Introduction

Once thought of as solely profit-maximizing entities that cater to the interests of shareholders, organizations in today’s complex business market must attend to a wide variety of stakeholders in order to be successful. Beyond wealth generation obligations, external stakeholders increasingly expect companies to be active corporate citizens through corporate social responsibility (CSR) initiatives (Aguilera et al. 2007). In return, research shows that stakeholders significantly reward companies that engage in CSR (Brønn and Vrioni 2001; Jones et al. 2014; Orlitzky et al. 2003). These considerations have made CSR an important item on most strategic agendas (Skarmeas and Leonidous, 2013). It is therefore not surprising that researchers have observed a dramatic rise in the number of companies engaging in CSR in the past few decades (Ruef and Scott 1998; BSR/Cone 2008).

To explain the effectiveness of CSR, researchers have drawn extensively on signaling theory (Connelly et al. 2011; Spence 1973). This theory argues that organizations (signalers) use CSR to communicate positive organizational qualities to stakeholders (receivers); for instance, that the company is trustworthy and genuinely concerned about the welfare of society (Connelly et al. 2011). In turn, stakeholders use signal information to make important decisions (e.g., investment and product purchase decisions) that positively impact the organization (Bergh and Gibbons 2011; Walsh et al. 2009). However, recent research indicates that stakeholders are becoming increasingly aware that while some companies engage in CSR out of genuine goodwill, others pursue CSR initiatives purely to secure the associated economic benefits (Skarmeas and Leonidous 2013; Vanhamme and Grobben 2009). Thus, CSR signals are becoming increasingly ambiguous because the positive messages that organizations wish to convey through their CSR initiatives are no longer being automatically interpreted as such by stakeholders. This has made CSR a potentially unreliable source of information for stakeholder decision making. In light of these issues, the goal of this study is to examine the psychological processes that unfold when stakeholders receive alternate signals that contradict the intended positive CSR signals of an organization. We integrate signaling theory and attribution theory to explore whether additional cues related to a CEO’s ethics—cues that either support or contract CSR signals—impact stakeholders’ decisions to personally support that company’s CSR. We further explore two psychological mechanisms, stakeholder CSR attributions and stakeholder cynicism, to better understand how CEO ethics signals ultimately impact stakeholders’ decisions to support CSR.

This study is motivated by a number of theoretical issues. First, Spence (2002) argues that organizations send many signals on a regular basis, often without awareness or intent. Stakeholders receive these unintended signals (along with intentional signals such as CSR) and must determine their value to decision making. In their review, however, Connelly et al. (2011) noted that there has been little theoretical and empirical development on the impact of such signals—particularly those that (a) are unknowingly transmitted, (b) conflict with intentional signals, and/or (c) carry negative information about the signaler. Thus, there is a need for better theoretical understanding of how stakeholders process these signals, especially since stakeholders have become increasingly skeptical of CSR (Skarmeas and Leonidous 2013). Integrating fundamentals of attribution theory (e.g., Kelley and Michela 1980) with signaling theory, we propose that when stakeholders receive additional signals that conflict with intentional signals, they enact an attribution process to re-evaluate the motives behind the intended signals (Bolino et al. 2008; Mills 2003). In particular, we suggest that alternate signals from top executives play an important role in stakeholders’ interpretations of CSR signals. Although there has been some research on factors that influence stakeholder’s perceptions of CSR motives (e.g., Ellen et al. 2006; Groza et al. 2011), the influence of top executives has received little attention to date. This is surprising because CEOs may be key signalers given that they are often featured in the media when the company does well, and similarly targeted for blame and retribution when the organization misbehaves (Jory et al. 2015). We know little about how signals about CEO ethics, as portrayed in the media, provide stakeholders with supportive or contradictory cues when forming attributions of why an organization engages in CSR. In this study, we test the impact of public media exposure to the morally questionable ethics of a CEO (vs. an ethical or “ethics-unknown” CEO) on stakeholder support for the company’s CSR.

Furthermore, we extend the limited literature by exploring psychological mechanisms through which CEO ethics affects stakeholder attitudes and behaviors toward the company’s CSR. We complement signaling theory’s macro-perspective on how organizations send CSR cues to outsiders with a micro-foundations framework (Devinney 2013) that explores the individual-level psychological processes that occur when the receiver acquires the company’s CSR message. Specifically, we draw on attribution theory to explore stakeholder attribution cognitions and subsequent cynicism that ensue after they are presented with information about CEO ethics that contradicts the company’s positive CSR message. We propose that these cognitions indirectly explain the impact of CEO ethics on stakeholder CSR support intentions and behaviors.

Our study makes important contributions to the literature. First, we contribute to signaling theory by examining how ambiguous signals (i.e., CSR) are interpreted and processed by receivers from an attributional theory perspective. CSR initiatives are ambiguous signals in terms of their underlying motives (moral vs. instrumental). We illustrate how stakeholders interpret these equivocal signals positively when additional cues from CEOs are genuine (i.e., ethical CEOs). In contrast, equivocal CSR signals are likely to be negatively interpreted when cues from CEOs (i.e., morally questionable CEOs) are inconsistent with the company’s CSR signal. An understanding of this process is important given the possibility that ambiguous signals may not always elicit the positive stakeholder reactions desired by the signaler.

Second, we extend the limited research on the antecedents of CSR motive attributions by examining the role of CEO ethics. Specifically, we test the impact of the ethics of a CEO on attributions of CSR motives, and subsequent effects on public support for the company’s CSR. To date, research on the antecedents of CSR motive attributions have focused primarily on the characteristics of the CSR messaging strategy, as well as those of the recipient of the message (e.g., Groza et al. 2011). As such, we extend this literature by considering how internal actors (i.e., CEOs) within an organization indirectly shape external stakeholder evaluations of CSR motives. Moreover, our study expands on the literature (e.g., Becker-Olsen et al. 2006) by exploring psychological mechanisms through which CEO ethics indirectly affect stakeholder attitudes and behaviors toward the company’s CSR. Specifically, we explore the mediating role of organizational cynicism in understanding how CSR motive attributions ultimately impact CSR support. We argue that stakeholder’s CSR motive attributions affect stakeholder cognitions (i.e., cynicism about the organization), and this has important implications for understanding stakeholder willingness to support a company’s CSR engagement.

Our theoretical model is summarized below (Fig. 1). In this model, we link morally questionable (versus ethical or ethics-unknown) CEOs to stakeholder attributions (i.e., instrumental CSR motive attributions versus moral CSR motive attributions), which subsequently influence organizational cynicism. In turn, we expect that organizational cynicism results in lower: (a) intentions to personally volunteer for the CSR; (b) financial donation behaviors in support of the CSR, and (c) product purchase intentions (a passive form of CSR support).

Fig. 1
figure 1

Theoretical model

Full size image

Theoretical Background

Signaling Theory

Signaling theory explains how organizations communicate to stakeholders using signals which deliver positive information about a company’s actions (Akerlof 1970). This theory suggests that organizations routinely send signals aimed at communicating their quality or character. In turn, these signals should influence stakeholders’ perceptions and behaviors toward the organization (e.g., Becker-Olsen et al. 2006). Several scholars conceptualize CSR as an important signal of a company’s quality and intentions in the marketplace (Branco and Rodrigues 2006; Turban and Greening 1997) and may be viewed as a strategic form of communication used by the company (e.g., Godfrey et al. 2009). Indeed, a plethora of research shows that organizations with positive CSR activities yield benefits in the form of better brand images, greater stakeholder loyalty, positive word-of-mouth advertising, and ultimately, increased financial performance (e.g., Brønn and Vrioni 2001; Orlitzky et al. 2003). However, since signal senders (i.e., an organization) and signal receivers (i.e., stakeholders) do not always have the same interests, a company may have an incentive to send false signals that do not actually reflect its motives (Connelly et al. 2011). This is problematic because information asymmetries—uneven access to information pertinent to signal interpretation—exist between signal sender and signal receivers, such as when companies withhold “insider” information about an initiative from stakeholders (Connelly et al. 2011). Thus, higher information asymmetries may lead to ambiguous signals that can be interpreted either positively or negatively by the receiver. We suggest that signal receivers process and interpret ambiguous CSR signals through an important psychological attribution process.

Attribution Theory

Although multifaceted in scope and application, the underlying theme of attribution theory (Kelley and Michela 1980) is that individuals (perceivers) observe and form judgements or interpretations of the cause of another’s behaviors (Kelley and Michela 1980). These interpretations, in turn, impact the perceiver’s subsequent attitudes and behaviors. This general principle aligns with signaling theory to the extent that stakeholders (as perceivers) routinely interpret the cause of a company’s actions that are encoded in the signals it sends. Upon receipt, stakeholders must decipher the meaning and intent of the organization’s signals, and in the process, critically evaluate the motive or cause of the company’s action (Love and Kraatz 2009). As such, when companies advertise CSR initiatives, stakeholders (e.g., the public) often consider the motive behind the initiative in an attempt to evaluate the quality or characteristics of the organization (Prabhu and Stewart 2001).

In today’s economic and social milieu, however, stakeholders’ normal attribution process is likely to be convoluted when evaluating a company’s CSR-related actions. Although companies typically aim to send positive signals about their character, noise or contextual differences may disrupt the receiver’s interpretation of the company’s signals (Connelly et al. 2011). Such signal disruption is likely in the CSR context because of increased corporate scandals that have heightened public mistrust of corporations in general, especially those corporate activities purported to be borne out of goodwill (Skarmeas and Leonidous 2013). Therefore, a company’s intentions in sending positive CSR signals may be different from the attributions stakeholders make upon receiving the signals (Brown et al. 2006). In other words, a mismatch may occur between a company’s intent to send a positive signal (i.e., by advertising its CSR initiative) and the receiver’s ultimate interpretation of the signal (i.e., CSR motives attributions).

Scholars have identified two broad categories of attributions regarding corporate motives behind CSR. An organization may have genuine intentions to contribute to the welfare of society through philanthropic activities beyond what is legally required (Carroll 1999). This is often described in the literature as a moral motive attribution for CSR (Aguilera et al. 2007), although other terms have also been used (e.g., other-serving or values-driven motives; Ellen et al. 2006). In contrast, organizations may also pursue CSR agendas purely to secure the economic benefits associated with perceived corporate goodwill. Such instrumental motive attributions (i.e., self-serving or egoistic motives) have also been documented (Aguilera et al. 2007). Research shows that stakeholders are able to distinguish between these “doing good” versus “doing well” motives, and this has implications for how they react to the organization (Skarmeas and Leonidou 2013). Furthermore, research shows that stakeholders reward organizations judged to have moral motives and punish those whose CSR involvement appears to be solely instrumentally motivated (see Ellen et al. 2006; Skarmeas and Leonidous 2013).

We theorize that the ambiguity associated with CSR signals sent by a company activates a stakeholder attribution process aimed at uncovering the underlying motive behind the CSR. According to Kelley and Michela (1980), a key aspect of the attribution process is the availability of relevant information to evaluate the signal sender’s behavior. This suggests that stakeholders are likely to search for other sources of information that provide reliable insights into the motive behind the company’s CSR signal. However, given the informational asymmetry that exists between companies and the public—i.e., companies generally tend to keep inauspicious information from the public (Connelly et al. 2011)—stakeholders rarely have direct access to unfiltered information specific to CSR initiatives. Thus, stakeholders are likely to seek and interpret additional signals about the company during the attribution process in order to determine how they will ultimately react to the company’s CSR. Below, we explore the role of the CEO as an important signaler of pertinent information that stakeholders may consider in forming attributions about the company’s CSR signals.

Hypothesis Development

CEOs as Signal Senders

Signals conveyed through the actions of CEOs are likely to offer unique and potentially important insights that may shape stakeholders’ CSR motive attributions. Our focus on CEOs is supported by research showing that they significantly impact the propensity and motivation of their companies to engage in CSR (Manner 2010; Oh et al. 2016; Waldman et al. 2006) and, therefore, should be considered key signal-sending actors. Yet, we know little about the extent to which CEOs shape outsider attributions of CSR motives. This is surprising because stakeholders now have greater exposure to corporate CEOs through traditional and social media platforms.

We theorize that stakeholders may use information about the ethicality of an executive leader—observed through the leader’s attitudes and actions in the public domain—in forming attributions about the company’s CSR motives. This is likely because stakeholders look to CEO’s as those who genuinely “walk the talk” and articulate the motivation of their organization. That is, the ethics of CEOs convey important information about the character of the organization as a whole (Hogg 2001). Stakeholders are likely to develop negative attributions about an organization’s CSR motives if its CEO’s ethics are deemed morally questionable—as is often inadvertently discerned during publicized speeches or through action/inaction on polarizing topics or events (Ogunfowora 2014). Indeed, the extant research to date shows that CEOs’ unethical misconduct is commonplace, and can have a negative impact on a company’s performance when information about the misconduct becomes public (Jory et al. 2015).

We therefore propose that publicity about a CEO’s morally questionable ethics serves as an additional cue/signal that contradicts the positive signal the company intends to send regarding its CSR activities. For example, In Abercrombie & Fitch’s CSR statement, the company declares its support for human rights, diversity, communities, environmental sustainability, and responsible business decisions. However, the veracity of these claims were challenged when the CEO, Mike Jeffries, created a global publicity crisis by excluding larger sized customers and catering only to “thin and beautiful” people. Similarly, Tim Hortons, a coffee company, intended to send a positive signal by announcing its involvement in different CSR initiatives, including community and environmental responsibility in its sustainability report (Tim Hortons 2014). However, media coverage of the CEO’s subsequent decision to cut the budget for many of these corporate responsibility initiatives (Crane and Matten 2015) likely served as an additional, contradictory signal that may shape stakeholder attributions of the company’s CSR motives.

As such, stakeholders may be more likely to attribute purely instrumental motives to a company’s CSR initiative, despite the company’s best intentions to send a positive message of its goodwill. In particular, we argue that stakeholders are likely to conclude that a morally questionable CEO’s unethical intent in one arena carries over to other arenas, including the establishment of a CSR initiative. As an example of this carry-over effect, Jory and colleagues (2015) argue that investors tend to believe that when unethical executives run a company, it will likely have difficulties preserving its wealth, thereby discouraging investors from investing in it. Stakeholders may further form attributions that the CSR initiative of a company run by an unethical CEO is based on self-serving motives, such as to bolster the CEO’s image, the company’s reputation, and to improve the firm’s (and thus, the CEO’s) financial success. Indeed, there is some evidence to suggest that some executives do promote CSR for the sake of their own benefit (e.g., Cai et al. 2012). Such CEOs engage in CSR for the economic benefits, even when doing so is a disingenuous attempt at “window dressing”—attempts to deceive stakeholders in hopes of establishing a (false) sense of legitimacy (Cai et al. 2012). Thus, stakeholder attributions of instrumental CSR motives in reaction to a morally questionable CEO may, in some cases, be a valid indicator of the organization’s character as a whole.

In contrast, if a CEO’s morality is shown to be ethical in the media, this signal conveys information that is consistent with the “goodwill” signal the company aims to portray through its CSR activities. Thus, contrary to the case of the morally questionable CEO, stakeholders should ascribe moral CSR motive attributions in reaction to the ethical CEO. Stakeholders may conclude that ethical CEOs are mainly driven by a sense of stewardship (Davis et al. 1997) and responsibility (Bansal and Roth 2000), and that they support CSR because of their moral principles and ethical standards (Aguilera et al. 2007, p. 842). Moreover, stakeholders are likely to believe that the ethical CEO’s company will pursue its CSR agenda even when there is no clear economic benefit to doing so. Indeed, existing studies suggest that stakeholders associate ethical CEO behaviors with social responsibility (De Hoogh and Den Hartog 2008; Waldman et al. 2006). Thus, we propose the following hypotheses:

Hypothesis 1a

Stakeholders will attribute significantly greater instrumental CSR motives when exposed to a firm headed by a morally questionable CEO, compared to that of an ethical or ethics-unknown CEO.

Hypothesis 1b

Stakeholders will attribute significantly less moral CSR motives when exposed to a firm headed by a morally questionable CEO, compared to that of an ethical or ethics-unknown CEO.

Consequences of Stakeholder CSR Motive Attributions

Attribution theory suggests that once stakeholders cognitively assign motives for an organization’s CSR involvement, these attributions significantly influence their ensuing attitudes and behaviors (Kelley and Michela 1980; Skarmeas and Leonidou 2013). In particular, CSR motive attributions are likely to engender a series of cognitive, affective, and behavioral responses in stakeholders. First, we propose that CSR motive attributions activate cognitive-based evaluations of the trustworthiness of the organization and its CSR initiative. In line with extensive research showing that stakeholders have become increasingly skeptical about CSR (Foreh and Grier 2003), we anticipate that CSR motive attributions will most proximally impact organizational cynicism toward the organization. Stakeholders who—following exposure to the morally questionable ethics of a CEO—attribute instrumental motives to the company’s CSR should experience increased skepticism about the intent and trustworthiness of the organization. This is because they are likely to infer from the morally questionable CEO’s character that the company is opportunistic and likely to be disingenuous in its dealings with others (Skarmeas and Leonidou 2013). This idea derives from the social identity perspective that organizational outsiders tend to surmise that the attributes of the top executive are prototypical of, and largely embody those of, the organization as a whole (Hogg 2001). Thus, instrumental motive attributions that result from exposure to a morally questionable CEO should increase organizational cynicism of the company. In contrast, attributions of moral CSR motives should be negatively related to organizational cynicism. This reasoning is supported by recent empirical evidence linking instrumental and moral forms of CSR motive attributions to organizational cynicism (Skarmeas and Leonidou 2013).

Hypothesis 2a

Stakeholder attribution of instrumental CSR motives is positively related to organizational cynicism.

Hypothesis 2b

Stakeholder attribution of moral CSR motives is negatively related to organizational cynicism.

Lastly, we propose that organizational cynicism is likely to result in negative attitudinal and behavioral reactions toward the CSR. To date, some studies have linked organizational cynicism to negative outcomes, such as word-of-mouth advertising and retail equity (e.g., Foreh and Grier 2003; Skarmeas and Leonidou 2013). To the best of our knowledge, however, there is a gap in our understanding of how CSR motive attributions affect stakeholders’ reactions to the CSR itself. The majority of existing research has focused on the positive benefits of CSR to the organization as a whole (Brønn and Vrioni 2001; Orlitzky et al. 2003), while little attention has been paid to factors that influence the success of the CSR initiative. We propose that CSR motive attributions, through cynicism, are likely to impact stakeholder’s support for the CSR initiative. Indeed, many organizations today encourage stakeholders to support their CSR initiatives in different ways. We argue that when stakeholders attribute instrumental (versus moral) motives to the company’s CSR initiative and become highly cynical of the firm, they should be strongly discouraged from supporting the CSR (Aguilera et al. 2007). When stakeholders become cynical about a company’s CSR motives, they are likely to experience strong emotional reactions (e.g., anger and resentment) to the firm’s apparent intentions to manipulate and take advantage of their goodwill (Johnson and O’Leary-Kelly 2003). This may make it difficult for the public to separate the duplicitous intent of the firm from the positive benefits that the CSR offers to society (Andersson 1996). Moreover, people may reason that the CSR is a “tactical maneuver” (Ellen et al. 2006, p. 155) that is unsustainable in the long run, and will be mismanaged in ways that ultimately result in a disservice to the intended beneficiaries of the CSR.

Thus, we expect an overall negative effect of organizational cynicism on CSR support. For instance, it is not an uncommon practice for companies to advertise that by buying their products and services, people will be supporting a particular CSR initiative. We therefore expect that stakeholders will withdraw passive forms of support for the CSR such as refusing to purchase the company’s products and services. Moreover, many companies rely on the public’s active involvement in their CSR initiatives as a way to generate interest and publicityFootnote 1 (e.g., through community events that rely on volunteers) and also to manage the costs of the CSR (e.g., matching donations made by the public). Thus, we anticipate that increased cynicism should also lead stakeholders to be less likely to actively support the CSR—either by personally volunteering to help with the CSR or by making financial donations to the CSR. In other words, we propose that instrumental and moral motive attributions differentially affect organizational cynicism, which in turn predicts different types of public support for the CSR. More broadly, we predict a causal mediatory chain that links CEO ethics to instrumental and moral motive attributions, which in turn influence organizational cynicism, and ultimately, stakeholder CSR support.

Hypothesis 3

Organizational cynicism is negatively related to a) purchase intentions, b) CSR volunteer intentions, and c) CSR financial donation behavior.

Hypothesis 4

The impact of CEO ethics (i.e., morally questionable CEO versus ethical or ethics-unknown CEO) on the public’s (a) purchase intentions, (b) CSR volunteering intentions, and (c) CSR financial donation behavior is mediated by stakeholder attributions of CSR motive attributions and cynicism, respectively.

Method

Participants

Two hundred and thirteen student participants from a mid-sized Eastern Canadian university in the business school participated in this study for bonus credit. Of these participants, 54.2 % were female and the average age was 22.17 years (SD = 3.34 years). In order to minimize socially desirable responding, students were informed that their professors were not made aware of their participation in the study until after final grades were approved. We considered business students appropriate for this research because, despite their membership in higher education institutions, they are also legitimate members of the public. Specifically, much like other demographics, they are avid consumers of the media, are equally exposed to CSR advertisements, and are equally solicited to support CSR initiatives in their daily interactions in the public domain.Footnote 2

Research Design and Procedure

We carried out a quasi-experimental study using a previously validated procedure (Ogunfowora 2014). As a cover story, participants were informed that they were participating in research designed to investigate “attitudes towards companies and their management.” At the beginning of the study, participants completed basic demographic and filler survey questions. Next, participants were verbally introduced to a (fictitious) company and provided with a number of materials to “give them more information about the company to better assist with their evaluations.” Similar to Ogunfowora (2014), we created a fictitious company called the MelGreen Group, an “energy and infrastructure firm” operating out of Western Canada. To enhance believability, we created a number of company materials, including a website and company brochures. The general brochure introduced participants to the company and its operations. In addition, the brochure informed participants that joint ventures are a core business area of the MelGreen Group. They were told that a particularly important MelGreen joint venture is FastGas, a chain of gas stations in Western Canada (with a small presence in Eastern Canada, where the study took place). We created a dedicated brochure for FastGas. This subsidiary was used to showcase MelGreen’s CSR initiative (described below). It is worth noting here that a significant number of students at the university where data were collected experience long, daily commutes to the main campus. This “commuter” university is situated at the edge of a geographically dispersed but easily accessible network of cities and townships, with about 55 % of the student population coming from distant cities in the region. As such, the gas retail industry is particularly relevant to the present context and sample.

We used previously developed press conference videos to introduce participants to the CEO of MelGreen and to manipulate participants’ perceptions of the CEO’s ethics. A White male professional actor in his mid-50s played the role of the CEO in all three video conditions. Other paid actors played the roles of PR officer, camera-persons, and members of the press. Participants were randomly assigned to three different versions of the press conference video that correspond to three types of CEO ethics conditions: Morally questionable CEO (n = 59), Ethical CEO (n = 64), and a control or “Ethics-unknown” CEO (n = 73).

At the start of the video, the CEO discusses the recent success and growth of the firm and indicates that the reason for the press conference is to announce the expansion of the company’s operations into two new Canadian provinces. After the announcement, reporters ask a series of questions about the new operations. About half way into the questioning period, a reporter inquired about an unrelated matter involving allegations against a corporate business partner. Specifically, this business partner, who had been operating overseas, has been accused of corruption and bribery of foreign government officials. The CEO’s responses to this question were manipulated to showcase his ethics-related beliefs, values, and behavioral tendencies. In the ethical leader condition, the CEO reprimands corporate corruption and promises to dissolve business ties with the partner if the allegations were true. The CEO’s resolve did not flounder even after the reporter highlighted that a substantial portion of MelGreen’s profits comes from its business with this partner. The CEO further describes his ongoing efforts to communicate and reinforce high ethical standards within his company. In the morally questionable leader condition, the CEO’s responses were designed to convey questionable ethics and attitudes toward bribery and corruption without coming across as entirely supportive of these actions. The morally questionable CEO was more sympathetic toward the accused business partner and promised to stand by the firm regardless of the allegations (e.g., “those foreign government officials are really to blame for demanding bribes… we cannot be seen as deserting our business partner during these difficult times … try explaining to stakeholders that they lost some of their investments because your business partner was convicted of corporate misconduct on another continent”). This strategy was designed to emulate real life, in which executives are highly unlikely to say things purposely in the media that show they are unethical. Finally, in the “Ethics-unknown” CEO video, there were no questions from the reporter about the allegations or any other issues pertaining to ethics. The videos were about 8–10 min in length. Immediately after the video, a confederate posing as a participant provided further “validation” of the CEO’s ethics portrayed in the video. The confederate claimed to have read a recent story in the newspaper that corroborates the relevant CEO ethics in each condition. We ensured no further discussion occurred among the participants following the confederate’s short remark.

Next, participants completed a survey about their attitudes toward the CEO and the company. Following this, they were introduced to MelGreen’s CSR initiatives through its FastGas subsidiary (brochures describing the CSR initiative were circulated). Participants learned that the CSR initiative, called Citizens Drive, provides transportation services to senior citizens and members of the community with disabilities. The transportation allowed individuals to get to places such as the doctor’s office, drug stores, health and wellness centers, and so on. They were also informed that a portion of every sale at FastGas stations goes to supporting Citizens Drive. Lastly, the brochure explained that the company relies on many volunteers in different capacities to ensure the smooth and efficient operation of Citizens Drive. Volunteers could help out as drivers, “drive buddies” (who accompany citizens to appointments), vehicle maintenance persons (e.g., cleaning the vehicles and filling them up with gas), and office support (e.g., those who schedule pickups and drop-offs). The brochure informed readers that “you can contribute to Citizens Drive by simply purchasing gas from any FastGas location and/or volunteering time to help out with Citizens Drive in whatever capacity you are interested in.”

After reviewing this brochure, participants completed the last part of the survey, which asked them about their intentions to support MelGreen/FastGas and the CSR initiative (i.e., product purchase intentions and intentions to volunteer for the CSR). Lastly, we assessed participants’ CSR financial donation behavior. Participants were told that a portion of the funding for this research had been set aside to support a community initiative in their area. We informed participants that in addition to the bonus course credits, we would donate $5 on behalf of every participant to one of two community support programs in their area. The first option was Citizens Drive, MelGreen’s CSR initiative. Participants were also given a second choice—PeopleCare, “a very similar program that supports senior citizens and those with disabilities operated by a relatively new not-for-profit agency.” Participants were then given an “independent financial donation” sheet with the above options. CSR financial support donation behavior was assessed by whether participants chose to donate the $5 to MelGreen’s CSR initiative or to the other similar (but unknown) CSR program. Importantly, we clearly emphasized that the funding for the donations did not come from the MelGreen group or any of its subsidiaries. At the end of data collection, participants were debriefed. The procedures used in this study were reviewed and approved by the university’s research ethics board.

Measures

Organizational Cynicism

Participants responded to a five-item measure of cynicism (α = 0.75) developed by Kanter and Mirvis (1989). They responded to the items on a 1 (strongly disagree) to 5 (strongly agree) scale. We replaced the term “people” with “the organization.” A sample item is “This organization will tell a lie if it can gain by it.”

CSR Motive Attributions

We assessed attributions of the company’s CSR motives using Ellen et al.’s (2006) measure. This measure assesses four types of motives, two of which align with moral and instrumental motives. We assessed moral motive attributions using four items from the value-driven motive scale (α = 0.77). Sample items include “the owners or employees of the company believe in the cause” and “this company is trying to give something back to the community.” We assessed instrumental motive attributions using the four-item egoistic motive scale (α = 0.74). Sample items include “this company is taking advantage of the compassion of customers to help its own business” and “this company wants to get publicity.” Participants responded to the items on a 5-point rating scale (1—strongly disagree; 5—strongly agree).

We also measured the two other motives attributions—strategic and stakeholder—assessed by Ellen et al.’s (2006) scale. We used these measures as a check on our decision to focus on moral and instrumental motives. Specifically, we did not expect the participants, based on their exposure to the CEO’s ethics, to evaluate the company along these two other attributions. This is because the strategic and stakeholder motives do not have explicit moral evaluative components to them. The strategy motive indicates that the company is concerned about performance and being successful, without clear reference to disingenuous intent (Ellen et al. 2006). A sample item is “the company will keep more of its customers by supporting this cause.” The stakeholder motive indicates that the company is concerned about the expectations of its stakeholders, including employees and stockholders (Ellen et al. 2006). A sample item is “the company feels its employees/stockholders expect it (i.e., its involvement in CSR).”

Purchase Intentions

Participants responded to a single-item measure of purchase intentions. Participants were told to “imagine that FastGas recently opened a location that was as convenient for you as other current gas stations, and offered the same quality, price, and service … How likely would you be to purchase gas from FastGas the next time you need to fill up?” They responded on a 5-point Likert-type scale (1—very unlikely; 5—very likely). Recall that participants were told that a portion of gas station sales went toward Citizens Drive.

CSR Volunteer Intentions

Participants responded to another question about their intentions to volunteer time to Citizens Drive. Specifically, they were asked, “In the absence of any scheduling conflicts, how likely would you be to volunteer a portion of your time to help with FastGas’ ‘Citizens Drive’ initiative?” They responded to this question on a 5-point Likert-type scale (1—very unlikely; 5—very likely).

CSR Financial Donation Behavior

Participants’ choice of donating to the company’s CSR initiative (Citizens Drive) versus the other initiative (PeopleCare) was indicative of their donation behavior. We coded a participant’s choice to donate to PeopleCare as “1” and donations to Citizens Drive as “2.”

Manipulation Check Test

The respondents evaluated the CEO’s ethics using seven items from the Ethical Leadership Scale (ELS; Brown et al. 2005). The selected items assessed CEO attributes that participants could reasonably judge from the videos (e.g., “the CEO discusses business ethics or values with employees” and “the CEO disciplines employees who violate ethical standards”). The results of a one-way analysis of variance (ANOVA) showed a statistically significant overall manipulation effect, F (2, 217) = 88.50, p < 0.001. The means of each condition (morally questionable CEO, ethical CEO, and ethics-unknown CEO) were significantly different from one another (M = 2.68, 4.02, and 3.53, ps < 0.01). Thus, the manipulation was successful. We also manipulated the order in which the manipulation check questions were presented (on the first or last page of the questionnaire). T test analyses showed no difference between the two versions on any of the study variables. Hence, the data were combined. Lastly, shortly after data collection ended, we sent participants an email asking them to describe the extent to which they believed that the company and its CEO were fictional or made up. We sent delayed emails rather than asking them directly at the end of the study because participants came from the same class. As such, we were concerned that they would relay any suspicions to other students who had yet to complete the study. We received responses from roughly 30 % of the respondents. The results showed that there was no significant difference in perceptions of believability across the three CEO ethics conditions, F (2, 62) = 0.31, p > 0.05.

Analytic Strategy

First, we carried out a series of confirmatory factor analyses (CFAs) to test our measurement model. These analyses included all variables that were measured in the survey questionnaire (i.e., all study variables except for CEO ethics). Second, we tested our hypotheses using structural equation modeling (SEM) in Mplus. For these analyses, we created two dummy-coded CEO ethics variables to allow for meaningful comparison. The first compared the ethical CEO to the morally questionable CEO (this variable is labeled Ethical vs. Morally questionable). The second compared the Ethics-unknown (or control) CEO condition to the morally questionable CEO (labeled Ethics-unknown vs. Morally questionable). These dummy-coded variables were entered into the SEM analyses as exogenous, categorical observed variables.

Prior to the main analyses, we regressed the dummy-coded CEO ethics variables on the four types of CSR motive attributions assessed by Ellen et al.’s (2006) measure. The goal of this analysis was to determine whether respondents differentiated the morally questionable CEO from the ethical and ethics-unknown CEOs along the four CSR motive attributions. The results showed that the [Ethical vs. Morally questionable] and [Ethics-unknown vs. Morally questionable] variables were both significantly and negatively related to the egoistic (i.e., instrumental) and values (i.e., moral) attributions. In other words, participants rated the morally questionable CEO’s company significantly lower than the ethical and ethics-unknown CEOs’ companies on these two CSR motives. In contrast, participants’ ratings did not differ significantly on the strategic and stakeholder motives, with the exception of one case (of the four possible associations, there was only a significant difference between morally questionable and ethical CEO on strategy motive). These findings generally support our decision to focus primarily on the moral and instrumental CSR motives.

Results

Measurement Model

In order to assess the distinctiveness of the measured constructs, we tested our hypothesized six-factor model, which includes instrumental motive, moral motive, cynicism, purchase intentions, volunteering intentions, and financial donation behavior. The results showed strong support for this model (see Model 1 in Table 1), with all items loading on their respective constructs, χ 2 (94) = 136.92, p < 0.05, CFI = 0.92, TLI = 0.90, and RMSEA = 0.05.

Table 1 Results of confirmatory factor analyses of study measures
Full size table

The proposed six-factor structure was further compared with other alternate models, including Model 2a (a five-factor model that combines instrumental and moral motives), Model 2b (a five-factor model that combines cynicism and moral motives), Model 2c (a five-factor model that combines cynicism and instrumental motives), Model 3 (a four-factor model that combines cynicism, instrumental, and moral motives), Model 4 (a two-factor model that allows the independent variables to load on one construct and the dependent variables on another), and Model 5 (a one-factor model). The results showed that the hypothesized six-factor model was a better fit to the data, compared to all the alternate models. These results are presented in Table 1. Lastly, we assessed the possibility of common method bias effects in our self-reported data (Podsakoff et al. 2003). We specified a five-factor model that included the self-reported attitudinal measures in our data (we excluded the donation behavior variable because it was not assessed using the same method). This five-factor model was compared to an alternate model that specified an additional sixth, uncorrelated method factor. All items loaded on their respective factors and on the method factor. The results showed that the addition of the method factor did not improve fit to the data, ∆χ 2 (16) = 13.21, p > 0.05, CFI = 0.89, TLI = 0.83, and RMSEA = 0.09. Moreover, the method factor accounted for 3.93 % of the variance in the items. This is much smaller than previous estimates (roughly 25 %) of the impact of common method bias in self-reported data (Podsakoff et al. 2003). Thus, these results indicate that common method bias is not a significant issue in the data.

Hypotheses Testing

The descriptive statistics and correlations among the study variables are presented in Table 2. To test our hypotheses, we carried out structural equation model (SEM) analyses. First, we tested our proposed theoretical model, which is a fully mediated path from CEO ethics to the dependent variables through CSR instrumental and moral motive attributions and public cynicism, respectively. The results showed an excellent fit to the data, χ 2 (126) = 201.99, p < 0.01, CFI = 0.94, TLI = 0.92, and RMSEA = 0.05. These results are shown in Fig. 2. Compared to the morally questionable CEO, participants attributed significantly lower instrumental CSR motives to the ethical CEO (β = − 0.41, p < 0.01) and ethics-unknown CEO (β = − 0.28, p < 0.01). This provides support for hypothesis 1a. In support of hypothesis 1b, participants attributed significantly higher moral CSR motives to the ethical (β = 0.54, p < 0.01) and ethics-unknown (β = 0.54, p < 0.01) CEOs, compared to the morally questionable CEO. In turn, instrumental (β = 0.31, p < 0.01) and moral (β = − 0.77, p < 0.01) CSR motive attributions were significantly related to cynicism about the organization. Hence, hypotheses 2a and 2b are supported. Next, participant cynicism was negatively related to purchase intentions (β = − 0.44, p < 0.01), CSR volunteering intentions (β = − 0.35, p < 0.01), and actual financial donation behavior (β = − 0.23, p < 0.05). These results provide support for hypotheses 3a, 3b, and 3c, respectively.Footnote 3

Table 2 Descriptive statistics: Means, standard deviations, and correlations
Full size table
Fig. 2
figure 2

Structural equation modeling (SEM) test of theoretical model. Factor loadings are not shown. Estimates reported are standardized values. *p < 0.05, **p < 0.01.

Full size image

Next, we tested the indirect effects proposed in hypothesis 4 using bootstrapping analyses in Mplus. We estimated 95 % confidence intervals around the estimated indirect effects from CEO ethics to each dependent variable. For the comparison between the ethical CEO and morally questionable CEO, there are two possible indirect effects to each dependent variable—that is, one indirect effect through instrumental motive attribution and the other through moral motive attribution. Hence, there are six indirect effects in total for all three dependent variables. Similarly, the comparison between the ethics-unknown and morally questionable CEO also has six proposed indirect effects. The results of the bootstrapping analyses are presented in Table 3. In support of Hypothesis 4, the results showed empirical support for all possible indirect effects from CEO ethics to the dependent variables. Media exposure to the morally questionable ethics of a CEO—compared to an ethical or ethics-unknown CEO—indirectly impacted passive and active support for the company’s CSR through the mediating effects of moral and instrumental CSR motive attributions and cynicism, respectively.

Table 3 Bootstrapping tests of proposed indirect effects using Bayesian procedures
Full size table

Lastly, we tested the plausibility of reverse causality in our theoretical model. We statistically evaluated the possibility that CEO ethics influences organizational cynicism, which in turn impacts CSR support intentions and behaviors through public attributions of CSR motives. In this SEM model, we swapped the positions of organizational cynicism and CSR motive attributions. The results showed that this model did not result in an improved fit to the data, χ 2 (142) = 204.53, p < 0.01, Δχ 2 (16) = 2.54, p > 0.05, CFI = 0.95, TLI = 0.94. Importantly, bootstrapping analyses showed weak support for this theoretical ordering of constructs. The indirect effects of CEO ethics on CSR financial support behavior was not supported in this causal ordering. In addition, indirect paths through instrumental CSR motive attributions were not supported. The complete results of the bootstrapping analyses are presented in Table 4. In sum, the results show greater support for the proposed theoretical ordering that places CSR motive attributions and organizational cynicism as proximal and distal mediators, respectively (and not vice versa).

Table 4 Bootstrapping tests of alternate model with reverse causal ordering of indirect effects
Full size table

Discussion

Organizations send signals to stakeholders as a means of communicating positive attributes of organizational actions. Signaling theory (e.g., Connelly et al. 2011) suggests that firms do this in order to make positive impressions on stakeholders. However, even though CSR may be framed as a positive signal, CSR initiatives have become equivocal signals to stakeholders because different motives exist behind them. Specifically, organizations may send information about CSR engagement, but this information exists alongside other signals from media coverage of questionable corporate practices, such as a company spending more on marketing its CSR initiative than on the initiative itself. With limited information at hand, stakeholders (e.g., consumers and members of the general public) today may assume that an organization’s investment in a CSR initiative is primarily driven by the desire to profit from the economic benefits of CSR (an instrumental motive) rather than by a genuine concern for social change (a moral motive). In other words, stakeholders have become increasingly doubtful of the sincerity of organizational motives for engaging in CSR (Skarmeas and Leonidou 2013).

In order to ascribe the motives behind the equivocal CSR signals, stakeholders enact an attribution process. In this study, we examined the role of the CEO’s ethics as an important cue in the stakeholder’s CSR motive attribution processes. Our findings suggest that stakeholders react to CSR signals positively when additional cues from CEOs are consistent with genuine CSR motives (i.e., in the case of ethical CEOs), while they perceive CSR signals negatively when cues from CEOs are inconsistent with the company’s message (i.e., morally questionable CEOs). We found that a morally questionable CEO is positively associated with stakeholder’s attribution of instrumental motives and negatively associated with moral motives, both of which differentially influence organizational cynicism. We further found that increased organizational cynicism reduced stakeholder CSR support, such as CSR-related purchase intentions, CSR volunteerism, and CSR financial donation behaviors. Collectively, our findings suggest that CEOs need to seriously consider public perceptions of why their companies are engaging in CSR and the role that they play in shaping these perceptions.

Contributions and Implications

In examining stakeholders’ attribution process and subsequent behaviors, we make a number of theoretical contributions to existing research on signaling theory, views on attribution process, and CSR. First, in examining CSR as equivocal signals, we contribute to signaling theory and attribution theory by linking two theoretical perspectives. While CSR initiatives may communicate equivocal signals in terms of the corporate motives behind them, only few recent studies (e.g., Skarmeas and Leonidou 2013) have begun to argue that CSR initiatives can be perceived negatively. Furthermore, our findings highlight the critical role of signal senders, but more importantly, the salience of signal receivers (i.e., stakeholders). Depending on the perceived CEO ethics, stakeholders’ attribution and subsequent behavior may not be the ones intended and desired by signal senders. Therefore, it is important to align the attributes of the signals (i.e., CSR initiatives) with those of the sending actors (i.e., the CEO ethics).

Second, we show the importance of the stakeholder’s role in interpreting equivocal signals. The potential influence of CEO ethics on stakeholder attributions of CSR motives has received little attention to date. Management scholars have examined the role of senior executive characteristics in understanding a firm’s involvement in CSR initiatives (e.g., Manner 2010; Oh et al. 2016). This line of research focuses on the role of executives as key decision makers who drive organizational actions related to social issues. Yet, no research to date has considered whether external stakeholders consider the executive leader when forming judgments about the motives behind an organization’s involvement in CSR. In other words, the role of the CEO from the stakeholder’s perspective is not well understood. Our findings suggest that—beyond basic considerations such as the nature of a firm’s CSR communication strategy—a CEO’s ethics can filter through a variety of social media platforms to shape stakeholder perceptions of the company’s CSR motives.

Moreover, our study provides some insight into the mechanism through which CEO ethics affect public support for CSR initiatives. Consistent with attribution theory (Kelley and Michela 1980), we showed that CEO-driven CSR motive attributions affect stakeholder cognitions (i.e., cynicism about the organization), and this ultimately influences their willingness to support the firm’s CSR. These findings contribute to ongoing endeavors to understand how stakeholders process and respond to information pertaining to an organization’s involvement in CSR. To the best of our knowledge, this study is the first to explicate the process through which the CEOs, as key signal-sending actors, shape the intentions and actions of external stakeholders in understanding public support for CSR. Our findings show that organizational cynicism plays an important mediating role in explaining how CEO-driven CSR motive attributions impact stakeholders’ reactions.

From a practical standpoint, this study provides important implications for the formulation and implementation of a company’s CSR policies. Many companies today participate in social issues as an essential organizational agenda (Skarmeas and Leonidou 2013). As a result, organizations increasingly invest substantial resources in CSR programs. While it is important for companies to proactively engage in CSR, our study suggests that it may be more important to convince stakeholders that their CSR engagement comes from a genuine motive to accomplish social change. Our findings emphasize the importance of having and signaling ethical motives (through the CEO) since the detrimental effects from stakeholder doubt may overwhelm the potential positive outcomes of the CSR. One way to achieve this is to establish a strong CEO public profile of integrity, respect, and commitment to business ethics. There are examples of ethical CEOs who frequently talk about the importance of ethics on various social media platforms (e.g., YouTube, TED talks). Perhaps more importantly, ethical CEOs are also routinely recognized by external bodies that monitor ethics in business (e.g., the Botwinick Prize in Ethics and the Ethisphere Awards).

Corporate executives should also pay attention to stakeholder’s CSR motive attributions after they implement new CSR programs. In particular, companies may wish to consider adopting monitoring systems that evaluate and track organizational cynicism. This should help organizations develop action plans to manage potential misperceptions and ultimately reduce unwarranted cynicism. For example, many hotel chains advertise towel reuse campaigns for the environmental concerns. However, there has been increasing cynicism about how savings are actually used by the companies. In this case, the CEOs of these hotels may need to highlight that the savings will be put toward a good cause (e.g., donation to environmental protection programs). Also, it is likely that towel reuse campaigns would be more effective to reduce stakeholder cynicism if the campaign also translates into benefits for stakeholders (e.g., hotel guests receive additional hotel reward points by their involvement in the towel reuse campaign).

Limitations and Future Directions

In spite of our study’s contributions, there are a few limitations. First, we used an experimental design to manipulate CEO ethics. Although our use of the press conference setting mirrors a real-life context in which the public is likely to learn about a CEO’s ethics, public exposure to CEOs remains relatively weak. New social media platforms such as twitter and LinkedIn (e.g., Lee et al. 2013) may, however, be helping reduce this gap.

Second, while we used an experimental design to manipulate CEO ethics, our other constructs were assessed using self-reported survey measures (except for financial donation behavior). This opens up the possibility of same-source, common method bias effects. However, our CFA analyses suggest that this is less likely to be a concern in our data analyses. Also, the cross-sectional design of the study—particularly in the administration of the survey measures—limits our ability to claim a causal relationship among the variables in the latter part of our model. For example, a reverse causality might occur such that organizational cynicism influences attributions of CSR motives (as opposed to the reverse direction proposed in our theoretical model). Although we found weak support for this alternate model in our analyses, the issue of reverse causality is most likely to be satisfactorily addressed using longitudinal research designs. Longitudinal research designs also have the added benefit of being able to clarify whether CSR motive attributions and organizational cynicism fade over time following corporate interventions suggested above.

Third, our findings may have stronger implications for companies operating in consumer goods and services industries, such as retailing, fashion, and personal financial services, where the public image of the CEO is critical in signaling to the stakeholders. In contrast, the results may have limited implications for companies in industrial goods and services markets, such as shipbuilding, heavy equipment, professional corporate service (e.g., accounting firms), and military service. Generally, members of the public are not usually exposed to, or play active roles as stakeholders of companies in these industries (compared to the retail industry, for instance); thus, our findings are likely to have limited generalizability. Future studies may benefit from examining the roles of top executives as signalers and stakeholder attribution processes in different industries.

Fourth, future studies may also benefit from examining the role of other individual, organizational, environmental, and institutional contextual factors that may alter the stakeholder attribution process; this would allow for a greater understanding of the external generalizability of our findings. For example, stakeholder judgments that a CEO is trustworthy and/or that a firm has an ethical corporate culture may provide a buffering effect in reducing stakeholder cynicism and subsequent negative behaviors. Also, given that CSR may be influenced by institutional systems (e.g., the political, economic, financial, and cultural environment), future studies may benefit from examining whether our findings can be observed in other institutional contexts.

Lastly, future research should test the generalizability of our findings by replicating our study in a field setting using a non-student sample. We do not expect, however, that our findings will be significantly different from replications in a field setting given that university students are also members of the public who are routinely involved in public discourse on CSR within and outside the academic environment. They are exposed to CSR advertising and their support is solicited in much the same way as it is with other adult groups.

Conclusion

In conclusion, organizations continue to face strong internal and external pressures to be good corporate citizens in society (Aguilera et al. 2007). As such, many organizations have begun to integrate CSR into their corporate strategic agendas as positive signals. Our findings illustrate that the ethics of CEOs, as key signal-sending actors, can be influential in shaping the perceptions that stakeholders have of the firm’s motive for engaging in CSR. Organizations need to pay greater attention to creating ethical awareness and responsibility in their executives as negative impressions of the CEO can significantly impact the success of the company’s CSR initiative. Ultimately, our hope is that this paper will spur future research that considers the role of CEO leadership in forming judgements about the motives behind an organization’s involvement in CSR.