Abstract
Apologies are a powerful way to restore trust and reduce punishment costs in bilateral settings. But what do we know about public apologies for large scale man-made disasters? Herein we report on results from an experiment with apologies in a multilateral setting: a firm-caused environmental disaster. Subjects read about an oil spill scenario, and learned whether the oil firm made a full apology, a partial apology, or no apology, and whether the firm had a good, bad, or no environmental reputation. A partial apology is one that fails to accept full material responsibility for damages, such as by shifting the blame to another party. We find that full apologies and better reputation reduce the demand for punishment. However, full apologies and reputation are substitutes, with reputation being significantly more important. Additionally, apologies do not reduce the demand for compensation and may increase it if the firm is clearly a bad actor, or if admission of guilt is the only information subjects have. Our results help explain corporate social responsibility investments and greenwashing, and why many public apologies over an environmental disaster are only partial apologies.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Mungan (2012) even designs a mechanism to explicitly price apologies in the legal system in order to give the signal a clear cost.
Effective reputation and crisis management is costly and valuable. Corporate crisis management specifically has been associated with long term stock price impacts (Knight and Pretty 1996). A more recent Deloitte survey of global executives found that reputation risk is rated as the most important strategic business risk, with growing importance as firms surveyed plan to invest more in reputation and crisis management, and allocate direct responsibility for reputation risk at the chief executive and board level (Deloitte 2014).
We recruited subjects from: Illinois, Indiana, Iowa, Kansas, Michigan, Minnesota, Missouri, Nebraska, North Dakota, Ohio, South Dakota, and Wisconsin. We purposefully chose to limit our subject pool to avoid surveying residents of the West Coast who would be more likely to know about any local oil spill. We acknowledge that this geographic targeting limits the external validity of our experiment. Future work that interacts our treatments with a wider array of subject characteristics (including, for example, a person’s proximity to the impacted area) could be useful.
OLS results are robust to using different numbers of the top-coded response. Additionally, preferences for environmental fines may not be monotonic in the population, so we also estimated the effects on fine size using an unordered discrete choice conditional logit model. In the conditional logit model, the left hand side variable is equal to one for subject i’s preferred fine and zero for every other fine option, and the right hand side variables include a continuous variable for fine size in addition to interactions between the continuous fine variable and the treatment dummies. Results from this model confirm those reported here and are available in the corresponding online appendix.
We use Wilner’s (2007) Stata package “wtpcikr” for simulating Krinsky and Robb confidence intervals, which is based on Krinsky and Robb (1986) as well as Loomis and Ekstrand (1998) and Poe et al. (2005). When confidence intervals are a nonlinear function of estimated standard errors, this method provides better estimates than the delta method. The method uses the estimated covariance matrix from the maximum likelihood model to simulate 5000 predicted mean and median WTA for different random draws of the error, and then chops off the smallest and largest 2.5% of the draws.
In debriefing questions, subjects were strongly split on whether they thought the apology affected their preferred fine and WTA, based on whether the firm’s reputation was good or bad. The subset of subjects who believe the apology was influential for them tended to believe it reduced their WTA and preferred fine. This is, however, not entirely consistent with how this subset actually responded to the treatments. Appendix A has a full discussion of these results.
In unreported results, we also estimated these effects by ordered probit and OLS with similar results which are available in the corresponding online appendix.
The sole exception being the likelihood of signing a petition. For this outcome, shifting the blame reduces the likelihood of a personal response relative to the Bad Reputation, No Apology treatment.
In a set of debriefing questions we asked subjects whether they believed the firm will change safety practices in the future. We estimate an additional logit model using this information. To the extent that subjects’ beliefs about the firm’s future practices reflect their trust in the firm, these results give a rough picture of whether the apologies and reputation information help restore or maintain trust. Table 3 shows the results which generally complement the findings for the personal responses. Bad reputational information decreases the likelihood that a subject reports believing the firm will change practices in the future, regardless of apology type. Apologies have no statistically significant impact, and therefore do not appear to help restore trust.
Subjects were asked in a follow up survey, “Do you think of yourself as an environmentalist?” Subjects that responded “Yes, very strongly” or “Yes, somewhat” were coded as being environmentalists. We also separated the sample according to whether subjects had visited a national park. These results yielded largely insignificant and confounding results, perhaps due to the fact that so many people have visited parks, regardless of their preference for marine ecosystems. These results are available in the corresponding online appendix.
In criminal trial settings, for example, many jurisdictions require statements of remorse from violent criminals upon conviction and before sentencing.
References
Abeler, J., Calaki, J., Andree, K., Basek, C. (2010). The power of apology. Economics Letters, 107(2), 233–235.
Barrage, L., Chyn, E., Hastings, J. (2014). Advertising, reputation, and environmental stewardship: Evidence from the BP oil spill. Ann Arbor, 1001, 48109–1220.
Behar, R. (1990). Interview with Lawrence Rawl: Exxon Strikes Back. Time Magazine. http://www.time.com/time/magazine/article/0,9171,969673,00.html. Accessed 1 March 2017.
Benoit, W.L. (1997). Image repair discourse and crisis communication. Public Relations Review, 23(2), 177–186.
Boothman, R.C., Blackwell, A.C., Campbell, D.A. Jr, Commiskey, E., Anderson, S. (2009). A better approach to medical malpractice claims? The University of Michigan experience. Journal Health of Life Sciences Law, 2(2), 125–159.
Bottom, W. P., Gibson, K., Daniels, S.E., Murnighan, J.K. (2002). When talk is not cheap: Substantive penance and expressions of intent in rebuilding cooperation. Organization Science, 13(5), 497– 513.
Cherry, M.A., & Sneirson, J.F. (2011). Beyond profit: Rethinking corporate social responsibility and greenwashing after the BP oil disaster. Tulane Law Review, 85(4), 983–1038.
Cohen, J.R. (1999). Apology and organizations: Exploring an example from medical practice. Fordham. Urban Law Journal, 27, 1447.
De Cremer, D. (2010). To pay or to apologize? On the psychology of dealing with unfair offers in a dictator game. Journal of Economic Psychology, 31(6), 843–848.
De Cremer, D., van Dijk, E., Pillutla, M.M. (2010). Explaining unfair offers in ultimatum games and their effects on trust: An experimental approach. Business Ethics Quarterly, 20(01), 107–126.
De Cremer, D., Pillutla, M.M., Folmer, C.R. (2011). How important is an apology to you? Forecasting errors in evaluating the value of apologies. Psychological Science, 22(1), 45–48.
Delmas, M.A., & Burbano, V.C. (2011). The drivers of greenwashing. California Management Review, 54(1), 64–87.
Deloitte. (2014). Global survey on reputation risk. http://www2.deloitte.com/content/dam/Deloitte/pl/Documents/Reports/pl_Reputation_Risk_survey_EN.pdf. Accessed 1 March 2017.
Desmet, P.T., De Cremer, D., van Dijk, E. (2011). In money we trust? The use of financial compensations to repair trust in the aftermath of distributive harm. Organizational Behavior and Human Decision Processes, 114(2), 75–86.
Fehr, E., & Gächter, S. (2000). Cooperation and punishment in public goods experiments. American Economic Review, 90(4), 980–994.
Fehr, E., & Gächter, S. (2002). Altruistic punishment in humans. Nature, 415(6868), 137.
Fischbacher, U., & Utikal, V. (2013). On the acceptance of apologies. Games and Economic Behavior, 82, 592–608.
Gold, G.J., & Weiner, B. (2000). Remorse, confession, group identity, and expectancies about repeating a transgression. Basic and Applied Social Psychology, 22(4), 291–300.
Haab, T.C., & McConnell, K.E. (2002). Valuing environmental natural resources: The econometrics of non-market valuation. Northampton: Edward Elgar Publishing.
Hearit, K.M. (2006). Crisis management by apology: Corporate response to allegations of wrongdoing. Routledge.
Ho, B. (2012). Apologies as signals: With evidence from a trust game. Management Science, 58(1), 141–158.
Ho, B., & Liu, E. (2011). Does sorry work? The impact of apology laws on medical malpractice. Journal of Risk and Uncertainty, 43(2), 141–167.
Jeanty. P.W. (2007). wtpcikr: Constructing Krinsky and Robb confidence interval for mean and median willingness to pay (wtp) using stata. North American Stata Users’ Group Meetings, August 2007.
Kim, P.H., Ferrin, D.L., Cooper, C.D., Dirks, K.T. (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.
Knight, R.F., & Pretty, D.J. (1996). The impact of catastrophes on shareholder value. Templeton College.
Kotchen, M., & Moon, J.J. (2012). Corporate social responsibility for irresponsibility. The BE Journal of Economic Analysis and Policy, 12(1), 1–23.
Korobkin, R., & Guthrie, C. (1994). Psychological barriers to litigation settlement: An experimental approach. Michigan Law Review, 93(1), 107–192.
Krinsky, I., & Robb, A.L. (1986). On approximating the statistical properties of elasticities. The Review of Economic and Statistics, 68(4), 715–719.
Loomis, J., & Ekstrand, E. (1998). Alternative approach for incorporating uncertainty when estimating willingness to pay: The case of the Mexican spotted owl. Ecological Economics, 27, 29– 41.
Martinez-Vaquero, L.A., Han, T.A., Pereira, L.M., Lenaerts, T. (2015). Apology and forgiveness evolve to resolve failures in cooperative agreements. Scientific Reports, 5(10639), 1–12.
McMichael, B.J., Van Horn, R.L., Viscusi, W.K. (2016). Sorry is never enough: The effect of state apology laws on medical malpractice liability risk. Available at SSRN: https://ssrn.com/abstract=2883693 or https://doi.org/10.2139/ssrn.2883693.
Minor, D., & Morgan, J. (2011). CSR as reputation insurance: Primum non nocere. California Management Review, 53(3), 40–59.
Mungan, M.C. (2012). Don’t say you’re sorry unless you mean it: Pricing apologies to achieve credibility. International Review of Law and Economics, 32(1), 178–187.
O’Hara O’Connor, E.A. (2004). Apology and thick trust: What spouse abusers and negligent doctors might have in common. Chicago-Kent Law Review, 79, 1055–1089.
O’Hara O’Connor, E.A. (2011). Organizational apologies: BP as a case study. Vanderbilt Law Review, 64(6), 1959–1991.
O’Hara O’Connor, E.A., & Yarn, D. (2002). On apology and consilience. Washington Law Review, 77, 1121–1192.
O’Hara O’Connor, E.A., & Robbins, M.M. (2009). Using criminal punishment to serve both victim and social needs. Law and Contemporary Problems, 72(2), 199–217.
Ohtsubo, Y., & Watanabe, E. (2009). Do sincere apologies need to be costly? Test of a costly signaling model of apology. Evolution and Human Behavior, 30(2), 114–123.
Pace, K.M., Fediuk, T.A., Botero, I.C. (2010). The acceptance of responsibility and expressions of regret in organizational apologies after a transgression. Corporate Communications: An International Journal, 15(4), 410–427.
Park, T., Loomis, J.B., Creel, M. (1991). Confidence interval for evaluating benefit estimates from dichotomous choice contingent valuation studies. Land Economics, 67, 64–73.
Pauly, J.J., & Hutchison, L.L. (2005). Moral fables of public relations practice: The Tylenol and Exxon Valdez cases. Journal of Mass Media Ethics, 20(4), 231–249.
Poe, G.L., Giraud, K.L., Loomis, J.B. (2005). Computational methods for measuring the difference of empirical distributions. American Journal of Agricultural Economics, 87, 353–365.
Robbennolt, J.K. (2003). Apologies and legal settlement: An empirical examination. Michigan Law Review, 102(3), 460–516.
Schniter, E., Sheremeta, R.M., Sznycer, D. (2013). Building and rebuilding trust with promises and apologies. Journal of Economic Behavior and Organization, 94, 242–256.
Silliman, B.R., van de Koppel, J., McCoy, M.W., Diller, J., Kasozi, G.N., Earl, K., Adams, P.N., Zimmerman, A.R. (2012). Degradation and resilience in Louisiana salt marshes after the BP-Deepwater Horizon oil spill. Proceedings of the National Academy of Sciences, 109(28), 11234–11239.
Small, W.J. (1991). Exxon Valdez: How to spend billions and still get a black eye. Public Relations Review, 17(1), 9–25.
Spence, M. (1973). Job market signaling. The Quarterly Journal Of Economics, 87(3), 355–374.
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(1), 51–73.
Williams, D.E., & Treadaway, G. (1992). Exxon and the Valdez accident: A failure in crisis communication. Communication Studies, 43(1), 56–64.
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
Below is the link to the electronic supplementary material.
Appendix A: Perceived versus actual effect of apologies and reputation information
Appendix A: Perceived versus actual effect of apologies and reputation information
We asked subjects who received an apology treatment whether they thought the apology affected their valuation decisions, and if so in what way — a smaller fine or WTA, a larger one, or neither. We also asked a similar question for the information treatments. In this appendix we compare these responses to the actual treatment effects, separately estimating the effects for the subgroup who felt the apology or reputation information influenced them.
Table 9 compares these results for the apology treatments, where treatment dummies are defined relative to the “Sorry” treatment (Full Apology only, No Reputation). The first column shows that good or bad reputation information makes subjects more or less likely to believe the apology affected their decisions, but that blame shifting does not have this effect. The second and third columns show that, among subjects who believe the apology affected their decisions, those receiving good reputation information also believe they are less likely to demand a larger fine. Consistent with the “size of fine” results in Table 7 discussed above, this effect is larger when the “Shift the Blame” message is paired with a good reputation. However, the results are much stronger here, in column 3 of Table 9, for the perceived effect on a subject’s preferred fine among subjects who believe the apologies matter to them. Compare this perceived effect to column 6 of Table 9 which separates the treatment effects on preferred fine size by whether or not the subject believes the apology mattered to them. In this column we reestimate (2) with an interaction between treatment dummies and a dummy that is equal to one if the subject stated the apology affected their decisions. In this column, the baseline group consists of subjects in the “Sorry” (Full Apology) treatment with no reputation information, who do not believe the apology mattered to them. Relative to that group, subjects in the same treatment group who believe the apology affected them demanded a $3.48 million smaller fine on average. However any additional reputation information works in the opposite direction for this group to partially undo the extent of their forgiveness. So while the apology-affected group did not demand larger fines on average or in any specific treatment, consistent with column 3, they seem to extend additional judgments when more information is provided. Notably, the “Good, Blame” treatment which we noted earlier when discussing Table 7 was most effective at reducing the demanded fine is most effective among the group who believes apologies did not matter to them; this group demanded a $2.26 million smaller fine in this treatment.
Similarly, the fourth and fifth columns of Table 9 show that subjects who believe the apology influenced them also say they are more likely to demand less compensation if the apology is combined with a good reputation, and less likely to demand more compensation if a firm with a bad reputation apologizes. Yet we saw in Table 8 that full apologies combined with reputation information had almost no effect on compensation demanded. We again compare this perceived effect in columns 4 and 5 of Table 9 to column 7. As with the fine size, in this column we reestimate (3) with an interaction between treatment dummies and a dummy for whether or not the subject believed the apology mattered to them. Subjects in the “Sorry” treatment with no reputational information were significantly more likely to accept a given compensation bid if they believe apologies matter - which suggests there is considerable heterogeneity in the response given that the average effect of the “Sorry” treatment (Table 8) was to increase WTA. However, as with the fine size, this effect is at least partially offset by any additional reputation information or alternative apology message, as the interactions of the treatment dummies with the “Apology Matter” dummy show.
Table 10 shows a similar cognitive dissonance between the perceived and actual effect of reputation information. The first column shows that treatments have no effect on whether subjects think the information influenced their decisions. The second and third columns show that subjects who believe they responded to reputation information also say that they are less likely to advocate for a lower fine if the firm has a bad reputation unless the firm offers some form of apology. This is consistent with the average treatment effects on the preferred fine described in Table 7 and the ranking of interaction coefficients in the second to last column of Table 10. However, the fourth and fifth columns of Table 10 show that subjects think they behave in the same way regarding compensation, which is clearly not consistent with the last column of interaction coefficients on the likelihood of accepting a given bid, or results from Table 8.
Rights and permissions
About this article
Cite this article
Gilbert, B., James, A. & Shogren, J.F. Corporate apology for environmental damage. J Risk Uncertain 56, 51–81 (2018). https://doi.org/10.1007/s11166-018-9276-4
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11166-018-9276-4