the highest mean education level, but they are also characterized by a relatively low number of annual board meetings. Given that most variables have mean values that fall in-between the mean values of the other clusters, boards in group 2 can be classified as intermediate boards.
Finally, the boards in cluster 3 emphasize frequent board meetings. Additionally, they have the highest levels of board educational specialization, the highest levels of educational background heterogeneity, and the lowest levels of board independence. Furthermore, the large board size and high mean education level suggest boards in this group can be classified as advice-oriented boards. I developed three dummy variables to capture the different board types.
3.3.4 Control variables
Firm size. Existing research has acknowledged that large firms typically possess more bargaining power
than small firms, which may help to pursue competitive strategies (e.g., Porter, 1980). Even so, Jensen and Meckling (1976) indicate that firm size may also lead to an increase in agency problems, which can increase costs and subsequently lower efficiency. Moreover, firm size potentially influences the benefit received from engaging in eco-innovation practices (Dixon-Fowler et al., 2013). For these reasons, I control for firm size, measured as the natural logarithm of the total assets (Desender et al., 2016).
Firm growth. To achieve and retain competitive advantage, it is crucial for firms to grow
(Schuler & Jackson, 1987). However, firm growth may also be associated with less intense competition and high firm growth may thus indicate little need for competitive strategies (Eisenhardt & Schoonhoven, 1990). So, I control for firm growth, measured as the one-year growth rate of sales (e.g., Brush, Bromiley, & Hendrickx, 2000).
Firm age. I also control for firm age. Firms may accumulate resources over years, which can
support firms in employing competitive strategies to gain competitive advantage (Barney, 1991). Moreover, older firms may be less likely to appoint nonaffiliated, outside directors as older firms can be more entrenched (Peng, 2004). Firm age is measured as the number of years since the firm’s inception (Manikandan & Ramachandran, 2015).
R&D intensity. Existing research argues that research and development (R&D) intensity can
support firms in employing competitive strategies (Baysinger & Hoskisson, 1989; Kotabe, Srinivasan, & Aulakh, 2002; Ito & Pucik, 1993). Moreover, R&D investments may lead to environmentally-friendly process and product innovations (McWilliams & Siegel, 2000, 2001). Hence, I control for R&D intensity, measured as a firm’s R&D expenditures divided by sales (e.g., McWilliams & Siegel, 2000).
Free cash flow. In addition, I include free cash flow as a control variable, which gauges the
21
with higher agency costs and may benefit more from control-oriented boards. I measure free cash flow as cash flow from operations less capital expenditures (Bushee, 1998).
Leverage. Given that a firm’s competitive strategy may be affected by its debt financing (Barton
& Gordon, 1987; Taylor & Lowe, 1995), I control for leverage, defined as the debt to assets ratio (e.g., Erickson, 1998; Minton & Schrand, 1999). Debt financing may ‘force’ managers to keep their promises to pay out future cash flows (Barton & Gordon, 1987). So, reduced leverage can increase cash flows available for managers to spend, which may support competitive strategies (Ho et al., 2011), or increase agency costs (Jensen, 1986).
Market share. Porter (1980) states that the essence of competitive strategy is coping with
competition. Market share may serve as a proxy for a firm’s bargaining power, which enhances the firm’s position vis-à-vis competitors and thus may be positively associated with competitive advantage (Szymanski, Bharadwaj, & Varadarajan, 1993; Porter, 1980; Barney, 1995). Accordingly, I control for market share, defined as the ratio of a firm’s sales to total industry sales (e.g., Szymanski et al., 1993).
Industry. When competitive strategy is the dependent variable of concern, the operating
environment plays a key role (Porter, 1980; Mauri & Michaels, 1998). Hence, I control for types of industry using four-digit Standard Industry Classification (SIC) codes, as Porter (1980) argues SIC classifications with “two-digit industries overly broad for most purposes, five-digit industries often too narrow, and four-digit industries usually about right” (p. 370).
Variable compensation. Prior studies have shown that incentive pay can support firms in
achieving competitive advantage (Collins & Clark, 2003; Pfeffer, 1995). Tying executives’ rewards to firm performance may mitigate agency problems (Baysinger & Hoskisson, 1990; Bhagat, Brickley, & Lease, 1985) and motivate executives to search for innovative products (Collins & Clark, 2003; Schuler & MacMillan, 1984). Moreover, as incentive pay helps to align executives’ and shareholders’ interests, the monitoring potential of the board of directors may be reduced (Rediker & Seth, 1995). So, I include variable compensation, that is, the proportion of variable executive pay to the total executive compensation (Mak & Li, 2001; Yermack, 1996).
Ownership concentration. The final control variable, ownership concentration, refers to the
22
3.4 Method of analysis
The baseline hypotheses and hypothesized moderating effects of different board types were tested by examining the four regression equations below for the equality of the regression coefficients (Sharma, Durand, & Gur-Arie, 1981). All four equations were examined for each dependent variable (Baron & Kenny, 1986). I performed robust regression to provide resistant results in the presence of outliers (Rousseeuw & Leroy, 2005). As recommended by Aiken and West (1991), I mean-centered the independent variables to test the interaction effects.
[1] 𝑦𝑖,𝑡+1= α + 𝛽1𝑐1,t+ 𝛽2𝑐2,t+ 𝛽3𝑐3,t+ 𝛽4𝑐4,t+ 𝛽5𝑐5,t+ 𝛽6𝑐6,t+ 𝛽7𝑐7,t+ 𝛽8𝑐8,t+ 𝛽9𝑐9,t+ 𝛽10𝑐10,t+ 𝜀 [2] 𝑦𝑖,𝑡+1= α + 𝛽1𝑐1,t+ 𝛽2𝑐2,t+ 𝛽3𝑐3,t+ 𝛽4𝑐4,t+ 𝛽5𝑐5,t+ 𝛽6𝑐6,t+ 𝛽7𝑐7,t+ 𝛽8𝑐8,t+ 𝛽9𝑐9,t+ 𝛽10𝑐10,t+ 𝛽11𝑥1,t+ 𝛽12𝑥2,t+ 𝜀 [3] 𝑦𝑖,𝑡+1= α + 𝛽1𝑐1,t+ 𝛽2𝑐2,t+ 𝛽3𝑐3,t+ 𝛽4𝑐4,t+ 𝛽5𝑐5,t+ 𝛽6𝑐6,t+ 𝛽7𝑐7,t+ 𝛽8𝑐8,t+ 𝛽9𝑐9,t+ 𝛽10𝑐10,t+ 𝛽11𝑥1,t+ 𝛽12𝑥2,t+ 𝛽13𝑧𝑡+ 𝜀 [4] 𝑦𝑖,𝑡+1= α + 𝛽1𝑐1,t+ 𝛽2𝑐2,t+ 𝛽3𝑐3,t+ 𝛽4𝑐4,t+ 𝛽5𝑐5,t+ 𝛽6𝑐6,t+ 𝛽7𝑐7,t+ 𝛽8𝑐8,t+ 𝛽9𝑐9,t+ 𝛽10𝑐10,t+ 𝛽11𝑥1,t+ 𝛽12𝑥2,t+ 𝛽13𝑧𝑡+ 𝛽14𝑥𝑧𝑡+ 𝜀
In these equations, 𝑦𝑖,𝑡+1 represents the value of the dependent variable of the succeeding year
(2013). The dependent variable shows the result of the succeeding year (t + 1) to reduce endogeneity issues. In the rest of the equations, α is the intercept, 𝛽𝑖 denotes the regression coefficients, 𝑐𝑖,𝑡 represents
the ten control variables, 𝑥𝑖,𝑡 indicates the two independent variables, and 𝑧𝑖,𝑡 is the moderating variable
referring to a dummy variable to indicate a specific board type, so that 𝑥𝑧𝑡 is used to show the interaction
effect of the independent variable with a certain board type. Following Sharma et al. (1981), I classify 𝑧 as a pure moderator only if equation 2 is not significantly different from equation 3, but both are significantly different from equation 4. I classify 𝑧 as a quasi-moderator if equations 2, 3, and 4 are all significantly different from each other. In the case that equations 3 and 4 are not significantly different, 𝑧 is not a moderator variable but simply an independent predictor variable. I performed likelihood ratio tests (LR test) to determine whether the equations significantly differed from each other.
23
4. Results
This section presents the results found in this study. First, the descriptive statistics and correlations are described. Next, the results from the regression analyses used to test the hypotheses are presented. The third part shows the results of the additional analyses that were performed to identify a potential competitive strategy that was not hypothesized, and to check the robustness of the theory-based clustering method.
4.1 Descriptive statistics and correlations
The descriptive statistics are presented in Table 4, showing the mean, standard deviation, minimum value, maximum value, and correlations of the variables in this study. As can be seen from this table, no high values of correlations (𝑟 < 0.8) are present between the variables, suggesting there is no reason to suspect the variables of exhibiting multicollinearity (Bryman & Cramer, 1997; Abdelsalam, El-Masry, & Elsegini, 2008), which complements the results of the VIF examination. Table 4 also indicates several significant correlations among the main variables. Green process innovation is significantly positively correlated with cost efficiency (𝑟 = 0.206, p < 0.01), indicating that high (low) values of green process innovation are associated with high (low) values of cost efficiency. In addition, green product innovation is significantly positively correlated with both cost efficiency (𝑟 = 0.176, p < 0.05) and selling intensity (𝑟 = 0.283, p < 0.01). Moreover, green product innovation shows a significant positive correlation with green process innovation (𝑟 = 0.535, p < 0.01), indicating that firms that score high (low) on one type of green innovation also tend to score high (low) on the other type of green innovation. Regarding the different board types, significant positive correlations are present between advice-oriented boards and selling intensity (𝑟 = 0.190, p < 0.05), and advice-oriented boards and green product innovation (𝑟 = 0.162, p < 0.05).
4.2 Regression results
To test the hypotheses, the regression equations previously mentioned were examined. Table 5 shows the results pertaining to the cost leadership strategy and Table 6 shows the results for the differentiation strategy. In these tables, model 1 only includes the control variables. In model 2, the independent variables are included to test the baseline hypotheses. Finally, models 3-8 include the moderators and show the results of the moderated regression analyses to test the main hypotheses.
24
Table 4. Descriptive statistics and correlations
Variables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Cost efficiency 1.000
2. Selling intensity 0.465*** 1.000
3. Green process innovation 0.206*** 0.108 1.000
4. Green product innovation 0.176** 0.283*** 0.535*** 1.000
5. Firm size (log) 0.012 -0.089 0.264*** 0.210*** 1.000
6. Firm growth -0.049 -0.122* 0.150* 0.138* 0.244*** 1.000
7. Firm age -0.308*** -0.184** 0.046 -0.042 0.364*** -0.079 1.000
8. R&D intensity 0.608*** 0.495*** 0.093 0.210*** -0.071 -0.082 -0.385*** 1.000
9. Free cash flow 0.166** -0.002 0.366*** 0.287*** 0.561*** 0.507*** 0.127* 0.011 1.000
25
Table 5. Regression results on cost efficiency
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Green process innovation 0.00347*** 0.00331*** 0.00354** 0.00360** 0.00407*** 0.00339*** 0.00224*
(0.00124) (0.00121) (0.00148) (0.00144) (0.00132) (0.00126) (0.00135)
Green product innovation 0.00094 0.00122 0.00123 0.00075 0.00064 0.00129 0.00128
(0.00169) (0.00184) (0.00183) (0.00166) (0.00169) (0.00179) (0.00180)
Firm size (log) -0.02644** -0.02622** -0.02457** -0.02441** -0.02343** -0.02290** -0.02753** -0.02673**
(0.01189) (0.01146) (0.01143) (0.01144) (0.01148) (0.01151) (0.01133) (0.01134) Firm growth -0.00001*** -0.00001*** -0.00001*** -0.00001*** -0.00001*** -0.00001*** -0.00001*** -0.00001*** (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) Firm age -0.00059* -0.00058* -0.00060** -0.00060** -0.00061** -0.00063** -0.00057* -0.00058* (0.00030) (0.00030) (0.00029) (0.00029) (0.00030) (0.00030) (0.00030) (0.00030) R&D intensity 1.41702*** 1.31363*** 1.29804*** 1.30836*** 1.28220*** 1.27321*** 1.33739*** 1.37271*** (0.21513) (0.21695) (0.21259) (0.21684) (0.22268) (0.21809) (0.20985) (0.21197)
Free cash flow 0.00002*** 0.00001*** 0.00001*** 0.00001*** 0.00001*** 0.00001*** 0.00001*** 0.00001***
(0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) Leverage -0.00998 -0.02280 -0.00047 -0.00198 -0.01569 -0.01961 -0.01570 -0.02542 (0.09325) (0.09014) (0.09020) (0.09077) (0.09195) (0.09153) (0.08932) (0.09009) Market share 0.00278 0.00407** 0.00319** 0.00312* 0.00349** 0.00349** 0.00398** 0.00376** (0.00169) (0.00160) (0.00161) (0.00161) (0.00160) (0.00161) (0.00160) (0.00161) Industry 0.00002** 0.00002** 0.00002** 0.00002** 0.00002** 0.00002* 0.00002** 0.00002* (0.00001) (0.00001) (0.00001) (0.00001) (0.00001) (0.00001) (0.00001) (0.00001) Variable compensation 0.31030*** 0.29922*** 0.29384** 0.29574** 0.29938** 0.30155** 0.29676*** 0.30602*** (0.11793) (0.11349) (0.11557) (0.11612) (0.11644) (0.11620) (0.11284) (0.11356) Ownership concentration -0.12311 -0.11901 -0.11798 -0.11588 -0.10556 -0.10965 -0.12894 -0.12648 (0.09297) (0.09087) (0.09041) (0.09030) (0.08915) (0.08913) (0.08994) (0.08982) Control-oriented board 0.05689** 0.05678** (0.02255) (0.02262)
Green process innovation x control-oriented board (H2a) 0.00066
(0.00158)
Intermediate board -0.04175 -0.04503*
(0.02645) (0.02591)
Green process innovation x intermediate board -0.00257**
(0.00109)
Advice-oriented board -0.03617 -0.03808
(0.02812) (0.02796)
Green process innovation x advice-oriented board (H3b) 0.00247
(0.00192) Constant 0.25621** 0.27248*** 0.22734** 0.22360** 0.26269*** 0.26046*** 0.28841*** 0.28373*** (0.10505) (0.09620) (0.10026) (0.09931) (0.09591) (0.09522) (0.09539) (0.09366) R2 0.4708 0.4966 0.5125 0.5127 0.5031 0.5073 0.5003 0.5033 Adjusted R2 0.4498 0.4724 0.4869 0.4851 0.4749 0.4793 0.4741 0.4751 F-statistic 20.07*** 24.03*** 25.75*** 25.01*** 24.12*** 32.58*** 22.54*** 22.00***
26
Table 6. Regression results on selling intensity
Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7 Model 8
Green process innovation 0.00072 0.00076 0.00082 0.00072 0.00152 0.00079 0.00094
(0.00117) (0.00113) (0.00117) (0.00114) (0.00119) (0.00112) (0.00099)
Green product innovation 0.00329* 0.00289* 0.00395* 0.00358* 0.00325* 0.00296* 0.00253*
(0.00187) (0.00173) (0.00218) (0.00172) (0.00192) (0.00179) (0.00131)
Firm size (log) -0.02511** -0.02495** -0.02590** -0.02491** -0.02463** -0.02458** -0.02373* -0.02215*
(0.01269) (0.01227) (0.01212) (0.01228) (0.01191) (0.01187) (0.01207) (0.01226) Firm growth -0.00001*** -0.00001* -0.00001* -0.00001* -0.00001* -0.00001 -0.00001* -0.00001* (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) Firm age 0.00007 0.00010 0.00012 0.00014 0.00011 0.00008 0.00009 0.00012 (0.00028) (0.00028) (0.00028) (0.00028) (0.00026) (0.00028) (0.00028) (0.00028) R&D intensity 1.11020*** 1.04221*** 1.05109*** 1.05647*** 1.05112*** 1.05144*** 1.01997*** 1.03586*** (0.11577) (0.12093) (0.11961) (0.11908) (0.12321) (0.12380) (0.12368) (0.12409)
Free cash flow 0.00001** 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000 0.00000
(0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) (0.00000) Leverage 0.03234 0.02638 0.01367 0.01097 0.02402 0.01479 0.01973 0.01087 (0.07938) (0.07671) (0.07579) (0.07633) (0.07803) (0.07829) (0.07454) (0.07583) Market share 0.00421*** 0.00514*** 0.00563*** 0.00548*** 0.00425*** 0.00524*** 0.00522*** 0.00502*** (0.00152) (0.00155) (0.00148) (0.00144) (0.00153) (0.00159) (0.00152) (0.00148) Industry 0.00001 0.00001 0.00000 0.00001 0.00001 0.00000 0.00001 0.00001 (0.00001) (0.00001) (0.00001) (0.00001) (0.00001) (0.00001) (0.00001) (0.00001) Variable compensation 0.14183 0.13479 0.13785 0.13701 0.13421 0.14098 0.13709 0.13809 (0.09777) (0.09420) (0.09083) (0.09137) (0.09357) (0.09324) (0.09297) (0.09341) Ownership concentration -0.06234 -0.06854 -0.06912 -0.06045 -0.07281 -0.07572 -0.05924 -0.04914 (0.07805) (0.07584) (0.07567) (0.07661) (0.07443) (0.07432) (0.07536) (0.07649) Control-oriented board -0.03229** (0.01934) -0.03251** (0.01935)
Green product innovation x control-oriented board (H2b) -0.00348*
(0.00262)
Intermediate board -0.04212*
(0.02426)
-0.01376 (0.02129)
Green product innovation x intermediate board -0.00521**
Advice-oriented board (0.00301) 0.08192*** (0.02630) 0.08379*** (0.02583)
Green product innovation x advice-oriented board (H3a) 0.00546*
(0.00285) Constant 0.26614*** 0.27685*** 0.30070*** 0.32067*** 0.28482*** 0.28847*** 0.24912*** 0.24197*** (0.09005) (0.08591) (0.08337) (0.08559) (0.08619) (0.08591) (0.08299) (0.08416) R2 0.3044 0.3325 0.3619 0.3749 0.3348 0.3362 0.3638 0.3691 Adjusted R2 0.2767 0.3003 0.3285 0.3394 0.2995 0.3011 0.3305 0.3333 F-statistic 16.74*** 15.19*** 17.15*** 18.54*** 15.38*** 15.23*** 15.53*** 14.32***
27
Table 5 also shows the results of hypothesis 2a and hypothesis 3b. In this table, model 3 includes the dummy variable of the control-oriented board, which shows a significant direct positive effect on cost efficiency (β = 0.05689, p < 0.05). In model 4, the interaction effect between the control-oriented board and green process innovation on cost efficiency is included. The results suggest that control-oriented boards have a significant direct positive effect on cost efficiency (β = 0.05678, p < 0.05), but do not significantly moderate the relationship between green process innovation and cost efficiency (p > 0.10), thus no support is found for the hypothesized positive moderating effect in hypothesis 2a. Based on these results, the control-oriented board can be classified as an independent variable for cost efficiency, as model 3 significantly differs from model 2 (LRχ² = 8.38, df = 1, p < 0.01) and model 4 also significantly differs from model 2 (LRχ² = 8.53, df = 2, p < 0.05). Model 8 shows the results of hypothesis 3b. No significant moderating effect is found for advice-oriented boards on the relationship between green process innovation and cost efficiency (p > 0.10), providing no support for the hypothesized negative moderating effect predicted in hypothesis 3b. In addition, model 6 shows both a direct negative effect of intermediate boards on cost efficiency (β = -0.04503, p < 0.10) and a negative interaction effect between intermediate boards and green process innovation on cost efficiency (β = -0.00257, p < 0.05). Although both models 5 (LRχ² = 3.26, df = 1, p < 0.10) and 6 (LRχ² = 5.60, df = 2, p < 0.10) significantly differ from model 2, the intermediate board can be classified as an independent predictor variable for cost efficiency, as model 6 is not significantly different from model 5.
28
the intermediate board can be classified as an independent predictor variable for selling intensity. Concerning the control variables, firm size shows a significant negative effect on both cost efficiency and selling intensity across all models. Firm growth, defined as the one-year growth in sales, also shows a significant negative effect on cost efficiency and selling intensity across all models, except for model 6 in Table 6. Firm age is found to have a significant negative effect across all models only for cost efficiency. R&D intensity shows a significant positive effect on cost efficiency and selling intensity across all models. Free cash flow, defined as cash flow from operations less capital expenditures, shows a significant positive effect across all models for cost efficiency. However, a significant positive effect on selling intensity is found only in model 1. In addition, no significant effects are found for leverage. Market share shows a significant positive effect in models 2-8 for cost efficiency and across all models for selling intensity. Industry type, based on the four-digit SIC codes, and variable compensation both show a significant positive effect on cost efficiency across all models, however no effect is found on selling intensity. The final control variable, ownership concentration, does not show a significant effect on any of the dependent variables.
4.3 Additional analyses
I analysed green innovation’s effect on a third competitive strategy, and the robustness of the theory-based clustering method as two additional topics. According to Porter (1980), firms can gain competitive advantage through cost leadership, differentiation, and focus strategies. The hypotheses in this study were based on strategies of cost leadership and differentiation. So, I performed an additional moderated regression analysis to examine whether engaging in green process innovation and green product innovation enables firms to achieve competitive advantage through focus strategies, and whether the board has a moderating effect on these relationships. To measure Porter’s (1980) focus strategy, the sales to target audience was measured (Hambrick, 1983; Miller & Friesen, 1986). This measure captures the ratio of sales generated by the firm’s largest customer (as reported by the firm) to total sales.
The results of the analysis show a significant direct positive effect of green product innovation on sales to target audience (β = 0.00418, p < 0.01). In addition, the results show that an intermediate board has significant positive interaction effects with both green process innovation (β = 0.00434, p < 0.01) and green product innovation (β = 0.01155, p < 0.01) on sales to target audience. After examining whether the corresponding models significantly differ from each other, the intermediate board can be classified as a pure moderator for sales to target audience. Concerning the control variables, only R&D intensity showed a significant positive effect on sales to target audience across all models.
29
was used to assign boards to one of these five clusters. The five board types identified were labelled as follows: control-oriented boards, less control-oriented boards, intermediate boards, advice-oriented boards, and active advice-oriented boards. In this classification, the less control-oriented board has mean values that mostly fall in-between the intermediate board and the control-oriented board. The distinguishing feature of active advice-oriented boards is their high number of meetings, hence the label ‘active’. I performed ANOVA and MANOVA analyses in order to test if the identified board types really exist (Fiegenbaum & Thomas, 1990). Similar to the process that identified three clusters, ANOVA statistics were significant for the same six out of eight variables, and the F-statistic of the MANOVA test (18.34) was significant at the 1 percent level.
Next, moderated regression analyses were performed to test the interaction effects with the five board types. The results supported the findings of the theory-based clustering procedure. Hypothesis 2a, which argues that control-oriented boards positively moderate the relationship between green process innovation and cost leadership, was not supported (p > 0.10). In addition, hypothesis 2b, which predicted a negative moderating effect of control-oriented boards on the relationship between green product innovation and differentiation, was supported (β = -0.00392, p < 0.10). Hypothesis 3a, which argued a positive moderating effect of advice-oriented boards on the relationship between green product innovation and differentiation, was also supported (β = 0.01060, p < 0.01). Finally, hypothesis 3b, which predicted a negative moderating effect of advice-oriented boards on the relationship between green process innovation and cost leadership, was not supported (p > 0.10). No significant effects were found for less control-oriented boards and active advice-oriented boards.
5. Discussion
The results in this study suggest that different types of boards have different effects on the relationship between green innovation and competitive strategies. Apparently, control-oriented boards do not positively moderate the green process innovation-cost leadership relationship, but may harm differentiation advantages resulting from green product innovations. Furthermore, advice-oriented boards seem to have no negative moderating effect on the relationship between green process innovation and cost leadership strategies, but they may strengthen the positive effects of green product innovation on differentiation strategies. This section first discusses the implications for both theory and practice of these main findings and additional results. The section concludes with a discussion on the limitations of this study and offers possible avenues for future research.
5.1 Theoretical implications
30
López et al., 2007), and customers may refuse to accept green products (Casey, 1992). However, in line with recent literature (Dixon-Fowler et al., 2013; Chen, 2008; Chen et al., 2006; Petruzzelli et al., 2011), this study found that (1) green process innovations can enable firms to achieve competitive advantage through cost leadership, and (2) green product innovations can enable firms to achieve competitive advantage through differentiation and focus strategies. An explanation for the first finding is that green process innovations can reduce costs through pollution reduction (Hart, 1995), improvements in the recycling of by-products of processes (Ashford, 1993), an increase in material savings, and reduction of energy consumption (King & Lenox, 2002; Klassen & Whybark, 1999; Chen, 2008). According to the RBV, green process innovations allow firms to develop strategic capabilities to effectively reduce costs, which are key for developing competitive advantage. (Hart, 1995; Shrivastava, 1995). The finding that green product innovation leads to differentiation advantages can be explained by the production of environmentally friendly products with enhanced quality, design, and reliability (Hart, 1995; McWilliams & Siegel, 2000, 2001), and the possibility to advertise the environmental benefits of those products (Christmann, 2000). Firms that engage in green product innovation can charge premium prices, which improves their competitive position. Moreover, following the RBV, green product innovations improve a firm’s pro-environment reputation (Russo & Fouts, 1997), which can be a strategic resource that enables firms to better differentiate their products (Rivera, 2002). An unexpected finding was that green product innovations can help firms gain competitive advantage by attaining focus strategies. A possible explanation is offered by Porter and van der Linde (1995), who suggest that green products generally appeal to environmentally informed customer that can easily be identified and reached by firms. Firms that succeed in identifying those customers are able to command considerable price premiums, which suggests a differentiation-focus strategy may be particularly successful. Moreover, from the notion of the RBV, firms that engage in green product innovation may develop strategic capabilities in targeting environmentally-conscious customers and become increasingly capable at fulfilling those customer’s needs (Laroche, Bergeron, & Barbaro-Forleo, 2001).
