Recent social science research suggests that one’s capacity to act on her stated moral values is often significantly compromised by psychological limitations. Psychologists often refer to these limits as “bounded ethicality”—an umbrella term used to describe a diverse variety of psychological mechanisms that putatively lead people astray (Bazerman and Tenbrunsel Reference Bazerman and Tenbrunsel2011). Bounded ethicality, one might think, has a significant implication for moral decision-making, which we call the Radical Behavioral Challenge (RBC):
Bounded ethicality makes it psychologically impossible for people to make choices that reflect the moral principles they endorse and the moral principles articulated in normative business ethical theory, creating a gap between what normative theory prescribes and what people can in fact do.Footnote 1
Although RBC has not been expressly endorsed by psychologists or others, some psychologists make claims that can only be understood in its terms. For example, Bazerman and Tenbrunsel (Reference Bazerman and Tenbrunsel2011, 137) assert that particular bounded ethicality phenomena make it “psychologically impossible” that auditors will satisfy the obligation to do an objective audit, an obligation that presumably normative ethical theory, as well as the professional codes that auditors embrace, assign to auditors. Psychologists thus suggest that bounded ethicality raises a fundamental doubt about the value of normative ethical theory for business.
In the remainder of this paper we proceed as follows. In Section 2, we further explain bounded ethicality and elaborate on RBC. We argue that the soundness of RBC depends on a traditional moral principle, “ought implies can (OIC),” which provides, roughly, that a person ought to do something only if she can. In Section 3, we observe that OIC is subject to different interpretations, not all of which are defensible. We argue that the most sensible version of OIC enables us to hold businesspersons under the authority of moral obligations, even when they are influenced by bounded ethicality; normative theorizing should continue to set high ethical standards. In Section 4, we review objections to our view. In Section 5, we discuss the implications of behavioral findings for normative ethics, as well as the implications of our argument for behavioral theorizing.
Normative and behavioral business ethics have co-existed for decades, often with a degree of tension (Donaldson Reference Donaldson1994; Robertson Reference Robertson1993; Treviño and Weaver Reference Treviño and Weaver1994; Weaver and Treviño Reference Weaver and Treviño1994; Werhane Reference Werhane1994). Given the recent prominence of research in behavioral ethics and, in particular, the findings surrounding bounded ethicality, it is more important than ever to understand the relation of the two approaches, and to examine how each can contribute to the other. The contribution of behavioral ethics can be recognized without detracting from the importance of normative ethics for pedagogy and practice, we will conclude.
2. Bounded Ethicality and the Radical Behavioral Challenge
2.1. Bounded Ethicality
The social science field of behavioral ethics aims to explain the behavior of people who face ethical decisions (Bazerman and Gino Reference Bazerman and Gino2012; Tenbrunsel and Smith–Crowe Reference Tenbrunsel and Smith‐Crowe2008; Treviño, Weaver, and Reynolds Reference Treviño, Weaver and Reynolds2006). Psychologists have confirmed the existence of many pernicious influences on ethical decision—often operating at an unconscious level. When these influences lead to unethical behavior that conflict with an actor’s moral beliefs and commitments, the phenomenon may be referred to as bounded ethicality (Bazerman Reference Bazerman2011; Bazerman and Tenbrunsel Reference Bazerman and Tenbrunsel2011; Chugh, Bazerman, and Banaji Reference Chugh, Bazerman, Banaji, Moore, Cain, Loewenstein and Bazerman2005; Tenbrunsel Reference Tenbrunsel, Moore, Cain, Loewenstein and Bazerman2005).Footnote 2
One example of bounded ethicality is “ordinary prejudice,” which reveals itself in implicit associations about gender, race, and other demographic groups (Bertrand, Chugh, and Mullainathan Reference Bertrand, Chugh and Mullainathan2005; Green et al. Reference Green, Carney, Pallin, Ngo, Raymond, Iezzoni and Banaji2007; Greenwald et al. Reference Greenwald, Poehlman, Uhlmann and Banaji2009; Rudman and Ashmore Reference Rudman and Ashmore2007). These associations can lead to unintentionally discriminatory results, such as discriminatory hiring practices (Bertrand, Chugh, and Mullainathan Reference Bertrand, Chugh and Mullainathan2005) and unwarranted discrepancies in the evaluation of the skills and competencies of workers (Bertrand and Mullainathan Reference Bertrand and Mullainathan2004; Reeves Reference Reeves2014), which mimic the results of deliberate discrimination. Other elements of bounded ethicality include “in-group favoritism,” “self-serving bias,” “illusion of control,” “(overly) discounting future consequences,” and “motivated blindness.”Footnote 3
The literature on bounded ethicality is vast. For simplicity, we focus on the stream within it that discusses a phenomenon particularly relevant for business, motivated blindness. This stream is exemplified in work by Moore, Tanlu, and Bazerman (Reference Moore, Tanlu and Bazerman2010). These authors conducted an experiment in which 100 participants were randomly assigned to one of four groups: buyer, seller, buyer’s auditor, or seller’s auditor. Using the same information, including information about how to estimate the worth of a firm, the auditors were asked to estimate a certain company’s value. The auditors were instructed, “Your goal is for this assessment to be as impartial as you can make it,” and were promised a monetary reward for the accuracy of their judgments. The results show that the auditors were aware that their roles had influenced their appraisals, but significantly underestimated this influence and were unable to correct for it. The same experiment was replicated with more than 100 full-time professional auditors from one of the Big Four accounting companies in the U.S. and found similar results. The experiments indicate that the mere role assignment influences auditors to be biased in the interests of their clients (Moore, Tanlu, and Bazerman Reference Moore, Tanlu and Bazerman2010).
The experiments above may be construed to practically imply that “audit failures are natural by-products of the auditor-client relationship” (Bazerman and Tenbrunsel Reference Bazerman and Tenbrunsel2011, 137). The experiments may also imply that “the current U.S. audit system makes it ‘psychologically impossible’ … for even the most honest auditors to make objective judgments” (Bazerman and Tenbrunsel Reference Bazerman and Tenbrunsel2011, 137; italics ours). Thus, “cases of audit failure are inevitable” (Bazerman, Morgan, and Loewenstein Reference Bazerman, Morgan and Loewenstein1997, 90).
2.2. The Radical Behavioral Challenge (RBC)
In this section, we review RBC, explaining how bounded ethicality—for instance, motivated blindness—arguably presents a challenge to normative business ethics. We begin with the following hypothetical, inspired by the experiments outlined above (Moore, Tanlu, and Bazerman Reference Moore, Tanlu and Bazerman2010):
Case 1: The Biased Auditors’ Defense
Suppose that you consider buying stock in a California-based smartphone company. You rely upon publicly certified auditors’ reports to determine your stock portfolio. However, you do not know that the auditors were unconsciously biased towards the client’s interests due to motivated blindness. The estimated valuation of the company as reported by the company’s auditors is about 26 percent higher than estimates by impartial experts. A national newspaper happens to discover that the audit is notably biased and discloses the fact to the public. The stock market immediately responds to the scandal and the value of the company’s stocks decreases dramatically. You argue that the auditors and the accounting firm committed a wrong by violating the auditors’ duty to be impartial, independent, and objective (Duska and Duska Reference Duska and Duska2003, chap. 5–6). A clever representative of the auditors, Trevor, denies the complaint, saying, “We did not do anything we should not have done.” Trevor explains that the auditors intended to be as independent and fair as possible and that no act can be wrong or duty-infringing if it is beyond their capacities to act in such a way. Trevor goes on to argue that it is the supervising partners of the accounting firm, the CEO of the client company, and the relevant policy makers who have the institutional resources to enable the auditors to counteract the influence of motivated blindness; therefore, only they wronged you.
Did the biased auditors do anything wrong? We think so. They breached a duty of impartiality. How could Trevor suggest otherwise?
Underlying Trevor’s defense is the philosophical doctrine that “ought” implies “can,” which we state in its most generic form as follows:
OIC: A moral principle binds an agent only if she can act on it.
The auditors committed a wrong only if they had a duty of impartiality, but Trevor, relying on OIC, thinks that the duty cannot and hence does not apply to the auditors.
In coming to terms with OIC, one must recognize that it contains at least two treacherous concepts. First, it is very difficult to say what it means for a principle to bind an agent. Intuitively, by saying that a principle binds a person, we mean that it gives her a reason to act on it and that it gives others a reason to view her critically if she fails to act on it. While this intuitive parsing will have to suffice for purposes of this paper, we note that it appeals to the idea of giving a person a reason to act, which is itself challenging, but outside our scope. The second problematic concept contained in OIC is that which a person can do. As we will argue, different interpretations of “what a person can do” have very different normative implications, and it is therefore important to settle on a sensible interpretation. Still, the basic idea of OIC seems plausible: if a moral agent ought to perform some action then it must be possible for her to do so. If it is impossible, the agent is not obligated to perform the action.
OIC is broadly accepted in the fields of moral philosophy (Altham Reference Altham1988; Donagan Reference Donagan1979; D. Levy Reference Levy and Kolb2008; Stern Reference Stern2004) and normative business ethics (Brenkert Reference Brenkert1995; Goodpaster Reference Goodpaster, Brenkert and Beauchamp2010; Jones and Felps Reference Jones and Felps2013; Kim Reference Kim2014; Phillips Reference Phillips2010). The plausibility of OIC seems clear when we reflect on the harshness of rejecting the principle. Thus, imagine that there is a baby in dire straits on Mars. If you could leap to Mars, you could save the baby. But you cannot make that leap, nor get to Mars in any other way. So it makes no sense to say that you do something wrong by failing to save the baby; because you cannot save the baby, it is untrue that you ought to save the baby. No doubt there are less exotic cases in which it makes sense to say that because you cannot do something, it is untrue that you ought to do it. For example, if you are a quadriplegic and see a baby drowning on the far side of a river, you have no obligation to swim to save the baby. The function of OIC is to enable us to respond intelligently to such cases.
As we have suggested, OIC can be interpreted in various ways. The first interpretation we consider is also the simplest:
OIC 1: A moral principle binds an agent only if she has the occurrent ability to act on it.Footnote 4
A person has an occurrent ability to perform some action, as we shall understand it, if it is true that if she tries to do that action, she will, and nothing prevents her from trying to perform that action. OIC 1 thus identifies what a person can do in terms of contingent facts about that person—what, in fact, happens when she tries to do something. We recognize that both what counts as a person trying to do something and what counts as preventing a person from trying are difficult notions, but we must leave them unanalyzed here.
OIC 1 has many applications. One can use it, for instance, to criticize utilitarianism (Griffin Reference Griffin1992, Reference Griffin1996). By pointing out that we cannot “perform the large-scale calculations of what is best that [utilitarianism] require[s]” (Griffin Reference Griffin1992, 130), one can argue that we are not under an obligation to maximize utility. OIC 1 also has more fundamental implications for moral theorizing in general. For example, James Griffin (Reference Griffin1992, 131) argues as follows:
Moral norms must be tailored to fit the human moral torso. They are nothing but what such tailoring produces. There are no moral norms outside the boundary set by our capacities. Moral norms regulate human action; a norm that ignores the limited nature of human agents is not an “ideal” norm, but no norm at all.
Thus, relying on the OIC 1, Griffin (Reference Griffin1992, Reference Griffin1996) proposes that determinations of right, wrong, duty, and obligation must be in part determined by contingent facts about human abilities. Later we argue that OIC 1 misconstrues the relevance of such facts.
We maintain that a plausible reconstruction of Trevor’s arguments defending the auditors must rely on OIC 1’s distinctive appeal to contingent facts about human abilities. Consider:
The Argument of Radical Behavioral Challenge (RBC):
Premise 1) An agent ought morally to act on a principle only if she has the occurrent ability to act on it. (OIC 1.)
Premise 2) The auditors in Case 1 do not have the occurrent ability to act on a principle requiring objectivity. (Empirical claim stemming from phenomenon of motivated blindness.)
Conclusion) Therefore, it is not true that the auditors in Case 1 ought to act on a principle requiring objectivity. Further, by not acting objectively, these auditors did nothing wrong.
Although as we have presented the argument for RBC, it is about motivational blindness and auditor independence, we will soon argue that the argument can be generalized to cover other obligations affected by other putative sources of bounded ethicality. Indeed, we doubt that any moral principle espoused in commonsense morality or by normative ethics theorists would escape a version of RBC, if RBC were a cogent argument.
Is the argument for RBC cogent? Not if either of its premises is false. We will argue, against Premise 1, that there are clear cases in which an agent lacks an occurrent ability to do x but still can and must do it. Before beginning that discussion, it is worth getting clear about why it is worth having. Thus, as a preliminary matter, we will discuss the significance of RBC. It is tempting to dismiss RBC as a straw person, but we think that it can be a real threat to normative ethics. We want to show that the straw person concern is unfounded. As we see it, there are two versions of the straw person concern: First, that RBC does not pose a problem for any normative theory, research, or claim; second, that nobody really embraces RBC. We address each in turn.
i) Which normative claims are challenged by RBC?:
RBC challenges many normative arguments, including any argument that maintains that auditors generally have an obligation to be impartial, objective, and fair. For example, Ronald Duska and Brenda Duska (Reference Duska and Duska2003, 31, 85, 95, 125, 203) defend the Code of Conduct of the American Institute of CPAs (AICPA), which stipulates that all members ought to adhere to “The principle of objectivity,” which “imposes the obligation to be impartial, intellectually honest, and free of conflicts of interest” (AICPA 2014). John Boatright (Reference Boatright2007) also justifies accountants’ moral obligation to be even-handed, which he refers to as the “gatekeepers’ obligation.” We suspect that most, if not all, business ethics teachers would accept that auditors have such a duty. RBC is inconsistent with all such positions.
But the challenge posed by RBC transcends auditor independence. For example, RBC challenges arguments in favor of fair treatment of minorities and other disadvantaged populations. Thomas Donaldson (Reference Donaldson1989) defends a list of fundamental international rights, including non-discriminatory treatment. Implicit biases against certain disadvantaged groups and in-group favoritism may thwart organizational efforts at non-discrimination.Footnote 5 Hence, RBC poses a challenge to the advocacy of rights against nondiscriminatory treatment.
More generally, RBC challenges moral guidance with respect to the values of fairness and justice in business organizations. For instance, Jeffrey Harrison and Andrew Wicks (Reference Harrison and Wicks2013) defend organizational justice as one of the four main values that any adequate stakeholders approaches should embody. Robert Audi and Patrick Murphy (Reference Audi and Murphy2006) explain that one important dimension of business integrity is the virtue of fairness. In fact, we suspect that almost all business ethics theories, except for those grounded in moral egoism, would submit that organizations should espouse and enhance the value and rules of fairness (e.g., Bowie Reference Bowie1991, Bowie Reference Bowie1998; McCall Reference McCall2001, McCall Reference McCall2003; Radin and Werhane Reference Radin and Werhane2003). However, certain forms of bounded ethicality, such as the self-serving bias or over-claiming credit, may make it difficult to see how one or one’s organization is behaving unfairly towards internal or external stakeholders, and can therefore impede fairness and justice from being realized (Brewer Reference Brewer1979; Dasgupta Reference Dasgupta2004; Epley and Caruso Reference Epley and Caruso2004; Ross Reference Ross and Berkowitz1977).
ii) Who asserts RBC?:
The second form that the straw person objection takes is as follows: RBC poses no challenge because nobody endorses it; there is no point arguing against RBC because it expresses a position that nobody takes.
Our response to the second form of the straw person argument relies on an empirical claim. We have heard the claim confirmed in the hallways of different business schools, and we feel confident that much of this journal’s readership shares our experience. The claim is that many business school colleagues and students dismiss normative business ethics as “impractical.” Anyone who has taught business ethics has likely encountered business students or practitioners who argue for some unarticulated versions of RBC. To our knowledge, however, no scholars have explicitly proposed RBC. Nonetheless, the mere fact that no scholar has actually proposed the challenge does not mean that we do not have reason to evaluate it. What is important is whether or not the challenge can be plausibly proposed, and, if so, whether it is forceful enough to challenge existing perspectives. We have shown that it is.
Second, the two premises of RBC are latent in two bodies of existing literature: philosophical ethics and behavioral ethics. The first premise of RBC is the OIC 1, which various scholars (e.g., Griffin Reference Griffin1992, Reference Griffin1996; Vranas Reference Vranas2007) have supported. The second premise of RBC is bounded ethicality, which has been well documented. Although no one has explicitly asserted RBC, its conclusion is contained in the two premises. In this article, we simply draw upon two areas of scholarship that do not often speak directly to one other in order to articulate the argument of RBC.
3. Rejecting the Radical Behavioral Challenge
In this section, we aim to show that RBC derives its plausibility from flaws in OIC 1. Kant had OIC correct, and, once one appreciates Kant’s analysis, RBC loses its appeal, as does OIC 1. The insight in Kant’s argument, we believe, is that it shows that judgments about what a person can do are typically more than empirical reports about individual propensities and occurrent abilities. People have rational resources that OIC 1 does not reflect.
3.1. Kant’s Version of OIC
We accept Robert Stern’s (Reference Stern2004) analysis of Kant’s (Reference Kant, Guyer and Wood1999, Reference Kant and Gregor1996b, Reference Kant and Gregor1996a) original articulation of OIC. On this interpretation, Kant did not mean for “can” to be interpreted according to individual agents’ contingent abilities. Instead, Kant argued that moral requirements are determined by “what obligations can be shown to apply to rational agents capable of acting rightly; and then, once the moral law is fixed, [Kant] uses ‘ought implies can’ to determine what we are capable of qua human beings, in so far as we fall under this law” (Stern Reference Stern2004, 60). On this view, the very fact that the moral law binds us indicates that, as rational beings, we are capable of following it. In a manner of speaking, according to Kant, whether we “can” is determined by the fact that we “ought.” One might then interpret Kant’s version of OIC as follows:
OIC 2: A moral principle binds an agent only if a rational agent acting out of respect for the moral law could act on it.
Notice that OIC 2 cannot be used to underwrite RBC. Presumably, for example, cognitively perfect, superhuman, rational moral agents could act on a principle of auditor objectivity, even if we cognitively flawed humans cannot.
However, OIC 2 may also pose problems for the critic of RBC. If “ought” implies “can” provides merely that we humans ought to do something only if perfect creatures could do that thing, it does not limit our obligations to do objective audits. As stated, OIC 2 seems implausibly demanding. For instance, it could conceivably underwrite an obligation to defy gravity, or perform some other physically impossible feat. Recall the Martian we discussed earlier. While our Martian was visiting us on Earth, if she noticed that a baby was in dire straits on Mars, who could be saved only by immediately popped to Mars, our superhuman Martian could save the baby. We cannot. That fact about us seems to support the idea that we have no obligation to save the baby. Ruling out the Martian case is something that a credible OIC principle should do, but OIC 2 fails to do it. Of course, as written, OIC 2 states a necessary condition for a morally binding principle, not a sufficient condition, and so it does not declare that we humans have an obligation to jump to Mars. Still, as a necessary condition, it is not informative about us humans.
A principle of charity of interpretation requires that we attribute to Kant a version of OIC that shows greater deference to human limits. Humans are subject to the laws of nature, just as they are subject to the moral law. One cannot ask human beings to overturn laws of nature, particularly those that govern human action. We thus propose the following interpretation of Kant’s OIC:
OIC 3: A moral principle binds a human agent only if, were she a rational human agent acting out of respect for the moral law and acting consistently with the laws of nature, she could act on it.Footnote 6
As no rational human agent could save the Martian baby, OIC 3 correctly avoids saying that it would be wrong to not save the baby. OIC 3 is otherwise promising. It generally does not deny the wrongfulness of an action as easily as OIC 1 does. Suppose that an adult responsible for childcare intentionally becomes so drunk that he cannot care for his child, and that the child consequently suffers a terrible accident. Arguably the adult does nothing wrong on OIC 1, because he did not have the occurrent ability to act on a principle dictating that one takes good care of his children. But he does something wrong on OIC 3, because a rational agent could act on a principle requiring that one take reasonable care of one’s child. How, if at all, do these differences matter for case involving the influence of bounded ethicality?
4. Comparing Versions of OIC
In this section we will argue that OIC 3 better illuminates issues of bounded ethicality than does OIC 1. In applying any OIC principle, much turns on the interpretation of the word “can” and its cognates. Can auditors, for example, act without bias? The fact, as we have seen, is that most don’t, and from that fact one might infer that most can’t, that bounded ethicality limits what most auditors can do. But we think that inference would be rash. In the moral assessment of the actions and auditors and others affected by bounded ethicality, options may be expansive. For all relevant moral purposes, we should conclude that auditors can do objective audits.
In some of the most influential psychological studies of phenomena that might be identified as bounded ethicality, most but not all subjects act against their moral obligations. Thus in Asch’s (Reference Asch1956), Darley and Batson’s (Reference Darley and Batson1973), and Milgram’s (Reference Milgram1974) works, many subjects act wrongly, succumbing to influences representing bounded ethicality, but other subjects do not. This suggests that the impact of bounded ethicality is fixed not by the laws of human nature but by human choice. One of Asch’s (Reference Asch1956) experiments found that 25% of the subjects under group pressure to conform to wrongful expectations refused to conform. While this is not a majority, it demonstrates that 25% of subjects did not succumb to the psychological influence of conformity. Similarly, in Darley and Batson’s (Reference Darley and Batson1973) study, seminary students were observed walking past a victim who needed help while they were on their way to give a talk. One group was not prompted to hurry, the second was prompted to hurry only moderately, and the third was prompted to hurry quite a lot by being told they were running late. Of the subjects who were not in any hurry, 63% helped the victim. A lower proportion, 45%, of those subjects under moderate-hurry conditions helped the victim in need. Of those under the high-hurry conditions, 10% helped. Again, there are individuals who, even under stressful conditions, can stop and do what seems most commendable. Milgram’s (Reference Milgram1974) experiments also feature individuals who did not succumb to obedience to authority. If these cases confirm the phenomenon of bounded ethicality, demonstrating that psychological influences can cause people to act wrongly, they also demonstrate that people can find ways to avoid or limit the effects of these influences.
Even Bazerman and Tenbrunsel (2011, chap. 8) acknowledge it is possible to help individuals correct their behaviors in the face of bounded ethicality. One suggested technique for countering the effects of bounded ethicality is the engagement of “pre-commitment devices,” through which individuals are encouraged to publicly share their intended ethical choices with impartial third-parties, as this helps to escalate their commitment and increases the likelihood that they will behave as they hope (see Brockner and Rubin Reference Brockner and Rubin1985 for reviews). Another suggestion is to learn to “recognize the signs that you are in a cognitive minefield, slow down, and ask for reinforcement from System 2 thinking,” which involves reflection and careful analysis, as opposed to relying on one’s intuitive System 1 thought processes (Kahneman Reference Kahneman2011, 417). To be fair, Kahneman believes this is more difficult than it sounds. To help the process along, Bazerman and Tenbrunsel suggest “focusing on the abstract principles that guide decisions” and “thinking about the values and principles that you believe should guide the decision” to help engage one’s System 2 processes (Bazerman and Tenbrunsel Reference Bazerman and Tenbrunsel2011, 157). Tools to do so include imagining what you would like said in your eulogy or thinking about whether you would feel comfortable sharing your decision with your mother (Bazerman and Tenbrunsel Reference Bazerman and Tenbrunsel2011, 156–157). While we should not exaggerate the extent to which these tactics help decision-makers do the right thing, the fact that the tactics often help individuals avoid the impact of bounded ethicality is evidence that bounded ethicality has limited impact on auditor decision-making and is not, as Bazerman and Tenbrunsel (Reference Bazerman and Tenbrunsel2011, 137) say, “inevitable." Further, if bounded ethicality is not inevitable, then it is not a law of nature.
One might object that we rely on a simplistic notion of how laws of nature affect human action. These laws can be probabilistic and not deterministic. If the laws are merely probabilistic, then they do not dictate that each auditor acts in a biased way, but rather that it is probable that for any particular auditor, she acts in a biased way, or that a particular proportion of auditors do so. So, the objection concludes, we should not be surprised when some people show themselves not to be victims of bounded ethicality; some people are victims, others are not, and victims cannot do objective audits. If one engages in a biased action as an expression of either a probabilistic law or a deterministic law, the normative significance is the same, one might argue: the action occurs as the result of a law of nature outside one’s control, and one cannot sensibly be expected to act on a principle requiring one to do otherwise, that is, one cannot be required to break the laws of nature in order act in an unbiased manner. And so, one might think, to the extent that psychological findings of bounded ethicality express even probabilistic laws of nature, they show that most cannot adhere to principles proscribing bias. That fact, combined with OIC 1, one might conclude, shows that these auditors do no wrong by issuing biased reports. In response to this objection, we maintain that it makes no difference whether psychological laws undergirding bounded ethicality and bias are probabilistic or deterministic; the same choices are available no matter what the structure of the law. In saying this, we invoke Bazerman, Tenbrunsel, and others who do research on bounded ethicality while advising people how to avoid its pitfalls. Bounded ethicality may indeed be a pernicious influence, but there is no reason to think that it is an influence that competent people can’t handle. Hence, bounded ethicality presents no reason in support for RBC.
When Bazerman and Tenbrunsel talk about interacting with people who have problematic biases, the approach they advocate demonstrates the superiority of OIC 3 to OIC 1. Recall that OIC 3 instructs us to assess what a person can do in terms of what a rational agent can do, while OIC 1 instead has us look at a person’s contingent propensities. Bazerman and Tenbrunsel approach people with biases as agents with the rational resources to improve their behavior, not as a set of propensities constrained to repeat behavior. Bazerman and Tenbrunsel’s practice is hopeful about human rational possibilities, and thus makes best sense on OIC 3, not OIC 1.
A proponent of RBC might say that we do not take bounded ethicality seriously enough, that we do not appreciate how it can make some moral choices in fact impossible; she might thus claim that OIC 3 is unrealistic. To answer this concern, we introduce a case in which bounded ethicality may seem more powerful than in the auditor case, a case in which we stipulate that the decision-maker simply cannot, in one important sense of that term, avoid biased judgment. Reflection on this case shows, we believe, that OIC 1, not OIC 3, is unrealistic:
Case 2: The Loan Officer
You head an Asian-American household living in a rental. To achieve your dream to be a homeowner, you apply for a home mortgage. Your loan officer was raised to dislike persons of East Asian origin and has long cultivated her racism by reading racist tracts and immersing herself in a racist subculture. Through her upbringing and actions, she has been rendered incapable of making decisions in racially impartial ways. There is no remedy available to the racist loan officer in the moment that she must decide on your application. She denies the loan, despite the fact that she has approved loans to similarly qualified Caucasian loan applicants. Therefore, you believe that she has unfairly discriminated against you, acting wrongly. The officer claims that she is incapable of applying the loan criteria in a race-neutral way; she says that it is untrue that she ought not to discriminate unfairly against Asians.
Let us suppose, for the sake of argument, that it makes sense to say that this loan officer was psychologically incapable of applying loan criteria in a race-neutral way. She disapproved the loan, thereby committing a paradigmatically wrongful, racist act. One might think that this particular act is not entirely the officer’s fault, because she was raised as a racist since childhood. Even so, we maintain, it makes sense to say to her, “You ought not to reject that loan application for racist reasons.”
Proponents of the OIC 1, such as Vranas (Reference Vranas2007), might deny that it makes sense to tell the loan officer that she should not reject the loan application for racist reasons, arguing that because the loan officer was incapable of acting on non-racist reasons, it makes no sense to complain when she acts on racist reasons. Still, Vranas might say that there exists a way to account for our judgment about what is appropriate to say to the racist. For instance, proponents might argue that we should not infer from the statements, “You ought not to be racist,” or “You ought to be race-neutral in applying the loan criteria,” an actual ascription of an obligation or duty to the racist. Instead, we ought to interpret these statements as merely pointing out what would be “fitting or good” in these circumstances (Vranas Reference Vranas2007, 179). In other words, this kind of “ought” statement is not really evaluative or prescriptive, but merely idealized or aspirational, and, as such, does not imply “can” (Altham Reference Altham1988, 281–282; Haines Reference Haines1972, 263). We disagree.
The proponents’ interpretation is not plausible in addressing the appropriateness of the demand, “You ought to be race-neutral in applying the criteria.” Assuming that this demand is appropriate given the moral facts at hand, it is odd to interpret the “ought” statements as merely issuing an aspirational statement about what would have been “fitting or good.” It makes much more sense to interpret the statements at face value, i.e., as a moral demand that the loan officer be impartial now. Thus, given our practices, it makes sense to reject OIC 1, and thereby to reject RBC.
In contrast to the OIC 1, OIC 3 does not necessitate a scramble to find the hidden meanings in appropriate-sounding “ought” statements. For the OIC 3, the proper point of reference for “can” is duties or obligations ascribed to a rational agent capable of acting rightly and operating under the laws of nature. Because denying a loan based on race is inconsistent with the moral value or rules of fairness, rational agents capable of acting rightly are under the obligation not to be racist; loan officers are obligated to decide loans in a race-neutral manner. OIC 3 confirms this obligation.
An implication of embracing the OIC 1 is that we lower moral standards for those confronting bounded ethicality and other social or psychological influences that act as impediments to doing the right thing. While it is important for normative theorizing and pedagogy to recognize those impediments, these impediments should not function as a justification for not doing the right thing. Conceding RBC and allowing bounded ethicality to partially determine moral standards of right and wrong would lead to a corruption of those standards, rendering them false and overly lax. It would also open the door to a subjectivism of morals, which makes it plausible for the racist loan officer and the auditors to reply, “Well, impartiality may be right for you, but it is not right for me.” These complaints about OIC 1, Vranas’ interpretation of OIC, show that it is normatively undesirable, resulting in an unattractive lowering of moral standards. He might respond that even if his position points to undesirable facts about humans, it does not follow that his position is false. To show that the position is false, we must show some flaw in OIC 1. The Loan Officer Case simplifies the task of finding a flaw, because it stipulates that the loan officer occurrently cannot act in an unbiased way.
It makes sense to criticize the loan officer’s action as wrong, we think, because a person’s failures of rationality do not limit his obligations. We will show this for the case of culpable drunkenness, and suggest a generalization to bounded ethicality. Bounded ethicality affects decision-making in morally different ways, and it is worth sorting through some of the varieties.
If, in fact, the loan officer in our scenario cannot make an unbiased decision, it is her fault that she cannot, as she has chosen to cultivate her biases. OIC 1 measures a person’s obligation purely in terms of her occurrent abilities, but that is implausible when it is her fault that her present abilities are defective. When it is one’s fault that one cannot avoid doing a wrong, one’s action may nonetheless be wrong and even blameworthy. Consider wrong acts committed while drunk. When a drunk driver recklessly injures a pedestrian, his action is nonetheless wrong; he may even be regarded as blameworthy for that action, as long as he voluntarily became drunk. His drunkenness has no bearing on what ultimately he should have done: drive cautiously, not recklessly, even though he could not have driven cautiously in the occurrent sense of “can” from OIC 1. It follows from reflection on the culpable drunkenness case, we think, that the sense of “can” in OIC 3 is what matters for morally right action, not the sense of “can” in OIC 1. In Case 2, the loan officer voluntarily put herself into a position in which her bias would lead her to do wrong. She is responsible for her bias and the wrong acts that issue from it, just as the culpable drunk is responsible for what he does while drunk. Our loan officer scenario is not unrealistic. A person who rises to the position of a loan officer has had many years to try to neutralize her bias. If she does not, she is complicit in that bias. Both the culpable drunkenness case and the self-cultivated bias case are interesting because they show that sometimes the fact that a person cannot occurrently engage in an action does not exclude her from an obligation to engage in that action. In both these cases, the fact that a person cannot occurrently fulfill her obligation is in part her fault, but we do not think that this fact is an essential feature of cases in which occurrent disabilities do not limit moral obligation.
Suppose, for the sake of argument, that the loan officer is somehow not complicit in her bias, and that she simply cannot, as a matter of her occurrent abilities, bring herself to make an unbiased decision on the East Asian’s application. She can still act rightly even if she cannot reach an unbiased verdict on the loan applications. For example, she can recuse herself, refusing to make a decision on the application, turning the application over to a better-equipped colleague. If the loan officer nonetheless makes a biased decision on the application, we can then reasonably complain that she made the morally wrong choice on the application, because she had another option available to her, turning the choice over to someone else. The highest moral standards properly apply to the loan officer. The best explanation why she should recuse herself is that it would be wrong for her to reach a biased verdict on the loan application.Footnote 7 A proponent of OIC 1 will, of course, respond that it makes no sense to say that the loan officer ought to act fairly because she (occurrently) can’t. But, as we have already seen in the drunk driver case, sometimes it makes sense to say that a person ought to do something even if he occurrently can’t. Culpable drunkenness constitutes one context in which an occurrent ability to do something (drive carefully) turns out to be no prerequisite for an obligation to do that thing (drive carefully), but it is only one example of a morally flawed exercise of rationality not limiting obligations. There are others. Indeed, even if one is nonculpably intoxicated because, for example, someone secretly put drugs in one’s soft drink, one must nevertheless drive carefully, not recklessly. A compromised state of mind, if relevant, may give one an excuse for doing something wrong, but does not make one’s action any less wrong. Just as we are all morally obligated not to drive recklessly in ways affected by intoxication, no matter what our occurrent abilities, we are also required not to act in ways affected by personal prejudices and other moral failures of rationality. The right action for the loan officer is to issue a fair verdict, no matter what her occurrent abilities. Doing so is what a rational person concerned about conforming to the moral law would do. Thus, OIC 3 captures the correct sense of “ought implies can,” OIC 1 does not, and RBC is defeated.
In this section, we respond to objections that might be raised against our arguments.
5.1. Objection 1: Case 2 and the Role of “Can”
One might doubt that Case 2 functions as a strong argument against RBC. Case 2 recommends that the loan officer take action against her bias. But if she can do this, then it would seem that she can be impartial after all. A case in which a person can simply avoid the biases of bounded ethicality is irrelevant to RBC, the objection might conclude.
In response, we would point out we stipulate that in Case 2 it is psychologically impossible for the auditor to avoid bias; her occurrent psychological state stands in the way of an impartial judgment. But while we hold that she cannot avoid bias, we maintain that as a practical matter she can avoid acting on her bias. It is important to see that her action is wrong so that it is clear that she should not do it. OIC 1 does not allow one to declare her action to be wrong. And that is sufficient reason to reject OIC 1 in favor of OIC 3.
5.2. Objection 2: Blame, Liability, and Punishment
One might invoke the following reductio as a way in order to reject OIC 3:
Premise 1) OIC 3.
Premise 2) OIC 3 justifies blaming, holding liable, or punishing someone for something that she cannot occurently do.
Premise 3) Blaming, holding liable, or punishing someone for something she cannot occurently do is unfair.
Conclusion) Therefore, OIC 3 should be rejected.
This argument is mistaken because Premise (2) is false. Our argument concerns act evaluation (e.g., Tony has an ex ante duty not to perform a particular act; or a particular act is impermissible). Act evaluation is conceptually distinct from agency evaluation (e.g., Tony is not culpable, blameworthy, or liable for performing a particular act) (Arpaly and Schroeder Reference Arpaly and Schroeder2014; Bennett Reference Bennett1995).Footnote 8 Consider, for instance, cases of criminal law where the punishment, which can be understood as a public expression of blameworthiness (Feinberg Reference Feinberg1965), is partly determined by mens rea of criminals and relevant situational factors. It is noteworthy, however, that the processes for attributing blame, liability, or punishment to agency does not affect the wrongness or the impermissibility of the criminal act itself. Whether someone intentionally hurts a child or does so through negligence or even unintentionally, the act of child abuse per se is wrong and impermissible, regardless of the mental states or any capabilities of the agent. The agent in both cases has an equally stringent duty not to abuse children.
Likewise, regardless of whether, for example, the biased auditors’ agencies are blameworthy/culpable/liable for the financial damage involved in Case 1, one can, as a separate matter, argue that they have an independent duty to be impartial. Even if they did something they should not have done, it is a different matter to define how blameworthy they are. Hence, the fact that OIC 3 implies that the auditors have a duty of fairness does not necessarily entail anything about agency evaluation, blame, punishment, or liability. Thus, Premise 2) must be rejected, which responds to the objection at hand.Footnote 9
We have presented an argument to defend normative business ethics against the Radical Behavioral Challenge (RBC). If we are correct about the status of RBC, normative ethics can continue to ascribe obligations and to generate “ought” statements that bind even those moral agents susceptible to the influence of bounded ethicality. That is, the obligations of moral actors remain the same regardless of their knowledge of bounded ethicality or their ability to effectively combat it. Naturally, the knowledge that bounded ethicality exists may generate new obligations for moral actors—the obligation, for example, to undertake effective means to combat the relevant social and psychological influences of bounded ethicality. It is a significant contribution of behavioral business ethics that it helps us to think about the necessity of recognizing these new obligations.
Interestingly, RBC challenges even behavioral business ethicists’ criticism about the insufficiency of normative business ethics for ethics education. Behavioral business ethicists’ pedagogical challenge to normative ethics is that theories of normative business ethics are irrelevant for ethics education because they cannot help individuals to make the right decision or act rightly (Bazerman and Tenbrunsel Reference Bazerman and Tenbrunsel2011, chap. 2). To overcome psychological impediments, behavioral ethicists suggest that the auditors, for instance, look to the findings of behavioral ethics. Note that this remedy assumes that the right thing to do is not to be biased in favor of their clients. In other words, behavioral ethicists, just like their normative counterparts, would assert a normative claim that the auditors ought not to be biased toward the clients. This clearly normative claim, which many behavioral business ethicists believe to be true, enables behavioral business ethicists to criticize normative business ethics arguments and pedagogy. If RBC is true, however, then the normative claim that the auditors ought not to be biased is untenable, and, accordingly, behavioral ethicists’ pedagogical criticisms of normative ethics are likewise untenable. Thus, in defending the usefulness of normative business ethics against RBC, we have bolstered the practical and pedagogical relevance of behavioral ethics, too.
What contribution can behavioral ethics make to normative business ethics?Footnote 10 Normative business ethics’ primary role is to study what values organizations should espouse, assess what obligations or duties to ascribe to individuals within those organizations, and determine which policies and regulations are appropriate for governing businesses and markets in a just society. As we mentioned, this requires normative business ethics theory to develop frameworks to understand organizational and individual obligations for handling problems of bounded ethicality. In addition, because normative ethics also concerns itself with evaluating whether agents are blameworthy or deserving of punishment, it is possible that bounded ethicality may affect such agential evaluations. Whether bounded ethicality can affect blameworthiness and how it might mitigate blame deserve serious investigation and present an area for fruitful future research (Holroyd Reference Holroyd2015; N. Levy Reference Levy2014; Machery Reference Machery2010; Saul Reference Saul, Jenkins and Hutchison2013).
Finally, we ask, does behavioral business ethics have anything to learn from normative ethics? The normative findings that we have revealed in this article have significant implications about behavioral theorizing. If the philosophical argument that we have defended in this article is correct, then we have good reason to maintain higher moral standards, even though each and every businessperson may not always be able to perform these duties or obligations. If an important goal of behavioral business ethics is to help businesspersons to perform these duties, they have two primary options. One option is to lower standards with the assistance of the OIC 1, which is a possibility we have rejected. The second and more promising option is to emphasize the existence of those who have cultivated or possess the ability to live bold, unbounded ethicality, that is, those who do not succumb to the influence of pessimistic, bounded ethicality. This suggests that it is worth paying closer attention to these individuals capable of breaking through the influence of bounded ethicality. We want to recall that Asch’s (Reference Asch1956) memorable study was titled “Studies of Independence and Conformity: I. A Minority of One Against A Unanimous Majority” (italics ours), and that 25% of the experiment participants were capable of independence in the face of pressures to conform. In this line of reasoning, we suggest that behavioral ethics scholars have good reason to study not only the bounded ethicality of the 75%, but to expend equal time and resources studying the unbounded ethicality of the 25%.
An earlier version of this article was presented at the Society for Business Ethics Annual Meeting in Philadelphia. We are indebted to Ken Goodpaster, Nien-hê Hsieh, Jessica Kennedy, Daryl Koehn, and Diana Robertson for their feedback. Several anonymous reviewers, along with BEQ editor Denis Arnold, provided excellent guidance. Thanks to all for comments and advice, including those who think that we are absolutely wrong.