Moral and Amoral Conceptions of Trust, with an Application in Organizational Ethics

Rousseau et al. (1998) found broad consensus on Mayer et al.’s definition in the management literature. And, moreover, Mayer et al. won an award for the best paper published in Academy of Management Review in the 1990s, which serves as another measure of the consensus around their definition.

Brenkert distinguishes between what he calls basic trust, guarded trust, and extended trust. As Brenkert uses those terms, basic trust applies in “impersonal, systematic relations” (1998, p. 303); extended trust can occur between individuals in close relationships, and it involves the exposure of significant vulnerabilities; guarded trust is a more limited exposure of vulnerability in the context of business contract.

The argument in the main text focuses on two versions of the epistemological conception of trust, Mayer et al. (1995) and Brenkert (1998). For both, trust is a disposition to act held on the basis of cognitive processes (beliefs and expectations). Jones offers an alternative version of the epistemological account that incorporates emotion. For Jones, trust is an affective attitude: trust “is an attitude of optimism that the goodwill and competence of another will extend to cover the domain of our interaction with her, together with the expectation that the one trusted will be directly and favorably moved by the thought that we are counting on her” (1996, p. 4). As an account of the psychological processes involved in trust, Jones’s account is plausible and consistent with recent work in management and empirical psychology (see Williams 2001; Dunn and Schweitzer 2005). But acknowledging the role of emotion in trust does not repair the epistemological account: according to Jones the affective component explains why the trusting party feels betrayed or wronged, but it cannot account for why the person is wronged. So the epistemological account—even admitting a role for emotion—is still inadequate. More generally, our central claim is that the trust relationship is a moral one, and this conceptual claim is independent of questions about the psychological processes involved. That said, the psychological processes involved in coming to trust on our moral conception will likely be complex and involve both cognitive and emotional components, some mixture that will vary according to the context. And given recent work on the role of emotion in moral judgment and moral action (such as Vetlesen 1994; Haidt 2001; Nichols 2007; Prinz 2009), we should expect emotion to play a central role in the psychological processes surrounding trust relationships.

Williamson refines Coleman’s economic analysis of this example, arguing that the loan was made because the London bank “had the most knowledge of the shipowner and the best prospect for future business” (1993, p. 470). He argues, further, that instances of trusting behavior that that seem to be “noncalculative”—meaning non-economic—can almost always be better explained in calculative terms. And he describes his method as follows: “The relentless application of calculative economic reasoning is the principal device that I employ to define and delimit the elusive notion of trust” (1993, p. 453). Mayer et al.’s definition of trust is “calculative” in Williamson’s sense of that term, and we could have used the term “calculative” as a label for the characterization of trust in the social science, management, and business ethics literatures, in place of “epistemological.” But we chose the term “epistemological” because it emphasizes the way expectations, beliefs, etc., serve to justify—in a practical sense—risk-taking behavior.

There is also reason to think that epistemological states are not necessary for trust, though this is a further critical point outside the scope of the present argument. We can trust when we do not hold reliance-inducing beliefs or expectations, for example, when we do not have a choice, as in some emergency situation; or when we lack information to make a tradeoff between potential costs and benefit; or in contexts in which cost/benefit analyses are inappropriate. Holton (1994) offers a series of examples of the third kind, suggesting that one could trust because, “perhaps you simply think it is the way you ought to treat one of your employees” (p. 63). In cases of this sort, a person can trust without any expectation about the other person’s trustworthiness, and without any belief about the other party, or even despite expectations and/or beliefs that the other party is untrustworthy. Flores and Solomon (1998) refer to examples of this sort as “blind trust.” Williamson (1993) also notes that cost/benefit or what he calls “calculative” considerations are inappropriate in close-personal relationships, because calculative considerations would “destroy the ‘atmosphere’” of those close, personal relationships [quote from Craswell’s 1993, p. 497, discussion of Williamson). As a result, for Williamson, trust in close, personal relationships requires that the usual epistemological states be left out of consideration. Craswell (1993) notices that we might extend Williamson’s point to commercial relationships. Williamson does not think that giving calculative accounts of trust behavior in close, personal relationships will be inaccurate. Instead, as Craswell puts it, “[Williamson’s] argument for excluding calculative theories from the personal sphere rests entirely on the undesirable effects that calculativeness would have if it were internalized by the participants in those relationships (Craswell 1993, p. 499). Craswell labels such effects “spillover,” because the conception of trust as a calculative phenomenon spills over and affects the relationship meant to be explained. Spillover in this sense is another motivation for the present project: by failing to acknowledge the moral dimension of trust, we could be encouraging persons to view their relationships with others in calculative rather than moral terms. And by extending Williamson’s point to the commercial domain, Craswell raises the question of whether we have this same risk of “spillover” in commercial relationships. If so, the epistemological conception of trust might actually be dangerous in some contexts.

The account of trust presented here is somewhat similar to the one developed in Nickel (2007). According to Nickel, “if one person [A] trusts another [B] to do something, then she [A] takes him [B] to be obligated to do that thing” (p. 310, our emphasis), and because A takes B to be obligated, A can hold B “blameworthy” if B fails to act as required. On this view, when A trusts B, A ascribes an obligation to B, meaning—for Nickel—that A acts as if B is so-obligated. But Nickel is careful to note that no actual obligation exists (see p. 312). His point is about the “conceptual commitments of the attitude of trust” (p. 312): person A must ascribe an obligation to B in order to make blame or punishment appropriate, but that does not require that there be an actual obligation (according to Nickel). Our view begins with the same intuition as Nickel, namely, that non-moral conceptions of trust are inadequate. But the problem with Nickel’s approach is that B is not blameworthy if no actual obligation exists, so B will not be blameworthy when A holds him or her to be so. Our view differs in that we take trust to put an actual obligation in place.

For a discussion of justification conditions in the philosophical literature, see Baier, who suggests that “reasonable trust will require good grounds for…confidence in another’s good will” (1986, p. 235). Also see Hardin (1993). For Hardin, “A full account of rational trust must be grounded in reasons for expecting another to fulfill a trust and in reasons for holding general beliefs about trustworthiness” (p. 526). We take Baier’s use of the term “reasonable” and Hardin’s reference to “rational” trust to amount to the same requirement, namely that trust be justified. For Hardin, the reasons we choose to trust will, to some degree, depend on our own emotional history, and these reasons are also based on experience with others or “vaguely on Bayesian generalizations” (p. 516). As mentioned (in footnote three), Jones (1996) explains our trusting behavior in terms of an affective attitude, and the affective component explains why trust is often insensitive to evidence (we often trust those likely to betray). In the sociology literature, Gambetta and Hamill’s (2005) book is an excellent empirical study of justification conditions.

Brenkert (1998) makes this same argument against Hosmer’s (1995) account of trust. According to Hosmer, trust is “the expectation by one person, group, or firm of ethically justifiable behavior…on the part of the other person, group or firm in a joint endeavor or economic exchange” (p. 399). But Brenkert notes that there can be trust among firms engaged in immoral practices (see pp. 301–302). Brenkert takes this to be a general argument against “moralized accounts of trust,” but note that the argument will not apply to our moral account of trust because we do not equate trust with general obligations to do what is right. Hosmer’s view is an epistemological one because it defines trust in terms of an expectation, and though expectations can be disappointed, there is no basis for saying that the trusting party is betrayed.

Jones also argues—against Baier—that we can trust when we do not have any basis for making an inference about the other person’s goodwill (1996, pp. 18–19). This argument is a variant of the one suggested in footnote five applied here: the epistemological states Baier cites as characterizing trust are not necessary. Separate from the argument in our main text, Holton also offers a series of examples that show why goodwill cannot do the work Baier requires, cases in which goodwill is not necessary or sufficient for trust (see Holton 1994 pp. 64–5). For example, “a trickster might rely on your goodwill without trusting you” (p. 64); this shows that relying on another’s goodwill is not sufficient for defining trust.

Smith (2001) also distinguishes between confidence and trust (and offers a helpful survey of literature on this distinction, beginning with Gambetta 1988a, b and Luhmann 1988). According to Smith, we can have confidence, or we can act with confidence, in situations that do not involve risk and vulnerability. Trust, in contrast, “arises, and is indeed necessary, when the inputs and outputs of social interaction cannot be certain” (p. 292). Trusting action therefore requires “some personal calculation about those attributes of a situation or relevant other, and about the acceptability of potential disappointment relative to an advantageous outcome, which generate the belief that it is worth taking a risk” (p. 292). Taken as a description of the preconditions of trusting action, we readily agree. But as we use the term “confidence,” it includes expectations in situations involving risk and vulnerability, and we argue that trust is not confidence in this sense. So acting on the basis of a risk/reward calculation is not yet to trust. Put another way, we can distinguish between acting with confidence (when there is no risk or vulnerability) and acting prudently (with a responsible or reasonable assessment of risk and reward) on one hand, and trust on the other, where trust is not to be confused with confidence or prudence.

Uslaner’s more general point is that moralistic trust on his definition is a precondition of civic life, not a product of civic or social involvement, in opposition to Robert Putnam’s work on social capital. Jones and Bowie (1998) also appeal to a moral obligation to trust to explain trust in temporary groups; so their “ethical account of trust”—like Uslaner’s—is an account of the moral justification conditions for trust, not an account of trust as a moral relationship.

Ghoshal’s specific target is Williamson’s (1985) version of transaction cost economics, which develops Coase (1937); Ghoshal also cites Coase’s (1988) retrospective discussion.

These two approaches to organizations are run together in Solomon’s (2004) work on ethical leadership. According to Solomon, “[trust] establishes a framework of expectations and agreements (explicit or not) in which actions conform or fail to conform” (p. 96). This account of trust suggests a conception of organizations as coordinated, and that in turn suggests that he has in mind using incentives and monitoring systems to better coordinate behavior among members of some organization, with the terms of relationships made explicit in contracts. It is difficult to see how this can serve as the foundation for an account of ethical leadership, but the moral conception of trust suggests a richer account of ethical leadership, one that involves cultivating moral relationships and personal connections among persons in the service of some organizational goal. Those relationships could give rise to the kind of expectations Solomon mentions, and that network of relationships could also shape or change the organizational goal.