BUSINESS APOLOGIES AND THE ETHICS AND NARRATIVE OF TRUST
[☛ watch the video]
The “invisible hand” that leads markets to work efficiently, and societies to derive the benefits from these markets, is invisible not only because if operates through a myriad of decentralized actors, decisions and behaviors guided by the price mechanism. More deeply, the success of a capitalistic, market-based economy depends on something even less visible or easy to define: the trust itself that people have in the system to operate correctly. Individuals and organizations interact under the assumption that trades will generate gains that the parties will somehow share; laws and regulations, political and low-enforcing institutions provide formal means to abide by these principles. But no law, regulation or contract is fully complete; therefore, parties often act on implicit agreements and assumptions founded on trust, credibility and reputation.
Paper and electronic money, for example, play their key role in the economy thanks to the trust that all agents assign value to it. More generally, members of a society accept an economic system insofar as all parties respect some criterion of fairness and no one systematically suffers exploitation. In fact, in recent years observers pointed out that practices such as the abuse of market power and “cronyism”, or the increase in income and wealth inequality often through acquired privilege rather than meritocracy, may hamper the overall faith that individuals have in the ability of a capitalistic, market-based and democratic system to produce welfare that everyone will enjoy. For-profit corporations, perhaps the most representative agents in a capitalistic society, are often blamed for exploiting consumers, workers as well as less powerful partners; in fact, public trust in corporations continue to fall according to numerous measures and accounts. Such a distrust, or even disenchantment, is also involving the young, high-tech, once revered companies in the Silicon Valley, for example: from representing forces of innovation, progress and connection, several accounts now depict them as almost hostile to society.
Apologies are defined as regretful acknowledgments of an offense of failure, with the undertaking of responsibility and the hope to restore trust between the parties. Can apologies issued by corporations help restoring this trust? In this essay, after providing some anecdotal evidence of business apologies, I will discuss three potential roles that they can have: as “costly signals” of the characteristics and culture of a company; as ways to (re)establish empathy or limit emotional negative responses by customers and communities more in general; and, finally, as parts of “narratives” meant to sustain societal trust into welfare-enhancing the role of corporations. I will then conclude with some proposals for a data-driven research agenda.
Companies Do Apologize
Anecdotal evidence suggests that companies do recur to public apologies, typically issued by CEOs, following episodes of alleged wrongdoing as well as activities that, although legal, are perceived as morally unacceptable. In 1982, Johnson & Johnson’s Chairman James Burke took personal responsibility for the deaths caused by cyanide-laced Tylenol capsules (even though J&J was not responsible for the tampering), and directly oversaw the $100 million product recall. The automobile company Volkswagen publicly apologized in the person of its CEO, in 2015, for their illicit use of software that would defeat emission controls. Reed Hastings, the co-founder and CEO of Netflix, in 2011 issued an apology after having increased prices by about 60% to customers who subscribed to both the streaming and DVD mailing services (Netflix stocks dropped drastically after the price increase announcements, and hundreds of thousands of customers unsubscribed from the service). Apologies for pricing strategies considered exploitative or morally outrageous came also from other companies, notably Uber in more than one occasion for applying “surge” pricing during such events as the lockdown in Sydney due to the siege of a coffee shop by a gunman in 2014, a terrorist attack in London (UK) in 2017, and protests in New York against President Trump’s immigration ban that included a strike by taxi drivers. Note that there was nothing illicit in the reliance on flexible pricing by Uber; in fact, one could argue that those strategies made economic sense because, in periods of shortages, high prices stimulate supply of rides on the one hand, and provide the service to the customers who value them the most. However, in many circumstances people perceive surge pricing as unethical. In fact, Uber has issued quite a large number of apologies over the past few years for actions for many reasons beyond the use of surge pricing; instances go from a rape of a woman by an Indian driver to cases of workplace harassment, to the behavior of the CEO Travis Kalanick that eventually led to him leaving the company. Other prominent cases include the apologies of United Airlines CEO for the treatment of a passenger who was violently removed from a plane in overbooking in 2017; and, in 2010, those by British Petroleum following the massive oil spill in the Gulf of Mexico.
Although apologies typically refer to a specific episode, there have been instances of managers and entrepreneurs expressing regret and taking responsibility for the overall direction that a company took, and for its consequences. In a post on September 30 2017, to celebrate the end of Yom Kippur, Facebook founder and CEO Mark Zuckerberg asked for forgiveness because, rather than being loyal to its recently stated mission to “bring the world closer together”, Facebook has been indicated as a main vehicle used to instill division and even hatred over the past few years. The events around the Presidential campaigns and 2016 election in the United States are of course a main example. Similarly, and in reaction to the same set of events, Twitter co-founder Evan Williams regretted that the world was not automatically becoming a better place because “everybody could speak freely and exchange information and ideas”. He apologized for any responsibility that social media and Twitter in particular might have had in affecting the outcome of the 2016 election, and more generally for the diffused harassments taking place on the platform.
In other cases, companies rather make a point of not apologizing; when Goldman Sachs was blamed for shorting securities that the firm itself sold to its client – essentially thus betting against them – CEO Lloyd Blankfein claimed that this is how things go in financial markets, and investors should be aware of it. Neither Mylan CEO Heather Bresch nor Turing Pharmaceutical CEO Martin Shkreli admitted any ethical trespasses when they sharply increased the prices of the allergy shot EpiPen and of Daraprim (a drug used to treat toxoplasmosis), respectively.
But what are the incentives for companies to issue an apology? In other words, what are the costs and benefits, or what’s an apology worth?
Apologies can be perceived as just “cheap talk”; they are minimally costly and, per se, do not represent any formal commitment or change. In fact, a more cynical view would interpret apologies by companies as a quick, deceiving “way out” that managers use. Also the demand for apologies may lack credibility; what would consumers, for example, do if a company does not apologize? Activities such as lawsuits or boycotts suffer from collective action problems that make them difficult to pursue and, as such, not a credible threat.
However the frequent release of apologies and the media attention that they often receive suggest that there may be more to them – in particular, higher costs for companies, and greater meaning for societies that demand them.
Apologies as costly signals
Apologies may actually be more costly than it seems. For one, they could represent an admission of responsibility that, at least in some jurisdictions, is even admissible in court. Moreover, the decision itself to apologize, and the content of the apology, may reveal information about other unobserved characteristics and actions of a company and its leaders. People may ask why a company is issuing an apology, and whether for example this is a way to selectively reveal information while hiding other relevant details, to focus attention only on some alleged wrongdoings in order to downplay or obfuscate others, and so on. People may also question the timing of an apology, and the time lag between an action, its discovery, and any statement by the firm. Apologies may also constitute, at least informally, a precedent for a given company and also for others, as they may change social norms and expectations, thus imposing potential costs also to other companies. Paradoxically, even rival businesses that might benefit from the disgrace of a company may prefer that company to stay “silent” in order to not increase scrutiny on competitors too.
Because of these potential consequences and complications, companies may need to incur additional costs in the form of legal counsel as well as professional PR assistance.
It is, however, the cost of an apology that would make such an apology more meaningful and consequential. In order to provide credible information about unobserved features of an agent, a signal must be costly: the agent is willing to give up some benefits in order to (re)obtain others. Firms, therefore, may release apologies precisely because they are costly and risky. Potential lawsuits, the risk of additional scrutiny, and the other ambiguities surrounding such statements may be worth the positive perception by the public of the intention to make amend for any mistakes or wrongdoing.
The psychology and economics of business apologies
Although the costs of an apology for illicit actions such as fraud, corruption or collusion can be identified and potentially quantified, it is more difficult to define the costs of (and, as a consequence, the benefits from) apologies for activities that are not formally illicit. What is the actual “signalling” content of apologies, for example, for relying on surge pricing, shorting stocks held by clients, having been rude to employees, claiming that fame, money and prestige is helping to date good looking women, or even apologies for not having apologized sooner or better? Arguments based on credibility and reputation fall short of explaining the role of apologies unless one clarifies the potential negative consequences for a company to issue one.
Psychological explanations might provide insights. A frequent tendency of individuals, for example, is to evaluate outcomes not on the basis of absolute terms, but relative to a “reference point”. By publicly admitting responsibility and regret, a company might change the point or yardstick with which potential customers and society more in general evaluate the firm; the admission of a violation of a shared societal value will lead individuals to now expect that such higher moral standards will become the normal behavior of the company. Moreover, aggrievances for the perception of an unfair treatment or behavior may lead people to reciprocate or retaliate even if there are costs of doing so, as numerous studies have shown. This tendency might counteract the lack of credibility that collective-action processes generate in these cases; even if not formally organized and coordinated, individuals might find it worthwhile to incur costs from punishing a company. In particular, less costly actions, and actions that can be implemented quickly and almost on a “whim”, can represent forms of successful, diffuse boycott. Following the typology that Daniel Kahneman elaborated in his 2011 book “Thinking, Fast and Slow”, these acts appeal more to “System 1” than “System 2” reasoning; people will mostly rely on psychological responses such as those mentioned above rather than to formal, more “thought-through” cost-benefits considerations.
One example of low cost, “emotional” response is unsubscribing from an app-based car service like Uber even if this implies spending more for transportation; this may contribute to explaining why, as mentioned above, Uber has been apologizing so often. Numerous analyses of what makes a “good” corporate apology (the role of certain terms or, say, the use of the passive vs. active voice, the role of one’s facial expression, and so on) stress the psychological reactions to perceived wrongdoings and how certain cues may make an apology more likely to be taken as genuine.
Apologies as narratives
Psychological and emotional reactions may help explaining why companies release apologies in some cases where such a signal would be overall uninformative; it is implausible, however, that these explanations may exhaust the reasons for the moral outrage following certain conducts and the frequency with which companies seems to be apologizing. I am going to argue that there can be more in a corporate apology – something that, although immaterial and admittedly ill-defined, reconnects to the overall role of apologies to help re-establish trust and sharing of societal values.
In his Presidential address at the 2017 Annual Meetings of the American Economic Association in Chicago, Robert Shiller claimed that economists have paid too little attention to narratives and how they can determine public perceptions and interpretation of events and, in turn, affect major decisions and their social outcomes by spreading and “contagion”, often independently of whether those perceptions and interpretations are supported by facts.
Narratives may also serve the purpose of reflecting or re-establishing shared values. Corporate apologies, therefore, may be parts of “stories” that managers and entrepreneurs tell the public to restore trust in the overall role of profit-seeking organization in a society, and more generally in the operating of a capitalistic system. Apologies thus may function as frequent reminders (both to companies and communities) about the underlying principles that make a capitalist, market economy accepted and nurtured; these values include fairness, equal opportunities, and respect of the law, among others. By contributing to a narrative whereby companies make amend for damaging certain consumers (even if within the law), firms contribute to spread a certain view of how an economy should work.
Firms may well opt for espousing a different narrative; in the case of Goldman Sachs, Mylan and Turing Pharmaceuticals mentioned above, for example, one may argue that the CEOs of those companies were essentially providing a different worldview whereby certain actions that many consider unfair are integral part of how for-profit corporations operate. In other words, there is not only one narrative about the working of a capitalistic society operates and what the role of companies is.
Note that I am not claiming that apologies are merely “stories” (which, by the way, is the original etymology of the word apology!), without any content or actual consequences. As Shiller points out, narratives gain their power when they spread broadly, thus providing a shared view that affects behaviors. To the extent that apologies affect how people interpret processes and communicate these views, and if these interpretations are consequential, then these statements can have major consequences for a single organization and for society as a whole.
Classifying and Understanding Apologies in Business: A Proposal for a Research Agenda and Methodology
The notes in the previous section identify several potential motives for companies to apologize, and mechanisms through which these apologies may be effective or not, either to amend a specific relationship or to contribute to restore a more general trust in business organizations and a capitalistic economy. As a consequence of this multitude of motives and mechanisms, it is hard to make empirical predictions and prescriptions about the role and impact of apologies in business.
Some limited evidence on the role of apologies provides insights. Laboratory experiments where individuals played versions of trust games or engaged in other forms of interactions found that sending a (costly) “I am sorry” message helped restoring trust; and that players exhibited “lying costs” that made them less likely to send false apologies. An experiment on EBay found that a message of apology from a seller was more likely than small monetary offer to persuade a buyer to remove a negative review. Although informative, this evidence may not easily generalize to actual situations where the trust toward a company, or more generally the corporate world, is endangered.
Systematic empirical studies of apologies in business that address some of the issues and questions that I discussed above are essentially non-existent. Thus, as much as I think that there is some substance in those arguments, I am not in a position to substantiate my claims with reliable empirical evidence. I will, however, propose some initial ideas for empirical analyses. In particular, I would propose as a first step to perform a comprehensive classification of business apologies through history. By doing so, one would go beyond anecdotal and sparse accounts and would provide a more reliable empirical basis for understanding the role of apologies in business and their ability to restore trust.
Although this might seem a daunting exercise, modern technologies provide tools to make the task feasible with a high degree of detail and precision. The digitization of information on the one hand, and text analysis techniques based on machine-learning methodologies, would allow automating most of the search. The availability of digitized corpora of relevant documents (news releases and articles from such databases as Lexis Nexis / Factiva, disclosure documents stored by securities commission, etc.) would provide the raw material for the analysis. Approaches ranging from basic “naïve Bayes” to more structured sentiment analyses or word vectors would be useful to isolate and interpret words and phrases that refer to the issue of a formal apology, as well as to retrieve details about the contingencies that led to issuing that apology.
Of course this will not just be a programming job: the definition of the proper algorithm will depend on what we are looking for, and the specific questions that will define the search and classification. The research agenda that I am sketching here is interdisciplinary in nature, spanning from economics to sociology, from cultural studies to computer science, from psychology to political science. I do hope to see, in particular, young scholars undertake this line of research as a collaborative endeavour between research fields.
Below I discuss what I see as some key questions that this search task should address.
What are companies apologizing for?
Classifying the reasons for apologies and determining their relative frequency is a key first step toward understanding the role and impact of apologies. In particular, the analysis of the frequency of different types may inform as to what kinds of corporate responsibilities individuals and societies pay particular attention, or to which “negative” behaviors companies decide to increase attention. Broad categories may include poor financial performance, outright illegal activities, or legal business practices that are perceived as breaching societal or moral values. More broadly, this analysis could give a preliminary view of what people perceive to be the roles of corporations in society. Comparisons of types and relative frequencies over time and across legal and geographical areas would provide further insights as to cultural differences as well as the evolution of the perception of the role of corporations over time, and the relevance of certain principles overall (e.g. attention to the environment, attitudes toward inequality, corruption, fairness, and so on).
How do companies express their apologies?
More detailed analyses of the wording and specific expressions that companies use in their apology statements would provide information about the specific message or “signal” that a company is conveying, and as a consequence, what the company is trying to reveal about itself. Text analysis may identify apologies with a clear attribution of responsibility as opposed to more generic references to external factors, and algorithms can be trained to discern sincere and insincere apologies based on wording and phrasing structure; these tools would be useful for an historical analysis as well as a predictive device to establish the credibility of apologies in “real time”.
The timing of apologies
There are different aspects of the timing of business apologies that a broad, automated text analysis could uncover. A first question concerns the temporal distance between the occurrence of the cause of the apology and the issuance of the apology itself, and the correlates of potentially heterogeneous time lags.
It would also be interesting to assess the frequency of apologies in different periods (while defining a proper baseline), and, also, whether there is evidence of “clustering” of apologies; if, for example, corporate apologies tend to occur in concentrated periods, one could study whether this is due to a response to changes in public perception of the unwritten pact with corporations, or because corporations tend to undertake certain behaviors in concentrated periods. Of interest would be also to explore whether the frequency of and reasons for apologies vary with the overall business cycle.
Sectoral differences in apologies
How often do companies in different industries apologize, and for what? The exploration of sectoral differences may inform about the underlying frequency of wrongdoings in different industries; or, controlling for available information on wrongdoings (again, through text analysis of news sources), we may infer the propensity to apologize in different sectors, which in turn may indicate that individuals and society have different expectations about the social responsibilities of companies producing different goods and services. Further investigations based on potential sectoral differences may highlight whether the level of competition in a sector or its overall technological sophistication relate to the frequency and types of issued apologies.
This is only a first, partial list of potential large-sample analyses; I see them as initial “empirical building blocks” to better understand, more broadly the implicit “deal” between corporations and society about the broader role and functions expected from businesses.
In an epoch of decline in public trust toward both the corporate world and political institutions, a deeper understanding of formal and informal mechanisms that enhance common values and understanding should be in the agenda of social scientists in different disciplines. In this essay I advanced the idea that, in addition to representing ways to recover public approval, either as costly signals or responses to psychological reactions by the public, corporate apologies may indeed be part of a broader narrative meant to maintain a common understanding and sharing of the underlying values of an economic system and social order. In this sense, what could be seen as cheap talk or empty words may contribute to spreading a particular “story” of a society and, as such, turn into a commitment for companies to abide by certain social values. Of course we need systematic theorizing and empirical work to establish whether this particular mechanism contributes to initiating virtuous circles of trust; I hope that the notes above will provide insights for such scholarly endeavours.
* University of Toronto, Institute of Management and Innovation (UofT-Mississauga), Centre for Ethics at UofT, Behavioural Economics @ Rotman, and National Bureau of Economic Research. Email: email@example.com.