During the last three decades, entrepreneurial orientation (EO) has become one of the most influential topics in the entrepreneurship literature, with more than one-hundred studies exploring the concept (Rauch, Wiklund, Lumpkin, & Frese, Reference Rauch, Wiklund, Lumpkin and Frese2009). EO captures the entrepreneurial aspects of firm strategic posture (Covin & Lumpkin, Reference Covin and Lumpkin2011), and may be characterized as an organization’s ‘strategy-making practices, management philosophies, and firm-level behaviors that are entrepreneurial in nature’ (Anderson, Covin, & Slevin, Reference Anderson, Covin and Slevin2009: 220). The defining components of EO include the manifestation of innovative, risk taking, and proactive firm processes and behaviors (Miller, Reference Miller1983; Covin & Slevin, Reference Covin and Slevin1989; Tang, Kreiser, Marino, Dickson, & Weaver, Reference Tang, Kreiser, Marino, Dickson and Weaver2009). Research exploring EO has focused intensely upon its relationship with firm financial performance, with most studies evidencing a positive relationship (Rauch et al., Reference Rauch, Wiklund, Lumpkin and Frese2009).
Nonetheless, key knowledge voids remain concerning the effects of EO; particularly in the context of the value generated by the firm that does not accrue to shareholders (e.g., Dess, Ireland, Zahra, Floyd, Janney, & Lane, Reference Dess, Ireland, Zahra, Floyd, Janney and Lane2003). It is fair to say that researchers have generally assumed ‘that the primary function of an EO is to enhance financial outcomes rather than to advance other goals that organizations and their managers may pursue’ (Rauch et al., Reference Rauch, Wiklund, Lumpkin and Frese2009: 780). Because of the increasing importance society ascribes to stakeholders – the set of actors who are impacted by, or are capable of affecting organizational outcomes (Freeman, Reference Freeman1984; Wood, Reference Wood1991), and their well-being, organizational performance metrics measuring stakeholder value (SV) have become an influential component of the ‘bottom line’ for the firm (Heath & Palenchar, Reference Heath and Palenchar2009). SV measures the degree to which corporate activities generate value for key stakeholders such as customers, buyers, suppliers, local government, community residents, and the natural environment (Clarkson, Reference Clarkson1995; Mitchell, Agle, & Wood, Reference Mitchell, Agle and Wood1997). For example, corporate activities directed towards managing relations (ensuring fair prices, improved work environments, nondiscrimination on the job, customer service, etc.) with employees, buyers, suppliers, community residents, and the government plus strategies which ensure that corporate actions do not inflict harm on the physical environment (reduction in emissions of harmful pollutants, preparation of sustainability reports and policies concerning disclosure of incidents, etc.) evidence greater SV. While firm value has several claimants, the stakeholder groups referenced above hold a salient position in the minds of CEO’s (Agle, Mitchell, & Sonnenfeld, Reference Agle, Mitchell and Sonnenfeld1999). Increasingly, indicators of SV are being instituted as metrics in the design of reward structures for managers (Berrone & Gomez-Mejia, Reference Berrone and Gomez-Mejia2009). In this vein, Dess et al. (Reference Dess, Ireland, Zahra, Floyd, Janney and Lane2003) also recommend exploring outcomes that benefit groups other than the firm’s financial shareholders when exploring entrepreneurial strategy-making processes. Heeding this call, we posit: Do more entrepreneurially oriented firms generate higher SV?
EO AND SV
EO is a strategic and managerial posture which has been explored in contexts that are financially driven (Lee & Chu, Reference Lee and Chu2013; Kollmann & Stöckmann, Reference Kollmann and Stöckmann2014; Saeed, Yousafzai, & Engelen, Reference Saeed, Yousafzai and Engelen2014), as well as nonprofit-oriented contexts where behaving entrepreneurially may help advance the social mission of a nonprofit organization (Morris, Webb, & Franklin, Reference Morris, Webb and Franklin2011). While the exploration of EO within nonprofit contexts represents a potentially fruitful area of inquiry, this study rather deals with the question of whether EO within commercially motivated enterprises influences the degree to which these institutions generally attend to the needs of their key stakeholder groups thereby increasing the overall value accrued by firm stakeholders.
In their seminal work, Lumpkin and Dess (Reference Lumpkin and Dess1996) offered a holistic perspective on possible EO–performance relationships, including proposing an effect of EO on stakeholder satisfaction. Nonetheless, a metaanalysis conducted by Rauch et al. (Reference Rauch, Wiklund, Lumpkin and Frese2009) on the EO–financial performance relationship observes limited prior research of EO on nonfinancial goals. Yet, their analysis suggests that the influence of EO on nonfinancial performance metrics (i.e., managerial goal attainment, satisfaction, global success ratings, etc.) may be comparable in strength to its influence on financial firm performance. As such, the present study examines EO’s effect on SV.
Central to the concept of SV is the recognition that key stakeholder groups demand and require managerial attention (Mitchell, Agle, & Wood, Reference Mitchell, Agle and Wood1997). CEO’s of large corporations have been shown to attend to the needs of stakeholder groups such as buyers, suppliers, employees, governmental regulators, and the communities within which the corporation operates (Ahmed, Balzarova, & Cohen, Reference Ahmed, Balzarova and Cohen2014). As these stakeholders control the flow of resources valuable to aiding firms in achieving a competitive advantage, managerial attention devoted to fulfilling their demands and managing good relationships with them is vital to a firm’s legitimacy, success, and survival (Clarkson, Reference Clarkson1995; Freeman, Wicks, & Parmar, Reference Freeman, Wicks and Parmar2004; Portney, Reference Portney2008). Managing stakeholder relationships requires the channeling of valuable firm resources such as managerial time and attention devoted to stakeholder communications, and a direct allocation of financial and/or nonfinancial resources to stakeholders’ legitimate causes (Harrison, Bosse, & Phillips, Reference Harrison, Bosse and Phillips2010). Thus, when firms invest in stakeholder-related activities they transfer some of the value generated within the firm to stakeholders.
Corporate activities that promote SV have attracted appreciable scholarly interest (Jones, Reference Jones1995; Jones & Wicks, Reference Jones and Wicks1999). Most advances made in this area have approached SV as an outcome of external pressures exerted on an organization by its stakeholders. In this vein, corporations face multiple demands from various actors in the business environment, for example, explicit requirements imposed by governmental agencies or expectations of socially legitimate behavior by society at large to act responsibly (e.g., Cambra-Fierro, Wilson, Polo-Redondo, Fuster-Mur, & Lopez-Perez, Reference Cambra-Fierro, Wilson, Polo-Redondo, Fuster-Mur and Lopez-Perez2013). Corporations respond to these external expectations by engaging in activities that enhance SV (Johnson & Greening, Reference Johnson and Greening1999; Aguilera, Rupp, Williams, & Ganapathi, Reference Aguilera, Rupp, Williams and Ganapathi2007; Campbell, Reference Campbell2007). However, it has also been suggested that intrafirm characteristics may be equally, or perhaps even more important in explaining corporate attention to SV (Logsdon & Yuthas, Reference Logsdon and Yuthas1997). A key assertion within this line of inquiry is that greater understanding of the organizational-level phenomena that drive SV is needed (Rupp, Williams, & Aguilera, Reference Rupp, Williams and Aguilera2011; Aguinis & Glavas, Reference Aguinis and Glavas2012). Accordingly, we explore EO as an influential firm strategic orientation (Covin & Lumpkin, Reference Covin and Lumpkin2011), which may help to explain increases in firm SV.
HYPOTHESES
A growing number of studies have begun to investigate the individual components of EO (i.e., Hughes & Morgan, Reference Hughes and Morgan2007; Kreiser & Davis, Reference Kreiser and Davis2010; Pearce, Fritz, & Davis, Reference Pearce, Fritz and Davis2010; Marques, Ferreira, Ferreira, & Lages, Reference Marques, Ferreira, Ferreira and Lages2013). In line with the previously discussed notion that the drivers of firms’ SV may be multifaceted, we explore the relationships between each of the three core dimensions of firm EO and corporate SV.
Innovativeness
The combination and recombination of resources into novel means–end relationships is a fundamental characteristic of the entrepreneurial process (Schumpeter, Reference Schumpeter1934; Shane & Venkataraman, Reference Shane and Venkataraman2000). Innovativeness has long been argued to represent a central element of entrepreneurial firm behavior (Stevenson & Gumpert, Reference Stevenson and Gumpert1985; Covin & Miles, Reference Covin and Miles1999). Indeed, Stopford and Baden-Fuller (Reference Stopford and Baden-Fuller1994: 522) observe that, ‘most authors accept that all types of entrepreneurship are based on innovations.’ Pearce, Fritz, and Davis (Reference Pearce, Fritz and Davis2010) provide empirical support for the effect of innovativeness upon firm performance within a sample of religious congregations, demonstrating innovativeness to have a stronger influence on performance than either risk-taking or proactive entrepreneurial behaviors.
At its core, innovativeness encourages managers to engage in creativity and experimentation when solving problems (Hamel & Prahalad, Reference Hamel and Prahalad1994). Innovativeness drives the discovery of new resources and opportunities for a wide variety of new organizational initiatives (Williams & Lee, Reference Williams and Lee2009). We expect that SV will increase in tandem with an increase in the intensity of investments towards innovation. Indeed, as firms seek solutions to lower costs, improve the value of products, and remove competitive ‘road blocks’ including those that come from stakeholders, they may enhance their competitiveness as well as generate additional SV (Porter & Linde, Reference Porter and van der Linde1995; Pavelin & Porter, Reference Pavelin and Porter2008). With greater innovativeness, we expect firms to identify more opportunities and rationale for enhancing SV.
Innovativeness manifests as a strong desire to explore new ideas and to exploit emerging opportunities for learning (Cohen & Levinthal, Reference Cohen and Levinthal1989). These ideas are not only internally focused but may also be engendered through interaction with firm stakeholders. Research confirms that innovation-driven exploitation of knowledge gained from such stakeholders not only enhances firm learning capabilities but also signals the mindfulness of the firm towards its key stakeholders thereby improving relationships with them (Waddock, Reference Waddock2001; Ayuso, Rodríguez, & Ricart, Reference Ayuso, Rodríguez and Ricart2006). In sum, innovativeness leads firms to discover new opportunities and rationale for engaging with stakeholders, encourages better stakeholder relationships through dialogs and knowledge transfers, and motivates increased engagement. Given that catering to stakeholder satisfaction and improving relationships with them enhances SV, we hypothesize the following:
Hypothesis 1: Firms with higher degrees of innovativeness within their EO generate greater SV.
Risk taking
Risk is a fundamental aspect of the entrepreneurial process as the rewards of entrepreneurial activity are, by definition, uncertain (Knight, Reference Knight1921; Arrow, Reference Arrow1974; McMullen & Shepherd, Reference McMullen and Shepherd2006). Risk taking refers to the commitment of a significant portion of organizational assets to uncertain endeavors (Baird & Thomas, Reference Baird and Thomas1985). From the perspective of an EO, risk taking has been defined as ‘the extent to which top managers are inclined to take business-related risks’ and/or ‘the degree to which managers are willing to make large and risky resource commitments—i.e., those which have a reasonable chance of costly failure’ (Miller & Friesen, Reference Miller and Friesen1978: 923; Covin & Slevin, Reference Covin and Slevin1988: 218). Many other authors have adopted definitions that are similar to the ones presented above, all emphasizing the role of corporate managers in choosing less risky or more risky investments in the pursuit of increased financial performance.
Risk-taking taxes a firm’s resource base as, by definition, significant portions of the organization’s resources are committed to uncertain opportunities. This suggests that with greater risk taking the organizational resources available to attend to the needs of all key stakeholder groups are diminished. Pursuing such uncertain opportunities demands a higher commitment of organizational resources, including top managers (Selznick, Reference Selznick1957; Barney & Arikan, Reference Barney and Arikan2001), and managerial attention is itself a limited firm resource (Simon, Reference Simon1947).
With greater risk taking through deeper investments, managers must devote more of their attention to managing the specific risks associated with the entrepreneurial endeavor at hand. If firms devote a sizable portion of their resources towards a risky initiative, then managers are likely to focus more on that initiative. Working to manage, mitigate, and control the risks associated with the highly risky initiative is likely to be very demanding upon managerial attention. In this type of environment, managers’ attention is also heavily focused on evaluating the possible outcomes (success, failure, or even extinction) for themselves and their firms (March & Sharpira, Reference March and Shapira1987, Reference March and Sharpira1992) rather than on the implications of their risky decisions for the firm’s stakeholders. As such, firm managers may be less inclined to expand their stakeholder involvement and value creation when risk taking is high.
Risk taking also may occur through the pursuit of additional entrepreneurial investments which serve to annex free resources within the organization plus reduce slack and strategic degrees of freedom. It has long been argued that increased risk taking through the undertaking of additional entrepreneurial initiatives is often symptomatic of managerial hubris (Li & Tang, Reference Li and Tang2010). In this regard, hubris may be manifest in the form of overconfidence that the organization is capable of high levels of risk taking in terms of undertaking additional entrepreneurial initiatives and the resource requirements which accompany them, while still robustly attending to the multitude of demands placed by extant stakeholders. It is, perhaps, not surprising that overconfident managers tend to underestimate resource necessities (Shane & Stuart, Reference Shane and Stuart2002). As such, high levels of firm risk taking may decrease the degrees of freedom within the firm to attend to diverse stakeholder needs and in doing so, lower overall SV. In this vein, many high risk-taking companies often fall into the trap of technological ‘myopia’ or focusing on their technology as opposed to the broader needs of the stakeholders, which influence their products ultimate adoption (Galbraith, Reference Galbraith1967; Unsworth, Sawang, Murray, Norman, & Sorbello, Reference Unsworth, Sawang, Murray, Norman and Sorbello2012). Therefore, we propose:
Hypothesis 2: Firms with higher degrees of risk taking within their EO generate lower SV.
Proactiveness
The recognition and exploitation of new opportunities ahead of competitors represents a central theme in entrepreneurship research (Shane & Venkataraman, Reference Shane and Venkataraman2000). Proactiveness captures an organization’s efforts to act opportunistically and assume an industry leadership role. It is also a strategic posture which positions the firm to preempt competition in the marketplace and shape the development of broader environmental trends (Lumpkin & Dess, Reference Lumpkin and Dess1996). With higher levels of proactiveness, we propose that firms are more likely to attend to the demands of key stakeholders and increase their value accrued as addressing stakeholder issues is a path to marketplace leadership, as well as a potential way to gain or maintain a competitive advantage.
To explain, proactively attending to external stakeholder needs can help organizations enhance their overall social legitimacy (Singh, Tucker, & House, Reference Singh, Tucker and House1986). Such activates facilitate the acquisition of valuable resources and information-based competitive advantages, which enable the organization to more effectively ‘influence important stakeholders’ (Oliver & Holzinger, Reference Oliver and Holzinger2008: 511). As such, gaining influence with key stakeholders enables firms to achieve better strategic positions in the marketplace from which to shape the development of (as opposed to simply reacting to) external trends. For example, firms with a more proactive strategic posture are more likely to expend efforts lobbying regulatory agencies for the enactment of favorable policies, or to attempt to establish legitimacy through seeking leadership roles in professional associations with the goal of affecting industry rules or norms (DiMaggio & Powell, Reference DiMaggio and Powell1983).
Beyond influencing key stakeholder groups, building upon Buysse and Verbeke (Reference Buysse and Verbeke2003), we argue that proactively oriented firms will also be concerned generally with elevating their organization’s salience in the eyes of important stakeholders in the hope of compelling these groups to reciprocate by attaching greater importance to the organization, its mission, and offerings – again improving resource acquisition and subsequent performance. Overall, these arguments suggest that a proactive strategic decision-making orientation will favor investing in stakeholder relationships in order to attain and maintain legitimacy, as well as to acquire reputational and competitive advantages. Therefore, we hypothesize:
Hypothesis 3: Firms with higher degrees of proactiveness within their EO generate higher SV.
DATA SOURCE AND SAMPLE CONSTRUCTION
We collected data from two different sources – the Kinder, Lydenburg, and Domini (KLD) database, and EO data from secondary sources. In order to develop a panel data set to provide a robust test of our hypotheses (Hsiao, 2003) we gathered data on all firms, which appeared in the KLD database between 2005 and 2008Footnote 1 . KLD Research and Analytics is a MSCI subsidiary that specializes in collecting, analyzing, and providing objective, sector-specific ratings on corporations’ management of their stakeholder relationships. ‘KLD maintains an independent research staff with industry and issue specialties … KLD analysis teams also closely follow the evolution of these and other issues within each industry KLD covers’ (Kinder, Reference Kinder2007: 4).
KLD analyzes data from a variety of sources including regulatory filings, company websites, direct company communications (e.g., annual surveys), industry and trade associations, government and nongovernment sources (e.g., NGO reports), plus media coverage. Independent expert analysts rate a company’s performance on a positive and negative scale to measure strengths and concerns, respectively, on several items within each dimension. As data is collected from corporations operating in different industries, KLD utilizes proprietary technology to assign sector-specific weights to the ratings, and then annually reviews, plus adjusts the weights according to changing risks and opportunities faced by corporations (KLD, 2008). More information on the KLD team of expert analysts is available on their archived website (KLD Web Page, 2009). Numerous studies on the topic of SV have used data from the KLD database because of its comprehensive independent and objective assessment of stakeholder issues (e.g., Waddock & Graves, Reference Waddock and Graves1997; Ruf, Muralidhar, & Paul, Reference Ruf, Muralidhar and Paul1998; Hillman & Keim, Reference Hillman and Keim2001; Kacperczyk, Reference Kacperczyk2009; Walls, Berrone, & Phan, Reference Walls, Berrone and Phan2012). Moreover, the continuous version of KLD data we have used in this study possesses superior measurement characteristics (Hart & Sharfman, Reference Hart and Sharfman2015) than the binary version that has already seen widespread acceptance as the standard source for SV data.
We supplemented our data on SV with financial data extracted from the Compustat database for the years 2004 to 2007. The 1 year lag in collecting objective data to construct EO variables is intentional and by design allows us to explore the impact of prior EO on current levels of SVFootnote 2 . Use of secondary data to operationalize EO follows the conceptual organizational resource allocations approach (Lyon, Lumpkin, & Dess, Reference Lyon, Lumpkin and Dess2000) and recent empirical precedent where researchers use archival data to create valid measures for EO (Miller & Le Breton-Miller, Reference Miller and Le Breton-Miller2011). Measuring how firms allocate scarce resources provides a conceptually valid and reliable proxy for firms’ strategy-making processes (Miller & Friesen, Reference Miller and Friesen1978) as it can ‘more clearly capture emergent or realized entrepreneurial behavior’ (Lyon, Lumpkin, & Dess, Reference Lyon, Lumpkin and Dess2000: 1075). Specifically, because the choice of measurement approach should coincide with the theoretical perspective appropriate for a given research question (Boyd, Dees, & Rasheed, Reference Boyd, Dees and Rasheed1993); in this instance, given that SV is tightly coupled with an organization’s resource endowments (McWilliams & Siegel, Reference McWilliams and Siegel2011), adopting a relative resource allocation perspective may provide a ‘more meaningful indicator of EO’ (Lyon, Lumpkin, & Dess, Reference Lyon, Lumpkin and Dess2000; 1077). From an empirical perspective as well, this direction builds upon recommendations by Miller (Reference Miller2011) to explore alternative operationalizations of EO using objective indicators for each of its components.
As the years of our objective indicators range from 2004 to 2007, the measurement of our independent variables begins and ends 1 year before the measurement of our dependent variable (i.e., 2005–2008). While most research on EO has been cross-sectional (Rauch et al., Reference Rauch, Wiklund, Lumpkin and Frese2009), prior research suggests that the effects of EO can take some time to manifest (Zahra & Covin, Reference Zahra and Covin1995; Wiklund, Reference Wiklund1999). By introducing a lag of 1 year between EO and SV, the present study creates assurances that EO has been permitted time to affect the dependent variable and that the direction of causality is from EO to SV. After retaining observations for which information on all study variables was available on each firm for at least 2 consecutive yearsFootnote 3 , we arrived at a panel data set of 1,015 public US firms in 53 industriesFootnote 4 (based on two-digit SIC codes) over 5 years. Appendix lists the industry composition of sample firms.
VARIABLES
SV
SV is reflected in corporate activities directed towards managing relations with stakeholders. Consistent with previous strategic management literature (e.g., Berman, Wicks, Kotha, & Jones, Reference Berman, Wicks, Kotha and Jones1999; Hillman & Keim, Reference Hillman and Keim2001), we use the ratings on five dimensions of attention to primary stakeholder groups compiled by KLD within the categories of community, diversity, employees, product, and environment to measure SVFootnote 5 . These areas encompass community relations (e.g., charitable giving, support for education and housing, compliance with federal, state, or local government), employee relations (e.g., worker health and safety plus union issues), product quality (a proxy for customer relations), workforce diversity (e.g., record on minority discrimination, treatment of the differently abled), and environmental performance.
In the continuous version of the data we used, KLD rates the various items within each stakeholder dimension on a scale ranging from 0 to 30 for strengths, and −30 to 0 for concerns. Using data from the years 2005 to 2008, we computed the average of nonmissing strengths and concerns ratings separately to construct two sets of scores for each of the five SV dimensions. The standardized scale reliability coefficient Cronbach’s α (Cronbach, Reference Cronbach1951) for the strengths and concerns scores computed over 1 year was 0.8 and 0.7, respectively. In social sciences research, Cronbach’s α values approaching 0.7 are indicative of acceptable levels of reliability of a composite variable (Nunnally & Bernstein, Reference Nunnally and Bernstein1994). The Cronbach’s α for a 2-item scale comprising the strengths and concerns scores was 0.7. The interitem correlation (r=0.52, p<.0001) among the strengths and concerns groups was well above 0.3, signaling good internal consistency, and convergence (Van de Ven & Ferry, Reference Van de Ven and Ferry1980; Mitchell & Jolley, Reference Mitchell and Jolley1988). Assured that the strengths and concerns dimensions are sufficiently correlated in our data and represent a singular underlying construct (Griffin & Mahon, Reference Griffin and Mahon1997), we continued by following recommendations in prior research and subtracted the concerns scores from the strengths scores to arrive at a composite measure of SV.
While we have followed the empirical precedent within SV research of creating a composite measure of SV, our approach is supported by other methodological experts who suggest that in situations where multiple indicators are correlated with each other, the use of an aggregated variable may alleviate multicollinearity problems in regression-based analyses (Kennedy, Reference Kennedy2003). Moreover, an aggregated SV measure constructed using multiple dimensions will enhance the generalizability of results (Chatterji, Levine, & Toffel, Reference Chatterji, Levine and Toffel2009).
EO
We focused on three dimensions of EO, innovativeness, risk taking, and proactiveness. The present study extends prior research by measuring EO using a more objective, secondary measure – a measurement approach which recently has been advocated as an important avenue to advancing the EO literature (Covin & Lumpkin, Reference Covin and Lumpkin2011) given that the vast majority of prior research has been reliant on primary data (Lyon, Lumpkin, & Dess, Reference Lyon, Lumpkin and Dess2000). All EO variables used in this study were constructed using objective indicators (Miller, Reference Miller2011). A description of each indicator’s suitability for assessing a component of EO is discussed by Miller and Le Breton-Miller (Reference Miller and Le Breton-Miller2011: 1061–1065), and we briefly outline each below.
Innovativeness
We measured this variable as the ratio of a firm’s research and development (R&D) expenses to its sales, a measure that is consistent with prior EO research (Deeds, DeCarolis, & Coombs, Reference Deeds, DeCarolis and Coombs1998; Miller & Le Breton-Miller, Reference Miller and Le Breton-Miller2011). Intriguingly, data on R&D expenses is available for far fewer firms than data on other accounting items such as firm revenues, or assets. Moreover, many firms do not spend measurable amounts of capital on R&D; hence, their R&D expenses are reported as zero. The challenge of collecting data on R&D in accordance with Miller and Le Breton-Miller (Reference Miller and Le Breton-Miller2011) therefore reduces the overall sample size, limiting statistical power to detect effects of the hypothesized variable(s) on the outcomes of interest (Cohen, Reference Cohen1992). In our initial sample over 52% of firms reported R&D expenditures allowing a reasonably large number of observations over which to conduct our analyses. Nevertheless, in order to address the abovementioned challenge posed by missing R&D observations, we followed prior research (Hanlon, Rajgopal, & Shevlin, Reference Hanlon, Rajgopal and Shevlin2003) and re-ran our analysis by filling the missing values of the innovativeness variable with zeroes. Following the recommendations of Hall and Reenen (Reference Hall and Reenen2000) we also created an indicator variable which took on values of 1 if missing R&D values were filled in with a zero, and 0 otherwise. Results from our supplemental analysis replicated our principal analysis; all statistically significant results retained their hypothesized direction and statistical significance. Moreover, the indicator variable failed to achieve statistical significance providing further assurances for the manner in which we constructed the innovativeness measure.
Risk taking
As investments made in the pursuit of uncertain opportunities have a higher chance of turning out to be costly for the firm (Miller & Friesen, Reference Miller and Friesen1978), risk taking is usually reflected in the unsystematic (or idiosyncratic) risk faced by the firm. Consistent with prior research, we approached a firm’s risk-taking behavior by observing the fluctuations in its market-valuation compared to other firms in its industry (Miller & Le Breton-Miller, Reference Miller and Le Breton-Miller2011). When firms embark on risky projects, enter new untested markets, or invest aggressively, their stock price is likely to exhibit more volatility in comparison to that of their more cautious industry counterparts. Indeed, higher unsystematic risk reflects the bold initiatives taken by a firm’s management in the pursuit of uncertain opportunities for financial gain (Miller & Le Breton-Miller, Reference Miller and Le Breton-Miller2011). Following Miller and Le Breton-Miller (Reference Miller and Le Breton-Miller2011), we measured this variable as the unsystematic component of a firm’s stock price fluctuations which are reflective of corporate managers’ strategic decision-making (Sanders & Carpenter, Reference Sanders and Carpenter2003; Sanders & Hambrick, Reference Sanders and Hambrick2007). We computed this variable for each firm by taking a rolling 5 year average of its monthly stock price volatility for each of the 4 years from 2004 to 2007.
Proactiveness
We followed prior EO research and measured proactiveness by the percentage of profits reinvested in the firm each year adjusted for industry competition. The variable was constructed by first calculating the mean industry level of percentage of profits reinvested per year for each firm’s industry, excluding the focal firm’s reinvested profits. Consistent with past research (e.g., Miller & Le Breton-Miller, Reference Miller and Le Breton-Miller2011), the variable we computed proactiveness by subtracting this industry average from each firm’s percentage of reinvested profits. The measure thus provides an overall proxy for the constructive moves made by management in pursuit of opportunities, calculated over time, adjusted for any industry-level factors or trends driving the decision to re-invest profits.
Control variables
We included several variables in our models to control for alternative influences upon a corporation’s SV. Larger firms, due to their sheer size, are more visible to stakeholders and hence have to create more SV (Orlitzky, Reference Orlitzky2001). To control for size effects, we computed firm size as the natural logarithm of a firm’s total assetsFootnote 6 . We also controlled for corporate governance which may have a direct impact on firm strategies concerning the handling of stakeholder issues (Coffey & Wang, Reference Coffey and Wang1998; Johnson & Greening, Reference Johnson and Greening1999). We approached the corporate governance variable by utilizing the items in the corporate governance dimension within the KLD dataFootnote 7 and constructed the measure by taking the mean of both the aggregated strengths and concerns scores in this categoryFootnote 8 .
Heterogeneous firm-level endowments can also explain variations in firm activities directed at enhancing SV. From an organizational slack perspective, a corporation’s SV may be driven by excess physical and financial resources available to managers that may be used to invest in attending to the needs of key stakeholders (McGuire, Sundgren, & Schneeweis, Reference McGuire, Sundgren and Schneeweis1988). We accounted for munificence-driven explanations for attention to stakeholders by controlling for several firm-level variables. We operationalized firm slack using both the quick and current ratios. Results are similar using either operationalization; in order to maintain consistency in the presentation of results all models used the current ratio operationalization of firm slack computed as the ratio of current assets to current liabilities. The firm performance variable we created using the return on assets ratio calculated by dividing net income by total assets. Using return on sales as an alternate measure, computed by dividing firms’ net income by sales, does not alter our results. To control for the impact of strategic allocation of resources on a corporation’s attention to stakeholders (Russo & Fouts, Reference Russo and Fouts1997) we created the variable capital intensity by first subtracting, for a given year, the current assets of a firm from its total assets, and then divided it by the total number of employees for that year.
Membership in different industries may have a significant impact on the level of SV generated by firms (Griffin & Mahon, Reference Griffin and Mahon1997). We controlled for time-invariant industry effects by introducing dummy variables, constructed using the two-digit SIC codes in Compustat. Industry fixed-effects control for variation in other study variables across industries; for example, the nature and reporting of R&D expenditures may vary across competitors in different industries and using industry dummies helps account for such differences. To control for time-varying industry effects we computed for each firm the variable mean industry SV by taking the average level of the dependent variable for each firm’s industry, per year, excluding the focal firm and included this as an explanatory variable. To control for the impact of business cycles, and to mitigate potential problems from contemporaneous correlation, we introduced year dummy variables which greatly improve the accuracy of panel data regression estimators in the presence of serial correlation (Certo & Semadeni, Reference Certo and Semadeni2006). As firms in KLD data are listed on different stock exchanges and confront varying degrees of risks and stakeholder pressures (Aupperle, Carroll, & Hatfield, Reference Aupperle, Carroll and Hatfield1985; Rehbein, Waddock, & Graves, Reference Rehbein, Waddock and Graves2004), we constructed dummy variables for the stock exchange membership for each firm. All sets of dummy variables were found to be jointly statistically significant; we ran all models retaining them as controls. We inspected all variables for extreme deviation from normality and where necessary transformed them using division by a constant to make the means and standard deviations similar (Cohen, Cohen, West, & Aiken, Reference Cohen, Cohen, West and Aiken2003).
ANALYSIS AND RESULTS
We adopted a random-effects panel data estimation technique to test our hypotheses. Using the random-effects estimator is intuitively appealing not only because we use a sample of a larger population (Wooldridge, Reference Wooldridge2002) of public US corporations but also because random-effects estimation yields relatively more efficient estimates compared to those generated by comparable panel data estimators (Greene, Reference Greene2008)Footnote 9 . The Hausman (Reference Hausman1978) difference-in-variance test which is generally employed by researchers to help make the choice between fixed-effects and random-effects estimation was inconclusive because our sample failed to meet the strict asymptotic assumptions of the test; a somewhat common occurrence in modern panel data sets (Davidson & MacKinnon, Reference Davidson and MacKinnon1993; Wadhwa & Kotha, Reference Wadhwa and Kotha2006). However, we found statistically significant random-effects (χ2=324, p<.0001) in our sample using the Breusch and Pagan (Reference Breusch and Pagan1980) Lagrange-Multiplier test and proceeded with the random-effects estimatorFootnote 10 . We estimated all models using STATA statistical software.
In Table 1 we present pair-wise correlations for corporations’ SV, the three dimensions of EO, and the control variables used in the study. As we use panel data, correlations were computed for the year 2007 (the year with the most observations); using another year for producing the correlation matrix yields qualitatively similar results. We inspected the condition index of the design matrix and the variance inflation factors after regressions. Both the condition number and the mean variance inflation factor never exceeded the critical limits of 30 and 2, respectively, indicating a low likelihood of misestimations due to multicollinearity (Belsley, Kuh, & Welsch, Reference Belsley, Kuh and Welsch1980; Neter, Wasserman, & Kutner, Reference Neter, Wasserman and Kutner1996; Greene, Reference Greene2008).
Table 1. Means, standard deviations, and correlationsFootnote a
Note. Means and standard deviations are computed over the entire sample of 3,796 observations.
a n=1,015 (year=2007–the year with the most observations).
b Variable multiplied or divided by a constant.
*p<.05; **p<.01; ***p<.001.
In Table 2 we provide the random-effects regression estimates for the impact of the EO dimensions on SV. Model 1 includes all of the control variables in our study. Results indicate that as evidenced in prior research, firm size is a significant predictor (p<.001) of firms’ SV. In Model 2 we tested the unique impact of innovativeness on SV. The coefficient was positive and statistically significant (p<.05), which supports Hypothesis 1; innovativeness has a positive effect on SV. The positive relationship between innovativeness and SV remains consistent in all the models. Moreover, because we report standardized coefficients, the magnitude of the effect of innovativeness on SV is very similar to that of corporate governance on SV. In Model 3, we introduced risk taking as a predictor of SV. The coefficient of risk taking is negative and statistically significant (p<.05) providing support for Hypothesis 2. Firm risk taking is indeed negatively related to managing primary/salient stakeholder relationships resulting in lower SV. The coefficients remains negative and statistically significant in all models providing consistent support for Hypothesis 2. Finally, in Model 4 we tested the relationship between proactiveness and SV. Model 4 is also the full model in which all predictor variables plus control variables are included. The coefficient of proactiveness is positive and statistically significant (p<.01). This result provides strong support for Hypothesis 3; proactiveness positively impacts SV. In addition, the impact of proactiveness on SV appears to be similar in magnitude to the effects of slack or financial performance on SV.
Table 2. Results of random-effects regression on stakeholder valueFootnote a
Notes. Constant term included, coefficients not reported.
Standardized β estimates shown.
Heteroskedasticity-robust standard errors are in parentheses.
a n=3,796 observations.
† p<.10; *p<.05; **p<.01; ***p<.001.
DISCUSSION
We proposed and tested a relationship between EO and SV using longitudinally collected secondary data from 1,015 public US firms. We advance prior research methodologically by going beyond cross-sectional investigations of EO–outcome relationships, expressly incorporating the notion that the effects of EO take time to manifest (Zahra & Covin, Reference Zahra and Covin1995; Wiklund, Reference Wiklund1999). We also extend prior research by employing secondary measures of EO which build upon observations that subjective and objective measures of entrepreneurial behavior are comparable (Jennings & Young, Reference Jennings and Young1990), and that their use enhances the comprehensiveness of EO research (Miller, Reference Miller2011). Thus, this work addresses previous calls for greater research exploring secondary measures of EO (Lyon, Lumpkin, & Dess, Reference Lyon, Lumpkin and Dess2000).
A principal implication of this study is that the relationship between firm-level EO and SV is not homogenous, but rather complex and divergent. In line with Lumpkin and Dess (Reference Lumpkin and Dess1996), a core contribution of this study lies in the explication of how the individual dimensions of EO might either contribute to constructive implications for SV (i.e., in the case of innovativeness and proactiveness), or adverse implications (i.e., in the case of risk taking). This nuanced relationship is in contrast to the vast majority of prior research exploring the dimensions of EO individually which has frequently observed the effects of the dimensions to be in the same direction – often with one or more components of EO manifesting a nonsignificant effect upon the dependent variable of interest (i.e., Covin, Green, & Slevin, Reference Covin, Green and Slevin2006; Hughes & Morgan, Reference Hughes and Morgan2007; Pearce, Fritz, & Davis, Reference Pearce, Fritz and Davis2010). In only a few studies have relationships been demonstrated in which dimensions of EO actually manifest divergent implications for the outcome of interest (e.g., Short, Broberg, Cogliser, & Brigham, Reference Short, Broberg, Cogliser and Brigham2010).
Moreover, there is on-going debate in the stakeholder literature as to the antecedents of corporation’s choices when dealing with its stakeholders (e.g., Chiu & Sharfman, 2011). While a variety of studies have posited internal and external antecedents to SV, the present results take this research in potentially a new direction. Our results suggest that internal EO behaviors have direct implications for stakeholder-related initiatives (i.e., nonfinancial outcomes). This set of results shows that rather than being separate issues, SV levels are critically intertwined with the entrepreneurial strategic orientation which permeates the firm. As such, analyzing the firm’s past entrepreneurially oriented behavioral patterns can give some insight into the firm’s present and likely near-term overall level of SV being generated. In doing so, we offer an additional, holistic firm performance variable to future studies examining the outcomes of EO.
Findings of this study also have managerial implicationsFootnote 11 . Given the rise in both the pressure on firms to enhance SV and the increased efforts firms are spending on this activity, managing such efforts effectively becomes paramount. Our results lead to some ideas managers could implement to enhance their SV creation efforts. Given that our first and third hypotheses confirmed positive relationships between innovativeness plus proactiveness and SV, it might seem that firms would simply want to be as innovative and proactive as possible. However, given that there is some evidence (Tang, Tang, Marino, Zhang, & Li, Reference Tang, Tang, Marino, Zhang and Li2008) that the relationship between EO and firm financial performance is at times curvilinear, managers first must be careful to implement additional innovation and proactivity efforts consistent with peers and organizational norms. While some outdistancing of competitors may be warranted for competitive purposes, mangers must be sure that these efforts are not distracting from key firm and stakeholder-related initiatives.
Second, given that innovative and proactive activities are related to enhanced SV, it is incumbent on management to insure that stakeholders are aware of these efforts. Given the vast increase in social media use across all manner of firms, those techniques might be quite helpful in informing stakeholders of specific innovative or proactive efforts by the firm. Finally, because risk taking was negatively related to SV, firms must be careful in how such efforts are portrayed. It is conventional wisdom that firms, particularly global firms, must take risks. However, firms have the ability to portray those risks in ways that will not affect stakeholders perhaps as negatively. Specifically, risks taken to preserve gains or to take advantage of opportunities, likely will be seen less negatively by stakeholders (Berman & West, Reference Berman and West1998). This is not to say that managers should falsely ‘spin’ risk-taking actions but if there is an opportunity that can be highlighted managers may find stakeholders more supportive or at least less resistive.
Implications notwithstanding, the contributions of this study should be considered in light of its research limitations. While additional dimensions of EO have been suggested within the literature, for the sake of parsimony, and consistency with recent research, the present study focused upon the three most often used dimensions of EO, which incidentally have been commonly included across the various alternative conceptualizations of EO (Wales, Gupta, & Mousa, Reference Wales, Gupta and Mousa2013). It is of course possible that the effects of additional dimensions of being entrepreneurial may also be significant and serve to further refine our understanding of this important relationship, such as those proposed by Lumpkin and Dess (Reference Lumpkin and Dess1996) and Antoncic and Hisrich (Reference Antoncic and Hisrich2003), among others. Moreover, while the measurement of EO at an organizational resource allocation level was relevant to our research question and empirically appropriate in our research setting, future research may find it valuable to evaluate these results through complementary operationalization and measurement of EO (Lyon, Lumpkin, & Dess, Reference Lyon, Lumpkin and Dess2000) or through analysis within a contingency framework (Van de Ven & Drazin, Reference Van de Ven and Drazin1985) to help enhance the validity and generalizability of the results.
Future research may extend the present study by exploring the extent to which the relationship between EO and SV is moderated by various organizational and environmental variables (Covin & Slevin, 1991; Lumpkin & Dess, Reference Lumpkin and Dess1996). While the present study broadens the boundaries of EO research to include outcomes concerning SV, questions for future research remain concerning the possibility of contextual influences upon the strength of the EO–SV relationship. In particular, factors such as economic prosperity, environmental uncertainty, and complexity may have meaningful impact on the relationships proposed within this study. While beyond the scope of the present study, examining a more extensive spectrum of outcomes from EO in future research including moderating influences, alternative social value indicators, and broader ethical considerations will ultimately provide an even better picture of EO’s effects upon meaningful firm outcomes beyond its well-established financial implications.
Acknowledgement
The authors gratefully acknowledge Professor Tom Lumpkin for his generous feedback and constructive comments on a previous version of this manuscript.
Appendix
Industry composition of sample firms
