Comparing the Effectiveness of Post-Scandal Apologies From Nonprofit and Commercial Organizations: An Extension of the Moral Disillusionment Model

Abstract

There is a double standard in public responses to scandals: Nonprofits are penalized more harshly than commercial organizations for the same transgression (the “moral disillusionment effect”). However, previous research—focused on commercial organizations—has sometimes shown that a positive reputation can insure organizations against the negative effects of scandals. In light of this, we asked whether a second double standard exists when it comes to trust repair: Can nonprofits regain trust and consumer support more quickly than commercial organizations after apologizing? Two experiments (combined N = 805), considering responses to sexual exploitation and fraud scandals, replicated and extended the moral disillusionment effect. Trust and consumer support were partially restored following an apology (and even a statement acknowledging the scandal without apologizing), but the rate of repair was the same for nonprofits and commercial organizations. Nonprofit managers should therefore implement internal controls to prevent violations and issue public responses when scandals emerge.

Keywords

trust scandal apology trust repair nonprofits

A series of high-profile charity scandals have sent shock waves through the nonprofit sector in recent years. These include scandals about sexual exploitation by employees of Oxfam and Save the Children (Scurlock et al., 2020), slow dispersion of funds to victims of the Haiti earthquake by the Red Cross (Sullivan, 2015), and charities’ high-pressure fundraising techniques allegedly contributing to an elderly donor’s suicide (Hind, 2017). Media outlets claim such transgressions have shaken the public’s trust in nonprofits, in turn making it difficult for such organizations to effectively fundraise and fulfill their social missions (e.g., Gaskin, 1999).

Compounding the problem, recent evidence suggests that there is a double standard in how the public respond to scandals emerging within different sectors. Specifically, experiments show that nonprofits are punished more harshly than commercial organizations for the same ethical transgression. Consistent with the Moral Disillusionment Model (MDM), this double standard is at least partly driven by the violation of expectations that occurs when an organization that you presume to be moral does something immoral (Hornsey et al., 2020). However, research on this phenomenon is in its infancy, and to date the boundary conditions for the effect are poorly understood. For example, it could be that nonprofits are penalized relatively harshly in the immediate aftermath of a scandal, but that their reputations bounce back relatively quickly after a trust repair effort. To answer this question, the current research consists of two experiments that test an extension of the MDM by comparing the impact of organizational apologies from nonprofit versus commercial organizations following a transgression. This comparison addresses the question of whether a nonprofit’s loss of trust following a transgression is temporary and recoverable, or rather a more permanent consequence of having lost their moral credentials.

The Impact of Moral Credentials

Research shows that the moral credentials of organizations can impact how the public responds to future events. Moral credentials—or a reputation for being good—can be indicated in a number of ways. The literature has focused primarily on commercial organizations that highlight their moral credentials through engagement in corporate social responsibility (CSR) programs, corporate philanthropy, or cause-related marketing campaigns. But an organization’s status as a nonprofit or social enterprise can also signal moral credentials.

The jury is out with regard to whether or not moral credentials serve as an asset in the face of scandal. On one hand, some research suggests that past good deeds may insure against the negative impact of future misdeeds. On the other hand, research sometimes shows boomerang effects, whereby organizations with positive reputations are evaluated more harshly when negative information comes to light. Following Hornsey et al. (2020), we use the terms moral insurance and moral disillusionment to refer to these effects, respectively.

Evidence of moral insurance effects comes primarily from the CSR literature. Broadly speaking, a history of CSR can protect firms from damage caused by scandal (Janney & Gove, 2011). Stock markets react less negatively to scandals involving firms with a history of CSR (Godfrey et al., 2009). By examining stock market responses to environmental announcements, for example, Flammer (2013) demonstrated that firms with a stronger history of environmental CSR experienced smaller drops in share value after announcements of environmentally harmful behavior than did firms with a weaker history of environmental CSR. Commercial organizations who engage in CSR also have fewer lawsuits filed against them (Barnett et al., 2018), indicating protection from scandal in the first place. However, even when scandals do emerge, firms who have engaged in higher levels of corporate philanthropy—another way to indicate moral credentials—experience less reputation loss after engaging in criminal activity (Williams & Barrett, 2000). In sum, there is a wealth of empirical evidence that moral credentials are sometimes a benefit, insuring commercial organizations against the fall out of scandals.

In contrast, there is also evidence of moral disillusionment effects, coming mostly from the CSR and corporate reputation literatures (e.g., Janssen et al., 2015; Sohn & Lariscy, 2015). For example, Chandler et al. (2020) studied incidences of major U.S. oil spills between 1985 and 2016. They found that press coverage of oil spills for firms that had a reputation for bad behavior (operationalized as a greater number of previous oil spills) was lower because a new spill was not considered newsworthy. This suggests greater scrutiny (likely resulting in harsher penalties) for firms who are first time offenders, or who otherwise have a reputation for integrity in this domain. Also, communicating CSR credentials to buffer against the negative effects of crisis may not be effective for firms who don’t have a long history of engagement in CSR, indicating that the public are skeptical of firms using good deeds to wallpaper over bad behavior (Vanhamme et al., 2015; Vanhamme & Grobben, 2008).

Moral disillusionment effects can be explained with reference to Expectancy Violation Theory (EVT; Burgoon & Le Poire, 1993). EVT proposes that negative reevaluations are generated when an entity engages in behavior that deviates from expectations. Furthermore, the more an entity’s behavior deviates from expectations, the more profoundly negative the reevaluations will be (Lin-Hi et al., 2015). Thus, EVT could explain observed moral disillusionment effects: People have higher expectations of organizations with strong moral credentials (i.e., those that engage in CSR or corporate philanthropy), and these higher expectations can be violated more profoundly by bad behavior.

In sum, moral credentials can sometimes insure an organization against the ill-effects of future scandals, and sometimes backfire by bringing stronger reprimand for transgressions that violate consumers’ expectations. This research, however, has been conducted almost entirely with commercial organizations. One notable exception is recent work that generated the MDM, which is summarized below.

The MDM of Organizational Transgressions

Hornsey and colleagues (2020) compared the effects of ethical transgressions that occurred within nonprofits and commercial organizations. Across three experiments—using examples of organizational transgressions involving fraud, sexual exploitation of women, and exploitative labor practices—they found consistent support for the moral disillusionment hypothesis. Nonprofits suffered greater losses of trust, consumer intentions, and word of mouth intentions than did commercial organizations that had committed the exact same transgression: in part because consumers have higher expectations of nonprofits. This finding demonstrated a double standard: Nonprofits suffer more than commercial organizations when scandals emerge.

The MDM shows that nonprofits suffer more negative effects after a transgression than do commercial organizations. However, empirical work on this question is still in its infancy. In the work by Hornsey and colleagues, participants evaluated the organizations immediately after hearing about a transgression. In reality, however, scandals are almost always followed up with some attempt at trust repair (e.g., an organizational apology). The MDM shows that an organization’s moral credentials impact the rate of trust loss after an organizational transgression. We test for the first time whether or not moral credentials also impact the rate of trust repair after a transgression, with a particular focus on apologies.

Organizational Apology as a Strategy for Repairing Trust

Trust repair efforts are actions that are intended to return a relationship to a positive state after trust has been lost or damaged (Kramer & Lewicki, 2010). Apologies are one of the most commonly studied forms of trust repair (Lewicki & Brinsfield, 2017) and can repair trust in two broad ways (Dirks et al., 2009; Kramer & Lewicki, 2010). First, from an attributional perspective, an apology may offset or change some of the negative inferences which resulted from the violation (Gillespie & Dietz, 2009; Kim et al., 2004). Second, from a social equilibrium perspective, an apology may serve to restore balance to a relationship and reaffirm the social norms that govern it in the aftermath of a violation (Dirks et al., 2009). Although results have sometimes been mixed (e.g., Davidow, 2000; de Ruyter & Wetzels, 2000), the majority of studies show apologies are relatively effective in terms of restoring trust, customer satisfaction, and positive consumer sentiment (e.g., Coombs & Holladay, 2008; Hegner et al., 2016; Kiambi & Shafer, 2016; Xie & Peng, 2009).

There has been debate in research within the fields of psychology and communication about what constitutes the key components of an apology (e.g., Lewicki & Polin, 2012; Schlenker & Darby, 1981). One simple definition of an apology is “a statement that acknowledges both responsibility and regret for a trust violation” (Kim et al., 2004, p. 105). Lewicki and Brinsfield (2017) propose there are six key components of an apology: expression of regret (e.g., saying “we are sorry”), explanation for the violation, acknowledgment of responsibility, declaration of repentance, offer of repair, and request for forgiveness. Apologies have been shown to be more effective when more components are used, and some components are more important than others (Lewicki et al., 2016). Two of the most effective components are an acknowledgment of responsibility and an offer of repair.

To date, research on the effectiveness of apology for trust repair has focused exclusively on responses to transgressions committed by commercial organizations. Research on trust maintenance among nonprofits has focused instead on strategies such as achieving accreditation (Becker, 2018; Bekkers, 2003), celebrity endorsements or using social media to promote positive news stories (Scurlock et al., 2020), and using organizational statements that deny or diminish the impact of a scandal (Hou et al., 2020).

No study has yet considered how trust in nonprofits can be repaired by way of an organizational apology. Although focused on a for-profit context, one article is relevant to the current research. Chung (2018) investigated how participants felt about a pharmaceutical company that had failed to adequately notify the national governance board about the negative side-effects of one of their drugs. When the apology referenced previous CSR work the company had done in the health care arena, perceptions of the company were more positive than when the apology did not reference their CSR efforts. However, the reverse was true when the CSR work was in a different domain to the transgression (e.g., on education).

The Current Research

The current research compares how the public respond to apologies from nonprofits and commercial organizations following a scandal. Competing predictions are plausible based on the literature reviewed above and the MDM. These competing predictions are summarized in Figure 1. The left side of the figure reiterates the moral disillusionment effect reported by Hornsey and colleagues (2020). The focus of the current article is to test the trust repair extension depicted on the right side of the figure.

Figure 1.

Conceptual model of the competing moral insurance (H1) and moral disillusionment (H2) hypotheses within the context of trust repair after a transgression.

If the past good deeds of nonprofits are taken into consideration, we would expect to see a stronger recovery for nonprofits (compared with commercial organizations) once they signal their remorse in the form of an apology. This would be consistent with evidence on corporate apologies that shows that organizations can benefit from moral credentials in the form of past CSR (Chung, 2018). Thus, the moral insurance hypothesis would anticipate the following:

Hypothesis 1 (H1): Trust and consumer intentions will rebound more effectively for nonprofits (vs. commercial organizations) following an apology.

Alternatively, if consumers have lost faith in nonprofits after their transgression, we would expect a weaker return of trust and consumer intentions despite trust repair efforts. The strongest version of the moral disillusionment hypothesis would therefore see nonprofits recovering trust and consumer intentions less effectively than commercial organizations. This would also be consistent with (untested) propositions from Bundy and Pfarrer (2015) that organizations with high social approval before a scandal may arouse feelings of skepticism and betrayal when they accept responsibility for a transgression. A weaker version of the same hypothesis is that improvements in trust and consumer intentions will be no different between nonprofits and commercial organizations. Nonprofits are penalized more harshly than commercial organizations immediately after a transgression (Hornsey et al., 2020). It may also be that, having lost their special status as moral organizations, nonprofits are subsequently treated as equivalent to commercial organizations, “locking in” the original disillusionment effect. Thus, the moral disillusionment hypothesis would anticipate the following:

Hypothesis 2 (H2): Trust and consumer intentions will rebound less effectively or equally for nonprofits (vs. commercial organizations) following an apology.

By considering whether organizational apologies are (equally) effective for both nonprofits and commercial organizations, the current research contributes to the literature in two key ways. First, it extends the MDM by examining consumer responses to trust repair efforts following scandals in nonprofit compared with commercial organizations. Second, it is the first research to assess experimentally the effectiveness of organizational apologies within the nonprofit sector.

Study 1

By looking not just at consumer responses to an ethical transgression, but also their responses to trust repair efforts, Study 1 considers the boundary conditions of consumers’ moral disillusionment in nonprofits after a scandal. We test the two competing hypotheses (H1 and H2) outlined above.

As a secondary question, we sought to nuance between two types of nonprofit transgressions. Past research (Hornsey et al., 2020, Study 2) has shown that moral disillusionment is stronger when the transgression directly subverts the stated mission of the nonprofit (e.g., the mission is to protect children and the transgression involves exploiting children) than when the transgression is mission-irrelevant (e.g., the mission is to protect the environment and the transgression involves exploiting children). Likewise, as described above, a history of CSR has a different impact on how people respond to an apology depending on whether the CSR activity is in the same or a different domain to the transgression (Chung, 2018). Given this, Study 1 required participants to respond to the same transgression attributed to one of three different organizations: a commercial organization, a nonprofit whose mission was relevant to the transgression, and a nonprofit whose mission was irrelevant to the transgression.

Method

Participants

American participants were recruited through Dynata, an online crowdsourcing survey platform, for a 9-min study. Of the 554 respondents who completed the study, 93 were removed from analysis as a result of failing one of three comprehension checks (65 at Time 1, 16 at Time 2, and 12 at Time 3—for details of these measures, see below).¹ This resulted in a usable sample of 461 participants. Participants were aged between 18 and 82 years of age (M_age = 46.94 years). The sample consisted of 206 males, 254 females, and one participant who identified as “other.”

Design and procedure

The experiment used a 3 (Organization Type: Commercial vs. Nonprofit Children vs. Nonprofit Women) × 3 (Time: Pre-transgression vs. Post-transgression vs. Post-apology) mixed groups design. Organization type was assessed between-subjects, and time was within-subjects.

Participants were first introduced to an online fashion store called GlobalDress that was described either as a commercial organization or as a nonprofit with a mission to protect children, or as a nonprofit with a mission to protect women from sexual exploitation. After reading the description of the organization, participants indicated both how much they trusted the organization and their consumer intentions. Next, participants read about an organizational transgression: Staff of GlobalDress had been involved in the sexual exploitation of women in developing countries. They again indicated their trust and consumer intentions. Finally, participants read a statement of apology from GlobalDress and once more indicated their trust and consumer intentions. All participants were then fully debriefed.

Manipulations

The experimental materials are summarized below. Full stimuli and questionnaires for both studies are also available on the Open Science Framework (OSF; https://osf.io/q4vth/).

Organization type

In the commercial condition, participants were introduced to GlobalDress. Although GlobalDress is fictitious, it was presented to participants as real. They were told that the organization was “an online-only fashion store that designs and sells a wide selection of fashion goods for both males and females.”

In the nonprofit conditions, the same organization was described as a “charitable, not-for-profit online-only fashion store” that “promises to direct 50% of their profits toward their charity arm.” Participants read a variant of the original text followed by a description of the charity’s mission, which was presented either as reducing exploitation of children or of women (i.e., “The charity works to reduce [exploitation of child labor/sexual exploitation of women] in developing countries”).

Transgression

All participants were then presented with the following information describing an organizational transgression committed by GlobalDress:

The CEO of GlobalDress—along with several other company directors—have been found to have been purchasing sex services in multiple developing countries. Investigations showed that the CEO, the Chief Financial Officer, Director of Operations, and the Director of Community Affairs were purchasing services from sex workers while on business trips.

The investigation revealed that this occurred while the staff members were in Indonesia, Bangladesh, Cambodia and Rwanda. The investigation indicated that these individuals paid for sex at brothels, but they used personal funds, not company funds.

Apology

All participants were then presented with an apology issued by GlobalDress. The full statement (available on the OSF) included an expression of regret, acknowledgment of responsibility, and an offer to repair trust by implementing preventive measures to avoid future offense. An extract of this apology, for example, reads as follows:

As many of you are aware, a recent investigation has found that a number of our senior management team have been paying for sex services while on business trips in developing countries. We assume responsibility for this failure and sincerely apologize. We have damaged your trust by failing in our practices. [. . .] We will ensure that we take all appropriate actions to assist them. Further, we promise to identify exactly how this could have happened, and will take the necessary steps to ensure that processes are in place to prevent this from happening again.

Comprehension checks

To ensure that the manipulations were effective, participants responded to a True or False question pre-transgression (“GlobalDress is an online fashion business that sells fashion goods for males and females”), post-transgression (“GlobalDress CEO and several company directors have been found to have purchased sex services in multiple developing countries”), and post-response (“GlobalDress promise to identify how this happened, and will take the necessary steps to prevent it from happening again”).

Measures

In this article, we report and analyze only a subset of the measures that were included in the study. Below we detail the relevant measures, including sample items for existing scales. The full questionnaire and supplementary analyses are also available on the OSF. Unless otherwise noted, items in both studies were measured on 7-point Likert-type scales (1 = Strongly disagree, 7 = Strongly agree) and were averaged with higher scores indicating more of the construct.

Trust was measured using four items from Nakayachi and Watabe’s (2005) perceived integrity scale (e.g., “This company is trustworthy).² This scale proved reliable at all time points (α_{pre-transgression} = .83, α_{post-transgression} = .87, α_post-apology = .88).

Consumer intentions were measured to assess the potential consequences of trust damage or repair. We selected a broad measure of consumer sentiment and intentions so that it would be relevant in both nonprofit and commercial contexts. Five items based on the “intention of usage” scale by Nakayachi and Watabe (2005) were used but adapted to refer to an online clothing store (e.g., “I would buy their products”; α_{pre-transgression} = .88, α_{post-transgression} = .96, α_post-apology = .96).

Results and Discussion

Trust

A 3 (Organization Type) × 3 (Time) analysis of variance (ANOVA) on trust revealed a significant main effect of time, F(2, 916) = 539.88, p < .001, $η_{p}^{2}$ = .54. Participants’ trust dropped from pre-transgression (M = 4.97, SD = 0.91) to post-transgression (M = 2.96, SD = 1.36) but partially recovered post-apology (M = 3.79, SD = 1.24). All comparisons were significant at p < .001.

As predicted, the main effect of time was qualified by a significant Time × Organization Type interaction, F(4, 916) = 10.92, p < .001, $η_{p}^{2}$ = .05, see Figure 2. Inspection of Figure 2 suggests an asymmetrical drop in trust from pre-transgression to post-transgression depending on organization type but a relatively even improvement in trust from post-transgression to post-apology regardless of organization type. This pattern replicates Hornsey and colleagues (2020) and is consistent with their moral disillusionment hypothesis.

Figure 2.

Changes in participant trust over time as a function of organization type (Study 1).

To formally test whether these differences were reliable, we ran a 2 (Time) × 3 (Organization Type) ANOVA, focusing just on scores pre-transgression compared with post-transgression. This revealed a significant interaction, F(2, 458) = 12.52, p < .001, $η_{p}^{2}$ = .05. Simple main effects of time were significant in all three conditions, although the effect sizes were much stronger in the nonprofit women condition, F(1, 458) = 356.35, p < .001, $η_{p}^{2}$ = .44, and the nonprofit children condition, F(1, 458) = 293.79, p < .001, $η_{p}^{2}$ = .39, than the commercial condition, F(1, 458) = 174.14, p < .001, $η_{p}^{2}$ = .28. This indicates that both nonprofits experienced sharper declines in trust post-transgression compared with the commercial organization: and that the nonprofit with a mission to protect women from sexual exploitation experienced the sharpest declines in trust.

To test the focal hypotheses, we then ran a 2 (Time) × 3 (Organization Type) ANOVA, this time comparing scores post-transgression with scores post-response. The interaction was nonsignificant, F(2, 458) = 1.11, p = .332, $η_{p}^{2}$ = .01, confirming that the effects of the apology were statistically equivalent regardless of type of organization. Thus, results supported the weaker version of the moral disillusionment hypothesis: Nonprofits were punished more harshly than commercial organizations for the same transgression, and this discrepancy was not reduced after an apology.

Consumer intentions

A main effect of time, F(2, 916) = 437.35, p < .001, $η_{p}^{2}$ = .49, reflected the fact that participants’ consumer intentions dropped from pre-transgression (M = 5.08, SD = 0.97) to post-transgression (M = 3.23, SD = 1.59) but partially recovered post-apology (M = 3.72, SD = 1.55). All comparisons were significant at p < .001.

Again, however, these effects were qualified by a significant Time × Organization Type interaction, F(4, 916) = 5.98, p < .001, $η_{p}^{2}$ = .03, see Figure 3. As for the trust analyses, we first examined whether we could replicate the findings of Hornsey and colleagues (2020). A 2 (Time) × 3 (Organization Type) ANOVA comparing scores pre-transgression with scores post-transgression revealed a significant interaction, F(2, 458) = 7.98, p < .001, $η_{p}^{2}$ = .03. Simple main effects of time were significant in all three conditions, and the trend in the effect sizes was the same as observed for trust. Specifically, effect sizes were greatest in the nonprofit women condition, F(1, 458) = 274.68, p < .001, $η_{p}^{2}$ = .38, followed by the nonprofit children condition, F(1, 458) = 202.19, p < .001, $η_{p}^{2}$ = .31, followed by the commercial condition, F(1, 458) = 140.73, p < .001, $η_{p}^{2}$ = .24. These results indicate that nonprofits experienced greater losses in consumer intentions after a transgression, and this difference was accentuated when the transgression was relevant to the nonprofit’s mission.

Figure 3.

Changes in consumer intentions over time as a function of organization type (Study 1).

To test the focal hypotheses, we then ran a 2 (Time) × 3 (Organization Type) ANOVA comparing scores post-transgression with scores post-response. As for trust, the interaction was nonsignificant, F(2, 458) = 0.82, p = .440, $η_{p}^{2}$ < .01, confirming that the effects of the apology did not differ across organization type. Thus, the pattern of results observed on consumer intentions was consistent with the moral disillusionment hypothesis.

Overall, results indicate that apologies are effective. For all types of organizations, trust and consumer intentions rebounded after exposure to an organizational apology. However, this recovery was not complete: All organizations finished at lower levels of trust and consumer intentions than before the transgression.

Moving to the focal hypotheses, the results are broadly consistent with the weaker version of the moral disillusionment hypothesis (H2). There was no difference in the effectiveness of trust repair for the nonprofits and commercial organizations. Nonprofit organizations experienced sharper declines in trust and consumer intentions immediately after an ethical transgression, especially when the transgression violated the nonprofit’s stated mission. However, the gains in trust and consumer intentions after apologizing were equivalent regardless of the type of organization that transgressed. Results suggest that past good deeds do not allow nonprofits to recover more effectively from scandal (i.e., no support for H1, the moral insurance hypothesis). Instead, moral disillusionment is unlikely to be easily overcome.

Study 2

To replicate the focal effect of nonprofit status on the effectiveness of organizational apologies, we ran a second study. Study 2 was designed to again test the competing hypotheses that nonprofits may repair trust more easily (H1) or repair trust either the same or to a lesser extent than commercial organizations (H2) after a transgression. To extend the generalizability of our conclusions and enhance the theoretical contributions of this research, we introduced two modifications to the design of Study 2. First, we used a different type of transgression to test the generalizability of the process. Second, we tested the relative effects of two types of trust repair responses. As in Study 1, we tested an apology in which the organization accepted responsibility for the transgression. In Study 2, we also tested a “stonewalling” response in which the organization acknowledged the events but admitted no responsibility and did not apologize. Including the “stonewalling” response provided a baseline against which apologies could be compared. Consideration of this additional organizational response also broadens our understanding of how different types of trust repair strategies are received depending on whether they come from a nonprofit or a commercial organization.

Study 2 also introduced a number of methodological improvements to enhance the robustness of conclusions. First, we conducted a priori power analysis to confirm sufficient power to detect effects. Second, we used a better sampling platform to retain more participants (having lost 17% through comprehension check in Study 1). Third, we added three new dependent measures to better understand the various ways consumers may evaluate organizations that transgress: word of mouth intentions, collective blame, and forgiveness. Finally, having already replicated the original moral disillusionment effect in Study 1, we simplified the design to home in on the trust repair part of the process.

Method

We report how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study (see best practice recommendations from Simmons et al., 2012). All materials, data, syntax, and supplementary analyses are available on the OSF.

Participants and procedure

In total, 362 Americans took part in an online experiment. Participants were recruited through the Prolific website and were paid £1.20 for completing the 8-min study. The sample size was determined by an a priori power analysis based on 90% power to detect a small-to-medium effect (f = .15), plus a 10% buffer for planned exclusions. In total, 18 participants were excluded prior to analysis: 13 for failing the attention check (i.e., “To ensure you are still reading this, please select option 2 [Disagree]”), and five for failing one of the two comprehension checks. The final sample (N = 344) consisted of 152 men, 191 women, and one nonbinary person. Participants ranged in age from 18 to 79 years (M_age = 35.33 years, SD = 12.61 years).

After completing demographics, participants were presented with information about an organization (nonprofit vs. commercial) and then told about an organizational transgression, whereby the organization’s CEO had been found guilty of fraud. Participants then indicated how much they trusted the organization, as well as their consumer and word of mouth intentions. Next, participants read about the organization’s official response (apology vs. stonewalling), and indicated how much this type of response met their expectations. After reading the response, participants rated the organization’s perceived remorse, blame, and likelihood to commit a similar act in the future. Participants again indicated their trust, consumer, and word of mouth intentions, as well as their forgiveness. Finally, participants were fully debriefed.

Design

A 2 (Organization Type: Nonprofit, Commercial) × 2 (Response Type: Apology, Stonewalling) × 2 (Time: Post-transgression, Post-response) mixed-factorial design was employed. Organization and response types were both between-subjects factors, and time was within-subjects.

Manipulations

Organization type

Participants read about GlobalDress, the same online fashion store as in Study 1. The descriptions were slightly adapted from Study 1, to remove any potential confounds in terms of length of the description or international reach of the organization. The description of the commercial organization was similar to that used in Study 1. In contrast, the description of the nonprofit organization highlighted that GlobalDress was “a charitable, nonprofit online-only fashion store, which sells clothing to raise money for people in need.” Then, participants read that “since its inception, 100% of GlobalDress profits have been donated to charity.”

Transgression

All participants were then exposed to the same transgression, an instance of fraud by the CEO:

CEO of GlobalDress, Daniel Field, has been found guilty of fraud, falsifying documents, misappropriating funds, and giving false information.

Mr Field was brought to court in July last year after inconsistencies in accounts emerged in late 2017. Investigations showed that Field signed a fraudulent voucher of $50,000, ostensibly used to purchase clothing materials for the company, but in reality was intended for his personal use.

The prosecution revealed in earlier court proceedings that Field had purchased property, a horse, a luxury car, and a golf-club membership in Australia over the years. He had also taken a $253,000 loan from GlobalDress in 2016 to purchase shares under his name. The money was not repaid until auditors looked into GlobalDress’s affairs.

Response type

Participants then read about the organization’s response to the scandal. Half of participants were randomly allocated to read a stonewalling response, where the organization acknowledged the violation but took no responsibility (see Dawar & Pillutla, 2000):

You may have read or heard that recently, Daniel Field has been found to have misappropriated company funds. We at GlobalDress do not condone nor support such acts. The executive committee was not aware that this was occurring until a recent internal audit.

Mr Field no longer works for GlobalDress.

Again, we were not aware of this misappropriation of funds. This will not affect our services and we will continue in our endeavor to provide our customers with the highest quality of service.

The other half of participants read a full apology from the organization:

You may have read or heard that recently, a member of our executive committee has been found to have misappropriated company funds. We assume responsibility for this failure and sincerely apologize. The executive committee was not aware that this was occurring until a recent internal audit.

We have damaged your trust by failing in our diligence. We, the Directors of GlobalDress, promise to find out exactly how this could have happened and we promise that this will never happen again.

This will not affect our services and we will continue in our endeavor to provide our customers with the highest quality of service.

Measures

Below we report focal measures for the current analysis and provide sample items for established scales.

Trust was measured as per Study 1. This scale was reliable at both time points (α_{post-transgression} = .84, α_{post-response} = .92).

Consumer intentions were also measured as per Study 1 and found to be reliable (α_{post-transgression} = .92, α_{post-response} = .94).

Word of mouth intentions were measured using the single-item Net Promoter Index (Reichheld, 2003). The item is “How likely is it that you would recommend GlobalDress to a friend or colleague?”: 0 = Not at all likely, 10 = Extremely likely.

Collective blame

Three items were used to measure how much the organization as a whole should be considered responsible for the CEO’s fraud (e.g., “GlobalDress as a whole should be accountable for what happened”; α = .82).

Forgiveness

Four items were adapted from Wohl and Branscombe (2005) to measure the degree to which participants were willing to forgive the organization (e.g., “I forgive GlobalDress for the misuse of company funds”; α = .93).

Results and Discussion

A series of 2 (Organization Type) × 2 (Response Type) × 2 (Time) mixed-factorial ANOVAs were run to assess the effects of the manipulations on trust, consumer intentions, and word of mouth intentions.

Trust

A strong main effect was found for time: Participants trusted GlobalDress more after reading the organization’s response (M = 3.95, SD = 1.34) than after reading about the transgression (M = 2.55, SD = 1.10), F(1, 340) = 410.79, p < .001, $η_{p}^{2}$ = .55. There were, however, no main or interactive effects of either organization type or response type on trust, Fs < 1.93, ps > .166. Furthermore, time did not moderate the effects of organization type or response, nor was there a three-way interaction, Fs < 1.22, ps > .269. Thus, consumers responded equally to organizations who attempted to repair trust after an ethical transgression, regardless of whether the organization was a nonprofit or commercial entity, and regardless of the type of trust repair strategy employed. These results are summarized in Figure 4.

Figure 4.

Changes in trust over time as a function of organization and response types (Study 2).

Consumer intentions

A main effect of time was found, F(1, 340) = 178.95, p < .001, $η_{p}^{2}$ = .35. Participants reported improved consumer intentions after reading the organization’s response (M = 4.12, SD = 1.48) compared with directly after reading about the transgression (M = 3.32, SD = 1.46). A main effect of organization type was also found, with participants overall reporting greater consumer intentions for the commercial organization that had committed fraud (M = 4.00) than for the nonprofit (M = 3.43), F(1, 340) = 16.12, p < .001, $η_{p}^{2}$ = .05. This main effect was qualified by a Time × Organization Type interaction, F(1, 340) = 5.43, p = .020. Participants reported improved consumer intentions toward both kinds of organizations after their trust repair response, but—in line with the moral insurance hypothesis—this improvement was larger for nonprofits (M_{post-transgression} = 2.95, M_{post-response} = 3.90), F(1, 340) = 121.25, p < .001, $η_{p}^{2}$ = .26, than for commercial organizations (M_{post-transgression} = 3.67, M_{post-response} = 4.34), F(1, 340) = 62.09, p < .001, $η_{p}^{2}$ = .15. Although nonprofits gained more from trust repair in terms of consumer intentions, this finding must be contextualized: The nonprofit started in a significantly worse position after the transgression and, despite their sharper gains in consumer intentions, the nonprofit did not recover to the same level as the commercial organization after response (see Figure 5). This pattern is therefore consistent with what was found in Study 1. No other significant main or interactive effects were found on consumer intentions, Fs < 1.16, ps > .282.

Figure 5.

Changes in consumer intentions over time as a function of organization and response types (Study 2).

Word of mouth intentions

Again, a significant main effect of time was found, F(1, 340) = 176.91, p < .001, $η_{p}^{2}$ = .34, with participants reporting greater willingness to spread positive word of mouth about the organization after reading the response (M = 4.64, SD = 2.67) compared with after the transgression (M = 3.29, SD = 2.30). Participants overall reported greater word of mouth intentions when GlobalDress was described as a commercial organization (M = 4.23) than when it was described as a nonprofit (M = 3.70), F(1, 340) = 4.57, p = .033, $η_{p}^{2}$ = .01. No other significant main or interactive effects were observed, Fs < 1.88, ps > .172 (see Figure 6). Thus, in terms of their willingness to spread positive word of mouth, consumers showed no differential response to trust repair efforts by nonprofit and commercial organizations who had transgressed. This again reinforces the pattern found in Study 1.

Figure 6.

Changes in word-of-mouth intentions over time as a function of organization and response types (Study 2).

Collective blame

Because collective blame was only measured after the post-transgression response, scores were analyzed using 2 (Organization Type) × 2 (Response Type) between-subjects ANOVA. Overall, participants blamed the organization more for the transgression when it was described as a nonprofit (M = 5.44, SD = 1.15) than when it was described as a commercial organization (M = 5.10, SD = 1.17), F(1, 340) = 7.70, p = .006, $η_{p}^{2}$ = .02. There was also a marginal effect of response type, whereby the apology was perceived to signal more blame (M = 5.38, SD = 1.14) than the stonewalling response (M = 5.15, SD = 1.19), F(1, 340) = 3.58, p = .059, $η_{p}^{2}$ = .01. No significant interaction between organization and response type was found for blame, F(1, 340) = 0.07, p = .789.

Forgiveness

A 2 (Organization Type) × 2 (Response Type) ANOVA revealed that participants were marginally less likely to forgive the nonprofit (M = 4.09, SD = 1.48) than the commercial organization (M = 4.38, SD = 1.27), F(1, 340) = 3.82, p = .051, $η_{p}^{2}$ = .01. No difference in forgiveness was found for type of response, and no interaction between type of organization and response was found, Fs < 1.71, ps > .192.

In sum, Study 2 improved the design over Study 1 in a number of ways and replicated the key findings. Again, trust repair efforts were effective in partially regaining losses of trust, consumer intentions, and word of mouth intentions following an ethical transgression. We found no evidence that apologies were more effective responses than responses that acknowledged the transgression without formally apologizing.

Turning to the focal hypotheses, we found little difference in the recovery of trust or word of mouth intentions between nonprofits and commercial organizations after an organizational response, indicating an extension of the moral disillusionment effect (supporting H2). Nonprofits were also blamed more and were marginally less likely to be forgiven, than commercial organizations that had committed the same transgression. Together, results support H2 (but not H1): Nonprofits have no moral advantage in terms of recovery from transgression after trust repair efforts.

General Discussion

Overall, the results do not paint a happy picture for nonprofits that find themselves embroiled in scandal. Nonprofits will be punished more harshly than commercial organizations for ethical missteps and will not have an easier time repairing damaged trust. Trust can be repaired to some extent when organizations either directly apologize or merely acknowledge the violation without taking responsibility. However, nonprofits cannot easily win back their initial trust advantage over commercial organizations once they have lost it. In Study 2, nonprofits regained consumer intentions more effectively than commercial organizations following an apology. However, the nonprofit started at a far lower level of trust after the transgression and the improvement did not see the nonprofit surpass or even equal consumer intentions toward the commercial organization. Furthermore, this moderation by organization type only emerged on one of the five dependent measures. Overall, this pattern of results confirms and extends the MDM prediction of continued consumer punishment of nonprofits who commit ethical violations (Hornsey et al., 2020) and supports H2 (vs. H1) of the current study.

Results are partly explained by the fact that consumers hold nonprofits to higher standards than they hold commercial organizations (see also Hornsey et al., 2020; Witesman & Fernandez, 2012). Results show that consumers believe the overall organization is more to blame for a transgression when that organization is a nonprofit. Consumers are also less likely to forgive a nonprofit that transgresses.

It is not all bad news, however. Results clearly indicate that trust repair strategies are effective. After reading organizational responses, all consumers showed increased trust, consumer, and word of mouth intentions. Apologies were found to be no more or less effective than responses that acknowledged the violation without apologizing or taking responsibility. Thus, trust repair efforts appear to be effective, and apologies can help rebound consumer sentiment following a scandal (echoing, for example, Boshoff & Leong, 1998; Martin & Smart, 1994).

Past evidence has shown that apologies can be effective trust repair strategies for commercial organizations (Gillespie et al., 2014; Hegner et al., 2016; Xie & Peng, 2009). Until now, however, no study has directly examined the effects of apologies for nonprofit transgressors. Here, we find that apologies are just as effective for nonprofits as they are for commercial organizations. We can therefore add apologies to the repertoire of effective trust repair strategies for nonprofits that have previously been evidenced, including denial, diminishing the violation, positive social media campaigns, and celebrity endorsement (Hou et al., 2020; Scurlock et al., 2020).

Strengths, Limitations, and Future Directions

The experimental design employed in the current research allows us to directly consider the causality of attributes. This design allows us to conclude (a) that it is the nonprofit status of an organization that results in greater trust losses and (b) that neither the type of organization nor its chosen response affect trust repair. By considering two types of transgressions (sexual exploitation and fraud) and two nonprofit business models (social enterprise in Study 1 and pure nonprofit in Study 2), we gain confidence in the generalizability of findings. Further confidence is gained by our replication of findings across two well-powered studies. In both studies, we considered how organization type affects trust repair trends alongside other factors: domain of transgression in Study 1 and type of response in Study 2. We see this as a theoretical strength of the current research, as it helps articulate the conditions under which moral disillusionment effects do and do not emerge. Nonetheless, future research may benefit from employing streamlined designs that test each independent variable separately.

Brand recognition could also be an important moderator of the effects we tested here. For example, people may trust a reputable organization over one that they had never heard of, and this may result in greater responsivity to trust repair efforts. We encourage future research to consider using real-world organizations and to examine the moderating effects of brand recognition.

We have examined the direct effects of nonprofit status during trust repair efforts. We found that moral disillusionment effects continue even after an organizational apology is issued. A fruitful avenue for future research would be to also examine indirect effects to identify the psychological mechanisms through which moral disillusionment continues to impact consumer reactions to nonprofits after scandals.

Finally, we focused specifically on the impact of apologies. Given our findings show no difference in the impact of apologies and stonewalling responses, future research may wish to consider testing other forms of trust repair. Other forms of crisis response and trust repair strategies that may warrant examination include denial, distancing, ingratiation, and even no response (Coombs, 1995).

Managerial Implications

Nonprofits must be extra vigilant because they have more to lose from transgressions than commercial organizations that behave in exactly the same way. Compared with commercial organizations, results show that nonprofits will lose more trust and suffer greater losses of consumer intentions following a scandal, and that these losses may be difficult for them to fully recover from. Nonprofit managers must therefore take proactive steps to prevent transgressions and ensure the organization is designed to reliably produce trustworthy behavior (see Hurley et al., 2013, for guidelines). Strategies for designing trustworthy organizations include ensuring effective governance measures are in place, such as background checks on staff, board accountability, internal controls, and auditing (see also Gibelman & Gelman, 2001).

Results also show that there are measurable benefits for organizations that acknowledge their mistakes, including those that apologize directly. Although nonprofits get no added advantage in terms of trust repair compared with commercial organizations, results show that efforts to repair damaged consumer trust will be effective. Nonprofit managers should be willing to publicly acknowledge any organizational mistake, as the process of acknowledging bad behavior appears to be a key step in repairing damaged trust. Because nonprofits suffer more from scandals, they must also be more willing to take proactive action to repair trust following ethical breaches.

Footnotes

Authors’ Note

Both studies received ethical clearance from the University of Queensland (clearance number 2019000131).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The current research was funded by a Discovery grant awarded by the Australian Research Council to the second and fourth authors (DP170103856).

ORCID iD

Cassandra M. Chapman

Notes

Author Biographies

Cassandra M. Chapman is a lecturer in marketing at the University of Queensland researching the psychology of charitable giving and nonprofit scandals.

Matthew J. Hornsey is a professor of management at the University of Queensland researching social influence.

Heidi Mangan is a masters of organizational psychology graduate from the University of Queensland.

Nicole Gillespie is professor of management at the University of Queensland whose research focuses on trust in organizations and institutions.

Stephen La Macchia is a lecturer in marketing at Deakin University whose research focuses on consumer trust and decision-making.

Steven Lockey is a postdoctoral research fellow in organizational trust at the University of Queensland whose research interests primarily relate to the development and repair of trust.

References

Barnett

M. L.

Hartmann

Salomon

R. M.

(2018). Have you been served? Extending the relationship between corporate social responsibility and lawsuits. Academy of Management Discoveries, 4(2), 109–126. https://doi.org/10.5465/amd.2015.0030

Becker

(2018). An experimental study of voluntary nonprofit accountability and effects on public trust, reputation, perceived quality, and donation behavior. Nonprofit and Voluntary Sector Quarterly, 47, 562–582. https://doi.org/10.1177/0899764018756200

Bekkers

(2003). Trust, accreditation, and philanthropy in the Netherlands. Nonprofit and Voluntary Sector Quarterly, 32, 596–615. https://doi.org/10.1177/0899764003258102

Boshoff

Leong

(1998). Empowerment, attribution and apologising as dimensions of service recovery. International Journal of Service Industry Management, 9(1), 24–47. https://doi.org/10.1108/09564239810199932

Bundy

Pfarrer

M. D.

(2015). A burden of responsibility: The role of social approval at the onset of a crisis. Academy of Management Review, 40(3), 345–369. https://doi.org/10.5465/amr.2013.0027

Burgoon

J. K.

Le Poire

B. A.

(1993). Effects of communication expectancies, actual communication, and expectancy disconfirmation on evaluations of communicators and their communication behavior. Human Communication Research, 20(1), 67–96. https://doi.org/10.1111/j.1468-2958.1993.tb00316.x

Chandler

Polidoro

Yang

(2020). When is it good to be bad? Contrasting effects of multiple reputations for bad behavior on media coverage of serious organizational errors. Academy of Management Journal, 63(4), 1236–1265. https://doi.org/10.5465/amj.2017.1248

Chung

(2018). Examining the effectiveness of using CSR communication in apology statements after bad publicity. Corporate Communications: An International Journal, 23, 357–376. https://doi.org/10.1108/CCIJ-06-2017-0055

Coombs

W. T.

(1995). Choosing the right words: The development of guidelines for the selection of the “appropriate” crisis-response strategies. Management Communication Quarterly, 8(4), 447–476. https://doi.org/10.1177/0893318995008004003

10.

Coombs

W. T.

Holladay

S. J.

(2008). Comparing apology to equivalent crisis response strategies: Clarifying apology’s role and value in crisis communication. Public Relations Review, 34(3), 252–257.

11.

Davidow

(2000). The bottom line impact of organizational responses to customer complaints. Journal of Hospitality & Tourism Research, 24(4), 473–490. https://doi.org/10.1177/109634800002400404

12.

Dawar

Pillutla

M. M.

(2000). Impact of product-harm crises on brand equity: The moderating role of consumer expectations. Journal of Marketing Research, 37(2), 215–226. http://www.jstor.org.ezproxy.library.uq.edu.au/stable/1558501

13.

de Ruyter

Wetzels

. (2000). Customer equity considerations in service recovery: A cross-industry perspective. International Journal of Service Industry Management, 11(1), 91–108. https://doi.org/10.1108/09564230010310303

14.

Dirks

K. T.

Lewicki

R. J.

Zaheer

(2009). Repairing relationships within and between organizations: Building a conceptual foundation. Academy of Management Review, 34(1), 68–84.

15.

Flammer

(2013). Corporate social responsibility and shareholder reaction: The environmental awareness of investors. Academy of Management Journal, 56(3), 758–781. https://doi.org/10.5465/amj.2011.0744

16.

Gaskin

(1999). Blurred vision: Public trust in charities. International Journal of Nonprofit and Voluntary Sector Marketing, 4(2), 163–178. https://doi.org/10.1002/nvsm.66

17.

Gibelman

Gelman

(2001). Very public scandals: Nongovernmental organizations in trouble. Voluntas, 12(1), 49–66.

18.

Gillespie

Dietz

(2009). Trust repair after an organization-level failure. Academy of Management Review, 34(1), 127–145.

19.

Gillespie

Dietz

Lockey

(2014). Organizational reintegration and trust repair after an integrity violation: A case study. Business Ethics Quarterly, 24(3), 371–410. https://doi.org/10.5840/beq2014437

20.

Godfrey

P. C.

Merrill

C. B.

Hansen

J. M.

(2009). The relationship between corporate social responsibility and shareholder value: An empirical test of the risk management hypothesis. Strategic Management Journal, 30(4), 425–445. https://doi.org/10.1002/smj.750

21.

Hegner

S. M.

Beldad

A. D.

Kraesgenberg

A.-L.

(2016). The impact of crisis response strategy, crisis type, and corporate social responsibility on post-crisis consumer trust and purchase intention. Corporate Reputation Review, 19(4), 357–370. https://doi.org/10.1057/s41299-016-0007-y

22.

Hind

(2017). New development: Fundraising in UK charities—Stepping back from the abyss. Public Money & Management, 37(3), 205–210. https://doi.org/10.1080/09540962.2017.1282238

23.

Hornsey

M. J.

Chapman

C. M.

Mangan

La Macchia

Gillespie

(2020). The moral disillusionment model of organizational transgressions: Ethical transgressions trigger more negative reactions from consumers when committed by nonprofits. Journal of Business Ethics, 172, 653–671. https://doi.org/10.1007/s10551-020-04492-7

24.

Hou

Zhang

Guo

(2020). How nonprofits can recover from crisis events? The trust recovery from the perspective of causal attributions. Voluntas: International Journal of Voluntary and Nonprofit Organizations, 31(1), 71–93. https://doi.org/10.1007/s11266-019-00176-7

25.

Hurley

R. F.

Gillespie

Ferrin

D. L.

Dietz

(2013). Designing trustworthy organizations. MIT Sloan Management Review, 54(4), 75–82.

26.

Janney

J. J.

Gove

(2011). Reputation and corporate social responsibility aberrations, trends, and hypocrisy: Reactions to firm choices in the stock option backdating scandal. Journal of Management Studies, 48(7), 1562–1585. https://doi.org/10.1111/j.1467-6486.2010.00984.x

27.

Janssen

Sen

Bhattacharya

C. B.

(2015). Corporate crises in the age of corporate social responsibility. Business Horizons, 58(2), 183–192. https://doi.org/https://doi.org/10.1016/j.bushor.2014.11.002

28.

Kiambi

D. M.

Shafer

(2016). Corporate crisis communication: Examining the interplay of reputation and crisis response strategies. Mass Communication and Society, 19(2), 127–148. https://doi.org/10.1080/15205436.2015.1066013

29.

Kim

Ferrin

Cooper

Dirks

(2004). Removing the shadow of suspicion: The effects of apology versus denial for repairing competence-versus integrity-based trust violations. Journal of Applied Psychology, 89(1), 104–118.

30.

Kramer

R. M.

Lewicki

R. J.

(2010). Repairing and enhancing trust: Approaches to reducing organizational trust deficits. Academy of Management Annals, 4, 245–277.

31.

Lewicki

R. J.

Brinsfield

(2017). Trust repair. Annual Review of Organizational Psychology and Organizational Behavior, 4(1), 287–313. https://doi.org/10.1146/annurev-orgpsych-032516-113147

32.

Lewicki

R. J.

Polin

. (2012). The art of the apology. In Kramer

R. M.

Pittinsky

T. L.

(Eds.), Restoring trust in organizations and leaders: Enduring challenges and emerging answers. Oxford Scholarship Online. https://doi.org/10.1093/acprof:oso/9780199756087.001.0001

33.

Lewicki

R. J.

Polin

Lount

R. B.

Jr. (2016). An exploration of the structure of effective apologies. Negotiation and Conflict Management Research, 9(2), 177–196. https://doi.org/10.1111/ncmr.12073

34.

Lin-Hi

Hrisch

Blumberg

(2015). Does CSR matter for nonprofit organizations? Testing the link between CSR performance and trustworthiness in the nonprofit versus for-profit domain. Voluntas, 26(5), 1944–1974. https://doi.org/10.1007/s11266-014-9506-6

35.

Martin

C. L.

Smart

D. T.

(1994). Consumer experiences calling toll-free corporate hotlines. Journal of Business Communication, 31(3), 195–212. https://doi.org/10.1177/002194369403100303

36.

Nakayachi

Watabe

(2005). Restoring trustworthiness after adverse events: The signaling effects of voluntary “Hostage Posting” on trust. Organizational Behavior and Human Decision Processes, 97(1), 1–17. https://doi.org/10.1016/j.obhdp.2005.02.001

37.

Reichheld

(2003). The one number you need to grow. Harvard Business Review, 81(12), 46–54.

38.

Schlenker

B. R.

Darby

B. W.

(1981). The use of apologies in social predicaments. Social Psychology Quarterly, 44(3), 271–278. https://doi.org/10.2307/3033840

39.

Scurlock

Dolsak

Prakash

(2020). Recovering from scandals: Twitter coverage of Oxfam and Save the Children scandals. Voluntas, 31(1), 94–110. https://doi.org/10.1007/s11266-019-00148-x

40.

Simmons

J. P.

Nelson

L. D.

Simonsohn

(2012). A 21 word solution. SSRN. https://doi.org/10.2139/ssrn.2160588

41.

Sohn

Y. J.

Lariscy

R. W.

(2015). A “buffer” or “boomerang?”—The role of corporate reputation in bad times. Communication Research, 42(2), 237–259. https://doi.org/10.1177/0093650212466891

42.

Sullivan

(2015, June 3). In search of the Red Cross’ $500 million in Haiti relief. NPR. https://www.npr.org/2015/06/03/411524156/in-search-of-the-red-cross-500-million-in-haiti-relief

43.

Vanhamme

Grobben

(2008). “Too good to be true!” The effectiveness of CSR history in countering negative publicity. Journal of Business Ethics, 85(2), 273. https://doi.org/10.1007/s10551-008-9731-2

44.

Vanhamme

Swaen

Berens

Janssen

(2015). Playing with fire: Aggravating and buffering effects of ex ante CSR communication campaigns for companies facing allegations of social irresponsibility. Marketing Letters, 26(4), 565–578. https://doi.org/10.1007/s11002-014-9290-5

45.

Williams

R. J.

Barrett

J. D.

(2000). Corporate philanthropy, criminal activity, and firm reputation: Is there a link? Journal of Business Ethics, 26(4), 341–350. https://doi.org/10.1023/A:1006282312238

46.

Witesman

E. M.

Fernandez

(2012). Government contracts with private organizations: Are there differences between nonprofits and for-profits? Nonprofit and Voluntary Sector Quarterly, 42(4), 689–715. https://doi.org/10.1177/0899764012442592

47.

Wohl

M. J. A.

Branscombe

N. R.

(2005). Forgiveness and collective guilt assignment to historical perpetrator groups depend on level of social category inclusiveness. Journal of Personality and Social Psychology, 88(2), 288–303. https://doi.org/10.1037/0022-3514.88.2.288

48.

Xie

Peng

(2009). How to repair customer trust after negative publicity: The roles of competence, integrity, benevolence, and forgiveness. Psychology and Marketing, 26(7), 572–589. https://doi.org/10.1002/mar.20289