- Email: [email protected]
Comparing apology to equivalent crisis response strategies: Clarifying apology's role and value in crisis communication
Public Relations Review 34 (2008) 252–257
Contents lists available at ScienceDirect
Public Relations Review
Comparing apology to equivalent crisis response strategies: Clarifying apology’s role and value in crisis communication W. Timothy Coombs ∗ , Sherry J. Holladay 1 Communication Studies, Eastern Illinois University, 600 Lincoln Avenue, Charleston, IL 61920, United States
a r t i c l e
i n f o
Article history: Received 24 February 2008 Accepted 11 April 2008 Keywords: Apology Crisis communication Crisis response strategies
a b s t r a c t Crisis communication has begun to systematically examine how people perceive the crisis response strategies using experimental designs. This is an important step in the evolution of crisis communication research. One problem with this research has been the over promotion of apology as “the” response. Too often the value of an apology is established by comparing how people react to an apology and to other less victim-centered/accommodative responses. This “unfair” comparison results in apology appearing to be “the” best alternative. This study compares apology to more equivalent crisis response strategies to more fairly determine if apology is “the” best strategy. The results show that people react similarly to any victim-centered/accommodative strategy meaning apology is not “the” best strategy. The implications of the results for crisis managers are discussed. © 2008 Elsevier Inc. All rights reserved.
After a crisis, organizations need to convey messages to stakeholders. These messages begin by telling stakeholders what to do to protect themselves from the crisis (instructing information) and to help them cope psychologically with the crisis (adjusting information). The next step is to address the reputational threat posed by the crisis (Sturges, 1994). Reputations are valuable resources that are threatened during a crisis (Davies, Chun, da Silva, & Roper, 2003; Fombrun & van Riel, 2003). Most post-crisis communication research has utilized case study methods. While providing useful descriptive data, case studies offer little insight into how stakeholders actually respond to crisis response strategies. Moreover, the case studies often provide minimal theoretical insight into crisis communication (Dawar & Pillutla, 2000; Dean, 2004). Crisis communication research has begun moving beyond case studies to experiments designed to assess systematically how people perceive crisis response strategies. Experimental research has begun to explore how crisis response strategies affect stakeholder perceptions and responses to the organizations in crisis and to build crisis communication theory (e.g., Arpan & Roskos-Ewoldsen, 2005; Coombs & Holladay, 1996; Dean, 2004; Huang, Lin, & Su, 2005). This experimental research is essential if we are to build evidence-based knowledge for crisis management. Evidence-based management is derived from evidence-based medicine and argues for the use of scientifically proven results to guide actions in medicine and now management (Rousseau, 2005). We need to move beyond the speculation of cases based on media reports if we are to create evidence-based crisis communication (Coombs, 2007). The empirical evidence for the effects of crisis response strategies is still in its infancy and has room for development. One area of potential growth is understanding the effects of different crisis response strategies utilized in the same crisis. Such comparisons can help crisis managers make choices when they face particular crises. However, problems arise when researchers offer comparisons of crisis communication strategies that “unfairly” pit extremely diverse strategies against one another. For example, apology is sometimes compared to responses such as denial or excuse that do little, if anything,
∗ Corresponding author. Tel.: +1 217 581 3324; fax: +1 217 581 5718. E-mail addresses: [email protected] (W.T. Coombs), [email protected] (S.J. Holladay). 1 Tel.: +1 217 581 6311; fax: +1 217 581 5718. 0363-8111/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.pubrev.2008.04.001
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
253
to address the concerns of victims. Such biased comparisons allow apology, a victim-centered/accommodative response, to easily “win” and provides a questionable foundation for its recommended use to crisis managers. Apology should be compared to more equivalent crisis response strategies to establish a more realistic assessment of its value in crisis communication. This study compares apology to other accommodative crisis response strategies in an effort to produce a richer and more rounded view of apology in crisis communication. 1. Literature review and research questions Research has shown that crisis response strategies – what an organization says and does after a crisis – serve to protect a reputation after a crisis (e.g., Coombs & Holladay, 1996). Researchers have over-emphasized the use of apology/mortification as the “best” crisis response (e.g., Benoit, 1995; Benoit & Drew, 1997) and have used widely varying definitions of apology (Patel & Reinsch, 2003). At its core, an apology is marked by the organization accepting responsibility for the crisis and asking for forgiveness (Benoit & Drew, 1997; Fuchs-Burnett, 2002). A variety of additional components can be added to this definition including expression of remorse/sympathy, expression of regret, preventative measures, and reparation (Benoit & Drew, 1997; Cohen, 1999; Fuchs-Burnett, 2002; Patel & Reinsch, 2003). Accepting responsibility is the centerpiece of an apology and makes it the most expensive response financially for an organization (Fuchs-Burnett, 2002; Patel & Reinsch, 2003; Tyler, 1997). When an organization offers an apology it opens itself to lawsuits and financial loss. The apology is used as evidence in court to win lawsuits against the organization. However, some evidence indicates that compensation and sympathy, two less expensive strategies, are as effective as an apology in shaping people’s perceptions of the organization taking responsibility for the crisis because these strategies focus on victims’ needs. The sympathy response expresses concern for victims while compensation offers victims something to offset the suffering. These are referred to as highly accommodative strategies (Coombs, 2006; Fediuk, 2002). If other highly accommodative response strategies are as effective as apology in protecting reputation, then this has important implications for the options crisis managers might pursue in protecting their reputations after a crisis. Previous experimental research that has established the value of apologies in crisis management has stacked the deck in favor of apologies. The effects of the apology strategy were compared to less accommodative strategies such as giving no comment, denial, excuse, or justification (Bradford & Garrett, 1995; Dean, 2004; Lyon & Cameron, 1998). It is not surprising that two of the studies found apology to be the preferred strategy, meaning the apology strategy had the strongest positive effect on perceptions of an organization’s reputation (Bradford & Garrett, 1995; Dean, 2004). We know little about how apology compares to other accommodative strategies such as compensation and expression of sympathy. One study that did compare apology and compensation used in a service failure situation (bad service at a restaurant) not a crisis. The service failure situation found apology to be more effective at increasing satisfaction (Wirtz & Mattila, 2003). While an expression of sympathy is more of an adjusting information strategy than a reputation repair strategy, it reflects a high level of accommodation (Sturges, 1994). To insure that the effects documented in this study were not created by simply offering any organizational response, the information only response was included. It is common for crisis managers to report only what has happened, a type of what Sturges (1994) calls instructing information. Previous research has found little effect for information only responses on people’s post-crisis perceptions (Coombs, 1998). The first research question in our study examines the effect of apology, compensation, sympathy, and information only on post-crisis reputations. RQ1: Do compensation, apology, and sympathy crisis responses produce similar post-crisis reputation evaluations and do they differ from information only response? Other important outcomes stem from a crisis including anger and negative word-of-mouth. Along with attributions, a crisis will produce affective responses (Coombs & Holladay, 2005). Stakeholder anger can damage the organization–stakeholder relationship or even lead stakeholders to end the relationship (Jorgensen, 1996; Stockmyer, 1996). Anger can also motivate stakeholders to say bad things about the organization to people they know—negative word-of-mouth (Coombs & Holladay, 2007). Organizations value word-of-mouth because of the benefits associated with positive word-of-mouth and the harm inflicted by negative word-of-mouth (Herr, Kardes, & Kim, 1991; Laczniak, DeCarlo, & Ramaswami, 2001). Crisis response strategies can affect the anger and negative word-of-mouth intentions created by a crisis. RQ2: Do compensation, apology, and sympathy crisis responses produce similar reports of anger and negative word-ofmouth and do they differ from information only response? Account acceptance refers to how respondents feel about the crisis response offered by the organization. Greater account acceptance indicates respondents believe the crisis response is appropriate. Any difference in how crisis response strategies affect reputation could be a result of differing levels of account acceptance. RQ3: Do compensation, apology, and sympathy crisis responses produce similar levels of account acceptance and do they differ from information only response?
254
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
2. Methods 2.1. Participants Participants in the study were 167 undergraduate students from a Midwestern university. The ages ranged from 18 to 56 (M = 20, S.D. = 1.55). The participants were 45% female (n = 75) and 55% male (n = 92). 2.2. Design and materials This study used four different response conditions: apology, compensation, sympathy, or information only. Participants read a series of three print news stories and completed the research instrument. Participants were randomly selected to receive only one of the four possible research conditions. Only one news story, the report of a chemical explosion, was of interest to this investigation and contained the crisis response manipulation. The other two stories were presented before and after the explosion story and served as distractors. The crisis stimulus described an actual chemical explosion at Marcus Oil. The Marcus Oil crisis was selected because (1) respondents were very unlikely to know about this company or to have any strong feelings about this small wax manufacturer and (2) the crisis would generate moderate attributions of crisis responsibility (feeling the organization is responsible for the cause of the crisis). Working from Situational Crisis Communication Theory (SCCT), crisis managers have a wider array of potential viable crisis response strategies when attributions of crisis responsibility are moderate to low (Coombs, 2007). The three news stories were reported to the respondents as coming from the Reuters News Service. Participants read the three brief news stories. In the crisis story, the crisis response was manipulated by including a quote from a company spokesperson. Appendix A contains the news stories and crisis response strategies used in the study. It should be noted that we used the basic definition of apology as accepting responsibility and asking for forgiveness. 2.3. Procedures The research was conducted in a classroom setting. Participants received a packet containing a consent form, cover page with directions, prior crisis reputation items, the set of three print news stories, and a three-page questionnaire. They also were verbally instructed to carefully read the news stories and then respond to the questions that followed. The administration took about 20–25 min. 2.4. Measures Prior reputation was assessed with a one item, global evaluation. The cover page asked participants to rate four different organizations, including Marcus Oil, on the item “Overall, my impression of ‘x’ is. . .” Responses were recorded on sevenpoint scales ranging from “very unfavorable” to “very favorable.” While crude, the global measure provided a general idea of how participants viewed each organizational reputation. Participants completed these global assessments along with the demographic information before reading the news stories. The post-crisis organizational reputation was measured using the 5-item version of Coombs and Holladay’s (2002) Organizational Reputation Scale and the same one item, global evaluation of reputation. Crisis responsibility was measured with two items from the personal control dimension of McAuley, Duncan, & Russell’s (1992) attribution scale and three items adapted from Griffin, Babin, & Darden’s (1992) responsibility measure. Account analysis was assessed with six items from Blumstein et al. (1974). Anger was assessed with a three-item scale from Jorgensen (1996). Negative word-of-mouth intention was measured using three items: “I would encourage friends or relative NOT to buy products from Marcus Oil,” “I would say negative things about Marcus Oil and its products to other people,” and “I would recommend Marcus Oil products to someone who asked my advice.” All items were assessed on seven-point scales ranging from 1 = “strongly disagree” to 7 = “strongly agree.” These measures were completed for the organization depicted in the news report. The survey included three manipulation check items pertaining to the crisis response strategies used by the company spokesperson: “Marcus Oil expressed concern for those affected by the crisis” (sympathy), “Marcus Oil accepted responsibility for the incident” (apology), and “Marcus Oil gave money to and provided housing for those affected by the incident” (compensation). 3. Results 3.1. Reliabilities The reliability coefficients (Cronbach’s alpha) were .84, .88, .90, and .76 for the Organization Reputation Scale, the anger scale, the account analysis scale, and the negative word-of-mouth scale, respectively. All represented acceptable reliability scores.
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
255
3.2. Manipulation checks The study involved a manipulation of crisis response strategy. Four crisis responses were used: sympathy, compensation, apology, and information only. A series of one-way ANOVAs were run to check the response manipulation. For the item “Marcus Oil accepted responsibility for the incident,” the apology condition (M = 5.70) was rated highest (F(3, 160) = 13.39, p < .001). Information only (M = 3.63) was significantly lower than the other three conditions while the sympathy (M = 4.70) and compensation (M = 4.95) conditions had similar scores on accepting responsibility. This showed the manipulation for apology was successful because respondents rated it highest on accepting responsibility, the strategy’s defining trait. For the item “Marcus Oil gave money and provided housing,” the compensation condition (M = 6.10) was rated significantly higher than the information only (M = 2.73), sympathy (M = 3.59) or apology (M = 3.65) conditions (F(3, 160) = 151.37, p < .001). This showed the response manipulation for compensation was successful because respondents rated it highest on giving money, the strategy’s defining trait. There was no significant difference between the sympathy and apology conditions. For the item “Marcus Oil expressed concern,” the information only condition (M = 3.63) was rated significantly lower than the apology condition (M = 4.87), sympathy condition (M = 5.15) or the compensation condition (M = 5.30) (F(3, 160) = 10.49, p < .001). There was no difference between the apology, sympathy or compensation conditions. This shows the manipulation for information only was successful because respondents rated it lowest on expression of concern, the trait that separates this strategy from the other three. Ideally, sympathy should have rated higher than compensation for expression of concern but respondents in other studies also have viewed compensation as an expression of concern (e.g., Coombs, 2006). The four conditions were compared to determine if they differed on their initial assessments of Marcus Oil’s reputation. A one-way ANOVA was run using the initial one-item reputation measure as the dependent variable and response as the independent variable. There was no significant effect for response condition (F(3, 162) = 1.43, p = .24). The four conditions started with similar prior reputation scores. 3.3. Research questions The three research questions were answered using one-way ANOVAs. For RQ1, there was a significant effect for response on reputation (F(3, 160) = 7.32, p < .001, 2 = .12, power = .98). However, the post hoc analysis revealed that the cause of the difference was the information only condition. The information only condition (M = 3.96) had a lower post-crisis reputation than sympathy (M = 4.82), compensation (M = 4.86), or apology (M = 4.82). There was no significant difference between the sympathy, compensation, or apology conditions. For RQ2, there was no effect for response in anger (F(3, 160) = 1.15, p = .35) or negative word-of-mouth intention (F(3, 160) = .34, p = .80). All four crisis response conditions produced the same amounts of anger and negative word-of-mouth intentions. For RQ 3, there was a significant effect for response on reputation (F(3, 160) = 8.01, p < .001, 2 = .13, power = .99). However, the post hoc analysis revealed that the cause of the difference was the information only condition. The information only condition (M = 3.03) had a lower post-crisis reputation than sympathy (M = 3.92), compensation (M = 4.09), or apology (M = 4.07). There was no significant difference between the sympathy, compensation, or apology conditions. 3.4. Limitations It is important to realize that the nature of crisis is a limitation to this study. The crisis for this study was an industrial accident where no cause had been determined. Typically this type of crisis generates low to moderate attributions of crisis responsibility and anger. The mean crisis responsibility score for the study was M = 4.03 and M = 3.30 for anger. There were no significant differences between the conditions for attributions of crisis responsibility. The crisis responsibility score is near the mid-point of four on the scale and the anger score is below the mid-point. It is possible that different results could be found if the attributions of crisis responsibility and anger are high. Such crises are strongly linked to organizations playing a significant role in the crisis occurring. Given the implied responsibility in those crisis situations, apology may be a more viable option. But that is a point for future research to determine. 4. Discussion In general, respondents had similar reactions to sympathy, compensation, and apology response strategies. There was a main effect for response strategy for reputation and account acceptance. The follow-up analysis revealed that in both cases, the information only condition was responsible for the differences. Apology, sympathy, and compensation were all rated the same for scores on post-crisis reputation account acceptance, anger, and negative word-of-mouth intention. Contrary to the advice that favors apology, sympathy and compensation can be just as effective in producing a favorable reaction from stakeholders who are not victims of the crisis. Anecdotal evidence suggests many victims want an apology but that question has yet to be fully explored and verified. Still, the vast majority of the targets for public crisis response strategies are stakeholders who are not victims of the crisis. All four crisis response strategies had similar effects on anger and negative word-of-mouth intentions. This is not surprising given the low level of anger created by the crisis. Previous research identified anger as a driver in negative word-of-mouth intentions (Coombs & Holladay, 2007) so the low anger affected both variables. Further research is necessary to determine
256
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
if utilizing crises that produce higher levels of anger will result in differences between information only, apology, sympathy, and compensation. The results have implications for crisis managers facing crises that generate low to moderate levels of crisis responsibility and anger. For non-victims, expressions of sympathy or compensation are just as effective as apology when it comes to post-crisis reputations, anger, account acceptance, and negative word-of-mouth intentions. Further research is needed to confirm if the same benefits are accrued in a more serious crisis. Given the higher costs associated with apologies, crisis managers can confidently offer compensation and/or express sympathy in the lower to moderate responsibility crises rather than relying on apology as the default. Ethically and pragmatically, if management knows it is at fault, an apology is advised. It is unethical to evade responsibility when it is known. However, not accepting responsibility (expression of sympathy and/or compensation) is an important and viable option to an apology when responsibility is unknown or ambiguous. Appendix A Text of the Different Crisis Response Strategies A spokesperson for Marcus Oil and Chemical Company had this to say: Insert spokesperson comments (four different versions—see below): Spokesperson comments for the Marcus Oil and Chemical Company story: Version 1—sympathy: We at Marcus Oil and Chemical Company are deeply saddened by this incident. The safety of our employees, local emergency personnel, and those living by our facility is of the utmost importance to us. Our thoughts and prayers go out to those affected by this incident. Version 2—compensation: We at Marcus Oil and Chemical Company will pay for housing and meals for those who had to evacuate their homes. To compensate for the inconvenience, we’ll also give $200 to each household that had to evacuate their homes or shelter-in-place. Version 3—apology: We at Marcus Oil and Chemical Company accept responsibility for last night’s explosion. We hope those who were affected by the incident can forgive us. Version 4—information only: A fire causing significant damage began around 5:50 p.m. near the main warehouse and was followed a series of explosions. One explosion damaged tank number 7. Two of our warehouse employees suffered minor injuries. The facility will be closed until the repairs are completed. References Alsop, R. J. (2004). The 18 immutable laws of corporate reputation: Creating, protecting, and repairing your most valuable asset. New York: Free Press. Arpan, L. M., & Roskos-Ewoldsen, D. R. (2005). Stealing thunder: Analysis of the effects of proactive disclosure of crisis information. Public Relations Review, 31, 425–433. Benoit, W. L. (1995). Accounts, excuses, and apologies: A theory of image restoration. Albany, NY: State University of New York Press. Benoit, W. L., & Drew, S. (1997). Appropriateness and effectiveness of image repair strategies. Communication Reports, 10, 153–163. Blumstein, P. W., Carssow, K. G., Hall, J., Hawkins, B., Hoffman, R., Ishem, E., et al. (1974). The honoring of account. American Sociological Review, 40, 551–566. Bradford, J. L., & Garrett, D. E. (1995). The effectiveness of corporate communicative responses to accusations of unethical behavior. Journal of Business Ethics, 14, 875–892. Cohen, J. R. (1999). Advising clients to apologize. Southern California Law Review, 72, 1009–1073. Coombs, W. T. (1998). An analytic framework for crisis situations: Better responses from a better understanding of the situation. Journal of Public Relations Research, 10, 177–191. Coombs, W. T. (2006). The protective powers of crisis response strategies: Managing reputational assets during a crisis. Journal of Promotion Management, 12. Coombs, W. T. (2007). Protecting organization reputations during a crisis: The development and application of situational crisis communication theory. Corporate Reputation Review, 10(3), 163–177. Coombs, W. T., & Holladay, S. J. (1996). Communication and attributions in a crisis: An experimental study of crisis communication. Journal of Public Relations Research, 8, 279–295. Coombs, W. T., & Holladay, S. J. (2002). Helping crisis managers protect reputational assets: Initial tests of the situational crisis communication theory. Management Communication Quarterly, 16, 165–186. Coombs, W. T., & Holladay, S. J. (2005). Exploratory study of stakeholder emotions: Affect and crisis. In N. M. Ashkanasy, W. J. Zerbe, & C. E. J. Hartel (Eds.), Research on emotion in organizations: The effect of affect in organizational settings (pp. 271–288). New York: Elsevier. Coombs, W. T., & Holladay, S. J. (2007). The negative communication dynamic: Exploring the impact of stakeholder affect on behavioral intention. Journal of Communication Management, 11(4), 300–312. Davies, G., Chun, R., da Silva, R. V., & Roper, S. (2003). Corporate reputation and competitiveness. New York: Routledge. Dawar, N., & Pillutla, M. M. (2000). Impact of product-harm crises on brand equity: The moderating role of consumer expectations. Journal of Marketing Research, 27, 215–226. Dean, D. W. (2004). Consumer reaction to negative publicity: Effects of corporate reputation, response, and responsibility for a crisis event. Journal of Business Communication, 41, 192–211. Dowling, G. (2002). Creating corporate reputations: Identity, image, and performance. New York: Oxford University Press. Fediuk, T. A. (2002). Concern and corrective action during an accident: An analysis of responses communicated in a crisis situation. In Paper present at the Annual Meeting of the National Communication Association. Fombrun, C. J., & van Riel, C. B. M. (2003). Fame & fortune: How successful companies build winning reputations. New York: Prentice Hall Financial Times. Fuchs-Burnett, T. (2002). Mass public corporate apology. Dispute Resolution Journal, 57(3), 26–32. Griffin, M., Babin, B. J., & Darden, W. R. (1992). Consumer assessments of responsibility for product-related injuries: The impact of regulations, warnings, and promotional policies. Advances in Consumer Research, 19, 870–877. Herr, P. M., Kardes, F. R., & Kim, J. (1991). Effect of word-of-mouth and product attribute information on persuasion: An accessibility-diagnostic perspective. The Journal of Consumer Research, 17, 452–462.
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
257
Huang, Y. H., Lin, Y. H., & Su, S. H. (2005). Crisis communicative strategies in Taiwan: Category, continuum, and cultural implication. Public Relations Review, 31, 229–238. Jorgensen, B. K. (1996). Components of consumer reaction to company-related mishaps: A structural equation model approach. Advances in Consumer Research, 23, 346–351. Laczniak, R. N., DeCarlo, T. E., & Ramaswami, S. H. (2001). Consumers’ responses to negative word-of-mouth communication: An attribution theory perspective. Journal of Consumer Psychology, 11, 57–73. Lyon, L., & Cameron, G. T. (1998). Fess up or stonewall? An experimental test of prior reputation and response style in the face of negative news coverage. Web Journal of Mass Communication Research, 1(4). Retrieved June 5, 2005 from http://www.scripps.ohiou.edu/wjmcr/vol01/1-4a.htm McAuley, E., Duncan, T. E., & Russell, D. W. (1992). Measuring causal attributions: The revised causal dimension scale (CDII). Personality and Social Psychology Bulletin, 18, 566–573. Patel, A., & Reinsch, L. (2003). Companies can apologize: Corporate apologies and legal liability. Business Communication Quarterly, 66, 17–26. Rousseau, D. M. (2005). Is there such a thing as “evidence-based management”? Academy of Management Review, 31, 256–269. Stockmyer, J. (1996). Brands in crisis: Consumer help for deserving victims. Advances in Consumer Research, 23, 429–435. Sturges, D. L. (1994). Communicating through crisis: A strategy for organizational survival. Management Communication Quarterly, 7, 297–316. Tyler, L. (1997). Liability means never being able to say you’re sorry: Corporate guilt, legal constraints, and defensiveness in corporate communication. Management Communication Quarterly, 11, 51–73. Wirtz, J., & Mattila, A. (2003). The effects of consumer expertise on evoked set and service loyalty. Journal of Services Marketing, 17(7), 649–665.
Contents lists available at ScienceDirect
Public Relations Review
Comparing apology to equivalent crisis response strategies: Clarifying apology’s role and value in crisis communication W. Timothy Coombs ∗ , Sherry J. Holladay 1 Communication Studies, Eastern Illinois University, 600 Lincoln Avenue, Charleston, IL 61920, United States
a r t i c l e
i n f o
Article history: Received 24 February 2008 Accepted 11 April 2008 Keywords: Apology Crisis communication Crisis response strategies
a b s t r a c t Crisis communication has begun to systematically examine how people perceive the crisis response strategies using experimental designs. This is an important step in the evolution of crisis communication research. One problem with this research has been the over promotion of apology as “the” response. Too often the value of an apology is established by comparing how people react to an apology and to other less victim-centered/accommodative responses. This “unfair” comparison results in apology appearing to be “the” best alternative. This study compares apology to more equivalent crisis response strategies to more fairly determine if apology is “the” best strategy. The results show that people react similarly to any victim-centered/accommodative strategy meaning apology is not “the” best strategy. The implications of the results for crisis managers are discussed. © 2008 Elsevier Inc. All rights reserved.
After a crisis, organizations need to convey messages to stakeholders. These messages begin by telling stakeholders what to do to protect themselves from the crisis (instructing information) and to help them cope psychologically with the crisis (adjusting information). The next step is to address the reputational threat posed by the crisis (Sturges, 1994). Reputations are valuable resources that are threatened during a crisis (Davies, Chun, da Silva, & Roper, 2003; Fombrun & van Riel, 2003). Most post-crisis communication research has utilized case study methods. While providing useful descriptive data, case studies offer little insight into how stakeholders actually respond to crisis response strategies. Moreover, the case studies often provide minimal theoretical insight into crisis communication (Dawar & Pillutla, 2000; Dean, 2004). Crisis communication research has begun moving beyond case studies to experiments designed to assess systematically how people perceive crisis response strategies. Experimental research has begun to explore how crisis response strategies affect stakeholder perceptions and responses to the organizations in crisis and to build crisis communication theory (e.g., Arpan & Roskos-Ewoldsen, 2005; Coombs & Holladay, 1996; Dean, 2004; Huang, Lin, & Su, 2005). This experimental research is essential if we are to build evidence-based knowledge for crisis management. Evidence-based management is derived from evidence-based medicine and argues for the use of scientifically proven results to guide actions in medicine and now management (Rousseau, 2005). We need to move beyond the speculation of cases based on media reports if we are to create evidence-based crisis communication (Coombs, 2007). The empirical evidence for the effects of crisis response strategies is still in its infancy and has room for development. One area of potential growth is understanding the effects of different crisis response strategies utilized in the same crisis. Such comparisons can help crisis managers make choices when they face particular crises. However, problems arise when researchers offer comparisons of crisis communication strategies that “unfairly” pit extremely diverse strategies against one another. For example, apology is sometimes compared to responses such as denial or excuse that do little, if anything,
∗ Corresponding author. Tel.: +1 217 581 3324; fax: +1 217 581 5718. E-mail addresses: [email protected] (W.T. Coombs), [email protected] (S.J. Holladay). 1 Tel.: +1 217 581 6311; fax: +1 217 581 5718. 0363-8111/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.pubrev.2008.04.001
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
253
to address the concerns of victims. Such biased comparisons allow apology, a victim-centered/accommodative response, to easily “win” and provides a questionable foundation for its recommended use to crisis managers. Apology should be compared to more equivalent crisis response strategies to establish a more realistic assessment of its value in crisis communication. This study compares apology to other accommodative crisis response strategies in an effort to produce a richer and more rounded view of apology in crisis communication. 1. Literature review and research questions Research has shown that crisis response strategies – what an organization says and does after a crisis – serve to protect a reputation after a crisis (e.g., Coombs & Holladay, 1996). Researchers have over-emphasized the use of apology/mortification as the “best” crisis response (e.g., Benoit, 1995; Benoit & Drew, 1997) and have used widely varying definitions of apology (Patel & Reinsch, 2003). At its core, an apology is marked by the organization accepting responsibility for the crisis and asking for forgiveness (Benoit & Drew, 1997; Fuchs-Burnett, 2002). A variety of additional components can be added to this definition including expression of remorse/sympathy, expression of regret, preventative measures, and reparation (Benoit & Drew, 1997; Cohen, 1999; Fuchs-Burnett, 2002; Patel & Reinsch, 2003). Accepting responsibility is the centerpiece of an apology and makes it the most expensive response financially for an organization (Fuchs-Burnett, 2002; Patel & Reinsch, 2003; Tyler, 1997). When an organization offers an apology it opens itself to lawsuits and financial loss. The apology is used as evidence in court to win lawsuits against the organization. However, some evidence indicates that compensation and sympathy, two less expensive strategies, are as effective as an apology in shaping people’s perceptions of the organization taking responsibility for the crisis because these strategies focus on victims’ needs. The sympathy response expresses concern for victims while compensation offers victims something to offset the suffering. These are referred to as highly accommodative strategies (Coombs, 2006; Fediuk, 2002). If other highly accommodative response strategies are as effective as apology in protecting reputation, then this has important implications for the options crisis managers might pursue in protecting their reputations after a crisis. Previous experimental research that has established the value of apologies in crisis management has stacked the deck in favor of apologies. The effects of the apology strategy were compared to less accommodative strategies such as giving no comment, denial, excuse, or justification (Bradford & Garrett, 1995; Dean, 2004; Lyon & Cameron, 1998). It is not surprising that two of the studies found apology to be the preferred strategy, meaning the apology strategy had the strongest positive effect on perceptions of an organization’s reputation (Bradford & Garrett, 1995; Dean, 2004). We know little about how apology compares to other accommodative strategies such as compensation and expression of sympathy. One study that did compare apology and compensation used in a service failure situation (bad service at a restaurant) not a crisis. The service failure situation found apology to be more effective at increasing satisfaction (Wirtz & Mattila, 2003). While an expression of sympathy is more of an adjusting information strategy than a reputation repair strategy, it reflects a high level of accommodation (Sturges, 1994). To insure that the effects documented in this study were not created by simply offering any organizational response, the information only response was included. It is common for crisis managers to report only what has happened, a type of what Sturges (1994) calls instructing information. Previous research has found little effect for information only responses on people’s post-crisis perceptions (Coombs, 1998). The first research question in our study examines the effect of apology, compensation, sympathy, and information only on post-crisis reputations. RQ1: Do compensation, apology, and sympathy crisis responses produce similar post-crisis reputation evaluations and do they differ from information only response? Other important outcomes stem from a crisis including anger and negative word-of-mouth. Along with attributions, a crisis will produce affective responses (Coombs & Holladay, 2005). Stakeholder anger can damage the organization–stakeholder relationship or even lead stakeholders to end the relationship (Jorgensen, 1996; Stockmyer, 1996). Anger can also motivate stakeholders to say bad things about the organization to people they know—negative word-of-mouth (Coombs & Holladay, 2007). Organizations value word-of-mouth because of the benefits associated with positive word-of-mouth and the harm inflicted by negative word-of-mouth (Herr, Kardes, & Kim, 1991; Laczniak, DeCarlo, & Ramaswami, 2001). Crisis response strategies can affect the anger and negative word-of-mouth intentions created by a crisis. RQ2: Do compensation, apology, and sympathy crisis responses produce similar reports of anger and negative word-ofmouth and do they differ from information only response? Account acceptance refers to how respondents feel about the crisis response offered by the organization. Greater account acceptance indicates respondents believe the crisis response is appropriate. Any difference in how crisis response strategies affect reputation could be a result of differing levels of account acceptance. RQ3: Do compensation, apology, and sympathy crisis responses produce similar levels of account acceptance and do they differ from information only response?
254
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
2. Methods 2.1. Participants Participants in the study were 167 undergraduate students from a Midwestern university. The ages ranged from 18 to 56 (M = 20, S.D. = 1.55). The participants were 45% female (n = 75) and 55% male (n = 92). 2.2. Design and materials This study used four different response conditions: apology, compensation, sympathy, or information only. Participants read a series of three print news stories and completed the research instrument. Participants were randomly selected to receive only one of the four possible research conditions. Only one news story, the report of a chemical explosion, was of interest to this investigation and contained the crisis response manipulation. The other two stories were presented before and after the explosion story and served as distractors. The crisis stimulus described an actual chemical explosion at Marcus Oil. The Marcus Oil crisis was selected because (1) respondents were very unlikely to know about this company or to have any strong feelings about this small wax manufacturer and (2) the crisis would generate moderate attributions of crisis responsibility (feeling the organization is responsible for the cause of the crisis). Working from Situational Crisis Communication Theory (SCCT), crisis managers have a wider array of potential viable crisis response strategies when attributions of crisis responsibility are moderate to low (Coombs, 2007). The three news stories were reported to the respondents as coming from the Reuters News Service. Participants read the three brief news stories. In the crisis story, the crisis response was manipulated by including a quote from a company spokesperson. Appendix A contains the news stories and crisis response strategies used in the study. It should be noted that we used the basic definition of apology as accepting responsibility and asking for forgiveness. 2.3. Procedures The research was conducted in a classroom setting. Participants received a packet containing a consent form, cover page with directions, prior crisis reputation items, the set of three print news stories, and a three-page questionnaire. They also were verbally instructed to carefully read the news stories and then respond to the questions that followed. The administration took about 20–25 min. 2.4. Measures Prior reputation was assessed with a one item, global evaluation. The cover page asked participants to rate four different organizations, including Marcus Oil, on the item “Overall, my impression of ‘x’ is. . .” Responses were recorded on sevenpoint scales ranging from “very unfavorable” to “very favorable.” While crude, the global measure provided a general idea of how participants viewed each organizational reputation. Participants completed these global assessments along with the demographic information before reading the news stories. The post-crisis organizational reputation was measured using the 5-item version of Coombs and Holladay’s (2002) Organizational Reputation Scale and the same one item, global evaluation of reputation. Crisis responsibility was measured with two items from the personal control dimension of McAuley, Duncan, & Russell’s (1992) attribution scale and three items adapted from Griffin, Babin, & Darden’s (1992) responsibility measure. Account analysis was assessed with six items from Blumstein et al. (1974). Anger was assessed with a three-item scale from Jorgensen (1996). Negative word-of-mouth intention was measured using three items: “I would encourage friends or relative NOT to buy products from Marcus Oil,” “I would say negative things about Marcus Oil and its products to other people,” and “I would recommend Marcus Oil products to someone who asked my advice.” All items were assessed on seven-point scales ranging from 1 = “strongly disagree” to 7 = “strongly agree.” These measures were completed for the organization depicted in the news report. The survey included three manipulation check items pertaining to the crisis response strategies used by the company spokesperson: “Marcus Oil expressed concern for those affected by the crisis” (sympathy), “Marcus Oil accepted responsibility for the incident” (apology), and “Marcus Oil gave money to and provided housing for those affected by the incident” (compensation). 3. Results 3.1. Reliabilities The reliability coefficients (Cronbach’s alpha) were .84, .88, .90, and .76 for the Organization Reputation Scale, the anger scale, the account analysis scale, and the negative word-of-mouth scale, respectively. All represented acceptable reliability scores.
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
255
3.2. Manipulation checks The study involved a manipulation of crisis response strategy. Four crisis responses were used: sympathy, compensation, apology, and information only. A series of one-way ANOVAs were run to check the response manipulation. For the item “Marcus Oil accepted responsibility for the incident,” the apology condition (M = 5.70) was rated highest (F(3, 160) = 13.39, p < .001). Information only (M = 3.63) was significantly lower than the other three conditions while the sympathy (M = 4.70) and compensation (M = 4.95) conditions had similar scores on accepting responsibility. This showed the manipulation for apology was successful because respondents rated it highest on accepting responsibility, the strategy’s defining trait. For the item “Marcus Oil gave money and provided housing,” the compensation condition (M = 6.10) was rated significantly higher than the information only (M = 2.73), sympathy (M = 3.59) or apology (M = 3.65) conditions (F(3, 160) = 151.37, p < .001). This showed the response manipulation for compensation was successful because respondents rated it highest on giving money, the strategy’s defining trait. There was no significant difference between the sympathy and apology conditions. For the item “Marcus Oil expressed concern,” the information only condition (M = 3.63) was rated significantly lower than the apology condition (M = 4.87), sympathy condition (M = 5.15) or the compensation condition (M = 5.30) (F(3, 160) = 10.49, p < .001). There was no difference between the apology, sympathy or compensation conditions. This shows the manipulation for information only was successful because respondents rated it lowest on expression of concern, the trait that separates this strategy from the other three. Ideally, sympathy should have rated higher than compensation for expression of concern but respondents in other studies also have viewed compensation as an expression of concern (e.g., Coombs, 2006). The four conditions were compared to determine if they differed on their initial assessments of Marcus Oil’s reputation. A one-way ANOVA was run using the initial one-item reputation measure as the dependent variable and response as the independent variable. There was no significant effect for response condition (F(3, 162) = 1.43, p = .24). The four conditions started with similar prior reputation scores. 3.3. Research questions The three research questions were answered using one-way ANOVAs. For RQ1, there was a significant effect for response on reputation (F(3, 160) = 7.32, p < .001, 2 = .12, power = .98). However, the post hoc analysis revealed that the cause of the difference was the information only condition. The information only condition (M = 3.96) had a lower post-crisis reputation than sympathy (M = 4.82), compensation (M = 4.86), or apology (M = 4.82). There was no significant difference between the sympathy, compensation, or apology conditions. For RQ2, there was no effect for response in anger (F(3, 160) = 1.15, p = .35) or negative word-of-mouth intention (F(3, 160) = .34, p = .80). All four crisis response conditions produced the same amounts of anger and negative word-of-mouth intentions. For RQ 3, there was a significant effect for response on reputation (F(3, 160) = 8.01, p < .001, 2 = .13, power = .99). However, the post hoc analysis revealed that the cause of the difference was the information only condition. The information only condition (M = 3.03) had a lower post-crisis reputation than sympathy (M = 3.92), compensation (M = 4.09), or apology (M = 4.07). There was no significant difference between the sympathy, compensation, or apology conditions. 3.4. Limitations It is important to realize that the nature of crisis is a limitation to this study. The crisis for this study was an industrial accident where no cause had been determined. Typically this type of crisis generates low to moderate attributions of crisis responsibility and anger. The mean crisis responsibility score for the study was M = 4.03 and M = 3.30 for anger. There were no significant differences between the conditions for attributions of crisis responsibility. The crisis responsibility score is near the mid-point of four on the scale and the anger score is below the mid-point. It is possible that different results could be found if the attributions of crisis responsibility and anger are high. Such crises are strongly linked to organizations playing a significant role in the crisis occurring. Given the implied responsibility in those crisis situations, apology may be a more viable option. But that is a point for future research to determine. 4. Discussion In general, respondents had similar reactions to sympathy, compensation, and apology response strategies. There was a main effect for response strategy for reputation and account acceptance. The follow-up analysis revealed that in both cases, the information only condition was responsible for the differences. Apology, sympathy, and compensation were all rated the same for scores on post-crisis reputation account acceptance, anger, and negative word-of-mouth intention. Contrary to the advice that favors apology, sympathy and compensation can be just as effective in producing a favorable reaction from stakeholders who are not victims of the crisis. Anecdotal evidence suggests many victims want an apology but that question has yet to be fully explored and verified. Still, the vast majority of the targets for public crisis response strategies are stakeholders who are not victims of the crisis. All four crisis response strategies had similar effects on anger and negative word-of-mouth intentions. This is not surprising given the low level of anger created by the crisis. Previous research identified anger as a driver in negative word-of-mouth intentions (Coombs & Holladay, 2007) so the low anger affected both variables. Further research is necessary to determine
256
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
if utilizing crises that produce higher levels of anger will result in differences between information only, apology, sympathy, and compensation. The results have implications for crisis managers facing crises that generate low to moderate levels of crisis responsibility and anger. For non-victims, expressions of sympathy or compensation are just as effective as apology when it comes to post-crisis reputations, anger, account acceptance, and negative word-of-mouth intentions. Further research is needed to confirm if the same benefits are accrued in a more serious crisis. Given the higher costs associated with apologies, crisis managers can confidently offer compensation and/or express sympathy in the lower to moderate responsibility crises rather than relying on apology as the default. Ethically and pragmatically, if management knows it is at fault, an apology is advised. It is unethical to evade responsibility when it is known. However, not accepting responsibility (expression of sympathy and/or compensation) is an important and viable option to an apology when responsibility is unknown or ambiguous. Appendix A Text of the Different Crisis Response Strategies A spokesperson for Marcus Oil and Chemical Company had this to say: Insert spokesperson comments (four different versions—see below): Spokesperson comments for the Marcus Oil and Chemical Company story: Version 1—sympathy: We at Marcus Oil and Chemical Company are deeply saddened by this incident. The safety of our employees, local emergency personnel, and those living by our facility is of the utmost importance to us. Our thoughts and prayers go out to those affected by this incident. Version 2—compensation: We at Marcus Oil and Chemical Company will pay for housing and meals for those who had to evacuate their homes. To compensate for the inconvenience, we’ll also give $200 to each household that had to evacuate their homes or shelter-in-place. Version 3—apology: We at Marcus Oil and Chemical Company accept responsibility for last night’s explosion. We hope those who were affected by the incident can forgive us. Version 4—information only: A fire causing significant damage began around 5:50 p.m. near the main warehouse and was followed a series of explosions. One explosion damaged tank number 7. Two of our warehouse employees suffered minor injuries. The facility will be closed until the repairs are completed. References Alsop, R. J. (2004). The 18 immutable laws of corporate reputation: Creating, protecting, and repairing your most valuable asset. New York: Free Press. Arpan, L. M., & Roskos-Ewoldsen, D. R. (2005). Stealing thunder: Analysis of the effects of proactive disclosure of crisis information. Public Relations Review, 31, 425–433. Benoit, W. L. (1995). Accounts, excuses, and apologies: A theory of image restoration. Albany, NY: State University of New York Press. Benoit, W. L., & Drew, S. (1997). Appropriateness and effectiveness of image repair strategies. Communication Reports, 10, 153–163. Blumstein, P. W., Carssow, K. G., Hall, J., Hawkins, B., Hoffman, R., Ishem, E., et al. (1974). The honoring of account. American Sociological Review, 40, 551–566. Bradford, J. L., & Garrett, D. E. (1995). The effectiveness of corporate communicative responses to accusations of unethical behavior. Journal of Business Ethics, 14, 875–892. Cohen, J. R. (1999). Advising clients to apologize. Southern California Law Review, 72, 1009–1073. Coombs, W. T. (1998). An analytic framework for crisis situations: Better responses from a better understanding of the situation. Journal of Public Relations Research, 10, 177–191. Coombs, W. T. (2006). The protective powers of crisis response strategies: Managing reputational assets during a crisis. Journal of Promotion Management, 12. Coombs, W. T. (2007). Protecting organization reputations during a crisis: The development and application of situational crisis communication theory. Corporate Reputation Review, 10(3), 163–177. Coombs, W. T., & Holladay, S. J. (1996). Communication and attributions in a crisis: An experimental study of crisis communication. Journal of Public Relations Research, 8, 279–295. Coombs, W. T., & Holladay, S. J. (2002). Helping crisis managers protect reputational assets: Initial tests of the situational crisis communication theory. Management Communication Quarterly, 16, 165–186. Coombs, W. T., & Holladay, S. J. (2005). Exploratory study of stakeholder emotions: Affect and crisis. In N. M. Ashkanasy, W. J. Zerbe, & C. E. J. Hartel (Eds.), Research on emotion in organizations: The effect of affect in organizational settings (pp. 271–288). New York: Elsevier. Coombs, W. T., & Holladay, S. J. (2007). The negative communication dynamic: Exploring the impact of stakeholder affect on behavioral intention. Journal of Communication Management, 11(4), 300–312. Davies, G., Chun, R., da Silva, R. V., & Roper, S. (2003). Corporate reputation and competitiveness. New York: Routledge. Dawar, N., & Pillutla, M. M. (2000). Impact of product-harm crises on brand equity: The moderating role of consumer expectations. Journal of Marketing Research, 27, 215–226. Dean, D. W. (2004). Consumer reaction to negative publicity: Effects of corporate reputation, response, and responsibility for a crisis event. Journal of Business Communication, 41, 192–211. Dowling, G. (2002). Creating corporate reputations: Identity, image, and performance. New York: Oxford University Press. Fediuk, T. A. (2002). Concern and corrective action during an accident: An analysis of responses communicated in a crisis situation. In Paper present at the Annual Meeting of the National Communication Association. Fombrun, C. J., & van Riel, C. B. M. (2003). Fame & fortune: How successful companies build winning reputations. New York: Prentice Hall Financial Times. Fuchs-Burnett, T. (2002). Mass public corporate apology. Dispute Resolution Journal, 57(3), 26–32. Griffin, M., Babin, B. J., & Darden, W. R. (1992). Consumer assessments of responsibility for product-related injuries: The impact of regulations, warnings, and promotional policies. Advances in Consumer Research, 19, 870–877. Herr, P. M., Kardes, F. R., & Kim, J. (1991). Effect of word-of-mouth and product attribute information on persuasion: An accessibility-diagnostic perspective. The Journal of Consumer Research, 17, 452–462.
W.T. Coombs, S.J. Holladay / Public Relations Review 34 (2008) 252–257
257
Huang, Y. H., Lin, Y. H., & Su, S. H. (2005). Crisis communicative strategies in Taiwan: Category, continuum, and cultural implication. Public Relations Review, 31, 229–238. Jorgensen, B. K. (1996). Components of consumer reaction to company-related mishaps: A structural equation model approach. Advances in Consumer Research, 23, 346–351. Laczniak, R. N., DeCarlo, T. E., & Ramaswami, S. H. (2001). Consumers’ responses to negative word-of-mouth communication: An attribution theory perspective. Journal of Consumer Psychology, 11, 57–73. Lyon, L., & Cameron, G. T. (1998). Fess up or stonewall? An experimental test of prior reputation and response style in the face of negative news coverage. Web Journal of Mass Communication Research, 1(4). Retrieved June 5, 2005 from http://www.scripps.ohiou.edu/wjmcr/vol01/1-4a.htm McAuley, E., Duncan, T. E., & Russell, D. W. (1992). Measuring causal attributions: The revised causal dimension scale (CDII). Personality and Social Psychology Bulletin, 18, 566–573. Patel, A., & Reinsch, L. (2003). Companies can apologize: Corporate apologies and legal liability. Business Communication Quarterly, 66, 17–26. Rousseau, D. M. (2005). Is there such a thing as “evidence-based management”? Academy of Management Review, 31, 256–269. Stockmyer, J. (1996). Brands in crisis: Consumer help for deserving victims. Advances in Consumer Research, 23, 429–435. Sturges, D. L. (1994). Communicating through crisis: A strategy for organizational survival. Management Communication Quarterly, 7, 297–316. Tyler, L. (1997). Liability means never being able to say you’re sorry: Corporate guilt, legal constraints, and defensiveness in corporate communication. Management Communication Quarterly, 11, 51–73. Wirtz, J., & Mattila, A. (2003). The effects of consumer expertise on evoked set and service loyalty. Journal of Services Marketing, 17(7), 649–665.