Marketing experience of CEOs and corporate social performance

Abstract

Corporate social performance (CSP) is increasingly becoming an important firm performance dimension in its own right. Since the CEO plays a pivotal role in setting the firm’s strategic actions, the examination of CSP’s antecedents has often focused on how CEO characteristics may impact CSP. According to upper echelons theory, one such key characteristic is the CEO’s functional background. As CEO experience in marketing may instill a CSP-supportive mindset in line with stakeholder theory, we examine how such CEO marketing experience may promote CSP and how situational factors may moderate this. Analyses of 3,569 CEOs from 1,999 firms from 2001 to 2016 reveal that CEO prior experience in marketing positively relates to CSP. This finding is robust to multiple analytical methods and endogeneity checks. Further, marketing experience’s effect is stronger than that of other functional experiences. Moderation results indicate this effect is associated more with executive discretion than job demands.

The idea, expressed in various forms (e.g., triple-bottom-line, corporate social responsibility), that firms should account for the interests and obligations of stakeholders beyond just shareholders has become more prominent over the past three decades (e.g., Smith, 2009). Recently, the Business Roundtable—an influential group of 200 top CEOs—confirmed this view, arguing that firms should invest in their employees, protect the environment, and deal ethically with customers and suppliers (Business Roundtable, 2019).

A considerable body of scholarly literature indicates that firms benefit from such corporate social responsibility (CSR)¹—organization actions that aim to advance social good beyond that which is required by law (e.g., McWilliams & Siegel, 2001). Extant research reveals an overall positive financial picture for firms investing in pro-social causes (e.g., Kang et al., 2016; Lenz et al., 2017). Corporate social performance (CSP) also delivers various consumer-level benefits for the firm, such as increased customer satisfaction and loyalty, customer-firm identification, and improved firm and brand image (Chernev & Blair, 2015; Lacey et al., 2015).

Moreover, evidence is increasing that CSP is becoming a key performance metric for firms (or KPI) in its own right. For instance, firms are increasingly incorporating CSP-related targets in top management team compensation packages (Ikram et al., 2019). CSP information is reported on quarterly earnings calls, as equity analysts consider this information an important factor in their valuation models (Ioannou & Serafeim, 2015). Investors and credit rating agencies are also using CSP when making investment decisions and to assess creditworthiness (e.g., Ikram et al., 2019; Nikolaeva & Bicho, 2011). Further, the Big Four accounting firms are promoting the importance of CSP metrics and encouraging firms to report these in their financial statements (Cohn, 2020). As a result, more than 90 percent of the 250 largest companies in the world annually report on their corporate social performance (McLymont, 2018).

As such, scholarly attention has begun to focus more on understanding the antecedents of CSP, frequently through the lens of stakeholder theory. A stakeholder, from this view, is any “group that the firm needs in order to exist, specifically customers, suppliers, employees, financiers, and communities” (Dunham et al., 2006). Management, therefore, has the critical task of balancing the often-conflicting interests of multiple stakeholders and identifying and supporting those groups which have an impact on a company’s ability to operate in the marketplace (Hillebrand et al., 2015). Researchers further suggest that a firm’s CSR strategies are one of the few levers that can be used to build and improve stakeholder relationships (Hoeffler et al., 2010; Sen et al., 2006). Therefore, factors that increase the salience and impact of stakeholders’ demands on the firm can lead to increased CSP (e.g., Brower & Mahajan, 2013).

In addition, scholars have also focused on the role of the CEO in CSP (Agle et al., 1999), because the CEO, as leader of the firm, shapes the values and orientation of an organization and plays a pivotal role in setting the firm’s strategic actions² (e.g., Hambrick, 2007). Since the CEO can thus be a critically important driver in shifting the firm to act in socially responsible ways (e.g., Wood, 1991), scholars have considered how various CEO characteristics may relate to CSP, drawing on upper echelons theory (UET) (e.g., see Whitler et al., 2021 for a review). UET holds that executives act on their individual interpretations of the situations they face, and as such, their experiences, values, and personalities result in the development of highly personalized construed realities, which in turn shape firm strategy setting and implementation. Recent empirical work has revealed that CEOs’ values, as measured through their political liberalism (e.g., Chin et al., 2013), as well as CEOs’ personality, in terms of CEO hubris (e.g., Kashmiri et al., 2017; Tang et al., 2015), impact CSP. Yet, this research, with a few exceptions (Manner, 2010; Slater & Dixon-Fowler, 2009; Thomas & Simerly, 1994), has paid less attention to the role of CEO prior functional experience on CSP.

However, as the CEO’s functional experience can strongly affect firm strategic decision making (Hambrick & Mason, 1984), we suggest that prior CEO experience in marketing is likely to influence CSP. This is because marketing experience can instill a mindset that leads the CEO to focus on a wider variety of audiences and stakeholders (Hult et al., 2011) and because it is likely to increase the belief that meeting the needs and obligations of such stakeholders—such as through CSP—creates critical resources for the firm (Pfeffer & Salancik, 1978).

Further, beyond the need to examine CEO functional experience’s, particularly marketing’s, influence on CSP, there remain unanswered questions in UET that we also hope to address. For instance, UET proposes two distinct perspectives on what leads executive backgrounds to shape strategic choices—the job demands perspective and the discretion perspective, and we specifically select moderators to help untangle these competing mechanisms. In addition, UET is largely silent on how the timing of prior functional experience may influence its effect, and how the alignment between the CEO’s functional experience and firm-level resources may facilitate discretion. Thus, we aim to provide new insights by addressing the following three research questions: (1) how do CEOs’ previous functional experiences, particularly experience in marketing, influence CSP? (2) what conditions moderate this relationship? and (3) as such, which of the two competing UET proposed mechanisms is stronger?

To examine these questions, we compile a comprehensive panel dataset of 3,569 CEOs from 1,999 public U.S. firms from 2001 to 2016, and through difference-in-difference (DID) analysis, we investigate the CSP impact of CEOs’ prior functional experience. Subsequently, we explore how internal and external factors may affect the CEO’s level of discretion (job demands), moderating the effect of marketing functional background. In this analysis, our model controls for important CEO (e.g., gender) and firm-level factors (e.g., firm size, slack resources) previously posited and shown to influence firm CSP (e.g., Brickson, 2007; David et al., 2007; Neubaum & Zahra, 2006), and we account for other factors (i.e., institutional and regulatory pressures) through industry-level CSP and industry/firm and year fixed effects.

As a result, this study makes the following contributions to theory and practice. First, we provide strong evidence for the relationship between CEO marketing experience³ and CSP. Other CEO functional experiences either do not positively impact CSP (i.e., finance, operations, CSR) or provide a weaker effect (general management). Thus, marketing experience appears to be the most influential CEO background experience for CSP. This enriches our understanding of CSP determinants and the role of CEO marketing experience in shaping strategic firm choices.

Second, we examine how specific internal and external factors may moderate the influence of the CEO’s marketing experience on CSP by affecting the CEO’s level of discretion. This helps to answer prior calls for more consideration of boundary conditions of discretion in the marketing literature (Whitler et al., 2021). In particular, we enrich the literature by exploring how discretion may be affected by firm resource-CEO functional experience alignment.

Third, we also offer two new insights for upper echelons theory. We are the first to explore how the timing of functional experience strengthens its influence. While all marketing experience facilitates CSP, results indicate that marketing experience early in the career has the strongest effect. Further, through our set of moderators, we can shed light on the process (Spencer et al., 2005) and distinguish whether the UET-proposed job demands or executive discretion pathway is stronger. Our results indicate that the discretion-based pathway has greater explanatory power, providing additional new knowledge for upper echelons theory.

Finally, our study provides valuable new guidance for firms about the antecedents of CSP. CEO marketing experience enhances CSP, and this effect strengthens by alignment between firm-level factors and the CEO’s marketing background, but it lessens with lower CEO compensation and in more certain market environments. Results thus offer pathways—in terms of promotion, hiring, and marketing-related discretion levers—for firms to enhance their CSP.

We proceed as follows. We first review key theoretical mechanisms proposed in the literature to lead to CSP, with a particular focus on UET and stakeholder theory. We then develop our conceptual model linking CEO functional experience to CSP (see Fig. 1). Following, we present our methodology and empirical results, and we conclude by discussing the theoretical and practical implications, as well as limitations of our research.

Fig. 1

A contingency model for the impact of CEO marketing experience on corporate social performance (CSP)

Theoretical background

In marketing, most scholars have focused on the consequences of corporate social responsibility for firm performance (e.g., Mishra & Modi, 2016). Recent empirical research in this vein (Kang et al., 2016) finds that corporate social responsibility relates to positive financial performance as social investments are viewed as a mechanism for “good management,” adding value for the firm as well as its stakeholders. Research in marketing, however, has paid limited attention to the determinants of CSP.

In the broad CSP space, research on CSP’s antecedents has used several theoretical bases (Table 1 provides a review of CSP’s key internal antecedents; a broader review of the most used theoretical antecedents is provided in Web Appendix Table A). Increasingly, the explanation for CSP has centered on the idea that as firms depend on society for their existence, continuity, and growth, they have certain obligations to give back. The prevailing theory in this domain is stakeholder theory (Clarkson, 1995).

Table 1 Internal antecedents of corporate social performance (CSP): Prior literature

Stakeholder theory

Stakeholder theorists, starting with Freeman (1984), argue for the necessity of a stakeholder orientation of the organization. The theory suggests that executives have a fiduciary relationship to stakeholders and not only to the owners of the corporation. Stakeholders are persons or groups with legitimate interests in procedural and or substantive aspects of corporate activity and the interests of all stakeholders are of intrinsic value (Freeman, 1984). From this perspective, the organization has a responsibility to all of its stakeholders driven by intrinsic moral concerns, and CSP is driven primarily by managerial motivations rather than organizational factors (Mena & Chabowski, 2015).

However, other stakeholder scholars recognize the potential external benefits that can accrue from CSP. In this regard, stakeholder responsiveness, and hence better stakeholder relationships, can be viewed as a valuable firm resource (Brower & Mahajan, 2013; McWilliams & Siegel, 2001; Mena & Chabowski, 2015). In this variant of stakeholder theory, the organization identifies which stakeholder interests are important and hence stakeholder salience is directly relevant (Mitchell et al., 1997). Thus, factors increasing the firm’s sensitivity to stakeholder demands, the diversity of such demands, and the firm’s exposure to stakeholder scrutiny can foster CSP. While prior studies utilizing stakeholder theory tend to emphasize either the intrinsic or the instrumental perspective, the literature recognizes that these approaches can operate in parallel or be coincident (Muller & Kolk, 2010).

Upper echelons theory

Another key explanation advanced for CSP has been upper echelons theory (Hambrick & Mason, 1984). UET focuses on how executives’ personal characteristics color their processing of information in ways that affect their strategic decision-making. As their influence is unrivaled, executives are not uniform in their personal orientations and cognitive decision-making frameworks which they bring to bear in their roles. Consequently, strategy setting is not a technical endeavor with preset rules and regulations, but rather an interpretive one: executives perceive their situations and alternatives through highly personalized lenses, which are formed by their experiences, personalities, and values; these constructs collectively shape their choices (Hambrick, 2007). Thereby, given the CEO’s crucial role in setting firms’ strategic actions (e.g., Whitler et al., 2021), CEO characteristics have also been considered key antecedents of CSP, as they shape the CEO’s construed reality, affecting the attention that the CEO places on particular issues and the preferred solutions (Wood, 1991).

Extant UET work on CSP indicates that executives’ personalities and values significantly affect the CSR policies and outcomes their companies achieve (e.g., Cha et al., 2019; Chin et al., 2013). Some evidence in business ethics indicates that stakeholder focused background experiences (Manner, 2010) and output functional experiences (i.e., functions emphasizing externally oriented activities such as R&D and marketing) are associated with CSP (Slater & Dixon-Fowler, 2009; Thomas & Simerly, 1994). However, this prior research has amalgamated functional experiences together (e.g., it empirically mixes the influence of marketing with other factors, particularly R&D experience), obscuring the effect of specific functional experiences such as marketing. Moreover, little attention has been paid to situational factors that may moderate the relationship between CEO functional experience and CSP. To fill these gaps, we focus on the potential role of marketing experience, as it provides a connecting link between stakeholder theory and UET perspectives.

Integrating UET and stakeholder theory and CSP: The role of attention

UET follows a general logic that executives’ characteristics affect the heterogeneity of attention they place on environmental stimuli, which shape their perceptions and outlooks and hence are reflected in their strategic outcomes. Prior career experiences therefore act as a screen or filter, influencing the cognitive frames that executives bring to bear, guiding executives to pay more attention to stimuli which are more compatible with familiar categories and stay away from unfamiliar ones (Graham et al., 2013). Thus, due to bounded rationality cognitive constraints, strategy setting is an interpretive process, and executives make strategic decisions for the firm on the basis of their personalized understanding of the situation, which is heavily influenced by the executive’s background and experiences (Hambrick, 2007).

Similarly, stakeholder theory considers that the stakeholder attributes in need of attention are not an objective feature in the environment; rather they are a socially constructed reality, as the firm’s managers determine which stakeholders are salient (Mitchell et al., 1997). Thus, to link the UET and stakeholder theory approaches, we suggest that prior marketing experience can create a mental schema that fosters CEO attention to all stakeholders, enhancing stakeholder salience and leading to CSP. Thus, the UET lens is particularly useful because it can explain why CEOs—through their construed reality arising from prior experience—might differentially attend to various stakeholders, explaining variation in firm CSP.

Though there is some evidence that CEOs can be swayed to emphasize CSP through irrational factors such as hubris and narcissism (e.g., Tang et al., 2015), we follow the majority view that the attention to CSP arising from executive background factors follows a cognitive, deliberative process (e.g., Agle et al., 1999; Basu & Palazzo, 2008; Connelly et al., 2011). Thus, we now consider how prior marketing experience may instill a mindset favoring CSP.

Hypotheses

The impact of CEOs’ marketing experience on CSP

A number of factors indicate that CEO prior marketing experience would be instrumental in shifting the CEO’s attention in ways that would promote CSP. Though customers are its priority, the marketing function has a unique perspective that focuses attention on stakeholders beyond the firm’s customers alone, in creating value for the firm and society at large (Hillebrand et al., 2015; Narver & Slater, 1990). Thus marketing experience provides a differentiated knowledge base (Feng et al., 2015), which leads to selective information processing, attention, and considered solutions. Further, supported by marketing’s shift to a relational paradigm (e.g., Palmatier et al., 2006), marketing experience tends to instill a longer-term goal orientation in the face of short-term demands than other functional areas (Feng et al., 2015; Verhoef & Leeflang, 2009). Marketing’s longer-term outlook also can facilitate a more holistic stakeholder view which supports meeting the needs of all external and internal stakeholders (Chabowski et al., 2011). It is therefore thought that marketing’s outward focus and longer-term goal orientation facilitates consideration of a wider range of non-shareholder groups (e.g., customers, suppliers, employees, regulators, and communities) (Hult et al., 2011), consistent with a contemporary stakeholder perspective as identified by stakeholder theory (Crittenden et al., 2011).

Thus, marketing experience is uniquely positioned to shift mindsets to identify, attend to, and consider opportunities related to all salient stakeholders (Smith et al., 2010). Thus, an executive’s schema from prior marketing experience is more well-aligned with—and conducive to—meeting the needs of multiple stakeholders through CSP to that acquired through other functional experiences. In contrast, other functional experiences like operations, accounting, and finance facilitate an inward attentional focus on production efficiency and cost management as well as shorter-term goal orientation (Finkelstein et al., 2009; Whitler et al., 2018) and thus are unlikely to produce schemas as sensitive to the needs of a variety of stakeholders.

In addition, marketing experience may also increase CSP by increasing the salience of its instrumental benefits. Marketing experience can fundamentally showcase the advantages—i.e., consumer receptivity to firm marketing efforts—that accrue when the firm enjoys a favorable corporate reputation and image (Grayson et al., 2008). CEOs with marketing experience would thus be more inclined to view the firm’s reputational capital—the organization’s stock of perceptual and social assets, the quality of the relationships it has established with stakeholders (Fombrun et al., 2004)—as a critical resource for the firm (Pfeffer & Salancik, 1978) and thereby would more strongly believe in the need to build and maintain this through socially responsible actions (e.g., Hoeffler et al., 2010; Luo & Bhattacharya, 2009). Because such CSP generates increased resources that are likely more critical for marketing than for other functions (Maignan & Ferrell, 2004), and because reputation management is often considered a core responsibility of the marketing function (Nikolaeva & Bicho, 2011), CSP may not be perceived as critical of a resource in other functional domains as it would be for marketing. Thus, marketing experience should also strengthen the CEO’s perception of the business case for CSP.

Therefore, to sum, we argue that marketing functional experience emphasizes supporting the needs of a wide range of stakeholders, and thus from a stakeholder theory perspective, increases their salience. Arguably, marketing experience attunes executives to adopt a longer-term goal orientation, which should further support CSR efforts. Further, marketing experience would also strongly support the view that such CSP would facilitate critical resources for the firm. Thus, adopting the logic of upper echelons theory, marketing experience is likely to promote a mindset that will make a CEO more likely to consistently attend to—and act upon—information about a broad array of the firm’s stakeholders, facilitating CSP. Thus,

H1a

All else equal, there is a positive relationship between CEO marketing experience and CSP.

H1b

All else equal, the relationship between CEO marketing experience and CSP is stronger than other functional experiences (i.e., operations, finance, legal, CSR, general management) and CSP.

Timing of marketing experience

The literature has paid scant attention to how the timing of functional experience may strengthen the functional experience’s influence. We posit that, ceteris paribus, early career marketing experience is likely to be more influential because the executive’s mental map begins to form during this time, and thus the early functional experience may become more foundational in the executive’s mindset. According to schema theory (Alba & Hasher, 1983), earlier information frames and filters the processing of later information (e.g., Day & Nedungadi, 1994; Minsky, 1975), limiting responsiveness to later signals (Hambrick, 1982). Thus, individuals can anchor on their initial representations, guiding future behavior in terms of what they search for and how they interpret what they encounter (Alba & Hasher, 1983), to reinforce the initial mental map. Additionally, early career experience is instrumental in forming the executive’s professional identity (Vakkayil, 2014). As such, such mental maps can be difficult to change, and earlier information is thought to have a stronger effect (Black & Gregersen, 2002).

On the other hand, executives may have marketing experience as one of their last roles prior to being promoted to CEO. Such recent marketing experience may be more influential because this experience is more top of mind (e.g., Kahneman et al., 1993). Such senior marketing experience may also more directly pertain to aspects, such as the firm’s corporate image, for which CSP may play a more crucial role. Alternatively, marketing experience that is neither early nor recent (i.e., inter-medial marketing experience) may align more with the establishment career stage (Cron, 1984), where increasing decision-making authority may add weight to this functional experience and cement it more strongly in the executive’s mental schema. However, given the crucial importance of early knowledge in schema formation and persistence (e.g., Minsky, 1975), we propose that early marketing experience has the strongest impact on CSP.

H2

All else equal, CEO marketing experience timing influences the effect of CEO marketing experience on CSP, such that early CEO marketing experience has a stronger impact.

Situational moderators of the relationship between CEO marketing experience and CSP

Further, upper echelons theory has proposed that situational contexts may moderate the connection between executive experiences and strategic firm choices (Hambrick, 2007). According to Finkelstein et al. (2009, p. 120), chief amongst these are the concepts of executive discretion and job demands.⁴ Executive discretion is the latitude of action and knowledge of the possible courses of action available to the executive. From this discretion-based perspective, when the external environment and internal organization factors confer wide latitude of action (e.g., Kim et al., 2016), executive characteristics are more likely to be reflected in firm choices.

The job demands perspective, on the other hand, suggests that executive characteristics are more likely to be reflected in firm choices based on the “degree to which an executive experience their job as difficult or challenging” (Hambrick et al., 2005). This perspective asserts that executives exhibit traits of bounded rationality, face multiple, at times conflicting goals and courses of action, as well as have varying personal aspirations, all of which combine to make it more difficult for executives to employ strategic rationality in decision making. Thus, as job demands increase, executives instead put more weight on past experiences and cognitive maps resulting from their prior backgrounds (Hambrick & Mason, 1984). However, research has yet to consider the relative worth of these two perspectives.

To test between them, we sought to select factors that would have differing predictions based on whether discretion or job demands were the operant mechanism. In other words, we sought moderating factors (market munificence, prior performance, marketing-aligned assets) which when high (low), would produce high (low) discretion and low (high) job demands, permitting a comparison of the two perspectives.⁵ Additionally, we also examine market uncertainty and CEO compensation, which have also been proposed as key sources of discretion and job demands (Hambrick & Finkelstein, 1987), and because CEO compensation has been found to affect CSP (Luo et al., 2012; McGuire et al., 2019).

External moderators of the CEO marketing experience-CSP link

Market munificence

Market munificence is the abundance of critical resources needed by firms operating within an environment that can support sustained growth (e.g., Feng et al., 2017), and it is positively associated with the range of strategic options available (e.g., Tushman & Anderson, 1986) and thus executive discretion (Hambrick & Finkelstein, 1987).

From this perspective, market munificence provides CEOs a more favorable operating environment, resulting in fewer resource constraints. This, in turn, allows the chief executive a wider latitude of action. Thereby, idiosyncratic executive factors (i.e., experiences) may become more reflected in firm outcomes (Finkelstein & Hambrick, 1990). Favorable environments also allow for reinforcing patterns of relationship building across all stakeholders, furthering the perceived potential for reputational capital benefits. These suggest market munificence should strengthen the connection between CEO marketing experience and CSP.

However, the job demands perspective would suggest the opposite, that the CEO marketing experience-CSP relationship would be stronger when market munificence is low. Less munificent environments are more difficult to operate in, and thus such environments would produce greater job demands (Finkelstein et al., 2009). According to that perspective, as jobs become more challenging, executives are less likely to employ strategic rationality in decision making and instead put more weight on past experiences and cognitive maps, resulting from their prior backgrounds (Hambrick et al., 2005). Thus, the job demands perspective suggests in less munificent environments, executives would be likely to rely on their schemas from their prior marketing experience to a greater extent—strengthening the connection between CEO marketing experience and CSP. Thus, the discretion and job demands perspectives produce competing predictions for how market munificence might affect the marketing experience-CSP relationship:

H3

Market munificence will amplify (weaken) the effect of CEOs with marketing experience on CSP, supporting the discretion (job demands) perspective.

Market uncertainty

Market uncertainty is the inability to predict or foresee the future (Dess & Beard, 1984), due to instability in the firm’s external environment. Such market uncertainty can arise from unpredictable environmental changes or instability in the tastes and preferences of consumers (Nath & Bharadwaj, 2020). Increasing market uncertainty results in a more complex operating environment for the firm, which in turn allows the chief executive a wider latitude of action (Hambrick & Finkelstein, 1987). Ceteris paribus, if conditions for such high discretion exist, executive factors (i.e., functional experience) may become more reflected in organizational outcomes (Finkelstein & Hambrick, 1990). As the requisite attention required for each stakeholder is less clear in uncertain environments, this allows CEO marketing experience to have more of an impact.

Moreover, when there is market uncertainty, executives face an abundance of complex and ambiguous information, increasing their job demands. UET theorists hold that in such weak-information situations without clear means-ends causal connections, executives are likely to put more weight on mental maps developed through their past experiences to make sense of this information (Hambrick et al., 2005). In particular, marketing experience may point to the importance of the firm’s reputation and image, which would be valuable regardless of how consumer trends evolve. Thus, both the job demands and discretion perspectives would suggest that in uncertain environments, CEO marketing experience’s impact on CSP would strengthen.

H4

Market uncertainty will amplify the effect of CEO marketing experience on CSP.

Internal moderators of the CEO marketing experience-CSP link

Prior financial performance

As financial performance improves, discretionary expenditures gain legitimacy. As such, prior literature suggests that positive prior financial performance may have a positive effect on CSP (Waddock & Graves, 1997).

From a discretion perspective, when prior performance is high, as there would be fewer resource constraints on CEO actions than when prior performance is poor, allowing programs and actions in support of more stakeholders. In addition, the positive feedback from such performance can lead executives to rely on the familiar, reinforcing the lens that their experience provides (Dearborn & Simon, 1958). Thus, CEO marketing experience may have a greater impact on the firm’s strategic choices such as CSP as prior firm performance increases.

On the other hand, when prior performance is poor, executives must work hard, searching for solutions to the performance problem (Cyert & March, 1963; Greve, 2003). As such, it is thought that executives in poorly performing firms face considerable pressure and greater job demands (e.g., Hambrick et al., 1993), leading them to rely on their prior experience to a greater extent. Thus from the job demands perspective, poor performance would strengthen the CEO marketing experience-CSP relationship. Therefore, there are competing predictions:

H5

Prior financial performance will amplify (weaken) the effect of CEO marketing experience on CSP, supporting the discretion (job demands) perspective.

CEO compensation

The evidence suggests that long-term compensation—which is where CEOs receive the bulk of their compensation—tends to promote CSP but an emphasis on short-term compensation hinders it (e.g., Deckop et al., 2006). Beyond these direct effects, we consider how total CEO compensation may moderate the influence of CEO functional experience.

UET and empirical findings (e.g., Rajagopalan & Finkelstein, 1992) suggest that CEO compensation reflects the level of strategic decision-making discretion available to top executives. It follows that, if CEOs with greater compensation enjoy higher levels of discretion, then their characteristics will become more reflected in strategy setting (Finkelstein et al., 2009), leading CEO compensation to strengthen CEO marketing experience’s effect on CSP.

Relatedly, logic also suggests that since most CEO compensation is equity-based, the larger the CEO’s compensation, the greater the challenge the CEO faces. Thus, the job demands perspective would similarly suggest that CEO compensation may strengthen the relationship between marketing experience and CSP. Such long-term inducements would be aligned with the longer-term orientation that accrues from marketing experience, fostering these effects. Thus,

H6

CEO compensation will amplify the effect of CEO marketing experience on CSP.

Marketing-aligned assets

Finally, the CEO’s perceived ability to enact preferred strategies is also likely dependent on the alignment between the firm’s resource position and the CEO’s mindset. Thus, when the CEO’s cognitive frame—arising from prior experience—aligns with firm resources, the CEO should perceive greater opportunities to enact their preferred strategies and solutions (i.e., a greater latitude of action), enhancing the CEO’s perceived discretion.

In other words, CEOs with marketing experience may perceive wider latitude of action when the firm possesses assets conducive to producing policies and programs aligned with a marketing experience mindset. Thus, as such aligned resources should enhance the marketing CEO’s perceived discretion and allow greater attention to the CEO’s priorities, greater such resource levels may strengthen the impact of CEO marketing experience on CSP. However, the job demands perspective would suggest an opposite moderating effect. Ceteris paribus, when the firm has lower levels of such resources, the CEO’s job would be harder, leading to greater reliance on the CEO’s background when such resources are low. Thus:

H7

The effect of CEO marketing experience on CSP will strengthen (weaken) when there are greater levels of marketing-aligned assets, supporting the discretion (job demands) perspective.

Data and method

Corporate social performance (CSP)

We collect information on corporate social performance (CSP) from the Kinder, Lydenberg, and Domini (KLD) database (www.msci.com) from 2000 to 2016. The KLD data provides annual independent ratings on corporate social performance measures pertaining to multiple dimensions such as community, human rights, environment, etc. A value of 1 or 0 is assigned under each category by trained independent observers and the scores are added up under each dimension to arrive at a separate ‘strength’ and ‘concern’ score. We provide additional details of the categories tracked by KLD as part of Web Appendix W2.⁶

Following others (Chin et al., 2013; David et al., 2007; Kotchen & Moon, 2012), we first calculate the net score of strengths and concerns of the firm for the fiscal year. The net ratings follow an approximately normal distribution without any transformation. To account for changes in the KLD rating methodology over time, following Kang et al. (2016), we standardize the net scores by subtracting average CSP scores in each year from original CSP scores and then divide by the standard deviation of CSP scores each year. This accounts for the varying number of CSP items over years in KLD and the varying number of firms in the database. We also include time fixed effects in our model to control for this change in KLD items.

Focal independent variable: CEO experience in marketing

We identify CEOs using the BOARDEX database (http://corp.boardex.com). The data includes information on current designations and roles and past positions of executives. We search for ‘CEO’ and ‘Chief Executive Officer’ keywords to identify CEOs. Since we want to build panel data with a firm-year structure, we exclude the deputy, interim, vice-, emeritus, assistant, associate, alternate, division, and regional CEOs.

To identify CEOs with marketing experience, we utilize data from BOARDEX database as well as LinkedIn and Bloomberg.com. We follow prior classification approaches to capture marketing background (Feng et al., 2015; Nath & Mahajan, 2008; Srinivasan & Ramani, 2019). In particular, we use a large set of keywords (including their variants: capitalizations, word-order sequence, and abbreviations) relating to general marketing (e.g., “marketing”, “sales”, “cmo”, and “chief marketing officer”), as well as keywords relating to functional marketing areas (e.g., “advertising”, “brand management”, “customer service”, and “merchandising officer”) to identify CEOs with any such previous experience as a CEO with marketing experience. We measure the years that a CEO has had experience in marketing-related positions prior to being appointed as CEO (i.e., MKTG_EXPR) as well as tenure or the total years of experience in any positions including as CEO (i.e., EXPR). Finally, we calculate marketing experience intensity (i.e., MKTG) using $\frac{\mathrm{ln}(MKTG\_EXPR+1)}{\mathrm{ln}(EXPR+1)}$ formula. This continuous measure captures the intensity or the relative dominance of marketing experience in a CEO’s career and ranges from 0 to 1.

Other functional experience measurement. To test H1b, we identify various other functional backgrounds. Following the approach by Guadalupe et al. (2014), we identify CEOs with general management experience through keywords (and their variants): “general manager”, “division manager”, “head of division”, “general director”, “regional director”, “head of group”, “group manager”, “regional head”, as well as “president” keyword if no additional specific description is provided, and R&D experience through “scientific”, “r&d”, “applications”, and “discovery”. For finance and operations experience, we developed a concurrent keyword analysis, which systematically identified the other terms in their textual background data that most frequently co-occurred with finance and operations. We use “finance”, “cfo”, “accounting”, “controller”, “financial”, and “treasurer” keywords and “operations”, “ops”, “coo”, and “plant” to identify finance and operations backgrounds.⁷ This text analysis approach affirms the keywords that we use for marketing as well. We provide the full list of keywords and more details on this analysis in Web Appendix W3. CSR experience keywords (“responsibility”, “sustainability”, “ESG”, “diversity”, “community”, “social”, “employee relations” and variants) were identified from CSR/Sustainability job postings from BSR.org, a leading networking organization for CSR professionals. For legal experience, we used the terms “lawyer”, “counsel”, “legal”, and “attorney”.

Control variables

We systematically collect a variety of variables to include as controls based on previous literature on the determinants of CSP. By including these in the model, we seek to eliminate omitted variable bias to the extent possible. In particular, we control for three levels of factors: firm, individual executive (CEO), and industry (market) characteristics.

Firm characteristics

Using COMPUSTAT data, we control for recent financial performance, measured as lagged return on assets (ROA), the ratio of income before extraordinary items divided by total assets (ib/at) (Neubaum & Zahra, 2006; Tang et al., 2015); firm size, measured as log of total assets (Manner, 2010; Neubaum & Zahra, 2006; Tang et al., 2015; Waddock & Graves, 1997); for slack resources, we used unabsorbed resource slacks as these are more deployable, measured as retained earnings divided by total assets (re/at) (Zuo et al., 2019); strategic flexibility can also affect the firm’s ability to pursue CSP, therefore, following the approach by Kurt and Hulland (2013), we measure it as the difference between the financial leverage of the focal firm ((ddlt + lct)/at) and that of the industry, excluding the focal firm, and standardized by the standard deviation of financial leverage in the industry.

In addition, we control for a marketing firm-level measure: marketing power – measured as the percent of board members with marketing experience, excluding the CEO (Whitler et al., 2018) – to further delineate the impact of CEO’s marketing experience on CSP. Individual board member marketing experience is determined using the same procedure as described for CEOs.

Furthermore, we capture the facilitating presence of marketing-aligned assets through two measures. First, we use the firm’s number of trademarks owned (from the U.S. Patents and Trademark Office). This is frequently identified as a key marketing resource (e.g., Wiles et al., 2012). Ceteris paribus, more trademarks allow for more latitude of action (Krasnikov et al., 2009). Second, we include the firm’s intangibles intensity, measured as 1 minus the firm’s ratio of physical assets to total assets (e.g., Tuli et al., 2010). Higher values indicate more intangible resources such as reputation, technology, and human capital at the CEO’s disposal.

CEO characteristics

Prior research suggests that CEOs with an output-focused background (i.e., externally oriented backgrounds such as marketing and R&D) can increase CSP (Manner, 2010; Thomas & Simerly, 1994), but this research amalgamated all output-focused backgrounds together. To isolate the impact of marketing experience, we separately account for CEO experience in marketing and R&D, as the mindsets from such experiences may conflict (e.g., Maltz & Kohli, 2000).

We also control for CEO total annual compensation⁸ in thousand dollars from Execucomp (item TDC1) (e.g., Deckop et al., 2006); CEO gender, a dummy variable equal to 1 if the CEO is female (Manner, 2010); CEO duality, a dummy variable equal to 1 if CEO is also a chairman/chairwoman according to BOARDEX (Chin et al., 2013); CEO work experience (e.g., Hitt & Tyler, 1991; Tang et al., 2015), measured as the total number of years a CEO has worked. In addition, we include CEO political donations, following Chin et al. (2013), measured as the difference between total donations (to the candidates, parties, and PACs) to the Democratic party and the Republican party in thousand dollars from opensecrets.org, and we control for the CEO’s career experience breadth, measured by the count of prior functional experiences (Manner, 2010).

Industry (market) characteristics

We control for market-average CSP, measured as the average CSP for all firms in the same 2-digit SIC industry (Chin et al., 2013). We also control for market uncertainty and market munificence as we test for their moderating effects. We follow Keats and Hitt (1988) and regress the total sales of firms operating in a 2-digit SIC industry against the past 5 years to estimate market uncertainty and growth for the sixth year, as follows:

$${\mathrm{SALES}}_{\mathrm{t}}={\mathrm{b}}_{0}+{\mathrm{b}}_{1}+{\mathrm{Year}}_{\left(\mathrm{t}-5:\mathrm{ t}-1\right)+}{\mathrm{e}}_{\mathrm{t},}$$

where SALES_t is the total sales in the industry in year t, Year _{(t-5: t-1)} represents the past five years, and e is the error term. We calculate market uncertainty as the standard error of b1 divided by total sales in the past five years and market munificence as b₁ divided by total sales in the past five years. Finally, using year dummies, we control for time fixed-effects to capture the macroeconomic changes over time.

We match the BOARDEX-LinkedIn-Bloomberg database to the Compustat and Execucomp databases by CIK codes. Due to missing data across some variables, the final data set used for estimation consists of 1,999 firms across 63 2-digit SIC codes and 18,997 firm-year observations, with 9.5 years of data for each firm on average from 2001 to 2016. Our final sample also consists of 3,569 CEOs, 19.98% of whom have marketing experience, 21.6% have finance experience, 43.59% have operations experience, and 70.47% have general management experience. Descriptive statistics are presented in Table 2 and Web Appendix W4, which also provides details on the industries in the sample.

Table 2 Descriptive statistics and correlations

Data analysis approach

To provide evidence for the main effect of marketing experience on CSP (i.e., H1ab and H2), we employ a difference in difference (DID) approach where the presence of a CEO with a particular background in a firm (e.g., marketing) is considered as the treatment variable. We use various firm-level variables (i.e., lagged ROA, firm size, unabsorbed slack, strategic flexibility, marketing power, finance power, operations power,⁹ number of trademarks, and intangibles intensity), industry (2-digit SIC), and year to identify appropriate neighbors (control group) using a propensity score matching technique (Abadie & Imbens, 2006).

In order to test the remaining hypotheses H3-H7, we first start with a simple pooled-OLS analysis. We then use panel data models that account for violations of key assumptions in the pooled model. We use random effects panel model followed by an industry fixed-effects panel model, and finally a dynamic panel model, system GMM. Preliminary tests confirm the presence of heteroscedasticity and serial correlation in the data. In addition, a firm’s recent (i.e., lagged) CSP affects its current CSP performance (David et al., 2007; Kang et al., 2016), therefore we introduce the lag of CSP in the system GMM¹⁰ model to reduce omitted variable bias and reduce serial correlation (Germann et al., 2015; Roodman, 2009; Wooldridge, 2015).

For the dynamic model, we utilize the level and first-differenced equations as a system of equations. In particular, we use the lagged first-differenced realizations of CSP as instruments for the level equation and lagged levels of CSP as instruments for the first-differenced equation in the system GMM estimation. We propose the following model specification:

$${\mathrm{CSP}}_{\mathrm{it}}={\upbeta }_{0}+{\upbeta }_{1} {\mathrm{MKTG}}_{\mathrm{it}}+{\upbeta }_{2} {\mathrm{MKTG}}_{\mathrm{it}}\times {\mathrm{MUN}}_{\mathrm{it}}+{\upbeta }_{3} {\mathrm{MKTG}}_{\mathrm{it}}\times {\mathrm{UNC}}_{\mathrm{it}}+{\upbeta }_{4} {\mathrm{MKTG}}_{\mathrm{it}}\times {\mathrm{LROA}}_{\mathrm{it}}+{\upbeta }_{5} {\mathrm{MKTG}}_{\mathrm{it}}\times {\mathrm{TC}}_{\mathrm{it}}+{\upbeta }_{6} {\mathrm{MKTG}}_{\mathrm{it}}\times {\mathrm{TRMK}}_{\mathrm{it}}+{\upbeta }_{7} {\mathrm{MKTG}}_{\mathrm{it}}\times {\mathrm{IAI}}_{\mathrm{it}}+{\upbeta }_{8-23}\mathrm{ CONT}+{\upbeta }_{24-39}\mathrm{ YEAR}+{\upeta }_{\mathrm{it}}+{\upvarepsilon }_{\mathrm{it}},$$

where η_i is the time-invariant unobservable firm effects, and ε_it is the i.i.d errors capturing the idiosyncratic shocks. MKTG refers to CEO’s marketing experience intensity (measured as total years of marketing experience to total years of experience). UNC, MUN, LROA, TC, TRMK, and IAI represent market uncertainty, market munificence, prior financial performance (lag of ROA), total compensation, number of trademarks, and intangibles intensity. CONT is a vector of control variables, consisted of firm-level variables (i.e., lagged CSP, lagged ROA, firm size, unabsorbed resource slack, strategic flexibility, marketing power, number of trademarks, and intangibles intensity), CEO-level variables (i.e., R&D experience, total compensation, gender, duality, work experience in years, political donations, and career experience breadth), and industry-level variables (i.e., uncertainty, munificence, and average industry-CSP), and YEAR is a vector of time (year) dummies. We also control for endogeneity due to unobserved time-invariant firm heterogeneity by adding the η_i term. Finally, by adding the lagged dependent variable, we control for omitted variable bias due to carry-over effects of the firm’s prior status. We allow for heteroscedasticity at industry level and utilize the robust or sandwich estimator of variance to produce valid standard errors.

Results and discussion

We use the entire sample of CEOs for whom we identified their experiences to test H1ab and H2 (results are presented in Table 3). This data consists of 29,808 observations and 6,147 CEOs. Simple mean comparisons indicate that that CSP is highest among firms that have CEOs with marketing experience (0.147), followed by general management experience (0.04), finance (0.038), and operations (0.012). Results from the DID analysis confirms that CEOs with marketing experience increase CSP (treatment effect = 0.170; p = 0.000), supporting H1a.

Table 3 DID analysis

To test H1b, we run multiple DID models with identical settings to examine the effect of the appointment of CEOs with other experiences. We find that the only statistically significant results are CEOs with general management (treatment effect = 0.119; p = 0.000)¹¹ and legal experience (treatment effect = -0.131; p = 0.030). The propensity score matching for CEOs with R&D experience fails to find any neighbors within an acceptable distance, thus we do not report the results of DID for R&D experience. Overall, our analysis reveals that indeed the effect of CEOs with marketing experience is stronger than those of CEOs with other experiences, supporting H1b.¹² We provide more details in Table 3 panel A.

We classified the timing of marketing experience according to whether the CEO had a marketing position in the early years of their career, in the last years prior to becoming CEO, or sometime in the middle of their career. We use various cut-off points (2–5 years) to build these groups. For example, using the 3-year cut-off point to assess early and recent experience, we find that 73.24% of marketing CEOs had experience in marketing in their first 3 years, 36.32% have experience in their 3 years prior to becoming CEO, and 49.67% fall into the inter-medial group (experience in-between the early and recent cut-offs). Using the same settings in H1ab, we utilize the propensity score matching method to find the treatment group among other CEOs with marketing experience who do not have early (inter-medial; recent) marketing experience. Providing support to H2, results reveal that starting one’s career in marketing (treatment effect = 0.184; p = 0.000) has the strongest impact on CSP outcomes. Inter-medial CEO marketing experience has a moderate effect (treatment effect = 0.102; p = 0.011), however, recent marketing experience has no additional effect than any marketing experience (treatment effect = 0.014; p = 0.744). We find similar results when we use various cut-off points (2-, 4-, and 5-year) in defining these categories. See Table 3 panel B for further details.

Moreover, since there are overlaps among these three groups, as a stronger test, we repeat this analysis by focusing on CEOs who fall into only one of these groups (e.g., only have early marketing experience). In particular, we run three DID models where only early, only inter-medial, and only recent marketing experience are respectively considered treatment. We also use the number of years of marketing experience as an additional condition in the propensity score matching to disentangle the effect of timing of experience from the length of experience. We find that the effect of only early marketing experience is positive and significant (0.169; p = 0.019) while the effects of only inter-medial and only recent marketing experience are not significant. This provides further evidence for H2 that early marketing experience has the strongest impact.

In contrast to the DID analysis, we use a continuous measure of the independent variable, CEO marketing experience intensity, to test the remaining hypotheses. Following Aiken et al. (1991), we de-mean covariates before generating the interaction terms. Variance inflation statistics (VIF) suggest that multicollinearity is not a concern (VIF values are from 1.02 to 5.29, mean of 2.32). We present the results in Table 4. The results for marketing experience are very consistent across the pooled-OLS, random effects, industry fixed-effects, and system GMM models. However, since controlling for the lag of the dependent variable is critical for reducing omitted variable bias and reverse causality, we focus on the system GMM estimation results.

Table 4 Results

The Hansen test of overidentification (J test) indicates that the moment conditions are valid, and our specification is not over-identified. To check for first-order serial correlation in levels, we look for second-order correlation in differences. The AR(2) test indicates that the second-order lags can be used as instrumental variables (p > 0.36). To avoid overidentification and biases associated with “too many instruments”, we limit the number of instrumental variables to only fifth-order lags in our main estimation (Roodman, 2009). As we discuss later, we also examine our results using different settings of lag structure and find our results to be robust. Moreover, we use the difference-in-Hansen test to examine whether the instruments of the model (i.e., GMM type and standard instruments in both first-differenced and level equations) are de facto exogenous. The difference-in-Hansen test is preferred to the Durbin-Wu-Hausman test because it can report test statistics that are robust to various violations of conditional homoscedasticity (Baum et al., 2003). The results from the difference-in-Hansen test of endogeneity reveals that the instruments are exogenous (p = 0.99).

Our results reveal that CEO marketing experience enhances firms’ CSP (0.088; p = 0.001), providing further support for hypothesis 1. Hypothesis 3 suggests that as market munificence increases, the effect of CEO marketing experience on CSP gets stronger or weaker, depending on whether discretion or job demands is salient. The interaction term between CEO marketing experience and market munificence is positive but not statistically significant (0.264; p = 0.520). Further, hypothesis 4 predicts that market uncertainty would amplify the positive relationship between the CEO’s marketing experience and CSP. The results suggest a significant positive interaction effect (5.253; p = 0.003), in support of H4.

Further, for hypothesis 5, the interaction between CEO marketing experience and prior performance is not statistically significant (-0.000; p = 0.965). We find support for hypothesis 6 as the interaction term between CEO marketing experience and CEO total compensation is positive and significant (0.013; p = 0.007). Thus, we reveal that there are interactive effects between executive compensation and executive characteristics on strategic outcomes. Finally, we find that the interaction of CEO marketing experience and intangibles intensity is significant (0.175, p = 0.034), and the interaction between CEO marketing experience and number of trademarks is significant in all the models except system GMM. These results indicate that marketing-aligned assets strengthen the positive impact of CEO marketing experience on CSP, in support of H7.

In summary, the results support the notion that CEO marketing experience delivers higher annual CSP scores even as we control for a wide array of confounding variables. That we find the effect of CEO marketing experience, even after controlling for the board’s marketing experience, strongly indicates that our results are due to an organic effect through the CEO, separate from any board influence. Further, this effect is stronger in uncertain environments, when the CEO receives more compensation, and when marketing-aligned assets are strong, providing support for the discretion-based perspective.

Robustness checks

Alternative identification strategies

We used a variety of alternative identification approaches to assess the effect of marketing experience intensity on CSP. Table 5 shows the results of these analyses. The results show that the main effect is robust to alternative specifications and further corroborate our findings of the DID analysis.

Table 5 Various identification strategies

First, we explore an instrument-free approach that exploits the structure of the joint-distribution of the endogenous regressor and the error term in a regression model. When the endogenous regressor can be shown to follow a non-normal distribution, the error term and the regressor can be modeled as following a joint-distribution specified by a Gaussian Copula to account for correlation and thus avoid bias owing to endogeneity as proposed by Park and Gupta (2012). The Shapiro–Wilk test confirms that our potentially endogenous regressor is nonnormally distributed (W = 0.95, p < 0.001), as required for identification. Specifically, we calculate a copula term for marketing experience intensity. This term is specified as MKTG_copula_it = Φ⁻¹(H(MKTG_it)) where Φ⁻¹ is the inverse of the normal cumulative function, and H(MKTG_it) represents the empirical distribution function of marketing experience intensity. That is the probability mass of observing a value less or equal to MKTG_it. To estimate the parameters of this model, we use bootstrapping with 1,000 replications. The MKTG_copula is the copula control function term, and it controls for the correlation between the error term and the potentially endogenous regressor (i.e., MKTG). The results of including the copula term in our proposed model are consistent with our main results and support the proposed positive impact of marketing experience intensity on CSP (0.097; p = 0.005).

Second, we explored identification using candidate instrumental variables in a two-stage least squares framework. In particular, we use two different instrumental variables: the board’s marketing experience and a peer-based instrument. Boards that consist of members with marketing experience should be more prone to appoint CEOs who also have marketing experience. Although the appointment of a CEO is a board prerogative, the day-to-day management of the firm is primarily vested on the CEO thus insulating it from influencing executive actions that influence CSP. The correlation between CEO marketing experience intensity and board marketing experience is 0.14. The effect of marketing experience intensity is statistically significant (1.69; p = 0.003).

Similarly, firms that operate in the same industry influence each other’s decisions because managers tend to resolve uncertainty in strategic choices by mimicking their peers (Spender, 1989). Therefore, if a firm operates in an industry where many firms have CEOs with marketing experience, it should be more likely that the firm also appoints a CEO with marketing experience. However, in order to avoid the peer-effects problem that threatens identification as articulated by Angrist (2014), we construct peers within the same industry that only partially overlap with each other so as to create sufficient variation in the instrumental variable. Adapting Lim et al. (2020), within the same industry, we consider another firm as a peer only when the difference between their sizes is less than one standard deviation of firm sizes in that industry. This results in firms having overlapping unique sets of peers that are all subgroups of their own industry. The average size of these groups to the average size of industries is about 0.52. The correlation between marketing experience intensity and the instrumental variable is 0.12, and the effect of marketing experience intensity is statistically significant (3.56; p = 0.000).

Third, we also employ a two-step control function approach as has been frequently used in the marketing literature (e.g., Petrin & Train, 2010). In the first stage, we estimate the correction term by regressing marketing experience intensity on a set of exogenous variables. We use board’s marketing experience and the estimated peer-based instrumental variables as they should satisfy the exclusion restriction condition.¹³ We incorporate the correction term in the second step in a system GMM model and find that the effect remains consistent (2.34; p = 0.016). These analyses provide strong additional support for CEO marketing experience’s CSP impact.

Other ways of measuring the marketing experience of CEOs

We re-estimate our model using three alternative measures of the CEOs’ marketing experience: a binary variable (1 if a CEO has marketing experience; 0 otherwise), number of years that CEOs have worked on marketing-related jobs (ln of MKTG_EXP, as it is heavily skewed), and marketing dominant, measured in a binary form (1 if marketing experience intensity is greater than or equal to 0.5; 0 otherwise). Results remain substantively unchanged, (Web Appendix W5, Table A, B, C), suggesting our findings are not sensitive to the type of measurement of marketing experience.

Other lag structures in the system GMM model specification

We run several models using different numbers of instrumental variables by changing the number of lagged endogenous variables; therefore, we limit the depth and thus the number of lags, beginning from the second-order lags. In addition, we run a lag-restriction-free model in which we use all the available lags as instrumental variables. The results (Web Appendix W5, Table D) are substantively similar, suggesting our findings are not sensitive to the setting of lag structure.

Alternative tests for the effect of timing of marketing experience

Instead of using a DID model, we introduce three dummy variables – accounting for early, inter-medial, and recent marketing experience – in our models, while we also control for the number of years of marketing experience. We continue to find that early marketing experience is more important in increasing CSP (Web Appendix W5, Table E).

Controlling for additional factors

We repeat our analysis including measures for finance and operations power as well as a variable to indicate whether they have earned a master’s degree in business administration (MBA), as some evidence suggests that this can enhance aspects of CSP (Manner, 2010). Results are not affected (Web Appendix W5, Table F).

Accounting for outliers

We winsorize the marketing intensity, CSP, and all the moderators at the 1% level. The results (Web Appendix W5, Table G) are substantively consistent with our original analysis; therefore, the results are independent of outliers.

Supplemental analyses

Alternative considerations of CSP

We repeat our difference-in-difference analysis for the effect of CEO marketing experience on alternative formulations of CSP: considering strengths only, concerns only, and net score and strengths only formulations of institutional and technical CSP (Groening & Kanuri, 2018) (see in Web Appendix W5, Table H). Findings for strengths replicate H1 (treatment effect = 0.137; p < 0.001). Further, we observe strong effects for marketing experience for institutional CSP (strengths, treatment effect = 0.140; p < 0.001 and technical CSP (strengths, treatment effect = 0.095; p < 0.001) and similar results for using net scores, indicating that marketing experience has a broad-based effect on CSP. Further, to test whether our findings are derived from a particular dimension of CSP, we also exclude dimensions of CSP one at a time and measure CSP by the other five dimensions. We find that results are very consistent across all the models, and marketing experience intensity always remains statistically significant.

Impact of the rank of CEO marketing experience

We also examine the effects of the hierarchical rank of marketing experience, motivated by Feng et al.’s (2015) classification. As such, we classify positions such as chief marketing officer (CMO), president, and global marketing as top-ranked; positions such as director of marketing, marketing manager, brand manager, etc. as mid-ranked positions; and the remaining positions such as sales representative, marketing analyst, as low-ranked. CEOs could fall into multiple groups. Findings indicate that lower rank marketing experience is more associated with CSP (see Web Appendix W5, Table I). This complements our prior analysis that showed the benefits of earlier marketing experience.

Implications and limitations

In this inquiry into the theoretical and empirical significance of CEO marketing experience, we explored its influence on CSP and delineated possible boundary conditions. Using difference-in-difference analysis, we provide strong evidence for the relationship between CEO marketing experience and CSP. This relationship is stronger for early career marketing experience, when there is market uncertainty, when the CEO’s total compensation is higher, and when aligned resources provide more marketing-related discretion. We now discuss the theoretical and practical implications of our research, as well as its limitations.

Theoretical implications

For the marketing literature, our findings enrich our understanding of the consequences of marketing experience by showing how CEO marketing experience can shape strategic firm choices. Specifically, we provide empirical evidence that suggests a robust relationship between CEO marketing experience and CSP, and this effect is stronger than any other functional background experience. Thus, marketing experience appears to have the strongest degree of influence over CSP, due to its emphasis towards a wider variety of stakeholders, and due to its greater recognition of, and appreciation for, the resources that this can generate. These findings enrich our understanding of CSP determinants and support the modern understanding of the marketing function, which suggests an increasing role for marketers as guardians and promoters of stakeholder voices (Hult et al., 2011; Morgan, 2012).

In addition, we draw out the natural synergies between UET and stakeholder theory—which can be linked through CEO marketing experience, and we identify conditions that moderate CEO marketing experience’s impact on CSP. In particular, we reveal that earlier marketing experience has the strongest impact. This offers new insight for marketing scholars, and for upper echelons theory, on how functional experience timing impacts strategic firm outcomes, suggesting the crucial importance of early, formative experience. However, those who only have late marketing experience produce more CSP than CEOs without marketing experience (indicating that late experience still affects the executives’ mindset). Nonetheless, the relatively weak effect for those with only late experience raises the question of whether these individuals would consider themselves marketers, an issue that warrants future research.

We also enrich upper echelons theory by being the first to empirically investigate how specific external environment and internal organization factors can moderate the degree to which CEO functional experiences influence strategic firm actions (here, in the CSP domain). Moreover, by testing these moderators, we shed light on the proposed process (Spencer et al., 2005). While results for market uncertainty and CEO compensation are consistent with both the discretion and job demands perspective, the positive moderation observed for the marketing-aligned factors—number of trademarks, intangible asset intensity—point to the superiority of the discretion explanation, providing new insight for UET theory. Further, that we observe results for the marketing-aligned factors, but not market munificence or prior performance, offers new insight into discretion, suggesting the crucial role of CEO mental schema-firm resource alignment in affecting perceived latitude of action.

For the CSP literature, our robust model provides new insight on factors that lead to CSP. We demonstrate that CEOs with marketing experience increase CSP and that market uncertainty, CEO compensation, and mental schema aligned resources amplify this relationship. Thereby, we provide strong support for executive discretion as a driver of CSP. In addition, our logic for the importance of marketing experience on CSP could also be extrapolated to other marketing-related dimensions. In line with this, we find support for the positive impact of other structural marketing factors (i.e., marketing power) in our results. This adds to our understanding of marketing’s role and the internal antecedents of CSP, particularly at the top management level.

Managerial implications

Our finding that the marketing experience of a CEO, compared to other CEO functional experience, helps firms enhance their CSP should be of particular relevance to top executives and boards of directors. The impact of marketing experience is managerially significant as it increases CSP by 41.38%.¹⁴ First, we hope it will encourage firms to pay closer attention to the potential benefits of marketing experience when evaluating CEO candidates—particularly as CSP continues to emerge as an important firm performance dimension in its own right. Further, marketing experience enhances both technical and institutional CSP, indicating marketing leads to a broad-based effect across the CSP dimensions and groups and that it does not have a harmful effect on particular CSP dimensions.

Second, results indicate firms interested in CSP should design their leadership pathways to emphasize marketing experience, particularly in earlier roles. Additionally, increasing CEO compensation can amplify the impact of prior CEO experience in marketing. Further, greater investment in marketing-aligned resources can also strengthen the positive impact of CEO marketing experience on CSP. We observe this positive impact for number of trademarks and intangible asset intensity, but logic suggests that this effect would extend to other marketing-related investments as well. These provide actionable guidance in terms of structuring CEO candidate evaluation, promotion pathways, and complementary resource investment for CSP.

Third, findings thus suggest that CEOs without marketing experience may need to be informed of their CSP blind spot. Firms also may need to intervene with marketing experience CEOs when the CEO’s perceived latitude of action may be lower due to the lack of marketing-related intangible resources, when there is lower CEO compensation, and in less uncertain environments—as these reduce CEO marketing experience’s benefits on CSP.

For socially-conscious investors and regulators, our findings, based on the largest sample to date, offer new guidance for what conditions facilitate CSP. Findings confirm that firm size is strongly associated with CSP, as this increases exposure to stakeholder scrutiny (Brower & Mahajan, 2013), consistent with stakeholder theory. Consistent with UET, findings reinforce the positive impact of female CEOs on CSP, but we observe no effect for CEO duality or career breadth, while liberal political donations and years of work experience tend to show a negative impact across the models. Supporting the slack resources explanation (e.g., Waddock & Graves, 1997), results indicate strong effects for prior performance and slack resources. Further, the effect of market-average CSP also suggests that institutional pressures also can contribute to CSP (e.g., Brower & Dacin, 2020).

Limitations and future research

There are several limitations of this study that should be considered in combination with our findings. First, our focus on CEOs exclusively may mask the group-based (e.g., TMT) decision making in some corporations. Examining the functional background experiences of the entire top management team and board of directors may provide complete information about the CSP impact of functional backgrounds at the upper echelons of the organization. However, our results suggest that CEO functional experience has a predominant impact. Further, CEO marketing experience influences CSP, even when controlling for board marketing experience.

Second, we do not empirically test the actual process through which CEO experiences translate into business decisions in the CSP domain. We have based our supposition that the attentional processes exist based on solid prior theoretical foundations. As such corroborating evidence would be mostly internal to the firm (e.g., meeting deliberations and notes), it would be difficult to provide further empirical evidence for it, as noted by prior researchers in this space (e.g., Whitler et al., 2018). Other proxies for CEO attention (i.e., shareholder letters) are likely less suitable in this situation, as such letters often reflect impression management concerns and may not accurately reflect CSP (Wickert et al., 2016). Future research may be able to utilize experimental approaches to assess these pathways and connections. Relatedly, our support for discretion’s role relies on testing this process through moderation. As such discretion perceptions would be internal to the CEO, they also would be difficult to access through historical or secondary sources, and a traditional mediation process is unlikely to be possible. Further research may delve deeper into the notion of discretion through experiments or CEO surveys.

Third, there may be particular alignment between the social performance dimensions KLD tracks and the lens that marketing experience provides regarding which stakeholders to focus on. While we find positive effects for marketing experience on both institutional and technical CSP, indicating a broad effect, our findings are limited to KLD’s dimensions of social performance. Future research may be needed if new dimensions of social performance emerge.

Fourth, extant theory holds that marketing’s support for a wide variety of stakeholders can be motivated by both intrinsic and instrumental considerations. Future research should disentangle which of these perspectives is stronger. Our stronger effects for institutional CSP provide some suggestive evidence that the intrinsic pathway may be more salient.

Finally, our research focuses on public North American based corporations. As such, the results may not extrapolate to smaller, privately held companies, or in international contexts.